INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

UNDER MOST CONDITIONS, the heart's utilization of glucose is modest relative to that of nonglucose substrates. Glucose has theoretical advantages as a substrate for the ischemic or hypoxic heart, however, including the ability to generate ATP nonoxidatively via glycolysis, greater oxygen efficiency of oxidative ATP formation, and reduced production of toxic products of fatty acid metabolism (2, 5, 6, 11, 12, 18). This has fostered interest in the therapeutic potential of augmenting myocardial glucose utilization by systemic infusion of supplemental glucose and/or insulin to patients with ischemic heart disease. Although a large body of work addresses the regulation of glucose metabolism in hearts perfused ex vivo, the relative influence of circulating glucose and insulin concentrations on myocardial glycolytic and oxidative utilization of an administered glucose load are not as well defined for the case of intact organisms.

Glycolytic and oxidative metabolism of exogenous glucose can be examined coordinately by measuring the accumulation of ¹³C-labeled intermediary metabolites in hearts supplied with [¹³C]glucose. Theoretical assumptions, supported by studies of isolated rat and guinea pig hearts, dictate that during perfusion with [1-¹³C]glucose the heart accumulates [3-¹³C]pyruvate in proportion to the fraction of glycolytic substrate supplied by exogenous glucose relative to alternative unlabeled substrate sources (e.g., endogenous glycogen) and -[4-¹³C]ketoglutarate in proportion to the fraction of tricarboxylic acid (TCA) cycle carbon flux supported by flux through pyruvate dehydrogenase (PDH), relative to other acetyl-CoA sources [e.g., free fatty acids (FFA); see Refs. 13, 16, 20, 23]. Pyruvate and -ketoglutarate are present in small quantities in muscle but are in isotopic equilibrium with the much larger alanine and glutamate pools (7, 23); thus measurement of relative steady-state ¹³C enrichments in plasma [1-¹³C]glucose and myocardial [3-¹³C]alanine and [4-¹³C]glutamate allows estimation of the contribution of circulating glucose to myocardial glycolytic and oxidative flux, relative to competing substrates (9, 10, 15, 21).

In the present study, we used this method to examine the influence of circulating glucose and insulin level on myocardial glucose metabolism in the intact rat. Rats were chosen because their small size allows labeling of intermediary metabolite pools to significant ¹³C enrichments in vivo, because they can be conveniently studied in the conscious, chronically catheterized state, avoiding potential artifacts of immobilization, anesthesia, or in vitro perfusion, and because they permit economical study of a number of different experimental conditions. We combined steady-state [1-¹³C]glucose infusion with standard euglycemic and hyperglycemic insulin clamp techniques to test 1) the degree to which raising circulating glucose and/or insulin level within their physiological ranges influences glycolytic and oxidative utilization of exogenous glucose by the heart, relative to other substrates; 2) whether the effects of the two are additive; 3) whether proportional utilization of circulating glucose increases further at supraphysiological levels; and 4) whether glycolysis, or alternatively PDH flux, sets the upper limit on the proportional contribution of glucose to oxidative flux.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Experimental Animals

Experimental Protocols

1

1

1

1

1

1

1

1

1

1

The last two groups examined whether raising glucose and insulin concentration to supraphysiological levels still increased relative glucose utilization further. A maximal insulin group (n = 8) received a 3-h infusion of regular insulin at 10 mU/min while D-[1-¹³C]glucose was infused at 25 mg · kg¹ · min¹ to maintain basal plasma glucose. A maximal insulin plus glucose group (n = 16) received the same 10 mU/min insulin infusion along with enough glucose (50 mg · kg¹ · min¹) to raise plasma glucose to approximately four times the fasting level. In this group, to also compare the reliability of our analytical methods at low vs. higher tissue metabolite ¹³C enrichments, seven rats received 2 mg · kg¹ · min¹ D-[1-¹³C]glucose along with 48 mg · kg¹ · min¹ unlabeled glucose, whereas the remaining eight received 8 mg · kg¹ · min¹ D-[1-¹³C]glucose and 42 mg · kg¹ · min¹ unlabeled glucose.

At intervals during each study, arterial plasma was sampled to define plasma glucose concentration and ¹³C enrichment in plasma glucose carbon-1. Five minutes before the end of each study, rats were anesthetized with pentobarbital sodium (50 mg/kg ip) and were placed on positive-pressure mechanical ventilation through a tracheostomy, taking care to continue all experimental infusions without interruption. A final arterial blood sample was taken, and the heart was then removed, blotted dry, and frozen by clamping between aluminum plates chilled in liquid nitrogen. Plasma and tissue were stored at 80°C.

The experimental protocol assumes achievement of isotopic steady state between plasma (glucose) and myocardial (alanine and glutamate) carbon pools during the 3-h [1-¹³C]glucose infusion. This assumption was based on the observation that rat hearts perfused in vitro with [¹³C]glucose reach isotopic steady state in the glutamate carbon-4 position within 30 min (23) and on our previous observation that isotopic steady state is achieved within 2 h during [1-¹³C]glucose intravenous infusion in intact canines (15). Nevertheless, to confirm that [3-¹³C]alanine and [4-¹³C]glutamate were at steady state relative to plasma glucose by 3 h, in an additional four rats, we extended the 2 mg · kg¹ · min¹ [1-¹³C]glucose infusion to 4 h. This comparison was made in fasting rats, since their rate of ¹³C equilibration between plasma in the myocardial glutamate pool should be the slowest among the groups studied.

Analytical Methods

Plasma substrates and insulin. Plasma glucose was measured with an automatic glucose oxidase analyzer (Statplus 2000; YSI). Plasma insulin was measured using a double-antibody RIA kit (New England Nuclear). Insulin administration affects the circulating level of the major nonglucose oxidative substrates (lactate and FFA). Accordingly, plasma lactate and FFA concentrations were measured in the basal state and during the final 15 min of insulin infusion at each of the three insulin levels studied (i.e., in fasting, hyperinsulinemia, and maximal insulin groups). Lactate was measured using an automated lactate oxidase analyzer (Statplus 2000; YSI). FFA were measured using a commercial colorimetric assay kit (Wako NEFA C test kit; Wako Chemicals, Neuss, Germany).

Myocardial glycogen concentration. Weighed portions (~30 mg) of frozen myocardium were dissolved in 30% KOH, glycogen precipitated with ethanol, and digested with amyloglucosidase (15). Glycogen concentration was calculated as micromoles glucose per gram wet weight of the myocardium.

Plasma and tissue ¹³C enrichments. Analyses were performed as reported previously (15). Frozen hearts were powdered under liquid nitrogen, and amino acids were extracted by homogenization in cold 6% perchloric acid. Alanine and glutamate in each homogenate were purified on a Dowex-50W (200-400 mesh) minicolumn equilibrated with ammonium formate (0.1 M, pH 3.0). The absolute enrichments of each alanine and glutamate carbon were determined algebraically from the absolute total enrichment [determined from gas chromatography-mass spectrometry (GC-MS)] and the relative enrichment at each carbon [determined from the ¹³C nuclear magnetic resonance (NMR) spectra]. All enrichments are reported as atoms percent excess (APE) above natural ¹³C abundance (assumed to be 1.1%).

NMR methodology. ¹³C NMR spectra were acquired at 125.76 kHz (AM 500; Bruker Instruments, Billerica, MA) using a standard ¹³C/¹H probe (21). Spectra were acquired using a 30° pulse, quadrature detection, digital resolution of 2.7 Hz/point and with a pulse program for inverse-gated heteronuclear WALTZ decoupling with a delay of 1 s between pulses.

GC-MS methodology. GC-MS analysis was performed with a gas chromatograph (HP-1 capillary column, 12 mm × 0.2 mm × 0.33 mm film thickness model 5890; Hewlett-Packard) interfaced to a HP 5971A mass detector operating in the positive chemical ionization mode with methane as reagent gas. Glucose was derivatized as the penta-acetate, and isotopic enrichment was determined from the ion intensities of mass-to-charge ratio (m/z) 331-334. Amino acids were derivatized and analyzed as the trifluoracetyl n-butyl ester (8). Isotopic enrichment of alanine was determined from the ion intensities of m/z 342-348, and glutamate m + 1 to m + 5 was determined from m/z 356-363.

Data Analysis

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Plasma Glucose and Insulin Concentration

1

1

1

1

1

1

1

1

View larger version (22K): [in this window] [in a new window]   Fig. 1.   Plasma glucose concentration (A; mM) and plasma [1-13C]glucose specific enrichment [B; atoms percent excess (APE)] during a 180-min intravenous infusion of [1-13C]glucose in experimental groups studied at 3 different levels of glycemia. Points represent means ± SD for 8 rats.

Plasma Lactate and FFA and Myocardial Glycogen Concentration

1

1

1

1

1

1

1

1

1

1

Plasma and Myocardial ¹³C Enrichments

View larger version (41K): [in this window] [in a new window]   Fig. 2.   Fraction of total glycolytic flux supplied by plasma glucose (A; filled bars) and fraction of total tricarboxylic acid (TCA) cycle flux supplied by pyruvate flux through pyruvate dehydrogenase (PDH; B; filled bars) for each of the 6 experimental conditions. Data were derived from the calculated steady-state 13C enrichment ratios [3-13C]alanine/[1-13C]glucose and [4-13C]glutamate/[3-13C]alanine, as discussed in text.

Verification of Isotopic Steady State

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The results of the study demonstrate that glucose imported from the circulation contributes only a minor portion of fasting glycolytic and oxidative flux in the heart of the conscious, intact rat. Its proportional contribution to these processes is increased equivalently when either its circulating concentration or that of insulin is raised within their physiological ranges, and increasing their concentrations to supraphysiological levels increases the plasma glucose contribution still further. However, although circulating glucose becomes essentially the sole glycolytic substrate at supraphysiological glucose and insulin levels, glycolytic pyruvate continues to supply only ~60% of myocardial TCA cycle flux. This suggests that the magnitude of the PDH response may impose an upper limit on the degree to which glucose can be made to support oxidative energy formation by the heart during glucose and insulin infusion.

In fasting animals in vivo, the myocardial glycolytic rate is so low that conventional techniques have not permitted determination of whether plasma glucose is the sole or even the most important glycolytic substrate. In the present study, the ratios in Table 2 indicate that during fasting the portion of myocardial glycolytic flux supported by glucose imported from the circulation averaged only 32% (i.e., [3-¹³C]alanine-to-[1-¹³C]glucose ratio = 0.16). This 32% estimate for the exogenous glucose contribution implies either a rather large contribution to fasting glycolytic flux from glycogen, isotopic dilution of the myocardial alanine pool by low-enrichment alanine and lactate imported from the circulation, or both. Although direct evidence that glycogen contributes significantly to myocardial glycolytic flux under aerobic conditions in intact animals is lacking, this interpretation generally agrees with recent evidence from pulse-chase studies that glycogen turnover may support as much as 40% (3, 4) to 60% (19) of energetic glucose utilization in the isolated working rat heart. Isolated heart preparations are inherently glycogenolytic however, a fact that has clouded the relevance of previous observations (3, 4, 19). The current data are among the first to provide evidence, albeit indirect, that glycogen turnover may contribute significantly to energetic substrate flux in the heart under conditions where its myocardial concentration is static. As such, they complement the earlier observation of Wisneski et al. (24) that the majority of the glucose taken up by the human heart in vivo enters a storage pool before undergoing glycolysis or oxidation. The implication that continuous glycogen turnover may represent a mechanism for supporting the low-level "homeostatic" glycolytic activity maintained in the heart during fasting is intuitively appealing, since in the fasting state plasma insulin levels are low and transmembrane transport of exogenous glucose is slow, but glycogen phosphorylase is maximally active and glycogen synthase largely inactive. We further note that it has long been a point of contention whether addition and removal of glucose residues from glycogen proceeds in an ordered "last-on, first-off" fashion (1) or is instead substantially random (4, 19). To the extent that glycogen-derived glucose was responsible for diluting the myocardial [3-¹³C]alanine pool in our experiments, our data would be more consistent with removal of glucose residues at random from the entire glycogen particle, whose average ¹³C enrichment should be lower than that of plasma glucose.

The observed [4-¹³C]glutamate-to-[3-¹³C]alanine ratio of 0.14 agrees with many previous demonstrations that the heart oxidizes little circulating glucose during fasting, when plasma levels of alternative oxidative substrates are high and glucose transport slow. Indeed, the absolute rate of myocardial glucose oxidation during fasting is so low that its contribution to oxidative flux has been difficult to accurately assess in intact animals, or even in isolated-perfused hearts, by the traditional technique of measuring ¹⁴CO₂ release in coronary effluent in hearts supplied with [¹⁴C]glucose. The observation in this study that pyruvate supports 14% of myocardial TCA cycle carbon flux during fasting, made at steady state with regard to circulating levels of glucose and competing substrates and in the absence of potential artifacts introduced by immobilization, anesthesia, or surgical trauma, should be a uniquely reliable estimate.

Energetic metabolism of circulating glucose by the heart requires the coordinated action of transmembrane transport, phosphorylation, glycolysis, and subsequent TCA cycle oxidation of glycolytic pyruvate. Raising the circulating glucose level would be predicted to increase glucose flux through this system by mass action. Raising the plasma insulin level should increase the capacity of the system by increasing sarcolemma glucose transporter density and its velocity by covalent modification of regulatory enzymes and should also relieve FFA-mediated suppression of glycolysis and PDH flux by lowering circulating FFA levels. The relative physiological importance of these effects in vivo has not been clearly defined. In the present study, raising the plasma glucose level alone from 5.6 to 11.1 mmol/l increased the proportional contribution of plasma glucose to myocardial glycolytic flux from 32 to 44% and increased the proportion of TCA cycle flux contributed by pyruvate ~2.5-fold. Euglycemic elevation in the circulating insulin level similarly increased the proportional plasma contribution to myocardial glycolysis from 32% at 84 pmol/l to 48% at ~500 pmol/l and the proportional pyruvate contribution to TCA cycle flux ~2.5-fold. Applying these conditions simultaneously yielded a small incremental increase. Thus the effects of modest, physiological elevation in circulating glucose and insulin levels on relative myocardial glycolytic and oxidative dependence on plasma glucose appear to be independent, of equal magnitude, and probably additive.

Increasing the plasma insulin concentration to well above its physiological range further increased the proportional contribution of circulating glucose to both the myocardial alanine and glutamate pool. This incremental effect of supraphysiological vs. physiological hyperinsulinemia may reflect the greater reduction in plasma FFA level observed during the 10 vs. the 3 mU · kg¹ · min¹ insulin infusion (17, 22). Adding supraphysiological hyperglycemia to this extreme level of hyperinsulinemia further increased the exogenous glucose contribution. The finding that exogenous glucose becomes the sole contributor to myocardial glycolytic flux under maximal insulin plus glucose conditions is not unexpected, since glucose transport and phosphorylation should be fully saturated and glycogenolysis fully suppressed. The failure of the [4-¹³C]glutamate-to-[3-¹³C]alanine ratio to simultaneously approach 1.0 has several potential explanations. The glycolytic stimulation imposed by extreme hyperglycemic hyperinsulinemia may simply have caused glycolysis and pyruvate oxidation to become uncoupled, due to a primary limitation in the capacity of PDH to metabolize the increased quantities of pyruvate formed. Furthermore, although maximal hyperinsulinemia lowered the circulating FFA level by 65%, this probably did not completely suppress all myocardial uptake and oxidation of FFA from plasma and resultant inhibition of PDH flux (17). Alternatively, the 60% upper limit on the [4-¹³C]glutamate-to-[3-¹³C]alanine ratio may reflect an obligate contribution to TCA cycle flux from turnover of endogenous nonglucose substrate (e.g., triglycerides) within the myocardium. This last possibility is supported by studies in the isolated rat heart demonstrating that the myocardial glutamate pool reaches only ~70% ¹³C enrichment, even during steady-state perfusion with 99% enriched [¹³C]lactate or [¹³C]pyruvate as sole substrates (20), consistent with Randle's earlier observation that oxidation of endogenous substrates continues to account for ~30% of oxygen consumption even when rat hearts are perfused with glucose as their sole substrate (17). Taken together, the current observations would be consistent with the general concept that turnover of endogenous substrate pools (glycogen, triglycerides) may support a significant fraction of myocardial glycolytic and oxidative substrate flux in the conscious rat under both fasted and fed conditions.

Methodological Considerations

1

1

1

1

Clinical Implications