INTRODUCTION

Top Abstract Introduction Materials Results Discussion References

IT IS WELL ESTABLISHED that chronic exposure to hypoxia results in pulmonary hypertension. To provide adequate perfusion of the lung, the right ventricle develops much higher pressures under hypoxic conditions than under normoxia. A response to this chronic functional overload is a compensatory increase in right ventricular mass. The increased vascular resistance during hypoxia also imposes a greater burden on the energy-producing pathways in matching ATP synthesis to ATP demand. Furthermore, this increased requirement for energy occurs during conditions of low O₂ availability.

In the well-oxygenated heart, >95% of the ATP supply is generated by oxidative phosphorylation (for review see Ref. 1) to support cycling of the contractile proteins, maintain ion gradients, and fuel biosynthetic reactions along with other ATP-dependent processes. O₂ consumption by heart cells in vitro is limited by PO₂ within the physiological range, i.e., <15 Torr (51). Measurements of PO₂ in animals ventilated with room air have yielded values of ~20-30 Torr in the epicardial microvasculature (49, 50), whereas substantially lower values, down to 4-6 Torr, were found within the cardiac myocytes (for review see Ref. 64). It might be expected, therefore, that even a modest lowering of O₂ delivery to the contracting cells would compromise functional capacity of the heart.

Although previous studies have evaluated morphological and biochemical parameters in the chronically stressed heart, with particular emphasis on the left ventricle, they have largely focused on events that occur in the presence of normal O₂ delivery (3). Few investigations (2) have addressed what, if any, alterations in energy metabolism result from a prolonged, abnormal elevation of cardiac work in the absence of "normal" availability of O₂. The present work was undertaken to characterize more comprehensively the progressive changes in energy metabolism in right and left ventricles of rats living in a normobaric atmosphere of 10% O₂. Our data indicate that adaptation to hypoxia results in significant alterations of substrate utilization. Oxidative capacity was adversely affected, for the most part, in the musculature of the left ventricle, although suppression of fatty acid oxidation, the preferred fuel for the heart, was common to both ventricles. Compensatory adjustments, i.e., enhanced capacity for glucose phosphorylation and changes in amino acid metabolism, were found to support cardiac work.

	METHODS AND MATERIALS

Top Abstract Introduction Materials Results Discussion References

Animals. Eighty 4-day-old, pathogen-free, Sprague-Dawley male rats were housed in standard cages in room air or in Plexiglas boxes in a 10% O₂ atmosphere. In the latter case, no effort was made to control the level of CO₂ (range 0.8-1.8%). The bedding was changed daily. Animals were permitted free access to water and standard rat chow.

Physiological measurements. After animals were anesthetized (urethan at 1.5 g/kg body wt im, followed 45 min later with 0.75 g/kg sc), the left common carotid and right femoral artery were catheterized for measurement of systemic pressure and for sampling of arterial blood gases, respectively. A pressure transducer (2-Fr, Millar, Houston, TX) was inserted into the right ventricle via the jugular vein for measurement of intraventricular pressure. The animals were tracheotomized but breathed room air spontaneously. At the end of each experiment, a bilateral thoracotomy was performed to remove the heart, which was rinsed in saline. The ventricles were separated, weighed, and dried overnight at 90°C for determination of dry weight.

Oxidative capacity. In a parallel study the heart was excised immediately after decapitation and placed in a vial containing ice-cold saline. The ventricles were separated on a plastic petri dish cooled in ice. The ventricular walls, excluding the septum, were minced and homogenized in ice-cold medium (1:10 wt/vol) containing 250 mM sucrose, 50 mM Tris, and 1 mM EDTA, pH 7.4. The latter process, which causes cellular disruption, involved two steps: an initial treatment in a Polytron (model PT 3000, Brinkman Instrument; 30-s pulse) followed by homogenization (6 passes) in a Teflon-glass Potter-Elvehjem homogenizer. To prevent proteolysis, the homogenates were kept on ice and used within 3-4 h of preparation.

Enzyme activities. Cytochrome-c oxidase activity was measured polarographically (17) as described above in a medium containing 50 mM KH₂PO₄ and 0.1 mM EDTA, pH 7.4. The substrate was 0.04 mM cytochrome c, 0.63 mM N,N,N',N'-tetramethyl-p-phenylenediamene, and 12.5 mM sodium ascorbate.

Amino acid analysis. Amino acids were determined in neutralized extracts of blood (plasma) and heart homogenates with HPLC (24). Blood was collected in heparinized tubes and centrifuged (Sorvall Instrument/DuPont, Newtown, CT) at low speed (1,500 g at 4°C) to separate serum. An aliquot of the latter was quenched with an appropriate volume of perchloric acid (final concentration 0.6 M) and, after removal of the precipitated protein, neutralized with 2.5 M KHCO₃. Samples were obtained from animals used for measurements of oxidative capacity. Protein was measured by the biuret reaction with BSA as a standard.

Measurements of mRNA abundance, RNA preparation, and analysis. Steady-state mRNA levels were measured in three independent determinations by Northern blot analysis. For each experiment, tissues were obtained from two to five rats that were adapted to hypoxia or had been kept in room air for equivalent periods of time (aged matched). Samples were stored at 80°C before use. Total RNA was extracted from the tissues by using Ultraspec Reagent (Biotecx, Houston, TX). The respective RNAs were then pooled, and their concentration was determined by absorbance at 260 nm. For Northern blotting, a 20-µg aliquot of total RNA was added to two volumes of medium containing 50% formamide, 2.2 M formaldehyde, 1× running buffer (40 mM MOPS, 10 mM sodium acetate, 1 mM EDTA, pH 7.0), 0.04% xylene cyanol, 0.04% bromphenol blue, 5% glycerol, and 10 mg/ml ethidium bromide, heated to 65°C, and electrophoresed through a 1.0% agarose gel. The latter contained 2.2 M formaldehyde and 1× running buffer. RNA quality and loading were checked by ultraviolet illumination. The gels were blotted and fixed to Hybond N⁺ positively charged nylon membranes (Amersham, South Clearbrook, IL). Northern hybridizations were performed as described previously (11, 18). Blots were air dried, exposed to autoradiography film (XAR, Kodak, Rochester, NY) at 80°C with intensifying screen for 1-7 days, and scanned (model SI densitometer, Molecular Dynamics, Sunnyvale, CA). Values (derived from densitometric pixel volume) were normalized to the signal generated from ribosomal protein L28. Data were expressed as relative mRNA levels calculated as percentages of the relevant control value (assigned a value of 100%).

Oligodeoxyribonucleotides. Five pairs of gene-specific oligonucleotides were synthesized using an ABI 392 Medium Throughput DNA/RNA Synthesizer (Perkin Elmer/Applied Biosystems, Foster City, CA; numbers correspond to description of probe preparation): 1) 5'-CAAAATGCCAAGGAAATCTTAACCC, 2) 5'-GACAGTAGCTTTGCTGTTGGTCT, 3) 5'-GAGAAGGCCTACCAAATCCTGATG, 4) 5'-AGGGGCGACCGCATGCGTCTC, 5) 5'-TCAGCTGATTTATAATCTTCTAAAGG, 6) 5'-AGAAGTCAGAGTCACCTTCACAA, 7) 5'-AAAACTCATTGCACCAGTTGCGG, 8) 5'-ATCATCCTTTAGCTTCTGGTTGATA, 9) 5'-ATGTCTGCGCATCTGCAATGGATG, and 10) 5'-TCAGGAGCTCTTGGTGGGGGAGG.

Probe generation. cDNA for ribosomal protein L28, rat hexokinase I, rat hexokinase II, rat lactate dehydrogenase A, and rat lactate dehydrogenase B were generated via RT-PCR using standard PCR conditions (39) and the following gene-specific primers: hexokinase I (485-bp PCR product, oligonucleotides 1 and 2), hexokinase II (477-bp PCR product, oligonucleotides 3 and 4), lactate dehydrogenase A (913-bp PCR product, oligonucleotides 5 and 6), lactate dehydrogenase B (919-bp PCR product, oligonucleotides 7 and 8), and ribosomal protein L28 (416-bp PCR product, oligonucleotides 9 and 10). All gene fragments were cloned into pT7Blue, and the identity of each insert was confirmed by sequence analysis. Probe DNA was prepared by PCR with use of purified plasmid containing the cloned cDNA fragment of interest as template and the appropriate gene-specific primer pair. PCR products corresponding to the expected sizes were purified from agarose gels and labeled to high specific activity with deoxy-[³²P]CTP (New England Nuclear, Boston, MA), as described previously (63).

Statistical analyses. Values are means ± SE unless stated otherwise. Statistical significance between groups was determined with Student's t-test or ANOVA followed by the Newman-Keuls test. Significant differences were established at the 0.05 level of confidence.

	RESULTS

Top Abstract Introduction Materials Results Discussion References

Physical characteristics. Table 1 compares some general characteristics and hemodynamic measurements of rats maintained in the hypoxic colony (experimental) with those of rats kept in room air (control). The animals were exposed to 10% O₂ for 7, 14, 21, and 28 days.

View larger version (19K): [in this window] [in a new window]   Fig. 1.   Effect of chronic hypoxia on right ventricular mass and its relation to right intraventricular systolic pressure. Measurements were obtained from urethan-anesthetized rats exposed to hypoxia for 7-36 days. Right ventricle was isolated, rinsed in saline, and dried overnight at 90°C. Each point represents an individual animal.

Oxidative capacity. Homogenates, and not isolated mitochondria, were used to determine rates of substrate utilization (per gram of tissue). This was done for the following reasons: 1) isolation of mitochondria yields a fraction of their total content, whereas the aim of the study was evaluation of the "total" oxidative capacity of the intact tissue in situ, 2) mitochondria isolated from "stressed" muscles may be more "fragile" and, consequently, might sustain more damage during isolation than the organelles from "healthy" tissues, and 3) the wet weight of the normoxic right ventricle, typically ~190 mg, precluded isolation of enough mitochondria to measure in duplicate oxidation of all substrates and activity of the cytochrome oxidase.

View larger version (24K): [in this window] [in a new window]   Fig. 2.   Effect of chronic hypoxia on substrate oxidation in rat heart. Values at time 0 represent combined measurements from all normoxic animals maintained in room air as controls for various periods of hypoxia. Malate (1 mM) was added in conjunction with each substrate to ensure that tricarboxylic acid cycle was not substrate limited. Substrate and ventricle orientation (i.e., right or left) are indicated. Data obtained from animals kept in hypoxia for 41-53 days were combined and designated 41 days. Values are means ± SE [n = 3-10 animals/group, except for palmitoyl-L-carnitine (Palm-L-carn) at 1 day of hypoxia, where n = 2].

1

1

1

1

1

1

1

1

Glycolytic activity. Because oxidative capacity reached an apparent nadir in the left ventricle after 14 days of hypoxia, markers of glycolytic capacity were evaluated beginning at this time point. Maximal velocity of hexokinase (one of the rate-limiting steps in the glycolytic pathway) was similar in homogenates of right and left ventricles (5.5 ± 0.6 and 5.6 ± 0.6 µmol/g wet wt, respectively) from normoxic rats. Animals adapted to hypoxia for 14 days showed a marked increase in myocardial enzyme activity (Fig. 3A); the rise was to 9.9 ± 0.9 µmol/g wet wt (P < 0.01), i.e., by 80%, in homogenates of the right ventricle and to 8.2 ± 0.7 µmol/g wet wt (P < 0.05), i.e., by 46%, in preparations from the left ventricle. Measurements in hearts of animals exposed to longer periods of hypoxia (28 and 40 days, each n = 1) were within the range of those shown in Fig. 3A.

View larger version (21K): [in this window] [in a new window]   Fig. 3.   Effect of chronic hypoxia on maximal activities of glycolytic enzymes in rat heart. Maximal activities of hexokinase (A) and lactate dehydrogenase (B) were measured spectrophotometrically at 21°C. Right ventricle was dissected free from left ventricle and septum (discarded) in ice-cold media. Isozyme profiles were not determined. Values are means ± SE (n = 3-5).

View larger version (36K): [in this window] [in a new window]   Fig. 4.   Abundance of mRNAs for hexokinase and lactate dehydrogenase in right and left ventricles of normoxic and hypoxia-adapted rats. Northern blots are representative of triplicate samples obtained from RNAs extracted from hearts of 2-5 animals. RV and LV, samples from right and left ventricle, respectively; HK and LDH, hexokinase (I or II) and lactate dehydrogenase (A or B). L28 served as a control.

Amino acid metabolism. A comparison of the total concentrations of the amino acids (Table 4) in plasma from normoxic and hypoxia-adapted rats showed no difference between the two groups. A small rise, 29%, in aspartate concentration and a slight decrease, 14%, in glutamine concentration in samples from hypoxic rats were statistically not significant. Nevertheless, these changes were consistent with those obtained from cardiac muscles. The sum of the collective amount of the amino acids in right and left ventricles was for the most part unchanged by long-term hypoxia (Table 4). For aspartate, however, levels in right and left ventricles were 57% (P < 0.01) and 91% (P < 0.01) greater than in the corresponding controls. Glutamate concentrations in neither plasma nor heart were significantly affected by hypoxic adaptation. The aspartate-to-glutamine ratio, however, increased in right and left ventricles by 70 and 94%, respectively, in response to long-term hypoxia. Right and left ventricular glutamine concentration was 36% (P < 0.01) and 10% lower than in the respective muscles of normoxic rats. Consequently, the glutamine-to-glutamate ratio declined from 1.6 to 1.1 in the right ventricle of the hypoxic rat, but essentially no change in this parameter occurred on the left side (1.3 to 1.2).

	DISCUSSION

Top Abstract Introduction Materials Results Discussion References

The heart, like other metabolically highly active tissues, is considered to be very sensitive to acute episodes of hypoxia. Nonetheless, mammals including humans can endure prolonged periods of O₂ deprivation, for example, during ascent to high altitude. Many species have adapted quite well during evolution to these seemingly harsh conditions. The purpose of the present investigation was to characterize alterations in fuel metabolism in the right and left ventricle caused by relatively long-term (days) exposure of animals to a low-O₂ environment. Our major findings were threefold: 1) The normally enhanced oxidative capacity of the left ventricle, relative to its right counterpart, was diminished by hypoxic adaptation. Oxidation of different carbon substrates, including pyruvate, palmitoyl-L-carnitine, and glutamate, was compromised in the left ventricle within 24 h of hypoxic exposure and remained reduced when animals were maintained in a low-O₂ atmosphere. By contrast, substrate oxidation in the right ventricle was, on average, unaffected by chronic hypoxia. The sole exception was catabolism of long-chain fatty acid, normally the preferred substrate in the heart, which was decreased. This change, unlike that found in the left ventricle, did not become apparent until 14 days of hypoxia, even though a marked elevation of right ventricular work expressed as a rise in intraventricular pressure and an associated increase in right ventricular mass was noted by 7 days. 2) The maximal velocity of one of the rate-limiting enzymes of glycolysis, hexokinase, was markedly enhanced in both ventricles, although the increase was modestly greater on the right side. A substantial rise in the relative abundance of mRNA for hexokinase I and II was also found, but the changes in enzyme activity and its transcripts were not temporally aligned. 3) The intracellular levels of aspartate and the aspartate-to-glutamate concentration ratio were increased in both hypoxic ventricles. Although the levels of glutamine decreased in these tissues, the glutamine-to-glutamate concentration ratio was changed only in the right ventricle, thereby distinguishing the effects of hypoxia from those of chronic pressure overload. Taken together, the alterations described above suggest that limitation of O₂ availability results in a reduced capacity to synthesize ATP via oxidative phosphorylation in the left, but not the right, ventricle. Perhaps more importantly, the changes in amino acid metabolism are indicative of significant disturbances within the metabolic pathways for energy production, which, to our knowledge, have not been previously described for these conditions. Hence, a simple enhancement of glycolysis is not the only change in substrate utilization in the heart resulting from long-term hypoxia.

Chronic hypoxic exposure leads to pulmonary hypertension, which, in turn, results in marked hypertrophy (for recent clinical examples see Refs. 35 and 38) and, as is often the case, eventual failure of the right ventricle. In the present study a linear correlation was obtained between RVSP and right ventricular dry weight. This compensatory increase in muscle mass is a typical response to a sustained elevation of pulmonary pressure that imposes a chronic physical and metabolic overload on the ventricle.

A first step toward understanding the potentially deleterious effects of chronic overload is an examination of mitochondrial morphology and/or function. Early investigations of the overloaded ventricle indicated a loss of mitochondrial volume per cell (4, 71) and disappearance of mitochondrial cristae (72). Mitochondrial oxidative capacity was decreased (34, 57, 70, 71) or unaffected in some animal models (46, 59, 61) and in biopsies of the human heart (6). The present study, unlike many earlier reports, evaluated the progression of metabolic changes in the stressed, overloaded right ventricle and the "nonoverloaded" left ventricle. Our findings show that the oxidative capacity of the former, measured with a range of substrates, is not altered by chronic pressure overload. These data are consistent with the lack of change in the maximal velocity of cytochrome-c oxidase, the terminal, rate-limiting reaction of the electron transport chain. Oxidation of palmitoyl-L-carnitine, however, was decreased in the right ventricle from 14 days of hypoxia onward, declining 42% compared with control after 41 days. This decline was also found in the hypoxic left ventricle, but a statistically significant change was noted as early as 7 days of hypoxia. Although the direction of these changes is in agreement with the results of Kinnula and Hassinen (29) on mitochondria from 7-day hypoxic rats, the latter authors had to use more severe limitation in O₂ delivery (40.8 kPa) to detect a significant fall in fatty acid oxidation.

The mechanism of the reduction of fatty acid catabolism is not known. Previous studies in models of pressure (47) and volume (8) overload showed a fall in capacity to oxidize long-chain fatty acids that was associated with reduced myocardial levels of carnitine. Oxidation of medium-chain fatty acids, which diffuse freely into the inner mitochondrial space, however, was unaffected by chronic overload (8). The latter finding suggested that the decrease in utilization of long-chain fatty acids was caused by the inability of these molecules to cross the inner mitochondrial membrane for lack within the cardiac myocyte of adequate carrier, carnitine (47, 68). Our data obtained in the presence of an adequate (5 mM) level of carnitine indicate that a component, or components, of fatty acid catabolism must also be adversely affected by long-term hypoxia. Mitochondrial metabolism of fatty acids is a complex process. It utilizes several proteins before the electrons enter the respiratory chain; they include two transporters, carnitine acyltransferases I and II, and four enzymes, acyl-CoA dehydrogenase (plus electron-transferring flavoprotein), enoyl-CoA hydratase, -hydroxyacyl-CoA dehydrogenase, and acyl-CoA acetyltransferase. In principle, each of these molecules could be influenced by prolonged hypoxia. However, because the enzymes involved in -oxidation resemble the proteins of the "proper" respiratory chain, whereas catabolism of palmitoyl-carnitine is the one that is specifically decreased in the right ventricle, one could postulate that chronic lack of O₂ and/or consequences thereof target one of the inner membrane transfer proteins. Indeed, changes in the activity of carnitine palmitoyltransferase I have been noted previously (66) after a 5-h incubation of neonatal cardiac myocytes in substrate- and (essentially) O₂-free media. Irrespective of the mechanism, the present data suggest that limitation of O₂ availability, rather than chronic pressure overload, is a more important factor in determining the final level of tissue utilization of long-chain fatty acids.

The left ventricle, unlike the right, underwent a moderate loss of oxidative capacity in response to chronic hypoxia. The reduction was such that absolute rates of oxidation became similar in both ventricles, thereby eliminating the differences between the chambers seen in the normoxic rats (see also Ref. 25). The general decline of oxidative capacity in the left ventricle of the hypoxia-adapted rat suggests that the number of mitochondria per myocyte is reduced or the activity (or content) of key enzymes within the metabolic pathways of the mitochondria is decreased (or downregulated) in an apparently coordinated manner. Supporting the latter notion are data which show that hypoxia simultaneously decreases the maximal velocity of several enzymes of the tricarboxylic acid cycle and cytochrome aa₃ content of rat skeletal muscle cells and mouse lung macrophages in culture (48). Exposure of mitochondria isolated from embryos of Artemia franciscana (brine shrimp) to anoxic media has been shown to result in rapid decline of protein synthesis (31); a similar effect, a reduction by 90%, was reported as an early response to hypoxia in turtle hepatocytes (32, 33; for review see Ref. 19). Protein biosynthesis is an ATP-demanding process that is very sensitive to a lack of the nucleotide and a decrease in the ratio of triphosphate to diphosphate nucleotides (23). It is possible, therefore, that a fall in the synthesis of the relevant polypeptides is responsible for the fall in oxidative capacity of the left ventricle described here. Interestingly, incubation of brine shrimp mitochondria with the respiratory chain inhibitors cyanide and antimycin A had little effect on protein synthesis, which suggests that the process may be regulated directly by O₂ concentration.

An alternative explanation for the decline in oxidative capacity of the left ventricle is that it occurs in response to a decrease in energy demand and/or substrate availability, which accompany, or result from, exposure to low PO₂. The content of mitochondrial enzymes within tissue is not necessarily an inherent property of a particular cell type but, rather, may be coupled, over the time average, to the cellular need for ATP (9, 52). For example, increasing energy demand by chronic endurance-type exercise (21) or by thyroid hormone treatment (41) enhances mitochondrial cytochrome content in skeletal muscle and in the heart, respectively. By contrast, limb immobilization lowers ATP requirements by the affected musculature, and, as a consequence, oxidative capacity falls (21), whereas severe restriction of substrate supply, i.e., starvation, results in a marked loss of mitochondrial proteins (60). It has been shown that ascent to high altitude results in weight loss, diminished food intake, altered absorption of nutrients, lethargy, and modifications of protein synthesis (for review see Ref. 27). Moreover, normobaric or hypobaric hypoxia was found to depress whole body O₂ consumption and temperature (14; for brief review see Ref. 37). The hypoxia-adapted animals also exhibited signs of growth retardation (Table 1). Although neither whole body O₂ consumption, motor activity, nor food intake was monitored in this study, it cannot be ruled out that locomotion and other specific ATP-requiring processes were suppressed during hypoxic adaptation. In this case, a fall in mitochondrial enzyme activity of the left ventricle might be expected to result from a lowering of cardiac work. This conclusion is consistent with the suggestion of Hochachka and co-workers (19) that suppression of metabolic work and sparing of O₂ for ATP generation is a more efficient survival strategy than enhancing ventilatory rate for O₂ convection.

During O₂ deprivation, substrate-level phosphorylations within the glycolytic pathway and the tricarboxylic acid cycle may supplement the need for ATP generated by oxidative mechanisms. In the present study, hexokinase activity was elevated by nearly twofold in the right ventricle, which should have augmented glycolytic energy production and compensated, to some extent, for the loss of oxidative capacity. The relative abundance of mRNA for hexokinase I and II was also increased in the same chamber, suggesting a rise in the content of the relevant protein. However, this followed, rather than preceded, the change in activity. Similar apparent lack of synchronization for hexokinase II has been reported previously for skeletal muscle subjected to chronic electrical stimulation (20). Such behavior indicates that a simple precursor-product relationship may not apply to all conditions. In response to changing energy demands, stimulation of hexokinase activity may arise from association of the enzyme with the mitochondrial membrane, thereby taking advantage of locally produced ATP (44). It is also possible that each enzyme operates best at a certain range of velocities, and only when that is exceeded must the amount of the protein itself increase (9). This ensures maintenance of almost constant, optimal activity per unit of enzyme and explains the temporal relationship between hexokinase activity and the amount of its message seen in the present work.

Enhancement of glycolytic capacity seen here in hypoxia-adapted animals likely represents a response to chronic ventricular overload and to hypoxia per se. Earlier studies have indicated that, depending on the prevailing hormonal state, phosphorylation of glucose is rate limiting for its utilization during sustained contractile activity (20, 26). Expression of GLUT-1, the minor transporter isoform, was shown previously to be greater in the hypoxic (14 days of adaptation) than in the normoxic rat heart and significantly higher in the right than in the left ventricle, suggesting an "additive" effect of pressure overload and hypoxia (58). A similar trend for hexokinase mRNA expression and enzyme activity was found in the present work. Increased glycolytic capacity resulting from chronic overload, but in the absence of hypoxia, has been reported previously by Taegtmeyer and Overturf (62). In the latter study, this was accompanied by a change in the proportion of lactate dehydrogenase isozymes, i.e., from the cardiac to the skeletal muscle type (see also Refs. 3 and 13). In our hands, the total activity of lactate dehydrogenase was unaffected. There was, however, an apparent upregulation in both ventricles of the A isoform of lactate dehydrogenase. Lactate dehydrogenase, which is "better" suited for anaerobic metabolism, has been shown to be induced by hypoxia in cultured cells (12, 48) and the human heart (15).

Our results show a marked, about twofold, increase of myocardial aspartate concentration and a lowering of glutamine concentration in both ventricles, which suggest that metabolism of these two amino acids is interrelated. Amino acids are not major substrates for generation of ATP in the heart, although it has been postulated that in hypoxia-ischemia myocardial glutamate helps provide high-energy phosphates through substrate-level phosphorylation (45, 54, 69). However, glutamate could also elicit beneficial effects in hypoxia by other mechanisms. This amino acid is produced in the cardiac muscle from glutamine [the high level of which in plasma ensures an abundant supply to the myocardium (30)] by a phosphate-activated glutaminase (40, 43) and can provide 2-oxoglutarate via the action of glutamate dehydrogenase or aminotransferases. Increased provision of 2-oxoglutarate protects against depletion of the tricarboxylic acid cycle intermediates and, hence, preserves oxidative capacity (7). Administration of glutamine has been shown to be cardioprotective, i.e., to preserve mechanical function and levels of high-energy phosphates, during acute periods of ischemia-reperfusion (28). The decrease in the glutamine-to-glutamate concentration ratio in the right ventricle of the hypoxic animals may be indicative of a greater contribution by glutamine to support the increased ATP requirements of the overloaded tissue. On the other hand, enhanced operation of alanine aminotransferase allows glycolysis to proceed without accumulation of lactic acid (with alanine as the end product that is readily lost via circulation), whereas generation of aspartate by aspartate transaminase supports the malate-aspartate shuttle. The latter is the main mechanism in the heart for transport of reducing equivalents from the cytosol to the mitochondrion (53). Our finding that the aspartate-to-glutamate concentration ratios in hearts from hypoxic animals were increased indicates (56) that flux through the malate-aspartate shuttle was decreased under O₂-limited conditions.

The molecular mechanisms and cellular processes responsible for transforming cells such as cardiac myocytes, which are normally sensitive to hypoxia, to cells that are more tolerant of O₂ deprivation are not known. Unicellular organisms are capable of adjusting the levels of energy-producing proteins, in particular the terminal oxidases, toward aerobic or anaerobic respiration (cytochrome bo to cytochrome bd), depending on the availability of O₂ (for review see Ref. 5). Recently, it has been reported that the DNA-binding activity of transcription factors involved in coordination of mitochondrial protein expression in eukaryotes, i.e., the nuclear respiratory factors NRF-1 and NRF-2 (10; for review see Ref. 55), is modulated by the redox state of the cell (36). In addition, low levels of O₂ activate a transcription factor, termed hypoxia-inducible factor, HIF-1. The latter, which affects the upregulation of erythropoietin, vascular endothelial growth factor, and a number of glycolytic enzymes (5, 42, 67), may also be under redox control (22). Therefore, it is conceivable that O₂ directly, or via its influence on the cellular metabolic state, regulates the expression of proteins within the energy-producing pathways, thus enabling a defensive response to hypoxic environments. A greater understanding of the molecular mechanism(s) underlying these adaptive changes may offer improved strategies for therapeutic interventions needed for O₂-related pathology.