Main

Glucose has been thought to be the primary substrate for energy metabolism in many tissues, including the developing lung(1). However, several studies have demonstrated that glutamine may be used as an energy substrate for the TCA cycle in several developing organs, particularly the gastrointestinal tract(2) and brain(3–5).

Utilization of alternate substrates depends on both the availability of the substrates and the ability of these substrates to be completely oxidized via the TCA cycle. Glutamine levels in plasma are about 0.5 mM, the highest concentration of any circulating amino acid(6). Therefore glutamine is abundant and available as an alternate substrate. For glutamine to be completely oxidized, either malic enzyme or phosphoenolpyruvate carboxykinase must be present to allow a portion of the accumulating malate to be converted to pyruvate, because the four carbons from oxaloacetate are maintained in the TCA cycle, whereas only the carbons entering from pyruvate (or acetyl CoA) are oxidized to CO2. Using the rat as an animal model, we have recently reported that both cytosolic and mitochondrial malic enzyme activities are present in the rat lung and that malic enzyme activity is significantly higher in the newborn(7). In the conversion of malate to pyruvate via malic enzyme, NADPH is produced. In the cytosol, this NADPH can be used in the production of fatty acids that are used in the production of surfactant(8).

Two important factors which contribute to the regulation of substrate utilization are the compartmentation of metabolic activity and the trafficking of substrates for energy metabolism within different areas of a cell, or between different types of cells within an organ. Metabolic trafficking and compartmentation of substrate oxidation has been reported between different cell types of the lung(9). An elegant method has been developed by McKenna et al.(10, 11) for evaluating this compartmentation as seen in Figure 1. With substrate competition studies, if there is a decrease in radioactive CO2 production from a 14C-labeled substrate in the presence of an added unlabeled substrate, this has been interpreted as indication that both compounds are metabolized in the same compartment. Conversely, if both compounds are known to be oxidized for energy, but do not affect each other's oxidative metabolism, it is interpreted that these compounds are metabolized in separate compartments. A third situation may exist where an unlabeled compound inhibits the oxidation of a labeled compound; however, when the reciprocal combination is examined, there is no decrease in the oxidation of the labeled second compound in the presence of unlabeled first compound, suggesting that the first compound is metabolized in both compartments, but the second compound is metabolized in only one. There are a number of cellular and subcellular factors which contribute to the compartmentation of metabolic activity. These may be different cell types or a number of subcellular factors. There may be unique transporter molecules at the cell membrane and at the subcellular level which also contribute to the regulation of substrate use in lung cells. In the brain, there is evidence of mitochondrial heterogeneity within a relatively pure population of brain cells(12–14), and this may also exist within type II cells. Another phenomenon which contributes to compartmentation is the association and dissociation of multienzyme complexes which are important in directing the flow of carbons from key compounds through the TCA cycle(15–18). The dynamic association and dissociation of these complexes (e.g. mitochondrial malate dehydrogenase and α-ketoglutarate dehydrogenase together bind to either citrate synthase or mitochondrial aspartate aminotransferase) is regulated by the concentration of key metabolites such as α-ketoglutarate and oxaloacetate in the intracellular milieu(15, 16, 18).

Figure 1
figure 1

Substrate competition studies (hypothetical rates of oxidation). This diagram explains the logic for using competition studies to demonstrate metabolic compartmentation. In the left column are two14 C-labeled substrates, A and B, incubated alone, or with the unlabeled other substrate. The right sided column is the possible relative rates of14 CO2 production for each combination. In situation #1, there is equal, reciprocal decrease in 14CO2 production, so the two substrates A and B share the same metabolic compartment, and are thus competitors. In situation #2, there is no reciprocal decrease from the baseline 14CO2 rates, so A and B are metabolized in separate compartments. In situation #3, unlabeled B has no effect on14 CO2 production from 14C-A, however, unlabeled A decreases 14CO2 production from labeled B, suggesting that A is metabolized in both compartments, but B is metabolized in only one. (Courtesy of Mary McKenna, Ph.D.)

Full size image

Plasma levels of alternate substrates such as glutamine, lactate, and ketone bodies are high during fetal and early neonatal periods. Furthermore, these substrates may be preferentially used for energy and for synthesis of surfactant in the type II pneumocyte, a cell type crucial to normal lung maturation and function.

For the present study, we measured rates of 14CO2 production from the oxidation of labeled glucose, glutamine, the ketone body 3-hydroxybutyrate, and lactate, and determined the effect of added unlabeled substrates as competitors on labeled substrate oxidation in d 19 fetal type II cells. We chose this gestational age because it is the time during which the rate of surfactant production increases(8) and because it is the earliest time in which adequate numbers of type II cells may be isolated to allow for these experiments to be performed. The results suggest that glutamine is oxidized in these cells, and perhaps surprisingly, that the presence of glucose, but not lactate, inhibited the oxidation of glutamine in these cells. The results clearly demonstrate that these cells oxidize the alternate substrates glutamine, lactate, and 3-hydroxybutyrate at rates considerably higher than the rate of glucose oxidation. In addition, these studies provide evidence for multiple compartments of TCA cycle activity in fetal type II cells.

METHODS

Biochemicals and reagents. Glutamine, 3-hydroxybutyrate, lactate, glucose, Tris-HCl, trypsin, and DNase were purchased from Sigma Chemical Co. (St. Louis, MO). The fluor used for liquid scintillation counting(Ready Solve EP) was purchased from Beckman Instruments, Inc. (Fullerton, CA). Solutions for the Pierce BCA microreagent protein assay were purchased from Pierce (Rockford, IL). All reagents and chemicals were of highest analytical grade, and Milli-Q water (Millipore Corp., Milford, MA) was used to prepare all reagent solutions. Radioactive compound, L-[U-14C]glutamine(>250 mCi/mmol), L-[U-14C]lactate (50-150 mCi/mmol), L-[U-14C]-3 hydroxybutyrate (10-35 mCi/mmol), and D-[U-14C]glucose were purchased from Amersham Corp. (Arlington Heights, IL) or DuPont NEN (Boston, MA).

Culture flasks were purchased from Nunc. Primaria flasks (Becton Dickinson, Franklin Lakes, NJ) were used for TII cell isolation. Culture medium (EMEM), antibiotic and antimycotic solution, and fetal bovine serum were obtained from Paragon Biotechnology, Inc. (Baltimore, MD). Solution I was composed of 140 mM NaCl, 5 mM KCl, 2.5 mM Na2HPO4, 10 mM HEPES, 6 mM glucose, and 0.2 mM EGTA at pH 7.4. Solution II was 140 mM NaCl, 5 mM KCl 2.5 mM Na2HPO4, 10 mM HEPES, 2.0 mM CaCl2 at pH 7.4.

Tissue preparation. The animal protocols in the study were approved by the University of Maryland Institutional Animal Care and Use Committee. Fibroblast and type II cells were isolated using the differential adhesion technique modified from Batenburg et al.(19).

Day 19 timed pregnant Sprague-Dawley rats (Harlan Laboratories) fed standard rat chow were anesthetized via intracardiac injection with ketamine and xylazine. The fetuses from eight to nine animals were delivered surgically. The pups were killed by decapitation before respiration could occur. The lungs were removed aseptically, and the visible bronchi were excised and then immediately placed in solution I on ice. Then the lungs were thoroughly minced, trypsinized in solution II, and then filtered through sterile 40-μm Nitex filter (Tetko, Inc., Lancaster, NY). The filtrates were centrifuged at 500 × g for 10 min at 20-25 °C. The supernatants were decanted, and the pellets were resuspended in EMEM with 20% fetal bovine serum. The centrifugation was repeated and the washed cells were resuspended in EMEM without fetal bovine serum to decrease attachment of type II cells. The cell suspension was plated in T25 flasks and incubated for 1 h at 37 °C in 95% air/5% CO2 to allow fibroblasts to adhere. To remove unattached cells, each cell suspension was triturated, combined with the cell suspension in the next flask, and triturated. The combined cell suspension from each group of six and a second wash of the same group were combined into one T75 flask. These flasks were incubated 1 h at 37 °C in the CO2 incubator. After incubation, unattached type II cells in the T175 flasks were transferred into a second set of T175 flasks and incubated for 1 h. The unattached type II cells and a wash were combined and centrifuged at 500 × g for 10 min at 20-25 °C. The pelleted type II cells were added to EMEM with 5% fetal bovine serum and plated in T25 Primaria flasks. The flasks containing the type II cell types were incubated 36-48 h at 37 °C in 95% air/5% CO2 atmosphere before the oxidation studies. The cell purity of type II cells isolated by this technique in our laboratory is 90% (J. T. Torday, unpublished data). They were judged viable by trypan blue exclusion at >95%

Oxidation experiments. Flasks containing type II cells were washed twice rapidly with PBS (pH 7.4). Flasks were then incubated for 1 h at 37 °C in buffer containing 14C-labeled substrate in 1 mM concentration in a rotary shaker. Appropriate controls were always included. The incubation was stopped by the addition of 0.3 mL of 10% TCA (vol/vol). The rate of 14CO2 was measured by trapping the 14CO2 in center wells containing methylbenzonium hydroxide. The center wells were transferred to vials, scintillation fluid was added, and activity was counted in a liquid scintillation spectrometer.

The concentration of the labeled substrates was 1 mM. Unlabeled alternate substrates were present at 2.5 mM concentration. Blanks were used as flasks without cells for each substrate and alternate substrate combination. This methodology was designed to eliminate the contribution of endogenous substrates, because, with a 1-h incubation and the concentration and amount of exogenous substrates, the effect of endogenous substrates is negligible.

Protein concentration. Protein concentrations were determined by dissolving the cells in flasks in NaOH (0.1 N). These aliquots were then assayed for protein concentration using the Pierce BCA microreagent protein assay method(20). Rates of oxidation were expressed in nanomoles of CO2 produced/mg of protein/min.

Statistical analysis. Within experiments, each substrate combination was performed using four to five flasks. The rates of oxidation for each substrate combination was repeated a minimum of three experiments. The mean ± SEM for each substrate is derived from experiments, not flasks. Data for the rates of 14CO2 production from14 C-labeled substrates were analyzed by analysis of variance and the Student-Newman-Keuls multiple range test to determine significant differences. In the competition experiments, the data are expressed as percent of control, which is the rate of oxidation of labeled substrate alone as control for each experiment. The actual rates of oxidation, rather than the percents of control, were tested for significance using an analysis of variance with Dunnett's test for multiple comparisons to control comparing the rate of14 CO2 production in the presence of each added unlabeled substrate to the control rate of 14CO2 production without added substrates. Statistical tests were performed using Sigmastat software package(Jandel Scientific, Corte Madera, CA).

RESULTS

The rates of oxidation of four 14C-labeled substrates by d 19 fetal rat lung type II cells are shown in Figure 2. The rates of glutamine oxidation was 24.36 ± 4.51 (nmol 14CO2/mg protein/min) (mean ± SEM). The ketone body, 3-hydroxybutyrate, was oxidized at a rate of 14.91 ± 1.93 (nmol 14CO2/mg protein/min). The rate of lactate oxidation was significantly higher than the glutamine oxidation rate, 40.29 ± 4.42 (nmol 14CO2/mg protein/min). Glucose was oxidized at a significantly lower rate than the other three substrates (2.13 ± 0.36 (nmol 14CO2/mg protein/min).

Figure 2
figure 2

Rates of substrate oxidation by fetal rat type II cells. The rates of substrate oxidation as measured by nanomoles of14 CO2 produced/h/mg of protein is shown for the substrates glutamine (24.36 ± 4.51), 3-hydroxybutyrate (14.91 ± 1.93), lactate (40.29 ± 4.42), and glucose (2.13 ± 0.36) (mean ± SEM). The concentration of the labeled substrate was 1 mM for each substrate. The flasks were incubated for 1 h at 37 °C. Within experiments, each oxidation was performed using four to five flasks. Each substrate was repeated a minimum of four experiments. The rate of lactate oxidation was significantly higher than the glutamine oxidation rate. The rate of glucose oxidation was significantly lower than that of glutamine. [*Compared with the rates of oxidation of glucose and glutamine; **compared with the rates of lactate and glutamine oxidation (p < 0.005).]

Full size image

Using the McKenna model as proposed in Figure 1(10, 11), we examined the effect of unlabeled substrates on labeled glutamine oxidation as shown in Figure 3. Glucose reduced the rate of glutamine oxidation by 40%. There was little effect of lactate on glutamine oxidation, as well as no effect of the ketone body, 3-hydroxybutyrate, on glutamine oxidation in the type II cells. This suggests that these substrates are oxidized in different compartments.

Figure 3
figure 3

Effect of unlabeled substrates on glutamine oxidation in d 19 fetal rat lung type II cells. The results are expressed as a percent change from the baseline rate of oxidation of glutamine alone compared with when 1 mM L-[U-14C]glutamine was incubated in the presence of unlabeled 2.5 mM lactate, glucose, and 3-hydroxybutyrate, respectively. Glucose had an effect on the 14CO2 production from labeled glutamine and decreased the rate by 40%. There was no effect of lactate or 3-hydroxybutyrate on the oxidation rate of glutamine. Data are shown mean ± SEM for a minimum of three experiments, four to five flasks/experiment. Experiments are carried out as described in “Methods.”

Full size image

In the reciprocal experiments (Fig. 4), when labeled glucose was incubated with unlabeled glutamine, in the type II cells, glucose oxidation was decreased from control by 25%. This effect was less than that seen by glucose on glutamine. Lactate, as expected, significantly inhibited glucose by 50%.

Figure 4
figure 4

Effect of unlabeled substrates on glucose oxidation in d 19 fetal rat lung type II cells. The results are expressed as a percent change in oxidation rate from the baseline rate of oxidation of 1 mM L-[U-14C]-glucose alone is shown when labeled glucose was incubated with unlabeled 2.5 mM glutamine or lactate, respectively. Glutamine decreased the rate of glucose oxidation by 20% from the baseline rate, whereas lactate decreased the rate of glucose oxidation by 50% (p < 0.05). The data are expressed as mean ± SEM for a minimum of three experiments, four to five flasks/experiment. Experiments are carried out as described in“Methods.”

Full size image

Conversely, in Figure 5, when unlabeled glucose was incubated with labeled lactate, the rate was decreased by 50%. Lactate oxidation was also significantly decreased by unlabeled glutamine by 20%; however, this effect was not reciprocal, because unlabeled lactate did not significantly decrease glutamine oxidation. Furthermore, unlabeled 3-hydroxybutyrate also significantly decreased lactate oxidation by 40%.

Figure 5
figure 5

Effect of unlabeled substrates on lactate oxidation in d 19 fetal rat lung type II cells. The results are expressed as a percent change in oxidation rate from the baseline rate of oxidation of 1 mM L-[U-14C]lactate alone compared when labeled lactate was incubated with unlabeled 2.5 mM glutamine, glucose, and 3-hydroxybutyrate, respectively. All three substrates significantly decreased lactate oxidation when compared with the baseline rate of lactate oxidation without any competitors. Unlabeled glutamine decreased the rate by 20% of control (p < 0.05), whereas 3-hydroxybutyrate decreased the lactate oxidation by 40% (p< 0.05). As expected, because lactate and glucose share common pathways into the TCA cycle, unlabeled glucose produced the greatest decrease from control (60%) (p < 0.05). The data are expressed as mean ± SEM for a minimum of three experiments, four to five flasks/experiment. Experiments are carried out as described in “Methods.”

Full size image

In type II cells the effect of two unlabeled substrates (glutamine and lactate) on labeled 3-hydroxybutyrate oxidation is shown inFigure 6. There was no effect of glutamine on 3-hydroxybutyrate oxidation. Although lactate decreased the rate of 3-hydroxybutyrate oxidation from control by 20%, this decrease was not significant.

Figure 6
figure 6

Effect of unlabeled substrates on 3-hydroxybutyrate oxidation in d 19 fetal rat lung type II cells. The results are expressed as a percent change in oxidation rate from the baseline rate of oxidation of 1 mM L-[U-14C]-3-hydroxybutyrate alone compared with when labeled 3-hydroxybutyrate was incubated with unlabeled 2.5 mM glutamine or lactate, respectively. Unlabeled glutamine had no effect on the oxidation of 3-hydroxybutyrate, whereas unlabeled lactate had a small, but not significant, decrease in the rate of the ketone body oxidation. The data are expressed as mean ± SEM for a minimum of three experiments, four to five flasks/experiment. Experiments are carried out as described in“Methods.”

Full size image

DISCUSSION

Alternate substrates other than glucose may be important sources for energy metabolism in the developing lung. Previous authors have examined the metabolism of lactate and ketone bodies in addition to glucose in both whole lung preparations(21–26) and isolated type II cells(27, 28). Many of these workers have focused on the role of these substrates as precursors to the phospholipids necessary for surfactant synthesis in the lung(24–26). Patterson and Rhoades(23) examined the rate of oxidation of glucose, lactate, and 3-hydroxybutyrate in d 19 and 21 fetal rat lung slices. They found that lactate in 5 mM concentration had oxidation rates significantly higher than that of glucose. 3-Hydroxybutyrate, at 2 mM concentration, had a rate lower than either glucose or lactate.

Our laboratory has chosen to focus our investigations on the role of alternate substrates as fuel for energy in isolated fetal type II cells. In addition to the three substrates examined by Patterson and Rhoades(23), glucose, lactate, and 3-hydroxybutyrate, we have also investigated the role of glutamine as an alternate substrate. Our data demonstrate that d 19 fetal rat type II cells use alternate substrates for energy metabolism. In contrast to the results of Patterson and Rhoades(23), we found in type II cells that glucose had the lowest rate of oxidation. Glutamine and lactate rates were at least 10 times higher than the rate of glucose oxidation in these cells.

It is possible that the rates of glucose and lactate oxidation may not reflect the overall utilization of these substrates, because glycogen content is high in type II cells of this gestational age(29). These glycogen stores are thought to be used as substrate for phospholipid synthesis that occurs at the end of gestation(29). It is possible that endogenous glycogen may dilute the observed glucose oxidation rate; however, it should also similarly affect lactate oxidation rates. Glycogen may be the precursor for both lactate and glucose, so it is unlikely that the glucose oxidation rate would be preferentially affected. In the competition studies, both lactate and glucose were reciprocally inhibited by 50% in the presence of the unlabeled other substrate. The measured rates of glucose oxidation was 1/20 that of lactate, thus, if glycogen were diluting both lactate and glucose pools, then the rates of these two should be somewhat comparable. Glutamine and 3-hydroxybutyrate which enter the TCA cycle at different points would be less likely to be altered by endogenous glycogen.

We are aware that other authors(30) have used the production of acetyl moieties to measure rates of substrate oxidation; however, because the calculation of production of acetyl moieties from glutamine would require some assumptions, we have chosen to express the rate of oxidation on the direct observation of the rate of 14CO2 production, especially because the liberation of CO2 before the formation of acetyl CoA units from either lactate, glucose, or glutamine yields significant energy.

In the fetus, lactate has been demonstrated to be in higher concentration than in the mother(31). In the type II cell, lactate had the highest rate of oxidation. Thus, our findings for lactate oxidation in isolated type II cells supports the earlier work by Patterson and Rhoades(23) in whole lung slices. However, our results for 3-hydroxybutyrate and glucose are at variance with their findings.

Our rate of 3-hydroxybutyrate may be higher than that reported by Patterson and Rhoades(23) because our studies have focused on a specific cell type, the type II cell, intimately involved in surfactant production in the developing lung. Yeh(24, 25) and others(26) have proposed that 3-hydroxybutyrate is a major substrate in the lung. In our studies, however, the oxidation rate of this compound was less than either glutamine or lactate. However, Yeh(25) had focused on the incorporation of ketone bodies into lung lipids. It may well be that there are different avenues of utilization for the various substrates.

We recognize that the examination of type II cells from other gestational ages may yield different results of substrate oxidation, but this time point was chosen because it corresponds to the time during which surfactant production in the type II cells is beginning to increase(8). Surfactant production requires NADPH. The oxidation of glutamine provides this cofactor, suggesting that glutamine oxidation may be essential in the normal development of the fetal lung.

The utilization of glutamine as an energy source by the lung has been a major focus of our laboratory. In order for glutamine to be completely oxidized, malic enzyme must be present, as previously shown by Zielkeet al.(32) in the skin fibroblast. Our previous findings have demonstrated that malic enzyme is present in both cytosolic and mitochondrial fractions in the rat lung and that both total and mitochondrial malic enzyme activity is significantly higher in the newborn than in the adult rat lung(7). In addition, we have examined rates of malic enzyme activity in fetal rat lung and the mitochondrial activity was about 2-fold higher than the cytosolic activity, and both the cytosolic and mitochondrial malic enzyme activities were significantly higher before birth than in the first 10 days of postnatal life(33). We have also measured malic enzyme activity in isolated d 19 fetal type II cells and found activity present in both cytosolic and mitochondrial fractions and the mitochondrial malic enzyme activity was higher than the activity in the cytosolic fraction (our unpublished data).

Glutamine has been shown to be an important energy source in developing gastrointestinal tract(2) and brain(3–5). Other workers have examined the role of glutamine in the lung(34–36). Examination of glutamine flux in the normal lung(34), in endotoxin-injured lung(35) and glucocorticoid-treated lung(36) have documented the increased efflux of glutamine after lung injury. Hautamaki et al.(37) examined glutamine transport into adult rat type II cells isolated in a comparable technique to the one we used. They found sodium-dependent transport into the cell with two saturable transport systems: low affinity and high affinity.

Our results above provide the first demonstration that glutamine is oxidized in type II cells in the fetal rat lung. In this cell type, the rate of glutamine oxidation was 10 times higher than the rate of oxidation of glucose. This suggests that glutamine may be a significant energy source in the developing lung. These results may be specific for type II cells, and other cell types in the lung may have different rates of oxidation of these substrates.

The use of labeled and unlabeled substrates has been used previously in our laboratory to demonstrate the presence of multiple compartments of energy metabolism in cells(4, 5, 10, 11). In the work reported in this publication, it is interesting that glucose, whose oxidation rate is 10-fold less than that of glutamine, would have such a profound effect on the oxidation of this amino acid. This suggests compartmentation within the type II cell, rather than simple dilution of substrates. This concept is supported by the finding that lactate had little to no effect on glutamine oxidation. This would further support the proposition that glucose, or one of its immediate metabolites, is acting as a specific regulator of glutamine oxidation. Competition for NADP between glucose (through the hexose monophosphate pathway), and for the conversion of malate to pyruvate, may also explain this reciprocal inhibition. Because one would expect lactate and glucose to use similar pathways into the TCA cycle, this lack of a similar effect on glutamine suggests multiple compartments of TCA cycle activity.

When 3-hydroxybutyrate and lactate were examined for the effects of competition, the effect of 3-hydroxybutyrate on lactate was significant, and the magnitude of the effect is greater when unlabeled 3-hydroxybutyrate was incubated with labeled lactate, rather than the reverse, suggesting that there is some compartmentation of these metabolites, because the mutual effects are not equivalent.

The reciprocal inhibition by glucose and lactate is not unexpected and provides evidence for the linkage between the Embden-Meyerhof pathway and the TCA cycle in these cells. Engle et al.(38) found that in the presence of lactate, glucose incorporation into total disaturated phosphatidylcholine was reduced by 35%. This would suggest that there is possible pools of oxidation and fatty acid synthesis that are in the same compartment.

Our data would suggest at least two separate compartments in the type II cells for substrate oxidation: one would be the compartment for glutamine metabolism, and a second would be the compartment for glucose. Alternatively, these data may provide indirect evidence for the specific inhibition of glutamine by glucose at one of the key enzymatic steps, either glutaminase or glutamic dehydrogenase.

This decrease of the rate of glutamine oxidation in the presence of glucose may well be connected with the phenomena in which, in times of stress, the lung has been described as a major source for glutamine(25), because, in the presence of physiologic levels of glucose, the rates of glutamine oxidation are greatly reduced.

In summary, we have documented that glutamine and other alternate substrates are oxidized preferentially over glucose for energy metabolism in the d 19 fetal rat lung type II pneumocyte. In addition, we have delineated some of the compartmentation that occurs within the developing type II cell which may determine how these substrates are used.