Alanine and glutamine play central roles in substrate cycling and interorgan nitrogen (N) metabolism. Alanine, which primarily arises de novo from the transfer of N from glutamate to pyruvate, represents the fundamental link between amino acid and carbohydrate metabolism and is a key substrate in glucose and urea production (12). Glutamine can also be derived from de novo synthesis from glutamate (equation 1), which in turn is derived from α-ketogluterate: Equation  where NH4+, ammonium; Pi, inorganic phosphate; and H+, hydrogen ion, and is coupled with the TCA cycle and glucose metabolism through its carbon skeleton, α-ketoglutarate (26). In addition, glutamine has unique functions, which range from modulating liver metabolism (16) and muscle protein synthesis (22) to serving as the main oxidative fuel for immunocytes and enterocytes (24).
During exercise there are increases in muscle protein turnover (6,10), the muscle efflux of ammonia (11) and alanine (1,11,13), and the oxidation and turnover of leucine (35). All these responses suggest that a significant increase in net protein catabolism occurs during exercise. Furthermore, since during exercise alanine and glutamine together represent ∼80% of the body's free amino-nitrogen (amino-N) pool (4) and the rate of appearance (Ra) of alanine approximately doubles (7), it seems likely that the total appearance of amino-N is also enhanced in exercise. It follows that if the availability of these amino acids increases, then the production of urea, the end product of nitrogen catabolism, should also increase. However, urea production is apparently unaffected by exercise (35). Whereas an infusion of alanine at rest stimulates ureagenesis (36), glutamine is generally the primary source of urea nitrogen (25). Since the appearance rate of glutamine in plasma has not been assessed under these conditions, it is possible that the sum total of amino-N available to the liver (represented by alanine + glutamine) for hepatic ureagenesis may not increase during exercise, if alanine formation is favored over glutamine synthesis. The possibility of this occurring during exercise is supported by the observation that alanine production is generally limited by pyruvate availability (12), and pyruvate availability should be increased during exercise. In this case, glutamine production would be limited by a decrease in the availability of glutamate, which could occur not only because of an increased transamination to alanine, but also because of a depletion of the TCA-cycle intermediate α-ketogluterate. The primary purpose of our study was therefore to quantify both alanine and glutamine kinetics during exercise of moderate intensity to determine the sum total of alanine and glutamine flux.
Subjects. Six male volunteers of average fitness (age 25 ± 2 yr, weight 68 ± 2.5 kg, V˙O2max 43 ± 2.4 mL·min−1·kg−1; means ± SE) participated in the study. Each subject passed a screening, which consisted of a medical history, physical examination, 12-lead electrocardiogram, blood count, plasma electrolytes, and liver and renal function tests. Informed consent was obtained after careful explanation of the study design, purpose, and possible risks. The experimental protocol was approved by the Institutional Review Board of the University of Texas Medical Branch at Galveston.
Preliminary testing. One or two days before the resting experiment, each subject's maximal oxygen uptake (V˙O2max) was tested by an incremental treadmill (Quinton 65, Seattle, WA) exercise test beginning at 6 mph at a 5% grade for 2 min with successive increases in speed and grade every minute until exhaustion.
The V˙O2max was considered to be attained when successive V˙O2 measurements failed to increase significantly despite increases in speed or grade (i.e., O2 demand) (31). V˙O2 was measured with an automated system for measuring oxygen and CO2 concentration and ventilation (Horizon Sensormedic, Anaheim, CA) using a noseclip and mouthpiece system.
Experimental design. Subjects were instructed to eat their normal diets in the days before the study and to not exercise. On the night before the study they were given a standard hospital meal and then fasted overnight until the start of the study the following day. A 20-gauge, 1.5-inch (Baxter Quick-Cath, Deerfield, IL) or, preferably, a 19-gauge 8-inch catheter (Intracath, Becton Dickinson, Sandy, UT) was inserted in an antecubital vein for the infusion of isotopes. The sampling catheter (20-gauge, 1.5-inch) was inserted into a vein near the wrist in the contralateral arm, and the hand and forearm were covered with a heating pad to permit sampling of "arterialized" venous blood (23). This procedure does not provide pure arterial blood, as reflected by an average PO2 of approximately 80-85 mm Hg, but nonetheless represents a well-mixed sample. The infusion and sampling protocol corresponded to this exercise study (see below).
The exercise infusion protocol was performed on a separate day within 2 wk of the resting protocol. For the exercise protocol, each subject began a fast-paced walk at a predetermined speed and grade that corresponded to ∼45% of their V˙O2max. The metabolic disequilibrium that occurs during the first few minutes of exercise renders curve-fitting untenable. On the other hand, within 5 min of exercise at this moderate intensity most physiological parameters such as V˙O2 kinetics have stabilized (29). Therefore, the first 10-15 min of exercise served as a preliminary period to allow equilibration of the cardiovascular and physiological parameters that rapidly change during the first few minutes of exercise. After this exercise equilibration period a background blood sample was drawn (10 mL) and the infusion of isotopes was started. Either primed-constant infusions (N = 3 each rest and exercise) or bolus injections (N = 3 each rest and exercise) of [2,3,3,4,4-2H2] glutamine (14 μmol·kg−1 prime, 0.15 μmol·min−1·kg−1 constant infusion, 40 μmol·kg−1 bolus) and [3-13C] alanine (22.5 μmol·kg−1 prime, 0.25 μmol·min−1·kg−1 constant infusion, 30 μmol·kg−1 bolus) were given. The boluses yielded enrichment decay curves for curve-fitting analysis and the plateau enrichments from the constant infusions were used for calculation of data. Both methods were used to calculate whole body rate of appearance (Ra) values, which are equal to their corresponding rate of disappearance (Rd) values (at steady state). Mathematically, these two tracer approaches yield the same value for Ra (8), and empirical results show that similar values are obtained with the two techniques (3,34) Blood was drawn once every minute up to 10 min, every ∼2.5 min up to 20 min and at 25, 30, 35, 45, 60, 75, 90, 105, 120, 150, and 180 min. The time corresponding to the midpoint between beginning and end of each blood draw (range 3 to 20 s/blood draw) was recorded for curve-fitting. Whole blood samples were immediately placed on ice. Within 1 h from the time that the blood was drawn, plasma was separated by centrifugation and frozen at −20°C until further processing. The protocol at rest was the same as the exercise protocol except that the subjects remained recumbent in bed throughout.
Sample analyses. Plasma proteins were precipitated with an equal volume of ice-cold 10% trichloroacetic acid (10% TCA). The supernatants were immediately loaded onto small (∼2 mL bed volume) prewashed cation-exchange columns (columns were prewashed successively with 5 mL each of 1 M NaOH, distilled and deionized water (ddH2O), 1 M HCl, and ddH2O before sample loading). After loading, each column was washed with 5-7 mL of ddH2O and the amino acids were eluted into labeled glass screw top tubes with 5 mL of 3 M ammonium hydroxide. The eluant was vortexed and half was added to another screw top tube so that half could be processed for alanine enrichment and the other half for glutamine enrichment. The samples were then quickly frozen and lyophilized to dryness.
Plasma alanine enrichments were determined using the N-acetyl,N-propyl ester (NAP) derivative as previously described (34). Chemical ionization gas chromatography/mass spectrometry (GC/MS) analysis was carried out on m/z 174.1 and 175.1.
Glutamine enrichment was determined using the tetratertiarybutyldimethylsilyl (TBDMS) derivative. Briefly, 100 μL each of acetonitrile (ACN) and N-methyl-N-(tertiary-butyldimethylsylil) trifluoroacetamide (MTBSTFA) was added to the dried eluant residue. The mixture was vortexed for 30 s and heated at 95°C for 1 h. Selected ion monitoring was carried out on ions at m/z 431.3 and 436.3 (33). Enrichment data are expressed as the tracer to tracee ratio (TTR). The TTR is analogous to the specific activity term used in radioactive tracer studies (34).
Calculations. The traditional single pool Ra values from the constant infusion experiments were determined from the alanine and glutamine enrichments (TTR) at plateau using the following equation: Equation  where I is the tracer infusion rate (μmol·min−1·kg−1) and E is the enrichment at plateau. This value represents the rate of amino acid entry into the blood.
For the bolus injection experiments, the first step in assessing the kinetics of alanine and glutamine was to fit the enrichment decay curve data to single, double, and triple exponentials (MLAB, Civilized Software, Bethesda, MD). The exponential equations were of the form: Equation  where Ai and λi are the macroparameters determined from curve-fitting. Next, the goodness of fit was evaluated visually, as well as by examining the coefficient of correlation (r2) and the Akaike information criterion (AIC; (2)).
The Ra into the sampled pool was then determined from the bolus dosage (μmol·kg−1) and the macroparameters of the exponential equation that best fit the enrichment decay data: Equation  where (A1/λ1 + A2/λ2 + ..) is equal to the area under the curve. The initial pool size (Q) was calculated from the bolus dosage divided by the TTR at time zero, which, from equation 3.2 at t = 0, is equal to (A1 + A2 + ...): Equation  where Q is the initial pool size.
Statistics. Paired t-tests were used to compare the rest and exercise data. Statistical significance was set at P < 0.05 for all tests. Values are means ± SE unless otherwise noted.
Enrichment versus time data curve-fitting. Representative examples of curve-fitting of the enrichment vs time data are shown in Figures 1 and 2. The coefficient of correlation (r2) was greater for the three exponential (0.999 ± 0.000) than for the two exponential (0.993 ± 0.002) curve-fits. More importantly, the three exponential curve-fits had much lower AIC values (see methods section) than the corresponding two exponential curve-fits (−222 ± 7 vs −133 ± 10, respectively). Thus, the kinetics of alanine and glutamine were better described by a sum of three exponentials than two exponentials.
Comparison of two tracer methods. There were no significant differences for alanine or glutamine kinetics for the two different tracer techniques. Consequently, the values were pooled, resulting in N = 6 for comparison of kinetic parameters at rest versus during exrcise.
Alanine and glutamine kinetics parameters (Table 1). Alanine Ra was significantly higher during exercise than at rest (P < 0.01). This difference was significant (P < 0.01) even if only the three subjects receiving the constant tracer infusion were considered. The values for glutamine Ra were not significantly different between methods or between rest and exercise. The sum total of alanine and glutamine N flux was significantly greater during exercise.
Catheter placement observation (Fig. 3). During a nonprimed constant infusion experiment in a resting subject, peripherally placed (forearm vein) 1.5 in catheters were associated with erratic and unusable enrichment data curves.
The results of the present study indicate that the sum of alanine + glutamine release into the circulation increases in exercise at 45% of V˙O2max. This is predominantly a result of the significant increase in alanine release, as there was not significant change in the rate of release of glutamine.
During exercise, accelerated amino acid oxidation (32,35) and purine nucleotide catabolism (28) produce an increased nitrogen load, which must be either incorporated into other compounds or released as free ammonia. Potentially toxic hyperammonemia is prevented by the incorporation of nitrogen into amino acids (primarily alanine and glutamine). The relative production rates of alanine and glutamine (5,13) and ammonia (11,29) appear to be determined to at least some extent by the intracellular availability of pyruvate, α-ketoglutarate, and other important nitrogen-acceptor carbon skeletons. Both the intramuscular availability of pyruvate from glycogenolysis and nitrogen from amino acid and purine nucleotide catabolism increase with increasing exercise intensity (e.g., (29)). Thus, it is not surprising that the absolute rate of de novo alanine production and the release of alanine from active muscle (11) both rise progressively as the exercise intensity is increased.
The lack of a significant exercise-induced stimulation of glutamine Ra is consistent with previous data in the literature. A recent study by Eriksson et al. (11) found no changes in the net release of glutamine from muscle during 35, 55, and 80% V˙O2max cycle ergometer exercise. Also, although the muscle glutamine concentration may fluctuate, the plasma glutamine concentration remains very stable relative to other amino acids and does not usually increase except in cases of hyperammonemia (15). Further, a limitation in the de novo synthesis of glutamine in muscle might be expected during exercise because the intramuscular concentration of the glutamine precursor, glutamate, drops several fold within a few minutes after exercise onset and remains low throughout exercise (1,17). On the other hand, a previous study suggested that glutamine efflux from muscle increases in exercise (21). If this is the case, then the unchanged whole-body Ra of glutamine observed in the present study could only be explained by the fact that it is offset by a decreased release from sources other than exercising muscle such as lung (27) or liver (16). For example, the finding that a reduced insulin/glucagon ratio simultaneously inhibits perivenous glutamine synthetase and stimulates periportal glutaminase in vitro (16) suggests that the reduced insulin/glucagon ratio brought about by exercise could act to diminish hepatic glutamine release in vivo. If the increased uptake of glutamine by the gut seen during exercise in dogs (30) occurs in humans, then the sum of glutamine uptake in extrasplanchnic regions must be concomitantly decreased, since the plasma concentration and total Rd of glutamine did not change.
At rest, alanine and glutamine together represent over 60% of the amino-N released from muscle (5). This is far in excess of their percent content in protein (20), which indicates that they primarily arise from de novo synthesis from other amino acids (14). Similar calculations were performed to determine whether this finding holds true during exercise as well. Data from the present study indicate that a total of about 200 mmol of alanine and 90 mmol of glutamine were released into the circulation over the 180 min of exercise (e.g., for alanine, 13.5 μmol·min−1·kg−1 × 68 kg × 180 min; Table 1). Active muscle is likely to be responsible for at least 60% of whole body alanine and glutamine Ra (10). Using an active muscle mass of ∼9 kg for ∼45% V˙O2max exercise (from comparison to the carefully calculated value of ∼10.8 kg active muscle during 65% V˙O2max cycling by Jansson et al. (18)) indicates that decreases of 12 and 6 mmol·kg−1 in free intramuscular alanine and glutamine stores, respectively, would have been necessary if the observed changes in Ra were exclusively derived from a depletion of intramuscular stores. Maximal rates of alanine and glutamine appearance from proteolysis (∼2 and 1 μmol·min−1·kg−1, respectively, e.g., from (19,37)) represent ∼10% of their respective Ra values during exercise. Therefore, similar to rest, the vast majority of alanine and glutamine released into the circulation over the 3 h of exercise was derived from de novo synthesis rather than from proteolysis.
Although we did not determine the fate of the accelerated disappearance of alanine from the circulation during exercise, it seems likely that the splanchnic region was primarily responsible for clearing alanine during exercise, presumably for increased gluconeogenesis (12). This is supported by the results of Ahlborg et al. (1), who found that the splanchnic fractional extraction of alanine increased from 40 to >90% after 4 h of exercise while the contribution of alanine to gluconeogenesis increased proportionately. Alanine utilization by the liver for gluconeogenesis, anaplerosis (replenishment of TCA cycle intermediates), or energy production could explain the fate of the carbon skeleton portion of the increased release of alanine. The fate of the amino-N of alanine is less clear, since in a previous study it was not found to appear in urea (37). Increased incorporation of alanine into acute phase proteins is one possible fate in the liver (9).
The most pertinent physiological parameter for the study of interorgan cycling is the rate of substrate release into the circulation. Venous infusion with "arterialized" venous sampling (i.e., v-a mode) was used in the present study because, aside from the v-a mode being more practical, it also more closely reflects the rate of release of substrate into the circulation than the a-v mode (arterial infusion with venous sampling). Noncompartmental analyses of enrichment data provide accurate and convenient measurements of the release of substrate into the sampled compartment (the circulation). In theory, the plateau enrichment method and the bolus injection curvefitting procedure should yield the same value since they both measure the appearance of substrate into the blood (34). Consistent with this expectation, the curve-fit Ra values were similar to the plateau Ra values obtained in the present study, as well as to previously published rest and exercise alanine Ra values (36).
Only data from subjects who received tracer through an 8 in catheter were used in this study. The erratic enrichments shown in Figure 3 indicate that infusions administered through a short (1.5 in) distally placed (antecubital vein) catheter are susceptible to intermittent venous stasis. This apparent opening and closing of peripheral venous valves can significantly complicate tracer studies in conditions such as at rest, when limb blood flow is relatively low. This observation, along with the finding that longer (8-inch) or more proximally placed catheters (e.g., biceps vein) avoid the intermittent blood flow problem, is of obvious practical significance to tracer infusion studies.
In summary, the present study represents the first measurements of whole-body glutamine Ra during exercise in humans. The curve-fitting and plateau enrichment methods provided similar results: whereas the glutamine Ra remained unchanged from rest to exercise, alanine Ra increased two- to three-fold. In addition, the derived sampled pool size of these amino acids was within the predicted size of its plasma pool. Evidence was provided for the contention that the majority of the release of these amino acids arises from de novo synthesis rather than proteolysis or a simple depletion of intracellular stores. Thus, alanine, rather than glutamine, appears to be the predominant N-carrier involved in the transfer of amino-N to the liver during moderate-intensity exercise. Furthermore, it appears that the previously-reported failure of urea production to increase acutely during exercise cannot be accounted for by a decrease in glutamine Ra to offset the increase in alanine Ra.
1. Ahlborg, G., P. Felig, L. Hagenfeldt, R. Hendler, and J. Wahren. Substrate turnover during prolonged exercise in man. J. Clin. Invest
2. Akaike, H. A new look at the statistical model identification. Trans. Automat. Contr.
3. Allsop, J. R., R. R. Wolfe, J. J. Destephano III, and J. F. Burke. The reliability of rates of glucose appearance in vivo calculated from single tracer injections. Canad J. Physiol. Pharm
4. Bergstrom, J., P. Furst, and E. Hultman. Free amino acids in muscle tissue and plasma during exercise in man. Clin. Physiol.
5. Bier, D. M. Intrinsically difficult problems: the kinetics of body proteins and amino acids in man. Diabetes/Metab Rev.
6. Biolo, G., S. P. Maggi, B. D. Williams, K. D. Tipton, and R. R. Wolfe. Increased rates of muscle protein turnover and amino acid transport after resistance exercise in humans. Am. J. Physiol.
7. Carraro, F., A. Naldini, J. M. Weber, and R. R. Wolfe. Alanine kinetics in humans during low-intensity exercise. Med. Sci. Sports Exerc.
8. Carson, E. R., C Cobelli, and L. Finkelstein. The Mathematical Modelling of Metabolic and Endocrine Systems.
New York: Wiley & Sims, 1983.
9. Carraro, F., W. H. Hartl, C. A. Stuart, D. K. Layman, F. Jahoor, and R. R. Wolfe. Whole body and plasma protein synthesis in exercise and recovery in human subjects. Am. J. Physiol. (Endocrinol. Metab.
10. Dohm, G. L., E. B. Tapscott, and G. J. Kasperek. Protein degradation during endurance exercise and recovery. Med. Sci. Sports Exerc.
11. Eriksson, L. S., S. Broberg, O. Bjorkman, and J. Wahren. Ammonia metabolism during exercise in man. Clin. Physiol.
12. Felig, P. The glucose-alanine cycle. Metabolism
13. Felig, P. and J. Wahren. Amino acid metabolism in exercising man. J. Clin. Invest
14. Gorski, J., D. A. Hood, O. M. Brown, and R. L. Terjung. Incorporation of 15N-leucine amine into ATP of fast-twitch muscle following stimulation. Biochem. Biophys. Res. Comm.
15. Harper, A. E., R. H. Miller, and K. P. Block. Branched chain amino acid metabolism. Ann. Rev. Nutr.
16. Haussinger, D. Liver glutamine metabolism. J. Parent Ent. Nutr.
17. Henriksson, J. Effect of exercise on muscle amino acid concentrations in skeletal muscle and plasma. J. Exp. Biol.
18. Jansson, E., and L. Kaijser. Substrate utilization and enzymes in skeletal muscle of extremely endurance-trained men. J. Appl. Physiol.
19. Kominz, D. R., A. Hough, P. Symonds, and K. Laki. The amino acid composition of actin, myosin, tropomyosin, and the meromyosin. Arch. Biochem. Biophys.
20. Krebs, H. A. Glutamine metabolism in the animal body. In: Glutamine: Metabolism, Enzymology, and Regulation,
J. Mora and R. Palacios (Eds.). New York: Academic, 1980, pp. 319-329.
21. MacLean, D. A., T. E. Graham, and B. Saltin. Branched chain amino acids augment ammonia metabolism while attenuating protein breakdown during exercise. Am. J. Physiol.
22. MacLennan, P. A., R. A. Brown, and M. J. Rennie. A positive relationship between protein synthetic rate and intracellular glutamine concentration in perfused rat skeletal muscle. FEBS Lett.
23. McGuire, E. A. H., J. H. Helderman, J. D. Tobin, R. Andres, and M. Berman. Effects of arterial vs venous sampling on analysis of glucose kinetics in man. J. Appl. Physiol.
24. Newsholme, E. A., P. Newsholme, R. Curi, M. A. Challoner, and M. S. M. Ardawi. A role for muscle in the immune system and its importance in surgery, trauma, sepsis, and burns. Nutrition
25. Nissim, I., C. Cattano, and M. Yudkoff. Relative role of glutaminase, glutamate dehydrogenase, and AMP-deaminase pathways in hepatic ureagenesis: studies with 15N. Arch. Biochem. Biophys.
26. Smith, R. J. Glutamine metabolism and its physiologic importance. J. Parent Ent. Nutr.
14(4 Suppl.):40S-44S, 1990.
27. Souba, W. W., K. Herskowitz, and D. A. Plumely. Lung glutamine metabolism. J. Parent Ent. Nutr.
28. Spencer, M. K. and A. Katz. Role of glycogen in control of glycolysis and IMP formation in human muscle during exercise. Am. J. Physiol.
29. Wagenmakers, A. J. M., E. J. Bechers, F. Brouns, et al. Carbohydrate supplementation, glycogen depletion, and amino acid metabolism during exercise. Am. J. Physiol.
30. Wasserman, D. H., R. J. Geer, P. E. Williams, T. A. Becker, D. B. Lacy, and N. N. Abumrad. Interaction of gut and liver on nitrogen metabolism during exercise. Metabolism
31. Whipp, B., K. Wasserman, J. A. Davis, N. Lamarra, and S. A. Ward. Determination of O2
kinetics during exercise in man. In: Exercise Bioenergetics and Gas Exchange,
P. Cerretelli and B. J. Whipp (Eds.). Amsterdam: Elsevier/North-Holland, 1980, pp. 175-185.
32. White, T. P. and G. A. Brooks. U-14C-glucose, alanine, and leucine oxidation in rats at rest and at two intensities of running. Am. J. Physiol.
33. Williams, B. D. and R. R. Wolfe. Determination of amino- and amide-15N glutamine enrichment with tertiarybutyldimethylsilyl (TBDMS) derivatives. Biol. Mass. Spectrom.
34. Wolfe, R. R. Radioactive and Stable Isotope Tracers in Biomedicine: Principles and Practice of Kinetic Analysis.
New York: Wiley-Liss, 1992, pp. 1-417.
35. Wolfe, R. R., R. D. Goodenough, M. H. Wolfe, G. T. Royle, and E. R. Nadel. Isotopic analysis of leucine and urea metabolism in exercising humans. J. Appl. Physiol.
36. Wolfe, R. R., F. Jahoor, and J. H. F. Shaw. Effect of alanine infusion on glucose and urea production in man. J. Parent Ent. Nutr.
37. Wolfe, R. R., M. H. Wolfe, E. R. Nadel, and J. H. F. Shaw. Isotopic determination of amino acid-urea interactions in exercise in humans. J. Appl. Physiol.
Key Words: ALANINE; GLUTAMINE; AMINO ACIDS; CARBOHYDRATE METABOLISM
© Williams & Wilkins 1998. All Rights Reserved.