Journal Logo

Basic Sciences: Original Investigations

Human erythrocyte and plasma amino acid concentrations during exercise


Author Information
Medicine & Science in Sports & Exercise: July 2000 - Volume 32 - Issue 7 - p 1244-1249
  • Free


In vitro studies have consistently shown that amino acid uptake by human erythrocytes is slow (27). It is generally believed that plasma rather than erythrocytes is the vehicle of amino acid transport between tissues (20) with the notion that the erythrocytes are of little if any significance in the interorgan transfer of amino acids (14). Consequently, most experimental measures of circulating amino acids have been restricted to the quantification of plasma concentrations (2,4,5,13,21).

Relatively few analytical data are available on the actual amino acid content of the erythrocyte, although the red blood cell is well established as a cellular model of amino acid “transporter” systems that could exist in a cell (28). Some researchers have proposed that erythrocytes are significantly involved in the transport of amino acids. In vivo studies have shown that amino acid uptake or delivery by/to the erythrocytes is faster in vivo than expected from in vitro, indicating that the erythrocyte may have a specific role in the interorgan transport of amino acids in dogs (10,11), sheep (16), and humans (3,7,14).

Differences in plasma arterial-venous concentrations are measured to establish fluxes in amino acids, yet the role and significance of the erythrocyte in these processes has not been seriously considered. The function of the erythrocyte as an amino acid transporter may be accentuated under exercise conditions as the concentration of the free amino acid pool becomes increased, particularly under conditions of glycogen depletion (6). Carbohydrates reduce but do not eliminate the exercise induced elevation in amino acid catabolism (8,18,19). Thus, the ingestion of carbohydrate during exercise is expected to decrease the fuel required from amino acids, including an attenuation of the glucose-alanine cycle, which is normally maintained during exercise (13). Carbohydrate ingestion also influences any insulin dependent erythrocyte amino acid fluxes as observed by Aoki et al. (3).

Because no studies have so far been undertaken concerning erythrocyte amino acids and exercise, nor the influence of carbohydrate ingestion on erythrocyte amino acids, this investigation was undertaken to provide the followinginformation. First, to provide data on the erythrocyte amino acids concentrations during exercise. Second, to determine whether the changes in plasma and erythrocyte amino acids during exercise and recovery follow a similar pattern, and finally, to assess the influence of ingesting carbohydrate upon both plasma and erythrocyte concentrations of amino acids during exercise and into recovery.



Eight healthy, trained males [mean age 23.3 ± 3.6 (SD) yr, body mass 70.2 ± 6.0 kg, height 175.6 ± 6.0 cm and peak oxygen uptake (V̇O2 peak) 4.64 ± 0.81 L min1] from the Liverpool John Moores University volunteered to be subjects. Before the first exercise trial each subject gave their written informed consent in accordance with the procedures approved by the Ethics Committees of the Liverpool John Moores University and the South Manchester Area Health Authority.


Subjects attended the laboratory on three different occasions separated by at least 1 wk. The circadian variation in circulating amino acid concentrations was controlled for by conducting the trials at the same time of day for each subject. All participants were instructed to refrain from strenuous activity for the preceding 24 h and to be in a postabsorptive state after a 12-h overnight fast.

Preliminary Test

The peak oxygen uptake (V̇O2 peak) of each individual was determined using a continuous incremental loading test to volitional exhaustion on an electrically braked cycle ergometer (Bosch erg 551). After a standard familiarization and warm-up procedure, each subject exercised for 2 min at 120 W, and thereafter the resistance was increased by 30 W at 2-min intervals until volitional exhaustion. Respiratory analysis was monitored throughout the test. Expired air was collected continuously by an automated gas analysis system (P. K. Morgan Ltd., Rainham, Kent). This incorporated a Fleisch no. 3 pneumotachograph for quantifying volume changes, an infra-red CO2 analyzer, and a paramagnetic O2 analyzer, interfaced with a microcomputer. The analysers were calibrated to a known volume of air, known concentrations of O2 and CO2, and to the atmospheric pressure before each exercise session. Values for O2 consumption (V̇O2) were obtained at 1-min intervals; peak oxygen uptake (V̇O2 peak) was determined as the highest V̇O2 value attained. The work rate corresponding to 65% V̇O2 peak was calculated from the test.

Experimental Protocol

Subjects were randomly assigned to the two 90-min exercise treatments using a counter-balanced design in a single-blind format. At predetermined time intervals, orange-flavored water (trial Pl) or an orange-flavored 10% maltodextrin solution (trial Md) was ingested. Trials Md and Pl consisted of 90-min exercise on a cycle ergometer at an intensity corresponding to 65% V̇O2 peak and a 25-min postexercise rest period. Immediately before commencing a trial and at 20, 40, 60, and 80 min of exercise, 150 mL of the appropriate solution was ingested. Expired air was collected from 0 to 15 min of each trial to check that the calculated O2 consumption corresponded to the estimated work intensity.

Upon arrival at the laboratory, an 18-gauge indwelling intravenous cannula (Venflon 2), flushed with saline solution (0.9% w/v), was cited to a forearm vein of the subjects right arm using a 2% lignocaine local anesthesia. After 20-min rest and at 25, 55, and 85 min of exercise and 25 min postexercise, 10-mL venous blood samples were taken for the determination of plasma glucose, plasma insulin, and plasma and erythrocyte amino acid concentrations. The hemoglobin (Hb) concentration and packed cell volume (PCV) were also measured. Blood was withdrawn directly into a blood collection tube (Starstedt monovette) containing lithium-heparin to prevent coagulation. In accordance with amino acid assay procedures developed by the Department of Clinical Chemistry of the Royal Liverpool University Hospital, 2 mL of whole blood was removed from the collection tube and centrifuged for 15 min at 4000 rev·min1. One mL of plasma was pipetted into a clean glass test tube, and 1 mL of distilled water was then added to the red cells, resulting in 2 mL of hemolyzed cells, which were transferred to a clean test tube. The remaining whole blood in the collection tube (8 mL) was centrifuged for 20 min at 2000 rev·min1, the plasma removed and stored at −20°C until assayed.

Blood Analyses

Hemoglobin and hematocrit.

The Hb concentration was determined using a β-Hb photometer (Hemocue, Sheffield, U.K.). Duplicate samples were taken and the mean value used in subsequent calculations. PCV was measured by the standard microcapillary centrifuge method; the mean value of four samples was used in subsequent calculations. Plasma volume (PV) changes were calculated using Hb and PCV measurements according to the methods proposed by Dill and Costill (9).

Plasma glucose.

The plasma concentration was determined using a multi-purpose single channel analyzer (Analox GM7, London, UK). The system’s oxygen electrode detected the O2 change in solution after a reaction of glucose with glucose oxidase. All plasma samples were measured in triplicate.

Plasma insulin.

A double antibody radioimmunoassay technique was used to determine plasma insulin concentrations using an insulin assay kit (RIA 100, Kabi Pharmacia Diagnostics, Lund, Sweden). All plasma insulin concentrations were measured in duplicate. The assay has a detection limit of <2 mU·L1. The intra-assay precision is <10% on all parts of the curve, and the inter-assay precision is 12% on values between 3 and 7 mU·L1 and 6% on all other parts of the standard curve.

Plasma and erythrocyte amino acids.

Plasma and hemolysate amino acids were measured as ortho phthalaldehyde (OPA) derivatives as detected by a fluorometer (Model 410 Jasco, Essex, UK) after separation on high performance liquid chromatography (HPLC) as incorporated into an automated procedure (Asted, Anachem, UK). The erythrocyte amino acid concentrations were calculated from the red cell hemolysate solutions using the following formula: MATH

CC = erythrocyte concentration of amino acid; HC = hemolyzed cell concentration (= erythrocyte + residual plasma + water); PC = plasma concentration; Hct = hematocrit.

The above calculation is based on the fact that 2 mL of whole blood was initially spun down and 1 mL of plasma removed for assay (PC). One mL of water was then added to the erythrocytes (which included some residual plasma) to re-establish the original 2 mL (HC). Because HC has been effectively diluted by 50%, there is a need to multiply HC by 2. The plasma amino acid concentration (PC) is then subtracted from HC, but taking into account the residual volume of plasma left after the initial removal of 1 mL of plasma (hence the need to take account of Hct).


Plasma amino acid concentrations were corrected for changes in plasma volume. The data for plasma glucose, insulin, and erythrocyte and corrected plasma amino acids were analyzed using the same methods. Resting values were subjected to one-sample t-tests to determine whether any significant differences existed between initial values. All values recorded at rest, during exercise, and after recovery were analyzed using a repeated measures analysis of variance (ANOVA). Two-way ANOVAs were performed on the raw data (absolute values) and on the ratio of exercise and recovery data using resting values as the divisor, the first factor comparing treatments (Pl vs Md) and the second factor being time (rest, 25, 55, and 85 min during exercise, and 25 min postexercise). The reason for using ratio data accrued after the discovery that the erythrocyte values at rest were significantly different between the treatments, and hence a need to examine changes in relation to baseline (12). A three-way ANOVA was undertaken specifically to compare changes in the plasma with the erythrocyte amino acid concentrations during the treatments, i.e., were the changes in plasma amino acids with respect to time significantly different to those found in the erythrocytes? This was achieved by converting all the raw data into ratios using the corresponding resting values as baseline. Significance was accepted when P < 0.05. Results were expressed as means ± SD, and all statistical analysis was performed using Minitab.


Plasma Glucose

Plasma glucose levels declined gradually during exercise with the Pl ingestion whereas Md elevated the plasma glucose level above the resting values during exercise. ANOVA revealed this difference to be significant (F = 5.94;P < 0.01). Table 1 highlights the mean glucose concentrations during both treatments. When compared with resting values, the postexercise glucose concentrations increased by 1.58 mmol·L1 and decreased by 0.40 mmol·L1 in Md and Pl, respectively. The increase in Md was significant (P < 0.001).

Table 1
Table 1:
Plasma glucose and insulin concentrations (±SD) at rest, during exercise (25–85 min), and 25 min post-exercise (Post-Ex) in the maltodextrin (Md) and placebo (Pl) trials.

Plasma Insulin

The plasma insulin concentrations decreased during exercise in Pl and increased in Md up to 55 min before decreasing slightly. Overall, insulin levels in Md were significantly greater than Pl during exercise (F = 3.68;P < 0.05). Insulin increased postexercise to a greater extent in Md than Pl (Table 1), the Δ values from rest to postexercise being significantly different (P < 0.001).

Plasma and Erythrocyte Amino Acids

The plasma and erythrocyte amino acid concentrations observed throughout the treatments are displayed in Tables 2 and 3, respectively. Data for asparagine and glutamine were not obtained and neither were there complete data sets for histidine, tyrosine, and arginine.

Table 2
Table 2:
Mean absolute plasma amino acid concentrations (±SD) at rest, during exercise (25–85 min), and 25 min post-exercise in placebo and maltodextrin trials.
Table 3
Table 3:
Mean absolute erythrocyte amino acid concentrations (±SD) at rest, during exercise (25–85 min), and 25 min post-exercise in placebo and maltodextrin trials.

The erythrocyte amino acid concentrations at rest were significantly higher than those found in plasma with the exceptions of arginine and tryptophan. Resting levels of erythrocyte amino acid varied significantly between treatments for glutamate, histidine, glycine, threonine, alanine, valine, methionine, isoleucine, and lysine (P < 0.05), whereas the difference in plasma amino acid concentrations between treatments at rest was not significant (P > 0.05).

Plasma amino acid concentrations changed significantly with time in both treatments for the amino acids alanine, tyrosine, valine, isoleucine, and leucine (P < 0.05). The trend was for an increase in concentration during exercise and then a decline to baseline in recovery (Table 2). The two-way ANOVAs further revealed significant differences between treatments for plasma tryptophan only (P < 0.05).

The erythrocyte amino acid concentrations showed a significant increase with time for all amino acids in both treatments (P < 0.05). Significant differences between treatments were found for all erythrocyte amino acids except aspartic acid, alanine, methionine, and isoleucine (P > 0.05). These changes over time and treatments can be seen in Table 3.

Because significant differences had already been observed between the resting values of nine erythrocyte amino acids and so could have resulted in the significant differences noted with the ANOVA for absolute measurements of the amino acids, it was decided to calculate ratios using baseline measures and reevaluate the time and treatment effects. These analyses identified significant time effects for plasma alanine and tyrosine (P < 0.05) but were not found for plasma valine, isoleucine, and leucine (P > 0.05), which were reported above (see Table 2). The revised analyses on ratio data also identified treatment main effects for plasma glycine and threonine (P < 0.05), whereas the previous analysis on absolute values showed a difference with tryptophan only.

Significant changes with time were found for all erythrocyte concentrations expressed as ratios (P < 0.05) except tyrosine (P > 0.05), whereas the results for the analyses of the absolute data (mentioned above) showed a significant time effect for all the amino acids. No significant differences were apparent between treatments for erythrocytes when expressed as ratios (P > 0.05), which is in contrast to the analyses of the absolute data where significant differences between treatments were found for nine of the amino acids (see Table 3).

One of the aims of this investigation was to examine whether the changes over time in concentration of the erythrocyte and plasma amino acids follow a similar pattern. This was achieved using a three-way ANOVA in which the changes with time were expressed as a ratio of the baseline (resting) concentrations. A significant 2-way interaction (plasma-erythrocyte × time) was observed for all amino acids (P < 0.05) except for aspartic acid, glycine, and ornithine (P > 0.05).


The plasma and erythrocyte showed a marked difference in post absorptive concentrations of amino acids taken at rest except for tryptophan. The relatively higher concentrations observed in the erythrocyte have been shown previously for dogs (10,11), sheep (16), and humans (1,15). The resting erythrocyte concentrations obtained on separate occasions were significantly different for nine of the amino acids. This anomaly was particularly highlighted because plasma concentrations were consistent at each trial. The difference may be due to minor variations in the protein content of the diets before subjects began the 12-h overnight fast. Erythrocyte amino acids have been reported to be elevated in dogs as a result of high protein intake before fasting (10).

Resting plasma amino acid concentrations were consistent and within the range of previously reported values (1,2,4,5,13,17,21,22,26). During exercise, plasma concentrations remained relatively constant for all amino acids except alanine and tyrosine for the ratio-calculated data, although changes in the branched chain amino acids valine, isoleucine, and leucine were also observed with the raw data. Alanine and tyrosine levels increased progressively during exercise irrespective of the treatment, whereas the branched chain amino acids decreased during exercise. The recovery period resulted in a restoration of preexercise levels. The changes in plasma alanine and branched chain amino acids support previous findings (25). The present study design did not allow us to establish whether the changes in alanine and branched chain amino acid concentrations were a result of greater release or reduced uptake from exercising muscle as a-v differences were not determined. Circulating plasma alanine has been found to increase during exercise (13,21) and is associated with the production and release of glucose from the liver (2). The postexercise decrease in plasma alanine concentrations is probably due to a combination of augmented hepatic uptake and reduced peripheral alanine production, because less pyruvate becomes available for transamination once muscle activity is reduced. Leucine, isoleucine, and valine concentrations increased postexercise, possibly due to a reduced uptake by muscle because blood flow to these regions would be diminished at this time.

The ingestion of maltodextrin significantly increased plasma glucose and insulin concentrations but did not influence the plasma amino acid concentrations.

The amino acid concentrations determined in erythrocytes at rest in our study were in the same range to those reported recently by Agli et al. (1) for Glu, Thr, Arg, Tyr, Ileu, and Orn, although large variations were apparent for the other amino acids. Two other investigations reporting baseline values of in erythrocytes have also exhibited variations in concentration of some of the amino acids (3,15). It is unlikely that these differences are due to interindividual variations but more likely to be the result of the process of separation, whether the cells had been washed with saline, the method of calculating erythrocyte concentration from whole blood, and the technique for identification of the amino acids. Until consistency of methodology is adopted, it would seem prudent to express changes in erythrocyte amino acids from rest/baseline values as percentages, Δ-values, or (as in our case) ratios.

During exercise, it was observed that erythrocyte amino acid concentrations increased significantly with no corresponding increase in plasma levels except for alanine and tyrosine. This provides further evidence that the measure of plasma levels alone are not a true representation of amino acid transport during exercise. The exclusive determination of plasma amino acids underestimates circulating levels because the packed cell volume contains significant levels of amino acids in the erythrocytes (16,17,22). We recommended that investigations in circulating amino acids assess the content of both the erythrocyte and plasma compartments. Furthermore, we suggest that the erythrocyte acts as a store of amino acids until such time it is required by metabolising tissues. Far from being relatively inert and slow in the uptake of amino acids (27), the erythrocyte would appear to sequester certain amino acids at an appreciable rate during exercise.

Ingestion of maltodextrin failed to produce significant changes in erythrocyte amino acid levels compared with placebo, a similar finding (with the exceptions of glycine and threonine) being noted for plasma. These findings would suggest that despite a significant increase in circulating glucose and insulin subsequent to Md ingestion, there was no stimulation of amino acid uptake by erythrocytes. Four passive transport systems and three active transport systems have been proposed for the uptake of amino acids by erythrocytes (23). It appears that none of these systems are influenced by increases in insulin or glucose.

In a recent study examining the effects of carbohydrate supplementation on plasma glutamine, glutamate, alanine, and the branched chain amino acids during exercise and into recovery, the authors concluded that no significant difference accrued from supplementation (24). Our results support these findings.

We can conclude that amino acid concentrations at rest are generally significantly higher in erythrocytes than in plasma and that during exercise there is an increase in erythrocyte levels that are not matched by similar increases in plasma concentrations. We therefore urge colleagues to assess erythrocyte levels of amino acids in addition to plasma levels when examining changes as a consequence of exercise or dietary interventions. Finally, we may conclude that carbohydrate ingestion leading to increases in plasma glucose and insulin does not affect plasma or erythrocyte concentrations of amino acids during exercise or short-term recovery.


1. Agli, A-N., A. Schaefer, B. Geny, F. Piquard, and P. Haberey. Erythrocytes participate significantly in blood transport of amino acids during the post absorptive state in normal humans. Eur. J. Appl. Physiol. 78:502–508, 1998.
2. Ahlborg, G., P. Felig, L. Hagenfeldt, R. Hendler, and J. Wahren. Substrate turnover during prolonged exercise in man: splanchnic and leg metabolism of glucose, free fatty acids and amino acids. J. Clin. Invest. 53:1080–1090, 1974.
3. Aoki, T. T., W. A. Muller, M. F. Brennan, and G. F. Cahill. Blood cell and plasma amino acid levels across forearm muscle during a protein meal. Diabetes 22:768–775, 1973.
4. Bergstrom, J., P. Furst, and E. Hultman. Free amino acids in muscle tissue and plasma during exercise in man. Clin. Physiol. 5:155–160, 1985.
5. Brodan, V., E. Kuhn, J. Pechar, and D. Tomkova. Changes of free amino acids in plasma of healthy subjects induced by physical exercise. Eur. J. Appl. Physiol. 35:69–77, 1976.
6. Brooks, G. A. Amino acid and protein metabolism during exercise and recovery. Med. Sci. Sports Exerc. 19:S150–S156, 1987.
7. Darmaun, D., P. Froguel, M. Rongier, and J. Robert. Amino acid exchange between plasma and erythrocytes in vivo in humans. J. Appl. Physiol. 67:2383–2388, 1989.
8. Davies, C. T. M., D. Halliday, D. J. Millward, M. J. Rennie, and J. R. Sutton. Glucose inhibits CO2 production from leucine during whole-body exercise in man. J. Physiol. 332:40–41, 1982.
9. Dill, D. B., and D. L. Costill. Calculation of percentage changes in volumes of blood, plasma and red cells in dehydration. J. Appl. Physiol. 37:247–248, 1974.
10. Elwyn, D. H. Distribution of amino acids between plasma and red blood cells in the dog. Fed. Proc. 25:854–861, 1966.
11. Elwyn, D. H., W. J. Launder, H. C. Parikh, and E. M. Wise. Roles of plasma and erythrocytes in the interorgan transport of amino acids in dogs. Am. J. Physiol. 222:1333–1342, 1972.
12. Everitt, B. S. The analysis of repeated measures: a practical review with examples. Statistician 44:113–135, 1995.
13. Felig, P., and J. Wahren. Amino acid metabolism in exercising man. J. Clin. Invest. 50:2703–2714, 1971.
14. Felig, P., J. Wahren, and L. Raf. Evidence of inter-organ amino acid transport by blood cells in humans. Proc. Natl. Acad. Sci. USA 70:1775–1779, 1973.
15. Hagenfeldt, L., and A. Arvidsson. The distribution of amino acids between plasma and erythrocytes. Clin. Chim. Acta 100:133–141, 1980.
16. Heitman, R. N., and E. N. Bergman. Transport of amino acids in whole blood and plasma of sheep. Am. J. Physiol. (Endocrinol. Metab. 2) 239:E242–E247, 1980.
17. Iob, V., M. McMath, and W. W. Coon. Intra-individual and inter-individual variations in plasma free amino acids in normal adults. J. Surg. Res. 3:85–89, 1963.
18. Lemon, P. W. R. Protein and exercise: update 1987. Med. Sci. Sports Exerc. 19:S179–S190, 1987.
19. Lemon, P. W. R., and J. P. Mullin. Effect of initial muscle glycogen levels on protein catabolism during exercise. J. Appl. Physiol. 48:624–629, 1980.
20. Munro, H. N. General aspects of the regulation of protein metabolism by diet and by hormones. In: Mammalian Protein Metabolism, H. N. Munro (Ed.). New York: Academic Press, 1970, pp. 299–375.
21. Poortmans, J. R., G. Siest, M. M. Galteau, and O. Houot. Distribution of plasma amino acids in humans during submaximal prolonged exercise. Eur. J. Appl. Physiol. 32:143–147, 1974.
22. Scriver, C. R., D. M. Gregory, D. Sovetts, and G. Tissenbaum. Normal plasma free amino acid values in adults: the influence of some common physiological variables. Metabolism 34:868–873, 1985.
23. Tunnicliff, G. Amino acid transport by human erythrocyte membranes. Comp. Biochem. Physiol. 108A:471–478, 1994.
24. van Hall, G., W. H. M. Saris, and A. J. M. Wagenmakers. Effect of carbohydrate supplementation on plasma glutamine during prolonged exercise and recovery. Int. J. Sports Med. 19:82–86, 1998.
25. Viru, A. Mobilisation of structural proteins during exercise. Sports Med. 4:95–128, 1987.
26. Wahren, J., P. Felig, R. Hendler, and G. Ahlborg. Glucose and amino acid metabolism during recovery after exercise. J. Appl. Physiol. 34:838–845, 1973.
27. Winter, C. G., and H. N. Christensen. Migration of amino acids across the membrane of the human erythrocyte. J. Biol. Chem. 239:872–878, 1964.
28. Young, J. D., and J. C. Ellory. Red cell amino acid transport. In: Red Cell Amino Acid Transport, J. C. Ellory and V. L. Lew (Eds.). New York: Academic Press, 1977, pp. 301–325.


© 2000 Lippincott Williams & Wilkins, Inc.