Ascorbic Acid Bioavailability in Humans Ascorbic Acid in Plasma, Serum, and Urine YAO-HUI WANG, KULDEEP R. DHARIWAL, AND MARK LEVINE" Laboratory of Cell Biology and Genetics National Institute of Diabetes and Digestive and Kidney Diseases National Institutes of Health Bethesda, Maryland 20892
INTRODUCTION The optimal amount of ascorbic acid for human health is unknown.' To solve this problem we developed the concept of in sifu kinetics.',? Its biochemical arm is determination in sit14 of reaction kinetics for different vitamin C biochemical functions in relation to in sit14 vitamin C concentration^.^,^ The clinical arm is to learn how ascorbic acid in humans is made available to tissues, so that the specific biochemical reactions can then occur.' The clinical component requires a new understanding of how vitamin C ingestion regulates its plasma concentration and subsequent urinary excretion. To perform the clinical experiments several prerequisite experiments were necessary. We therefore investigated these issues: the presence of ascorbic acid and its oxidation product, dehydroascorbic acid, in plasma and urine; stability of ascorbic acid in blood, plasma, and urine; and protein binding of ascorbic acid in plasma.
RESULTS Ascorbic acid in plasma and serum was measured in 5 normal women volunteers, as shown in TABLE1 The upper values in the first column are ascorbic acid in plasma; the lower values are from serum of the same volunteers. The data indicate that ascorbic acid in plasma and serum are identical and that either preparation can be used. The second column is the sum of dehydroascorbic acid and ascorbic acid. If dehydroascorbic acid were present, the values in the second column should be higher than the values for ascorbic acid alone in the first column. As shown, the values in the first and second columns are identical for each volunteer, and in plasma and serum. These data indicate that dehydroascorbic acid is not present in fresh plasma and serum. To investigate protein binding, plasma and serum were subjected to rapid centrifugal ultrafiltration. If ascorbic acid were protein-bound, the concentration
.'
a Address correspondence to Mark Levine, M.D., Building 8, Room 415, National institutes of Health, Bethesda, MD 20892. 383
ANNALS NEW YORK ACADEMY OF SCIENCES
384
should differ in the retentate and the filtrate. As shown, the values are the same (TABLEI , columns 3 and 4). These data indicate that ascorbic acid is free in plasma and serum. Ascorbic acid stability in whole blood and plasma was studied.5 Ascorbic acid was stable in whole blood in the dark at 4°C for 24 hours. Ascorbic acid in proteinprecipitated plasma was stable at - 70°C for at least one month and was also stable for 2 hours in an autosampler at 4°C. Without cooling, however, stability was lost: after 30 minutes, dehydroascorbic acid formed in protein-precipitated plasma. Plasma prior to protein precipitation was stable at 4°C for 5 hours, but not longer. Ascorbic acid stability was determined in unprocessed urine samples at 4"C, in unprocessed samples at room temperature, and in samples diluted with methanol/ EDTA ;it 4°C (FIG.IA). Ascorbic acid in urine samples was stable in unprocessed
Ascorbic Acid, Dehydroascorbic Acid, and Ascorbic Acid Protein Binding in Plasma and Serum of Normal Women"
TABLE I .
Ascorbic Acid
Ascorbic Acid and Dehydroascorbic Acid
(ILM)
Subject Plasma
I 2 3 4 5 Serum 1 2 3 4 5
Ascorbic Acid Retentate Filtrate
70.8 2 85.0 ? 68.2 2 67.7 ? 68.5 t_
0.6 0.3 0.2 0.3 0.7
72.3 C 1.2 88.0 2 1.0 67.3 _t 0.7 68.2 2 1.2 67.4 t 0.7
69.1 C 1.0 83.6 t 0.0 66.2 C 0.2 67.3 t 0.9 68.5 t 2.2
68.6 2 0.0 83.0 t 0.7 65.0 C 0.3 66.1 t 0.1 68.0t 0.8
66.1 ? 86.0 t 64.6 ? 71.2 5 68.5 2
0.1
64.3 2 86.3 C 64.0 ? 71.5 ? 67.9 t
62.8 t 0.0 85.2 5 0.9 65.3 5 1.1 71.3 t 0.2 70.6 ? 0.2
64.0 2 85.0 ? 60.7 k 71.4 5 67.2 2
1.1 0.5 0.2 1.1
1.0 1.8 0.0 1.1
1.5
0.2 0.2 0.6 0.4 0.4
' I Plasnia and serum were prepared from fresh blood from 5 normal women.s Ascorbic acid was determined by using HPLC with coulometric electrochemical detection." Dehydroascorbic acid was first reduced to ascorbic acid by 2,3-dimercapto-l-propanol,and the total ascorbic .icid was then measured.' Dehydroascorbic acid content was the difference between the two measurements. For protein binding, plasma and serum samples were subjected to centrifugal ultrafiltration at 6000 x g for 30 minutes at 4°C using a membrane with a molecular weight cutoff of 3000.' Ascorbic acid was measured in the filtrate and the retentate.
and processed samples for at least 24 hours as long as the samples were at 4°C. Methanol-treated samples were also stable in an autosampler at 4°C for at least 12 hours (FIG.IB), indicating that these samples are suitable for unattended analysis. Methanol-treated urine samples were stable at - 70°C for at least one week. Dehydroascorbic acid was not detectable in urine samples. Ascorbic acid content of these samples ranged from 54.7-2015 p M .
CONCLUSIONS In plasma and serum, ascorbic acid is present only in the reduced form as ascorbic acid and not in the oxidized form as dehydroascorbic acid. Ascorbic acid
WANG et al.: ASCORBIC ACID
385
25 +-
c = 200 0 c
\
E,
v
2
8
.-V f0 0
B
100
1501
50t 04 0
4
8
12
16
20
28
24
32
Time (hour)
FIGURE 1A. Urine was stored at 4°C ( 0 )or at room temperature (0)in an amber vial. At the times indicated, one volume of urine was mixed with two volumes of 90% methanol saturated with EDTA. Samples were iced for 10 minutes and then centrifuged at 12,000 x g for 5 minutes at 4°C. The supernatant was aliquoted in triplicate and injected. Methanolprecipitated urine was also stored at 4°C; at different times, samples were placed in an autosampler and immediately injected (A).
0
0
2
4
6
8
10
12
Time (hour) FIGURE 1B Urine samples were treated with methanol/EDTA as in FIG. 1A and placed in a refrigerated autosampler at 0-2°C. Injections in triplicate were made at the times indicated.
386
ANNALS NEW YORK ACADEMY OF SCIENCES
is free in the circulation and not protein-bound. Ascorbic acid, but not dehydroascorbic acid, is present in urine. Ascorbic acid is stable in urine at 4°C for at least 24 hours. Similarly, whole blood can be stored at 4°C for 24 hours without ascorbic acid loss. These studies provide a foundation for determining how ascorbic acid concentrations in plasma, serum, and urine change as a function of diet. Ascorbic acid pharmacokinetic studies in normal humans can now be undertaken. REFERENCES , 1986. N. Engl. J . Med. 314: 892-902. 1. L E V I N EM. P. WASHKO,Y-H. WANG,R. WELCH,P. BERGSTEN & M., K. R . DHARIWAL, 2. LEVINE, E. J . BUTLER.1991. Am. J . Clin. Nutr. 54: 1157S-1162S. K . R . , P. WASHKO, W. 0. HARTZELL & M. LEVINE.1989. J . Biol. Chem. 3. DHARIWAL. 264: 15404-15409. K., M. SHIRVAN & M. LEVINE.1991. J . Biol. Chem. 266: 5384-5387. 4. DHARIWAL, K., W. HARTZELL & M. LEVINE.1991. Am. J . Clin. Nutr. 54: 712-716. 5 . DHARIWAL, P.. W. 0. HARTZELL & M. LEVINE.1989. Analyt. Biochem. 181: 276-282. 6. WASHKO, 7. DHARIWAL K.. P. WASHKO& M. LEVINE.1990. Anal. Biochem. 89: 18-23.
INTRODUCTION The optimal amount of ascorbic acid for human health is unknown.' To solve this problem we developed the concept of in sifu kinetics.',? Its biochemical arm is determination in sit14 of reaction kinetics for different vitamin C biochemical functions in relation to in sit14 vitamin C concentration^.^,^ The clinical arm is to learn how ascorbic acid in humans is made available to tissues, so that the specific biochemical reactions can then occur.' The clinical component requires a new understanding of how vitamin C ingestion regulates its plasma concentration and subsequent urinary excretion. To perform the clinical experiments several prerequisite experiments were necessary. We therefore investigated these issues: the presence of ascorbic acid and its oxidation product, dehydroascorbic acid, in plasma and urine; stability of ascorbic acid in blood, plasma, and urine; and protein binding of ascorbic acid in plasma.
RESULTS Ascorbic acid in plasma and serum was measured in 5 normal women volunteers, as shown in TABLE1 The upper values in the first column are ascorbic acid in plasma; the lower values are from serum of the same volunteers. The data indicate that ascorbic acid in plasma and serum are identical and that either preparation can be used. The second column is the sum of dehydroascorbic acid and ascorbic acid. If dehydroascorbic acid were present, the values in the second column should be higher than the values for ascorbic acid alone in the first column. As shown, the values in the first and second columns are identical for each volunteer, and in plasma and serum. These data indicate that dehydroascorbic acid is not present in fresh plasma and serum. To investigate protein binding, plasma and serum were subjected to rapid centrifugal ultrafiltration. If ascorbic acid were protein-bound, the concentration
.'
a Address correspondence to Mark Levine, M.D., Building 8, Room 415, National institutes of Health, Bethesda, MD 20892. 383
ANNALS NEW YORK ACADEMY OF SCIENCES
384
should differ in the retentate and the filtrate. As shown, the values are the same (TABLEI , columns 3 and 4). These data indicate that ascorbic acid is free in plasma and serum. Ascorbic acid stability in whole blood and plasma was studied.5 Ascorbic acid was stable in whole blood in the dark at 4°C for 24 hours. Ascorbic acid in proteinprecipitated plasma was stable at - 70°C for at least one month and was also stable for 2 hours in an autosampler at 4°C. Without cooling, however, stability was lost: after 30 minutes, dehydroascorbic acid formed in protein-precipitated plasma. Plasma prior to protein precipitation was stable at 4°C for 5 hours, but not longer. Ascorbic acid stability was determined in unprocessed urine samples at 4"C, in unprocessed samples at room temperature, and in samples diluted with methanol/ EDTA ;it 4°C (FIG.IA). Ascorbic acid in urine samples was stable in unprocessed
Ascorbic Acid, Dehydroascorbic Acid, and Ascorbic Acid Protein Binding in Plasma and Serum of Normal Women"
TABLE I .
Ascorbic Acid
Ascorbic Acid and Dehydroascorbic Acid
(ILM)
Subject Plasma
I 2 3 4 5 Serum 1 2 3 4 5
Ascorbic Acid Retentate Filtrate
70.8 2 85.0 ? 68.2 2 67.7 ? 68.5 t_
0.6 0.3 0.2 0.3 0.7
72.3 C 1.2 88.0 2 1.0 67.3 _t 0.7 68.2 2 1.2 67.4 t 0.7
69.1 C 1.0 83.6 t 0.0 66.2 C 0.2 67.3 t 0.9 68.5 t 2.2
68.6 2 0.0 83.0 t 0.7 65.0 C 0.3 66.1 t 0.1 68.0t 0.8
66.1 ? 86.0 t 64.6 ? 71.2 5 68.5 2
0.1
64.3 2 86.3 C 64.0 ? 71.5 ? 67.9 t
62.8 t 0.0 85.2 5 0.9 65.3 5 1.1 71.3 t 0.2 70.6 ? 0.2
64.0 2 85.0 ? 60.7 k 71.4 5 67.2 2
1.1 0.5 0.2 1.1
1.0 1.8 0.0 1.1
1.5
0.2 0.2 0.6 0.4 0.4
' I Plasnia and serum were prepared from fresh blood from 5 normal women.s Ascorbic acid was determined by using HPLC with coulometric electrochemical detection." Dehydroascorbic acid was first reduced to ascorbic acid by 2,3-dimercapto-l-propanol,and the total ascorbic .icid was then measured.' Dehydroascorbic acid content was the difference between the two measurements. For protein binding, plasma and serum samples were subjected to centrifugal ultrafiltration at 6000 x g for 30 minutes at 4°C using a membrane with a molecular weight cutoff of 3000.' Ascorbic acid was measured in the filtrate and the retentate.
and processed samples for at least 24 hours as long as the samples were at 4°C. Methanol-treated samples were also stable in an autosampler at 4°C for at least 12 hours (FIG.IB), indicating that these samples are suitable for unattended analysis. Methanol-treated urine samples were stable at - 70°C for at least one week. Dehydroascorbic acid was not detectable in urine samples. Ascorbic acid content of these samples ranged from 54.7-2015 p M .
CONCLUSIONS In plasma and serum, ascorbic acid is present only in the reduced form as ascorbic acid and not in the oxidized form as dehydroascorbic acid. Ascorbic acid
WANG et al.: ASCORBIC ACID
385
25 +-
c = 200 0 c
\
E,
v
2
8
.-V f0 0
B
100
1501
50t 04 0
4
8
12
16
20
28
24
32
Time (hour)
FIGURE 1A. Urine was stored at 4°C ( 0 )or at room temperature (0)in an amber vial. At the times indicated, one volume of urine was mixed with two volumes of 90% methanol saturated with EDTA. Samples were iced for 10 minutes and then centrifuged at 12,000 x g for 5 minutes at 4°C. The supernatant was aliquoted in triplicate and injected. Methanolprecipitated urine was also stored at 4°C; at different times, samples were placed in an autosampler and immediately injected (A).
0
0
2
4
6
8
10
12
Time (hour) FIGURE 1B Urine samples were treated with methanol/EDTA as in FIG. 1A and placed in a refrigerated autosampler at 0-2°C. Injections in triplicate were made at the times indicated.
386
ANNALS NEW YORK ACADEMY OF SCIENCES
is free in the circulation and not protein-bound. Ascorbic acid, but not dehydroascorbic acid, is present in urine. Ascorbic acid is stable in urine at 4°C for at least 24 hours. Similarly, whole blood can be stored at 4°C for 24 hours without ascorbic acid loss. These studies provide a foundation for determining how ascorbic acid concentrations in plasma, serum, and urine change as a function of diet. Ascorbic acid pharmacokinetic studies in normal humans can now be undertaken. REFERENCES , 1986. N. Engl. J . Med. 314: 892-902. 1. L E V I N EM. P. WASHKO,Y-H. WANG,R. WELCH,P. BERGSTEN & M., K. R . DHARIWAL, 2. LEVINE, E. J . BUTLER.1991. Am. J . Clin. Nutr. 54: 1157S-1162S. K . R . , P. WASHKO, W. 0. HARTZELL & M. LEVINE.1989. J . Biol. Chem. 3. DHARIWAL. 264: 15404-15409. K., M. SHIRVAN & M. LEVINE.1991. J . Biol. Chem. 266: 5384-5387. 4. DHARIWAL, K., W. HARTZELL & M. LEVINE.1991. Am. J . Clin. Nutr. 54: 712-716. 5 . DHARIWAL, P.. W. 0. HARTZELL & M. LEVINE.1989. Analyt. Biochem. 181: 276-282. 6. WASHKO, 7. DHARIWAL K.. P. WASHKO& M. LEVINE.1990. Anal. Biochem. 89: 18-23.
