|
|
||||||||
Drug Monitoring and Toxicology |
1 Research Centre and Pediatric Department, Sainte-Justine Hospital, University of Montreal, Montreal, Quebec, Canada.
2 Department of Nutrition, University of Montreal, Montreal, Quebec, Canada.
3 Division of Neonatology, Childrens and Womens Health Centre of British Columbia, Vancouver, British Columbia, Canada.
aAddress correspondence to this author at: Research Centre, Hospital Ste-Justine, 3175 Chemin Côte Ste-Catherine, Montreal, Quebec, H3T 1C5 Canada. Fax 514-345-4801; e-mail jean-claude.lavoie{at}umontreal.ca.
| Abstract |
|---|
|
|
|---|
Methods: Mass spectrometry was used to document the formation of byproducts of ascorbic acid in solutions containing a MVP, vitamin C + riboflavin, and vitamin C + H2O2 + Fe2+. The involvement of ascorbic acid and H2O2 in the formation of organic peroxides was tested by measuring peroxide concentrations in solutions containing H2O2 with or without ascorbic acid and with or without Fe2+ before and after addition of catalase.
Results: The loss of ascorbic acid in photo-exposed MVPs was associated with the concomitant generation of byproducts different from DHA and DKG. Among them, one mass fingerprint was particularly observed with solutions of vitamin C + riboflavin exposed to ambient light as well as with the solution of vitamin C + H2O2 + Fe2+, suggesting a Fenton-like reaction. This fingerprint was associated with the formation of catalase-resistant peroxides.
Conclusion: Exposure of MVPs to light leads to the rapid loss of ascorbic acid and generation of specific byproducts that differ from DHA and DKG. The conversion of vitamin C into byproducts could be of biological importance in accounting for the decrease in ascorbic acid concentrations and the generation of organic peroxides in light-exposed MVPs.
| Introduction |
|---|
|
|
|---|
Conditions favoring the formation of such ascorbic acid derivatives may be encountered in solutions for total parenteral nutrition (TPN). Indeed, these solutions contain multivitamin preparations (MVPs), which when exposed to ambient light generate peroxides (10)(11)(12). Photo-excited riboflavin catalyzes the electron transfer between ascorbic acid and oxygen, with formation of H2O2 (12). We showed previously that 1520% of peroxides formed in MVPs are resistant to catalase (10).
A substantial proportion of ascorbic acid disappears from the TPN solution exposed to light (13). We hypothesize that this loss is accounted for by light-induced oxidation of ascorbic acid. In this case, concomitant increases in DHA and DKG might be expected. The aim of this study was to document by mass spectrometry (MS) the formation of byproducts of vitamin C found in MVPs and to conform whether they accounted for the reported decrease in ascorbic acid in TPN solutions exposed to light.
| Materials and Methods |
|---|
|
|
|---|
ms
Briefly, a 10-µL aliquot of each tested solution was directly injected into a VG Quattro II triple quadrupole tandem mass spectrometer (MS-MS; Micromass Canada Inc.) with electrospray ionization and operated in the negative mode. The MS source and quadrupole temperatures were set at 70 °C. Argon was the collision gas at a pressure of 0.25 Pa. The ions were monitored at a cone voltage of 25 V, using the scan mode in the range of m/z 50250.
peroxide determinations
Peroxides were measured by determining the activity of glutathione peroxidase (11)(14). Briefly, 100 µL of tested samples was added to 900 µL of a solution containing 50 mmol/L Tris (pH 7.4), 2 mmol/L EDTA, 5 mmol/L reduced glutathione, 5 U of glutathione reductase, and 5 mmol/L NADPH. The decrease in NADPH concentration was monitored at 340 nm for 2 min before and 10 min after the addition of 0.1 U of glutathione peroxidase. H2O2 was used to construct curves (0100 µmol/L). Results were expressed as mean (SD) for three different samples.
protocols
Protocols were designed to investigate the development of specific byproducts of vitamin C formed in a solution of MVPs. The objective of the first protocol was to document the effect of light, as a function of the duration of light exposure, on mass spectra of a solution of multivitamins. Because the most abundant byproduct formed was different from the classic products of oxidation of vitamin C (DHA and DKG), the second protocol was designed to find its origin. In the third protocol, the implication of components of MVP involved in the light-induced formation of peroxides was tested, and the fourth protocol tested the implication of a Fenton reaction leading to the formation of an organic peroxide.
Effect of light on the stability of vitamin C in MVPs (protocol 1).
MS analyses were performed to determine the function of time and light exposure in the following solutions: (a) three freshly prepared solutions containing 1% MVPs (1 mL of MVP solution in 100 mL of water) in water exposed to ambient light (corresponding to 32 foot-candles in the laboratory) for 0, 0.25, 1, 2, and 3 h before their injection into the MS; (b) three freshly prepared photo-protected solutions containing 1% MVPs in water injected into the MS at time 0 and 24 h after their preparation.
Origin of light-dependent byproducts detected in protocol 1 (protocol 2).
The following analyses were performed: (a) Parent ions of m/z 135 were determined. (b) Fragmentation spectra of molecules at m/z 191 (DKG); 207 (2,3-diketo-4,5,5,6-tetrahydroxylhexanoate, as proposed by Deutsch et al. (6)(7); and 209 (hydrated DKG) derived from 3 h of photo-exposure of solutions containing 1% MVPs were compared with the spectra from photo-exposed MVPs. (c) Fragmentation spectra of the ions at m/z 209 and 191 derived from the analysis of 1% photo-exposed MVPs were compared with the spectrum of DKG, synthesized and purified according to the method of Koshiishi et al. (9).
Verification of whether light-dependent byproducts found in protocol 1 were related to interaction between ascorbic acid and riboflavin (protocol 3).
MS analyses were performed in the following solutions after 3 h of incubation at room temperature for solutions of 2 mmol/L ascorbic acid, pH 7.4, with or without 15 µmol/L riboflavin [same concentrations as in 1% MVPs, leading to the generation of peroxides measured in MVPs (12)].
Investigation of whether reactions leading to the formation of the molecule with the fingerprint observed in MVPs were compatible with the previously observed generation of catalase-resistant peroxides (10).
The following reactions were performed: (a) MS analyses and peroxide determinations were carried out in the following solutions: 40 µmol/L H2O2 + 2 mmol/L ascorbic acid, with or without 15 µmol/L FeCl2, pH 7.4, incubated for 30 min at room temperature. (b) Organic peroxides were sought in the following mixture: 40 µmol/L H2O2 + 2 mmol/L ascorbic acid + 15 µmol/L FeCl2 + 50 or 100 U/L catalase. To determine H2O2 formation, catalase was added after a 30-min incubation, and peroxide concentrations were measured 15 min later. (c) To confirm that the enzymatic assay measured only peroxides, peroxides were determined in the following solution: 40 µmol/L H2O2 + 2 mmol/L ascorbic acid + 15 µmol/L FeCl2 + 200 or 400 µmol/L sodium bisulfite, which has antiperoxide properties (15). Sodium bisulfite was added after a 30-min incubation, and peroxide concentrations were determined 5 min later. We verified that there was no interaction between bisulfite and the enzymatic determination of peroxides.
| Results |
|---|
|
|
|---|
|
|
To further investigate the origin of the ion at m/z 135, the major ion of the byproducts formed during light exposure, we performed MS-MS analyses and found that its parent was the m/z 207 ion (Fig. 3A
). The MS fragmentation spectrum of the m/z 207 ion (Fig. 3B
) was compatible with the data, suggesting that the MVP-derived ion at m/z 135 was a fragment of m/z 207. Indeed, further fragmentation after trapping of m/z 207 generated the daughter ions m/z 59, 75, 89, 117, and 135, which were also observed in the time course experiments (Fig. 1
). Deutsch (16) has proposed that the ion at m/z 135 is threonate, which can be derived from DKG (m/z 191). Our fragmentation spectra of the m/z 191 and 209 ions (possibly hydrate forms of DKG) as well as the synthesized DKG were analyzed. In all of these analyses, the major daughter ions (Fig. 4
) were at m/z 57, 59, 87, and 99. The ions at m/z 209 (Fig. 4C
) and 191 (Fig. 4B
) produced very similar fragment patterns, strongly suggesting that the ion at m/z 209 is a hydrate form of DKG (m/z 191 + 18 = 209).
|
|
As shown in Fig. 5
, the light-exposed solution of ascorbic acid + riboflavin generated the same ions as photo-exposed MVPs (Fig. 1
). Because the reaction required the presence of both riboflavin and light, the influence of H2O2 was tested. The mass spectra obtained from a solution of ascorbic acid + H2O2 + Fe2+ was comparable to that observed with ascorbic acid + riboflavin + light (Fig. 5
). Only the ions at m/z 115 and 175 were observed in the absence of Fe2+ or riboflavin (data not shown).
|
Because the MVP solution is contaminated with H2O2 and organic peroxides, defined as catalase-resistant peroxides (10), peroxides were measured in the same experiments, yielding the data shown in Fig. 5A
. The data from Fig. 6s
how that the addition of H2O2 to ascorbic acid is not sufficient to induce endogenous generation of peroxides. However, the addition of Fe2+ led within 30 min to the formation of peroxides, which were catalase-resistant, suggesting a Fenton-like reaction.
|
| Discussion |
|---|
|
|
|---|
Deutsch (6) also reported that the mixture of ascorbic acid + H2O2 produced ions at m/z 89, 117, 135, and 207. He proposed that the ion at m/z 135 derived from DKG (m/z 191), but this interpretation differs from our data documenting that the m/z 135 ion derives from m/z 207 (Fig. 3
). MS-MS analyses of ions at m/z 191 and 207 were different. The discrepancies between both reports may be explained by differences in experimental protocols. Indeed, Deutsch used concentrations of ascorbic acid/DHA (50 vs 0.9 mmol/L) and H2O2 (3003000 vs 0.20.4 mmol/L) that were several orders of magnitude higher than in the present study.
Although Fig. 3
shows that the m/z 135 ion derives only from m/z 207, the present experiments were not designed to differentiate between the formation of threonate and the m/z 207-derived ion at m/z 135. Ion suppression phenomena (17) may also explain why threonate did not appear. However, the fact that the fingerprint obtained from the incubation of only ascorbic acid with riboflavin (+ light) or with hydrogen peroxides + ferrous ions (Fig. 5
) was similar to that observed with light-exposed MVPs reduced the probability that the matrix or others molecules from MVPs interfered with the detection of threonate.
The MS-MS ions observed in the fragmentation spectra of the molecule at m/z 207 (mostly m/z 59, 89, 117, and 135; Fig. 3
) were found in photo-exposed MVPs (Fig. 1
) but also in the solutions containing only ascorbic acid + riboflavin or ascorbic acid + H2O2 + Fe2+. The fact that, under our conditions, H2O2 and Fe2+ were essential to generate these species suggested that a Fenton-like reaction was involved. The hypothesis that the molecule at m/z 207 is an organic peroxide generated in photo-exposed MVPs is supported by the data in Fig. 6
, which shows that the combination of ascorbic acid + H2O2 + Fe2+ produced a catalase-resistant peroxide. Because the amount of the peroxides formed was greater than the initial amount of H2O2, this suggests a radical chain reaction that consumes vitamin C. Moreover, the ratio (37%) of organic peroxides [3060 µmol/L (10)] to ascorbic acid (0.9 mmol/L) in a solution containing 1% MVPs corresponds to the actual percentage [12 ( (3))%] of relative abundance of m/z 207 at 3 h based on ascorbic acid (m/z 175) at time zero (Figs. 1
and 2
). This suggests that the m/z 207 ion could be a peroxide form of ascorbic acid. Because of the potential clinical impact of ascorbic acid conditionally becoming an organic peroxide, this hypothesis must be confirmed. Further studies should be conducted to identify the chemical structure of these molecules. Deutsch et al. (6)(7) also reported a molecule at m/z 207, i.e., the 2,3-diketo-4,5,5,6-tetrahydroxyhexanoate, which was detected after incubation of DHA with H2O2. Because the very different experimental conditions, it can not be excluded that both molecules might exist.
Faced with problems linked to the immature antioxidant defenses of premature newborn infants, clinicians might be tempted to increase the infusion of multivitamin solutions because they contain antioxidant vitamins that have documented antiradical properties (18)(19). However, this process could increase the infusion of these byproducts of vitamin C. The present study confirms the findings of Silvers et al. (13) that the concentration of ascorbic acid decreases within minutes after exposure to light (Fig. 4
). Thus, infants may be infused with a large dose of peroxides (20) accompanied by a amount of the antioxidant ascorbic acid lower than that prescribed. The observed loss of ascorbic acid, without concomitant generation of a comparable amount of DHA (Fig. 4
), could also account for the absence of modification in plasma concentrations of vitamin C after multivitamins are administered with parenteral nutrition to premature infants (21). In these same infants, the vitamin C concentration doubled after initiation of oral feedings. It would be of interest to measure plasma concentrations of vitamin C after introducing photo-protected multivitamins with TPN.
The potential clinical impact of these byproducts of vitamin C is supported by concerns raised about the harmful effects of high vitamin C concentrations in cell culture (22) as well as in preterm infants (23). However, others have refuted these concerns by documenting that high plasma concentrations of ascorbic acid in the presence of an iron overload did not cause oxidative damage to lipids and proteins (24). These different findings could be related to different experimental conditions. In an animal model of TPN, the infusion of a multivitamin solution exposed to ambient light was associated with H2O2-dependent (25) and -independent(18) biological effects in the lungs and liver, respectively. Findings in liver suggest the infusion of toxic compounds generated under specific conditions such as those found in complex solutions of multivitamins. It remains to be verified whether the byproducts of vitamin C mediate these hepatotoxic effects.
| Acknowledgments |
|---|
| Footnotes |
|---|
1 Nonstandard abbreviations: DHA, dehydroascorbate; DKG, 2,3-diketogulonic acid; TPN, total parenteral nutrition; MVP, multivitamin preparation (Multi-12 Pediatric); and MS, mass spectrometry. ![]()
| References |
|---|
|
|
|---|
The following articles in journals at HighWire Press have cited this article:
![]() |
M. Khashu, A. Harrison, V. Lalari, J.-C. Lavoie, and P. Chessex Impact of shielding parenteral nutrition from light on routine monitoring of blood glucose and triglyceride levels in preterm neonates Arch. Dis. Child. Fetal Neonatal Ed., March 1, 2009; 94(2): F111 - F115. [Abstract] [Full Text] [PDF] |
||||
![]() |
L. Knafo, P. Chessex, T. Rouleau, and J.-C. Lavoie Association between Hydrogen Peroxide-Dependent Byproducts of Ascorbic Acid and Increased Hepatic Acetyl-CoA Carboxylase Activity Clin. Chem., August 1, 2005; 51(8): 1462 - 1471. [Abstract] [Full Text] [PDF] |
||||
![]() |
J.-C. Lavoie, T. Rouleau, and P. Chessex Interaction between Ascorbate and Light-Exposed Riboflavin Induces Lung Remodeling J. Pharmacol. Exp. Ther., November 1, 2004; 311(2): 634 - 639. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |