A method to measure the oxidizability of both the aqueous and lipid compartments of plasma

Free Radical Biology & Medicine, Vol. 31, No. 9, pp. 1043–1050, 2001 Copyright © 2001 Elsevier Science Inc. Printed in the USA. All rights reserved 0891-5849/01/$–see front matter

PII S0891-5849(01)00684-0

Original Contribution A METHOD TO MEASURE THE OXIDIZABILITY OF BOTH THE AQUEOUS AND LIPID COMPARTMENTS OF PLASMA GIANCARLO ALDINI,* KYUNG-JIN YEUM,† ROBERT M. RUSSELL,†

and

NORMAN I. KRINSKY†‡

†

*Istituto Chimico Farmaceutico Tossicologico, University of Milan, Milan, Italy; USDA Human Nutrition Research Center on Aging at Tufts University, Boston, MA, USA; and ‡Department of Biochemistry, Tufts University School of Medicine, Boston, MA, USA (Received 22 June 2001; Accepted 2 August 2001)

Abstract—The lipophilic radical initiator (MeO-AMVN) and the fluorescent probe C11BODIPY581/591 (BODIPY) were used to measure the lipid compartment oxidizability of human plasma. Aqueous plasma oxidizability was initiated by the aqueous peroxyl radical generator, AAPH, and 2⬘,7⬘-dichlorodihydrofluorescein (DCFH) was employed as the marker of the oxidative reaction. The distribution in aqueous and lipid compartments of the two radical initiators was determined by measuring the rate of consumption of the plasma hydrophilic and lipophilic endogenous antioxidants. In the presence of AAPH (20 mM), the order of consumption was: ascorbic acid ⬎ ␣-tocopherol ⬎ uric acid ⬎ ␤-carotene, indicating a gradient of peroxyl radicals from the aqueous to the lipid phase. When MeO-AMVN was used (2mM), ␤-carotene was consumed earlier than uric acid and almost at the same time as ␣-tocopherol, reflecting the diffusion and activation of MeO-AMVN in the lipophilic phase. The rate of BODIPY oxidation (increase in green fluorescence) significantly increased after the depletion of endogenous ␣-tocopherol and ␤-carotene, whereas it was delayed for 180 min when AAPH was used instead of MeO-AMVN. The measurement of lipid oxidation in plasma was validated by adding to plasma the two lipophilic antioxidants, ␣-tocopherol and ␤-carotene, whose inhibitory effects on BODIPY oxidation were dependent on the duration of the preincubation period and hence to their lipid diffusion. DCFH oxidation induced by AAPH only began after uric acid, the main hydrophilic plasma antioxidant, was consumed. In contrast, when MeO-AMVN was used, DCFH oxidation was delayed for 120 min, indicating its localization in the aqueous domain. In summary, the selective fluorescence method reported here is capable of distinguishing the lipophilic and hydrophilic components of the total antioxidant capacity of plasma. © 2001 Elsevier Science Inc. Keywords—Azo-initiator, Free radicals, Plasma oxidizability, Fluorescence, Antioxidants

INTRODUCTION

ous plasma compartment oxidizability, these methods are unable to determine the antioxidant capacity of the lipid compartment. It is not surprising that most of the methods used to measure the total antioxidant capacity of plasma such as TRAP and FRAP did not show any effects following a diet enriched with carotenoids [7,8]. Also the oxygen radical absorbance capacity [9], which only measures the antioxidant capacity in the aqueous compartment of plasma, failed to show any direct correlation between fat-soluble antioxidant nutrient levels and antioxidant capacity in circulation. These results can be explained by considering that the plasma carotenoids, being deeply embedded in the lipoprotein core, are not available for reacting with the aqueous radical species or with the ferric complexes used in the FRAP method. The selective measurement of lipid compartment oxidizability could be relevant not only to study the effect of

Several applications based on the oxidizability of biological fluids have been reported recently. In particular, plasma oxidizability has been used to study the total antioxidant capacity [1,2], the antioxidant activity of synthetic and natural compounds [3,4], and the reactivity of aqueous and lipophilic antioxidants [5]. However, most of the methods currently used to measure plasma oxidizability are based on the aqueous radical initiator AAPH and on hydrophilic markers of oxidizability [6]. As the oxidation process is induced primarily by hydrophilic radicals and measures the aqueAddress correspondence to: Dr. Norman I. Krinsky, Department of Biochemistry, Tufts University School of Medicine, 136 Harrison Avenue, Boston, MA 02111-1837, USA; Tel: (617) 636-6861; Fax: (617) 636-2409; E-Mail: [email protected]. 1043

1044

G. ALDINI et al.

a diet supplemented with lipophilic antioxidants, but also to investigate the lipoprotein oxidation in whole plasma [10] and the activity and mechanism of lipid-soluble antioxidants. Therefore, we focused our interest on developing a method capable of measuring and distinguishing the aqueous and lipid plasma compartment oxidizability. The aim of this paper was to develop a fluorometric method to selectively measure both aqueous and lipid compartment oxidizability in plasma. In particular, we used the azo compound 2,2⬘-azobis(4-methoxy-2,4-dimethylvaleronitrile) (MeO-AMVN), as a lipid soluble radical initiator, and 4,4-difluoro-5-(4-phenyl-1,3-butadienyl)-4-bora-3a,4a-diaza-s-indacene-3-undecanoic acid (C11-BODIPY581/591) as a lipophilic fluorescence probe to monitor the lipid compartment plasma oxidation. MeO-AMVN has been recently reported as a new lipophilic radical initiator for the oxidation of membranes and lipoproteins [11–13]. At the same time, the aqueous plasma oxidative reaction was monitored using AAPH as the aqueous peroxyl radical generator and 2⬘,7⬘-dichlorodihydrofluorescein (DCFH) as the marker of the oxidative reaction. MATERIALS AND METHODS

Chemicals The radical initiators AAPH, AMVN, and MeOAMVN were obtained from Wako Chemicals (Richmond, VA, USA). The fatty acid analogue C11-BODIPY 581/591 (BODIPY) and 2⬘,7⬘-dichlorodihydrofluorescein diacetate (DCFH-DA) were obtained from Molecular Probes (Eugene, OR, USA). Bovine serum albumin (BSA), ␣-tocopherol and ␤-carotene (type II) were purchased from Sigma (St. Louis, MO, USA). All the other reagents were of analytical grade. Human plasma oxidation induced by water- and lipid-soluble radical inducers After an overnight fast (10 –12 h), blood from two healthy donors (32 and 35 years old) was collected in ethylenediaminetetraacetic acid (EDTA)-containing tubes. In order to reduce the variability of different donors, blood samples from these two subjects were collected weekly for the duration of the experiment. Immediately after collection, the samples were placed on ice and protected from light. Plasma was obtained by centrifugation at 800 ⫻ g for 20 min at 4°C and immediately used for the in vitro studies. Aqueous and lipid plasma oxidation was induced at a constant rate by the two azo-initiators: (i) AAPH as a water-soluble peroxyl radical generating system, (ii)

AMVN and the analogue MeO-AMVN as lipid-soluble peroxyl radical initiators. In order to compare the consumption of endogenous antioxidants induced by AAPH and MeO-AMVN, the amount of free radicals generated was kept constant by adjusting the concentration of the two azo-initiators. In the presence of 10 –20 mM of AAPH, the flux of aqueous radicals calculated on the basis of the known rate of free radical generation from AAPH at 37°C (Ri ⫽ 1.36 ⫻ 10⫺6 [AAPH] mol/liter/s) [14] was, respectively, 1.36 and 2.72 ⫻ 10⫺8 mol/liter/s. Because the rate of peroxyl radical formation from MeO-AMVN is 14.2 ⫻ 10⫺6 mol/liter/s (calculated in micelles) [11], the concentration of the lipophilic azo-initiator was reduced 10-fold to 1–2 mM, to reach the same order of free radical flux. AAPH was prepared in phosphate-buffered saline (50 mM, pH 7.4, PBS) and stored at ⫺20°C, while AMVN and MeO-AMVN were prepared, respectively, in EtOH and CH3CN, immediately before use. In order to obtain homogeneous incorporation, the lipid-soluble initiators were added slowly to the samples with a microsyringe (10 ␮l) with stirring. The samples were then vortexed for 10 s and incubated at 37 ⫾ 1°C under aerobic conditions. Determination of hydrophilic and lipophilic plasma antioxidants Because the fluorescence probes did not affect the plasma concentration of antioxidant nutrients (data are not shown), the probes were not added in the incubation for the antioxidant nutrient analysis. Plasma:PBS (1:5, by vol) was incubated at 37°C up to 4 h in the presence and absence of the hydrophilic radical generator, AAPH (10 mM and 20 mM) or the hydrophobic radical generator, MeO-AMVN (1 mM and 2 mM). The fat-soluble antioxidant nutrients, such as ␤-carotene and ␣-tocopherol, were measured at 30 min, 1 h, 2 h, 3 h, and 4 h. ␤-Carotene and ␣-tocopherol in plasma were extracted and measured using the HPLC method described earlier [15]. A 100 ␮l aliquot of the reaction mixture was extracted for ␤-carotene and ␣-tocopherol analysis. Echinenone in ethanol was added as an internal standard. The mixture was extracted with CHCl3: CH3OH (2:1, by vol) containing 0.2% BHT and hexane containing 0.1% BHT, dried under nitrogen, redissolved in ethanol, and injected into an HPLC system with a C30 column (3 ␮m, 150 ⫻ 4.6 mm; YMC, Wilmington, NC, USA). A Waters 994 programmable photodiode array detector was set at 450 nm for carotenoids and 292 nm for ␣-tocopherol analyses. The major water-soluble antioxidants (ascorbic acid and uric acid) were measured at 5 min, 15 min, 30 min, 1 h, 2 h, 3 h, and 4 h. For water-soluble antioxidant measurement, the mixtures were immediately deprotein-

Measurement of plasma oxidizability

1045

ized with perchloric acid (250 mM). Ascorbic acid and uric acid in plasma was analyzed by HPLC using an electrochemical detector (Bioanalytical System, Inc., N. Lafayette, IN, USA) as described earlier [16]. Results are expressed as percentages with respect to control samples prepared without the azo-compounds.

Measurement of plasma oxidation Plasma oxidation was measured fluorometrically using two different fluorescent probes: DCFH and BODIPY. DCFH-DA and BODIPY stock solutions were prepared in EtOH and dimethylsulfoxide, respectively, stored under nitrogen at ⫺20°C and used within 2 months. DCFH was prepared from DCFH-DA by basic hydrolysis. Briefly 500 ␮l of DCFH-DA stock solution (1 mM) was mixed with 2 ml of NaOH (0.01 N at 4°C) for 20 min while protected from the light. The mixture was then neutralized with 2 ml of HCl (0.01 N), diluted with PBS to a final concentration of 10 ␮M and stored in ice for no longer than 8 h (working solution); an aliquot of 100 ␮l was added to 200 ␮l of plasma and then diluted to a final volume of 1 ml with PBS. Aqueous plasma oxidation was measured monitoring the 2-electron oxidation of DCFH to the highly fluorescent compound 2⬘,7⬘-dichlorofluorescein (DCF). The excitation wavelength (␭ex) was set at 502 nm (slit 5 nm) and emission (␭em) at 520 nm (slit 5 nm). For BODIPY incorporation into the lipid plasma compartment, 25 ␮l of the BODIPY stock solution (2 mM) were diluted 100-fold with PBS. Aliquots of 100 ␮l were then added to 200 ␮l of plasma and 100 ␮l of PBS, vortexed for 20 s and then incubated under aerobic conditions for 10 min at 37°C. The final volume was adjusted to 1 ml with PBS yielding BODIPY at a final concentration of 2 ␮M. Lipid plasma oxidation was determined by monitoring both the red fluorescence decay (␭ex ⫽ 580, ␭em ⫽ 600 nm) of BODIPY and the green fluorescence increase (␭ex ⫽ 500, ␭em ⫽ 520 nm) of the oxidation product. In the experiments using ␤-carotene, to avoid the filtering effect due to the carotenoid, the oxidation product of BODIPY was also detected at ␭ex ⫽ 520 and at ␭em ⫽ 540 nm. The fluorescence measurements were carried out using a Perkin Elmer spectrofluorometer (model 650-10s) with 1 cm path length fluorescence cuvettes. In order to evaluate the intra-assay precision of the method, six replicates of the same plasma sample were analyzed by a single individual while the inter-assay repeatability was carried out by four different individuals. The precision was evaluated as coefficient of variation (CV).

Fig. 1. Effect of AAPH and MeO-AMVN on hydrophilic antioxidant levels in human plasma. The initial concentrations of ascorbic acid (AA) and uric acid (UA) were, respectively, 48 ␮M and 220 ␮M. The azo-compounds were added to plasma samples (1:5 with PBS) and incubated at 37°C in the dark. At fixed times, aliquots were withdrawn and the concentration of AA and UA assayed by HPLC as described in the text. Legend to symbols: AAPH (20 mM): ■ ⫽ AA; 䊐 ⫽ UA; MeO-AMVN (2 mM): ● ⫽ AA; E ⫽ UA. Values are mean ⫾ SD of three independent experiments.

␣-Tocopherol and ␤-carotene plasma enrichment Plasma was supplemented with ␣-tocopherol and ␤-carotene according to Bowen and Omaye [17] with minor modifications. Briefly, ␤-carotene (dissolved in stabilized THF, 10 mg/ml) or ␣-tocopherol (4.3 mg/ml in EtOH) were added to plasma to reach a final concentration of 50 ␮M; the samples were then vortexed for 30 s and incubated at 37 ⫾ 1°C for 1 to 6 h under nitrogen. After the preincubation period, plasma samples were diluted 5-fold with PBS to give a final concentration of ␣-tocopherol and ␤-carotene of 10 ␮M; the final amount of the solvents was always less than 0.8% v/v. Controls were prepared in the same way using solvent only. RESULTS

Plasma antioxidant consumption induced by AAPH and MeO-AMVN The major hydrophilic (ascorbic acid and uric acid) and lipophilic (␣-tocopherol and ␤-carotene) plasma antioxidants were consumed in a time-dependent manner in the presence of AAPH or MeO-AMVN. As expected by the solubility of the radical inducers, the hydrophilic antioxidants were consumed more rapidly when AAPH was used, in contrast to MeO-AMVN. As shown in Fig. 1, ascorbic acid and uric acid were completely consumed

1046

G. ALDINI et al.

Fig. 3. Oxidation of DCFH to DCF induced by AAPH or MeO-AMVN. The reaction mixture consisted of DCFH (1 ␮M final concentration), the azo-compound and human plasma (1:5 with PBS). Samples were incubated at 37°C in the dark and at fixed times the DCF content measured by fluorescence (␭ex ⫽ 502 nm, ␭em ⫽ 520 nm). Legends to symbols: * ⫽ AAPH, 20 mM, no plasma addition; ⽧ ⫽ AAPH, 10 mM; ■ ⫽ AAPH, 20 mM; 䊐 ⫽ MeO-AMVN, 2 mM. Values are mean ⫾ SD of five independent experiments.

depleted after 60 min of incubation. The rate of consumption was significantly lower at 1 mM MeO-AMVN. In contrast to the consumption of ascorbic acid, uric acid, and ␣-tocopherol, the kinetics of ␤-carotene depletion was faster in the presence of 2 mM MeO-AMVN as compared to that of 10 –20 mM AAPH (Fig. 2B). Measurement of plasma aqueous compartment oxidation Fig. 2. Effect of AAPH and MeO-AMVN on ␣-tocopherol (a) and ␤-carotene (b) levels in human plasma (1:5 with PBS). The initial concentration of the lipophilic antioxidants was 25 ␮M (␣-tocopherol) and 3 ␮M (␤-carotene). Legends to symbols: ■ ⫽ AAPH 10 mM; 䊐 ⫽ AAPH 20 mM; ● ⫽ MeO-AMVN 1 mM; E ⫽ MeO-AMVN 2 mM . Values are mean ⫾ SD of three independent experiments.

within 15 min and 180 min, respectively, using 20 mM AAPH. The consumption of these antioxidants was significantly slower in the presence of 2 mM MeO-AMVN, because total disappearance of ascorbic acid and uric acid was observed after 30 min and 300 min, respectively. In the presence of 10 –20 mM AAPH, the lipophilic antioxidant ␣-tocopherol was almost completely consumed within 30 min (Fig. 2A), whereas there was little oxidation of ␤-carotene in this period (Fig. 2B). In the presence of 2 mM MeO-AMVN, the ␣-tocopherol content was reduced by 42% at 30 min, and almost totally

In the absence of plasma, 20 mM AAPH rapidly oxidized a solution of DCFH in PBS as shown in Fig. 3, where a rapid increase of fluorescence was observed, which increased linearly with time. In the presence of plasma, a lag time was observed whose length was dependent on the amount of AAPH added. The propagation phase started at 90 min with 20 mM AAPH and at 180 min with 10 mM AAPH, corresponding to the depletion of both ascorbic acid and uric acid (Fig. 1). MeO-AMVN (2 mM) induced the propagation phase only after 270 min of incubation. No significant DCF formation was observed in the absence of the radical initiators until 5 h of incubation (data not shown). Measurement of lipid compartment plasma oxidation As previously reported, BODIPY is a lipophilic fluorescence probe, suitable to monitor the oxidation pro-

Measurement of plasma oxidizability

Fig. 4. Time curves of red fluorescence (␭ex ⫽ 580 nm, ␭em ⫽ 600 nm) and green fluorescence (␭ex ⫽ 500 nm, ␭em ⫽ 520 nm) of BODIPY in human plasma (1:5 with PBS) in the presence of AMVN and MeO-AMVN. BODIPY red fluorescence: 䊐 ⫽ 2 mM AMVN; ■ ⫽ 2 mM MeO-AMVN; BODIPY green fluorescence: { ⫽ 2 mM AMVN; ⽧ ⫽ 2 mM MeO-AMVN. Values are mean ⫾ SD of five independent experiments.

cess in organic solvents and liposomes [18] as well as living cells [19]. When BODIPY was added to plasma, a linear dose-dependent red fluorescence increase was observed (r2 ⫽ 0.996), indicating the incorporation of the fatty acid analogue in the plasma lipid compartment (data not shown). Only a negligible fluorescence intensity (less than 5–10% with respect to plasma) was observed when BODIPY was added to PBS or a BSA solution (1 g/dl in PBS). Initially, we used AMVN, a typical generator of lipid peroxyl radicals, to induce the oxidative reaction in the lipid compartment. At 2 mM AMVN, we did not observe any change in the BODIPY fluorescence (Fig. 4), due to the low efficiency of free radical generation by AMVN in a viscous lipophilic compartment at 37°C. When the concentration of AMVN was increased to 4 mM, we observed the formation of a cloudy precipitate that interfered with the fluorometric analysis. To overcome this problem, we used MeOAMVN, a new lipophilic radical initiator characterized by a higher efficiency of free radical generation with respect to AMVN (the rate constant is about 15 times larger under the same conditions) [11]. When plasma containing BODIPY was incubated in the presence of 2 mM MeO-AMVN, a linear and timedependent decrease of red fluorescence was observed, accompanied by an increase of green fluorescence (Fig. 4). As previously reported [19], this effect is due to the oxidation of the diene bond with a consequent loss of

1047

Fig. 5. Time-course of BODIPY green fluorescence in human plasma (1:5 with PBS) in the presence of 2 mM MeO-AMVN (䊐) or 20 mM AAPH (■). Values are mean ⫾ SD of five independent experiments.

conjugation between the phenyl moiety and the boron dipyromethen difluoride core which, in isolated form, exhibits a green fluorescence. The green fluorescence increase was significant after 30 min of incubation and increased linearly until 90 min (slope ⫽ 0.072 ⫾ 0.002 F.U. ⫻ min⫺1). Between 90 and 120 min, we observed a significant change of the slope (0.125 ⫾ 0.004 F.U. ⫻ min⫺1) that correlated with the consumption of ␣-tocopherol and ␤-carotene (see Fig. 2). No change of BODIPY fluorescence was observed in the presence of 2 mM AMVN or in the absence of the radical initiators for 4 h (data not shown). When 20 mM AAPH was used with human plasma, BODIPY oxidation was delayed 180 min (Fig. 5). Oxidation was observed after 240 min, presumably as a consequence of the loss of ␤-carotene (Fig.2) and the subsequent initiation of the lipid peroxidation process. BODIPY oxidation began immediately after addition of 2 mM MeO-AMVN. The intra-assay variation of plasma samples in repeated measurements resulted in less than 5% using either fluorescent probe. The CV calculated in the interassay precision resulted in 6.4% when DCFH was used and in 8.7% for BODIPY. Effect of plasma preincubation with ␣-tocopherol and ␤-carotene on lipid oxidizability To validate the determination of lipid plasma oxidizability using BODIPY as the fluorescence lipophilic

1048

G. ALDINI et al.

Fig. 6. Effect of time of preincubation of human plasma (1:5 with PBS) with ␣-tocopherol or ␤-carotene (10 ␮M final concentration) on lipid plasma oxidizability. Results are expressed as percentage inhibition of BODIPY oxidation induced by MeO-AMVN (2 mM) after 4 h of incubation. Legend to pattern: blank ⫽ no preincubation; dotted ⫽ 1 h preincubation; lines ⫽ 6 h preincubation. Values are mean ⫾ SD of five independent experiments. Statistical analysis: one-way ANOVA with Tukey’s post-test; *p ⬍ .05, **p ⬍ .01.

probe and MeO-AMVN as the lipophilic radical inducer, we studied the effect of adding the membrane soluble antioxidants, ␣-tocopherol (in EtOH) and ␤-carotene (in THF), to plasma. Both of these fat-soluble antioxidants were found to be effective in protecting the lipophilic probe against radical-initiated oxidation. As shown in Fig. 6, the effect was dependent on the duration of preincubation of the antioxidants with plasma.

DISCUSSION

In the present paper, a selective fluorescent method to measure oxidation of the aqueous and lipid compartments of plasma is described. In particular, a lipidsoluble radical initiator, MeO-AMVN, together with a lipid fluorescence probe, BODIPY, was selected to study the plasma lipid oxidizability. To monitor aqueous phase oxidizability, AAPH as the hydrophilic radical initiator coupled to DCFH oxidation was used. The distribution of the two radical initiators was determined by measuring the rate of consumption of hydrophilic and lipophilic endogenous antioxidants in plasma. In the presence of AAPH, we observed the following order of disappearance: ascorbic acid ⬎ ␣-tocopherol ⬎ uric acid and ␤-carotene. As suggested by others [20], ascorbic acid could effectively trap hydrophilic peroxyl radicals in the aqueous phase of plasma

before they are able to diffuse into the lipid phase. Similar consumptions of uric acid and ␤-carotene indicate that once ascorbic acid has been completely consumed, the remaining water-soluble antioxidants provide only a partial trap for the aqueous peroxyl radicals, which are then free to diffuse into the lipoproteins. When MeO-AMVN was used as the radical inducer, the order of disappearance was partially reversed with ␣-tocopherol ⬵ ascorbic acid ⬎ ␤-carotene ⬎⬎ uric acid. These results confirm the diffusion and activation of MeOAMVN into the lipoproteins. The consumption of ascorbic acid by the lipophilic radical inducer, MeO-AMVN, would strongly support previous findings that ascorbic acid can repair the ␣-tocopheroxyl radical, thereby regenerating ␣-tocopherol, and permitting it to function again as a free radical chain-breaking antioxidant [21, 22]. In accordance with the findings of Frei et al. [23], ␣-tocopherol appears to be unable to trap the MeOAMVN-derived lipid peroxyl radicals efficiently enough to prevent them from either attacking plasma lipids or from diffusing into the aqueous compartment. The consumption of uric acid by MeO-AMVN indicates that consumption of the fat-soluble antioxidants (e.g., ␣-tocopherol and ␤-carotene) probably resulted in movement of lipid radicals from lipid compartment to aqueous compartment [24]. The oxidation of ␣-tocopherol at a more rapid rate by AAPH than by MeO-AMVN can be explained by considering the orientation of ␣-tocopherol in the lipid compartment. The chroman head group of tocopherol is oriented toward the membrane interfacial region, whereas the phytyl side chain is embedded within the hydrocarbon region of lipid compartment. Because the head group is responsible for scavenging radicals, it would be expected to react more rapidly with the aqueous radicals generated from AAPH than with the radicals produced by MeO-AMVN, as the latter diffuses into the core of the lipoproteins. DCFH is a water-soluble indicator of radical-mediated oxidation. It has recently been used [25] to assess the total antioxidant activity of human serum. DCFH was used in the presence of AAPH to measure aqueous plasma oxidation. The selectivity of the method was confirmed inasmuch as DCFH oxidation only started after uric acid, the main hydrophilic plasma antioxidant, was consumed. In addition, when MeO-AMVN was used as the radical inducer, DCFH oxidation was significantly delayed, indicating its main localization in the aqueous domain. To study the lipid oxidation process induced by MeOAMVN, we used BODIPY as a lipophilic fluorescence probe for the following reasons: (i) it is characterized by a high fluorescence quantum yield limited to the lipid phase, (ii) it is stable for several hours in biological fluids

Measurement of plasma oxidizability

at 37°C, (iii) it absorbs/emits in the visible region, (iv) it was found to be a sensitive and selective indicator of lipid oxidation in plasma, and (v) the initial peroxidation rate is similar to that observed for arachidonic acid [19]. Immediately after MeO-AMVN addition, we observed the BODIPY oxidation whose rate constant significantly increased after the depletion of ␣-tocopherol and ␤-carotene, whereas it did not appear to be related to the levels of the hydrophilic antioxidants. When AAPH was used as the radical initiator, BODIPY oxidation was significantly delayed suggesting its localization in the lipid phase of plasma, and inaccessibility to the water-soluble peroxyl radicals generated from AAPH. To measure oxidizability of plasma lipids, a lipophilic radical generator coupled to a selective method capable of detecting lipid peroxidation should be used. The azocompound 2,2⬘-azobis(2,4-dimethylvaleronitrile) (AMVN) has been the most frequently used lipid-soluble radical initiator. However, the rate of free radical generation from AMVN is slow under physiological conditions, due in part to a lower efficiency of free radical generation in the viscous lipophilic compartment [26]. As such, high concentrations of AMVN (20– 40 mM) are usually required to induce and sustain the lipid peroxidation process in biological fluids [4]. However, these high concentrations of AMVN are not recommended because they form a cloudy precipitate and alter the physical-chemical properties of the lipid phase of plasma [11]. To detect lipid peroxidation induced by AMVN, the choice of analyte is usually limited to lipid hydroperoxides or TBARS analysis because the presence of the precipitate limits most spectrophotometric methods. However, lipid hydroperoxide analysis requires great care to avoid possible artifacts due to the presence of the hydroperoxides generated by the radical inducer itself [27]. As a consequence, TBARS analysis is still widely used to measure lipid plasma oxidizability, even though it lacks sensitivity and specificity. MeO-AMVN was found to be a suitable lipophilic radical-inducer, because it functioned at concentrations not interfering with the spectroscopic measurement. In contrast, the popular radical initiator AMVN was found to be ineffective at the same concentrations, while even a 4 mM concentration interfered with the analysis due to the formation of a cloudy precipitate. The measurement of lipid plasma oxidation was validated by adding two lipophilic antioxidants, ␣-tocopherol and ␤-carotene, to plasma samples. As previously reported [17], preincubation with these antioxidants improves the enrichment of the plasma lipid compartments where the lipid radicals generated by MeO-AMVN are primarily localized. The protective effect was found to be dependent on the duration of the preincubation period, suggesting a slow insertion of ␣-tocopherol and ␤-carotene into the lipid compartment when added under in

1049

vitro conditions. These results agree with the findings of Dugas et al. [28], who reported that in vivo enrichment of ␤-carotene is markedly more effective in inhibiting LDL oxidation than in vitro enrichment. Other studies have also indicated that the in vitro addition of ␤-carotene to LDL, without a preincubation period, will not offer antioxidant protection [29]. In conclusion, we report a fluorescence method to distinguish the oxidizability of both the aqueous and lipid compartments of plasma, that is characterized by sensitivity, specificity, and ease of determination. This method is different from the other conventional methods for measuring total antioxidant capacity, because other methods only measure the aqueous compartment of plasma, whereas the present method analyzes both the aqueous and the lipid compartments. This method will be useful in the evaluation of potential antioxidants and in particular to study the lipophilic component of the total antioxidant capacity of plasma. Further studies are in progress to fully validate this method as an analytical tool for assessing the hydrophilic/lipophilic antioxidant capacity of human plasma/serum. Acknowledgements — This research has been supported in part by the U.S. Department of Agriculture, under agreement number 1950-51000048-01A. The contents of this publication do not necessarily reflect the views or policies of the U.S. Department of Agriculture, nor does mention of trade names, commercial products, or organizations imply endorsement by the U.S. Government. We thank B. Gindelsky and E. Seyoum at Nutrition Evaluation Laboratory (HNRC) for the HPLC analyses of uric acid and ascorbic acid. Dr. Aldini was partially supported by Indena, Milan, Italy.

REFERENCES [1] Ghiselli, A.; Serafini, M.; Maiani, G.; Assini, E.; Ferro-Luzzi, A. A fluorescence-based method for measuring total plasma antioxidant capability. Free Radic. Biol. Med. 18:29 –36; 1995. [2] Cao, G.; Verdon, C. P.; Wu, A. H. B.; Wang, H.; Prior, R. L. Automated assay of oxygen radical absorbance capacity with the COBAS FARA II. Clin. Chem. 41:1738 –1744; 1995. [3] Murase, H.; Moon, J. H.; Yamauchi, R.; Kato, K.; Kunieda, T.; Yoshikawa, T.; Terao, J. Antioxidant activity of a novel vitamin E derivative, 2-(␣-D-glucopyranosyl)methyl-2,5,7,8-tetramethylchroman-6-ol. Free Radic. Biol. Med. 24:217–225; 1998. [4] Lolito, S. B.; Fraga, C. G. (⫹)-Catechin prevents human plasma oxidation. Free Radic. Biol. Med. 24:435– 441; 1998. [5] Massaeli, H.; Sobratte, S.; Pierce, G. N. The importance of lipid solubility in antioxidants and free radical generating systems for determining lipoprotein peroxidation. Free Radic. Biol. Med. 26: 1524 –1530; 1999. [6] Prior, R. L.; Cao, G. In vivo total antioxidant capacity: comparison of different analytical method. Free Radic. Biol. Med. 27: 1173–1181; 1999. [7] Pellegrini, N.; Riso, P.; Porrini, M. Tomato consumption does not affect the total antioxidant capacity of plasma. Nutrition 16:268 – 271; 2000. [8] Castenmiller, J. J. M.; Lauridsen, S. T.; Dragsted, L. O.; van het Hof, K. H.; Linssen, J. P. H.; West, C. E. ␤-Carotene does not change markers of enzymatic and nonenzymatic antioxidant activity in human blood. J. Nutr. 129:2162–2169; 1999. [9] Cao, G.; Booth, S. L.; Sadowski, J. A.; Prior, R. L. Increases in human plasma antioxidant capacity after consumption of con-

1050

[10]

[11]

[12]

[13] [14]

[15]

[16]

[17] [18] [19]

G. ALDINI et al. trolled diets high in fruit and vegetables. Am. J. Clin. Nutr. 68:1081–1087; 1998. Spranger, T.; Finckh, B.; Fingerhut, R.; Kohlschutter, A.; Beisiegel, U.; Kontush, A. How different constituents of human plasma and low density lipoprotein determine plasma oxidizability by copper. Chem. Phys. Lipids 91:39 –52; 1998. Noguchi, N.; Yamashita, H.; Gotoh, N.; Yamamoto, Y.; Numano, R.; Niki, E. 2,2⬘-azobis(4-methoxy-2,4-dimethylvaleronitrile), a new lipid-soluble azo initiator: application to oxidations of lipids and low-density lipoprotein in solution and in aqueous dispersions. Free Radic. Biol. Med. 24:259 –268; 1998. Kawamura, M.; Miyazaki, S.; Teramoto, T.; Ashidate, K.; Thoda, H.; Ando, N.; Kaneko, K. Gemfibrozil metabolite inhibits in vitro low-density lipoprotein (LDL) oxidation and diminishes cytotoxicity induced by oxidized LDL. Metabolism 49:479 – 485; 2000. Yamashita, H.; Nakamura, A.; Noguchi, N.; Niki, E.; Kuhn, H. Oxidation of low density lipoprotein and plasma by 15-lipoxygenase and free radicals. FEBS Lett. 445:287–290; 1999. Niki, E. Free radical initiators as source of water- or lipid-soluble peroxyl radicals. In: Packer, L.; Glazer, A. N., eds. Methods in enzymology. Oxygen radicals in biological systems (vol. 186). New York: Academic Press; 1990:100 –108. Yeum, K.-J.; Booth, S. L.; Sadowski, J. A.; Liu, C.; Tang, G.; Krinsky, N. I.; Russell, R. M. Human plasma carotenoid response to the ingestion of controlled diets high in fruits and vegetables. Am. J. Clin. Nutr. 64:594 – 602; 1996. Behrens, W. A.; Madere, R. A highly sensitive high-performance liquid chromatography method for the estimation of ascorbic and dehydroascorbic acid in tissues, biological fluids, and foods. Anal. Biochem. 165:102–107; 1987. Bowen, H. Z.; Omaye, S. T. Oxidative changes associated with ␤-carotene and ␣-tocopherol enrichment of human low-density lipoproteins. J. Am. Coll. Nutr. 17:171–179; 1998. Naguib, Y. M. A. Antioxidant activities of astaxanthin and related carotenoids. J. Agric. Food Chem. 48:1150 –1154; 2000. Pap, E. H. W.; Drummen, G. P. C.; Winter, V. J.; Kooij, T. W. A.; Rijken, P.; Wirtz, K. W. A.; Op den Kamp, J. A. F.; Hagel, W. J.; Post, J. A. Ratio-fluorescence microscopy of lipid oxidation in

[20]

[21] [22] [23]

[24] [25] [26]

[27]

[28]

[29]

living cells using C11-BODIPY581/591. FEBS Lett. 453:278 – 282; 1999. Frei, B.; Ames, B. U. Relative importance of vitamin E in antiperoxidative defenses in human blood plasma and low-density lipoprotein (LDL). In: Packer, L.; Fuchs, J., eds. Vitamin E in health and disease. New York: Marcel Dekker Inc.; 1993:131– 139. May, J. M. Is ascorbic acid an antioxidant for the plasma membrane? FASEB J. 13:995–1006; 1999. Buettner, G. R. The pecking order of free radicals and antioxidants: lipid peroxidation, ␣-tocopherol, and ascorbate. Arch. Biochem. Biophys. 300:535–543; 1993. Frei, B.; Stocker, R.; England, L.; Ames, B. N. Ascorbate: the most effective antioxidant in human blood plasma. In: Emerit, I.; Packer, L.; Auclair, C., eds. Antioxidants in therapy and preventative medicine. New York: Plenum Press; 1990:155–163. Carr, A.; Frei, B. Does vitamin C act as a pro-oxidant under physiological conditions? FASEB J. 13:1007–1024; 1999. Valkonen, M.; Kuusi, T. Spectrophotometric assay for total peroxyl radical-trapping antioxidant potential in human serum. J. Lipid Res. 38:823– 833; 1997. Kigoshi, M.; Sato, K.; Niki, E. Oxidation of lipids induced by dioctadecyl hyponitrite and di-t-butyl hyponitrite in organic solution and aqueous dispersions. Effect of reaction medium and size of radicals on efficiency of chain reaction. Bull. Chem. Soc. Jpn. 66:2954 –2959; 1993. Zhang, J. R.; Cazers, A. R.; Lutzke, B. S.; Hall, E. D. HPLCchemiluminescence and thermospray LC/MS study of hydroperoxides generated from phosphatidylcholine. Free Radic. Biol. Med. 18:1–10; 1995. Dugas, T. R.; Morel, D. W.; Harrison, E. H. Dietary supplementation with ␤-carotene but not with lycopene, inhibits endothelial cell-mediated oxidation of low-density lipoprotein. Free Radic. Biol. Med. 26:1238 –1244; 1999. Gaziano, J. M.; Hatta, A.; Flynn, M.; Johnson, E. J.; Krinsky, N. I.; Ridker, P. M.; Hennekens, C. H.; Frei, B. Supplementation with ␤-carotene in vivo and in vitro does not inhibit low density lipoprotein (LDL) oxidation. Atherosclerosis 112:187–195; 1995.