High-Throughput Fluorescence Screening of Antioxidative Capacity in Human Serum

Analytical Biochemistry 297, 144 –153 (2001) doi:10.1006/abio.2001.5333, available online at http://www.idealibrary.com on

High-Throughput Fluorescence Screening of Antioxidative Capacity in Human Serum Birgit Mayer,* Martin Schumacher,† Helga Brandsta¨tter,* Franz S. Wagner,‡ and Albin Hermetter* ,1 *Department of Biochemistry, Technische Universita¨t Graz, A-8010 Graz, Austria; †Department of Internal Medicine, Division of Cardiology, Karl-Franzens University Graz, Graz, Austria; and ‡Department of Food Technology and Nutrition, Integrated Rural Development, Styria, Austria

Received February 23, 2001; published online September 12, 2001

Diphenylhexatriene-labeled phosphatidylcholine and propionic acid have been established as selective fluorescence markers for the continuous determination of oxidation processes in the lipid and aqueous phases of unfractionated human serum. Oxidation of the respective fluorophores leads to a decrease in fluorescence intensity from which the time-dependent degradation of the marker molecule can be determined. The lag times preceding the propagation of oxidation are representative for the antioxidative capacity of the system, which may be influenced by exogenous factors, e.g., the antioxidants from the diet. Supplementation of human serum by quercetin, rutin, vitamin E, vitamin C, or total apple phenolics in vitro led to a decrease in oxidizability depending on the oxidation marker and the hydrophobicity of the antioxidant. Quercetin and vitamin E showed a higher in vitro capacity of protecting lipoproteins against oxidation. In contrast, rutin and vitamin C were more efficient as inhibitors in the aqueous phase. The same effect on serum was found after dietary consumption of apples. This result is in line with the known observation that intake of plant polyphenols leads to an increase in serum levels of hydrophilic antioxidants. © 2001 Academic Press

Key Words: quercetin; rutin; nutrition; diphenylhexatriene-phospholipids; apple flavonoids; free radicals.

Evidence has been accumulated indicating that regular intake of dietary antioxidants may protect against various diseases (1– 4). This effect is not only attributed to the action of the “classic” antioxidants like 1 To whom correspondence should be addressed. Fax: 43-316-8736952. E-mail: [email protected].

144

vitamin C and vitamin E. In addition, flavonoids and other phenolic compounds are components that contribute to the high antioxidative capacity measured in the diet, e.g., fruits and vegetables (5– 8). Some of their antioxidative activities are as good or even stronger than those of vitamins C and E (9, 10). Many data have been published on the inhibitory effect of flavonoids on lipid oxidation which is likely to play a central role in the development of cardiovascular diseases and atherosclerosis (11). It has been inferred from these results that polyphenols should counteract the fatal physiological consequences (12). Intervention studies based on measurements of ex vivo inhibition of low-density lipoprotein (LDL) 2 oxidation after flavonoid intake do not support a general consensus about a potential antioxidative effect (13–15) although in vitro supplementation studies clearly showed a high antioxidative potential of these compounds (16, 17). While the respective studies focused on the effect of flavonoids on lipoprotein oxidation, little is known about their effect on protein oxidation, although quercetin in plasma is supposed to bind preferentially to human serum albumin (18). Accumulation of oxidized protein is associated with a number of diseases (19) and brief exposure to oxidative stress leads to alterations in protein structure similar to those associated with aging (20, 21). The determination of suitable biomarkers for oxidative damage of proteins is being used as a routine analysis procedure (22). How2 Abbreviations used: LDL, low-density lipoprotein; BSA, bovine serum albumin; HSA, human serum albumin; PBS, phosphate-buffered saline; AAPH, 2,2⬘-azobis(2-amidinopropane)hydrochloride; DPHPA, 1,6-diphenylhexatriene propionic acid; DPHPC, 1-palmitoyl-2-((2-(4-(6-phenyl-trans-1,3,5-hexatrienyl)phenyl)ethyl)-carbonyl-sn-glycero-3-phosphocholine; HDL, high-density lipoprotein; VLDL, very low density lipoprotein.

0003-2697/01 $35.00 Copyright © 2001 by Academic Press All rights of reproduction in any form reserved.

FLUORESCENCE SCREENING OF SERUM ANTIOXIDANT CAPACITY

ever, these methods cannot be used for the analysis of unfractionated serum without time-consuming separation procedures. Therefore, it was the objective of our study to develop specific markers for the hydrophilic and the hydrophobic phase in unfractionated human serum without the need of isolation of the oxidation products. We applied the new markers to find out whether and to what extent apple polyphenols inhibited oxidation in serum and thus improved its antioxidative capacity. Data obtained with serum after supplementation with isolated flavonoids and apple extracts in vitro as well as serum samples taken after sustained dietary apple intake are clearly in support of an inhibitory effect on serum oxidation. MATERIALS AND METHODS

Materials Quercetin, rutin, BSA (bovine serum albumin, fraction V, essentially fatty acid-free), HSA (human serum albumin, fraction V, essentially fatty acid-free), Barbital buffer, and Dulbecco’s phosphate-buffered saline (PBS buffer) were purchased from Sigma (Vienna, Austria). AAPH (2,2⬘-azobis(2-amidinopropane)hydrochloride) was from PolyScience (Warrington, PA). DPHPA (1,6-diphenylhexatriene propionic acid) was purchased from Molecular Probes (Eugene, OR). DPHPC (1-palmitoyl-2-((2-(4-(6-phenyl-trans-1,3,5-hexatrienyl)phenyl)ethyl)-carbonyl)-sn-glycero-3-phosphocholine) was prepared as described (23) (Fig. 1). LDL and highdensity lipoprotein, fraction 3 (HDL 3), were prepared as described (24). Standard laboratory chemicals were purchased from Merck (Darmstadt, Germany) and Sigma (Vienna, Austria) and were of analytical grade. Fluorescence Oxidation Assay An ethanolic solution (10 ␮L) of fluorophore (DPHPC or DPHPA, 20 nmol) was added to 1 mL of an argonsaturated (lipo)protein sample. For labeling with DPHPC, the reaction mixture was kept under argon at 37°C for 12 h, whereas for DPHPA labeling the mixture was incubated under argon at room temperature for 1 h. A 90-␮L aliquot of the labeled sample was diluted with 210 ␮L of air-saturated PBS buffer. Oxidation was started by the addition of 30 ␮L of 300 mM AAPH (30 mmol/L final concentration) dissolved in PBS and monitored by following the time-dependent decrease in fluorescence intensity at 430 nm (excitation at 354 nm) at 37°C using a fluorescence well-plate reader Perkin– Elmer LS50B and 96-multiwell plates from Perkin– Elmer Corp. (Norwalk, CT). Tryptophan fluorescence as a measure of BSA oxidation was determined in PMMA cuvettes from Semadeni AG, Switzerland. BSA concentration in PBS buffer

145

was 1.5 mg/mL. The reaction volume was 2 mL. Protein oxidation was started with AAPH (20 mmol/L assay concentration) and monitored at 37°C by continuous recording of fluorescence intensity at 344 nm (excitation at 280 nm) using a Shimadzu RF-5301 spectrofluorophotometer (Shimadzu Corp., Kyoto, Japan). Agarose Gel Electrophoresis SeaKem Agarose-LE (FMC BioProducts, Rockland, ME) (0.3 g) was boiled in 50 mL Barbital buffer (pH 8.6) and used for the preparation of electrophoresis gels (7 mm). Electrophoresis was carried out at 56 V for 2 h in an electrophoresis chamber (SCIE-PLAS, Warwickshire, UK) at room temperature. LDL (1 mg protein/mL), HDL (1 mg protein/mL), HSA (4 mg/mL), or human serum (4 mg protein/mL) was incubated with DPHPA or DPHPC as indicated before and 10 ␮L of each sample was applied onto the gel. Ten microliters of native, nonlabeled HSA (4 mg/mL) was also analyzed. After electrophoresis, the agarose gel was inspected using a gel documentation system (EX ⫽ 365 nm) Herolab EASY (Herolab, Wiesloch, Germany) to visualize the fluorescent bands. After taking pictures of the fluorescence pattern, the protein bands were stained with Coomassie blue for 30 min. Determination of Total Phenolics Total phenolics were determined using the Folin– Ciocalteau reagent (25). One gram of a freeze-dried apple sample was extracted with 40 mL of 80% methanol containing 1% hydrochloric acid at room temperature for 2 h using an orbital shaker set at 200 rpm. The mixture was centrifuged at 1500g for 10 min and the supernatant was decanted. The pellet was extracted again under identical conditions. The supernatants were combined and analyzed for total phenolics according the following procedure. One-hundred microliters of extract was mixed with 750 ␮L of Folin–Ciocalteau reagent (diluted 10-fold with distilled water) and left at room temperature for 5 min. Seven-hundred fifty microliters of sodium bicarbonate (60 g/L) solution was added to the mixture. After 90 min, absorbance was measured at 725 nm. Results were expressed as ferulic acid equivalents. In Vitro Supplementation of Serum with Polyphenols Human serum was prepared from blood freshly drawn from 40 healthy donors. Venous blood samples were collected, using vacuum tubes, and aliquots of isolated serum were stored at ⫺20°C. Stock solutions of rutin, quercetin, and vitamin E were prepared in ethanol as a solvent whereas vitamin C was solved in water. Apple extracts were prepared in 80% MeOH containing 1% HCl. To study the effect of exogenously

146

MAYER ET AL.

added antioxidants and total apple phenolics on serum oxidation, serum samples were supplemented with 10-␮L ethanolic solutions containing appropriate amounts of the respective antioxidants or 2.5, 5, or 10 ␮L of the above methanolic apple extract. The supplemented sera were labeled with DPHPA or DPHPC and label oxidation was measured as described above. Only vitamin C was added to the prelabeled serum immediately before oxidation. The indicated lag times are means ⫾ SD (n ⫽ 3). Apple Intervention Study The study protocol was approved by the Ethics Committee Institutional Review Board of the General and University Hospital of Graz and the Faculty of Medicine of the Karl-Franzens-University of Graz. The procedure and the purpose of the study were fully explained to the participants, who gave their written, informed consent. Design Nineteen women and 28 men participated in the study. All participants were healthy and did not take any medication during the whole study. The subjects followed a nearly antioxidant-free diet for 1 week prior to the supplement diet. For this purpose, a list of authorized foods was handed out to the participants. During supplementation, compliance with the above diet was monitored. Supplements consisted of 1 kg apples daily (Jonagold, Golden Delicious, Elstar and Gala grown in the southeast of the province of Styria, Austria). The apples were donated by the Fruit Production Organization of Styria, Austria, and were stored under controlled atmosphere conditions (ULO, ultralow oxygen storage). For total polyphenol determination apples were lyophilized. Venous blood samples were collected using vacuum tubes, and aliquots of isolated serum were stored at ⫺20°C. Immediately before apple consumption, blood samples were taken (t ⫽ 0 h). The second blood sample was taken after a period of 3 h, during which 1 kg apples was consumed (t ⫽ 3 h). Blood samples were then collected 24 h later without further apple intake (t ⫽ 24 h). After an additional 4 days of daily apple intake, another was taken 24 h after the last apple consumption (t ⫽ 24* h). Statistics and Analytical Evaluation For the in vitro studies, measurements were carried out in triplicate. The indicated lag times are means ⫾ SD. In the apple intervention study, data are reported as means ⫾ SE. Results have been found to fit a normal distribution and were further compared using a paired two-population t test. The statistical analyses were performed with Microcal Origin, Version 5.0 software.

The level of significance was set P ⬍ 0.05. Linear regression analyses of lag times versus prooxidant concentrations were carried out using a linear fit ( y ⫽ kx ⫹ d). For in vitro experiments and the analytical evaluation of the test system, pooled serum from 40 healthy subjects was used, whereas in the apple intervention study individual serum samples were analyzed. Reagent stability. The stock solution of the fluorescent marker (DPHPA) is stable for at least 3 years if stored at ⫺20°C in the dark, whereas DPHPC stability is lower (1 year). Both markers are hydrolyzed after 4 weeks at 4°C. Linearity. Under the experimental conditions described above lag times decrease linearly with prooxidant concentrations from 10 up to 60 mM AAPH (Fig. 4, y ⫽ ⫺5.83x ⫹ 328.3, correlation coefficient ⫽ 0.995). Precision. Results of precision studies are shown in Table 1. Within-run precision was determined from 20 replicate determinations for both fluorescent markers. Between-day precision was determined from data from 10 measurements on the same serum samples. Between each experiment, the sera were stored at ⫺20°C. Sensitivity. Sensitivity depends on the amount of label used (here it is 0.1 ␮M, which is 1 order of magnitude above the detection limit). Interferences. Serum components such as bilirubin, hemoglobin, and other chromophores do not interfere with our assay system since our assay is based on the measurement on relative changes of fluorescence intensities giving the same lag times irrespective of the initial absolute fluorescence intensity. RESULTS

In Vitro Studies Labeling of (lipo)protein samples. DPHPA and DPHPC (Fig. 1) are fluorescent analogs of carboxylic acids and phospholipids. Upon oxidation, their DPH fluorophores are destroyed and from the decrease in fluorescence intensity the rates of label oxidation can be determined. The fluorescent compounds readily insert into the surface of isolated serum lipoproteins (DPHPA and DPHPC) or bind preferentially to excess serum albumin (DPHPA) even in the presence of lipoproteins. Figure 2 shows the fluorescently labeled samples after agarose gel electrophoresis (LDL, HDL 3, mixtures of both lipoproteins, HSA, and serum), each labeled either with DPHPC or with DPHPA. In Fig. 2A all samples had been preincubated with DPHPC as a fluorescent probe which incorporated only into the surface of lipoproteins but not into proteins. As a consequence, DPHPC was detected in all lipoprotein samples (lanes 1–3). Lane 4 shows a mixture of LDL and

FLUORESCENCE SCREENING OF SERUM ANTIOXIDANT CAPACITY

147

FIG. 1. Chemical structures of redox-sensitive fluorophores. (A) DPHPA (1,6-diphenylhexatriene propionic acid). (B) DPHPC (1-palmitoyl2-((2-(4-(6-phenyl-trans-1,3,5-hexatrienyl)phenyl)ethyl)-carbonyl)-sn-glycero-3-phosphocholine).

HSA. In this case, only the LDL fraction was labeled. The same tendency is shown in lane 5. When pure HSA was preincubated with DPHPC, no fluorescence was observed. In human serum, LDL and, to a lesser extent, HDL 3 were labeled (lane 6). Unlabeled HSA did not bind DPHPC (lane 7). Therefore, this lane was not visible under the conditions employed for DPH fluorescence detection. In Fig. 2B samples 1– 6 were preincubated with DPHPA. This fluorescence marker binds to isolated lipoproteins and albumin as well (lanes 1– 4). In an LDL/HDL 3 mixture (lane 3), the fluorophore preferentially binds to the LDL fraction. In a mixture of HSA and isolated LDL, both components were labeled to a similar extent (lane 4). In the pure HSA sample (lane 5) and the human serum sample (lane 6) DPHPA was entirely confined to the abundant albumin fraction. Significant binding of DPHPA to HDL in the presence of albumin can be excluded, since binding to HDL is weak (lane 2), whereas binding to albumin is stronger (lane 5). Unlabeled HSA was again nonfluorescent (lane 7). Protein-containing bands as detected by fluorescence and Coomassie blue staining were identical. Isolated very low density lipoprotein (VLDL) migrates only slightly faster than LDL (data not shown). LDL and VLDL in a mixture of both or in serum show only a single (broadened) band under the same experimental conditions. VLDL is also efficiently labeled by DPHPC. Therefore, DPHPC monitors phospholipid oxidation in the surface of both VLDL and LDL in serum. VLDL in pure form is also labeled by DPHPA. But in the presence of excess HSA (corresponding to its serum concentration) no VLDL labeling is observed. Under these conditions, only the watersoluble protein is labeled. In summary, DPHPA in human serum can be considered an oxidation marker for the aqueous phase

(most likely in albumin-bound form), whereas DPHPC is a marker for the lipid phase (lipoproteins). Antioxidant effects of in vitro supplementation with antioxidants and total apple phenolics. We used both labels to monitor label oxidation in lipid (lipoprotein) and protein (albumin) environments and determined the protective effect of apple polyphenols against oxidative stress in biological fluids (serum). The time course of oxidative DPH degradation in the labeled samples was determined from the decrease of their initial fluorescence intensity. Figure 3 shows typical oxidation experiments with DPHPA (Fig. 3A) and DPHPC (Fig. 3B) as fluorescence markers in human serum. After addition of AAPH (30 mmol/L) as a prooxidant at 37°C a lag phase is observed. Lag times as determined from the intersection of the tangents are much shorter for DPHPA (Fig. 3A) as compared to DPHPC (Fig. 3B). The observed lag times linearly depend on AAPH concentrations (Fig. 4) and were obtained with reasonable within-run and between-day precision (Table 1). The influence of various antioxidants on the oxidation of human serum components was quantified from the prolongation of the lag times. If used in parallel experiments, quercetin, rutin, vitamin E, and vitamin C inhibited serum oxidation to a different extent. In Fig. 5 the effect of exogenously added antioxidants on the oxidation of DPHPA and DPHPC in human serum is shown. Quercetin (Fig. 5A) and vitamin E (Fig. 5B), being the more lipophilic antioxidants, exerted an antioxidative effect on the lipoprotein marker DPHPC, whereas oxidation of DPHPA in serum was much less affected. Rutin (Fig. 5C), being the more polar glycoconjugate of quercetin and vitamin C (Fig. 5D) did not influence DPHPC lag times so much, while DPHPA, which binds to the protein fraction of serum, was more

148

MAYER ET AL.

FIG. 3. Oxidation assays. AAPH-mediated oxidation of DPHPA (A) and DPHPC (B) in human serum. At time zero, AAPH (final concentration 30 mmol/L) was added to the labeled serum samples and oxidation was monitored at 37°C by following the decrease of fluorescence intensity at 430 nm (EX ⫽ 354 nm). Lag times were determined graphically as indicated. FIG. 2. Agarose gel electrophoresis of fluorescent lipoproteins, HSA, and human serum. Lipoproteins, HSA, or mixtures of both were labeled with DPHPC (A) or DPHPA (B) and samples were made visible under a gel documentation device using an UV lamp (EX ⫽ 365 nm): LDL (lane 1), HDL (lane 2), LDL/HDL (lane 3), LDL/HSA (lane 4), HSA (lane 5), human serum (lane 6), and unlabeled HSA (lane 7). Samples (lanes 1–7) were applied onto slots at different starting lines A and B, respectively, and simultaneously run on the same gel.

efficiently protected against oxidation. The inhibition of DPHPC oxidation by the hydrophobic compounds in human serum is higher compared to the effect of the hydrophilic ones. For DPHPA oxidation in serum the opposite effect was observed. We must emphasize that under our experimental conditions (storage at ⫺20°C and thawing of the serum samples before use; fluorescence labeling of serum at 37°C overnight) vitamin C is destroyed and therefore does not affect our results. Only if vitamin C was added immediately before the experiment did it improve the antioxidative capacity of serum as is shown in Fig. 5. In contrast, quercetin, rutin, and vitamin E were stable under the conditions indicated above (data not shown). Total apple phenolics exogenously added to DPHPC labeled serum showed also a concentration-dependent inhibition of lipid oxidation at the same concentrations as observed with the pure polyphenols quercetin and rutin (Fig. 6).

Antioxidant capacity as determined by the fluorescence method has also been compared to different established techniques. Lag times of the fluorescent lipid marker (DPHPC) are in good agreement with antioxidant effects that have been observed using the TEAC test (26), which is based on the ABTS ⫹ radical as a reporter molecule (M. Serafini, G. Maiani, B. Mayer, A. Hermetter, P. Castilla, M. A. Lasuncio´n, J. Wilczak, P.

FIG. 4. Effect of AAPH concentrations on DPHPA oxidation of human serum. Plot of lag times versus AAPH concentration. Regression coefficients: y ⫽ kx ⫹ d, where d (328.2 ⫾ 1.6), k (⫺5.83 ⫾ 0.03), and x are the intercept, slope, and concentration (mM), respectively. Multiple correlation coefficient ⫽ 0.995.

149

FLUORESCENCE SCREENING OF SERUM ANTIOXIDANT CAPACITY TABLE 1

Precision of Lag Times as Determined by the Fluorescence Oxidation Assay

DPHPA (DPH-propionic acid) DPHPC (DPH-phosphatidylcholine)

Ostaszewski, R. Pulido, and F. Saura-Calixto, manuscript in preparation). Oxidation rates obtained with DPHPA as an albumin-binding marker in serum were compared with kinetic results from the continuous monitoring of tryptophan fluorescence during protein oxidation (28, 29). Figure 7 shows the effect of quercetin on the loss of BSA tryptophan fluorescence as induced by AAPH as a prooxidant (Fig. 7A). Quercetin concentrations of 2 ␮mol/L already protected BSA against oxidation to a significant extent and addition of 10 ␮mol/L quercetin to the BSA solution completely inhibited tryptophan oxidation for at least 90 min. Addition of quercetin to DPHPA-labeled BSA (Fig. 7B) inhibited oxidation to a similar extent compared

Within-run (n ⫽ 20) mean ⫾ SD (min)

Between-day mean ⫾ SD (min)

200 ⫾ 10.4 273 ⫾ 9.8

205 ⫾ 8.9 270 ⫾ 10.6

to tryptophan oxidation. Quercetin concentrations of 2 ␮mol/L again showed a measurable antioxidative effect. No oxidation of the BSA-bound label occurred in the presence of 10 ␮mol/L quercetin. Thus, both experiments essentially gave the same results confirming the assumption that DPHPA is a reliable marker for oxidation processes in the aqueous phase of serum. This is important insofar as DPHPA fluorescence can unambiguously be measured in serum whereas tryptophan emission cannot. Apple Intervention Study DPHPA and DPHPC were used as hydrophilic and hydrophobic markers to study the effect of dietary ap-

FIG. 5. Effect of exogenously added antioxidants on serum oxidation in vitro. Effect of quercetin (A), vitamin E (B), rutin (C), and vitamin C (D) on oxidation of DPHPC (䊐) and DPHPA (■) in human serum. Oxidation was started by addition of AAPH (final concentration 30 mmol/L) to the labeled serum at 37°C. Lag times were determined as indicated in the legend to Fig. 3. The indicated values are means ⫾ SD (n ⫽ 3).

150

MAYER ET AL.

FIG. 6. Effect of total apple phenolics on DPHPC oxidation in serum. Concentrations of apple phenolics were 0 (}), 1.7 ␮g/mL (E), 3.4 ␮g/mL (Œ), and 6.8 ␮g/mL (⫻). After addition of apple extracts to serum, followed by labeling with DPHPC, oxidation was started by addition of AAPH (final concentration 30 mmol/L) to the labeled serum at 37°C and DPH fluorescence was monitored at 430 nm (EX ⫽ 354 nm).

ple consumption on the antioxidant capacity of serum in the aqueous and lipoprotein phases, respectively. For this purpose, blood serum was collected from volunteers, followed by labeling with either probe and determination of lag times of label oxidation, using AAPH as a prooxidant. The total phenolic content of the apples given in the intervention study was 2.71 ⫾ 0.15 g/kg of fresh apples. The results obtained with DPHPA (hydrophilic marker) and DPHPC (hydrophobic marker) are shown in Figs. 8 and 9, respectively. In general, lag times of DPHPA oxidation in serum were shorter than lag times of DPHPC oxidation. Dietary consumption of apples led to a statistically significant reduction in the susceptibility of DPHPA toward oxidation (Fig. 8). Specifically, 3 h after intake of 1 kg apples, DPHPA lag times increased from 224.7 ⫾ 5.7 to 241.6 ⫾ 6.1 min (P ⬍ 0.01). Serum, prelabeled with the lipoprotein-specific marker DPHPC, showed a smaller increase of lag times under the same conditions (from 282.3 ⫾ 9.1 to 290.0 ⫾ 8.7 min, P ⫽ 0.05) (Fig. 9). Twenty-four hours after apple consumption, lag times for DPHPC and DPHPA oxidation returned to the initial values observed before apple consumption (see Figs. 8 and 9). Thus, the antioxidant capacity of serum is short-lived (3 h) and does not lead to a sustained improvement of the antioxidant status if intervals between apple intake are longer. DISCUSSION

We analyzed the influence of apple consumption on the antioxidative capacity of human serum in compar-

FIG. 7. Effect of quercetin on BSA oxidation. (A) Oxidation of BSA tryptophans (ⴱ, no quercetin; 䊐, 2 ␮mol/L quercetin; }, 10 ␮mol/L quercetin). Tryptophan fluorescence intensity was measured at 344 nm (EX ⫽ 280 nm). (B) Oxidation of BSA-bound DPHPA (ⴱ, no quercetin; 䊐, 2 ␮mol/L quercetin, } 10 ␮mol/L quercetin). DPH fluorescence was followed at 430 nm (EX ⫽ 354 nm). Oxidation was induced by 20 mmol/L AAPH at 37°C. BSA and DPHPA concentrations were 1.5 mg/mL and 4 nmol/mL, respectively.

ison with in vitro effects of exogenously added apple flavonoids, vitamins, and total apple phenolics on serum oxidizability under identical experimental conditions. The experiments were carried out using two fluorescent probes, whose oxidation susceptibility is very similar compared to natural polyunsaturated fatty acids and albumin (30). They exhibit high fluo-

FIG. 8. Effect of apple consumption on DPHPA oxidation in human serum at 37°C. Oxidation was started by addition of AAPH (final concentration 30 mmol/L) to the labeled serum sample. The indicated values are means ⫾ SEM (n ⫽ 47) and correspond to serum samples collected immediately before and 3 and 24 h after consumption of 1 kg apples and 24* h after 4 days of daily apple consumption (* P ⬍ 0.01). DPHPA concentration was 20 nmol/mL serum.

FLUORESCENCE SCREENING OF SERUM ANTIOXIDANT CAPACITY

FIG. 9. Effect of apple consumption on DPHPC oxidation in human serum at 37°C. Oxidation was started by addition of AAPH (final concentration 30 mmol/L) to the labeled serum. The indicated values are means ⫾ SEM (n ⫽ 47) and correspond to serum samples collected immediately before and 3 and 24 h after consumption of 1 kg apples and 24* h after 4 days of daily apple consumption (P ⬎ 0.05). DPHPC concentration was 20 nmol/mL serum.

rescence intensities in diluted form and lose their fluorescence when oxidized. If incubated with lipoproteins or serum, DPHPC is incorporated into the LDL, VLDL, and HDL fractions. In serum, DPHPA preferentially binds to albumin. Therefore, fluorescent DPHPC can be used as a marker for the oxidation of polyunsaturated phosphatidylcholine in the surface of lipoproteins, whereas DPHPA is an appropriate reporter for the aqueous phase in human serum. Oxidation of lipids (DPHPC) may very much depend on the influence of the metabolic state of the donor, which is the subject of current studies in our laboratory. Chylomicrons played only a marginal role in our oxidation experiments, since serum was prepared from blood of fasting donors. Under these conditions, serum contains less than 0.1 g/L of this lipoprotein class compared to 2–3.5 g LDL/L and 0.5–2 g VLDL/L (31). In addition, phospholipid only represents 5% of total chylomicron mass (31). Therefore, fluorescence due to labeling of the surface of these particles (the core is not labeled by amphipathic DPH probes) is negligible compared with DPHPC fluorescence from LDL, HDL, and VLDL. The oxidation kinetics of DPHPC as determined from the gradual decrease in fluorescence intensity are identical to the time-dependent formation of lipid dienes in natural unlabeled lipoproteins (32). On the other hand, the oxidation kinetics, as determined from DPHPA and tryptophan fluorescence, are comparable, too. Thus, we report for the first time a continuous spectroscopic method which enables us to determine the oxidizability of hydrophobic and hydrophilic components in the same biological system. Using a fluorescence plate reader, this technique can be used as a high-throughput screening method for the efficient analysis of oxidation processes and antioxidant effects. The new methodology opens new possibilities in light of a recent hypothesis. According to Heinecke (33), protein oxidation may play a more important role in atherosclerosis than previously anticipated, and the

151

ability of antioxidants to inhibit protein oxidation may differ markedly from their ability to inhibit lipid oxidation. Therefore, it seems to be crucial to have selective methods to study both lipophilic and hydrophilic effects under oxidative stress in a physiological environment. Our data show that the lag times of the DPHPA marker in serum were always shorter than the lag times of the DPHPC marker. Presumably, proteins are less efficiently protected by their surrounding water-soluble antioxidants compared to lipoproteins which contain, in addition, lipophilic antioxidants (29). Thus, proteins might be preferential targets for watersoluble free radicals. Quercetin and, to a lesser extent, vitamin E, if exogenously added to serum, exert a much higher antioxidative effect on the lipoprotein (DPHPC) compared to the DPHPA marker. In contrast, supplementation by rutin, which is rendered more hydrophilic by its sugar moiety, and vitamin C was more efficient in protecting the hydrophilic marker against oxidation. These results are in line with the assumption that water-soluble proteins are protected mainly by water-soluble antioxidants (rutin is more polar than quercetin and thus more potent in this respect), whereas lipoproteins are affected differently. The hydrophilic antioxidants cannot scavenge lipophilic radicals within the lipid phase (34) but help restore their lipophilic antioxidant pool on the lipoprotein surface (vitamin E) and shield the particle from being attacked from free radicals in solution. In this respect, flavonoids are even more potent antioxidants compared to vitamins (up to 100-fold better protection). Serum as a system for studies of antioxidant effects is superior to isolated LDL as a substrate, if the antioxidant(s) to be studied are water-soluble. During the process of LDL isolation, the water-soluble components including the hydrophilic antioxidants are removed, leaving the “pure” lipoprotein particle. As a consequence, only marginal inhibition, if any, is observed, making the significance of the results questionable. For instance, apple flavonoids such as quercetin exist in glycoconjugated form in apples and other foods. In the body they are further modified giving rise to the formation of the more polar quercetin glucuronides and quercetin-3-O-sulfate which finally show up in the serum (35, 36). Thus, either direct uptake of polyphenol glycoconjugates into the blood stream or eventual modification of the aglycone by glucuronidation and sulfatation should increase the water-soluble antioxidant pool. As a consequence, isolated LDL is not an ideal system to study antioxidant effects of flavonoids as components of the daily diet. Relevant antioxidative effects on serum can only be measured if oxidation in the lipid and aqueous phase are determined in serum preserving the natural (or dietary-enhanced) concentrations of water-soluble antioxidants. The in vitro antioxidant effects of exogenously added quercetin and

152

MAYER ET AL.

rutin in serum, both representing important flavonoids in apples, have been observed at potentially physiological concentrations (37, 38). Apple extracts containing a mixture of different phenolics showed antioxidant effects at the same concentrations. In addition to in vitro supplementation of serum with apple polyphenols, consumption of apples also increased the resistance of human serum against oxidation. Again, the effects on the aqueous and lipoprotein phase in serum were different. Three hours after apple consumption we measured only a small increase of lag times for lipid oxidation, whereas the oxidation of DPHPA was efficiently inhibited. Thus, the more pronounced protection of the water-soluble compounds in human serum is in line with the in vitro supplementation studies, demonstrating that hydrophilic polyphenols (sugar conjugates and metabolites, see above) more efficiently inhibit oxidation of the water-soluble components (proteins). The time dependence of the physiological (antioxidative) effects as determined in our intervention study is in good agreement with results from analytical studies on flavonoid intake from flavonoid-rich meals (37). A maximum of the antioxidative effect was reached 3 h after apple intake. After 24 h, lag times of DPHPA and DPHPC oxidation returned to the initial values measured before the intervention. Even after an additional 4 days of daily apple intake, the sample taken 24 h after the last apple consumption showed no effect. Therefore, our data support the hypothesis that potentially beneficial effects of apple consumption are only short-lived and long-term effects are only warranted by regular apple intake.

6.

7.

8. 9.

10.

11. 12. 13.

14.

15.

16.

ACKNOWLEDGMENTS We thank the students of the Technische Universita¨t Graz and the students of the Fruit Production School Wetzawinkel, Austria, for participating in the study. We are indebted to W. Sattler for providing HDL and VLDL samples. Financial support from the province government of Styria/Austria is gratefully acknowledged.

17.

18.

19.

REFERENCES 1. Hertog, M. G., Kromhout, D., Aravanis, C., Blackburn, H., Buzina, R., Fidanza, F., Giampaoli, S., Jansen, A., Menotti, A., Nedeljkovic, S., et al. (1995) Flavonoid intake and long-term risk of coronary heart disease and cancer in the seven countries study [published erratum appears in Arch. Intern. Med., 1995, Jun 12; 155(11): 1184]. Arch. Intern. Med. 155, 381–386. 2. Carr, A. C., and Frei, B. (1999) Toward a new recommended dietary allowance for vitamin C based on antioxidant and health effects in humans. Am. J. Clin. Nutr. 69, 1086 –1107. 3. Hollman, P. C. H., Hertog, M. G. L., and Katan, M. B. (1996) Analysis and health effects of flavonoids. Food Chem. 57, 43– 46. 4. Keli, S. O., Hertog, M. G., Feskens, E. J., and Kromhout, D. (1996) Dietary flavonoids, antioxidant vitamins, and incidence of stroke: The Zutphen study. Arch. Intern. Med. 156, 637– 642. 5. Guo, C., Cao, G., Sofic, E., and Prior, R. (1997) HPLC coupled with coulometric array detection of electroactive components in

20.

21.

22.

23.

24.

fruits and vegetables: Relationship to oxygen radical absorbance assay. J. Agric. Food Chem. 45, 1787–1796. Heinonen, I. M., Meyer, A. S., and Frankel, E. N. (1998) Antioxidant activity of berry phenolics on human low-density lipoprotein and liposome oxidation. J. Agric. Food Chem. 46, 4107– 4112. Velioglu, Y. S., Mazza, G., Gao, L., and Oomah, B. D. (1998) Antioxidant activity and total phenolics in selected fruits, vegetables and grain products. J. Agric. Food Chem. 46, 4113– 4117. Halliwell, B. (1996) Antioxidants in human health and disease. Annu. Rev. Nutr. 16, 30 –50. Cao, G., Booth, S. L., Sadowski, J. A., and Prior, R. L. (1998) Increases in human plasma antioxidant capacity after consumption of controlled diets high in fruit and vegetables. Am. J. Clin. Nutr. 68, 1081–1087. Cao, G., Sofic, E., and Prior, R. L. (1997) Antioxidant and prooxidant behavior of flavonoids: Structure–activity relationships. Free Rad. Biol. Med. 22, 749 –760. Steinberg, D. (1991) Antioxidants and atherosclerosis: A current assessment. Circulation 84, 1420 –1425. [editorial] Faggiotto, A., Poli, A., and Catapano, A. L. (1998) Antioxidants and coronary artery disease. Curr. Opin. Lipidol. 9, 541–549. Princen, H. M., van Duyvenvoorde, W., Buytenhek, R., Blonk, C., Tijburg, L. B., Langius, J. A., Meinders, A. E., and Pijl, H. (1998) No effect of consumption of green and black tea on plasma lipid and antioxidant levels and on LDL oxidation in smokers. Arterioscler. Thromb. Vasc. Biol. 18, 833– 841. Fuhrman, B., Lavy, A., and Aviram, M. (1995) Consumption of red wine with meals reduces the susceptibility of human plasma and low-density lipoprotein to lipid peroxidation. Am. J. Clin. Nutr. 61, 549 –554. [see comments] Chopra, M., Fitzsimons, P. E., Strain, J. J., Thurnham, D. I., and Howard, A. N. (2000) Nonalcoholic red wine extract and quercetin inhibit LDL oxidation without affecting plasma antioxidant vitamin and carotenoid concentrations. Clin. Chem. 46, 1162– 1170. Lairon, D., and Amiot, M. J. (1999) Flavonoids in food and natural antioxidants in wine. Curr. Opin. Lipidol. 10, 23–28. Pearson, D. A., Tan, C. H., German, J. B., Davis, P. A., and Gershwin, M. E. (1999) Apple juice inhibits human low density lipoprotein oxidation. Life Sci. 64, 1913–1920. Boulton, D. W., Walle, U. K., and Walle, T. (1998) Extensive binding of the bioflavonoid quercetin to human plasma proteins. J. Pharm. Pharmacol. 50, 243–249. Berlett, B. S., and Stadtman, E. R. (1997) Protein oxidation in aging, disease, and oxidative stress. J. Biol. Chem. 272, 20313– 20316. Starke, P. E., Oliver, C. N., and Stadtman, E. R. (1987) Modification of hepatic proteins in rats exposed to high oxygen concentration. FASEB J. 1, 36 –39. Starke-Reed, P. E., and Oliver, C. N. (1989) Protein oxidation and proteolysis during aging and oxidative stress. Arch. Biochem. Biophys. 275, 559 –567. Davies, M. J., Fu, S., Wang, H., and Dean, R. T. (1999) Stable markers of oxidant damage to proteins and their application in the study of human disease. Free Rad. Biol. Med. 27, 1151–1163. Kalb, E., Paltauf, F., and Hermetter, A. (1989) Fluorescence lifetime distributions of diphenylhexatriene-labeled phosphatidylcholine as a tool for the study of phospholipid-cholesterol interactions. Biophys. J. 56, 1245–1253. Schumaker, V. N., and Puppione, D. L. (1986) Sequential flotation ultracentrifugation. Methods Enzymol. 128, 155–170.

FLUORESCENCE SCREENING OF SERUM ANTIOXIDANT CAPACITY 25. Singleton, V. L., and Rossi, J. A. (1965) Colorimetry of total phenolics with phosphomolybdic-phosphotungstic acid reagent. Am. J. Enol. Vitic. 16, 144 –158. 26. Rice-Evans, C., and Miller, N. J. (1994) Total antioxidant status in plasma and body fluids. Methods Enzymol. 234, 279 –293. 27. Sommer, A., Prenner, E., Gorges, R., Stutz, H., Grillhofer, H., Kostner, G. M., Paltauf, F., and Hermetter, A. (1992) Organization of phosphatidylcholine and sphingomyelin in the surface monolayer of low density lipoprotein and lipoprotein(a) as determined by time-resolved fluorometry. J. Biol. Chem. 267, 24217–24222. 28. Minetti, M., Mallozzi, C., Di Stasi, A. M., and Pietraforte, D. (1998) Bilirubin is an effective antioxidant of peroxynitrite-mediated protein oxidation in human blood plasma. Arch. Biochem. Biophys. 352, 165–174. 29. Dean, R. T., Hunt, J. V., Grant, A. J., Yamamoto, Y., and Niki, E. (1991) Free radical damage to proteins: The influence of the relative localization of radical generation, antioxidants, and target proteins. Free Rad. Biol. Med. 11, 161–168. 30. Hofer, G., Lichtenberg, D., and Hermetter, A. (1995) A new fluorescence method for the continuous determination of surface lipid oxidation in lipoproteins and plasma. Free Rad. Res. 23, 317–327. 31. Kostner, G. M., and Laggner, P. (1989) in Human Plasma Liproteins (Fruchart, J. C., and Shepherd, J., Eds.), pp. 23–54, Walter de Gruyter, Berlin.

153

32. Esterbauer, H., Striegl, G., Puhl, H., and Rotheneder, M. (1989) Continuous monitoring of in vitro oxidation of human low density lipoprotein. Free Rad. Res. Commun. 6, 67–75. 33. Heinecke, J. W. (1999) Is lipid peroxidation relevant to atherogenesis. J. Clin. Invest. 104, 135–136. [comment] 34. Niki, E., Yamamoto, Y., Komuro, E., and Sato, K. (1991) Membrane damage due to lipid oxidation. Am. J. Clin. Nutr. 53, 201S–205S. 35. Manach, C., Morand, C., Crespy, V., Demigne, C., Texier, O., Regerat, F., and Remesy, C. (1998) Quercetin is recovered in human plasma as conjugated derivatives which retain antioxidant properties. FEBS Lett. 426, 331–336. 36. Morand, C., Crespy, V., Manach, C., Besson, C., Demigne, C., and Remesy, C. (1998) Plasma metabolites of quercetin and their antioxidant properties. Am. J. Physiol. 275, R212–R219. 37. Hollman, P. C., van Trijp, J. M., Mengelers, M. J., de Vries, J. H., and Katan, M. B. (1997) Bioavailability of the dietary antioxidant flavonol quercetin in man. Cancer Lett. 114, 139 – 140. 38. Young, J. F., Nielsen, S. E., Haraldsdottir, J., Daneshvar, B., Lauridsen, S. T., Knuthsen, P., Crozier, A., Sandstrom, B., and Dragsted, L. O. (1999) Effect of fruit juice intake on urinary quercetin excretion and biomarkers of antioxidative status. Am. J. Clin. Nutr. 69, 87–94.