Original Contribution

INFLUENCE OF VITAMIN E AND C SUPPLEMENTATION ONLIPOPROTEIN OXIDATION IN PATIENTS WITH ALZHEIMER’S DISEASE

ANATOL KONTUSH,* ULRIKE MANN,† SONKE ARLT,* A MAAR UJEYL,* CHARLOTTE LUHRS,*TOMAS MULLER-THOMSEN,† and ULRIKE BEISIEGEL*

*Medical Clinic and†Psychiatric Clinic, University Hospital Eppendorf, Hamburg, Germany

(Received15 February2001;Accepted1 May 2001)

Abstract—Because increased oxidation is an important feature of Alzheimer’s disease (AD) and low concentrations ofantioxidant vitamins C and E have been observed in cerebrospinal fluid (CSF) of AD patients, supplementation withthese antioxidants might delay the development of AD. Major targets for oxidation in brain are lipids and lipoproteins.We studied whether supplementation with antioxidative vitamins E and C can increase their concentrations not only inplasma but also in CSF, and as a consequence decrease the susceptibility of lipoproteins to in vitro oxidation. Twogroups, each consisting of 10 patients with AD, were for 1 month supplemented daily with either a combination of 400IU vitamin E and 1000 mg vitamin C, or 400 IU vitamin E alone. We found that supplementation with vitamin E andC significantly increased the concentrations of both vitamins in plasma and CSF. Importantly, the abnormally lowconcentrations of vitamin C were returned to normal level following treatment. As a consequence, susceptibility of CSFand plasma lipoproteins to in vitro oxidation was significantly decreased. In contrast, the supplementation with vitaminE alone significantly increased its CSF and plasma concentrations, but was unable to decrease the lipoproteinoxidizability. These findings document a superiority of a combined vitamin E1 C supplementation over a vitamin Esupplementation alone in AD and provide a biochemical basis for its use. © 2001 Elsevier Science Inc.

Keywords—Alzheimer’s disease, Oxidation, Vitamin E, Vitamin C, Lipoproteins, Cerebrospinal fluid, Plasma, Anti-oxidants, Free radicals

INTRODUCTION

Alzheimer’s disease (AD) is a neurological disorder withincreasing prevalence in the Western world. The etiologyof sporadic AD, the most common form of this disease,is multifactorial [1]. Pathological oxidation has beenproposed to be among the most important factors in itspathogenesis [2–4]. Brain is especially vulnerable tooxidative stress as a result of its high oxygen consump-tion as well as high concentrations of easily oxidizablelipids and transition metal ions, which are capable ofproducing reactive oxygen species [5,6]. Consistent withthis, brain tissue from AD patients has been shown topossess higher levels of oxidized proteins [7], advancedglycation end products [8], 4-hydroxynonenal-derivedadducts [9], and products of lipid peroxidation [10,11],

than tissue from nondemented elderly controls. In addi-tion, the transition metals Cu(II) and Fe(III) are elevatedin AD brain [12,13].

These data imply that supplementation with antioxi-dants might delay development of AD. A first largeclinical trial did find a beneficial effect ofa-tocopheroland selegiline by slowing the progression of the disease[14]. a-Tocopherol (the main form of vitamin E) isconsidered as a major lipophilic antioxidant in man andis essential for the normal brain function [15]. It has beenshown to accumulate in the brain and thereby to decreaseperoxidation of brain lipids in animal models [16,17].The antioxidative efficiency ofa-tocopherol can be con-siderably increased by a co-supplementation with ascor-bate (vitamin C), which is a co-antioxidant for the former[18]. a-Tocopherol also reduces the neurotoxicity ofamyloid b peptide (Ab), a major component of senileplaques, in neuronal cell culture [19]. However, thedistinct molecular basis for the efficiency ofa-tocoph-erol supplementation in AD still needs to be elucidated.

Address correspondence to: Prof. Dr. Ulrike Beisiegel, Klinik fu¨rInnere Medizin, Universitaetsklinikum Hamburg Eppendorf, Martinis-traße 52, 20246 Hamburg, Germany; Tel:149 (40) 42803-3917; Fax:149 (40) 42803-4592; E-Mail: beisiegel@uke.uni-hamburg.de.

Free Radical Biology & Medicine, Vol. 31, No. 3, pp. 345–354, 2001Copyright © 2001 Elsevier Science Inc.Printed in the USA. All rights reserved

0891-5849/01/$–see front matter

PII S0891-5849(01)00595-0

345

The major targets for oxidation in brain include pro-teins and lipids. We have recently shown that oxidationof lipoproteins may also be important in AD [20,21].Lipoproteins in the density range of plasma high densitylipoproteins (HDL) have been found in human cerebro-spinal fluid (CSF) [22–24]. They contain polyunsaturatedfatty acids (PUFA), the major substrate for lipid peroxi-dation, as well as tocopherols, the major lipophilic anti-oxidants, and can be easily oxidized in vitro [25]. In AD,the chemical composition of CSF lipoproteins is changed[26] and they are more susceptible to in vitro oxidationthan lipoproteins from control subjects [20,21,27]. Inaddition, CSF also contains vitamin C, which is presentin about a 3-fold higher concentration than in plasma[21] and can efficiently recycle tocopherols in lipopro-teins [28]. Recently we have shown that CSF levels ofvitamin C are decreased in AD patients as compared tocontrols [21]. Importantly, CSF lipoproteins also carryAb [29,30] and could therefore be involved in the plaqueformation.

To elucidate molecular mechanisms of antioxidantaction in AD, we studied the effect of vitamins E and Csupplementation in AD patients. The concentrations ofboth vitamins in CSF and plasma were measured beforeand after supplementation as well as the susceptibility ofCSF and plasma lipoproteins to in vitro oxidation.

MATERIALS AND METHODS

Subjects

Twenty AD patients were recruited in the psychiatricclinic of Hamburg University Hospital. The patientswere all seen in the outpatient “memory clinic” anddiagnosed as having “clinically probable AD” accordingto the NINCDS-ADRDA criteria for primary degenera-tive dementia, Alzheimer type [31]. All AD patientswere in a mild to moderate stage of the disease, mobile,not hospitalized or in residential homes, and in a goodgeneral nutritional state, as demonstrated by a high nor-mal mean BMI of 25 kg/m2 (Table 1) and normal lipidconcentrations in plasma (Table 2) for both groups. Noneof the patients took antioxidant supplements. Informedconsent according to the declaration of Helsinki wasobtained before each lumbar puncture and the study wasapproved by the local Ethics Committee, Hamburg.

Vitamin supplementation

The patients were divided into two groups of 10. Thefirst group was supplemented with 400 IU vitamin E and1000 mg vitamin C per day for 1 month (vitamin E1C-supplemented group). The second group was supple-mented for 1 month with 400 IU/d vitamin E alone

(vitamin E-supplemented group). The selection of a doseof 400 IU per day instead of 2000 IU vitamin E, as usedby Sano et al. [14], was based on earlier studies in ourlaboratory in which we observed saturation of the plasmalevels of vitamin E at a dose of 400 IU per day (data notshown) and on the fact that high levels of tocopherol canfunction as pro-oxidant instead of its normal antioxidantaction [32].

Vitamin E (La Roche, Basel, Switzerland) was givenas a single dose in the morning together with a meal.Vitamin C (Jenapharm, Jena, Germany) was given astwo doses of 500 mg each in the morning and in theevening.

Sample collection and preservation

From each patient CSF and blood samples were takenat the same visit at both the beginning and end of thestudy. The CSF sampling was performed by lumbarpuncture between L3 und L4 in a horizontal positionafter 1 h resting in this position. One ml CSF and 10 mlethylenediaminetetraacetic acid blood were sampled andimmediately placed on ice. Blood was centrifuged at 4°Cfor 10 min at 2500 rpm to obtain plasma and the cellularbuffy coat for DNA preparation. CSF and plasma werefreshly frozen under argon or nitrogen at280°C, notlater than 30 min after puncture. The buffy coat wasstored at220°C. Samples were not stored longer than 6months. The samples were thawed at room temperaturedirectly before analysis. Patient compliance was ascer-tained by an intake-protocol filled in by the patientstogether with their caregivers.

Vitamin E, vitamin C, and other antioxidants inplasma and CSF

To ensure the efficiency of the vitamin supplementa-tion, CSF and plasma levels of vitamin E and vitamin Cwere measured.a-Tocopherol was determined as a majorform of vitamin E both in CSF and plasma. In addition,a- andb-carotene, and ubiquinol-10, were measured asother important lipophilic antioxidants in plasma. Be-cause lipid levels in CSF are about 100–500 times lowerthan in plasma, in CSF only the level ofa-tocopherolcould be determined. All lipophilic antioxidants werequantified by reversed-phase high-performance liquidchromatography (HPLC) with electrochemical detectionas described elsewhere [33], except that the system wascalibrated using an external standard method.

CSF and plasma vitamin C (ascorbate) was measuredphotometrically at 520 nm after its reaction with 2,6-dichlorphenolindophenol [34]. Urate, bilirubin, and totalsulfhydryl (SH)-groups were measured as other impor-

346 A. KONTUSH et al.

tant hydrophilic antioxidants in plasma. Urate was mea-sured using a commercially available enzymatic kit(Merck, Darmstadt, Germany). Bilirubin was measuredat 530 nm after its reaction with ethyl anthranilate indimethyl sulfoxide [35]. Total SH-groups were deter-mined spectrophotometrically at 412 nm after their reac-tion with dithionitrobenzene [36].

Plasma and CSF lipids

CSF fatty acids and cholesterol were measured bycapillary gas chromatography with flame ionization de-tection as described elsewhere [21], except that 300mlCSF was used for a single measurement. Heptadecanoicacid and 5a-cholestane were used as internal standardsfor fatty acids and cholesterol, respectively, whereasbutylated hydroxytoluene was added as antioxidant. CSFsaturated fatty acids (SFA) were calculated as a sum ofpalmitic and stearic acids, monounsaturated fatty acids(MUFA) were oleic and vaccenic acid and PUFA as asum of linoleic and arachidonic acids.

Plasma fatty acids were measured by capillary gaschromatography with flame ionization detection as de-scribed elsewhere [37]. Plasma total cholesterol, triglyc-erides, and HDL cholesterol were quantified by commer-cially available enzymatic kits (Boehringer Mannheim,Mannheim, Germany). Low-density lipoprotein (LDL)cholesterol was calculated using the Friedewald formula.

Plasma and CSF oxidation kinetics

Oxidation of CSF and plasma was monitored as achange in the sample absorbance at 234 nm. This param-eter has been shown to reflect the level of lipid hydroper-oxides in isolated LDL oxidized under in vitro conditions[38]. Lipid hydroxides, which have conjugated dienestructure, also specifically absorb at 234 nm. However,they make only a small percentage of hydroperoxidesformed during lipoprotein oxidation [39]. When lipopro-teins are oxidized in diluted plasma or CSF, change inthe absorbance at 234 nm correlates with other indices oflipid peroxidation, such as consumption of PUFAs andaccumulation of cholesterol linoleate hydroperoxide [25,40]. Furthermore, when oxidized plasma was treatedwith sodium borohydride to eliminate hydroperoxidesand then extracted with hexane, no increase in the ab-sorbance of the extract at 234 nm was found in compar-ison with unoxidized sample (data not shown). Thesedata justify using absorbance at 234 nm as a specificmeasure for the accumulation of lipid hydroperoxides inthe lipoproteins.

To register oxidation kinetics, CSF was diluted 10-fold with phosphate-buffered saline (PBS), containing

0.6 M NaCl, pH 7.4, treated with Chelex 100 ion-ex-change resin (Bio-Rad, Munich, Germany) for 1 h toremove transition metal ions. The samples were oxidizedat 37°C either in the absence (autoxidation) or in thepresence of the exogenous oxidant 2,29-azobis-(2-amidi-nopropane) hydrochloride (AAPH; Polysciences, Inc.,Warrington, PA, USA) at 100mM. The absorbance wasmeasured spectrophotometrically at 1 min intervals over50 h at 37°C in quartz cuvettes equipped with screw capssealed with a resin to avoid evaporation. In agreementwith our recent data [21,25] CSF oxidation kinetics re-vealed three consecutive phases, the lag, propagation,and plateau phases. The lag and propagation phases wereapproximated using two straight lines produced usinglinear regression analysis of both phases as describedbefore for the calculation of the lag phase of lipoproteinoxidation [21,25]. The abscissa at the intersection wasused as a lag phase duration.

Plasma was diluted 150-fold with PBS and incubatedat 37°C for 20 h in the absence of exogenous oxidants(autoxidation) or in the presence of AAPH (330mM)[41]. The absorbance was measured at 234 nm as de-scribed for CSF and the formation of conjugated dieneswas quantified by the mean oxidation rate during thelinear phase (between 50 and 360 min of incubation).

Plasma apolipoproteins

Plasma levels of apolipoproteins A-I and B weremeasured turbidimetrically using commercially availablekits. Plasma apolipoprotein E (apoE) was measured us-ing a sandwich ELISA developed in our laboratory(Mann, A., and Beisiegel, U., unpublished data). Amonoclonal antibody AK EE7 (DAKO, Hamburg, Ger-many) and polyclonal anti-apoE antibody [42] were em-ployed as a capture and detection antibody, respectively.The apoE genotype was determined using the restrictionisotyping method [43].

Statistical analysis

Differences between the two time-points within thesame group were analyzed by Wilcoxon’s matched pairstest. Differences in continuous variables between the twosupplementation groups at the same time-point wereanalyzed by Mann-Whitney U-test. Between-group dif-ferences in dichotomous variables were analyzed byFisher’s exact test. Pearson’s moment-product correla-tion coefficients were calculated to evaluate relationshipsbetween variables. All results are expressed as means6standard deviations. The quality of the assays was con-trolled by measuring the assay variability, which was nothigher than 8% for all the parameters measured [25,33,44].

347Vitamin supplementation and lipoprotein oxidation in AD

RESULTS

Characterization of patients

Clinical data of 20 AD patients recruited for the studyare given in Table 1. No parameter of potential interest inthe context of oxidative stress, such as smoking habits,presence of coronary heart disease, hypertension, anddiabetes, showed a significant difference between thevitamin E1 C- and vitamin E-supplemented groups. Thefrequency of the«4 allele of apolipoprotein E and themean Mini Mental Status Examination score were alsocomparable. The only exception was age, which wasslightly higher in the group supplemented with vitaminE 1 C. Accordingly, the mean age of onset of the diseasewas also higher in the vitamin E1 C-supplementedgroup reaching significance (Table 1).

Vitamin E, vitamin C, and other antioxidants inplasma and CSF

Vitamin E significantly increased in both the vita-min E 1 C- and vitamin E-supplemented groups, bothin CSF and plasma (Fig. 1). In plasma, supplementa-tion with vitamin E1 C raiseda-tocopherol level by35%, whereas supplementation with vitamin E re-sulted in a slightly higher increase by 45% (Fig. 1A).Interestingly, this relationship was reversed in CSF,wherea-tocopherol increased by 56% in the vitaminE 1 C-supplemented group but only by 23% in thevitamin E-supplemented group (Fig. 1B). Differencesin the vitamin E increase were not related to differentinitial levels of the vitamin, which were similar in bothgroups.

Vitamin C (ascorbate) significantly increased in thevitamin E1 C-supplemented group, both in plasma andCSF (Fig. 2). In plasma, ascorbate increased by 87% andin CSF by 17%. Importantly, the supplementation with

vitamin E 1 C was able to return vitamin C (which isdecreased in AD [21]) to its normal levels, both inplasma (normal level of 506 18 mM measured inage-matched control subjects in our laboratory [21]) andin CSF (normal level of 2006 28 mM measured in thecontrol subjects [21]).

Unexpectedly, a small but significant increase inplasma ascorbate of 14% was observed following thesupplementation with vitamin E alone (Fig. 2A). In CSF,ascorbate decreased insignificantly by 11% after the vi-tamin E supplementation (Fig. 2B). Similarly to vitaminE, differences in the vitamin C increase were not relatedto different initial levels of ascorbate, which were similarin both groups.

Of other antioxidants, plasmab-carotene signifi-cantly decreased from 0.396 0.25 to 0.316 0.17mM(p , .05,n 5 9) following vitamin E supplementation,and the supplementation with vitamin E1 C showeda not significant increase from 0.236 0.16 to 0.3060.24mM. No significant change in plasma SH-groups,bilirubin, urate, ubiquinol-10, anda-carotene was ob-served as a result of the supplementation with eithervitamin E or vitamin E1 C (data not shown).

Table 1. Study Population

Vitamin E 1C-supplemented

group(n 5 10)

Vitamin E-supplemented

group(n 5 10)

Age 69.96 7.1 64.56 8.2Sex (M/F) 2/8 6/4BMI (kg/m2) 24.86 2.9 24.86 2.5Smoking (Y/N) 2/8 2/8Coronary heart disease (Y/N) 0/10 1/9Hypertension (Y/N) 4/6 2/8Diabetes (Y/N) 0/10 1/9Mini Mental Status Examination

score18.96 4.2 20.36 4.4

Age of AD onset 67.16 7.6§ 60.06 8.3Apo E «4 allele frequency 0.45 0.40

§ p , .05 vs. vitamin E-supplemented group.

Fig. 1. a-Tocopherol in plasma (A) and CSF (B) of AD patientssupplemented with vitamin E1 C or vitamin E.a-Tocopherol wasmeasured using HPLC with electrochemical detection. *p , .05 and** p , .01 vs. corresponding value before supplementation.

348 A. KONTUSH et al.

Plasma and CSF lipids

All measured plasma lipids, such as total cholesterol,triglycerides, fatty acids (Table 2), LDL cholesterol, andHDL cholesterol (data not shown), were similar in bothpatient groups, both before and after the vitamin supple-mentation. Nor was any difference found when majorclasses of fatty acids (SFA, MUFA, and PUFA) wereexpressed as a percentage of the total.

Of plasma apolipoproteins, apoE was significantly

lower in the vitamin E1 C than in the vitamin E-sup-plemented group, at the beginning of the study (Table 2).No between- or within-group differences in plasma lev-els of apolipoproteins A-I and B were observed (data notshown).

In CSF, a significant increase in total cholesterol wasfound in the vitamin E1 C-supplemented group after thesupplementation (Table 3). No between- or within-groupdifferences in CSF levels of TFA and PUFA were found.MUFA were significantly lower and SFA significantlyhigher in the vitamin E1 C-supplemented group both atthe beginning and at the end of the study.

Plasma and CSF oxidation kinetics

Plasma oxidation kinetics observed in the presentstudy were in accordance with previously reported data[21]. The mean oxidation rate in AD samples was con-siderably higher than that measured in healthy subjects[41]. Supplementation with vitamin E and C significantlydecreased the rate of plasma autoxidation but had nosignificant effect on the rate of oxidation by AAPH (Fig.3). Supplementation with vitamin E alone did not influ-ence plasma oxidizability in vitro. It should be noted thatAD patients given vitamin E had significantly higherplasma rates of oxidation, both before and after thesupplementation, than patients given vitamin E and C(Fig. 3). Although we attempted to match the two patientgroups as much as possible (see Table 1), a completematching is impossible in practice, especially when awide array of parameters is under investigation, as wasthe case in our study. The between-group difference inthe basal plasma autoxidation rate could be related tosignificantly lower levels of urate, which is an importantantioxidant in human plasma, in the vitamin E- thanvitamin E1 C-supplemented group (2646 53 vs. 350643 mM, p , .01, at the beginning and 2756 33 vs.351 6 50 mM, p , .01, at the end of the study). Thisassumption was consistent with significant negative cor-

Table 2. Lipids in Plasma of AD Patients Supplemented with Vitamin E1 C or Vitamin E

Vitamin E 1 C-supplemented group(n 5 10)

Vitamin E-supplemented group(n 5 10)

Before After Before After

Cholesterol (mg/dl) 2306 21 2376 20 2206 51 2226 57Triglycerides (mg/dl) 1966 88 1676 76 1506 63 1546 86TFA (mg/dl) 5936 138 5646 125 6516 368 6776 366PUFA, %a 38.46 4.4 38.06 4.6 40.06 5.9 39.56 3.9MUFA, %a 27.46 3.4 27.66 3.6 25.86 3.3 26.76 2.9SFA, %a 34.16 1.5 34.46 1.9 34.26 3.0 33.76 2.1ApoE (mg/l) 676 25§ 766 30 1066 36 1086 49

a Weight percentage of TFA;§p , .05 vs. vitamin E-supplemented group.

Fig. 2. Ascorbate in plasma (A) and CSF (B) of AD patients supple-mented with vitamin E1 C or vitamin E. Ascorbate was measuredphotometrically at 520 nm after its reaction with 2,6-dichlorphenolin-dophenol. *p , .05 vs. corresponding value before supplementation.

349Vitamin supplementation and lipoprotein oxidation in AD

relations between urate and plasma oxidation rates (datanot shown).

CSF oxidation kinetics revealed three consecutivephases, the lag, propagation, and plateau phases (Fig. 4).This was in agreement with our recent data [21,25]. Thelag and propagation phases were approximated using twostraight lines produced using linear regression analysis ofthe both phases as described before for the calculation ofthe lag phase of lipoprotein oxidation [34,40,41]. The

abscissa at the intersection was used as a lag phaseduration. The nonlinear regression analysis of the wholeoxidation curve (its nonlinear approximation using twolinear functions) revealed very close values for the lagphase duration (r 5 0.93 for the correlation between thetwo methods obtained in a subset of 10 oxidation curves,data not shown) thereby justifying our approach.

The supplementation with vitamin E and C delayedCSF oxidation by AAPH, shifting the curve to the right(Fig. 4A). In contrast, the supplementation with vitamin

Fig. 3. Rate of autoxidation (A) and AAPH-induced oxidation (B) ofplasma of AD patients supplemented with vitamin E1 C or vitamin E.Plasma was diluted 150-fold with PBS and incubated at 37°C in theabsence of exogenous oxidants (autoxidation) and in the presence ofAAPH (330mM). *p , .01 vs. corresponding value before supplemen-tation; §§p , .01 vs. vitamin E1 C-supplemented group.

Fig. 4. Absorbance increase at 234 nm during AAPH-induced oxidationof CSF of AD patients supplemented with vitamin E1 C (A; n 5 10)or vitamin E (B; n 5 8). Oxidation kinetics were averaged for allsubjects in each group. Time points were taken every 5 min. CSF wasdiluted 10-fold with PBS and incubated at 37°C in the presence ofAAPH (100 mM).

Table 3. Lipids in CSF of AD Patients Supplemented with Vitamin E1 C or Vitamin E

Vitamin E 1 C-supplemented group(n 5 10)

Vitamin E-supplemented group(n 5 10)

Before After Before After

Cholesterol (mg/l) 3.326 1.26§ 3.906 1.07* 5.256 1.40 4.296 1.07TFA (mg/l) 6.286 2.81 6.396 2.73 5.276 0.99 5.836 2.60PUFA, %a 11.36 4.7 11.76 4.9 13.56 6.1 11.96 5.7MUFA, %a 37.16 4.7§ 33.26 7.5§ 40.86 4.8 42.16 5.9SFA, %a 51.66 5.3§ 55.06 9.9§ 45.66 3.1 46.06 4.6

a Weight percentage of TFA;§ p , .05 vs. vitamin E-supplemented group; *p , .05 vs. same group before vitaminsupplementation.

350 A. KONTUSH et al.

E alone was unable to significantly influence AAPH-induced CSF oxidation (Fig. 4B). When rates of CSFoxidation within the lag and propagation phase werecalculated, a significant inhibition of the oxidation byvitamin E 1 C, but not vitamin E alone, was alsoobserved (Fig. 5). In accordance with this, propagationphase was significantly prolonged by vitamin E1 C,whereas vitamin E alone was ineffective (Fig. 6). Similarbetween-group differences were observed, when CSFwas subjected to autoxidation; however, they did notreach significance (data not shown).

DISCUSSION

In this study we found that the supplementation of ADpatients with a combination of 400 IU vitamin E and1000 mg vitamin C for 1 month significantly increasedthe concentrations of both vitamins in CSF and plasma.Accordingly, the susceptibility of CSF and plasma li-poproteins to in vitro oxidation was significantly de-creased. In contrast, the 1 month supplementation with400 IU vitamin E alone significantly increased its CSFand plasma concentrations but was unable to decreasethe lipoprotein oxidizability.

We have recently shown that lipoproteins of humanCSF are oxidatively modified during its incubation at37°C [25] and that the in vitro oxidizability of both CSFand plasma samples from AD patients is significantlyhigher than in controls [20,21]. These data have beenconfirmed by others [27]. In parallel, CSF PUFA, themain substrate for lipid peroxidation, and antioxidantvitamins were found relatively reduced in AD, whichwas also in accordance with data published by others[26]. These results support the concept of oxidation as animportant factor in the pathogenesis of AD. This isconsistent with recent studies showing that AD braintissue has higher amounts of oxidatively modified bi-omolecules, as well as higher basal levels of lipid per-oxidation, than tissue from control subjects [7–11]. Im-portantly, the current concept of oxidation in AD isextended to include lipoproteins as a new target foroxidation. CSF lipoproteins are essential for the transportof lipids in brain [22,45]. Oxidation can impair theirnormal function and brain development and oxidizedlipoproteins can be toxic for neuronal cells [27,46].

In addition to the increase in lipoprotein oxidizability,a decrease in antioxidant levels in both CSF and plasmacould be demonstrated in AD. Ascorbate, the majorhydrophilic antioxidant, was significantly decreased anda-tocopherol, the major lipophilic antioxidant, tended tobe lower in CSF of AD patients [20,21]. This was animportant rationale for us to choose these two antioxi-dants for the supplementation.

Our present data indicate that the antioxidative effectsof the vitamin E1 C supplementation can compensatefor the increased lipoprotein oxidizability in AD, bring-ing it back to values measured in control subjects. How-ever, the supplementation with vitamin E alone wasinsufficient to achieve protection. These findings docu-ment a superiority of a combined vitamin E1 C over

Fig. 5. Rate of AAPH-induced oxidation in the lag (A) and propagation(B) phases measured in CSF of AD patients supplemented with vitaminE 1 C or vitamin E. CSF was diluted 10-fold with PBS and incubatedat 37°C in the presence of AAPH (100mM). * p , .05 vs. correspond-ing value before supplementation.

Fig. 6. Change in the duration of the lag and propagation phasesmeasured of AAPH-induced oxidation in CSF of AD patients supple-mented with vitamin E1 C or vitamin E. CSF was diluted 10-fold withPBS and incubated at 37°C in the presence of AAPH (100mM). * p ,.05 vs. vitamin E-supplemented group.

351Vitamin supplementation and lipoprotein oxidation in AD

vitamin E supplementation alone in AD and provide abiochemical basis for its use. However, in a processingdisease like AD, we would presuppose that longer dura-tion or higher dose of antioxidant supplementation maybe necessary for protective effects, such as those clini-cally seen in the study of Sano et al. [14].

Low levels of antioxidants in AD can be due to theirconsumption as a consequence of high levels of oxida-tive stress in vivo as well as to their insufficient supplyfrom diet. Regardless of this question, our data show thatsupplementation with therapeutically reasonable dosescan effectively increase vitamins E and C in both CSFand plasma of patients to the levels sufficient to decreaselipoprotein oxidizability in vitro.

The high oxygen consumption in the central nervoussystem implies a potentially increased production of ox-ygen radicals and there is a high need for antioxidativemolecules. However, only ascorbate is 3–5 times higher[47] in CSF, while all lipophilic molecules are around100–500 times less concentrated in CSF as compared toplasma. Therefore, ascorbate might be of special rele-vance in antioxidative protection of lipoproteins in CSF.

Ascorbate is well known as a major hydrophilic an-tioxidant in human CSF [21,47]. It is the antioxidant firstconsumed during plasma [48] and CSF [25] oxidationand builds the first line of their antioxidative defense. Weused an absorbance increase at 234 nm as a measure ofCSF oxidation. This parameter reflects oxidation of CSFlipoproteins [25]. We showed that the end of the lagphase of the oxidation of CSF lipoproteins correspondsto a complete loss of ascorbate, which, as long as it ispresent, completely protects them [25]. This protection isdue to the recycling of lipoprotein-locateda-toco-pheroxyl radicals, as well as direct scavenging of otherfree radicals [18]. Supplementation with ascorbate invitro is accordingly able to prolong the lag phase of CSFoxidation [25].a-Tocopherol is a major lipophilic anti-oxidant in humans. The major mechanism of its antioxi-dative action includes inactivation of one free radical byone molecule ofa-tocopherol with a subsequent scav-enging of a second radical by thea-tocopheroxyl radicalformed [32]. This mechanism is operative when freeradicals are formed at a relatively high rate, i.e., understrong oxidative conditions. However, the antioxidantactivity of a-tocopherol evolves into pro-oxidant undermild oxidative conditions. If no additional free radicalhits the lipoprotein particle for a certain time and inter-acts with thea-tocopheroxyl radical, the latter can di-rectly oxidize lipoprotein PUFA moieties. This schemeof thea-tocopherol-mediated peroxidation [32] suggeststhat the fate of thea-tocopheroxyl radical is crucial forthe neta-tocopherol activity towards oxidation.a-Toco-pheroxyl radicals must be efficiently eliminated from thelipoprotein particle to allowa-tocopherol to develop its

antioxidant activity. Elimination ofa-tocopheroxyl rad-ical by recycling it directly back intoa-tocopherol rep-resents an important mechanism of action of a widegroup of compounds calleda-tocopherol co-antioxidants[49]. Ascorbate seems to be physiologically the mostimportant amongst them because of its high concentra-tions in human plasma and, especially, CSF. Taken to-gether, these data explain the superiority of the combinedsupplementation with vitamins E and C over the supple-mentation with vitamin E alone. This conclusion is inaccordance with results obtained with plasma lipopro-teins, where a combined supplementation with vitamin Eand its co-antioxidants was considerably more effectivein inhibiting oxidation than a supplementation with vi-tamin E alone [50]. Another possible reason for the lackof an effect in the vitamin E-treated group might be thatthe age of onset was significantly lower in these patients.This could indicate a more severe stage of the disease andthereby explain the insufficient effect of the treatment.

In summary, our study demonstrates that the com-bined supplementation of AD patients with vitamins Eand C is able to increase their concentrations in CSF andplasma, return vitamin C to its normal levels, and therebydecrease in vitro oxidizability of CSF and plasma li-poproteins. These data on a protective effect of antioxi-dants in AD might provide a biochemical rationale forthe beneficial results in the recently published trial on theantioxidative treatment of AD patients [14] with vitaminE or selegiline. It remains to be shown whether supple-mentation with antioxidative vitamins alone or in com-bination can be sufficient to protect lipoproteins fromoxidation in vivo in long-term studies and whether thiscan slow down the development of AD.

Acknowledgements— We thank Dr. David Evans for the criticalreading of the manuscript. This study was performed in the frameworkof the Research Group “Molecular Pathomechanisms in Alzheimer’sDisease” and supported by the grant FOR 267/2-1 of the DeutscheForschungsgemeinschaft.

REFERENCES

[1] Hardy, J. Amyloid, the presenilins and Alzheimer’s disease.Trends Neurosci.20:154–159; 1997.

[2] Mark, R. J.; Blanc, E. M.; Mattson, M. P. Amyloid beta-peptideand oxidative cellular injury in Alzheimer’s disease.Mol. Neuro-biol. 12:211–224; 1996.

[3] Markesbery, W. R. Oxidative stress hypothesis in Alzheimer’sdisease.Free Radic. Biol. Med.23:134–147; 1997.

[4] Multhaup, G.; Ruppert, T.; Schlicksupp, A.; Hesse, L.; Beher, D.;Masters, C. L.; Beyreuther, K. Reactive oxygen species andAlzheimer’s disease.Biochem. Pharmacol.54:533–539; 1997.

[5] Spector, R. Vitamin homeostasis in the central nervous system.N. Engl. J. Med.296:1393–1398; 1977.

[6] Bush, A. I. Metals and neuroscience.Curr. Opin. Chem. Biol.4:184–191; 2000.

[7] Smith, C. D.; Carney, J. M.; Starke Reed, P. E.; Oliver, C. N.;Stadtman, E. R.; Floyd, R. A.; Markesbery, W. R. Excess brainprotein oxidation and enzyme dysfunction in normal aging and in

352 A. KONTUSH et al.

Alzheimer disease.Proc. Natl. Acad. Sci. USA88:10540–10543;1991.

[8] Smith, M. A.; Sayre, L. M.; Vitek, M. P.; Monnier, V. M.; Perry,G. Early AGEing and Alzheimer’s.Nature374:316; 1995.

[9] Montine, K. S.; Olson, S. J.; Amarnath, V.; Whetsell, W. O. Jr.;Graham, D. G.; Montine, T. J. Immunohistochemical detection of4-hydroxy-2-nonenal adducts in Alzheimer’s disease is associatedwith inheritance of APOE4.Am. J. Pathol.150:437–443; 1997.

[10] Lovell, M. A.; Ehmann, W. D.; Butler, S. M.; Markesbery, W. R.Elevated thiobarbituric acid-reactive substances and antioxidantenzyme activity in the brain in Alzheimer’s disease.Neurology45:1594–1601; 1995.

[11] Sayre, L. M.; Zelasko, D. A.; Harris, P. L.; Perry, G.; Salomon,R. G.; Smith, M. A. 4-Hydroxynonenal-derived advanced lipidperoxidation end products are increased in Alzheimer’s disease.J. Neurochem.68:2092–2097; 1997.

[12] Hershey, C. O.; Hershey, L. A.; Varnes, A.; Vibhakar, S. D.;Lavin, P.; Strain, W. H. Cerebrospinal fluid trace element contentin dementia: clinical, radiologic, and pathologic correlations.Neurology33:1350–1353; 1983.

[13] Dedman, D. J.; Treffry, A.; Candy, J. M.; Taylor, G. A.; Morris,C. M.; Bloxham, C. A.; Perry, R. H.; Edwardson, J. A.; Harrison,P. M. Iron and aluminum in relation to brain ferritin in normalindividuals and Alzheimer’s-disease and chronic renal-dialysispatients.Biochem. J.287:509–514; 1992.

[14] Sano, M.; Ernesto, C.; Thomas, R. G.; Klauber, M. R.; Schafer,K.; Grundman, M.; Woodbury, P.; Growdon, J.; Cotman, C. W.;Pfeiffer, E.; Schneider, L. S.; Thal, L. J. A controlled trial ofselegiline, alpha-tocopherol, or both as treatment for Alzheimer’sdisease.N. Engl. J. Med.336:1216–1222; 1997.

[15] Vatassery, G. T. Vitamin E and other endogenous antioxidants inthe central nervous system.Geriatrics 53(Suppl. 1):S25–S27;1998.

[16] Clement, M.; Bourre, J. M. Graded dietary levels of RRR-gamma-tocopherol induce a marked increase in the concentrations ofalpha- and gamma-tocopherol in nervous tissues, heart, liver andmuscle of vitamin-E-deficient rats.Biochim. Biophys. Acta1334:173–181; 1997.

[17] Meydani, M.; Macauley, J. B.; Blumberg, J. B. Effect of dietaryvitamin E and selenium on susceptibility of brain regions to lipidperoxidation.Lipids 23:405–409; 1988.

[18] Stocker, R. Lipoprotein oxidation: mechanistic aspects, method-ological approaches and clinical relevance.Curr. Opin. Lipidol.5:422–433; 1994.

[19] Behl, C.; Davis, J.; Cole, G. M.; Schubert, D. Vitamin E protectsnerve cells from amyloid beta protein toxicity.Biochem. Biophys.Res. Commun.186:944–950; 1992.

[20] Schippling, S.; Kontush, A.; Arlt, S.; Daher, D.; Buhmann, C.;Sturenburg, H. J.; Mann, U.; Mu¨ller-Thomsen, T.; Beisiegel, U.Lipoprotein oxidation and Alzheimer’s disease. In: Igbal, K.;Swaab, D. F.; Winblad, B.; Wisniewski, H. M., eds.Alzheimer’sdisease and related disorders.New York: Wiley; 1999:471–477.

[21] Schippling, S.; Kontush, A.; Arlt, S.; Buhmann, C.; Sturenburg,H.; Mann, U.; Muller-Thomsen, T.; Beisiegel, U. Increased li-poprotein oxidation in Alzheimer’s disease.Free Radic. Biol.Med.28:351–360; 2000.

[22] Pitas, R. E.; Boyles, J. K.; Lee, S. H.; Hui, D.; Weisgraber, K. H.Lipoproteins and their receptors in the central nervous system.J. Biol. Chem.262:14352–14360; 1987.

[23] Koch, S.; Beisiegel, U. Lipoproteins in the brain: a new frontier?In: Betteridge, D. J., ed.Lipids and vascular diseaseLondon:Martin Dunitz; 2000:51–64.

[24] Borghini, I.; Pometta, D.; James, R. W. Lipoprotein complexes inhuman cerebrospinal fluid.Schweiz. Arch. Neurol. Psychiatr.144:207–209; 1993.

[25] Arlt, S.; Finckh, B.; Beisiegel, U.; Kontush, A. Time-course ofoxidation of lipids in human cerebrospinal fluid in vitro.FreeRadic. Res.32:103–114; 2000.

[26] Montine, T. J.; Montine, K. S.; Swift, L. L. Central nervoussystem lipoproteins in Alzheimer’s disease.Am. J. Pathol.151:1571–1575; 1997.

[27] Bassett, C. N.; Neely, M. D.; Sidell, K. R.; Markesbery, W. R.;Swift, L. L.; Montine, T. J. Cerebrospinal fluid lipoproteins aremore vulnerable to oxidation in Alzheimer’s disease and areneurotoxic when oxidized ex vivo.Lipids 34:1273–1280; 1999.

[28] Thomas, S. R.; Neuzil, J.; Mohr, D.; Stocker, R. Coantioxidantsmake alpha-tocopherol an efficient antioxidant for low-densitylipoprotein.Am. J. Clin. Nutr.62:1357S–1364S; 1995.

[29] Koudinov, A. R.; Koudinova, N. V.; Kumar, A.; Beavis, R. C.;Ghiso, J. Biochemical characterization of Alzheimer’s solubleamyloid beta protein in human cerebrospinal fluid: associationwith high density lipoproteins.Biochem. Biophys. Res. Commun.223:592–597; 1996.

[30] Ladu, M. J.; Reardon, C.; Van Eldik, L.; Fagan, A. M.; Bu, G.;Holtzman, D.; Getz, G. S. Lipoproteins in the central nervoussystem.Ann. N.Y. Acad. Sci.903:167–175; 2000.

[31] McKhann, G.; Drachman, D.; Folstein, M.; Katzman, R.; Price,D.; Stadlan, E. M. Clinical diagnosis of Alzheimer’s disease:report of the NINCDS-ADRDA Work Group under the auspicesof Department of Health and Human Services Task Force onAlzheimer’s Disease.Neurology34:939–944; 1984.

[32] Bowry, V. W.; Stocker, R. Tocopherol-mediated peroxidation.The pro-oxidant effect of vitamin E on the radical-initiated oxi-dation of human low-density lipoprotein.J. Am. Chem. Soc.115:6029–6044; 1993.

[33] Finckh, B.; Kontush, A.; Commentz, J.; Hubner, C.; Burdelski,M.; Kohlschutter, A. High-performance liquid chromatography-coulometric electrochemical detection of ubiquinol-10, ubiqui-none-10, carotenoids, and tocopherols in neonatal plasma.Meth-ods Enzymol.299:341–348; 1999.

[34] Spranger, T.; Finckh, B.; Fingerhut, R.; Kohlschutter, A.; Beisie-gel, U.; Kontush, A. How different constituents of human plasmaand low density lipoprotein determine plasma oxidizability bycopper.Chem. Phys. Lipids91:39–52; 1998.

[35] Trotman, B. W.; Roy, C. J.; Wirt, G. D.; Bernstein, S. E. Azo-dipyrroles of unconjugated and conjugated bilirubin using diazo-tized ethyl anthranilate in dimethyl sulfoxide.Anal. Biochem.121:175–180; 1982.

[36] Motchnik, P. A.; Frei, B.; Ames, B. N. Measurement of antioxi-dants in human blood plasma.Methods Enzymol.234:269–279;1994.

[37] Kontush, A.; Meyer, S.; Finckh, B.; Kohlschutter, A.; Beisiegel,U. Alpha-tocopherol as a reductant for Cu(II) in human lipopro-teins.J. Biol. Chem.271:11106–11112; 1996.

[38] Esterbauer, H.; Striegl, G.; Puhl, H.; Rotheneder, M. Continuousmonitoring of in vitro oxidation of human low density lipoprotein.Free Radic. Res. Commun.6:67–75; 1989.

[39] Lenz, M. L.; Hughes, H.; Mitchell, J. R.; Via, D. P.; Guyton, J. R.;Taylor, A. A.; Gotto, A. M. J.; Smith, C. V. Lipid hydroperoxyand hydroxy derivatives in copper-catalyzed oxidation of lowdensity lipoprotein.J. Lipid Res.31:1043–1050; 1990.

[40] Regnstrom, J.; Strom, K.; Moldeus, P.; Nilsson, J. Analysis oflipoprotein diene formation in human serum exposed to copper.Free Radic. Res. Commun.19:267–278; 1993.

[41] Kontush, A.; Beisiegel, U. Measurement of oxidizability of bloodplasma.Methods Enzymol.299:35–49; 1999.

[42] Havekes, L. M.; de Knijff, K. P.; Beisiegel, U.; Havinga, J.; Smit,M.; Klasen, E. A rapid micromethod for apolipoprotein E pheno-typing directly in serum.J. Lipid Res.28:455–463; 1987.

[43] Hixson, J. E.; Vernier, D. T. Restriction isotyping of humanapolipoprotein E by gene amplification and cleavage with HhaI.J.Lipid Res.31:545–548; 1990.

[44] Kontush, A.; Spranger, T.; Reich, A.; Baum, K.; Beisiegel, U.Lipophilic antioxidants in blood plasma as markers of atheroscle-rosis: the role ofa-carotene andg-tocopherol.Atherosclerosis144:117–122; 1999.

[45] Fagan, A. M.; Bu, G.; Sun, Y.; Daugherty, A.; Holtzman, D. M.Apolipoprotein E-containing high density lipoprotein promotesneurite outgrowth and is a ligand for the low density lipoproteinreceptor-related protein.J. Biol. Chem.271:30121–30125; 1996.

[46] Draczynska-Lusiak, B.; Doung, A.; Sun, A. Y. Oxidized lipopro-

353Vitamin supplementation and lipoprotein oxidation in AD

teins may play a role in neuronal cell death in Alzheimer disease.Mol. Chem. Neuropathol.33:139–148; 1998.

[47] Lonnrot, K.; Metsa Ketela, T.; Molnar, G.; Ahonen, J. P.; Latvala,M.; Peltola, J.; Pietila, T.; Alho, H. The effect of ascorbate andubiquinone supplementation on plasma and CSF total antioxidantcapacity.Free Radic. Biol. Med.21:211–217; 1996.

[48] Frei, B.; England, L.; Ames, B. N. Ascorbate is an outstandingantioxidant in human blood plasma.Proc. Natl. Acad. Sci. USA86:6377–6381; 1989.

[49] Bowry, V. W.; Mohr, D.; Cleary, J.; Stocker, R. Prevention oftocopherol-mediated peroxidation in ubiquinol-10-free humanlow density lipoprotein.J. Biol. Chem.270:5756–5763; 1995.

[50] Thomas, S. R.; Neuzil, J.; Stocker, R. Cosupplementation withcoenzyme Q prevents the pro-oxidant effect of alpha-tocopheroland increases the resistance of LDL to transition metal-dependentoxidation initiation. Arterioscler. Thromb. Vasc. Biol.16:687–696; 1996.

ABBREVIATIONS

AD—Alzheimer’s diseaseAb—amyloid bAAPH—azobis-(2-amidinopropane) hydrochlorideCSF—cerebrospinal fluidHDL—high density lipoproteinsHPLC—high-performance liquid chromatographyLDL—low density lipoproteinMUFA—monounsaturated fatty acidsPBS—phosphate-buffered salinePUFA—polyunsaturated fatty acidsSFA—saturated fatty acidsTFA—total fatty acids

354 A. KONTUSH et al.