Pii: s0891-5849(99)00247-6

Free Radical Biology & Medicine, Vol. 28, No. 3, pp. 351–360, 2000 Copyright 2000 Elsevier Science Inc.
PII S0891-5849(99)00247-6
Original Contribution
INCREASED LIPOPROTEIN OXIDATION IN ALZHEIMER’S DISEASE SVEN SCHIPPLING,* ANATOL KONTUSH,* S¨ONKE ARLT,* CARSTEN BUHMANN,† HANS-J¨ORG ST¨URENBURG,† ULRIKE MANN,‡ TOMAS M¨ULLER-THOMSEN,‡ and ULRIKE BEISIEGEL* *Medical Clinic, †Neurological Clinic, and ‡Psychiatric Clinic, University Hospital Hamburg, Hamburg, Germany; (Received 19 October 1999; Accepted 16 November 1999) Abstract—Oxidation has been proposed to be an important factor in the pathogenesis of Alzheimer’s disease (AD) and
amyloid ␤ is considered to induce oxidation. In biological fluids, including cerebrospinal fluid (CSF), amyloid ␤ is found
complexed to lipoproteins. On the basis of these observations, we investigated the potential role of lipoprotein oxidation
in the pathology of AD. Lipoprotein oxidizability was measured in vitro in CSF and plasma from 29 AD patients and
found to be significantly increased in comparison to 29 nondemented controls. The levels of the hydrophilic antioxidant
ascorbate were significantly lower in CSF and plasma from AD patients. In plasma, ␣-carotene was significantly lower
in AD patients compared to controls while ␣-tocopherol levels were indistinguishable between patients and controls. In
CSF, a nonsignificant trend to lower ␣-tocopherol levels among AD patients was found. Polyunsaturated fatty acids, the
lipid substrate for oxidation, were significantly lower in the CSF of AD patients. Our findings suggest that (i) lipoprotein
oxidation may be important in the development of AD and (ii) the in vitro measurement of lipid peroxidation in CSF
might become a useful additional marker for diagnosis of AD.
Keywords—Alzheimer’s disease, Antioxidants, Cerebrospinal fluid, Free radicals, Lipid peroxidation, Lipoproteins,
Plasma
INTRODUCTION
[8,9]. Evidence for increased oxidative damage in ADincludes studies showing that brain tissue from AD pa- Alzheimer’s disease (AD) is a demential disorder with tients has higher levels of oxidized proteins [3], ad- increasing prevalence in the elderly population in the vanced glycation end products [10,11], and 4-hy- Western world. Senile plaques and neurofibrillary tan- droxynonenal-derived adducts [12,13] than tissue from gles are salient features in AD brains at autopsy and the nondemented elderly controls. Lipid peroxidation is in- histopathological hallmarks of clinical dementia. Apart creased in the brain in AD [14] and the transition metal from rare cases of early onset AD with causative muta- ions Cu(II) and Fe(III), capable of producing reactive tions in the amyloid precursor protein (APP) or preseni- oxygen species, have been shown to be elevated in AD lin (PS 1 and PS2) genes [1], the etiology of AD is brain tissue [15,16]. In addition, APP can reduce Cu(II) multifactorial [2]. In the framework of such a concept, to Cu(I), a highly reactive form [17].
several authors have proposed a pivotal role for oxida- The oxidation hypothesis is supported by the initial tion in the pathogenesis of AD in recent years [3–7]. The large clinical trial that proposed a beneficial effect of central nervous system is especially vulnerable to oxida- ␣-tocopherol and selegiline by slowing the progression tive stress as a result of the brain’s high oxygen con- of the disease [18]. Alpha-tocopherol is also known to be sumption, abundant lipid content, and relative paucity of effective against lipid peroxidation and to reduce the antioxidant compounds compared with other tissues neurotoxicity of amyloid ␤ (A␤), a major component ofsenile plaques [19]. The exact mechanisms responsible Address correspondence to: Dr. Anatol Kontush, Biochemisches for increased oxidation in AD brain remain unclear and Labor, Pav. 39, Medizinische Kern- und Poliklinik, Universita¨tskran- there is not yet enough evidence to decide whether oxi- kenhaus Eppendorf, Martinistraße 52, 20246 Hamburg, Germany; dation is a primary phenomenon inducing neurodegen- Tel: ϩ49 (40) 42803-4449; Fax: ϩ49 (40) 42803-4592; E-Mail:[email protected].
eration or is secondary to cell death and loss of neurons.
A␤ has been implicated as an oxidant involved in Sample collection and preservation the pathogenesis of AD [5,6,20]. Two facts make A␤potentially important for lipid peroxidation: soluble From each patient, 1 ml of CSF (obtained as a A␤ is found in biological fluids like cerebrospinal surplus of diagnostic lumbar puncture) and 10 ml of fluid (CSF), which is in direct contact with the brain ethylenediaminetetraacetic acid (EDTA)-anticoagu- [21], and in both CSF [22] and plasma [23], A␤ is lated blood were sampled at the same visit and imme- diately placed on ice. Blood was centrifuged at 4°C for Lipoproteins in the density range of plasma high- 10 min at 2500 rpm to obtain plasma and the cellular density lipoproteins (HDL) have been found in CSF buffy coat for DNA preparation. CSF and plasma were [24,25]. They contain polyunsaturated fatty acids freshly frozen under argon or nitrogen at Ϫ80°C, not (PUFA), the major substrate for lipid peroxidation, as later than 30 min after puncture. The buffy coat was well as lipophilic antioxidants such as tocopherols.
stored at Ϫ20°C. Samples were not stored longer than Plasma lipoproteins have been shown to be highly sus- 3 months. The samples were thawed at room temper- ceptible to oxidative modifications [26], a mechanism playing a crucial role in the pathogenesis of atheroscle-rosis. Changes in the chemical composition of CSF li- poproteins have been recently reported in AD [27]. Werecently found that lipoproteins of human CSF are easily Oxidation of CSF and plasma was monitored as a oxidized in vitro [28]. The increased oxidation damage in change in the sample absorbance at 234 nm. This param- the brain of AD patients led us to believe that lipoprotein eter has been shown to reflect the level of lipid hydroper- oxidation could be important in AD as it is in athero- oxides in isolated low-density lipoprotein (LDL) oxi- sclerosis. To assess this hypothesis, we analyzed plasma dized under in vitro conditions [30]. Lipid hydroxides are and CSF samples from 29 AD patients and 29 nonde- other products of LDL oxidation that have conjugated mented controls for their lipid content, the amount of diene structure and specifically absorb at 234 nm. How- lipophilic and hydrophilic antioxidants, and the oxidiz- ever, they comprise only a small percentage of hydroper- ability of the samples in vitro. We found that CSF oxides formed during lipoprotein oxidation [31]. When lipoproteins in AD patients were more susceptible to lipoproteins are oxidized in diluted plasma or CSF, oxidation than those from nondemented controls. The changes in the absorbance at 234 nm correlate with other increased lipoprotein oxidation in CSF can provide a indices of lipid peroxidation, such as consumption of useful additional marker in the clinical diagnosis of AD PUFAs and accumulation of cholesterol linoleate hy- and in particular it might allow a verification of the droperoxide [28,32]. Furthermore, when oxidized plasma was treated with sodium borohydride, to eliminate hy-droperoxides, and then extracted with hexane, no in-crease in the absorbance of the extract at 234 nm was MATERIALS AND METHODS
found in comparison with unoxidized samples (data not shown). These data justify the use of absorbance at 234nm as a specific measure for the accumulation of lipid AD patients (n ϭ 29) and control subjects (n ϭ 29) were recruited in the psychiatric clinic and the neurolog- To register oxidation kinetics, CSF was diluted 10- ical clinic of Hamburg University Hospital. The AD fold with phosphate-buffered saline (PBS), containing patients were all seen in the outpatient “memory clinic” 0.6 M NaCl, pH 7.4, treated with Chelex 100 ion-ex- and diagnosed as “clinically probable” according to the change resin (Bio-Rad, Munich, Germany) for 1 h to NINCDS-ADRDA and DSM-IV criteria for primary de- remove transition metal ions. The samples were oxidized generative dementia, Alzheimer type [29]. All AD pa- at 37°C either in the absence (autoxidation condition) or tients were in an early stage of the disease, mobile, in a in the presence of the exogenous oxidant 2,2Ј-azobis-(2- good general nutritional state, and did not take antioxi- amidinopropane) hydrochloride (AAPH; Polysciences, dant supplements. The control subjects attended the neu- Inc., Warrington, PA, USA) at 100 ␮M. The absorbance rological clinic and underwent lumbar puncture for di- was continuously registered spectrophotometrically at 5 agnostic purpose. Patients with degenerative disorders min intervals over 50 h at 37°C in quartz cuvettes tightly were excluded, as were all patients with clinically evi- sealed with Nescofilm to prevent evaporation.
dent cognitive impairment. Informed consent according Plasma was diluted 150-fold with PBS and incubated to the declaration of Helsinki was obtained before lum- at 37°C for 20 h in the absence of exogenous oxidants bar puncture and the study was approved by the Ethical (autoxidation condition) or in the presence of AAPH (330 ␮M) [33]. The absorbance was measured at 234 nm as described for CSF and the formation of conjugated (UV) detection at 267 nm [36]. One hundred microliters dienes was quantified by the mean oxidation rate during CSF were diluted 1:1 with 10% meta-phosphoric acid and centrifuged for 3 min at 13,000 rpm and 4°C. Onehundred microliters of the supernatant were injected intothe HPLC system using a solution of 0.1 M Na HPO , 2.5 mM EDTA, and 2.0 mM tetrahexyl ammonium chlo- CSF and plasma fatty acids and CSF cholesterol were ride, pH 3.0, as a mobile phase, running at 1.0 ml/min.
measured by capillary gas chromatography with flame Plasma ascorbate was measured photometrically as de- ionization detection [34]. One hundred microliters of CSF were mixed with 2 ml chloroform/methanol (2:1vol/vol), and 100 ␮l of heptadecanoic acid (200 mg/l) and 25 ␮l 5␣-cholestane (100 mg/l) were added as in-ternal standards; 25 ␮l butylated hydroxytoluene (BHT, The apolipoprotein (Apo) E genotype was determined 0.2 M) was added as antioxidant. The chloroform extract using the restriction isotyping method as described else- was evaporated under nitrogen, the dried lipids were dissolved in 250 ␮l toluene, and fatty acids were deri-vatized with 500 ␮l of 0.5 M anhydrous sodium methox- ide for 15 min at 50°C. The mixture was neutralized with1 ml 2.5% acetic acid and extracted with 250 ␮l hexane.
Between-group differences in continuous variables were The supernatant was evaporated under nitrogen and 100 analyzed by Student’s t-test for independent groups. Differ- ␮l dimethylformamide were added. Cholesterol was si- ences in dichotomous variables were analyzed by Fisher’s lylated by incubation with N,O-bis(trimethylsilyl)triflu- exact test. Pearson’s moment-product correlation coeffi- oroacetamide for 30 min at room temperature. After a cients were calculated to evaluate relationships between final evaporation, the pellet was dissolved in 20 ␮l tol- variables. Multiple regression was performed on oxidation uene, 2 ␮l of which were injected into a Hewlett-Packard parameters to elucidate the extent to which the oxidizability 5890 Series II gas chromatograph (Hewlett Packard, Palo was specifically influenced by the presence of the disease, Alto, CA, USA). CSF saturated fatty acids (SFA) were rather than by other independent factors. All results are calculated as the sum of palmitic, stearic, and arachidic expressed as means Ϯ standard deviations. The quality of acids; monounsaturated fatty acid (MUFA) was defined the assays was controlled by measuring the assay variabil- as oleic acid; and PUFA was defined as the sum of ity, which was not higher than 8% for all the parameters Plasma fatty acids were measured by capillary gas chromatography with flame ionization detection as de- scribed elsewhere [34]. Plasma cholesterol and triglyc-erides were quantified by commercially available enzy- matic kits (Boehringer Mannheim, Mannheim, Germany).
Clinical data of 29 AD patient and 29 controls are given in Table 1. The AD patients had a mean Mini Mental Status Examination score of 19 Ϯ 5. Due to thesignificantly higher mean age of the AD patients (72 vs.
Alpha-tocopherol, ␣- and ␤-carotene, ubiquinol-10, 55 years), we performed a subgroup analysis including and ubiquinone-10 were measured as the main lipophilic only individuals Ն 60 years to eliminate a possible antioxidants in plasma. For CSF, only the levels of age-related influence on the data. Twenty-six patients ␣-tocopherol and ␤-carotene were determined. The an- remained in the AD group and 14 in the age-matched tioxidant content was quantified by reversed-phase high- control subgroup. All data were analyzed for the total performance liquid chromatography (HPLC) with elec- group and the individual subgroups. The mean age of trochemical detection as described elsewhere [35], onset was 68 years in the total AD group and 70 years in except that the system was calibrated using an external the AD subgroup, and the mean time since diagnosis of the disease was 3.6 years in both groups.
As expected, the frequency of the ␧4 allele of Apo E was significantly higher in the AD patients, .36 vs.
.07 in controls. This frequency is in good accordance Ascorbate, the major hydrophilic antioxidant in CSF, with epidemiological data [39]. The between-group was measured by reversed-phase HPLC with ultraviolet comparison revealed a significantly higher rate of cur- * p Ͻ .01, † p Ͻ .05 vs. corresponding control group.
rent smokers among the controls. No other parameter ical fluids, in which lipids are present as lipoproteins. We of potential interest in the context of oxidative stress, previously developed a method to record lipoprotein such as presence of coronary heart disease (CHD), oxidation in human plasma via photometric measure- hypertension and diabetes or plasma lipids showed a ment of the kinetic of conjugated diene formation [33], a significant between-group difference (Table 1).
protocol that was subsequently adapted for the charac-terization of lipoprotein oxidation in CSF [28].
A typical time course of the absorbance at 234 nm exhibited three consecutive phases: the lag phase, during Given the difficulties in measuring oxidative stress in which the oxidation rate was close to zero; the propaga- vivo, we decided to assess lipid peroxidation in biolog- tion phase, representing rapid accumulation of lipid per-oxides; and the plateau phase with an oxidation rate closeto zero again (Fig. 1A). The absorption of oxidizing CSFwas found to parallel the time course of phosphatidyl-choline hydroperoxides accumulation measured byHPLC with UV detection (Fig. 1B) and was correlatedwith the consumption of antioxidants in the sample [28].
Typical oxidation kinetics from three AD patients and three controls are shown in Fig. 2. AD patients exhibitedan increased oxidation rate in the lag phase and a clearlyshorter lag phase duration, indicating a more rapid oxi-dation of CSF lipoprotein in AD patients compared withcontrols. To compare the kinetics for all subjects, themean oxidation rate during the initial phase and the lag Fig. 1. (A) Typical absorbance increase of CSF at 234 nm during itsautoxidation. Time points were taken every 5 min. CSF was diluted10-fold and incubated at 37°C in the sealed cuvettes. (B) Lag, propa-gation, and plateau phases of the oxidation. As in (A), but the accu- Fig. 2. Typical absorption increase at 234 nm of CSF obtained from mulation of lipid hydroperoxides was measured by absorption increase three representative AD patients (ᮀ) and three representative control at 234 nm (ᮀ) and HPLC (■); fewer time points were sampled, as subjects (●). CSF was diluted 10-fold and incubated at 37°C in the absence of exogenous oxidants to determine autoxidation.
Fig. 3. (A) Rate and (B) lag phase of autoxidation of CSF obtained from AD patients and control subjects. CSF was diluted 10-foldwith PBS and incubated at 37°C in the absence of exogenous oxidants. Bars correspond to the mean values calculated for each group.
Each open circle (E) and open square (ᮀ) corresponds to one subject older and one subject younger than 60 years, respectively.
Significance of the difference between the groups is shown as a p value.
phase duration were calculated for each curve. When There was no significant difference between the CSF both parameters were compared between all AD patients samples of patients with AD and controls in the levels of and all controls, the oxidizability of CSF from patients total protein, cholesterol, total fatty acids (TFA), and with AD was found to be significantly higher. The mean MUFA (Table 2). These data indicate that differences oxidation rate during the initial phase expressed as nano- observed in other parameters cannot be due to differ- moles of dienes liter per minute (nM/min), was signifi- ences in CSF volume. The relative level of PUFA in CSF cantly higher (p ϭ .0008; Fig. 3A) and the duration of the (expressed as a percentage of TFA) was significantly lag phase was significantly shorter (p ϭ .00003; Fig. 3B) lower in the patients with AD (p ϭ .001), whereas SFA in the AD group. The age of the patients had no influence were found to be relatively increased (p ϭ .02; Table 2).
on the sample oxidizability, so that the difference re- There was a negative correlation between the relative mained similarly significant in the age-matched sub- PUFA content and both the autoxidation rate and the groups. The differences were similarly pronounced when oxidation rate with AAPH (r ϭ Ϫ0.29, p ϭ .03; r ϭ AAPH, a chemical initiator of the oxidation, was added Ϫ0.43, p ϭ .001, respectively).
To complement these methods, the levels of hydro- Table 2. Protein, Lipids, and Antioxidants in CSF of AD Patients and Control Subjects a Weight percentage of TFA. * p Ͻ .001; † p Ͻ .01; ‡ p Ͻ .05 vs. corresponding control group.
philic and lipophilic antioxidants were measured in the in patients without AD [33]. When the mean initial same samples (Table 2). Ascorbate, the major hydro- oxidation rates were compared, the rates in AD plasma philic antioxidant in CSF [40], was significantly lower in samples were significantly higher. This was the case for the both the total group and the age-matched AD sub- both autoxidation and oxidation by AAPH, which in- group compared to controls (p ϭ .006 and .007, respec- creased the oxidation rates in both groups in parallel tively). Additionally, as expected, ascorbate levels were negatively correlated with CSF autoxidation rate (r ϭ When plasma lipids of patients with AD and controls Ϫ0.32, p ϭ .02). Alpha-tocopherol, the major lipophilic were compared, neither triglycerides and total choles- antioxidant in lipoproteins, tended to decrease in AD terol (Table 1), nor total fatty acids, saturated, monoun- patients’ CSF but this difference did not reach statistical saturated, and polyunsaturated fatty acids (data not significance. The difference in ␤-carotene values was The levels of antioxidants in plasma were in good accordance with the results of the oxidation kinetics (Table 3). In the total AD group, the hydrophilic antiox- The plasma oxidation kinetics observed in the present idant ascorbate was significantly decreased (p ϭ .02), but study were in accordance with previously reported data was only slightly lower in the age-matched AD subgroup Fig. 4. Initial oxidation rate of plasma obtained from AD patients and control subjects. Plasma was diluted 150-fold with PBS andincubated at 37°C in the absence of exogenous oxidants (autoxidation) and in the presence of AAPH (330 ␮M). Bars correspond withthe mean values calculated for each group. Each open circle (E) and open square (ᮀ) corresponds to one subject older and one subjectyounger than 60 years, respectively. Significance of the difference between the groups is shown as a p value.
Table 3. Antioxidants in Plasma of AD Patients and Control Subjects * p Ͻ .001; † p Ͻ .05 vs. corresponding control group.
(p ϭ .40). The only lipophilic antioxidant that showed DISCUSSION
significant difference between the groups was ␣-carotene(p ϭ .00001). There were no differences in the levels of Our study revealed a highly significant increase in ␣-tocopherol, ␤-carotene (Table 3), ubiquinol-10, and lipoprotein oxidizability in vitro for both CSF and ubiquinone-10 (data not shown). These results were con- plasma samples from AD patients compared with con- sistent with the age-matched analysis (Table 3) and the trols. In parallel, a decrease in antioxidant levels in both lipid-normalized values expressed as picomoles per mil- biological fluids was demonstrated. These data support ligram (pmol/mg) total cholesterol and triglycerides the concept of oxidation as an important factor in the (data not shown). The latter can be explained by the pathogenesis of AD and might provide a useful addi- comparable lipid levels in both groups. The plasma au- tional tool in the diagnosis of the disease.
toxidation rate was negatively correlated with the content Oxidation of plasma lipoproteins has been studied of ascorbate (r ϭ Ϫ0.29, p ϭ .05) and with the lipophilic quite extensively [26,41] and data obtained in plasma antioxidants ␣-tocopherol and ␣-carotene (r ϭ Ϫ0.37, samples of patients with CHD and hyperlipidemia versus Ϫ0.73; p ϭ .01, Ͻ .001, respectively), as was plasma controls demonstrated differences between these groups oxidation rate by AAPH with the levels of ascorbate and [33]. Human CSF contains lipoproteins with properties ␣-carotene (r ϭ Ϫ0.30 and Ϫ0.51; p ϭ .04 and Ͻ .001, similar to those of HDL from blood plasma [24,25,42].
CSF lipoproteins have, however, not yet been fully char-acterized, and no data on their oxidation are available.
We have recently shown that lipoproteins of human CSF Correlations and results of multiple regression are oxidatively modified during CSF incubation at 37°C[28]. Our present data show reproducible differences To elucidate a potential relationship between the ex- between AD patients and controls in the time course of tent of oxidizability of CSF and plasma, we examined the CSF oxidation in vitro. In addition, CSF PUFA, the main correlation between CSF and plasma oxidizability pa- substrate for lipid peroxidation, was found relatively rameters. The autoxidation rate in CSF correlated posi- reduced in AD, which was in accordance with data tively with plasma oxidation rates under both oxidizing published by others [27]. While the increased formation conditions (autoxidation: r ϭ 0.56, p Ͻ .001 and oxida- of conjugated dienes can also be shown in plasma, the tion by AAPH: r ϭ 0.49, p Ͻ .001). CSF lag phase relative reduction of PUFA cannot be demonstrated in duration also showed a negative correlation with plasma plasma samples. This observation can be due to the much autoxidation rate (r ϭ 0.45, p ϭ .001).
higher amount of fatty acids in plasma.
To detect a potential influence of factors other than It has been previously shown that the oxidizability of AD on the oxidation kinetics, multiple regression was plasma LDL is determined by three major factors: (i) performed on oxidation parameters as dependent vari- levels of antioxidants; (ii) levels of substrate for oxida- ables, using the categoric variables AD, sex, Apo E ␧4 tion, such as PUFA or other lipids; and (iii) levels of allele, current smoking, increased alcohol intake, pres- preformed oxidation products or other substances able to ence of CHD and hypertension, and continuous vari- accelerate LDL oxidation [43,44]. Increased oxidizabil- able age as independent variables. The presence of AD ity of plasma and CSF lipoproteins in AD might there- significantly influenced every oxidation parameter in fore be theoretically related to (i) lower levels of anti- CSF and plasma. From the other parameters only sex oxidants, (ii) higher levels of substrate for oxidation, and entered the final regression equation once. The rela- (iii) presence of (yet unidentified) oxidants. We did find tively small number of patients did not, however, lower levels of antioxidants in CSF and plasma from AD allow a correlation between the severity of disease and patients. However, the CSF level of oxidation substrate (PUFA) was significantly lower in AD. This finding could be due to elevated oxidation of CSF lipids in AD in accordance with recent studies showing that AD brain in vivo, i.e., to higher “preoxidation” of CSF samples tissue has higher amounts of oxidatively modified bi- used to measure the oxidizability. This phenomenon omolecules, as well as higher basal levels of lipid per- might be associated, in turn, with elevated levels of oxidation, than tissue from control subjects [3,10 –14].
oxidation products, or other substances able to accelerate Alternatively, low levels of antioxidants in AD may in vitro oxidation, in CSF from AD patients. The finding result from insufficient dietary supply. However, this is that CSF oxidizability was increased in AD despite lower unlikely to be the case in our study, since a subset of 10 PUFA levels, might therefore be explained by increased AD patients showed normal plasma levels of vitamin B12 levels of these, yet unidentified, oxidants.
and normal body mass index values (data not shown). In We have previously reported that diluted CSF can be addition, we recruited AD patients who were in the early oxidized without adding exogenous oxidants, i.e., auto- stage of disease, as indicated by the short time since its catalytically [28]. The autoxidation of CSF was com- diagnosis. This suggests the patients with AD had a pletely inhibited by EDTA, indicating that it was cata- lyzed by transition metal ions, such as Cu(II) and/or Because APP, A␤, and senile plaques, the main mark- Fe(III). These ions are present in native CSF as redox- ers for AD, have been shown to cause increased oxida- inactive complexes with the metal-binding proteins cer- tive stress [5,6,17,56], oxidation seems to be secondary uloplasmin and transferrin [45] but can be released, when to the formation of A␤ and senile plaques, thus promot- intact protein structure is disturbed under some patho- ing the course of the disease rather than being causal.
logic conditions, which may occur in vivo [46,47]. Pro-longed incubation in vitro at 37°C is likely to disturb the Oxidation can induce protein dysfunction and cell death, structure of metal-binding proteins, enabling release of and neuronal cell death contributes to further oxidation.
Cu(II) and Fe(III) in a catalytically active form. Cu(II) Therefore, oxidation might be both a cause and conse- and Fe(III) [15,16] as well as their carriers ceruloplas- quence of the degenerative processes in AD brains.
min, transferrin, and p97 [45,48-50] have all been shown The correlation found between plasma oxidation rates to be elevated in AD. These data implicate transition and antioxidant levels underlines the value of the in vitro metal ions as potential oxidants for CSF lipoproteins in measurement of oxidation as a marker for oxidative stress in vivo. The multiple regression analysis showed Reduction of catalytically active protein-bound tran- that the increased oxidation levels were specifically cor- sition metal ions must take place to initiate oxidation, related with the presence of the disease rather than with and reductants, such as ascorbate or tocopherol, are able other factors. A comparison of CSF oxidation param- to mediate reduction [34,53]. However, it has been eters between AD patients and patients with other shown that this particular pro-oxidative activity of ascor- degenerative neurological diseases, such as Parkin- bate does not results in its total pro-oxidative action on son’s disease and amyotrophic lateral sclerosis, is lipoprotein oxidation; the well-known radical-scaveng- ing action of ascorbate appears to prevail and ascorbate In conclusion, it should be noted that AD, a multifac- remains an antioxidant even in the presence of redox- torial disease, is neither genetically fully elucidated, nor active transition metal ions [54,55]. The lower levels of are all the factors influencing its pathogenesis known.
ascorbate we measured in the CSF of AD patients are The present study indicates that increased lipoprotein therefore in accordance with higher CSF oxidizability.
oxidation in CSF may be an additional important factor The high oxygen consumption in the central nervous in the progression of the disease. The recently published system implies a potentially increased production of ox- trial on the antioxidative treatment of patients with AD, ygen radicals and there is a elevated requirement for proposing a positive effect on the course of the disease, antioxidative molecules. However, ascorbate alone is supports this hypothesis [18]. In vitro measurements of three to five times higher in CSF, while all lipophilic lipoprotein oxidation in CSF might be useful as an ad- molecules are around 100 –500 times less concentrated in ditional tool for the clinical diagnosis of AD. They can CSF as compared to plasma. Therefore, ascorbate might also be of particular interest as biochemical markers for be of special relevance in antioxidative protection of the evaluation of antioxidant treatment of AD patients.
lipoproteins in CSF. Antioxidant levels could be ex-pected to be decreased in the CSF of AD patients as aconsequence of high levels of oxidative stress in vivo.
Acknowledgements — We thank Dr. David Evans for the critical read- Our finding that ascorbate is significantly decreased and ing of the manuscript and Diana Daher for accurate measurement of lipophilic antioxidants are slightly decreased among AD some of the samples. This study was performed in the framework of theResearch Group “Molecular Pathomechanisms in Alzheimer’s Dis- patients therefore strongly suggests that oxidation is a ease,” which was initiated by Prof. Roger Nitsch and is supported by current feature of AD pathology in vivo. This finding is the Grant Ni 486/2-1 of the Deutsche Forschungsgemeinschaft.
REFERENCES
lation and quantification of soluble Alzheimer’s beta-peptide from [1] Haass, C. Presenilin because of presenilin: the presenilin genes biological fluids. Nature 359:325–327; 1992.
and early onset Alzheimer’s disease. Curr. Opin. Neurol. 9:254 –
[22] Koudinov, A. R.; Koudinova, N. V.; Kumar, A.; Beavis, R. C.; Ghiso, J. Biochemical characterization of Alzheimer’s soluble [2] Hardy, J. Amyloid, the presenilins and Alzheimer’s disease.
amyloid beta protein in human cerebrospinal fluid: association Trends. Neurosci. 20:154 –159; 1997.
with high density lipoproteins. Biochem. Biophys. Res. Commun. [3] Smith, C. D.; Carney, J. M.; Starke Reed, P. E.; Oliver, C. N.; 223:592–597; 1996.
Stadtman, E. R.; Floyd, R. A.; Markesbery, W. R. Excess brain [23] Koudinov, A.; Matsubara, E.; Frangione, B.; Ghiso, J. The soluble protein oxidation and enzyme dysfunction in normal aging and in form of Alzheimer’s amyloid beta protein is complexed to high Alzheimer disease. Proc. Natl. Acad. Sci. USA 88:10540 –10543;
density lipoprotein 3 and very high density lipoprotein in normal human plasma. Biochem. Biophys. Res. Commun. 205:1164 –
[4] Dyrks, T.; Dyrks, E.; Hartmann, T.; Masters, C.; Beyreuther, K.
Amyloidogenicity of beta A4 and beta A4-bearing amyloid pro- [24] Pitas, R. E.; Boyles, J. K.; Lee, S. H.; Hui, D.; Weisgraber, K. H.
tein precursor fragments by metal-catalyzed oxidation. J. Biol. Lipoproteins and their receptors in the central nervous system.
Chem. 267:18210 –18217; 1992.
J. Biol. Chem. 262:14352–14360; 1987.
[5] Mark, R. J.; Blanc, E. M.; Mattson, M. P. Amyloid beta-peptide [25] Montine, K. S.; Bassett, C. N.; Ou, J. J.; Markesbery, W. R.; and oxidative cellular injury in Alzheimer’s disease. Mol. Neuro- Swift, L. L.; Montine, T. J. Apolipoprotein E allelic influence on biol. 12:211–224; 1996.
human cerebrospinal fluid apolipoproteins. J. Lipid Res. 39:2443–
[6] Markesbery, W. R. Oxidative stress hypothesis in Alzheimer’s disease. Free Radic. Biol. Med. 23:134 –147; 1997.
[26] Steinberg, D.; Parthasarathy, S.; Carew, T. E.; Khoo, J. C.; [7] Multhaup, G.; Ruppert, T.; Schlicksupp, A.; Hesse, L.; Beher, D.; Witztum, J. L. Beyond cholesterol. N. Engl. J. Med. 320:915–
Masters, C. L.; Beyreuther, K. Reactive oxygen species and Alzheimer’s disease. Biochem. Pharmacol. 54:533–539; 1997.
[27] Montine, T. J.; Montine, K. S.; Swift, L. L. Central nervous [8] Spector, R. Vitamin homeostasis in the central nervous system.
system lipoproteins in Alzheimer’s disease. Am. J. Pathol. 151:
N. Engl. J. Med. 296:1393–1398; 1977.
[9] Vatassery, G. T.; Nelson, M. J.; Maletta, G. J.; Kuskowski, M. A.
[28] Arlt, S.; Finckh, B.; Beisiegel, U.; Kontush, A. Time-course of Vitamin E (tocopherols) in human cerebrospinal fluid. Am. J. oxidation of lipids of human cerebrospinal fluid. Free Radic. Res. Clin. Nutr. 53:95–99; 1991.
[10] Smith, M. A.; Richey, P. L.; Taneda, S.; Kutty, R. K.; Sayre, [29] McKhann, G.; Drachman, D.; Folstein, M.; Katzman, R.; Price, L. M.; Monnier, V. M.; Perry, G. Advanced Maillard reaction end D.; Stadlan, E. M. Clinical diagnosis of Alzheimer’s disease: products, free radicals, and protein oxidation in Alzheimer’s report of the NINCDS-ADRDA Work Group under the auspices disease. Ann. N.Y. Acad. Sci. 738:447– 454; 1994.
of Department of Health and Human Services Task Force on [11] Smith, M. A.; Sayre, L. M.; Vitek, M. P.; Monnier, V. M.; Perry, Alzheimer’s Disease. Neurology 34:939 –944; 1984.
G. Early ageing and Alzheimer’s. Nature 374:316; 1995.
[30] Esterbauer, H.; Striegl, G.; Puhl, H.; Rotheneder, M. Continuous [12] Montine, K. S.; Olson, S. J.; Amarnath, V.; Whetsell, W. O., Jr.; monitoring of in vitro oxidation of human low density lipoprotein.
Graham, D. G.; Montine, T. J. Immunohistochemical detection of Free Radic. Res. Commun. 6:67–75; 1989.
4-hydroxy-2-nonenal adducts in Alzheimer’s disease is associated [31] Lenz, M. L.; Hughes, H.; Mitchell, J. R.; Via, D. P.; Guyton, J. R.; with inheritance of APOE4. Am. J .Pathol. 150:437– 443; 1997.
Taylor, A. A.; Gotto, A. M. J.; Smith, C. V. Lipid hydroperoxy [13] Sayre, L. M.; Zelasko, D. A.; Harris, P. L.; Perry, G.; Salomon, and hydroxy derivatives in copper-catalyzed oxidation of low R. G.; Smith, M. A. 4-Hydroxynonenal-derived advanced lipid density lipoprotein. J. Lipid Res. 31:1043–1050; 1990.
peroxidation end products are increased in Alzheimer’s disease.
[32] Regnstrom, J.; Strom, K.; Moldeus, P.; Nilsson, J. Analysis of J. Neurochem. 68:2092–2097; 1997.
lipoprotein diene formation in human serum exposed to copper.
[14] Lovell, M. A.; Ehmann, W. D.; Butler, S. M.; Markesbery, W. R.
Free Radic. Res. Commun. 19:267–278; 1993.
Elevated thiobarbituric acid-reactive substances and antioxidant [33] Kontush, A.; Beisiegel, U. Measurement of oxidizability of blood enzyme activity in the brain in Alzheimer’s disease. Neurology plasma. Methods Enzymol. 299:35– 49; 1999.
45:1594 –1601; 1995.
[34] Kontush, A.; Meyer, S.; Finckh, B.; Kohlschutter, A.; Beisiegel, [15] Hershey, C. O.; Hershey, L. A.; Varnes, A.; Vibhakar, S. D.; U. Alpha-tocopherol as a reductant for Cu(II) in human lipopro- Lavin, P.; Strain, W. H. Cerebrospinal fluid trace element content teins. J. Biol. Chem. 271:11106 –11112; 1996.
in dementia: clinical, radiologic, and pathologic correlations.
[35] Finckh, B.; Kontush, A.; Commentz, J.; Hubner, C.; Burdelski, Neurology 33:1350 –1353; 1983.
M.; Kohlschutter, A. Sensitive HPLC techniques for the simulta- [16] Dedman, D. J.; Treffry, A.; Candy, J. M.; Taylor, G. A.; Morris, neous determination of tocopherols, tocotrienols, ubiquinols and C. M.; Bloxham, C. A.; Perry, R. H.; Edwardson, J. A.; Harrison, ubiquinones in biological samples. Methods Enzymol. 299:341–
P. M. Iron and aluminium in relation to brain ferritin in normal individuals and Alzheimer’s-disease and chronic renal-dialysis [36] Barja, G.; Hernanz, A. Vitamin C, dehydroascorbate, and uric patients. Biochem. J. 287:509 –514; 1992.
acid in tissues and serum: high-performance liquid chromatogra- [17] Multhaup, G.; Schlicksupp, A.; Hesse, L.; Beher, D.; Ruppert, T.; phy. Methods Enzymol. 234:331–337; 1994.
Masters, C. L.; Beyreuther, K. The amyloid precursor protein of [37] Spranger, T.; Finckh, B.; Fingerhut, R.; Kohlschutter, A.; Beisie- Alzheimer’s disease in the reduction of copper(II) to copper(I).
gel, U.; Kontush, A. How different constituents of human plasma Science 271:1406 –1409; 1996.
and low density lipoprotein determine plasma oxidizability by [18] Sano, M.; Ernesto, C.; Thomas, R. G.; Klauber, M. R.; Schafer, copper. Chem. Phys. Lipids 91:39 –52; 1998.
K.; Grundman, M.; Woodbury, P.; Growdon, J.; Cotman, C. W.; [38] Hixson, J. E.; Vernier, D. T. Restriction isotyping of human Pfeiffer, E.; Schneider, L. S.; Thal, L. J. A controlled trial of apolipoprotein E by gene amplification and cleavage with HhaI. J. selegiline, alpha-tocopherol, or both as treatment for Alzheimer’s Lipid Res. 31:545–548; 1990.
disease N. Engl. J. Med. 336:1216 –1222; 1997.
[39] Roses, A. D.; Saunders, A. M. APOE is a major susceptibility [19] Behl, C.; Davis, J.; Cole, G. M.; Schubert, D. Vitamin E protects gene for Alzheimer’s disease. Curr. Opin. Biotechnol. 5:663–
nerve cells from amyloid beta protein toxicity. Biochem. Biophys. Res. Commun. 186:944 –950; 1992.
[40] Lonnrot, K.; Metsa Ketela, T.; Molnar, G.; Ahonen, J. P.; Latvala, [20] Davis, J. B. Oxidative mechanisms in beta-amyloid cytotoxicity.
M.; Peltola, J.; Pietila, T.; Alho, H. The effect of ascorbate and Neurodegeneration. 5:441– 444; 1996.
ubiquinone supplementation on plasma and CSF total antioxidant [21] Seubert, P.; Vigo, P. C.; Esch, F.; Lee, M.; Dovey, H.; Davis, D.; capacity. Free Radic. Biol. Med. 21:211–217; 1996.
Sinha, S.; Schlossmacher, M.; Whaley, J.; Swindlehurst, C. Iso- [41] Heinecke, J. W. Mechanisms of oxidative damage of low density lipoprotein in human atherosclerosis. Curr. Opin. Lipidol. 8:268 –
[52] Sayre, L. M.; Perry, G.; Smith, M. A. Redox metals and neuro- degenerative disease. Curr. Opin. Chem. Biol. 3:220 –225; 1999.
[42] Borghini, I.; Barja, F.; Pometta, D.; James, R. W. Characterization [53] Lynch, S. M.; Frei, B. Reduction of copper, but not iron, by of subpopulations of lipoprotein particles isolated from human human low density lipoprotein (LDL). J. Biol. Chem. 270:5158 –
cerebrospinal fluid. Biochim. Biophys. Acta 1255:192–200; 1995.
[43] Frei, B.; Gaziano, J. M. Content of antioxidants, preformed lipid [54] Berger, T. M.; Polidori, M. C.; Dabbagh, A.; Evans, P. J.; Halli- hydroperoxides, and cholesterol as predictors of the susceptibility well, B.; Morrow, J. D.; Roberts, L. J.; Frei, B. Antioxidant of human LDL to metal ion-dependent and -independent oxida- activity of vitamin C in iron-overloaded human plasma. J. Biol. tion. J. Lipid Res. 34:2135–2145; 1993.
Chem. 272:15656 –15660; 1997.
[44] Kontush, A.; Hubner, C.; Finckh, B.; Kohlschutter, A.; Beisiegel, [55] Kontush, A.; Finckh, B.; Karten, B.; Kohlschutter, A.; Beisiegel, U. How different constituents of low density lipoprotein deter- U. Antioxidant and prooxidant activity of alpha-tocopherol in mine its oxidizability by copper: a correlational approach. Free human plasma and low density lipoprotein. J. Lipid Res. 37:
Radic. Res. 24:135–147; 1996.
[45] Loeffler, D. A.; DeMaggio, A. J.; Juneau, P. L.; Brickman, C. M.; [56] Subbarao, K. V.; Richardson, J. S.; Ang, L. C. Autopsy samples Mashour, G. A.; Finkelman, J. H.; Pomara, N.; LeWitt, P. A.
of Alzheimer’s cortex show increased peroxidation in vitro.
Ceruloplasmin is increased in cerebrospinal fluid in Alzheimer’s J. Neurochem. 55:342–345; 1990.
disease but not Parkinson’s disease. Alzheimer. Dis. Assoc. Dis-
ord. 8:190 –197; 1994.
[46] Lamb, D. J.; Leake, D. S. Iron released from transferrin at acidic ABBREVIATIONS
pH can catalyse the oxidation of low density lipoprotein. FEBS
Lett. 352:15–18; 1994.
[47] Mukhopadhyay, C. K.; Ehrenwald, E.; Fox, P. L. Ceruloplasmin enhances smooth muscle cell- and endothelial cell-mediated lowdensity lipoprotein oxidation by a superoxide-dependent mecha- nism. J. Biol. Chem. 271:14773–14778; 1996.
AAPH—2,2Ј-azobis-(2-amidinopropane) hydrochloride [48] Kennard, M. L.; Feldman, H.; Yamada, T.; Jefferies, W. A. Serum levels of the iron binding protein p97 are elevated in Alzheimer’s
disease. Nat. Med. 2:1230 –1235; 1996.
[49] Loeffler, D. A.; Connor, J. R.; Juneau, P. L.; Snyder, B. S.; Kanaley, L.; DeMaggio, A. J.; Nguyen, H.; Brickman, C. M.; LeWitt, P. A. Transferrin and iron in normal, Alzheimer’s disease,
and Parkinson’s disease brain regions. J. Neurochem. 65:710 –
HPLC— high-performance liquid chromatography [50] Loeffler, D. A.; LeWitt, P. A.; Juneau, P. L.; Sima, A. A.; Nguyen, H. U.; DeMaggio, A. J.; Brickman, C. M.; Brewer, G. J.;Dick, R. D.; Troyer, M. D.; Kanaley, L. Increased regional brain concentrations of ceruloplasmin in neurodegenerative disorders.
Brain Res. 738:265–274; 1996.
[51] Smith, M. A.; Harris, P. L.; Sayre, L. M.; Perry, G. Iron accu- mulation in Alzheimer disease is a source of redox-generated free radicals. Proc. Natl. Acad. Sci. USA 94:9866 –9868; 1997.

Source: http://www.biotech-europe.de/editorials/ed424/beisiegel/Increasedlipoprotein_oxidation.pdf