ApoE plays an important role in cardiovascular homeostasis and is a recognised mediator of many metabolic processes underlying atherogenesis and thrombosis. This 34 kDa protein, mainly produced by the liver, was first identified as a regulator of lipoprotein metabolism being involved in hepatic lipoprotein secretion, lipoprotein metabolism in the circulation and serving as a high-affinity ligand for cellular lipoprotein uptake. apoE is also secreted by many other tissues, and extra-hepatic apoE has been shown to protect against atherosclerosis independently of its effects on lipid metabolism (Yamada et al. Reference Yamada, Inoue and Kawamura1992; Shimano et al. Reference Shimano, Ohsuga, Shimada, Namba, Gotoda, Harada, Katsuki, Yazaki and Yamada1995; Fazio et al. Reference Fazio, Babaev, Murray, Hasty, Carter, Gleaves, Atkinson and Linton1997; Hasty et al. Reference Hasty, Linton, Brandt, Babaev, Gleaves and Fazio1999). The protein is abundant in the atherosclerotic lesions, where resident macrophages are its principle source. The role of this locally produced apoE is not fully understood, but some identified effects include modulation of immunological functions in lymphocytes (Pepe & Curtiss, Reference Pepe and Curtiss1986; Kelly et al. Reference Kelly, Clay, Mistry, Hsieh-Li and Harmony1994; Mistry et al. Reference Mistry, Clay, Kelly, Steiner and Harmony1995), inhibition of platelet aggregation (Riddell et al. Reference Riddell, Graham and Owen1997), regulation of smooth muscle cell migration and proliferation (Zeleny et al. Reference Zeleny, Swertfeger, Weisgraber and Hui2002), and down-regulation of adhesion molecule expression in endothelial cells (Stannard et al. Reference Stannard, Riddell, Sacre, Tagalakis, Langer, von Eckardstein, Cullen, Athanasopoulos, Dickson and Owen2001). Apart from the effects on the surrounding cells, apoE is also thought to play an important role in macrophage metabolism acting as a modulator of cholesterol efflux (Bellosta et al. Reference Bellosta, Mahley, Sanan, Murata, Newland, Taylor and Pitas1995; Shimano et al. Reference Shimano, Ohsuga, Shimada, Namba, Gotoda, Harada, Katsuki, Yazaki and Yamada1995; Yoshida et al. Reference Yoshida, Hasty, Major, Ishiguro, Su, Gleaves, Babaev, Linton and Fazio2001; Hara et al. Reference Hara, Matsushima, Satoh, Iso-o, Noto, Togo, Kimura, Hashimoto and Tsukamoto2003) and NO production (Colton et al. Reference Colton, Czapiga, Snell-Callanan, Chernyshev and Vitek2001, Reference Colton, Brown, Cook, Needham, Xu, Czapiga, Saunders, Schmechel, Rasheed and Vitek2002). In addition, increasing evidence shows that apoE protects against oxidative reactions (Hayek et al. Reference Hayek, Oiknine, Brook and Aviram1994; Pratico et al. Reference Pratico, Lee, Trojanowski, Rokach and Fitzgerald1998; Kitagawa et al. Reference Kitagawa, Matsumoto, Kuwabara, Takasawa, Tanaka, Sasaki, Matsushita, Ohtsuki, Yanagihara and Hori2002), which may partly explain its anti-atherogenetic effects on vascular cell metabolism.
The apoE gene is polymorphic with two missense mutations resulting in three common allelic isoforms, apoE2, apoE3 and apoE4, which differ in the amino acids at positions 112 and 158 of the mature protein. apoE3 has Cys-112 and Arg-158, apoE2 has two cysteines and apoE4 has two arginines in these positions. apoE4 carriers (E3/E4, E4/E4), which represent 15–25 % of Caucasian populations, have been shown to have a relative risk of CVD of 1·4–1·5 relative to the common E3/E3 genotype (Song et al. Reference Song, Stampfer and Liu2004). The increased CVD burden has traditionally been attributed to higher LDL cholesterol in E4 carriers. However, it is becoming increasingly apparent that the modest 3–8 % elevation in LDL cholesterol (Minihane et al. Reference Minihane, Khan, Leigh-Firbank, Talmud, Wright, Murphy, Griffin and Williams2000) is unlikely to be the sole explanation for the difference in CVD incidence observed, and highlights the fact that the lipoproteins' apoE-mediated mechanisms are likely to contribute. Given that macrophages represent a significant source of apoE protein, which is a key regulator of the metabolism of the macrophages themselves and surrounding endothelial and smooth muscle cells and platelets, it is hypothesised that part of the disease differential is mediated by an impact of genotype on localised vascular wall metabolism.
A number of studies conducted in different models of Alzheimer's disease (AD) and atherosclerosis have reported a decreased antioxidant capacity of the apoE4 protein relative to apoE3 (Miyata & Smith, Reference Miyata and Smith1996; Jolivalt et al. Reference Jolivalt, Leininger-Muller, Bertrand, Herber, Christen and Siest2000; Tamaoka et al. Reference Tamaoka, Miyatake and Matsuno2000; Humphries et al. Reference Humphries, Talmud, Hawe, Bolla, Day and Miller2001; Dietrich et al. Reference Dietrich, Hua, Block, Olano, Packer, Morrow, Hudes, Abdukeyum, Rimbach and Minihane2005; Talmud et al. Reference Talmud, Stephens, Hawe, Demissie, Cupples, Hurel, Humphries and Ordovas2005). Macrophage oxidative state, which is mainly determined by the balance between cellular oxygenases and macrophage-associated antioxidants (Aviram & Fuhrman, Reference Aviram and Fuhrman1998), is a key regulator of macrophage metabolism. The differential influence of the apoE isoforms on this balance remains largely unknown. The aim of the present study was to characterise murine macrophages which have been stably transfected with either the human apoE3 or apoE4 gene, in terms of indices of oxidative status.
Materials and methods
Materials
Specialised chemicals were obtained from Sigma-Aldrich (Taufkirchen, Germany): Griess-Reagent, lipopolysaccharide (LPS; Salmonella enteriditis), buthionine sulphoximine (BSO), phorbol myristate acetate (PMA), reduced l-glutathione (GSH) and Sanger reagent. α-Tocopherol used in cell culture was obtained from BASF (Ludwigshafen, Germany) and α- and δ-tocopherol used as standards were obtained from Calbiochem (La Jolla, CA, USA). Neutral Red dye, nitroblue tetrazolium and dimethyl sulphoxide were purchased from Carl Roth (Karlsruhe, Germany). C11-BODIPY581/591 was obtained from Molecular Probes, Invitrogen (Karlsruhe, Germany). Specialised HPLC solvents were purchased from Mallinckrodt Baker BV (Deventer, The Netherlands). RAW 264·7 cells lines stably transfected with either human apoE3 or apoE4 were kindly given by Dr B. Pitas (Gladstone Institute, UCSF, USA). This model was chosen since the murine RAW 264·7 cells have been previously used to study differential effects of human apoE isoforms on monocyte function as these cells do not produce appreciable amounts of apoE (Miyata & Smith, Reference Miyata and Smith1996; Smith et al. Reference Smith, Miyata, Ginsberg, Grigaux, Shmookler and Plump1996). Cell culture medium, fetal bovine serum, penicillin/streptomycin and G-418 were from PAA Laboratories (Coelbe, Germany).
Cell culture
RAW 264·7-apoE3 and -apoE4 cells were cultured in Dulbecco's modified Eagle medium supplemented with 10 % fetal bovine serum, 4 mm-l-glutamine, 100 μg/ml penicillin and 100 μg/ml streptomycin, and 100 μg/ml G-418. Cells were grown in a humidified incubator at 37°C and 5 % CO2.
Cell lines assessment for apoE (genotype and protein levels)
For genotype cells, DNA was extracted from cell pellets using the QIAamp DNA Blood Mini Kit (Qiagen Ltd, Crawley, UK). The apoE region, which includes the two apoE polymorphic sites, was amplified by PCR using the following primers: 5′-ACA GAA TTC GCC CCG GCC TGG TAC AC-3′ (forward) and 5′-TAA GCT TGG CAC GGC TGT CCA AGG A-3′ (reverse). The amplified sequence was digested using the HhaI restriction enzyme, and the resultant fragments were separated and characterised by gel electrophoresis as described by Hixson & Vernier (Reference Hixson and Vernier1990).
ApoE protein concentrations were measured in cell supernatants to ensure that secreted levels were similar among the two clones. Protein was concentrated by using Vivaspin 20 columns (Sartorius, Epsom, UK). apoE concentration was determined by an automated turbimetric immunoassay kit (Apolipoprotein E-HA Wako, Alpha Laboratories Ltd, Eastleigh, UK) on the ILAB 600 automatic analyser (Instrumentation Laboratories UK Ltd, Warrington, UK). Total cell protein was measured with the BCA kit (Pierce Biotechnology, Rockford, IL, USA).
Assessment of peroxide-induced cell death
RAW 264·7-apoE3 and -apoE4 cells were seeded at 1·5 × 105 cells/well in twenty-four-well plates and grown for 48 h. Cells were then challenged with either tert-butyl hydroperoxide or H2O2 at increasing concentrations (100 μm to 2 mm) in serum-free media for 3 h. Cell viability was measured by the Neutral Red assay (Borenfreund & Puerner, Reference Borenfreund and Puerner1985) with slight modifications. Cells were incubated with the dye for 2 h and washed twice with PBS before extraction of the Neutral Red into solution. Cell viability (Neutral Red uptake) was determined by reading absorbance at 540 nm (Labsystems iEMS Reader; Labsystems, Helsinki, Finland). Results were expressed as absorbance observed as a percentage of control cultures.
Oxidation of C11-BODIPY581/591
The lipid peroxidation reporter molecule C11-BODIPY581/591 (Pap et al. Reference Pap, Drummen, Winter, Kooij, Rijken, Wirtz, Op den Kamp, Hage and Post1999) was used to examine the oxidation of membranes in macrophages. The C11-BODIPY581/591 dye is a lipophilic substance which incorporates into membranes and is useful to assess membrane oxidation since its oxidised and its reduced forms fluorescence at different wavelengths. C11-BODIPY581/591 oxidation was assessed under basal conditions and after oxidation. Macrophages were incubated with C11-BODIPY581/591 (10 μm) for 30 min at 37°C in growth medium. After labelling, cells were washed with PBS, and oxidation was induced by addition of cumene hydroperoxide (80 μm) and hemin (80 nm) in PBS and incubated for 1 h at 37°C. Control cells were incubated with PBS only. A multi-well spectrofluorometer (Optima; BMG Labtech, Offenburg, Germany) was used to measure green (485/520 nm) and red (544/590 nm) fluorescence intensity. Oxidation of C11-BODIPY581/591 was calculated as the ratio between green fluorescence (oxidised) and total fluorescence (oxidised plus reduced), to normalise for dye incorporation into cellular membranes.
α-Tocopherol and buthionine sulphoximine cytotoxicity test
In order to ensure that the α-tocopherol and BSO concentrations used for further experiments were not cytotoxic, cells were subcultured in twenty-four-well plates (1·5 × 105 cells/well) 24 h before experimental treatment. Cells were incubated with test substances at increasing concentrations (up to 100 μm for α-tocopherol and 200 μm for buthionine sulphoximine) for 24 h and the Neutral Red assay was carried out.
Glutathione levels
To determine whether apoE genotype had an influence on GSH levels under basal conditions, after treatment with α-tocopherol and upon GSH depletion, cells were incubated for 24 h with vehicle (0·1 % ethanol), α-tocopherol (50 μm in 0·1 % ethanol) or BSO (50 μm in PBS), an inhibitor of γ-glutamylcysteine synthetase, the rate-limiting enzyme for GSH de novo synthesis. After treatments, cells were washed twice and scraped on ice with cold PBS and then centrifuged. Cell pellets were frozen in liquid nitrogen and stored at − 80°C. Pellets were resuspended with 100 μl PBS and sonicated, an aliquot was used to determine protein concentration by the BCA kit, and 80 μl of sample was prepared for GSH determination (Hernandez-Montes et al. Reference Hernandez-Montes, Pollard, Vauzour, Jofre-Montseny, Rota, Rimbach, Weinberg and Spencer2006).
Briefly, after protein precipitation with TCA (5 %) containing 0·2 mm-desferal, GSH was derivatised by addition of iodoacetic acid to yield 5-carboxymethyl glutathione, and this was followed by complexation with dinitrofluorobenzene (Sanger Reagent). The resulting complex was analysed by HPLC analysis with UV detection by a modification of the method of Yoshida (Reference Yoshida1996).
Vitamin E concentration
To investigate whether the apoE genotype had an influence on the intracellular accumulation of α-tocopherol after its supplementation, macrophages were subcultured in six-well plates and grown for 24 h. Treatments with α-tocopherol and collection of samples were similar to that described by Gao et al. (Reference Gao, Stone, Huang, Papas and Qui2002). In summary, medium was enriched with α-tocopherol (5 and 50 μm) for 6 h. Adherent cells were washed with PBS and were overlaid with 420 μl lysis buffer (1 % Triton X-100 in pH 7·5, 20 mm-HEPES buffer with 0·1 mm-EDTA) and scraped. One aliquot was collected for protein determination and 2,6-di-tert-butyl-4-methylphenol was added to the other aliquot samples, which were stored for tocopherol analysis. δ-Tocopherol was used as internal standard. Cell samples were mixed with ethanol–1 % ascorbate (1 : 1·5) and with hexane (1 : 2·5) and were vortexed for 30 s. After phase separation, the upper layer was removed and dried under N2. Samples were resuspended in methanol. For the HPLC analysis, the mobile phase was methanol–water (98:2) isocratically delivered at a flow rate of 1·2 ml/min. The tocopherol content was analysed using a Jasco HPLC system on a Waters Spherisorb ODS-2 3 μm column (100 × 4·6 mm) with the detector set to an excitation wavelength of 290 nm and emission wavelength of 325 nm.
NO production
NO production was assessed by determining the amount of its stable metabolite, nitrite 
in culture supernatants using the Griess reaction. Cells were pretreated for 24 h with α-tocopherol (50 μm) or ethanol (0·1 %) and then stimulated for 4 h with LPS (1 μg/ml). Removal of LPS was carried out by washing twice with PBS, and fresh culture medium was added. After 20 h, supernatants were collected for nitrite measurement by the Griess reaction. Samples were processed as described elsewhere (Rimbach et al. Reference Rimbach, De Pascual-Teresa, Ewins, Matsugo, Uchida, Minihane, Turner, VafeiAdou and Weinberg2003) and nitrite concentration was determined by reading its absorbance at 540 nm (Labsystems iEMS Reader) using a standard curve for sodium nitrite.
Determination of intracellular superoxide anion production
A colorimetric assay was used to compare the amount of superoxide anion production by RAW 264·7-apoE3 and -apoE4 after stimulation with PMA. Superoxide anion was determined by the quantitative nitroblue tetrazolium assay (Choi et al. Reference Choi, Kim, Cha and Kim2006) with the only modification being that cells were grown in twenty-four-well plates and were stimulated for 1 h with increasing concentrations of PMA (0·25–1 μg/ml) in serum-free medium containing 0·05 % nitroblue tetrazolium (total volume of 500 μl/well). After dissolving the deposited nitroblue tetrazolium inside the cells, the solution was transferred into ninety-six-well plates and read at 620 nm (Labsystems iEMS Reader).
Statistical analysis
Statistical calculations were conducted with SPSS version 13.0 (SPSS Inc., Chicago, IL, USA). t Tests (for independent samples) were performed to compare the outcomes between apoE3 and apoE4 cells. In the absence of normal distributed data, the Mann–Whitney U test was used. Results are expressed as means and their standard errors, and significance was accepted at P < 0·05.
Results
Cell lines assessment for apoE
Genotyping confirmed that the RAW 264·7 cell lines contained the gene coding for either human apoE3 or apoE4. apoE protein levels secreted in the medium in 24 h were found to be similar in both clones (1·38 (sem0·38) μg/mg cell protein for apoE3, and 1·36 (sem0·38) for apoE4) and within the human physiological range (Wang-Iverson et al. Reference Wang-Iverson, Gibson and Brown1985).
α-Tocopherol and buthionine sulphoximine cytotoxicity test
Incubation of cells with α-tocopherol up to 100 μm, and BSO up to 200 μm for 24 h did not produce any changes in cell viability in RAW 264·7-apoE3 and -apoE4 as determined by the Neutral Red test.
Assessment of peroxide-induced cell death
At a concentration of 750 μm-H2O2, a modest loss of cell viability was evident (16 % for apoE3 and 20 % for apoE4). At a concentration of 1 mm-tert-butyl hydroperoxide a similar amount of cell death occurred (6 % for apoE3 and 20 % for apoE4) (Fig. 1(A, B)). At higher exposure of both substances, viability decreased in a dose-dependent manner. However, no significant differences between genotypes could be observed in the cytotoxicity dose–response curves.
Viability of apoE3 (◇) and apoE4 (■) stably transfected macrophages in response to peroxide challenge. RAW 264·7-apoE3 and -apoE4 cells were treated with either H2O2 (A) or tert-butyl hydroperoxide (B) at increasing concentrations (100 μm to 2 mm) in serum-free medium for 3 h. Cell viability was measured using the Neutral Red assay. Values, expressed as % viability relative to controls, are means with their standard errors depicted by vertical bars (two independent experiments performed in triplicate).
Oxidation of C11-BODIPY581/591
In cells where oxidative stress was not externally induced (baseline conditions), RAW 264·7-apoE4 showed an approximately 40 % increase (P < 0·001) in the oxidation ratio as compared to RAW 264·7-apoE3 (Fig. 2). One hour after induction with cumene hydroperoxide/hemin, the ratio increased considerably in both cases, but no differences could be observed between the two isoforms.
Comparison of C11-BODIPY581/591 dye oxidation in apoE3 (□) and apoE4 (
) macrophages. Cells were labelled with C11-BODIPY581/591 and oxidation was induced by addition of cumene hydroperoxide (CumOOH; 80 μm) and hemin (80 nm) for 1 h at 37°C. Baseline cells were incubated with PBS only. Fluorescence intensity was measured at green (485/520 nm) and red (544/590 nm) wavelengths. Oxidation of C11-BODIPY581/591 is expressed as the ratio between green fluorescence (oxidised) and total fluorescence (oxidised plus reduced). Values are means with their standard errors depicted by vertical bars (three or more independent experiments performed in quadruplicate). Mean values were significantly different from those of the apoE3 group: ***P < 0·001.
Glutathione levels
Intracellular glutathione levels were analysed in untreated cells (controls), and in cells incubated for 24 h with either α-tocopherol (50 μm) or BSO (50 μm). Although there was a trend for cells expressing apoE3 to exhibit higher levels of GSH, no significant differences could be observed between genotypes (Fig. 3). The amount of glutathione in untreated RAW 264·7-apoE3 and -apoE4 was 8·9 (sem 2·1) and 8·5 (sem 3·5) nmol/mg protein, respectively. A modest 16 % (P = 0·38) and 6 % (P = 0·75) increase in GSH levels were evident in apoE3 and apoE4 cells following α-tocopherol supplementation. BSO almost completely abolished GSH synthesis, so that only 8·7 and 5·1 % of the GSH levels found in controls could be detected in cells expressing apoE3 and apoE4, respectively.
Determination of reduced l-glutathione (GSH) levels in apoE3 (□) and apoE4 () macrophages. Cells were treated with 0·1 % ethanol (control), α-tocopherol (50 μm) or buthionine sulphoximine (BSO; 50 μm) for 24 h. Intracellular GSH levels were assessed as outlined in the Materials and Methods. Values are means with their standard errors depicted by vertical bars (three independent experiments performed in duplicate).
Vitamin E accumulation
Incubation with α-tocopherol for 6 h at 5 μm resulted in an accumulation of vitamin E in macrophages of 1·8 (sem 0·5) and 1·3 (sem 0·4) nmol α-tocopherol/mg protein in apoE3 and apoE4 cells, respectively (Fig. 4). Enrichment of medium with 50 μm-α-tocopherol for the same time period resulted in an approximately 10-fold increase of vitamin E detected in cell pellets (16·2 (sem 3·7) and 13·8 (sem 3·9) nmol α-tocopherol/mg protein for apoE3 and apoE4, respectively). Although in both tested concentrations there was a trend for apoE3 macrophages to accumulate more α-tocopherol, differences were not statistically significant.
Comparison of α-tocpherol accumulation in apoE3 (□) and apoE4 () macrophages. RAW 264·7-apoE3 and -apoE4 were incubated with α-tocopherol (5 and 50 μm) for 6 h and then intracellular concentration of vitamin E was analysed by HPLC as indicated in the Materials and Methods. Values are means with their standard errors depicted by vertical bars (three independent experiments performed in duplicate).
NO production
Twenty hours after stimulation of macrophages with LPS (1 μg/ml) for 4 h, macrophages showed a substantial increase in NO production (Fig. 5). The nitrite concentration in the medium from RAW 264·7-apoE4 was 6 % higher than in RAW 264·7-apoE3 (P = 0·028). Pre-incubation with α-tocopherol (50 μm) for 24 h reduced NO production in a genotype-dependent manner. RAW 264·7-apoE3 were more sensitive to vitamin E, with a 20 % reduction in nitrite, compared to a 7 % decrease evident in RAW 264·7-apoE4.
Evaluation of NO production in apoE3 (□) and apoE4 () macrophages. NO was assessed by determining the amount of its stable metabolite, nitrite in culture supernatants using the Griess reaction. Cells were pretreated for 24 h with α-tocopherol or ethanol (0·1 %) and then stimulated for 4 h with lipopolysaccharide (LPS) (1 μg/ml). Supernatants were collected 20 h later for nitrite analysis. Values are means with their standard errors depicted by vertical bars (three independent experiments performed in duplicate). Mean values were significantly different from those of the apoE3 group: *P < 0·05; **P < 0·01.
Superoxide production
After 1 h stimulation with PMA at increasing concentrations, the superoxide anion radical production increased in a dose-dependent way in both genotypes (Fig. 6). The amount of superoxide produced by the apoE4 macrophages was consistently higher in all the PMA doses tested. At 0·25 μg/ml PMA, cells expressing apoE4 produced 30 % more than the apoE3 macrophages (P = 0·005). At 0·5 and 1 μg/ml PMA, RAW264·7-apoE4 showed an increase of almost 60 % in production (P = 0·002 and P = 0·001).
Comparison of superoxide anion radical production between RAW 264·7-apoE3 (□) and -apoE4 (). Attached cells were stimulated with increasing concentrations of phorbol myristate acetate (PMA; 0·25–1 μg/ml) in the presence of 0·05 % nitroblue tetrazolium for 1 h. Superoxide anion production was determined by the quantitative nitroblue tetrazolium assay of Choi et al. (Reference Choi, Kim, Cha and Kim2006 Values are means with their standard errors depicted by vertical bars (three independent experiments performed in triplicate). Mean values were significantly different from those of the apoE3 group: **P < 0·01; ***P < 0·001.
Discussion
In addition to its well-described role in lipoprotein metabolism, apoE is becoming increasingly recognised for its role in numerous metabolic pathways within the vascular intima. Although the impact of apoE may be mediated by a cell-signalling process stimulated by interaction of apoE with cell surface apolipoprotein receptors, it is likely to be in part attributable to its antioxidant capacity within the relatively antioxidant-depleted environment of the vascular wall. In vitro studies have consistently reported differences in the antioxidant capacity of the different isoforms (Miyata & Smith, Reference Miyata and Smith1996). In macrophages, oxidative stress triggers a cycle of events which exacerbates lesion formation. The impact of genotype independently of apoE concentration on the oxidative status of macrophages is unknown and was the topic of the current investigation.
Since cellular apoE location is mainly associated with surface-connected compartments of the plasma cell membrane (Bellosta et al. Reference Bellosta, Mahley, Sanan, Murata, Newland, Taylor and Pitas1995; Duan et al. Reference Duan, Lin and Mazzone1997), the possible differential effects of apoE3 and apoE4 in protecting against membrane damage induced by H2O2 and tert-butyl hydroperoxide were investigated. The Neutral Red assay was chosen since it is a viability test sensitive to cell membrane injury. A shift in the cytotoxicity dose–response curve between apoE isoforms was not evident in our macrophage cell model. This differs from the results of Miyata & Smith (Reference Miyata and Smith1996), who showed protection of apoE3>E4 against H2O2-mediated cytotoxicity in rat neuronal B12 cell line after incubation with conditioned media containing either apoE3 or apoE4 (Miyata & Smith, Reference Miyata and Smith1996). In addition to possible cell-specific differences, an explanation to these contradictory results could be that our cell model exposed the cells to endogenously produced apoE, which is arguably a more physiological approach, rather than adding exogenous apoE to the cell culture medium.
To further determine whether the different apoE isoforms have an effect on the rate of membrane oxidation, cells were loaded with the C11-BODIPY581/591 dye, which can be incorporated readily into lipid bilayers by its fatty acyl chain and monitors the oxidation of unsaturated fatty acids in the membranes (Naguib, Reference Naguib1998; Pap et al. Reference Pap, Drummen, Winter, Kooij, Rijken, Wirtz, Op den Kamp, Hage and Post1999; Drummen et al. Reference Drummen, Makkinje, Verkleij, Op den Kamp and Post2004). Spontaneous oxidation of the dye (baseline levels) was significantly higher in RAW 264·7-apoE4 than in RAW 264·7-apoE3. Following stress induced by a free radical generation system (cumene hydroperoxide/hemin), lipid oxidation considerably increased in a genotype-independent manner. This is likely due to an inability of either isoform to protect against oxidation in a highly oxidising saturated environment. Similar results have been observed in the frontal cortex of AD brains, as measured by thiobarbituric acid reactive substances (Ramassamy et al. Reference Ramassamy, Averill, Beffert, Theroux, Lussier-Cacan, Cohn, Christen, Schoofs, Davignon and Poirier2000), in which under basal conditions, there was a genotype effect of apoE in lipid peroxidation (E4>E3), but after complete oxidation of the samples, no difference could be observed between genotypes. In any case both CVD and AD pathology are a result of chronic modest insults and it is likely that the genotype differences in lipid oxidation observed could make a meaningful contribution to the development of these diseases. Previously, the limited data in man was suggestive that genotype differences in oxidative status may be attributable to the lower circulating apoE protein levels in E4 carriers. As both E3 and E4 cells lines used in the current study secrete equivalent apoE protein levels, the data indicates that differences in the antioxidant capacity of the E3 and E4 isoforms may also make a significant contribution.
ApoE4 carriers have been suggested as a population that could potentially benefit from vitamin E supplementation (Peroutka & Dreon, Reference Peroutka and Dreon2000). α-Tocopherol is the most biologically active form of vitamin E and is known to be effectively accumulated by RAW 264·7 cells (Gao et al. Reference Gao, Stone, Huang, Papas and Qui2002; McCormick & Parker, Reference McCormick and Parker2004). Furthermore, in the current study its impact on macrophage oxidative status was determined. Accumulation studies indicated a trend towards higher concentrations of α-tocopherol in apoE3 compared to apoE4 transfected cells. In vivo LDL and HDL represent the major vitamin E carriers in the bloodstream, with apoE mediating the cellular uptake of HDL to scavenger receptor class B type 1 (Mardones et al. Reference Mardones, Strobel, Miranda, Leighton, Quinones, Amigo, Rozowski, Krieger and Rigotti2002). apoE4 has preferential affinity to VLDL rather than HDL, which could lead to a decreased uptake of vitamin E in the macrophages (or other tissues), and increased concentration in plasma. In fact, higher plasma vitamin E levels in apoE4 carriers have previously been reported (Lodge et al. Reference Lodge, Hall, Jeanes and Proteggente2004). However, any trends observed in the cell culture system may not be directly attributed to this potential mechanism as vitamin E was delivered in ethanol and not in lipoproteins. Furthermore, due to the nature of the experiment, it cannot be elucidated whether these small differences are due to lower uptake or to higher metabolism of vitamin E in apoE4 macrophages.
Glutathione is an important intracellular antioxidant found mainly in the cytosol, and it plays a pivotal role in maintaining the macrophage redox balance. In particular, GSH participates in the detoxification of peroxides, and in addition, it can determine the rate of LDL oxidation in these cells (Gotoh et al. Reference Gotoh, Graham, Nikl and Darley-Usmar1993). It was shown that mouse peritoneal macrophages from apoE-deficient mice had reduced GSH levels and increased lipid peroxides (Rosenblat et al. Reference Rosenblat, Coleman and Aviram2002). In addition, α-tocopherol has been shown to increase GSH in mouse peritoneal macrophages (Rosenblat et al. Reference Rosenblat, Coleman and Aviram2002). For these reasons, intracellular GSH concentrations were measured in RAW 264·7-apoE3 and -apoE4 under basal conditions and after incubation of cells with α-tocopherol at 50 μm, a concentration which was not cytotoxic and which falls within the reachable concentrations in plasma (Jialal et al. Reference Jialal, Fuller and Huet1995). In the present study, only a small decrease in GSH could be observed in E4 in respect to E3 under basal conditions. These results are in accordance with the findings of Ramassamy et al. (Reference Ramassamy, Averill, Beffert, Theroux, Lussier-Cacan, Cohn, Christen, Schoofs, Davignon and Poirier2000) and Fernandes et al. (Reference Fernandes, Proenca, Nogueira, Grazina, Oliveira, Fernandes, Santiago, Santana and Oliveira1999), who found no differences in the GSH concentration according to apoE genotype either in the frontal cortex or in the plasma of AD patients. Incubation of cells with α-tocopherol (50 μm) for 24 h induced a small increase in the GSH concentrations, which may be in part attributable to the small differences found in vitamin E accumulation. Furthermore, cells from both genotypes responded similarly to GSH depletion by incubation with BSO (50 μm), which inhibits the de novo synthesis of GSH.
Activated macrophages generate NO and superoxide anion which react to produce peroxynitrite, a strong oxidant, which is centrally involved in the pathogenesis of atherosclerosis. The rate of this reaction is faster than the rate of superoxide anion radical scavenging by superoxide dismutase (Beckman & Koppenol, Reference Beckman and Koppenol1996); therefore, NO can influence the fate of superoxide. NO is produced by nitric oxide synthase, with the induction of nitric oxide synthase occurring by activation of NF-κB, but also by other mechanisms, including the activation of NADPH oxidases (Zhou et al. Reference Zhou, Struthers and Lyles1999). It has been shown that apoE regulates the response to LPS in vivo (Ali et al. Reference Ali, Middleton, Pure and Rader2005), and modulates NO production in macrophages, in a genotype-dependent way (E4>E3; Colton et al. Reference Colton, Needham, Brown, Cook, Rasheed, Burke, Strittmatter, Schmechel and Vitek2004). Additionally α-tocopherol has been shown to inhibit NF-κB activation by LPS (Asehnoune et al. Reference Asehnoune, Strassheim, Mitra, Kim and Abraham2004). The present results confirm that RAW 264·7-apoE4 produce more NO than RAW 264·7-apoE3 following LPS stimulation. In addition, we show that apoE3 macrophages respond better to a pretreatment with α-tocopherol, since NO production was reduced by 20 % in apoE3 in comparison to 7 % in apoE4. The molecular mechanism by which apoE modulates LPS effects, and its possible synergism with vitamin E, need to be further investigated.
On the other hand, to assess the production of , cells were stimulated with PMA, an activator of the NADPH oxidase. It has been reported that NADPH oxidase-dependent superoxide produced by infiltrated monocytes is associated with atherosclerosis, as well as with thickness of carotid intima-media, even before appearance of clinical atherosclerotic disease (Bokoch & Knaus, Reference Bokoch and Knaus2003). The present results show a significant increase in superoxide anion radical production in apoE4 in comparison to apoE3 macrophages, in the three doses of PMA tested. The present data suggest that differential modulating effects of apoE3 and apoE4 on the NADPH oxidase system could partly contribute to the higher risk of CVD observed in apoE4 carriers. The hypothesis that apoE may influence the NADPH oxidase activity is supported by the fact that in apoE-deficient mice, mouse peritoneal macrophages secreted twofold more superoxide anions than the mouse peritoneal macrophages from the wild type (Rosenblat & Aviram, Reference Rosenblat and Aviram2002).
Peroxidation of cellular lipids and the formation of oxysterols have been considered responsible for the increased NADPH oxidase activity in apoE-deficient mice (Rosenblat & Aviram, Reference Rosenblat and Aviram2002). Increased content of oxysterols in E4 is plausible since differential oxidation of membrane lipids was observed in the present study (E4>E3), and this could partly contribute to the increased NADPH oxidase activity in apoE4 cells.
The mechanism by which apoE acts as an antioxidant, and why it is isoform-dependent, is not entirely understood. There are several putative mechanisms which include metal binding activity (Miyata & Smith, Reference Miyata and Smith1996), the presence of a thiol group (Cys 112) in apoE3, and recent work indicating that the most plausible mechanism relies on the radical scavenging activity of positively charged amino acids found in the receptor binding domain of the apolipoprotein (Pham et al. Reference Pham, Kodvawala and Hui2005). Due to the exchange of a Cys by an Arg at position 112 in apoE4, a domain interaction arises which leads to a less stable protein, with folding intermediates which are more rapidly degraded and which influence the affinity of apoE for different lipoproteins (Hatters et al. Reference Hatters, Peters-Libeu and Weisgraber2006). This could also have an impact on the antioxidant capacity of the protein (Pham et al. Reference Pham, Kodvawala and Hui2005).
The current study is suggestive that an impact of apoE genotype on monocyte oxidative status may contribute to the higher CVD and AD risk observed in man. However, it needs to be taken into account that cell culture models have limitations, and it is not possible to mimic the course of a chronic disease in vitro. Further clarification of the impact of apoE genotype on oxidative status using in vivo animal models and human trials is needed.
Conclusions
Taken together, current data indicate that apoE genotype affects several parameters which determine the oxidant/antioxidant status of macrophages. Differences (apoE4>apoE3) in membrane oxidation under basal conditions and in NO and superoxide anion radical production upon macrophage stimulation with LPS and PMA, respectively, were observed. Moreover, pretreatment with α-tocopherol was able to decrease LPS-induced NO production in E3>E4. On the other hand, small differences were observed in the concentrations of intracellular GSH and α-tocopherol (apoE3>apoE4). Taken together the data are suggestive of increased oxidative stress in macrophages which may be an important contributing factor to the higher CVD risk in apoE4 v. apoE3 carriers. Finally, the findings of the current trial are suggestive that antioxidant interventions, such as increased vitamin C, vitamin E or polyphenols, could serve to reduce the 40–50 % increased risk of CVD evident in E4 carriers.
Acknowledgements
We thank Dr B. Pitas (Gladstone Institute, USA) who kindly provided us with the RAW 264·7 cells expressing apoE3 and apoE4. The work was partially funded by the Spanish/German Programme Acciones Integradas Hispano-Alemanas (ref. HA2004-0091) and by a Ramon y Cajal grant to Sonia de Pascual Teresa.
