Oxidized LDL Induces Enhanced Antibody Formation and MHC Class II–Dependent IFN-γ Production in Lymphocytes From Healthy Individuals
Abstract The early stages of atherosclerosis are characterized by penetration into the arterial intima by both T lymphocytes and monocytes. Some of these T lymphocytes show signs of activation, though the mechanisms by which they become activated are not known. The monocytes develop into macrophages and subsequently into foam cells filled with oxidized LDL (oxLDL)-derived lipids. OxLDL has been found to exert several proinflammatory effects, including enhanced adhesiveness of endothelial cells and monocytes, chemotaxis of monocytes and T cells, and T-cell activation. The enzyme-linked immunospot (ELISPOT) assay has been shown to be a sensitive method for detection of single cells secreting antibodies or cytokines. Here we have used this method to characterize the T-cell cytokine secretion pattern after exposure to oxLDL in vitro. In peripheral blood mononuclear cells from healthy donors (n=27), a significantly enhanced number of INF-γ–producing cells was detected by ELISPOT (P<.001) after stimulation with 5 μg/mL oxLDL. In contrast, production of interleukin-4 was not significantly enhanced after stimulation with oxLDL. OxLDL-induced IFN-γ secretion and T-cell proliferation were completely inhibited by major histocompatibility complex (MHC) class II antibodies. Furthermore, oxLDL was found to enhance the antibody secretion, indicating B-cell activation. Our results indicate that oxLDL activates T cells by an MHC class II–dependent mechanism. In healthy individuals, oxLDL induces IFN-γ, which is produced by T helper type 1–like cells. These findings demonstrate that oxLDL induces a cell-dependent immune reaction, which may play an important role in the development of atherosclerosis.
- Received November 14, 1994.
- Accepted June 20, 1995.
During recent years, it has become clear that atherosclerosis has many characteristics in common with chronic inflammatory diseases. At the earliest stages of the disease, monocytes and T lymphocytes adhere to the endothelium and penetrate into the arterial intima. The monocytes differentiate into macrophages and subsequently develop into foam cells, filled with oxLDL or other forms of modified LDL–derived lipids. Some T lymphocytes in the lesion also become activated, with an enhanced expression of interleukin-2 (IL-2) receptors and HLA-DR. Both CD4+ and CD8+ T cells have been described in the atherosclerotic lesion.1 2 3
OxLDL has been shown to exert several potentially proatherogenic effects on the cells participating in the development of the atherosclerotic lesion. Different forms of oxLDL induce enhanced adhesiveness both in endothelial cells4 5 and monocytes.6 OxLDL is chemotactic for monocytes7 and T lymphocytes,8 stimulates differentiation of monocytes,6 and induces activation of T lymphocytes.9 OxLDL also has cytotoxic properties, and an important role of the macrophage scavenger receptor may be to remove this cytotoxic substance.10
LDL may be oxidized by endothelial cells, macrophages, smooth muscle cells,11 and T lymphocytes.9 The oxidation is believed to take place mainly in the artery wall, but the mechanisms by which cells oxidize LDL are not clear.11 In experiments with animals, antioxidants such as butylated hydroxytoluene12 have been found to inhibit development of atherosclerosis. Furthermore, oxLDL has been demonstrated in atherosclerotic lesions.13 Taken together, these findings favor the notion that oxLDL may play an important role in the development of atherosclerosis.
T cells are generally believed to be primary participants in the pathogenesis of several autoimmune diseases, including rheumatoid arthritis14 and multiple sclerosis.15 A preferential induction of Th1 response in rheumatoid arthritis16 and multiple sclerosis17 has been described, with expression of IFN-γ and IL-2, but no or little expression of IL-4 or IL-10. T-cell cytokines and antibody secretion may be determined by use of the ELISPOT technique, which has been reported as more sensitive than ELISA.18 Still little is known concerning the role of T cells in atherosclerosis. In the present article we report that exposure of PBMCs to low concentrations of oxLDL results in enhanced antibody formation and production of IFN-γ, which is MHC class II–dependent. The possible implications for oxLDL-mediated immune reactions in atherogenesis are discussed.
PBMCs were isolated from human buffy coats obtained from healthy blood donors by sterile technique. The buffy coat was diluted in PBS at a ratio of one to four, layered onto Ficoll-Hypaque, and centrifuged for 20 minutes at 1500g. Interface cells were washed twice in PBS. They were then counted and resuspended in RPMI 1640 medium supplemented with HEPES buffer, penicillin, and streptomycin and 10% fetal calf serum (complete medium) at a cell concentration of 1×106 cells/mL. Cell viability was determined by trypan blue exclusion and exceeded 95% in all experiments.
Determination of DNA Synthesis
PBMCs were prepared as described above, suspended in complete medium, and diluted to a cell concentration of 2×105 cells/mL. After the addition of native or oxLDL to the relevant cultures, 200 μL cell suspension was added to each well in round-bottomed 96-well cell culture plates (Nunc) and incubated for the indicated time periods in a 37°C humid cell incubator. Subsequently, 1 μCi of 3H-thymidine was added to each well. After 6 more hours of incubation, DNA was precipitated onto glass fiber filters by means of an automatic cell harvester, and the amount of incorporated 3H was determined in a liquid scintillation counter and expressed as counts per minute. Parallel cultures were made in all experiments.
Preparation of LDL
Venous blood from healthy donors was drawn after overnight fasting into precooled vacuum tubes containing Na2EDTA (1 mg/mL). Plasma was recovered by means of low-speed centrifugation (1400g, 20 minutes) at 1°C and kept at this temperature throughout the separation procedures. LDL was isolated from plasma in the density interval 1.025 to 1.050 kg/L by sequential preparative ultracentrifugation19 in a 50.3 Ti Beckman fixed-angle rotor (Beckman L8-80 ultracentrifuge) for 20 hours. The total protein content of the LDL preparation was determined by the Lowry technique.20
Oxidation of LDL by Copper
Isolated LDL was dialyzed in 0.02 mol/L phosphate/0.16 mol/L NaCl buffer, pH 7.4 for 15 hours at 4°C to remove EDTA. Copper-mediated oxidation of LDL was performed by incubating 0.2 mg/mL of EDTA-free LDL in medium F-10 containing 10−5 mol/L CuSO4 overnight at 37°C. Each preparation of oxLDL was used within 1 week of the oxidation. The presence of endotoxins in the lipoprotein preparations was analyzed with the Limulus assay (Kabi-Pharmacia AB). All endotoxin levels were below 0.5 ng/mL in the stock solutions and below 5 pg/mL in the test samples. There was no difference in endotoxin levels between native and oxLDL. The lipid peroxide contents of oxidized and native LDL were determined by analyzing TBAR substances and were expressed as MDA equivalents.21
Detection of IFN-γ– and IL-4–Producing Cells
For detection of IFN-γ–producing cells, 96-well nitrocellulose plates (Millititre HA, Millipore Co) were used. The plates were prewet with 200 μL/well fresh sterile PBS 30 minutes before being coated. Then a Millititer vacuum filtration holder (Millipore) was used to remove the solution. After three cycles of washing and vacuum suction using PBS, the plates were coated with 15 μg/mL, 50 μL/well anti-human IFN-γ catcher-mAb (1-DIK; Mabtech AB) diluted with fresh PBS. Plates were then left for coating in humid chambers at 4°C at least overnight.
PBMCs were prepared from healthy laboratory personnel and blood donors, and then oxLDL, native LDL, or PHA (Wellcome Diagnostics) at the indicated concentrations was added to the respective cell suspensions. Subsequently, the cells were cultured in round-bottomed cell culture plates (Nunc) at 37°C in 5% CO2 incubator for 72 hours.
The coated plates were then washed three times with sterile PBS with the vacuum device and 100 μL cell suspension was transferred to the wells of the precoated detection plates. The cells were then again incubated at 37°C, 5% CO2 overnight.
For the detection of IFN-γ, the nitrocellulose plates were washed two times and the PBS flicked off, followed by washing and vacuum suction as indicated above. Fifty microliters of a detector-mAb (7-B6-1-Biotin; Mabtech AB) at a concentration of 1 μL/mL was then added to each well in PBS. Thereafter, the plates were kept at +4°C overnight. After washing as above, we added 50 μL of avidin-alkaline phosphatase (Sigma Chemical Co) at 0.6 μg/mL in PBS, which was left for 2 hours in room temperature. After three more cycles of washing and vacuum suction, 50 μL of a phosphatase substrate solution producing an insoluble product (BCIP; Sigma) was added for 1 hour at room temperature. Thereafter, the plates were washed three times with distilled water and left to dry at room temperature. Spots were enumerated using a low-magnification (×20) dissection microscope, each spot representing one IFN-γ–producing cell.
For detection of IL-4, conventional ELISA plates (Immulon II, Dynatech) were used. The plates were coated with 15 μg/mL, 50 μL/well anti-human IL-4 mAb (Mabtech AB). These coated plates were kept at +4°C at least overnight in a moist chamber before use.
The precoated ELISA plates were then washed three times with sterile PBS. Thereafter, cell suspensions prepared as described above were transferred to wells (100 μL/well). The plates were then kept at 37°C in 5% CO2 overnight. Thereafter, the plates were washed three times with sterile PBS, and 50 μL/well of a detector-mAb (IL4-II-Biotin; Mabtech AB) at a concentration of 1 μL/mL was added. The plates were then kept in moist chambers at 4°C overnight. Reactions were developed as for IFN-γ ELISPOTs, except for the manual washing steps without vacuum device and the fact that BCIP had to be left in the wells for 4 to 5 hours to make the spots readily visible. Spots corresponding to cells that had secreted IL-4 were enumerated under low magnification (×20) in an inverted microscope, each spot representing one Il-4–producing cell.
Anti-MHC Class II Antibodies
For inhibition of MHC class II–dependent reactions, a pool of three different mAb reacting against human MHC class II was used at a maximal total final concentration of 1 μg/mL. This concentration had in preliminary experiments been shown to block PBMC proliferation and IFN-γ ELISPOT production to background levels using a tetanus toxoid as model antigen (data not shown). The pool consisted of equal amounts of 2.06 (mouse IgG1, anti-HLA DR; American Type Culture Collection [ATCC]), IVA 12 (IgG1, anti-HLA DR, DP, DQ; ATCC), and 9,3F10 (IgG2a, anti-HLA DR, DQ; ATCC).
As control antibodies for the MHC class II–blocking experiments, two monoclonals directed against keyhole limpet hemocyanin were used, clone H5 (IgG1) and clone 7-B4 (IgG2a).24 Control antibodies were used in the same concentrations and ratios of IgG subclasses as for the blocking antibodies in each experiment.
As another control, mAb directed toward an irrelevant antigen (human von Willebrand molecule) were used (IgG1, Immunotech).
Detection of Antibody Production
PBMCs from healthy donors were prepared and suspended in complete medium at a concentration of 2×105 cells/mL. OxLDL or LDL was added at the indicated concentrations, and cell suspensions were incubated at 37°C. After 16 hours, cells were washed three times in PBS, and 200 μL cell suspension at a concentration of 105/well or lower was transferred to the precoated ELISPOT wells. The frequency of cells producing antibodies was determined by a modified version of the ELISPOT technique exactly as recently described from our laboratory.25 26 Spots were counted under low magnification (×20) with an inverted microscope.
Analysis of Cellular Composition of PBMCs
Standard flow cytometry was used for the quantification of lymphocyte subsets in a FACScan from Becton Dickinson & Co. The following directly fluorochrome–conjugated antibodies were used: anti-CD3 (Leu-4)-fluorescein isothiocyanate (FITC)-conjugated, anti-CD4 (Leu-3a) phycoerythrin (PE)-conjugated, anti-CD8 (Leu-2a)-PE all purchased from Becton Dickinson, and anti-CD19 (HB37), A/S anti-HLe-1 (anti-CD45), Leu-M3 (anti-CD14) purchased from DAKO. For the automatic lymphocyte gating, the compound reagent Simultest Leucogate consisting of anti-CD45 (anti-HLe-1)-FITC and anti-CD14 (Leu-M3)-PE was used, and as negative control the compound reagent Simultest Control g1/g2a was used (both from Becton Dickinson). Cells, 200 000/tube in PBS, were stained with antibodies for 15 minutes in room temperature. After staining, the cell suspension was washed once in PBS and then directly analyzed on a FACSort flow cytometer, using the simulset software for automatic lymphocyte gating. T lymphocytes were defined as CD3+ positive cells, T helper lymphocytes as CD3+CD4+ double positive cells, T suppressor/killer lymphocytes as CD3+CD8+ double positive cells, and B lymphocytes as CD19+CD3− single positive cells.
Conventional methods were used for calculation of means and standard deviations. Coefficients of skewness and kurtosis were calculated to test deviations from a normal distribution. Differences in individual experiments between control samples and samples stimulated with oxidized or native LDL were analyzed by Student’s t test. Effects of native and oxLDL in the whole group of individuals tested were analyzed by Wilcoxon signed-rank test.
Exposure of native LDL to 10 μmol/L of Cu2+ for 16 hours at 37°C resulted in an increase in the amount of TBAR material present in the LDL preparation from 1.5±0.4 to 46.0±1.8 MDA equivalents/mg protein and in an increased mobility during agarose gel electrophoresis (data not shown).
In accordance with our previous results, the proliferation of PBMCs as determined by DNA synthesis was significantly enhanced after stimulation with copper-oxidized LDL. The optimal concentration of oxLDL for T-cell activation varied between individuals and different experiments, but in general, maximal stimulation was obtained at oxLDL concentrations between 1 and 10 μg/mL. Also, native LDL induced proliferation of PBMCs, but in general, higher concentrations (10 μg/mL) of native LDL were needed for maximal stimulation (data not shown).
To determine if oxLDL-induced T-cell stimulation is dependent on MHC class II, we tested the effect of mAb directed against MHC class II on oxLDL-induced PBMC proliferation. We found that a total concentration of 1 μg/mL of three pooled anti-MHC class II mAb added to the cell cultures abolished the T-cell response to 5 μg/mL oxLDL (Fig 1⇓). A similar effect was obtained when tetanus toxoid or native LDL was used as antigen (data not shown). The control antibody had no inhibitory effect (data not shown).
In contrast, antibodies to MHC class II at the same concentration that completely inhibited the response to oxLDL and tetanus toxoid had no effect on the T-cell response to the mitogen PHA, which has a strong stimulatory effect on T cells (Fig 2⇓).
Effect of OxLDL and Native LDL on Cytokine Pattern of PBMCs
To allow a quantitative determination of the induction of cytokines at the single-cell level, the ELISPOT assay was used. In 27 healthy blood donors tested, oxLDL at 5 μg/mL induced a 167% (SEM±32%; P<.001) increase. LDL at 10 μg/mL induced a 52% (SEM±42%) increase, but this difference was not significant. For comparison, incubation of a strong antigen, tetanus toxoid, at 5 U/mL with PBMCs from 6 healthy blood donors gave a 269% (SEM±51%; P<.01) increase in the number of IFN-γ–producing cells. In most individuals tested, a maximal effect was obtained at 5 μg/mL oxLDL and 10 μg/mL LDL. Only in 1 individual was a decrease in IFN-γ secretion noted after stimulation with oxLDL. A representative experiment demonstrating the effect of native and oxidized LDL on IFN-γ secretion is shown in Fig 3⇓. OxLDL had no effect on secretion of IL-4 in 16 of 18 healthy individuals tested. In 2 individuals, an enhanced IL-4 secretion was detected after stimulation with oxLDL (data not shown). LDL had no effect on the production of IL-4. A representative experiment showing IL-4–producing cells after stimulation with PHA or oxLDL is demonstrated in Fig 4⇓.
Effect of Antibodies to MHC Class II on OxLDL-Induced IFN-γ Secretion at the Single- Cell Level
To determine whether oxLDL-induced IFN-γ secretion is dependent on MHC class II, we tested the effect of mAb to MHC class II on oxLDL-induced IFN-γ secretion. We found that pooled anti-MHC class II antibodies added to the cell cultures completely abolished the oxLDL-induced IFN-γ secretion, while the control antibodies at the same concentrations and the same Ig isotypes had no effect (Fig 5⇓). A similar effect was obtained when tetanus toxoid was used as antigen (data not shown). We then determined if a monoclonal control antibody to another irrelevant antigen (von Willebrand factor) had any effect. In a representative experiment, 5 μg/mL oxLDL induced an increase from 22±2.0 to 34±1.1 of IFN-γ–producing cells/105. When 1 μg/mL control antibody was present, oxLDL induced an increase from 23.5±4.0 to 37±6.1 IFN-γ–producing cells/105. The irrelevant antibody thus had no effect on oxLDL-induced IFN-γ production. Trace amounts of copper are present in the cell cultures, but copper had no effect on production of IFN-γ (data not shown).
Effect of OxLDL on Antibody Secretion at the Single-Cell Level
To determine whether oxLDL-induced T-cell activation also involved a B-cell response, we tested the production of antibodies in PBMCs after treatment with oxLDL. We found that oxLDL induced enhanced antibody production of IgG, IgA, and IgM class, as determined by the ELISPOT technique (Fig 6⇓). Antibody specificity was not investigated in these experiments.
Cellular Composition of PBMCs
The cellular composition of PBMCs was analyzed by FACScan (Fig 7⇓). Our data indicate that the proportion of CD4+ and CD8+ T cells and B cells did not change after 3 days of culture and transfer to nitrocellular plates for determination of cytokine secretion by the ELISPOT technique. Furthermore, no difference between cells cultured in 5 μg/mL oxLDL and cells cultured in medium was detected.
The present study demonstrates that oxLDL-induced T-cell activation in healthy individuals is characterized by enhanced secretion of IFN-γ. Antibodies to oxLDL have been detected in both healthy individuals and patients with clinical manifestations of atherosclerosis,27 and we here demonstrate that oxLDL induces B-cell activation, as determined by enhanced secretion of immunoglobulins of IgG, IgA, and IgM classes. Cytokine and antibody formation was determined by use of the highly sensitive ELISPOT technique. The immune reaction initiated was dependent on MHC class II, since antibodies to MHC class II prevented both oxLDL-induced PBMC proliferation and IFN-γ secretion. An irrelevant control antibody had no effect. We found that incubation with anti-MHC class II antibodies had slight inhibitory effects on spontaneous IFN-γ secretion. This finding is in accordance with earlier reports and depends on T cells reactive with autologous MHC class II molecules.28
Both the basal IFN-γ secretion and the response to oxLDL varied between different individuals but was highly significant in the whole group (n=27) of healthy donors tested, indicating that an immune response to oxLDL is common in the population. In accordance with earlier findings, native LDL also induced proliferation and IFN-γ secretion in PBMCs, although not as strongly as oxLDL.
Oxidation of LDL is believed to be a key factor in the development of atherosclerosis, and during recent years the immune mechanisms in the development of atherosclerosis have gained considerable interest.1 2 3 The mechanisms by which oxLDL may activate T cells are largely unknown. In principle, oxLDL may induce T-cell stimulation by unspecific mechanisms, but it is also possible that activation by means of a conventional antigen or a superantigen is involved.
The possibility that the effect of oxLDL on PBMCs is due to an endotoxin contamination in the lipoprotein preparations cannot be completely excluded, since both LDL and oxLDL contain trace amounts of endotoxin (<5 pg/mL in the test samples). However, native LDL containing amounts of endotoxin similar to those in oxLDL had no significant effect on IFN-γ expression in the group tested. It is therefore not likely that endotoxin is responsible for oxLDL-induced enhanced expression of IFN-γ.
IFN-γ may be produced by T cells and natural killer cells. However, the finding that antibodies to MHC class II inhibit IFN-γ formation indicates that T cells are the main source of IFN-γ.
Oxidation of LDL is initiated by abstraction of a hydrogen atom from a polyunsaturated fatty acid in a cell membrane or lipoprotein. This leads to a chain reaction with formation of lipid hydroperoxides and aldehydes, fragmentation of the carrier protein apo B-100 and exposure of novel epitopes, and profound changes in the properties of the LDL molecule including increased negative charge, density, and electrophoretic mobility and decreased content of polyunsaturated fatty acids and vitamin E.29 OxLDL but not native LDL is taken up in macrophages after binding by the scavenger receptor, leading to foam cell formation.30
One possibility is that novel epitopes on oxLDL are presented on monocytes in the context of MHC class II and recognized as foreign by T cells, leading to a specific immune response.
We have recently found that oxLDL induces HSP 60 in monocytic cells (J.F. et al, unpublished observation, 1995). Immunity to HSP 60 has been implicated in autoimmune diseases such as rheumatoid arthritis and recently also atherosclerosis.31 32 HSP 60–reactive T cells are present also in healthy control individuals.33 Another possibility is thus that oxLDL activates T cells indirectly, by inducing HSP 60, which in its turn activates T cells in an MHC-dependent manner.
OxLDL-induced T-cell activation may also be dependent on a superantigen. Superantigens can activate T cells by crosslinking V regions of the T-cell receptor with MHC molecules on accessory cells.34 T-cell activation induced by superantigens is in general strong, and oxLDL-induced T-cell stimulation is comparatively weak, which argues against this possibility.
The most likely interpretation of our data is that oxLDL induces a cell-mediated immune reaction with a preferential induction of IFN-γ, a Th1 cytokine. Cells are activated by an MHC class II–dependent mechanism and thus require the presence of antigen-presenting cells. Signs of B-cell activation as determined by induction of immunoglobulin production were seen after incubation of PBMCs with oxLDL. The mechanisms behind this B-cell activation and the specificity of antibodies produced are not known, but this finding is compatible with recent data indicating the presence of antibodies to oxLDL in both normal control subjects and patients with clinical manifestations of atherosclerosis.26
T cells may in principle both inhibit and promote further development of atherosclerosis. IFN-γ, induced by oxLDL from T cells, may exert dual effects on the development of the disease. IFN-γ may inhibit smooth muscle cell growth and also suppress atherosclerosis in experimental animal models.35 On the other hand, IFN-γ and other T-cell–produced cytokines activate macrophages to secrete inflammatory mediators as well as growth factors for smooth muscle cells, which may lead to enhanced plaque growth and also a perpetuation of the inflammatory disease process.35 36 Another cytokine, TGF-β, has been proposed to play an important role in the atherosclerotic lesion, by effects possibly acting both to suppress and promote disease. TGF-β may both inhibit and stimulate smooth muscle cell growth, depending on the concentration used.37 38 The outcome of a human inflammatory disease may depend on the subpopulation of T cells that predominates at the site of inflammation. An intriguing possibility is thus that the Th1/Th2 balance in the atherosclerotic lesion may influence the disease progress, and it is possible that the balance between the various reactions initiated in the early inflammatory atherosclerotic lesion may influence the final outcome.
Selected Abbreviations and Acronyms
|ELISA||=||enzyme-linked immunosorbent assay|
|MHC||=||major histocompatibility complex|
|PBMC||=||peripheral blood mononuclear cells|
|Th1||=||T helper type 1|
This work was supported by the Swedish Medical Research Council (B95-19X-11244-01), the Foundation for Old Servants, King Gustaf V 80th Birthday Fund, the Swedish Society of Medicine, Ostermans Fund, the Swedish Association Against Rheumatism, and the Nanna Svartz Fund. The authors thank Dr Kerstin Carlson for preparation of LDL.
Berliner JA, Territo MC, Sevanian A, Ramin S, Kim JA, Bamshad B, Esterson M, Fogelman AM. Minimally modified low density lipoprotein stimulates monocyte endothelial interactions. J Clin Invest. 1990;85:1260-1266.
Frostegård J, Nilsson J, Haegerstrand A, Hamsten A, Wigzell H, Gidlund M. Oxidized low density lipoprotein induces differentiation and adhesion of human monocytes and the monocytic cell line U937. Proc Natl Acad Sci U S A. 1990;87:904-908.
Quinn MT, Parthasarathy S, Fong LG, Steinberg D. Oxidatively modified low density lipoproteins: a potential role in recruitment and retention of monocyte/macrophages during atherogenesis. Proc Natl Acad Sci U S A. 1987;84:2995-2998.
McMurray HF, Parthasarathy S, Steinberg D. Oxidatively modified low density lipoprotein is a chemoattractant for human T lymphocytes. J Clin Invest. 1993;92:1004-1008.
Frostegård J, Wu R, Giscombe R, Holm G, Lefvert AK, Nilsson J. Induction of T-cell activation by oxidized low density lipoprotein. Arterioscler Thromb. 1992;12:461-467.
Hessler JR, Morel DW, Lewis LJ, Chisolm GM. Lipoprotein oxidation and lipoprotein-induced cytotoxicity. Arteriosclerosis. 1983;3:215-222.
Björkhem I, Henriksson-Freyschuss A, Breuer O, Diczfalusy U, Berglund L, Henriksson P. The antioxidant butylated hydroxytoluene protects against atherosclerosis. Arterioscler Thromb. 1991;11:15-22.
Palinski W, Rosenfeld ME, Ylä-Herttuala S, Gurtner GC, Socher SS, Butler SW, Parthasarathy S, Carew TE, Steinberg D, Witztum JL. Low density lipoprotein undergoes modification in vivo. Proc Natl Acad Sci U S A. 1989;86:1372-1376.
Havel RJ, Eder HA, Bragdon JH. The distribution and chemical composition of ultracentrifugally separated lipoproteins in human serum. J Clin Invest. 1955;34:1345-1353.
Lowry O, Rosebrough NJ, Farr AL, Randall RJ. Protein measurement with the Folin phenol reagent. J Biol Chem. 1951;193:265-275.
elGhazali GE, Paulie S, Andersson G, Hansson Y, Holmquist G, Sun JB, Olsson T, Ekre HP, Troye-Blomberg M. Number of interleukin-4- and interferon-gamma-secreting human T cells reactive with tetanus toxoid and the mycobacterial antigen PPD or phytohemagglutinin: distinct response profiles depending on the type of antigen used for activation. Eur J Immunol. 1993;23:2740-2745.
Bensussan A, Meuer SC, Schlossman SF, Reinherz EL. Delineation of an immunoregulatory amplifier population recognizing autologous Ia molecules: analysis with human T cell clones. J Exp Med. 1984;159:559-576.
Parthasarathy S, Printz DJ, Boyd D, Joy L, Steinberg D. Macrophage oxidation of low density lipoprotein generates a modified form recognized by the scavenger receptor. Arteriosclerosis. 1986;6:505-510.
Xu Q, Dietrich H, Steiner HJ, Gown AM, Schoel B, Mikuz G, Kaufmann SH, Wick G. Induction of arteriosclerosis in normocholesterolemic rabbits by immunization with heat shock protein 65. Arterioscler Thromb. 1992;12:789-799.
Kiessling R, Grönberg A, Ivanyi J, Söderström K, Ferm M, Kleinau S, Nilsson E, Klareskog L. Role of hsp60 during autoimmune and bacterial inflammation. Immunol Rev. 1991;12:91-111.
Wilson AC, Schaub RG, Goldstein RC, Kuo PT. Suppression of aortic atherosclerosis in cholesterol-fed rabbits by purified rabbit interferon. Arteriosclerosis. 1989;10:208-214.
Geng Y, Hansson GK. Interferon-g inhibits scavenger receptor expression and foam cell formation in human monocyte-derived macrophages. J Clin Invest. 1992;89:1322-1330.