Extent of lesion development was analyzed by computer-assisted morphometry on oil red O-stained and hematoxylin-counterstained serial sections of the aortic root. The immunization protocol started in 8-week-old female apolipoprotein E knockout mice and was scheduled every second week. Mice (n = 8/group) received intraperitoneal injections of either phosphate-buffered saline (PBS) or the adjuvant coupled to either keyhole limpet hemocyanin (KLH) or phosphorylcholine (PC)-KLH. Mice were killed after 4 months of treatment. (Top) Quantification of lesion size in cross-sections of the aortic root at the specified distance from the appearance of the first cusp. (Bottom) Representative micrographs of oil red O-stained images from the 3 groups at the specified distance from the appearance of the first cusp. *p < 0.01 versus PBS or KLH (2-way analysis of variance).

Enzyme-linked immunosorbent assays were performed on sera collected at death (similar results were obtained on sera collected by retro-orbital puncture 9 weeks after the first immunization) to measure total immunoglobulin (Ig)G and IgM as well as anti-phosphorylcholine (PC) IgG and IgM titers. Individual sera from each mouse were tested at 1:10 dilution for specific antibodies and at 1:50 dilution for total antibodies. Data are represented as mean ± SEM for each group of treatment. *p < 0.05 versus PBS or KLH (1-way analysis of variance). OD = optical density; other abbreviations as in Figure 1.

(Left and center) Enzyme-linked immunosorbent assays (ELISA) were performed on sera collected at death to measure IgG and IgM anti-oxidized low-density lipoprotein (oxLDL) titers. Individual sera from each mouse were tested at increasing dilutions (from 1:6.25 to 1:100). Data are represented as mean ± SEM for each group of treatment. Reactivity against native LDL was not different from background signal (data not shown). *p < 0.05 versus phosphate-buffered saline or keyhole lympet hemocyanin (1-way analysis of variance). (Right) Competitive ELISA was performed by coating 2 µg/ml PC-bovine serum albumin (BSA) onto enzyme immunoassay plates and using serum dilutions corresponding to 50% of maximal binding on 2 µg/ml PC-BSA as determined by preliminary experiments. Increasing concentrations of oxLDL were coincubated with the individual sera for 1 h at room temperature. As shown, the level of binding of anti-PC IgG decreased with increasing concentrations of oxLDL, suggesting competition of the 2 antigens for the same antibodies. Abbreviations as in Figures 1 and 2.

Thioglycollate-elicited peritoneal macrophages were cultured on LabTek slide chambers during 24 h with complete DMEM medium supplemented with 10% fetal calf serum (FCS) and 20% L-929 conditioned medium. Macrophages were then deprived of FCS for 12 h before a 15-h incubation with or without oxidized low-density lipoprotein (50 µg/ml) and with or without serum (50 µg/ml) from phosphate-buffered saline (Ctrl) or PC-immunized mice (PC-KLH). Internalized LDL were stained with oil red O, and 3 random fields per condition were observed with phase-contrast microscopy (magnification x200). A customized program written with Qwin (Leica) was used to quantify the intracellular oil red O staining. The mean (±SEM) of intracellular staining per field is expressed in arbitrary units (au). A representative field and cell (frame) is shown for each condition. *p < 0.05 versus oxLDL without serum (1-way analysis of variance). Abbreviations as in Figures 1 and 3.

Representative micrograph showing immunohistochemistry from atherosclerotic lesions of phosphorylcholine (PC)-immunized apolipoprotein E knockout mice. T lymphocytes were detected using an anti-CD3 antibody, B cells with an anti-CD19 antibody, MHC II⁺ cells with an anti–I-A^b antibody, and T15 antibodies deposited in the lesions with the F6 anti-T15 antiidiotypic antibody (a kind gift from Dr. John Kearney, University of Alabama, Birmingham). Positive cells are stained red (hematoxyilin counterstaining). Original magnification x200. A = adventitia; L = lumen.

Immunohistochemistry was performed using anti–I-A^b (inflammatory cells) and anti-CD19 (B lymphocytes) primary antibodies. Positive cells were revealed by alkaline phosphatase-conjugated secondary antibodies and fast red substrate. Slides were counterstained by hematoxylin. Positive (red) staining was quantified by computer-assisted image analysis, and data are expressed as percentage of total cells. Inflammatory cells were significantly reduced in plaques of PC-immunized apolipoprotein E (apoE) knockout (KO) mice. In contrast, the number of infiltrating B lymphocytes was increased in PC-immunized apoE KO mice. *p < 0.05 versus PBS; **p < 0.01 versus KLH or PBS (1-way analysis of variance). Abbreviations as in Figure 1.