Previous work has implicated PPAR gamma in the regulation of CD36 expression and macrophage uptake of oxidized LDL (oxLDL). We provide evidence here that in addition to lipid uptake, PPAR gamma regulates a pathway of cholesterol efflux. PPAR gamma induces ABCA1 expression and cholesterol removal from macrophages through a transcriptional cascade mediated by the nuclear receptor LXR alpha. Ligand activation of PPAR gamma leads to primary induction of LXR alpha and to coupled induction of ABCA1. Transplantation of PPAR gamma null bone marrow into LDLR -/- mice results in a significant increase in atherosclerosis, consistent with the hypothesis that regulation of LXR alpha and ABCA1 expression is protective in vivo. Thus, we propose that PPAR gamma coordinates a complex physiologic response to oxLDL that involves particle uptake, processing, and cholesterol removal through ABCA1.

No abstract

Epidemiologic evidence has established a relationship between microbial infection and atherosclerosis. Mammalian TLRs provide clues on the mechanism of this inflammatory cascade. TLR2 has a large ligand repertoire that includes bacterial-derived exogenous and possibly host-derived endogenous ligands. In atherosclerosis-susceptible low-density lipoprotein receptor-deficient (Ldlr-/-) mice, complete deficiency of TLR2 led to a reduction in atherosclerosis. However, with BM transplantation, loss of TLR2 expression from BM-derived cells had no effect on disease progression. This suggested that an unknown endogenous TLR2 agonist influenced lesion progression by activating TLR2 in cells that were not of BM cell origin. Moreover, with intraperitoneal administration of a synthetic TLR2/TLR1 agonist, Pam3CSK4, disease burden was dramatically increased in Ldlr-/- mice. A complete deficiency of TLR2 in Ldlr-/- mice, as well as a deficiency of TLR2 only in BM-derived cells in Ldlr-/- mice, led to striking protection against Pam3CSK4-mediated atherosclerosis, suggesting a role for BM-derived cell expression of TLR2 in transducing the effects of an exogenous TLR2 agonist. These studies support the concept that chronic or recurrent microbial infections may contribute to atherosclerotic disease. Additionally, these data suggest the presence of host-derived endogenous TLR2 agonists.

Many reactive products may be formed when LDL undergoes lipid peroxidation, which in turn can react with lipids, apoproteins, and proteins, generating immunogenic neoepitopes. Autoantibodies recognizing model epitopes of oxidized low density lipoprotein, such as malondialdehydelysine, occur in plasma and in atherosclerotic lesions of humans and animals. Because apo E-deficient mice develop particularly high titers of such autoantibodies, we used their spleens to clone 13 monoclonal antibodies to various epitopes of oxidized LDL ("E0 antibodies"). Binding and competitive RIAs demonstrated significant differences in fine specificity even between E0 antibodies initially selected for binding to the same screening antigen. For example, some E0 antibodies selected for binding to malondialdehyde-LDL also recognized copper oxidized LDL, acrolein-LDL, or LDL modified by arachidonic or linoleic acid oxidation products. Circulating IgG and IgM autoantibodies binding to copper-oxidized LDL, 4-hydroxynonenal-LDL, acrolein-LDL, and LDL modified with arachidonic or linoleic acid oxidation products were found in apo E-deficient mice, suggesting that the respective antigens are formed in vivo. Epitopes recognized by some of the E0 monoclonal antibodies were also found on human circulating LDL. Each of the E0 monoclonal antibodies immunostained rabbit and human atherosclerotic lesions, and some of them yielded distinct staining patterns in advanced lesions. Together, this suggests that the natural monoclonal antibodies recognize different epitopes of complex structures formed during oxidation of lipoproteins, or epitopes formed independently at different lesion sites. Our data demonstrate that a profound immunological response to a large number of different epitopes of oxidized lipoproteins occurs in vivo. The availability of "natural" monoclonal autoantibodies should facilitate the identification of specific epitopes inducing this response.( J. Clin. Invest. 1996. 98:800-814.)

Increasing evidence indicates that low density lipoprotein (LDL) has to be modified to Induce foam cell formation. One such modification, oxidation of LDL, generates a number of highly reactive short chain-length aldehydlc fragments of oxidized fatty acids capable of conjugating with lyslne residues of apoprotein B. By Immunizing animals with homologous malondlaldehyde-modlfled LDL (MDA-LDL), 4-hydroxynonenal-LDL (4-HNE-LDL), and Cu ++ -oxldlzed LDL, we developed polyvalent and monoclonal antibodies against three epitopes found In oxldatively modified LDL The present article characterizes an antiserum and monoclonal antibody (MAL-2 and MDA2, respectively) specific for MDA-lyslne, and an antiserum and monoclonal antibody (HNE-6 and NA59, respectively) specific for 4-HNE-lyslne. In addition, a monoclonal antibody (OLF4-3C10) was developed against an as yet undefined epltope generated during Cu ++ oxldatJon of LDL With these antibodies, we demonstrated that MDA-lyslne and 4-HNE-lyslne adducts develop on apo-llpoproteln B during copper-Induced oxidation of LDL in vitro. The application of these antibodies for immunocytochemlcal demonstration of oxidized lipoprotelns In atherosclerotic lesions of progressive severity Is described in the companion article. These antibodies should prove useful In studying the role of oxldatively modified lipoprotelns as well as other oxldatively modified proteins In atherogenesls.