Oxidatively modified LDLs are antigenic and elicit the generation of autoantibodies often detected in plasma and within plaques of atherosclerotic patients. Although Cu(2+)-oxidized LDL and malondialdehyde (MDA)-modified LDL are usually used as antigens in immunoassays, other, still unrecognized epitopes may be formed in vivo during LDL oxidation and may induce antibody production. Antibodies recognizing LDL oxidatively modified by Cu2+, 2,2'-azobis-(2-amidino propane) hydrochloride (AAPH), and the combination of horseradish peroxidase and H2O2 (HRP) were detected in serum of a group of 90 unselected patients. HRP-oxidized LDL was the antigen that revealed the highest IgG titers, although the extent of LDL oxidation (evaluated as conjugated diene formation, loss of tryptophan fluorescence, production of fluorescent aldehydic adducts, and change in electrophoretic mobility) was comparable to that obtained with Cu2+ and AAPH. There was a highly statistically significant correlation between the IgG titers detected using Cu(2+)- and AAPH-oxidized LDLs as antigens, but no correlation was found between the IgG titers revealed by HRP and Cu2+ or AAPH. In addition, the antibody titers against MDA-modified LDL exhibited a significant correlation with those against Cu(2+)- or AAPH-oxidized LDL but did not correlate with those against HRP-oxidized LDL. Finally, immunocompetition experiments revealed that IgG recognizing HRP-oxidized LDL did not cross-react with Cu(2+)-oxidized LDL and vice versa. The possibility that lipid peroxidation-independent modifications could play a role in HRP-induced formation of antigenic epitopes in LDL was supported by two lines of evidence. First, in probucol-enriched LDL, despite the complete inhibition of lipid peroxidation, HRP, but not Cu2+ and AAPH, was still able to generate epitopes that were recognized by the same sera reacting with HRP-oxidized native (not probucol-enriched) LDL. In addition, the presence of autoantibodies against Cu(2+)- and AAPH-oxidized LDLs was negatively correlated with serum alpha-tocopherol concentration, whereas the titers against HRP-oxidized LDL did not exhibit any statistically relevant correlation with alpha-tocopherol levels. Together, these findings indicate that peroxidase(s)-dependent mechanisms can trigger peculiar lipid peroxidation-independent modifications of LDL in vivo.