Previous studies have established the presence of high numbers of activated T lymphocytes and "aberrant" expression of major histocompatibility complex class II antigens by endothelial and smooth muscle cells in human atherosclerotic lesions, implicating the involvement of a local cellular immune response. The identity of the antigen(s) eliciting this immune response, the extent of their effect, and the atherogenic stage at which they occur remain to be determined. In the present studies, 120 normocholesterolemic New Zealand White rabbits were immunized one or more times with various antigens, with or without adjuvants. The antigens and adjuvants included human or rabbit atherosclerotic lesion proteins, ovalbumin, Freund's complete and/or incomplete adjuvants, recombinant mycobacterial heat shock protein 65 (hsp65), and two hsp-free adjuvants, Ribi complete adjuvant and lipopeptide. In addition, some groups received a high-cholesterol diet. Sixteen weeks after the first immunization the animals were killed, and arteriosclerotic lesions in the intima of the aortic arch were found to have developed only in those animals immunized with antigenic preparations containing hsp, either in the form of whole mycobacteria or as purified recombinant hsp65, although their serum cholesterol levels were normal. No arteriosclerotic changes exceeding those of controls were found in the other groups, irrespective of the antigen used. Immunohistopathologic examination revealed that the lesions contained 20% T cells, 10-30% macrophages, and 10-40% smooth muscle cells. Analysis of the peripheral blood T-lymphocyte proliferative responses revealed that the occurrence of lesions was positively correlated with the presence of hsp65-reactive T cells, suggesting that hsp65 is involved in the induction of arteriosclerotic lesions. Furthermore, combined immunization with hsp-containing material and a cholesterol-rich diet provoked development of significantly more severe atherosclerosis and the appearance of characteristic foam cells. We conclude that an (auto)immune response to hsp may initiate the development of atherosclerosis and that a high blood cholesterol level is only one albeit a very important risk factor.