Acetaldehyde, the major metabolite of ethanol, reacts with lysine and other free amino groups on proteins to form acetaldehyde-protein adducts. The presence of antibodies which recognize such acetaldehyde-protein adducts in sera from alcoholics has been attributed to an immune response to such adducts. Complicating this conclusion is the finding that sera from non-alcoholic control subjects also contain antibodies which recognize acetaldehyde-protein adducts. In the current research we sought to determine whether antibodies which recognize epitopes formed by the reaction of a protein with acetaldehyde can be formed in response to a protein modified with a structurally related protein adduct. We modified lysine residues on apolipoprotein (apo) B-100 with acetaldehyde and formaldehyde under reducing conditions, to form epsilon-N-methyl- and epsilon-N-ethyl-lysine residues, and with acetic anhydride to form epsilon-N-acetyl-lysine residues, and made antibodies against these modified proteins in guinea-pigs. In ELISA assays antibodies made against methylated apoB-100 (Me-apoB) cross-reacted effectively with ethylated apoB-100 (Et-apoB), while antibodies made against acetic anhydride-modified apoB-100 did not cross-react. We conclude that methyl-lysine shares one or more immunoreactive epitopes with ethyl-lysine, and that antibodies which recognize acetaldehyde-modified proteins can be formed in response to formaldehyde-modified proteins. We demonstrate that sera from both alcoholics and non-drinkers contain antibodies which recognize Me-apoB and Et-apoB and that the titres of these antibodies are comparable. These data raise the possibility that some human serum antibodies which recognize acetaldehyde-modified protein epitopes may have been made against formaldehyde-modified protein epitopes.(ABSTRACT TRUNCATED AT 250 WORDS)