Oxidant membrane injury by the neutrophil myeloperoxidase system. I. Characterization of a liposome model and injury by myeloperoxidase, hydrogen peroxide, and halides.

Neutrophils and other phagocytes can injure cells by means of oxygen-dependent mechanisms, particularly the myeloperoxidase (MPO)-H2O2-halide system. The extent of such damage depends in part on the antioxidant defenses of the target cell. To facilitate the study of this phenomenon, we developed a model system in which we employed liposomes as targets for the myeloperoxidase system. The most useful species of liposomes employed 51Cr as the aqueous space marker and phosphatidyl choline with or without dicetyl phosphate and cholesterol as the structural lipid. Marker entrapment was established on the basis of 1) resolution of free from lipid-associated 51Cr by gel exclusion chromatography, 2) latency of 51Cr on rechromatography of detergent-treated liposomes, and 3) a correlation between entrapment and surface charge density. Exposure of liposomes to the complete MPO system resulted in release of 50 to 75% of the entrapped 51Cr. Release was abrogated by omission of myeloperoxidase or H2O2, heating of MPO, or addition of azide, cyanide, or catalase. Reagent H2O2 could be replaced by glucose plus glucose oxidase. Kinetic studies indicated a rapid process, lysis reaching half-maximal levels in less than 2 min. The addition of cyanide at various times interrupted lysis at once, indicating a requirement for ongoing myeloperoxidase-dependent reactions. Liposome disruption by the MPO system was pH dependent, increasing dramatically as pH was decreased from neutrality to 6.0. In the absence of halides, no lysis was observed. Maximum lysis was found with chloride at 10 to 100 mM, although at 1 mM concentrations, iodide, bromide, and thiocyanate were more active than chloride. Fluoride was inactive. Antagonism between halide species was demonstrated in that low concentrations of iodide or bromide inhibited the effect of optimal concentrations of chloride. Using 125I, we found that exposure of liposomes to the MPO system resulted in an association between iodide and liposomes; moreover, there was a close correspondence between this phenomenon and 51Cr release, suggesting that halogenation may be one mechanism of injury. These studies establish the usefulness of the liposome as a model of oxidant injury by a physiologically relevant system. They bear a striking parallel to work being done on MPO-mediated injury to eukaryotic and prokaryotic cells. By using this simplified model system, it should be possible to explore a number of determinants of target cell injury at a biochemical and molecular level.