Hypericin photosensitization in aqueous model systems.

Photosensitization of lysozyme, liposomes, ghosts and intact red blood cells (RBC) was investigated for aqueous hypericin. The effects of azide ion, 1,4-diazabicyclo(2.2.2)octane, and superoxide dismutase on photosensitized inactivation of lysozyme in 0.5% Triton X-100 indicate that singlet oxygen is the major inactivating intermediate with a contribution from superoxide. The singlet oxygen quantum yield (phi delta) scaled to methylene blue is 0.49 +/- 0.06 at monochromatic wavelengths from 514 to 600 nm. Relative values of phi delta based on lysozyme inactivation for different vehicles are: 0.5% Triton X-100 (1.13), human serum albumin (0.65), Cremophor-EL (0.76), Cremophor-RH40 (0.98), egg phosphatidylcholine (EPC) liposomes (0.04), hydrogenated soy phosphatidylcholine (HSPC) liposomes (< 0.01). Hypericin photosensitized lipid peroxidation of EPC liposomes and RBC ghosts. Extensive cross-linking of ghost membrane proteins occurred during the initial stages of lipid peroxidation. Prompt photohemolysis was used as the assay of RBC membrane damage. The photohemolysis curves are modeled with multihit target theory based on the "hit number" (n) and the target cross section (v). The values of v and the conventional "1/t50" parameter are equivalent determinants of the photohemolysis rate. The photohemolysis curves are in good agreement with n = 15 for incubation in phosphate-buffered saline at different hypericin concentrations and with additives. The measurements for other vehicles led to n = 19 for Cremophor-EL and n = 3 for EPC and HSPC liposomes, indicating that the kinetics of photohemolysis depend on the conditions of incubation.