Excited-State Photophysics of Hypericin and Its Hexamethoxy Analog: Intramolecular Proton Transfer as a Nonradiative Process in Hypericin

Liposome binding constants and singlet oxygen quantum yields of hypericin, tetrahydroxy helianthrone and their derivatives: studies in organic solutions and in liposomes

The spectroscopy and photophysics of several hypericin and helianthrone derivatives were studied in methanol and when bound to liposomes. The singlet oxygen quantum yields (phi(delta)) were measured indirectly relative to Rose Bengal and hematoporphyrin IX, employing 9,10-dimethylanthracene as a singlet oxygen trap. Hypericin was found to have a phi(delta) of 0.39+/-0.01 in methanol, and 0.35+/-0.05 in lecithin vesicles, in agreement with literature values. A heavy atom effect was evident upon bromination, resulting in phi(delta) for tetrabromohypericin of 0.72+/-0.02, presumably due to enhanced intersystem crossing. Elimination of the anionic hydroxyls by methylation also enhanced phi(delta) to 0.81+/-0.01. Conversely, addition of anionic sulfate groups drastically reduced phi(delta) resulting in phi(delta)'s of 0.12+/-0.01, 0.052+/-0.003 and 0.40+/-0.01 for hypericin disulfonate, hypericin tetrasulfonate and hexamethyl hypericin tetrasulfonate, respectively. The non-sulfonated helianthrones exhibited low phi(delta)'s in solution. The liposome binding constants, Kb, were measured using a spectroscopic assay. Except for hexamethyl hypericin, all non-sulfonated compounds bound well with Kb's ranging from 15.5+/-0.1 to 48.7+/-3.9 (mg/ml)(-1). None of the tetrasulfonated compounds bound, however the hypericin disulfonate had a Kb of 4.1+/-0.2 (mg/ml)(-1). The phi(delta)'s of the compounds capable of binding were measured and, in the case of the hypericin derivatives, were found not to vary dramatically from those in the free state. Liposome-bound helianthrone and dimethyl tetrahydroxy helianthrone both exhibited high phi(delta)'s, i.e. >0.5. The variations in binding constant and sensitization efficiencies are explained in conjunction with the molecular structure. The relevance of the above data to photodynamic therapy is briefly discussed.

Multidimensional reaction coordinate for the excited-state H-atom transfer in perylene quinones: importance of the 7-membered ring in hypocrellins A and B

The excited-state intramolecular H-atom transfer reactions of hypocrellins B and A are compared by using time-resolved absorption and fluorescence upconversion techniques. The hypocrellin B photophysics are well described by a simple model involving one ground-state species and excited-state forward and reverse H-atom transfer with a nonfluorescent excited state. We suggest that excited-state conformational changes are coupled to the H-atom transfer in hypocrellin B just as gauche/anti changes are coupled to the H-atom transfer in hypocrellin A.

The role of oxygen in the antiviral activity of hypericin and hypocrellin

The light-induced antiviral activity of hypericin and hypocrellin in the presence and absence of oxygen was examined under experimental conditions where the effect of oxygen depletion could be quantified. There was a significant reduction of light-induced antiviral activity of hypericin and hypocrellin under hypoxic conditions. Interestingly, antiviral activity of hypocrellin was not observed at low oxygen levels at which hypericin retained measurable virucidal activity. This suggests that additional pathways, such as the generation of protons from excited states of hypericin, may enhance the biological activity of activated oxygen species.

Hypocrellin A photosensitization involves an intracellular pH decrease in 3T3 cells

The fluorescent pH probe carboxy-seminaphtorhodafluor-1 (C-Snarf-1) has been used for laser microspectro-fluorometric assays of intracellular pH in 3T3 mouse fibroblasts treated with hypocrellin A. These results are compared to those previously obtained with the structurally related hydroxylated polycyclic quinone, hypericin (Sureau et al., J. Am. Chem. Soc. 118, 9484-9487, 1996). A mean local intracellular pH drop of 0.6 units has been observed in the presence of 1 microM hypocrellin A after 90 s of exposure to 0.1 microW of laser irradiation at 514.5 nm. The time evolution of the cytoplasm acidification for hypocrellin A-treated cells is faster than that for cells treated by hypericin. Thus, release of protons from an excited state of hypocrellin A appears to be more efficient than that from hypericin. In addition, the pH dependence of the quenching of C-Snarf-1 fluorescence in 3T3 cells under continuous irradiation has been observed. It is shown here that under continuous illumination, a pH decrease is able to induce a modification of the intracellular binding equilibrium of C-Snarf-1 that results in an increase of C-Snarf-1 fluorescence intensity. This latter observation suggests that the protons generated upon the photoexcitation of hypericin or its analogs may be involved in the production of other photoreactive species. Finally, we suggest that, just as for hypericin, this pH drop may be involved in the antiviral and antitumor activity of hypocrellin A.

Antioxidant enzyme response to hypericin in EMT6 mouse mammary carcinoma cells

Antioxidant enzyme activities were measured following exposure to hypericin +/- irradiation in EMT6 cells. CuZnSOD and catalase activities peaked within 0.5 h following irradiation for nontoxic 0.5 microM hypericin and toxic 1.0 microM hypericin. Catalase remained elevated up to 3 h for 1.0 microM hypericin + light. MnSOD activity was elevated immediately following irradiation for both doses. These levels returned to control by 1 h for 0.5 microM hypericin, but were depressed after 1 h for 1.0 microM hypericin. This suggests that mitochondria impairment may be a critical factor in hypericin phototoxicity. Glutathione reductase was inhibited immediately following irradiation with 1.0 microM hypericin, suggesting that an altered status of the glutathione pool contributed to cytotoxicity. Glutathione peroxidase activities were elevated following irradiation but returned to control levels within 0.5 h for both doses, implicating hydroperoxide formation as an early event in hypericin phototoxicity. Inhibition by hypericin in the dark was demonstrated for purified CuZnSOD, Se-dependent glutathione peroxidase, glutathione S-transferase, and glutathione reductase activities in vitro. Irradiation did not potentiate hypericin-mediated glutathione reductase inhibition and decrease inhibition for the other enzymes. Collectively, these data demonstrate an antioxidant enzyme response to hypericin photoactivation and confirm a role for oxygen in hypericin phototoxicity.