Redox Proteomics Sheds Light on Photodynamic Treatment of Cancer

Abstract

Cancer is the second most common cause of death among humans in the world, exceeded only by heart disease. One of the promising modalities for the treatment of cancer is photodynamic therapy (PDT). It is based on the concept that (1) certain light-sensitive compounds (photosensitizers) can be localized in neoplastic tissue, and (2) subsequently, these photosensitizers can be activated at an appropriate wavelength of light to generate reactive oxygen species, such as free radicals and singlet oxygen that are toxic for cells and tissues. Despite the wide acknowledgement of the susceptibility of proteins to oxidation, a thorough understanding of the involvement of protein oxidation in PDT-induced apoptosis is severely lacking. In this thesis we studied the effects of PDT on human cancer cells, with the emphasis on identification of PDT-oxidized proteins and uncovering their role in PDT-induced apoptosis. Two types of oxidative modifications, i.e. carbonylation and oxidation of protein thiols, are of particular interest. Carbonylation, i.e. the formation of carbonyl groups as a result of oxidative modifications of lysyl, arginyl, and prolyl residues, is an irreversible modification and usually serves as a marker of oxidative damage. In contrast, cysteine thiols, the most reactive natural amino acids, can be oxidized both reversibly and irreversibly. In this thesis we described the first application of a sensitive proteomic approach for identification of both carbonylated and oxidized thiol proteins in one system. We showed that subjection of tumor cells to PDT led to oxidative modifications of more than 300 cellular proteins, including cytosolic, nuclear, ER, mitochondrial, and other proteins. A large fraction of the identified proteins (281) underwent reversible thiol oxidation. At the same time, only 119 proteins were heavily oxidized with the formation of carbonyl groups. Hsp90, a protein that in response to PDT undergoes both carbonylation and oxidation of thiols, is of particular interest. It is overexpressed in many human tumors and functions by stabilizing and activating a set of diverse pro-survival proteins. We showed that PDT induces Hsp90 monomers to cross-link covalently with the formation of dimers, thereby disrupting the interactions between Hsp90 and Cdc37, Hsp90’s kinase targeting factor. Although an immediate PDT-induced disruption of Hsp90/Cdc37 complexes prevents the activation of newly synthesized anti-apoptotic substrates of Hsp90, we suggest that PDT-induced disruption of Hsp90/Cdc37 does not drastically contribute to the apoptotic cell death in vitro. The reason for that is a very fast nature of PDT-mediated apoptosis, which occurs within a time too short to significantly down-regulate Hsp90 substrates. This suggestion is in line with our observation that pre-incubation of cells with Hsp90 inhibitor, which decreases the levels of pro-survival Hsp90 clients, such as Akt and Raf-1, followed by a subsequent subjection of these cells to PDT, dramatically increased the apoptotic rates. Taken together, our data provide clear evidence that combination with Hsp90 inhibitors increases PDT-mediated apoptotic responses in vitro and may represent the general strategy for enhancing the efficiency of PDT cancer therapy as well as various other cancer treatment modalities in vivo.