Why is Rose Bengal More Phototoxic to Fibroblasts In Vitro Than In Vivo?

Abstract

Photosensitized protein cross-linking has been recently developed to seal wounds and strengthen tissue. Although the photosensitizing dye, Rose Bengal (RB), is phototoxic to cultured cells, cytotoxicity does not accompany RB-photosensitized tissue repair in vivo. We investigated whether the environment surrounding cells in tissue or the high irradiances used for photo-cross-linking inhibited RB phototoxicity. Fibroblasts (FB) grown within collagen gels to mimic a tissue environment and monolayer cultured FB were treated with RB (0.01-1 mm) and the high 532 nm laser irradiances used in vivo for tissue repair (0.10-0.50 W cm(-2)). Monolayer FB were substantially more sensitive to RB photosensitization: the LD50 was >200-fold lower than that in collagen gels. Collagen gel protection was associated with increased Akt phosphorylation, a prosurvival pathway. RB phototoxicity in collagen gels was 25-fold greater at low (0.030 W cm(-2)) that at high (0.50 W cm(-2)) irradiances. Oxygen depletion at high irradiance only partially accounted for the irradiance dependence of phototoxicity as replacing air with nitrogen only increased the LD50 by four-fold in monolayers. These results indicate that the lack of RB phototoxicity during in vivo tissue repair results from upregulation of prosurvival pathways in tissue cells, oxygen depletion and irradiance-dependent RB photochemistry.