TOXICOL METHODS. Toxicology Methods 2001 ;11 (1):1-19
P.A. Jones, SEAC Toxicology Unit, Unilever Research Colworth, Colworth House, Sharnbrook, Bedford MK44 1LQ EML: [email protected]
In vitro testing for phototoxic potential using the epiderm 3-D reconstructed human skin model
An in vitro photocytotoxicity model using mouse 3T3 fibroblasts (the 3T3 NRU PT) has recently been validated for the detection of the phototoxic potential of test substances. Although the 3T3 NRU PT has been demonstrated to detect the phototoxic potential of both strong and weak phototoxins, irrespective of their aqueous solubility, in cases of low solubility a further confirmatory test may be required. In vitro, three-dimensional, reconstructed human skin models may fulfill this requirement. The purpose of this study was to develop a protocol for identifying the phototoxic hazard of test substances using the EpiDerm(trademark) skin model. The assay differed from published methods in that an ultraviolet (UV) B filter was used to provide a spectrum similar to that of sunlight and organic solvents (ethanol, dimethylsulfoxide, and acetone) were used as the dosing vehicles to apply substances directly to the culture surface. Chemicals of known in vivo phototoxicity and ingredients (of low solubility previously tested in the 3T3 NRU PT) were investigated. A testing strategy was developed to minimize the number of cultures required. First, the toxicity was assessed in the dark; then the phototoxicity of two concentrations - the maximum dark, nontoxic concentration (up to 100 mg/mL) and a 1/2-log10 dilution - was assessed. The criterion for phototoxicity was a 30% difference between dark and light toxicity at one of the concentrations tested. A reverse-UV protocol was also used to probe any additive effects compared with phototoxicity. This strategy correctly identified the known phototoxicity of the chemicals tested. Relative potency could also be identified by a dose/response approach. Ingredients were found to be nonphototoxic at the concentrations tested, which were between 10 and 10,000 times greater than those tested previously in vitro. The testing of product formulations was also investigated and was found to be feasible, provided that the correct control experiments were performed to avoid false-positive results.