Evaluation of in vitro reporter gene induction assays for use in a rapid prescreen for compound selection to test specificity in the Tg.AC mouse short-term carcinogenicity assay.

Abstract

Under ICH guidelines, short-term carcinogenicity assays such as the Tg.AC assay are allowed alternatives for one species in the 2-year rodent bioassay. The Tg.AC transgenic mouse, which carries the v-Ha-ras oncogene under control of the zeta-globin promoter, develops skin papillomas in response to dermal application of carcinogens and tumor promoters. The appropriate specificity of the Tg.AC model for testing pharmaceuticals has not been systematically evaluated. The selection of candidate test compounds among noncarcinogenic pharmaceuticals would be aided by a high-throughput in vitro prescreen correlative of activity in the in vivo Tg.AC assay. Here we describe the development of a prescreen based on correct response to 24 compounds tested previously in Tg.AC mice. The in vitro prescreens, chosen to reflect molecular pathways possibly involved in Tg.AC papilloma formation, consisted of a zeta-globin promoter-luciferase construct stably expressed in K562 cells (Zeta-Luc) and three of the stress-response element-chloramphenicol acetyltransferase (CAT) fusion constructs stably expressed in HepG2 cells that are part of the CAT-Tox (L)iver assay. The stress response elements chosen were the c-fos promoter, the gadd153 promoter, and p53 response element repeats. Of the four assays, the gadd153-CAT assay showed the strongest concordance with activity in the Tg.AC assay, correctly classifying 78% of Tg.AC positive and 83% of Tg.AC negative compounds. The correlation was further improved by adding the Zeta-Luc assay as a second-stage screen. These cell-based assays will be used in a novel approach to selecting candidate compounds that challenge the specificity of the Tg.AC assay toward pharmaceuticals.