Evaluation of retroviral vector design in defined chromosomal loci by Flp-mediated cassette replacement

Successful retroviral vector construction is still empirical. Test systems for vector efficiency are based on statistical comparison of numerous infec tants with single proviral integrates, since their expression depends on th e chromosomal surrounding. More reliable data would be obtained if differen t vector constructs were studied in an identical chromosomal context. Here, we demonstrate the use of a new method, in which chromosomal sites are pro virally tagged in such a way that they can be targeted with other expressio n cassettes. The original tagging integrate is replaced in one step by the targeting element. This permits a reliable comparison of different retrovir al vector configurations, eliminating the influence of neighboring chromoso mal elements. We compared different retroviral vector types for coexpressio n of two genes: a vector containing an internal promoter and a vector with an internal ribosome entry site (IRES) element, In contrast to bicistronic retroviral vectors, dual-promoter proviruses exhibited rapid inactivation o f the long terminal repeat (LTR)-driven gene expression. Targeted exchange of the dual-promoter provirus with a bicistronic retroviral cassette result ed in gain of expression stability. The reverse experiment confirmed this p romoter interaction phenomenon since initial expression stability from a si ngle-promoter bicistronic provirus was lost by targeted exchange with a dua l-promoter cassette. In addition, targeting exchange of the dual-promoter p rovirus, replacing the LTR with an artificial (Tet) promoter restored expre ssion stability. These observations, valid for various integration sites, p rove the strong interaction between the LTR and the internal promoter. Our results have implications for retroviral vector design and suggest that ret roviral coexpression of two genes is more predictable in the bicistronic co nfiguration.