Molecular mechanisms regulating the tumor-targeting potential of splice-activated gene expression

Previous studies have suggested that differences in the ability of normal and malignant cells to process certain alternatively spliced pre-mRNA transcripts can be exploited as a potentially powerful means of targeting the expression of therapeutic genes to tumor cells in vivo and in vitro. Specifically, it was shown that efficient processing of minigene constructs containing the alternatively spliced CD44 exons v9 and v10 only occurs in tumor cells that express CD44 isoforms that incorporate these exons (e.g. CD44R1). In the present study, efforts were made to define the molecular mechanisms that underlie the apparent specificity of this process. RT-PCR analysis and DNA sequencing were used to characterize the various splicing events that occur between CD44 exons v8, v9 and v10 following transfection of minigene constructs containing these various exons into CD44R1-positive (PC3) and CD44R1-negative (T24) cell lines. The results obtained confirm that although the v8-v9 intron is efficiently removed in both CD44R1-positive and CD44R1-negative cells, the corresponding v9-v10 intron is accurately spliced and the exons appropriately joined only in lines that express v10-containing CD44 isoforms (e.g. PC3). In CD44R1-negative cell lines (e.g. T24) alternative 5' and 3' splice sites located within the v9-v10 intron are preferentially used, resulting in various portions of the intron being retained within the final processed mRNA product. It is proposed that identification of these functionally important intronic sequence elements will facilitate the development of second generation "splice activated gene expression" vectors that may prove useful in various cancer gene therapy applications.