573. Investigation into the Frequency of Post-Integrative Gene Silencing Using a Non-Selective Sleeping Beauty Transposition System

The Sleeping Beauty (SB) transposon represents an important vehicle for in vivo gene delivery because it can stably integrate into the genomes of mammalian cells. Although SB is the most efficient non-viral integrating system described to date, all previous estimations of SB's integrative potential have relied on the use of antibiotic selection schemes. Previous work in other integrating systems (e.g. retroviruses) has demonstrated significant post-integrative gene silencing, leading us to investigate how genomic integration influences transposon expression in human cells. To do this we devised a new strategy to monitor SB integration that is not dependent on transgene expression. We constructed a single plasmid (pT/RSV-YFP.CMV-SB) containing both a YFP-marked transposon and a separate transposase expression cassette (external to the transposon), then transfected the plasmid into HeLa cells. Shortly thereafter, cells were single-cell sorted by FACS to generate approximately 100 clonal cell lines. After 8 weeks of growth (>50 cell divisions) 30% of the cell lines continue to show stable transgene expression, as determined by fluorescence microscopy and FACS analyses. We have screened these cell lines for the presence of transposon sequences by both PCR and Southern blot analyses, using various restriction enzymes to determine (i) the copy number and (ii) the molecular forms of the transposable element. Results indicate approximately 50% of the total cell lines contain an integrated transposon sequence(s), of which roughly 20% show little or no YFP expression. Our early results, based upon Hpa II (a methylation sensitive restriction enzyme) cleavage of genomic DNA, suggested the lack of YFP expression was not due to DNA methylation but to some other regulatory system. However, more recent data (including the use of the methylation specific McrBC enzyme) indicates that some of the integrated transposons do in fact become methylated over time. Possibly related, most individual cell lines show heterogeneous YFP levels throughout the cell population. Using aphidicolin (a cell cycle inhibitor) we determined that the varied YFP expression is not cell cycle dependent. We are continuing to explore post-integrative gene expression/silencing using a variety of approaches including a detailed analysis of how CpG concentration within the transposon expression cassette affects long-term gene expression. Overall, our transposon clones are proving to be an important new reagent to investigate the mechanism(s) by which integrating genetic elements (e.g. transposons and retroviruses) are silenced by the host cell, and should further our understanding of transcriptional silencing of integrating gene transfer vectors.