High resolution dynamic maps of the timing of DNA replication using TimEX

Abstract

The development of lentiviral vectors greatly improved the prospects of gene therapy. Several groups have used such vectors to cure hemoglobinopathies in mouse models1. Nevertheless, viral integrations are random causing insertional mutagenesis, which remains a major concern. Insertional mutagenesis is mainly caused by enhancers used in the vectors that-regulate genes surrounding the integration site inducing cancers. Illustrating these risks, recent gene therapy clinical trials have revealed cases of leukemias induced by insertional mutagenesis. This risk can theoretically be decreased by blocking enhancer activity. This can be done by improving the design of the therapeutic cassettes, for instance using insulator elements. My project was to design cassettes for gene therapy for hemoglobinopathies; and to determine if insulator elements can be used to shield the site of integration from the influence of the regulatory elements present in the transgene. To answer that question, a strategy based on the Recombinase-Mediated Cassette Exchange (RMCE) methods was used to test various insulators in murine erythroleukemia cells (MEL). RMCE allows insertion of various cassettes in two different orientations at the same location in the genome. Cassettes containing an enhancer, the Locus Control Region or LCR, increases the level of transcription of genes surrounding the integration site called Random Locus 5 (RL5) in MEL. Using this system, we compared the enhancer blocking activities of various insulators.;We found that chicken HS4 (cHS4) was the only insulator with a significant ability to shield the integration sites from the effect of inserting a transgene, but even this element could only shield some of the genes in the flanking sequence. We also found that insertion of cHS4 in the vector decreased expression of the transgene and that the overall effect of this element was much more complex than what is suggested by existing models. This complexity precluded any form of generalization. We concluded from that part of the study that in the context of gene therapy, cHS4 has a potentially useful activity but that a much larger study would be necessary to determine if inclusion of this insulator in gene therapy vector would have an overall beneficial effect.;Previous work from the lab had shown that the level of expression of transgenes and their silencing depend on many factors including transcriptional interferences. It was also previously shown that at some of the sites of integration, the earliest epigenetic event preceding silencing was a switch of the timing of replication from early to late.;We therefore hypothesized that a map of the timing of replication genome-wide would help us understand some silencing events and maybe some of the effect of insulators on transgene expression.;We therefore developed methods to produce detailed maps of the replication timing and of primary transcription in mammalian using Nimblegen tiling array technology and massively parallel sequencing.;The methods that we have developed should have a wide range of application. The maps that we developed provided important insight as to the basic regulation of transcription in mammalian cells.;Our most important findings for this part of the studies reported in this thesis are that we have produced genome-wide high-resolution dynamic maps of the timing of replication in human erythroid, mesenchymal and embryonic stem cells using TimEX, a new method that we developed to measure DNA copy number variations during S phase with tiling arrays or massively-parallel sequencing.;We have shown that timing of replication is highly regulated, that origins are programmed to fire in defined narrow temporal windows during S phase and that many genomic segments are replicated in temporal transition regions devoid of initiation where replication forks progress unidirectionally from origins that can be hundreds of kilobases away.;We observed a strong inverse relationship between timing of replication and distance to the closest highly expressed gene. This relationship can be used to predict timing of replication profiles from expression data. This indicates that early origins of replication are preferentially located near expressed genes.