Development of non-integrating lentiviral vectors for safer gene therapy

During the last few years gene therapy has been used to treat children with adenosine deaminase (ADA) deficiency and common ?-chain, or X-linked, severe combined immunodeficiency (SCID). Retroviral vectors were used to deliver the therapeutic gene, which was able to integrate into the genome of host cells and cure the disease. Unfortunately, three children in a French trial for X-linked SCID have developed T cell leukaemia following the integration of retroviral DNA into the human genome. The leukemia in these patients was caused by the therapeutic gene integrating into a dangerous location in a host cell chromosome. Integration of the gene into this location caused an event called insertional mutagenesis and resulted in cell transformation and cancer. The outcome of this clinical trial emphasizes the need to develop safer gene therapy vectors. This can be achieved by preventing the integration of a therapeutic gene into an unwanted and dangerous area of the patients' genome. Instead of retroviruses, our group studies lentiviruses as gene therapy vectors. Lentiviruses can package large amounts of DNA, and unlike retroviruses they can infect non-dividing as well as dividing cells. These two properties mean that properly engineered lentiviruses will have vast potential to treat genetic diseases of many tissue and organ types. Like retroviruses, lentiviruses integrate into the patients own genome. This is beneficial because it means that long-term correction of a genetic defect can theoretically be achieved with just one administration of the lentivirus. However, as discussed above, there is a risk of disrupting normal DNA in the patient's genome. Therefore, we have proposed to engineer lentiviruses that eliminate this risk of insertional mutagenesis while retaining all the helpful properties that make lentiviruses such a promising vehicle for gene therapy. Insertional mutagenesis can be prevented by inhibiting integration of the lentivirus DNA into the patient's genome. Lentiviruses integrate by using an integrase protein. We are in the process of generating lentiviruses with an altered, ineffective integrase protein. These newly engineered lentiviruses will be tested for their ability to safely deliver genes into various non-dividing cell types. Since the cells do not divide, we expect the newly delivered gene to stay in the cell but not to integrate into the cell's genome. This therapeutic gene will be maintained as a separate piece of DNA that will not interfere with normal cellular DNA functions. We expect these non-integrating lentivirus vectors to be most useful in cells such as retina and muscle. Furthermore, since the therapeutic gene will be maintained indefinitely inside the cell, only one dose of the lentivirus is necessary. Such a system will translate into a safer and simpler approach than the currently available gene therapy techniques.

The lentiviral genome non-specifically integrates into the human genome and this can lead to insertional mutagenesis including cell transformation. We are developing strategies to eliminate the risk of insertional mutagenesis thereby improving the safety and efficacy of lentiviral gene therapy vectors. The main objective of this project is to engineer non-integrating lentiviral vectors that stably express a transgene in post-mitotic cells. Several integration-deficient vectors that contain mutations in the lentiviral integrase protein or mutations in the integrase attachment sites will be generated. Three integration-deficient lentiviral vectors have so far been developed: a D64V integrase catalytic mutant, an integrase attachment site mutant (delatt) and a double mutant containing both mutations, D64V/delatt. These mutations all inhibit integration but the vectors maintain expression of the transgene in vitro. Furthermore, eGFP expression from a vector containing the D64V mutation has been shown to persist in non-dividing cells for at least 6 months in vivo. We plan to generate at least three more vectors with mutations in integrase: W235E, N120K and K264,266,273R. The non-integrating vectors will be characterized by assessing the structure of the episomal lentiviral DNA and measuring long-term transgene expression. Calculating the level of integration will give an indication of the relative safety of the vectors in comparison to vectors containing the wild-type integrase. To test vector safety directly, a cellular transformation assay will be developed. This will involve transduction of primary cells in vitro and assessing cellular immortality. If the new vectors show reduced integration efficiency and still retain transgene expression, then the mutations will be used to engineer the next generation of vectors with double and triple mutations. Following in vitro analyses the vectors will be tested in vivo in a well-developed retinal model system.