Semliki forest virus-derived packaging system for production of retroviral vectors

Abstract: Gene therapy is a treatment of disease by transfer of genetic material into somatic cells of an individual. Successful gene therapy depends on a delivery system which can efficiently transfer the gene into target cells and lead to high level expression of the gene. Because retroviruses can integrate their genome into the chromosomes of infected cells, they have been used as vectors for stable expression of heterologous genes. Usually, the retroviral vectors are produced by transfection of vector DNA into a packaging cell line that synthesizes all retroviral proteins. One major drawback of this vector system is that it is limited to transduce the cDNA forms of processed mRNAs. These minigenes lack important control elements for gene expression and, therefore, often support low level of expression. For example, it is difficult to use the vector system for transduction of genes containing introns because any intron present in an engineered provirus will be removed from the transcribed RNA by the nuclear splicing machinery. In this thesis, I describe the development of a Semliki Forest virus (SFV)- derived packaging system for efficient production of retroviral vectors that can carry a foreign gene with an intron. We first inserted the gag-pol, ecotropic env and amphotropic env genes into individual SFV1 expression plasmids and transcribed the corresponding replication-competent SFV vector RNAs in vitro. The latter were introduced into the cytoplasm of BHK-21 cells by electroporation. We showed that the transfected cells synthesized a retroviral proteins, that is Gag proteins, Env proteins and viral enzymes, and that these were used for the assembly of mature virus like particles. To test if this system could produce infectious retroviral vectors, a recombinant Moloney murine leukemia virus genome containing the neomycin phosphotransferase gene (LTR-[psi]+-neoR-LTR) was inserted into SFV1 plasmid and the in vitro made RNA was cotransfected into BHK-21 cells together with the SFV1/gagpol RNA and SFV1/env RNA. We found that infectious retroviral vectors were produced from the transfected cells at titers up to 4 x 106 colony-forming units (cfu)/mil during a 5- hour incubation period. In this SFV-based system, the RNAs are synthesized in the cytoplasm of the cell, not in the nucleus. We reasoned that intron-containing genes can be packaged into retroviral vectors with the system. This was tested with a recombinant Molony murine leukemia virus genome containing the chloramphenicol. acetyltransferase (CAT) gene in association with an intron. When this RNA was cotransfected into BHK-21 cells together with the SFV1/gag-pol RNA and SFV1/env RNA, 1.3 X 106 cfu/ml were produced during a 5-hour incubation period. These vectors faithfully transduced the intron-containing gene into NIH 3T3 cells and HeLa cells, where the intron-CAT RNA was subjected to efficient splicing and used for high level expression. I conclude that we have developed a retroviral vector packaging system that is based on RNA replication in cell cytoplasm and can be used for transduction of intron-containing genes into mammalian cells. The fact that no helper virus could be detected in the virus stocks makes the system very interesting for gene therapy. In this thesis, I also present a study of human bone marrow stromal cells as target cells for retroviral-mediated gene transfer. Ninety one percent of the cells in a stromal cell culture were transduced by four rounds of retroviral vector infection without cytokine stimulation. The transduced stromal. cells could settle in the bone marrow of SCID mice and survive for at least two months. This suggests that human bone marrow stromal cells might be used as target cells for certain types of clinical gene therapy.