Non-Random Integration and Clone Selection by Lentiviral SIN-LTR Vectors.

The need for safer gene therapy vectors was highlighted by the occurrence of three cases of retroviral vector-induced leukemia in children after the cure of severe combined immunodeficiency by gene therapy. These severe adverse events enhanced the development of new gene therapy vector systems, which aim to reduce influence of insertions on integrity and expression of genomic host DNA. Lentiviral vectors with self-inactivating (SIN) long terminal repeats (LTRs) have no LTR enhancer activity and are not expected to produce insertional gene activation. In our study we analyzed the integrations sites of three different lentiviral SIN-HIV-based- vectors. These vectors differed in their internal elements (Promotor, Transgene, WPRE) which allowed us to investigate whether internal elements are influencing target site selection and clone survival. A total of 1422 integration sites were analyzed at three different time points (1 day, 30 days, 60 days) after transduction of identical HeLa cells by LAM- PCR. Our results showed similar gene involvement and chromosomal distribution of integration sites for all vector types analyzed, independent of vector composition. 62–63% of the integrations were detected in Refseq genes, 82–86% were found within Refseq genes and their surrounding 10kb. Surprisingly, 271 of 1422 integration sites were clustered as common integration sites (CIS). Computer simulations allowed us to show that this high number of CIS was significantly different from a modeled random distribution of integration sites. Gene ontology analysis showed no difference in significantly overrepresented gene categories between the distinct vector types. Interestingly, we detected a time dependent increase or decrease in significance. Specific gene categories like phosphorylation, protein kinase or ATP binding activity increased in significance from freshly transduced cells (1 day) to 30 days. This observation was even more pronounced at the latest time point analyzed (60 days). Other gene categories showed the exactly opposite effect. Gene ontology analysis of freshly transduced cells showed a significant overrepresentation of genes involved in cell cycle regulation, whereas the analysis of the later time points did not show overrepresentation of this gene category. These results suggest that lentiviral SIN-HIV-based vectors may induce clonal selection in vitro independently of internal vector elements. The character of such effects as well as any putative relevance of such genotoxicity for the in vivo situation will have to be investigated in detail for the role of individual vector/cell type configurations.