Gene expression of murine hematopoietic stem and progenitor cells transduced with integrating vectors

Gene therapy has emerged as a promising therapeutic strategy, but efficient in vitro assessment of vector safety remains challenging. To create a predictor of vector induced genotoxicity, we developed the Surrogate Assay for Genotoxicity Assessment (SAGA), which uses genetic algorithm-enhanced machine learning to detect the deregulation of a specific gene expression signature in murine hematopoietic stem and progenitor cells (HSPC) transduced with potentially genotoxic vectors. For the SAGA assay, murine lineage negative (Lin-) HSPC are isolated and transduced with high multiplicities of infection (MOI) to reach at least 3 vector copies per cell. Post-transduction, bulk cultures are expanded for a total of 15 days in medium supplemented with cytokines promoting myeloid differentiation. Total RNA is prepared from the bulk cultures on day 15 and gene expression profiles are analyzed on microarrays.SAGA classifies samples based on their gene expression profile by using a support vector machine with a radial kernel on 9 features deduced by a genetic algorithm. SAGA is more robust, sensitive and faster than previous assays and reliably predicts a mutagenic risk for vectors that led to leukemia in clinical trials. Murine Lineage-negative bone marrow cells were isolated and transduced with integrating vectors followed by ex vivo culture for 15 days. RNA was hybridized to a custom versions of the Agilent Whole Mouse Genome Oligo Microarray 4x44K v2. All arrays used in this study represent refined versions of the Agilent Whole Mouse Genome Oligo Microarray 4x44K v2 (Design ID 026655) comprised of all probes of this array in quadruplicates, but a different number of custom control probes. For the Gene expression matrix only the original Agilent Probes (A_XX_XXXXX) were used, which are the same on all arrays. ID-048306 (GPL19795) was developed by the Research Core Unit Genomics (RCUG) of Hannover Medical School. All non-control probes of design ID 026655 were printed four times within one 180K region. Design ID 066423 (048306On1M) and ID-084107 (048306On1M_V3 /GPL24228) were also developed by RCUG, using a 1x1M design format. All non-control probes of design ID 026655 were printed four times within a region comprising a total of 181560 Features. Four of such regions were placed within one 1M region. We have assigned all of these samples to GPL24228, a platform which contains the probe data consistent across all revised versions of Design ID 026655. The Agilent Design ID for each sample is available in the 'description' field and in the header of the raw data file.