291. A Novel Safety Assay for Retroviral Vectors That Reproduces Lmo2 Proto-Oncogene Insertional Activation Events

Gene therapies for X-linked severe combined immunodeficiency (SCID-X1) and Wiskott-Aldrich syndrome (WAS) with gammaretroviral vectors have caused a number of cases of acute lymphoid leukemias due to insertional activation of oncogenes, mostly LMO2. Transduction of murine bone marrow hematopoietic cells and subsequent culture for myeloid immortalization or transplant into recipient mice detected the oncogenic activity of the spleen focus-forming virus (SFFV) vector, which was used in the WAS trial, but failed to detect the oncogenic activity of the SCID-X1 gene therapy vector MFG-gc. The lack of the most relevant Lmo2 insertion by gammaretroviral vectors in these assays demands more relevant and sensitive vector safety assays. It has been shown that murine early thymic progenitor cells (ETP), when cultured on OP9-DL1 stroma cells, progress through the DN1, DN2, DN3, DN4, DP and SP differentiation stages, and that overexpression of LMO2 in ETP cells caused a differentiation block at the DN2 stage (CD4-CD8-CD25+CD44+). We tested whether transduction of ETP cells with gammaretroviral or lentiviral vectors could reproduce Lmo2 integrations and induce DN2 differentiation block. We tested three different gamma-retroviral vectors MSCV-GFP, SFFV-GFP and MFG-gc and two lentiviral vectors Cl20-SFFV-mCherry and Cl20i4r-SFFV-mCherry. The two self-inactivating lentiviral vectors contain a single internal SFFV long terminal repeat but differ in that one contains the cHS400 chromatin insulators. We transduced ETP cells with these vectors and cultured them on OP9-DL1 cells for up to 40 days. The range of vector copy number was 1.07-10.78 for the MSCV-GFP vector, 0.24-1.22 for the SFFV-GFP vector, 0.59-7.98 for the MFG-gc vector, 3.97-13.04 for the Cl20-SFFV-mCherry vector and 4.57-18.21 for the Cl20i4r-SFFV-mCherry vector at days 10-11. Between days 20-35, a distinct DN2-blocked cell subpopulation was clearly detected by flow cytometry in all the 20 gamma-retroviral groups and also in 3/9 lentiviral groups. None of the 6 mock groups had any evidence of DN2 blockade. Insertion site mapping of sorted DN2 cells showed that in the 10 MSCV-GFP groups, 3 had Lmo2 insertions, 3 had Mef2c insertions, and 3 had both Lmo2 and Mef2c insertions; in the 6 SFFV-GFP groups, 3 had Lmo2 insertions; in the 4 MFG-gc groups, 3 had Lmo2 insertions; both Cl20-SFFV-mCherry groups had Mef2c insertions and the one Cl20i4r-SFFV-mCherry group had Lmo2 insertion. These insertions occurred either in the introns or within +/- 50kb window of the gene. Transplant of the DN2-blocked cells from selected MSCV-GFP groups into recipient mice led to acute lymphoid leukemias. These results show that the transduction of ETP cells and the resultant DN2 blockade represent a more relevant and sensitive assay for vector safety assessment. Our result also suggests that an ETP-like subpopulation may be present in the bone marrow of SCID-X1 and WAS patients and explain the enhanced propensity of these disorders to oncogenic transformations.