102. Activation of the LMO2 Gene in Human Lymphoid Cells by rAAV-Mediated Targeted Insertion of a Single LTR Expression Cassette into the First Intron

Abstract

The pathogenic activation of the LMO2 proto-oncogene contributed to leukemia in two children otherwise successfully treated for Severe Combined Immunodeficiency (SCID) in a gene therapy trial (Science 302: 415-419, 2003). We have reproduced the retroviral vector insertion into the first intron of the LMO2 gene that occurred in one of the SCID patients in a human lympholeukemia cell line (Jurkat) using recombinant adeno-associated viral (rAAV) vector mediated gene targeting. The rAAV-LMO2 vector was constructed as follows. First, 1.3kb fragments containing LMO2 intron sequences from upstream and downstream of the retroviral insertion site in the SCID patient were amplified by PCR from human genomic DNA and sequenced to confirm identity to the genomic locus. Next, a gene cassette was designed to have the GFP coding sequences under the control of a single retroviral LTR that was used in the gene therapy trial followed by a polyadenylation signal. Then this 1.5kb single LTR-GFP cassette was placed in a reverse orientation between the 5' and 3' segments from LMO2 intron 1 and the assembled three fragments were cloned between AAV ITRs to derive the single stranded rAAV targeting vector. After packaging into serotype 1 capsid proteins, particles were purified by iodixanol gradient and gel filtration chromatography followed by ultrafiltration concentration. The Jurkat T cell line was transduced with rAAV-LMO2 particles at three multiplicities of infection (MOI). The percentage of GFP expressing cells was maximal (77%) at 3 days post-transduction and gradually decreased to stabilize at 2.2% at the highest MOI after 21 days. GFP expressing cells were selected by FACS from this cell population at 9 days post-transduction and clones derived from single cells were recovered. We initially screened 60 GFP-positive clones for targeting into the LMO2 locus and nine clones were positive by PCR using one primer in the GFP expression cassette and the other in the LMO2 intron just 3' to the region of homology with the targeting vector. These clones were analyzed by Southern hybridization using BglII or DraIII digestion with GFP and LMO2 probes. Four out of nine clones showed the predicted bands in addition to wildtype bands indicating successful gene targeting without rearrangement into one of the LMO2 alleles. Next, LMO2 expression was compared to that in control Jurkat cells by real-time quantitative RT-PCR using the comparative CT method. Expression of LMO2 mRNA was at a very low level in untransduced Jurkat cells but was highly upregulated in all 4 clones ranging from an 1100 to 3500-fold increase. Thus we have reproduced an LTR insertion at the exact location as occurred in one of the SCID patients and showed that a single LTR is sufficient to both maintain GFP expression and activate the upstream LMO2 promoter. This system should prove useful for evaluating the relative safety of specific retroviral vector designs with respect to the potential for protooncogene activation.