PLASMID DNA VECTOR DESIGN AND PRODUCTION administered systemically. Here we report that a codon-optimized piggyBAC transposase generated higher transposition efficiency than an inactive mutant in mammalian cells. Following hydrodynamic tail-vein injection, we showed that firefly luciferase expression driven by the murine albumin enhancer/human alpha anti-trypsin promoter persisted up to eight months in C57/Blk6 liver. Luciferase expression persisted following fifty percent liver tissue resection in animals treated with codon-optimized transposase while expression in those treated with the inactive mutant dropped to background levels. This finding suggests the expression seen in the former group was due to integrated transgene. We extended these studies and delivered the human α anti-trypsin (hAAT) gene to murine liver and measured stable hAAT expression in serum by ELISA. Furthermore, we demonstrated that modifying the piggyBAC transposase by adding the DNA binding domain of the Lac Repressor to its C-terminus did not reduce transposition efficiency in mammalian cells. This is the first demonstration of in vivo gene transfer using the piggyBAC transposon system. This work has implications for liver-directed gene transfer and directing integration to specific genomic loci thereby reducing the risk of insertional mutagenesis. This work is supported by R01 HL 79023 and the Roy J. Coarver Charitable Trust.
744. A Novel Method for Analyzing Zinc Finger Nuclease Specificity In Vivo by LAM-PCR of Integrase Defective Lentiviral Vector (IDLV) Captured by DNA Double-Strand Breaks (DSB)
Richard Gabriel*,1 Angelo Lombardo*,2 Anne Arens,1 Pietro Genovese,2 Michael C. Holmes,3 Philip D. Gregory,3 Manfred Schmidt,1 Luigi Naldini*,2 Christof von Kalle*.1 1 National Center for Tumor Diseases, Heidelberg, Germany; 2 San Raffaele Telethon Institute for Gene Therapy, Milan, Italy; 3 Sangamo BioSciences, Inc., Richmond, CA.
Zinc Finger Nucleases (ZFN) target the formation of a DSB at a predetermined sequence in the genome with high efficiency. However, identification of potential ZFN off-target sites is confounded by DSB repair by non-homologous end-joining (NHEJ) – a rapid process that leaves no consistent signature of action. While the in vitro DNA binding specificity of a given ZFN has been successfully employed to prospectively determine potential sites of ZFN cleavage, an unbiased in vivo method for genome-wide detection of ZFN action is lacking. Here we present a potential solution to this problem. We hypothesized that IDLV accumulating in the nucleus may be captured and ligated into the genome via NHEJ at the site of ZFN-induced and naturally occurring DSBs, introducing the integrated IDLV as a stable genetic marker at a transient DSB. To test this approach, we treated K562 cells with IDLV encoding CCR5-targeting ZFN and GFP-expressing donor IDLV bearing sequence homology flanking the ZFN target site resulting in ∼50% of K562 cells expressing GFP. Molecular analysis confirmed targeting of either a single GFP expression cassette or vector concatemers within the CCR5 target site in the vast majority of the bulk population and in all single cell clones. Next we mapped off-target DSBs that harbored IDLV integration by determining genomic sequences flanking the IDLV LTR by LAM-PCR and deep-sequencing. IDLV integration, e.g. at the CCR5 target locus, was accompanied by micro and macro deletions of the ZFN cleavage site and LTRs consistent with repair via NHEJ. Analysis of >200 integration sequences revealed several sites of near complete or partial homology with the ZFN target site where IDLV had been joined into the genome in multiple instances, clustered in up to 5 kb around each homology locus. As expected, the most frequent of such common integration sites (CIS) was the CCR5 gene, in which integrated IDLV were found in a 3 kb region centered on the ZFN target site. Most of these integrated IDLV were close to the ZFN cleavage site. The sequences identified by LAM-PCR were compared to a dataset of in silico predicted CCR5 ZFN target sites S284
in the human genome. Importantly, several other CIS including the CCR2 and ABLIM2 genes, both previously described as partially homologous off-target sites for cleavage with the CCR5 targeted ZFN, were observed multiple times in different samples, serving to further validate these results. Although retrieval of IDLV integrations is influenced by the relative proportion of target sequences in the target populations and LAM-PCR efficiency, our results confirm and extend a hierarchy of CCR5 ZFN target sites previously defined by SELEX. Our studies demonstrate that NHEJ of DSB efficiently traps IDLVs, allowing for an unbiased genome-wide assessment of ZFN specificity in vivo that greatly facilitates optimization of ZFN design for gene therapy. *equal contribution.
745. Gene Therapy of Murine Mucopolysaccharidosis Type I Using the Sleeping Beauty Transposon System
Elena L. Aronovich,1 Jason B. Bell,1 Shaukat A. Khan,2 Lalitha R. Belur,1 Roland Gunther,3 Brenda L. Koniar,3 Josh B. Parker,4 Pankaj Gupta,2 Cathy S. Carlson,4 R. Scott McIvor,1 Perry B. Hackett.1 1 Genetics, Cell Biology and Development, University of Minnesota, Minneapolis, MN; 2Hematology/Oncology Section, VA Medical Center, Minneapolis, MN; 3Research Animal Resources, University of Minnesota, Minneapolis, MN; 4Masonic Cancer Center, University of Minnesota, Minneapolis, MN. The Sleeping Beauty (SB) transposon system, a non-viral vector that directs transgene integration into vertebrate genomes, is a promising candidate for gene therapy of mucopolysaccharidosis type I (MPS I), an inherited disease of glycosaminoglycan (GAG) metabolism, which is caused by deficiency of a lysosomal enzyme α-L-iduronidase (IDUA). SB transposons were constructed for high-level expression of human IDUA under the regulation of either CAGGS or a liver-specific ApoEhAAT promoter. Plasmids with the transposon contained, or not, SB transposase. Either 25 or 5 µg of plasmids were hydrodynamically injected into MPS I mice that were on a NOD/SCID background to eliminate complications of immune responses. In treated mice plasma IDUA activity persisted for 18 weeks (termination point) at levels up to several hundred-fold WT activity, depending on DNA dose and gender. There was a pronounced gender effect in IDUA levels, especially at the 5µg DNA dose, where IDUA activities and mRNA copy numbers in males were on average 30 times higher than in females, ±SB transposase, respectively. Expression continued from episomes in both sexes. Our data indicate that differences in IDUA activity between males and females reflect transcriptional regulation since the ratios of IDUA activity to mRNA copy number were within the same range in all groups. Neither gender nor dose of DNA affected the transposition process. The ApoEhAAT liver-specific promoter, which produced levels of IDUA comparable to those from the CAGGS promoter, eliminated the gender differences observed with the CAGGS promoter in the NOD/SCID MPS I mice. Mice treated with 25 µg CAGGS-IDUA transposons were studied for effects of treatment. IDUA activity was present in all examined somatic organs, and even in the brain, and correlated with both GAG reduction in these organs and level of expression in the liver, the target of transposon delivery. IDUA activity was higher in the treated males than in females. Clinical effects included correction of hepatomegaly, thickening of zygomatic arch and accumulation of foamy macrophages in bone marrow and synovium. Thus, by demonstrating correction of MPS I disease in mice, we have validated that the SB transposon system is an efficacious gene therapy vector for MPS I and is ready for testing in large animal models.
Molecular Therapy Volume 17, Supplement 1, May 2009 Copyright © The American Society of Gene Therapy