Search

Transposons are mobile genetic elements that are found in nearly all organisms, including humans. Mobilization of DNA transposons by transposase enzymes can cause genomic rearrangements, but our knowledge of human genes derived from transposases is limited. In this study, we find that the protein encoded by human PGBD5, the most evolutionarily conserved transposable element-derived gene in vertebrates, can induce stereotypical cut-and-paste DNA transposition in human cells. Genomic integration activity of PGBD5 requires distinct aspartic acid residues in its transposase domain, and specific DNA sequences containing inverted terminal repeats with similarity to piggyBac transposons. DNA transposition catalyzed by PGBD5 in human cells occurs genome-wide, with precise transposon excision and preference for insertion at TTAA sites. The apparent conservation of DNA transposition activity by PGBD5 suggests that genomic remodeling contributes to its biological function. DOI: http://dx.doi.org/10.7554/eLife.10565.001, eLife digest Transposons are mobile genetic elements that can be cut out of and inserted into DNA. They are present in most living things and make up almost half of the human genome. Transposons help to rearrange and increase the variety of DNA sequences, which can drive evolution and regulate the expression of genes. Enzymes called transposases help to move transposons, but very few genes that encode these enzymes have been studied in humans. PiggyBac transposase—which was first discovered in the cabbage looper moth—helps to move transposons of the piggyBac family. Humans and many other animals have genes that encode similar enzymes. In particular, the gene that encodes the human PGBD5 transposase is expressed in the developing embryo and particular areas of the brain and is highly similar to genes found in other vertebrate animals. These intriguing features prompted Henssen et al. to investigate PGBD5. The experiments reveal that PGBD5 is able to move piggyBac-like transposons in human cells and insert them into sites that contain similar DNA sequences that are preferred by other PiggyBac transposases. Henssen et al. compared human PGBD5 to the piggyBac transposases from other organisms, including insects, bats, and frogs. They found that PGBD5 is deeply conserved among vertebrate organisms, and is distinct from other piggyBac transposases. These findings suggest that PGBD5 is indeed a fully working piggyBac transposase. Further work is needed to understand what portions of the human genome may be rearranged by PGBD5, and how this may contribute to human brain development or disease. DOI: http://dx.doi.org/10.7554/eLife.10565.002

The inherent risks associated with vector insertion in gene therapy need to be carefully assessed. We analyzed the genome-wide distributions of Sleeping Beauty (SB) and piggyBac (PB) transposon insertions as well as MLV retrovirus and HIV lentivirus insertions in human CD4(+) T cells with respect to a panel of 40 chromatin states. The distribution of SB transposon insertions displayed the least deviation from random, while the PB transposon and the MLV retrovirus showed unexpected parallels across all chromatin states. Both MLV and PB insertions are enriched at transcriptional start sites (TSSs) and co-localize with BRD4-associated sites. We demonstrate physical interaction between the PB transposase and bromodomain and extraterminal domain proteins (including BRD4), suggesting convergent evolution of a tethering mechanism that directs integrating genetic elements into TSSs. We detect unequal biases across the four systems with respect to targeting genes whose deregulation has been previously linked to serious adverse events in gene therapy clinical trials. The SB transposon has the highest theoretical chance of targeting a safe harbor locus in the human genome. The data underscore the significance of vector choice to reduce the mutagenic load on cells in clinical applications.

The DNA transposon piggyBac is a potential therapeutic agent for multiple genetic diseases such as cystic fibrosis (CF). Recombinant piggyBac transposon and transposase are typically codelivered by plasmid transfection; however, plasmid delivery is inefficient in somatic cells in vivo and is a barrier to the therapeutic application of transposon-based vector systems. Here, we investigate the potential for hybrid piggyBac/viral vectors to transduce cells and support transposase-mediated genomic integration of the transposon. We tested both adenovirus (Ad) and adeno-associated virus (AAV) as transposon delivery vehicles. An Ad vector expressing hyperactive insect piggyBac transposase (iPB7) was codelivered. We show transposase-dependent transposition activity and mapped integrations in mammalian cells in vitro and in vivo from each viral vector platform. We also demonstrate efficient and persistent transgene expression following nasal delivery of piggyBac/viral vectors to mice. Furthermore, using piggyBac/Ad expressing Cystic Fibrosis transmembrane Conductance Regulator (CFTR), we show persistent correction of chloride current in well-differentiated primary cultures of human airway epithelial cells derived from CF patients. Combining the emerging technologies of DNA transposon-based vectors with well-studied adenoviral and AAV delivery provides new tools for in vivo gene transfer and presents an exciting opportunity to increase the delivery efficiency for therapeutic genes such as CFTR.

The development of robust nonviral vectors could facilitate clinical gene therapy applications and may overcome some of the immune complications of viral vectors. Nevertheless, most nonviral gene deliver approaches typically yield only transient and/or low gene expression. To address these caveats, we have explored piggyBac transposons to correct hemophilia B by liver-directed factor IX (FIX) gene therapy in hemophilic mice. To achieve this, we combined the use of: (i) a hyperactive codon-optimized piggyBac transposase, (ii) a computationally enhanced liver-specific promoter, (iii) a hyperfunctional codon-optimized FIX transgene (FIX R338L Padua), and (iv) a modification of the transposon terminal repeats. This combination strategy resulted in a robust 400-fold improvement in vector performance in hepatocytes, yielding stable supraphysiologic human FIX activity (>1 year). Liver-specific expression resulted in the induction of FIX-specific immune tolerance. Remarkably, only very low transposon/transposase doses were required to cure the bleeding diathesis. Similarly, PB transposons could be used to express supraphysiologic factor VIII levels using low transposon/transposase doses. PB transposition did not induce tumors in a sensitive hepatocellular carcinoma-prone mouse model. These results underscore the potency and relative safety of the latest generation PB transposons, which constitutes a versatile platform for stable and robust secretion of therapeutic proteins.

The ability to direct gene delivery to a user-defined chromosomal location would greatly improve gene transfer applications. The piggyBac transposon system is a nonviral gene transfer system proven effective in a variety of cells and tissues, including human cells. We fused a highly site-specific synthetic zinc-finger DNA-binding domain (ZFP) to the N-terminus of the piggyBac transposase and evaluated site-directed genomic integration in human cells. Chimeric ZFP-piggyBac transposase exhibited robust gene transfer activity, targeted binding to a cognate endogenous chromosomal ZFP site in the human genome, and site-directed transposon integration into an episomal plasmid target containing a single ZFP site in human cells. We evaluated the ability of ZFP-piggyBac to direct gene integration into an engineered chromosomal ZFP target site in the human genome and consistently observed a higher degree of ZFP-piggyBac site-directed genomic integration when compared to native piggyBac. Chromatin immunoprecipitation (ChIP) experiments revealed binding of native piggyBac to our engineered TTAA-containing chromosomal target which supported integration, but not a TTAA-deficient chromosomal target which lacked integration. Our results offer insight into the requirements for using a chimeric zinc finger-piggyBac transposase to direct integration into a user-defined chromosomal location.

It has been previously shown that integrase-defective HIV-1-based gene vectors can serve, with moderate efficiency, as substrate for DNA transposition by a transiently expressed Sleeping Beauty (SB) transposase. Here, we describe the enhanced gene transfer properties of a HIV-1/SB hybrid vector that allows efficient DNA transposition, facilitated by the hyperactive SB100X transposase, from vector DNA intermediates in primary human cells. Potent transposase-dependent integration of genetic cargo carried by the hybrid HIV-1/SB vector (up to 160-fold above background) is reported in human cell lines as well as in primary human fibroblasts and keratinocytes. The efficiency of transgene integration in context of the newly developed hybrid vector is comparable with that of conventional lentiviral vectors (LVs). Integration profiles of integrating HIV-1-derived vectors and SB transposons mobilized from LVs are investigated by deep sequencing of a large number of integration sites. A significant bias of lentiviral integrations in genes is reported, confirming that biological properties of the viral integration machinery facilitate preferred insertion into actively transcribed genomic regions. In sharp contrast, lentiviral insertions catalyzed by the SB100X transposase are far more random with respect to genes. Based on these properties, HIV-1/SB vectors may become valuable tools for genetic engineering and therapeutic gene transfer.

In this study, we compared the genomic integration efficiencies and transposition site preferences of Sleeping Beauty (SB or SB11), Tol2, and piggyBac (PB) transposon systems in primary T cells derived from peripheral blood lymphocytes (PBL) and umbilical cord blood (UCB). We found that PB demonstrated the highest efficiency of stable gene transfer in PBL-derived T cells, whereas SB11 and Tol2 mediated intermediate and lowest efficiencies, respectively. Southern hybridization analysis demonstrated that PB generated the highest number of integrants when compared to SB and Tol2 in both PBL and UCB T cells. Tol2 and PB appeared more likely to promote clonal expansion than SB, which may be in part due to the dysregulated expression of cancer-related genes near the insertion sites. Genome-wide integration analysis demonstrated that SB, Tol2, and PB integrations occurred in all the chromosomes without preference. Additionally, Tol2 and PB integration sites were mainly localized near transcriptional start sites (TSSs), CpG islands and DNaseI hypersensitive sites, whereas SB integrations were randomly distributed. These results suggest that SB may be a preferential choice of the delivery vector in T cells due to its random integration site preference and relatively high efficiency, and support continuing development of SB-mediated T-cell phase I trials.

The stable introduction of therapeutic transgenes into human cells can be accomplished using viral and nonviral approaches. Transduction with clinical-grade recombinant viruses offers the potential of efficient gene transfer into primary cells and has a record of therapeutic successes. However, widespread application for gene therapy using viruses can be limited by their initially high cost of manufacture at a limited number of production facilities as well as a propensity for nonrandom patterns of integration. The ex vivo application of transposon-mediated gene transfer now offers an alternative to the use of viral vectors. Clinical-grade DNA plasmids can be prepared at much reduced cost and with lower immunogenicity, and the integration efficiency can be improved by the transient coexpression of a hyperactive transposase. This has facilitated the design of human trials using the Sleeping Beauty (SB) transposon system to introduce a chimeric antigen receptor (CAR) to redirect the specificity of human T cells. This review examines the rationale and safety implications of application of the SB system to genetically modify T cells to be manufactured in compliance with current good manufacturing practice (cGMP) for phase I/II trials.

Lentiviral vectors enter cells with high efficiency and deliver stable transduction through integration into host chromosomes, but their preference for integration within actively transcribing genes means that insertional mutagenesis following disruption of host proto-oncogenes is a recognized concern. We have addressed this problem by combining the efficient cell and nuclear entry properties of HIV-1–derived lentiviral vectors with the integration profile benefits of Sleeping Beauty (SB) transposase. Importantly, this integration enzyme does not exhibit a preference for integration within active genes. We generated integrase-deficient lentiviral vectors (IDLVs) to carry SB transposon and transposase expression cassettes. IDLVs were able to deliver transient transposase expression to target cells, and episomal lentiviral DNA was found to be a suitable substrate for integration via the SB pathway. The hybrid vector system allows genomic integration of a minimal promoter-transgene cassette flanked by short SB inverted repeats (IRs) but devoid of HIV-1 long terminal repeats (LTRs) or other virus-derived sequences. Importantly, integration site analysis revealed redirection toward a profile mimicking SB-plasmid integration and away from integration within transcriptionally active genes favored by integrase-proficient lentiviral vectors (ILVs).

Udgivelsesdato: 2008-Nov-4 The Sleeping Beauty (SB) transposon system represents an important alternative to viral integrating vector systems but may, as its viral counterparts, be subject to transcriptional silencing. To investigate shielding of SB-delivered transgene cassettes against transcriptional repression, we establish silencing assays in which SB vector-containing F9 murine teratocarcinoma cell clones are identified by strategies that include or exclude selection for transgene expression. Among clones carrying one or more SB transposon vectors, more than one-third are immediately silenced, and most of the remaining clones move toward silencing during prolonged passage. In line with the lack of an intrinsic ability of SB to resist silencing, we show that the stable transfection rate of SB vectors in F9 cells is significantly improved by flanking the transgene with heterologous 5'-HS4 chicken beta-globin (cHS4) insulators. In approaches based on drug selection and subsequent flow-cytometric detection of transgene expression, clones containing cHS4-insulated vectors are to a much higher degree protected against transcriptional silencing, resulting in long-term expression of the fluorescent marker. Our findings demonstrate that SB vectors, prone for transcriptional silencing by positional effects in F9 cells, are protected by insulators. We believe that insulated SB-derived vectors will become useful tools in transposon-based transgenesis and therapeutic gene transfer.Molecular Therapy (2008); doi:10.1038/mt.2008.224.

The Sleeping Beauty (SB) transposable element shows efficient transposition in human cells, and provides long-term transgene expression in preclinical animal models. Random chromosomal insertion of SB vectors represents a safety issue in human gene therapeutic applications, due to potential genotoxic effects associated with transposon integration. We investigated the transcriptional activities of SB in order to assess its potential to alter host gene expression upon integration. The untranslated regions (UTRs) of the transposon direct convergent, inward-directed transcription. Transcription from the 5'-UTR of SB is upregulated by the host-encoded factor high-mobility group 2-like 1 (HMG2L1), and requires a 65-base pair (bp) region not present in commonly used SB vectors. The SB transposase antagonizes the effect of HMG2L1, suggesting that natural transposase expression is under a negative feedback regulation. SB transposon vectors lacking the 65-bp region associated with HMG2L1-dependent upregulation exhibit benign transcriptional activities, at a level up to 100-times lower than that of the murine leukemia virus (MLV) long terminal repeat (LTR). Incorporation of chicken beta-globin HS4 insulator sequences in SB-based vectors reduces the transactivation of model promoters by transposon-borne enhancers, and thus may lower the risk of transcriptional activation of host genes situated close to a transposon insertion site.

Spliceosome-mediated RNA trans-splicing (SMaRT) is an emerging technology for the repair of defective pre-messenger RNA (pre-mRNA) molecules. It is especially useful in the treatment of genetic disorders involving large genes. Although viral vectors have been used for achieving long-lasting expression of trans-splicing molecules, the immunogenicity and suboptimal safety profiles associated with viral-based components could limit the widespread application of SMaRT in the repair of genetic defects. Here, we tested whether the non-viral Sleeping Beauty (SB) transposon system could mediate stable delivery of trans-splicing molecules designed to correct the genetic defect responsible for severe combined immune deficiency (SCID). This immunological disorder is caused by a point mutation within the 12.4 kilobase (kb) gene encoding the DNA protein kinase catalytic subunit (DNA-PKcs) and is associated with aberrant DNA repair, defective T- and B-cell production, and hypersensitivity to radiation-induced injury. Using a novel SB-based trans-splicing vector, we demonstrate stable mRNA correction, proper DNA-PKcs protein production, and conference of a radiation-resistant phenotype in a T-cell thymoma cell line and SCID multipotent adult progenitor cells (MAPCs). These results suggest that SB-based trans-splicing vectors should prove useful in facilitating the correction of endogenous mutated mRNA transcripts, including the DNA-PKcs defect present in SCID cells.

Transposons are natural gene delivery vehicles. The Sleeping Beauty (SB) transposon shows efficient transposition and long-term transgene expression in the cells of vertebrates including humans. SB transposition into chromosomal DNA occurs in a fairly random manner. This is clearly not desirable in human gene therapeutic applications because there are potential genotoxic effects associated with transposon integration. In this study we set out to manipulate the selection of SB's target sites for targeted transposition into predetermined chromosomal regions. We evaluated experimental strategies based on engineered proteins composed of DNA-binding domains fused to (i) the transposase; (ii) another protein that binds to a specific DNA sequence within the transposable element; and (iii) another protein that interacts with the transposase. We demonstrated targeted transposition into endogenous matrix attachment regions (MARs) and a chromosomally integrated tetracycline response element (TRE) in cultured human cells, using targeting proteins that bind to the transposon DNA. An approach based on interactions between the transposase and a targeting protein containing the N-terminal protein interaction domain of SB was found to enable an approximately 10(7)-fold enrichment of transgene insertion at a desired locus. Our experiments provide proof-of-principle for targeted chromosomal transposition of an otherwise randomly integrating transposon. Targeted transposition can be a powerful technology for safe transgene integration in human therapeutic applications.

Liver-targeted gene therapy using adeno-associated virus-based vectors (rAAV) shows therapeutic promise in animal models and adult-focused clinical trials. This promise, however, is not directly translatable to the growing liver where rapid clearance of rAAV episomal genomes occurs in concert with hepatocellular proliferation. This phenomenon has been shown to underpin progressive loss of therapeutic efficacy, initially achieved in the neonatal period, in mouse models of early-onset urea cycle defects (UCDs). We have developed a hybrid rAAV/piggyBac transposon vector system combining the highly efficient liver-targeting properties of rAAV with stable piggyBac-mediated transposition of the transgene into the hepatocyte genome. Transposition efficiency was first tested using an EGFP expression cassette following delivery to newborn wild-type mice, with a 20-fold increase in stably gene-modified hepatocytes observed 4 weeks post-treatment compared to traditional rAAV gene delivery. We then modeled the therapeutic potential of the system in the context of severe early onset UCDs. A single treatment in the perinatal period was sufficient to confer robust and stable phenotype correction in the ornithine transcarbamylase (OTC) deficient Spfash mouse and neonatal lethal argininosuccinate synthetase (ASS) knock-out mouse. The system was next utilized in a murine model of chronic liver disease, progressive familial intrahepatic cholestasis type 3 (PFIC3), in which intervention prior to disease onset is necessary due to the development of chronic liver pathology which impedes vector delivery and hepatocyte transduction. Stable persistent expression extending into adulthood was achieved with increased mean biliary phosphatidylcholine concentration and dramatically reduced liver fibrosis. Finally, transposon integration patterns were analysed, revealing 127,286 unique integration sites which conformed to previously published piggyBac data. Conclusion: Using a hybrid rAAV/piggyBac transposon vector system we achieved stable therapeutic protection in two UCD mouse models and a murine model with chronic liver pathology. This strategy has the potential for treating a wide array of inherited liver diseases with early onset. A clinically conceivable early application of this technology in the management of severe UCDs could be as a bridging therapy while awaiting liver transplantation. Further improvement of the system will result from the development of highly human liver-tropic capsids, the use of alternative strategies to achieve transient transposase expression, and engineered refinements in the safety profile of piggyBac transposase-mediated integration.

New generations of plasmid vectors used for non-viral gene therapy have a reduced size, thus allowing an increase in transfection efficiency and expression levels. We designed a novel vector, called pFAR4, which is Free of Antibiotic Resistance markers. In bacteria, the production of pFAR4 derivatives relies on a plasmid-borne function that suppresses a nonsense mutation introduced into an essential Escherichia coli gene.Thanks to its reduced size, pFAR4 appears to be an efficient gene vector as it displays a superior transgene expression level either after transfection of cultured B16 melanoma cells or electrotransfer into skin or tumors. Furthermore, hydrodynamic delivery into liver revealed that pFAR4 allowed a prolonged expression level, in comparison with kanamycin resistant plasmids that promote transgene silencing.We took advantage of the pFAR4 superiority in liver to assess a novel therapeutic approach for the Mucopolysaccharidosis type IIIA (MPS-IIIA) or Sanfilippo A syndrome. MPS-IIIA is a lysosomal storage genetic disease which results from the deficiency of the N-sulfoglucosamine sulfohydrolase (SGSH) protein, a sulfamidase required for the degradation of heparan sulfate glycosaminoglycans (GAGs). The accumulation of these macromolecules leads to somatic organ pathologies, severe neurodegeneration and patients’ death. In MPS-IIIA mice, hydrodynamic delivery of a pFAR4 derivative expressing the deficient enzyme from a liver-specific promoter allowed high and prolonged serum levels of sulfamidase protein that was efficiently taken-up by neighboring organs after its endocytosis by mannose 6-phosphate receptors-expressing cells. This led to the correction of GAG accumulation in all peripheral organs as well as in the brain, albeit at early stages of the disease. Improvement of our approach is provided by fusing the SGSH protein with peptides mediating transcytocis across the blood-brain barrier.Having established that pFAR4 is a potent expression vector in a non-integrative approach, our next step consisted in combining the pFAR4 technology and the Sleeping Beauty (SB) transposon system, which is composed of a transposase that excises the transgene-containing transposon from a donor plasmid and mediates its integration into the host cells’ genome. The reduced size of pFAR4 allowed an efficient delivery of SB components into human cells. Our current goal is to take the pFAR4 and Sleeping Beauty technologies to the clinic, in the context of an ex vivo gene therapy approach involving the transfection of autologous primary pigment epithelial cells for the treatment of neovascular age-related macular degeneration disease, in two phase I/II clinical trials.

Since the first human Gene Therapy clinical trial was initiated 25 years ago, many technological advances have dove-tailed into the field, moving personalized genetic treatments from concept towards routine reality. In particular, widespread adoption of next-generation sequencing technologies have enabled faster vector integration profiling to assay for potential genotoxities. Genomic data generation since 2007 has grown faster than Moore's law for computing power, now making data analysis the new bottleneck in the workflow. Newer and accelerated solutions are needed to manage exponential increases in data accumulation. Our goal is to compress sequence data generation and downstream data analysis within 5 work days, enabling the utility of genotoxicity assay to support time-sensitive clinical decision making. Time and cost savings in integrated and streamlined data processing should translate into easier and faster implementation in clinical settings. We use non-viral transfections, based on the Sleeping Beauty transposition system, to reprogram T cells for Adoptive Immunotherapy. We exploit the fact that Sleeping Beauty transposons invariably inserts into “TA” sites via a cut-and-paste mechanism driven by cognate transposases, to develop an analysis platform, SBhashmap, to quickly locate and unlock insights associated with the genomic integration sites. We chose hash tables over other table data structures for speed, an advantage which is more apparent when the number of entries is large. SBhashmap consists of pre-populated hash tables with over 150 million entries corresponding to the “TA” sites from the human genome and hash functions to match for quick site matching. By cataloging the TA sites ahead of time, we can cut the mapping time significantly. To benchmark SBhashmap, we compared transposon integration mapping to BLAST and BLAT matching for human genomic sequences, and found a 20-fold reduction in processing time, while maintaining similar levels of accuracy. This kind of hash maps is particularly efficient because the maximum number of entries is known in advance, such that the bucket array can be allocated once with the optimum size fixed (not resized, with no insertions and deletions necessary). With collision-free hash functions, the keys are not stored in the tables. With minor adaptations, SBhashmap can process genomic data from experimentally relevant model organisms such mice, worms, and fruitflies.

DNA transposons such as piggyBac and Sleeping Beauty employ a cut-and-paste mechanism of transposition that has proven effective for numerous gene transfer and cellular reprogramming applications. Therapeutic transposons are carried into cells usually on plasmid DNA vectors either on the same plasmid as the transposase expression cassette in “cis” or on another plasmid in “trans.” We examined the fate of the remaining plasmid backbone after transposition. It has been assumed that the recircularized plasmid backbone remains episomal and is quickly lost from dividing cells. We delivered transposons containing a neomycin resistance gene and the transposase in either “cis” or “trans” to human embryonic kidney (HEK-293) cells, then selected the cells for 8 weeks and evaluated transposase expression by immunoblot. All of the piggyBac and 5 out of 6 of the Sleeping Beauty “cis” transfections resulted in transposase-positive clones. In contrast, none of the piggyBac “trans” transfections resulted in transposase-positive clones. We were concerned that the prolonged expression of the transposase might result in some instability of the transgenes delivered “cis” in vivo; however, we found that mice that received a piggyBac “cis” luciferase transposon vector by hydrodynamic injection had long-term luciferase values that were not statistically different from mice receiving the plasmids in “trans.” We next evaluated if piggyBac “trans” delivery could be modified to reduce the amount of backbone integration and found that clones that were transfected with lower amounts of transposase showed a decrease in the frequency of backbone integrations. We found no evidence that transposition increases the frequency of integration of co-delivered linear DNA. We sequenced backbone integration events and found that that the majority of the time, once the transposon is excised from its parent plasmid, the plasmid backbone is repaired by re-ligation and integrated into the genome at a site other than the excision site. Alternatively, a rare event may occur in which the linearized plasmid backbone can be integrated without excision repair. None of the integration events occurred at the TTAA tetranucleotide sequence, indicating that they are not directly mediated by piggyBac transposase. We next devised a strategy to measure and eliminate the backbones from human cells. We created a plasmid containing the neomycin-resistance gene in the transposon portion of the plasmid and the eGFP marker in the backbone. Transfected HEK-293 cells were selected with geneticin and sorted for eGFP. While the selected cells contained approximately ten copies of the transposon per diploid cell, unsorted cells typically had one to two copies of plasmid backbone per diploid genome. eGFP sorting effectively eliminated recovery of backbone-negative transgene-only cells. We expect further improvements to improve recovery of transgene-only cells. We will also investigate the extent of backbone integration in therapeutically relevant cell types and in vivo.

Sleeping Beauty (SB) is a DNA transposon capable of mediating gene insertion and long-term expression in vertebrate cells when co-delivered with a source of transposase. In all previous reports of SB-mediated gene insertion in somatic cells, the transposase component has been provided by expression of a co-delivered DNA molecule that has the potential for integration into the host cell genome. Integration and continued expression of a gene encoding SB transposase could be problematic if it led to transposon re-mobilization and reintegration. We addressed this potential problem by supplying the transposase-encoding molecule in the form of mRNA. We show that transposase-encoding mRNA can effectively mediate transposition in vitro in HT1080 cells and in vivo in mouse liver following co-delivery with a recoverable transposon or with a luciferase transposon. We conclude that in vitro-transcribed mRNA can be used as an effective source of transposase for SB-mediated transposition in mammalian cells and tissues.

The Sleeping Beauty transposon system (SB) has been shown to mediate nonviral integration of expression constructs resulting in long-term gene expression in several mammalian targets. Often, however, it is difficult to discern long-term expression resulting from transposition vs nonhomologous chromosomal recombination or maintenance of plasmid DNA in an extrachromosomal form. We have designed a system to silence expression from nontransposed sequences, making it possible to determine more specifically the amount of expression resulting from transposition. A transposon plasmid, pT2F/Cage (carrying a murine erythropoietin (Epo) gene transcriptionally regulated by the ubiquitously expressed CAGS promoter), was engineered to contain LoxP sites positioned so as to interrupt expression upon Cre-mediated recombination. Upon transposition these sites become segregated, thus protecting the expression construct from Cre-mediated recombination and subsequent silencing. Interferon-inducible Mx1Cre mice were administered pT2F/Cage with or without transposase-encoding plasmid. At 2 to 4 weeks postinjection, in the absence of SB transposase, Cre induction reduced Epo expression to about 1% of that seen in the group that was administered transposase-encoding plasmid, which maintained Epo levels near those of the uninduced groups. Southern hybridization analysis and plasmid rescue of transfected tissue supported the efficient Cre-mediated silencing of nontransposed sequences. These results indicate a substantial level of DNA-mediated expression not associated with transposition, but which can be quantitatively distinguished from transposition by its sensitivity to Cre recombinase. The results also provide additional evidence for the effectiveness of the Sleeping Beauty transposon system as an in vivo DNA-mediated gene transfer strategy for achieving long-term expression.

Transposable elements have enormous potential to overcome one of the major hurdles in nonviral gene delivery, namely the lack of long-term gene expression. The Sleeping Beauty (SB) transposon is a promising vector system for nonviral gene therapy as it has the highest transposition activity of all known DNA transposons within mammalian cells. In an effort to generate a more efficient delivery vehicle, we conducted a systematic evaluation of several novel and previously identified SB transposase mutants. The results indicate that certain combinations of mutants do not enhance transposition, whereas others give a synergistic response. The most active mutant, designated HSB17, shows nearly 17-fold higher transposition activity compared to the original transposase SB10 when tested within the same expression cassette. In addition, synergistic activity is observed when this hyperactive mutant is combined with an improved transposon. Animal studies utilizing the hyperactive transposase show enhanced long-term reporter gene expression. These modifications further expand the utility of this transposon-based gene transfer system.

In this study we performed a head-to-head comparison of the integrase phiC31 derived from a Streptomyces phage and the Sleeping Beauty (SB) transposase, a member of the TC1/mariner superfamily of transposable elements. Mouse liver was cotransfused with a vector containing our most robust human coagulation factor IX expression cassette and the appropriate recombinase recognition site and either a phiC31- or a SB transposase-expressing vector. To analyze transgene persistence and to prove somatic integration in vivo we induced cell cycling of mouse hepatocytes and found that the transgene expression levels dropped by only 16 to 21% and 56 to 66% in mice that received phiC31 and SB, respectively. Notably, no difference in the toxicity profile was detected in mice treated with either recombinase. Moreover we observed that with the integrase-mediated gene transfer, transgene expression levels were dependent on the remaining noncoding vector sequences, which also integrate into the host genome. Further analyses of a hot spot of integration after phiC31-mediated integration revealed small chromosomal deletions at the target site and that the recombination process was not dependent on the orientation in which the phiC31 recognition site attached to the pseudo-recognition sites in the host genome. Coupled together with ongoing improvements in both systems this study suggests that both nonviral vector systems will have important roles in achieving stable gene transfer in vivo.

Endothelial cells have complex roles in the pathophysiology of vascular and heart disease and are increasingly being recognized as targets for gene therapy. The intravenous administration of plasmid DNA complexed to lipid tends to target transfection of endothelial cells within the lung; however, expression from the transgene remains transient. Here we utilize the integrating capability of the Sleeping Beauty (SB) transposon for durable gene transfer within lung endothelia. To restrict expression of the transgene, an endothelial cell-specific promoter, endothelin-1, was placed within the transposon. Further refinements to the transposon increased in vitro transposition efficiency by 3.6-fold. Utilizing this optimized transposon we evaluated the expression of two reporter molecules, secreted alkaline phosphatase (SEAP) and intracellular GFP, following administration of DNA–polyethylenimine complexes to mice. Long-term expression (>2 months) of SEAP occurred only with cotransfection of adequate amounts of transposase. Localization studies using the GFP reporter, at 3 days and 6 weeks postinjection, demonstrated that the majority of transgene-expressing cells were of endothelial origin, while the second most abundant cell type was type II pneumocyte. These results suggest that the SB transposon can be adapted to target particular cell types, in this case, endothelial cells. Such an approach may be useful for gene therapy paradigms involving the long-term modulation of vascular and endothelial cell biology.

The Sleeping Beauty (SB) transposable element is a promising vector for transgenesis in vertebrates and is being developed as a novel, nonviral system for gene therapeutic purposes. A mutagenesis approach was undertaken to improve various aspects of the transposon, including safety and overall efficiency of gene transfer in human cells. Deletional analysis of transposon sequences within first-generation SB vectors showed that the inverted repeats of the element are necessary and sufficient to mediate high-efficiency transposition. We constructed a "sandwich" transposon, in which the DNA to be mobilized is flanked by two complete SB elements arranged in an inverted orientation. The sandwich element has superior ability to transpose >10-kb transgenes, thereby extending the cloning capacity of SB-based vectors. We derived hyperactive versions of the SB transposase by single-amino-acid substitutions. These mutations act synergistically and result in an almost fourfold enhancement of activity compared to the wild-type transposase. When combined with hyperactive transposons and transiently overexpressed HMGB1, a cellular cofactor of SB transposition, hyperactive transposases elevate transposition by almost an order of magnitude compared to the first-generation transposon system. The improved vector system should prove useful for efficient gene transfer in vertebrates.

Adoptive therapy with T cells that were modified with gamma-retroviral and lentiviral (LV) vectors to express a CD19-specific chimeric antigen receptor (CAR) has shown remarkable efficacy in pilot clinical trials. However, there are persistent concerns with the use of viral vectors with regard to safety, as well as the cost and scale of vector production required for making CAR T-cell therapy available to patients on a global level. In this study, we have refined non-viral Sleeping Beauty (SB) transposition to provide an effective and broadly applicable gene-transfer strategy with superior safety profile. We demonstrate that SB transposition of CAR transgenes from minicircle (MC) DNA vectors enables dramatically improved transposition rates compared to conventional plasmids. MC-derived CD19-CAR transposons display a highly favorable integration profile and confer stable CAR expression that permits potent anti-tumor functions and rapid manufacture of CAR T-cell products.We prepared MC transposon donor vectors encoding a CD19-CAR or eGFP from corresponding parental pT2 plasmids through site specific recombination, and a MC encoding SB100X transposase. Each MC contained only the promotor and gene of interest, and was devoid of antibiotic resistance gene and bacterial origin of replication. We then performed transfections into CD8+ and CD4+ T cells of healthy donors (n=7) and analyzed transposition rate and stability that could be accomplished when CD19-CAR transposon and SB100X transposase were encoded by MCs vs. equimolar amounts of corresponding conventional plasmids. In each donor, we achieved a significantly higher transposition rate with MCs (mean 49.8% on day 14 post transfection) compared to plasmids (mean 12.8%; 4.4-fold difference; p

Introduction We previously showed that the piggyBac transposon system is a promising genetic tool for stable, non-viral gene engineering of primary human T cells. Our initial culture strategy for expanding piggyBac-modified CD19 CAR-T cells required magnetic cell isolation due to low CAR expression and 21 days of culture and our CAR included an IgG1-Fc spacer that may limit CAR-T cell expansion in vivo. We have now improved these process and construct to avoid magnetic cell isolation, shorten the culture period, and increase in vivo efficacy. Methods We used a 4D-Nucleofector device to electroporate 1 × 107 peripheral blood mononuclear cells from 9 healthy donors with a CD19.CD28.ζ–CAR transposon plasmid, with or without an IgG1-CH2CH3 spacer and a piggyBac transposase plasmid. We compared 2 new culture conditions with our original method for expansion of CD19 CAR-T cells. All cells were cultured in serum-free medium (TexMACS) containing IL-7 and IL-15 in 24-well plates. Electroporated cells were immediately transferred to irradiated autologous activated T-cells (ATCs), either pulsed with 4 viral peptide pools (PepTivator; AdV5 Hexon, CMV pp65, EBV EBNA-1, and BZLF1) (ACE), or unpulsed (non-ACE). The next day, non-ACE cells were transferred to CD3 and CD28 antibody-coated plates for 5 days. In the original method, electroporated cells were transferred to CD3/CD28-coated plates one day after electroporation without ATCs. On day 7, all cells were transferred into G-Rex10 culture flasks with ACE-pulsed or unpulsed irradiated ATCs or no ATCs for ACE, non-ACE and original cells respectively. On day 14, we collected CAR-T cells from all conditions. Results We obtained 4.7 ± 3.0 × 107 ACE CAR-T cells, 6.7 ± 2.1 × 107 non-ACE CAR T cells, and 5.3 ± 2.3 × 107 original CAR T cells after 14 days of culture. ACE CAR-T cells showed significantly higher expression of the CAR transgene (33.0 ± 9.7% and >50% for CD19.CAR with and without the CH2CH3 spacer respectively) than non-ACE (10.7 ± 7.8%) or original CAR-T cells (4.6 ± 3.2%). All CAR-T-cells inhibited the growth of CD19+ tumors (95%, 60% and 51% for ACE, non-ACE and original CAR-T at an E:T ratio of 1:50; ACE CAR-T cells contained 29.6 ± 17.3% CD4+ cells and 64.3 ± 16.8% CD8+ cells, with 79.9 ± 5.9% cells co-expressing CD45RA and CCR7 and 62.6 ± 20.9% cells co-expressing CD45RO and CD62L. In an in vivo mouse model, ACE CAR-T cells delayed tumor growth by 30 days compared to no T-cells. Conclusions We markedly increased CAR expression from piggyBac-mediated CD19 CAR-T cells without cell selection, reducing the culture period to 14 days using viral antigen stimulation. Deletion of IgG1-Fc spacer led to further increase of CAR expression. Together with the simple manufacturing process and cost-effectiveness of DNA transposon technology, piggyBac-based gene transfer provides an alternative to retro- or lentiviral gene transfer for CAR T-cell therapy.

Transposon-based vectors represent promising new tools for chromosomal transgene insertion and establishment of persistent gene expression in vivo. Here, we report the development of helper-independent transposon-transposase (HITT) vectors, which contain on single plasmids (i) a Sleeping Beauty (SB) transposon containing the transgene and (ii) a SB transposase expression cassette. To obtain an optimal level of transposase expression from HITT vectors, we determined the relative strength of a panel of different promoters in mouse liver and used these promoters to drive transposase expression from injected HITT vectors carrying a human alpha(1)-antitrypsin (hAAT) expression cassette flanked by transposon inverted repeats. By correlating promoter strength with stabilized serum hAAT levels, a narrow expression window supporting high-level transposition in the liver was defined. Peak levels of long-term gene expression were obtained with promoters 30- to 40-fold less active than CMV in mouse liver, whereas reduced stable levels of hAAT were detected with both weaker and stronger promoters. Injected HITT vectors induced transposase-dependent insertion of transposon DNA into the genome of at least 5-6% of transfected hepatocytes, generating levels of persistent hAAT expression that were 2- to 4-fold higher than with an optimized two-plasmid approach. In addition, we show that HITT vectors carrying a human factor IX (hFIX)-containing transposon support (i) long-term hFIX expression in normal mice and (ii) partial phenotypic correction in a mouse model of hemophilia B. SB-based HITT vectors represent a major advance in the establishment of persistent transgene expression from nonviral gene delivery systems and should prove useful for gene transfer to tissues or cell types in which transfection efficiencies are low.

Gene transfer to the lung could provide important new treatments for chronic and acquired lung diseases such as cystic fibrosis, α1-antitrypsin deficiency, emphysema, and cancer. DNA-mediated gene transfer to the lung has been previously demonstrated, but anticipated effectiveness has been limited by low gene transfer efficiencies and by transient expression of the transgene. Here, we combine plasmid-based gene transfer with the integrating capacity of the nonviral Sleeping Beauty (SB) transposon vector system to mediate gene insertion and long-term gene expression in mouse lung. We observed transgene expression after 24 h in lungs of all animals injected with the luciferase transposon (pT/L), but expression for up to 3 months required codelivery of a plasmid encoding the Sleeping Beauty transposase. We also observed long-term expression in pT/L-injected animals transgenic for SB transposase. Transgene expression was localized to the alveolar region of the lung, with transfection including mainly type II pneumocytes. We used a linker-mediated PCR technique to recover transposon flanking sequences, demonstrating transposition of pT/L into mouse chromosomal DNA of the lung.

The Sleeping Beauty (SB) transposon system, derived from teleost fish sequences, is extremely effective at delivering DNA to vertebrate genomes, including those of humans. We have examined several parameters of the SB system to improve it as a potential, nonviral vector for gene therapy. Our investigation centered on three features: the carrying capacity of the transposon for efficient integration into chromosomes of HeLa cells, the effects of overexpression of the SB transposase gene on transposition rates, and improvements in the activity of SB transposase to increase insertion rates of transgenes into cellular chromosomes. We found that SB transposons of about 6 kb retained 50% of the maximal efficiency of transposition, which is sufficient to deliver 70-80% of identified human cDNAs with appropriate transcriptional regulatory sequences. Overexpression inhibition studies revealed that there are optimal ratios of SB transposase to transposon for maximal rates of transposition, suggesting that conditions of delivery of the two-part transposon system are important for the best gene-transfer efficiencies. We further refined the SB transposase to incorporate several amino acid substitutions, the result of which led to an improved transposase called SB11. With SB11 we are able to achieve transposition rates that are about 100-fold above those achieved with plasmids that insert into chromosomes by random recombination. With the recently described improvements to the transposon itself, the SB system appears to be a potential gene-transfer tool for human gene therapy.

Gene therapy applications of naked DNA constructs for genetic disorders have been limited because of lack of permanent transgene expression. This limitation, however, can be overcome by the Sleeping Beauty (SB) transposable element, which can achieve permanent transgene expression through genomic integration from plasmid DNA. To date, only one example of an in vivo gene therapy application of this system has been reported. In this report, we have further defined the activity of the SB transposon in vivo by analyzing the expression and integration of a fumarylacetoacetate hydrolase (FAH) transposon in FAH-deficient mice. In this model, stably corrected FAH(+) hepatocytes are clonally selected and stable integration events can therefore be quantified and characterized at the molecular level. Herein, we demonstrate that SB-transposon-transfected hepatocytes can support significant repopulation of the liver, resulting in long-lasting correction of the FAH-deficiency phenotype. A single, combined injection of an FAH-expressing transposon plasmid and a transposase expression construct resulted in stable FAH expression in approximately 1% of transfected hepatocytes. The average transposon copy number was determined to be approximately 1/diploid genome and expression was not silenced during serial transplantation. Molecular analysis indicated that high-efficiency DNA-mediated transposition into the mouse genome was strictly dependent on the expression of wild-type transposase.

The non-viral integrating vector the Sleeping Beauty (SB) transposon system is efficient in treating systemic monogenic diseases in mice including mucopolysaccharidosis (MPS) types I and VII caused by α-iduronidase (IDUA) and β-glucuronidase (GUSB) deficiency, respectively. More recently we have used modified approaches of the hydrodynamic procedure to deliver therapeutic transposons to dog liver. Reproducible delivery and transposition in dogs are about 1% the levels in mice. Using a transgenic canine reporter secreted alkaline phosphatase (cSEAP), in the absence of immune suppression we can detect transgenic protein for up to six weeks post infusion using catheter-mediated hydrodynamic delivery. A proof-of-principle immunomodulation using GdCl3 to block macrophages in liver and spleen prolonged the presence of the cSEAP protein in circulation from 6 weeks to up to at least 5 months after a single vector infusion. We achieved stabilized activity in one dog at about 2-fold of baseline values. Durability of cSEAP in serum was inversely correlated with transient increase of liver enzymes ALT and AST in response to the vector delivery procedure, pointing to the deleterious effect of hepatocellular toxicity on transgene maintenance. However, GdCl3 immunomodulation was ineffective for repeat vector infusions, suggesting a possibility of an alternative, more potent immunosuppression regimen. Evidence of transposition was obtained with the most efficient transposase SB100X but not with SB11. For transgenic IDUA and GUSB, therapeutic activity in serum peaked at 50-350% of wild-type at 2-4 days post-treatment, but lasted only a few days. The differences in levels and duration of detection of cSEAP in the blood compared to those of IDUA and GUSB may be in part due to the facilitated uptake of lysosomal enzymes into cells compared to cSEAP. Longer endurance of transgenic proteins at therapeutic levels may be possible in SB-treated dogs using alternative immunosuppressive regimens.

Duchenne muscular dystrophy (DMD) is a genetic neuromuscular disorder caused by the absence of dystrophin. We developed a novel gene therapy approach based on the use of the piggyBac (PB) transposon system to deliver the coding DNA sequence (CDS) of either full-length human dystrophin (DYS: 11.1 kb) or truncated microdystrophins (MD1: 3.6 kb; MD2: 4 kb). PB transposons encoding microdystrophins were transfected in C2C12 myoblasts, yielding 65±2% MD1 and 66±2% MD2 expression in differentiated multinucleated myotubes. A hyperactive PB (hyPB) transposase was then deployed to enable transposition of the large-size PB transposon (17 kb) encoding the full-length DYS and green fluorescence protein (GFP). Stable GFP expression attaining 78±3% could be achieved in the C2C12 myoblasts that had undergone transposition. Western blot analysis demonstrated expression of the full-length human DYS protein in myotubes. Subsequently, dystrophic mesoangioblasts from a Golden Retriever muscular dystrophy dog were transfected with the large-size PB transposon resulting in 50±5% GFP-expressing cells after stable transposition. This was consistent with correction of the differentiated dystrophic mesoangioblasts following expression of full-length human DYS. Alternatively, dystrophic mesoangioblast-like cells were generated from iPS of DMD patients. These iPS-derived mesoangioblasts, constitute an essentially unlimited supply of stem/progenitor cells that could be genetically corrected using PB transposons expressing dystrophin. These results pave the way toward a novel non-viral gene therapy approach for DMD using PB transposons underscoring their potential to deliver large therapeutic genes.

The PiggyBac (PB) transposon system is a potent nonviral gene delivery tool with relevance in both gene therapy and cell reprogramming for production of induced pluripotent stem cells (iPSCs). Moreover, by taking advantage of the ability of the PB transposase to excise transposon-embedded gene cassettes from the genome without leaving footprints, the PB system is unique in facilitating seamless genome editing. The need for intracellular production of the transposase, however, raises concerns related to delivery and to cytotoxicity caused by sustained transposase expression and insertional mutagenesis. Furthermore, transposon re-integration may decrease the overall efficiency of the PB-mediated excision as well as increasing the risk of adverse secondary insertions. Based on our previous work, we present a new approach for lentivirus-based delivery of PB transposase. By fusing the hyperactive PB transposase, hyPBase, to the C-terminus of the GagPol polyprotein, we show robust incorporation and subsequent release of the transposase in matured lentiviral particles. Furthermore, in an effort to limit transposon re-integration, we engineered a hyPBase variant carrying three missense mutations. This novel hyPBase variant, hyPBaseExc+/Int−, demonstrates integration levels very close to background levels, thus limiting the risk of reintegration. Notably, the ability of hyPBaseExc+/Int− to excise transposons from plasmid-borne as well as from genomically integrated PB transposon cassettes is increased up to 6.3-fold relative to the original hyPBase transposase. By fusing the hyPBaseExc+/Int− to the C-terminus of GagPol, transposase protein can be efficiently delivered to cells by lentiviral protein transduction and, in our model system, performs seamless genomic excision of PB transposon cassettes in a copy number and dose-dependent manner, resulting in excision in up to 23.6% of virus-treated cells. Using a transposon containing the puro-deltaTK transgene cassette, we furthermore show that cells with successful excision can be enriched 17-fold by negative selection with FIAU. We believe that protein transduction of hyPBaseExc+/Int− may increase the applicability and safety of transposase-directed genomic excision in iPSCs and in hard-to-transfect cell types including hematopoietic cells.

The piggyBac (PB) transposon system is a highly active non-viral gene delivery system capable of integrating defined DNA segments into the genomes of human cells. We have previously demonstrated biased PB integration into an artificial target site engineered into human cells (Kettlun et al., Molecular Therapy, 2011). We undertook the current investigation to see if we could engineer PB to target integration into a user-defined chromosomal locus in human cells. We chose to target the hypoxanthine phosphoribosyltransferase (HPRT) gene located on chromosome X. HPRT-deficient cells can be selected by growth in 6-thioguanine (6TG), so PB-mediated targeted knockout of HPRT can be used to isolate cells with targeted integration events. We used HT-1080 human fibrosarcoma cells which harbor one × chromosome for our studies. We engineered a series of 4 zinc finger proteins (ZFP) and 4 transcription activator-like (TAL) proteins fused to the PB transposase. The transposon used contained a splice-acceptor sequence followed by a beta-galactosidase-neomycin resistance fusion gene (b-Geo) which allowed measurement of gene transfer activity (measuring neomycin resistant colonies after transposition) as well as HPRT knockout efficiency (due to the splice-acceptor sequence). All HPRT-targeted ZFP-PB and TAL-PB proteins exhibited transposition activity. Chromatin immunoprecipitation assays revealed varied targeting-binding of the different PB chimeras to the HPRT locus. One ZFP-PB and one TAL-PB chimera demonstrated higher HPRT gene targeting via reproducible production of a higher number of 6TG resistant cells. Deep sequencing analysis revealed integration efficiency of 0.42% and 1% for the ZFP-PB and TAL-PB chimeric transposases in 6TG-resistant HT-1080 cells. Our study provides new insights into engineering PB to target locations in the human genome which will permit further refinements to improve targeting efficiency. Additionally, selection of cells with targeted integration events should allow isolation of “targeted-only” cells for consideration of therapeutic purposes.ss

Since the ex vivo strategies using lentiviral vectors have shown great success, vectors need to be explored aiming at long-term transgene expression in vivo in dividing cells. In order to attain this objective, we are developing two independent viral vector systems. Both hybrid systems are based on third generation high-capacity adenoviral vectors (HCAdV), which mainly have three advantages: Highest transgene capacity of all viral vectors (36kb), low toxicity (no viral genes encoded), and tissue specificity (several serotypes are available). Since adenoviral vectors are predominantly episomally maintained we combined them either with a S/MAR (scaffold/matrix attachment region) system for episomal persistence or SB100X, a transposase based integration machinery. HCAdV-SB100X is the latest version of a series of HCAdV-SB hybrid vectors developed from our group. The gene of interest (GOI) is flanked by inverted repeat sequences and after FLP mediated circularization it is mobilized and integrated in a “copy and paste” manner into one out of ~2,000,000 AT dinucleotides present in the human genome. After performing colony forming assays we could show that our latest vector HCAdV-SB100 is more than a 100-fold more active compared to the negative control HCAdV-mSB and it shows nearly 10-fold increased activity compared to the precursor version HCAdV-HSB5 in forming blasticidin resistant colonies. We performed experiments in U-87 glioblastoma and A549 adenocarcinoma cells. HCAdV-SB100X showed for all conditions an increased number of copiess per cell of the GOI compared to HCAdV-HSB5. We observed a strong “overproduction inhibition effect” for SB100X also described in the literature, which means that by increasing the ratio between SB and the GOI less integration events leading to resistant colonies occur. This is why we are currently testing optimal ratios between transposon and transposase to achieve a maximum amount of resistant colonies. The second vector system (HCAdV-SMAR) carrying the identical transgene as for the SB100X system is thought to function via sequence-moderated binding to matrix proteins mediating extrachromosomal retention and replication. Using an improved molecular design we could show that HCAdV-SMAR is forming a significantly higher number of mitotically active balsticidin resistant colonies in various cell lines compared to a negative control without SMAR. For both vector systems results from qPCR studies revealed significantly higher transgene copy numbers per cell for the SB100X system (up to 20 copies per cell) and for the S/MAR system (up to 10 copies per cell). By combining both vectors with newest adenoviral targeting technologies and strategies for adenoviral immune escape we are now able to evaluate both vectors carrying the same transgene under different in vivo conditions in a direct manner for different disease models.

X-linked chronic granulomatous disease (X-CGD) is an immune deficiency resulting from lack of production of microbicidal reactive oxygen species (ROS) by phagocytic cells. Mutations causing X-CGD can occur throughout the >30-kb CYBB gene encoding gp91phox, with the majority of patients exhibiting a single causative mutation within one of the 13 exons or adjoining intronic splice sites, resulting in a loss of gp91phox protein expression. We previously demonstrated targeted “safe-harbor” gene therapy of X-CGD in iPSCs through insertion of a codon-optimized CYBB minigene into the AAVS1 locus, resulting in constitutive gp91phox expression from a CAG promoter. Another approach targeting insertion of the codon-optimized minigene or normal CYBB cDNA to the start site of endogenous CYBB resulted in little or no gp91phox expression or ROS activity in iPSC-derived granulocytes, suggesting that regulatory elements downstream of the CYBB promoter may be required for efficient expression at this locus. In order to maintain normal regulation of gp91phox expression with minimal alterations in corrected cells, we now demonstrate a strategy for seamless targeted repair of CYBB exon 5 mutations using an exon 5-specific TALEN pair or CRISPR, along with a donor plasmid containing the corrected 146-bp exon 5 interrupted by a piggyBac transposon cassette containing a PGK promoter and puroDtk gene for positive/negative selection, and flanked by ~900-bp upstream and downstream intron sequences for homologous recombination. iPSCs from X-CGD patients with exon 5 458T>G mutation or 461A deletion were nucleofected with donor and TALEN or CRISPR expression plasmids. Both TALENs and CRISPR enabled efficient targeted insertion in 82-98% of puromycin-resistant clones, 25-65% of which contained off-target inserts, for an overall efficiency of 29-75% of selected clones containing only the targeted insert. Targeted iPSC clones lacking off-target inserts were nucleofected with “excision-only” piggyBac transposase expression plasmid and selected with ganciclovir for removal of the PGK-puroDtk cassette, leaving only a silent codon change to accommodate the TTAA sequence left behind after piggyBac excision, without additional alterations detected beyond correction of the CYBB mutation. Corrected iPSCs maintained pluripotency and upon in vitro granulocyte differentiation exhibited restoration of gp91phox expression and ROS production comparable to normal blood neutrophils (3100% of normal gp91phox expression and 85-99% of normal ROS activity by mean fluorescence intensity in DHR assay). Our findings demonstrate efficient and seamless targeted repair of exon 5 mutations in X-CGD iPSCs with an exon replacement strategy, resulting in normal regulation of gp91phox expression and functional correction of the disease defect in iPSC-derived granulocytes. The development of similar approaches for the other exons of CYBB will enable seamless repair of the majority of causative mutations found in X-CGD patients.

Hemophilia A and hemophilia B are X-linked, genetic disorders that are caused by defective or deficient coagulation factor VIII and IX, respectively, resulting in the inability to form blood clots and sustained bleeding after trauma or injury. Recombinant clotting factor protein is currently used to treat hemophilia at a high cost per patient. As a therapeutic approach for hemophilia, gene transfer has the potential to provide more consistent levels of circulating clotting factor over an extended period of time for more cost-effective treatment. Moreover, only modest levels of FVIII or FIX expression (2%-5% of normal) can improve clinical outcomes. We are using the Sleeping Beauty (SB) transposon system to engineer autologous human B cells for secretion of clotting factors as a cellular therapy for hemophilia. An in vitro system for expansion and differentiation of memory B cells into plasma cells has been developed. Plasma cells are suitable for sustained delivery of FVIII or FIX, since they secrete high levels of protein and may survive for years in vivo. For human FIX expression, we assembled an SB transposon with the human FIX coding sequence (codon-optimized with R338L mutation for enhanced potency) regulated by the CAGS promoter. Elevated levels of hFIX in cultures of B lymphoblastoid cells required co-electroporation of the hFIX transposon along with an SB transposase encoding plasmid, demonstrating the role of transposition in achieving extended hFIX expression. There have been remarkable advances recently in the treatment of hemophilia B by systemic AAV8-hFIX administration, but similar treatment of hemophilia A presents a significant challenge due to the size of the FVIII-encoding sequence and the complexity of the protein. We previously demonstrated B-domain deleted hFVIII expression and correction of clotting dysfunction in FVIII deficient mice by hydrodynamic delivery using the SB transposon system (Ohlfest et al, Blood 105: 2691, 2005). Current studies are focused on identifying conditions for effective hFVIII transposon delivery and long term expression in primary human B cells after Sleeping Beauty-mediated transposition. Results from these studies will be applicable to the development of a clinical protocol for treatment of human hemophilia by infusion of B cells genetically engineered using the Sleeping Beauty transposon system.

Top of pageAbstract The Sleeping Beauty (SB) transposon system can direct the integration of precise DNA sequences into human and mouse genomes. The SB vector system has two components: 1) a transposon containing a genetic cargo and 2) the catalytic transposase enzyme that mediates the cut-and-paste reaction from a plasmid into a vertebrate chromosome. In most cases, the source of SB transposase has been a gene that can integrate into chromosomes via illegitimate recombination. There is a concern that random integration and extended expression of a transposase gene might lead to remobilization of inserted transposons in the same genomes. Consequently, we have examined the relative levels of SB transposon genes and the duration of expression of SB transposase when delivered via plasmids to cells in livers of mice. The SB11 transposase gene, under direction of a ubiquitin (Ub) promoter was expressed from either of two cis plasmids, pT2/mCAGGS-Luc//Ub-SB11 or pT2/mCAGGS-IDUA//Ub-SB11, which express luciferase (Luc) and |[alpha]|-L-iduronidase (IDUA), respectively, under direction of an mCAGGS promoter. The lysosomal enzyme, IDUA, escapes from cells whereas Luc is confined to the cytoplasm. Consequently, they may elicit different immunological responses that could affect duration of expression of SB11. We used hydrodynamic, tail-vein injection to deliver these constructs to the livers of WT C57Bl/6 mice. The injections were evaluated 24-hr post injection for expression of Luc using the Xenogen|[trade]| imaging system to evaluate efficiency of gene delivery. Mice were re-imaged and euthanized at 1, 2, 3, 4, 5, 6, 7, 10, 14, 17 days and 3, 4, 6, 8, 10, 12, 14, 16, 20, 24 weeks. Expression of SB transposase was measured by western blotting over several weeks using anti-SB monoclonal antibodies, by RT-PCR, excision-PCR and, for finer resolution, by immunohistochemical staining for SB11 using a monoclonal antibody. The presence of the SB11 and either Luc or IDUA was measured by QPCR. Our results indicate that we lose detectable SB transposase protein expression within the first few weeks. We also saw, using PCR, loss of DNA and RNA in the IDUA-injected mice, which suggested an immunological response was eliminating transgenic liver cells (see Aronovich et al. abstract). We conclude that long-term residual SB transposase activity in livers of mice is below minimal detection and, accordingly, that delivery of the gene for the transposase does not represent a measurable risk.

Many approaches for treating hemophilia via gene transfer have been attempted in large animal models but all have potential drawbacks. Helper-dependent (HD) adenoviral vectors devoid of all viral coding sequences offer high transduction efficiencies but were hampered by transient phenotypic correction at a non-toxic dose. For persistent hemostatic correction in hemophilia B dogs, our current study utilized a novel gene-deleted adenoviral vector system which combines adenoviral vectors for high transduction efficiencies and for the first time in a larger animal the Sleeping Beauty (SB) transposase for somatic integration and stable transgene expression. We generated the HD adenoviral vector FTC/TcFIX/attB/(FRT)2 in which a transposon with a canine factor IX (cFIX) expression cassette is flanked by FRT sites. In the presence of Flp recombinase and the recently developed hyperactive transposase HSB5 (Yant et al., unpublished), the transgene undergoes Flp mediated excision followed by HSB5 mediated somatic integration.To analyze transgene persistence in vivo C57Bl/6 mice were co-transfused with 2x10e9 transducing units (TU) of the vector FTC/TcFIX/attB/(FRT)2 and 7×10e8 TU of a second HD vector which encodes HSB5 and Flp. Control mice received an inactive version of SB, respectively. Rapid cell cycling of mouse hepatocytes was induced by performing a two-thirds partial hepatectomy and by injecting CCl(4). Seventy-five days post-injection we detected serum levels of cFIX of up to 2000 ng/ml in the active SB group and no cFIX in the control mice. This indicated that integration of the transgene from the episomal adenoviral vector genome into the host genome occurred.To test for persistence of hemostatic correction of the bleeding diathesis in hemophila B dogs we have co-injected 5.4x10e11 TU of the HD adenoviral vector FTC/TcFIX/attB/(FRT)2 and 2.6x10e11 TU of the HSB5 and Flp encoding vector. This equals a total dose of 1.6×10e13 viral particles (vps) and 1x10e12 vps/kg body weight. We measured plasma cFIX levels of up to 3900ng/ml (normal level = 5000ng/ml) and observed complete phenotypic correction of the factor IX deficiency. The whole blot clotting time (WBCT) was reduced from 60min to 18.5min on day 35, the length of the study to date. This was in sharp contrast to our previous studies in hemophilia B dogs in which we used a non-integrating HD vector. In that study we observed a 3-fold increase of the WBCT 35 days post-injection which indicates that in our current study transposition stabilized phenotypic correction. In contrast to previous studies using early generation adenoviral vectors and keeping in mind that for the first time the in vivo performance of the SB transposase in a larger animal was evaluated, we observed no vector-related elevation of liver enzymes and no fall in platelet counts. Taken together, this study demonstrates that this adeno-transposon hybrid vector system will be important for treating genetic diseases.

Short-term, reversible expression of foreign proteins could be useful to modify cell fate. We have previously shown that retroviral vector mutants that are unable to initiate reverse transcription of their plus-stranded mRNA genomes mediate transient and highly efficient delivery of the site specific recombinase Cre into human and mouse fibroblasts (Galla et al., Mol Cell 2004). Cre protein activity was detected in >95% of target cells with a single treatment. Cre activity was particle-mediated and required the retroviral mRNA packaging signal as well as the expression of both Gag and Env in the packaging cell. Furthermore we could exclude plasmid contamination and Cre protein transduction as the underlying mechanisms for this mode of Cre delivery. Our data thus suggest that retrovirally delivered mRNA serves as an immediate translation template if not being reverse transcribed. When delivering the Sleeping Beauty (SB) transposase (cooperation with Z. Ivics, Berlin) by this approach, RT-deficient retroviral vector mutants were sufficient to mediate transposition of transfected plasmids containing transposon DNA. However, the efficiency of stable integration was only 3-fold above background levels, suggesting that the timing and/or expression levels have to be improved. Side-by-side comparison with the delivery of episomal DNA by lentiviral vectors revealed that RT-deficient retroviral mutants express foreign proteins for a much shorter duration (max. 4-5 days in replicating cells as opposed to >7 days for episomal lentiviral vectors) and lower levels (one to two orders of magnitude below the activity provided by episomal lentiviral vectors). As many receptor molecules require relatively low levels of expression to confer biological effects, we tested RT-deficient retroviral vectors to deliver the sialomucin CD34 or the murine cationic acid transporter mCAT-1, which serves as the receptor for the ecotropic murine leukemia virus. More than 80% of the exposed cells could be transduced following a single exposure to RT-deficient vectors encoding these receptors, and mCAT-1 expression conferred susceptibility of Jurkat cells to transduction with retroviral or lentiviral vectors pseudotyped with the murine ecotropic envelope protein. We conclude that RT-deficient retroviral vectors hold great promise for applications where low and transient expression of proteins achieves striking biological effects.

Top of pageAbstract The Sleeping Beauty (SB) transposon represents an important vehicle for in vivo gene delivery because it can stably integrate into the genomes of mammalian cells. Although SB is one of the most efficient non-viral integrating systems described to date, all previous estimations of its integrative potential have relied on the use of antibiotic selection schemes. Based on previous studies demonstrating post-integrative gene silencing with other gene transfer systems (e.g. retroviruses), we investigated how genomic integration influenced transposon expression in human cells. To do this, we devised a novel strategy to monitor SB integration irrespective of transgene expression. We first constructed single plasmids (pT/RSV- YFP.CMV-SB, pT/EF1|[alpha]|-eGFP.CMV-SB, and pT/dmEF1|[alpha]|- dmGFP.CMV-SB) encoding a GFP variant-marked transposon and an externally-located transposase expression cassette and then transfected each of these plasmids into HeLa cells. Shortly thereafter, cells were single-cell sorted by FACS and then grown in the absence of antibiotic selection to generate a large number of clonal cell lines, each of which was screened via Southern blot analysis for transposon integrations. Results indicate that in the absence of selection, the Sleeping Beauty transposition system has a true integration efficiency of 41-52%, a value that is significantly higher than the 3-10% estimates previously generated using antibiotic selection-based approaches. In addition, we find that, irrespective of promoter usage or CpG content, many transposon copies undergo post-integrational gene silencing during a 12 week time period, with silencing in 13-18% of single-integration cell lines. Furthermore, upon continued passage of the pT/RSV-YFP.CMV-SB-derived lines, the frequency of transposon silencing had increased substantially from 18% to 71% after a 42 week time period. Studies utilizing the methylation- sensitive restriction enzyme McrBC indicate that DNA methylation contributes substantially to transgene repression in the vast majority of integrated transposon copies, a finding which was further substantiated through experimental application of the DNA methyltransferase inhibitor, 5-aza-2|[prime]|-deoxycytidine. Histone deacetylation was also found to play a role in transgene repression using the histone deacetylases inhibitor trichostatin A. Overall, these transposon clones are proving invaluable in helping unravel the molecular mechanism(s) governing transcriptional gene silencing, and may someday provide the insight needed to design an appropriate integrating vector system that is capable of maintaining long-term gene expression regardless of its integration site.

Transposon systems permit non-viral delivery of genetic cargo into the genome of human cells and hold promise for eventual use in gene therapy. Transposon systems with activity in vertebrates include Sleeping Beauty (SB) and Mos1 of the Tc1/Mariner family and piggyBac which represents a different class of transposable element. We undertook evaluation of these three different transposable elements in human cells in order to compare the properties between each transposon system. This permitted direct comparison of the activity of native and hyperactive SB, native and hyperactive Mos1, and native piggyBacin plasmid vectors differing only by the transposase cDNA and inverted repeat (IR) elements. Mos1 required full internal sequences to be present for measurable transposition activity in human cells. The piggyBac transposable element could be reduced to a 311bp left IR and a 236bp right IR with retention of full activity. Evaluation of overall transposable activity revealed Mos1 << SB < piggyBac in cultured human cells. Hyperactive mutants of the Mos1 transposase were evaluated and showed little increase in transposition over native Mos1 in human cells. Interestingly, piggyBac showed more activity than native or even hyperactive SB (two-to-ten fold more active than hyperactive SB12 transposase combined with hyperactive pT3 transposon). Additionally, overproduction inhibition, a known major limitation of SB, was lacking with piggyBac. We evaluated piggyBac transposition following transient transfection of 50ng, 100ng, and 2|[mu]|g transposon DNA while varying the transposase DNA amount. In all cases, piggyBac lacked overproduction inhibition while SB transposition could be driven to very low levels at higher amounts of transposase DNA. Such results revealed that piggyBac transposition was limited only by the amount of transposon DNA present. Finally, we evaluated the addition of a zinc finger DNA binding domain element to the N-terminus of the different transposase enzymes. Zinc finger DNA binding domain addition ablated Mos1 activity and decreased SB activity which was only reliably measured using a combination of hyperactive transposase and transposon mutants. However, chimeric-piggyBac transposase activity appeared unaffected by N-terminal zinc finger DNA binding domain addition when compared to the native system. These results demonstrate that the piggyBac transposon system has advantageous properties when compared to other transposon systems with activity in human cells thereby making it a viable alternative system for gene therapy.

Sleeping Beauty (SB) is a DNA transposon capable of mediating chromosomal integration and stable expression in vertebrate cells when co-delivered with a source of transposase. In all pre-clinical reports where SB-mediated gene insertion in somatic cells has been used to correct mouse models of human disease, the transposase component has been provided as a co-delivered DNA molecule that has the potential for integration into the host cell genome. Integration and continued expression of a gene encoding SB transposase could be problematic if it led to remobilization and reintegration of transposons. Such continued expression of transposase is a key safety concern in development of the SB transposon system for clinical applications. As an alternate source of transposase, we have previously shown that in vitro transcribed transposase-encoding messenger RNA (mRNA) can effectively mediate transposon insertion both in vitro and in mouse liver (Wilber et al., Molecular Therapy, 2005, in press). Here, we test the use of transposase- encoding mRNA plus transposon DNA for gene therapy of hereditary tyrosinemia type I by first evaluating several parameters for systemic delivery and expression of mRNA in mice. We also introduce a method to quantitatively track repopulating liver cells by in vivo bioluminescence imaging after co-delivery of a DNA or RNA source of transposase with a bi-functional transposon encoding both mouse fumaryl acetoacetate hydrolase (FAH) and firefly luciferase (luc) genes in FAH deficient mice by rapid, high-volume injection into the tail vein. Liver repopulation was quantitatively monitored over time by increasing luc activity, measured as light emitted from the liver. Using this method, we determined that supplying SB transposase in the form of mRNA results in selective repopulation of corrected hepatocytes with stable co-expression of both FAH and luc. Plasma taken from animals 5 months after co-infusion with transposase mRNA contained levels of succinylacetone (the clinical determinant of tyrosinemia) that were nearly normalized. Amino acid levels were also normalized, suggesting normal liver metabolism of catabolized protein products (including urea and glucose). We further demonstrated the stability of integration by transplanting hepatocytes (250,000) into FAH deficient recipient mice. All transplanted animals survived NTBC withdrawal and gained weight consistently over a period of 90 days and demonstrated stable expression of luc. In summary, we demonstrate for the first time that transposase-encoding mRNA can be used to mediate non-viral gene therapy resulting in complete phenotypic correction that is stable and not associated with any incidents of cellular transformation.

Top of pageAbstract The purpose of this project was to evaluate the Sleeping Beauty (SB) transposon for delivery of the human |[alpha]|-L-iduronidase ( IDUA) gene to chromosomes in the livers of IDUA-deficient mice for long- term expression and correction of MPS type I. We chose the hydrodynamics-based DNA injection, which targets mainly the liver. We constructed SB transposon, plasmid-based vectors and injected them into idua-/-mice. Because our goal was in part to determine the efficacy of transposition as a way of providing long-term expression of IDUA protein, control groups of MPS I mice did not get the transposase. Blood samples for plasma isolation were collected 1 day after treatment and once every 2 weeks thereafter. Plasma IDUA activities reached >100-fold of wild type levels on day 1 following treatment, but were essentially gone in all mice 4 weeks following gene delivery. IDUA activity was not detected in the livers of these mice 3 months after plasmid administration. We examined the duration of the transposon-delivered human IDUA DNA maintenance and enzyme activity over 6 months in the liver of wild type (WT) C57Bl/6 mice that express IDUA. As in the MPS I mice, plasma and liver IDUA activity reached supra-normal levels on day 1 and remained at this level for the first week, but still dropped dramatically within 2 weeks and by 4 weeks were indistinguishable from background. Using DNA PCR, we found that the maintenance of the IDUA transgene mirrored the IDUA activity time-line. Even though transposition was confirmed by an excision assay, the PCR product was detectable only for the first two weeks, but not thereafter. It appears that cells that express the IDUA gene are cleared from the liver of treated mice by 4 weeks following injection, which suggests induction of an immune response either to the therapeutic protein or/and the cells that express the therapeutic gene. This hypothesis is supported by our observations that in all cyclophosphamide immune-suppressed MPS I mice, the initial 1-day plasma activity levels dropped more than 100-fold by 2 weeks, but then persisted at 20-500% WT activity in some mice. IDUA levels were stable (up to 5-fold higher than in the WT mice) in SB-transposase-treated MPS I mice whereas without SB, IDUA levels declined to near-background over 3 months. These persistent IDUA activity levels were sufficient to reduce the number and size of pathologic inclusions in the liver seen by toluidine blue staining. Secondary elevation of murine |[beta]|-glucuronidase, another lysosomal enzyme, was reduced in the liver, spleen and kidney and correlated with degree of reduction of lysosomal pathology in these organs. Thus, with immune suppression, a single dose of the SB transposon system can result in partial to complete biochemical correction of IDUA activity in the livers of treated adult MPS I mice.

The Sleeping Beauty (SB) transposon is a Tc1/mariner family transposon that is being examined for its efficacy as a non-viral, integrating vector for human gene therapy. Integration of SB transposons, into TA dinucleotide base pairs, with respect to genes (exons and introns) and intergenic regions appears fairly random at the genomic level, but not on a micro-scale, with a previously reported consensus sequence around the TA-integration sites [Vigdal et al. (2002) J. Mol. Biol. 323: 441|[ndash]|452; Carlson et al. (2003) Genetics 165: 243|[ndash]|256]. In an earlier study [Liu et al. (2005) J. Mol. Biol. 346: 161|[ndash]|173], we analyzed target-site preferences of SB transposons using an intra-plasmid transposition assay. We found that preferential sites had a unique deformation pattern that was characterized by an off-center angling, non-uniform twisting of the DNA helix and an increase in the distance between the central base pairs at preferred target sites. Particular deformations of the double helix could be predicted by the Vstep algorithm, which distinguished three categories of TA sites that varied by about 16-fold in their acceptance of SB transposons: preferred, intermediate-preferred, and non-preferred. Here we report our results of integrations into TA sites in the ninth intron of the mouse B-raf gene. Transpositions occurred as a result of remobilization, by SB10 transposase, of chromosomally-resident |[ldquo]|insertionally oncogenic|[rdquo]| transposons in somatic tissues of transgenic mice. This resulted in oncogenic selection of 25 gain-of-function insertions into B-raf in p19Arf|[minus]|/|[minus]| mice [Collier and Carlson et al., submitted]. There are 349 potential TA-integration sites in the 4069-bp intron, of which 22 were targets for 25 unique insertions. The Vstep algorithm predicted preferential patterns for 44 of the 349 possible sites, of which 7 were hit and three were hit twice (a 23% hit rate). There were 107 predicted intermediate-preferred sites of which 11 were targets (10% hit rate), and 198 non-preferred sites of which 4 were hit (a 2% hit rate). These data strongly suggest about an 11-fold preference for integration at certain TA dinucleotide base pairs in chromatin, which can be predicted by the Vstep algorithm. This data suggests that a V-step algorithm can predict preferential insertion sites for transposon vectors during non-viral, SB-mediated gene transfer protocols.

Primary hematopoietic stem cells (HSC), capable of repopulating the entire hematopoietic and immune system with the full complement of cell types, are an attractive target for gene therapy strategies to treat diseases of blood cells. Gene modification of a patient's HSC could be used for autologous transplantation to treat inherited genetic disorders, including Severe Combined Immune Deficiency (SCID) and other immune deficiencies. Non-viral strategies have become available for gene transfer into cells that result in integration of therapeutic gene cassettes. For example, Sleeping Beauty (SB) transposes directly from one DNA locus to another through the activity of the transposase enzyme recognizing and performing excision and integration of the DNA sequences between specific inverted repeat (IR) sequences of the transposon. Transposase-mediated integration of the IR-flanked expression cassette into chromosomes provides the basis for long-term transgene expression. Here we used the transposon-based gene delivery strategy composed of a two plasmid system: a plasmid containing the SB transposase gene (for transient production of the transposase), and a second plasmid containing an expression cassette with a reporter gene flanked by the SB IR (inverted repeat) sequences. For this study, we evaluated the potential use of the SB transposon system to deliver genes in a stable manner to human hematopoietic cells by electroporation using the Amaxa nucleoporator. One plasmid carrying an enhanced Green Fluorescent Protein (eGFP) reporter cassette (to facilitate FACS analysis of resultant cells) flanked by the SB IR sequences for permanent integration into the target cell chromosomes was co-nucleoporated with a second plasmid carrying the SB transposase expression cassette. We first used the K562 human erythro-leukemia cell line to optimize several parameters of gene delivery, achieving high levels (50-80%) of stable gene integration with optimal amounts and ratios of the two plasmids. Southern analysis of K562 clonal populations showed that 1 to 6 copies of the transposon can be detected. We then nucleoporated the SB-based system into human CD34+ progenitor cells isolated from umbilical cord blood and achieved initial gene transfer to these primary cells at moderate levels (10-30%) that led to stable expression of the reporter gene (1-4%). Results from these studies demonstrate the potential for using the SB transposon system for gene therapy using HSC.

We are moving forward with development of our proprietary Sleeping Beauty Transposon (SBT) system for the treatment of Fanconi anemia (FA) via ex vivo electroporation of hematopoietic stem cells. FA is an inherited recessive disorder caused by deficiency in one of a number of genes whose products form a complex involved in DNA repair. The primary hematological hallmarks of the disease are aplastic anemia, bone marrow failure, and increased susceptibility to leukemias thought to be caused by defective cellular mechanisms for DNA repair. Allogeneic bone marrow transplantation is currently the only curative treatment for these aspects of the disease and despite improvements in clinical protocols over the years difficulties with allotransplants for FA still persist. The primary goal of this study is to evaluate the potential of SBT vectors carrying the FANCC gene to effectaffect transposition and establish long-term expression in FANCC-deficient cell populations after introduction by electroporation. Additionally, we show that integration of FANCC transposons can correct the sensitivity of lymphoblastoid cell lines (LCL) derived from FANCC patients to DNA damaging agents such as mytomycin C. We constructed a transposon containing a PGK promoter to regulate transcription of the normal human FANCC cDNA sequence. The vector containing the transposon also includes a ubiquitin promoter regulating expression of the SB transposase gene, encoded outside of the transposon. The SBT-FANCC vector was introduced via electroporation into FANCC deficient and normal control LCLs. Long-term maintenance of FANCC complementation in the FANCC-deficient cells was observed as a sustained reduction in sensitivity to mitomycin C similar to normal controls. In addition, the isolation and molecular characterization of clonal populations of these FANCC-complemented cells demonstrated FANCC transposon integration into the genome by SB mediated transposition. Preliminary results using SBT vectors in K562 cells suggest that a transposase provided on a separate plasmid may increase the efficiency of transposition. As a result, studies are also underway to characterize the efficacy of the SBT-FANCC vector described above compared to a similar vector in which the transposase is supplied either by a separate plasmid or by a SB-encoding mRNA. These studies demonstrate molecular feasibility of the SB transposon system in combination with electroporation-based cell loading technology for genetic therapy of Fanconi anemia.

Deficiency of fumaryl acetoacetate hydrolase (FAH) in hereditary tyrosinemia type I results in oxidative damage to hepatocytes. Treatment includes restriction of dietary intake of tyrosine and administration of 2-(2-nitro-4-trifluoromethylbenzyol)-1,3-cyclohexanedione (NTBC), which inhibits an enzyme upstream of FAH in tyrosine catabolism. FAH-/- knockout mice recapitulate the human disorder and require NTBC for survival. The mice have been used as a model system to study liver repopulation by wild-type hepatocytes or by FAH mutant hepatocytes that were corrected with wild-type FAH via viral or Sleeping Beauty (SB)-mediated gene transfer. Here we introduce a method to quantitatively track repopulating liver cells in vivo by bioluminescent imaging. We assembled a SB transposon plasmid with the strong CAGGS promoter regulating murine FAH expression. To the region 3' of the FAH coding sequence we attached an internal ribosome entry site (IRES) sequence and the luciferase transgene such that FAH and luciferase are translated from the same mRNA. Also, a SB transposase gene regulated by the PGK promoter was inserted into the plasmid outside of the transposon. For gene delivery to hepatocytes, we used rapid, high-volume injection into the tail vein of FAH-/- mice. Liver repopulation by SB-transposed cells was initiated by withdrawal of NTBC, which induces tyrosinemia. Six weeks after DNA injection the mice are steadily increasing in weight without NTBC, suggesting that the metabolic defect has been corrected. We also studied these mice for expression of the luciferase reporter gene contained on the same transposon sequence as the FAH gene. At various times after injection of tranposon DNA, the mice were injected i.p. with luciferin, and luciferase expression was visualized and quantified in vivo using the IVIS|[reg]| imaging system (Xenogen Corp.). We found that luciferase expression increased by 10-fold during two weeks of selective repopulation. We have engineered additional transposon constructs to link FAH expression with that of any transgene of interest, such as a mutant alpha1-antitrypsin gene. One potential application of this method is to generate liver-transgenic models without the necessity of germline transgenesis. In conclusion, we have developed a method to quantitatively track repopulating liver cells in vivo using luciferase as a reporter, and demonstrate that FAH-/- mice can be used for transgenesis of the adult mouse liver.

The Sleeping Beauty (SB) transposon system has been demonstrated to mediate gene insertion and long-term expression, suggesting the potential application of this plasmid based system for non-viral gene therapy. All reported studies of SB-mediated gene transfer to somatic cells have supplied the transposase by expression of co-injected plasmid DNA. Although successful for both in vitro and in vivo transposition, there is a lack of control with respect to the transient and long-term levels of transposase expressed in a target cell. There is also the potential for genomic integration of the transposase encoding DNA, resulting in sustained transposase expression and potential destabilizing effects on the integrated transgene or host genome integrity. We address these problems by supplying the transposase-encoding molecule in the form of mRNA. Rapid, high pressure injection of m7G-capped luciferase (Luc) mRNA through the tail vein of mice resulted in efficient luciferase expression with respect to a DNA element where the Caggs promoter controlled luciferase transcription. Using this system as a model for efficient gene transfer in the liver, we have taken steps to optimize delivery/expression of the mRNA by generating constructs to promote stability and translation. Real-time whole body in vivo imaging of Luc mRNA translation was conducted to measure the effect of an IRES and UTR sequences on luciferase expression. These studies determined that incorporation of 5' and 3' UTR elements from the Xenopus -globin gene were essential to achieve peak luciferase expression levels from an otherwise unprotected transcript. Co-delivery of these mRNAs with an inhibitor of or a substrate for endogenous RNases provided luciferase expression sufficient to mimic levels achieved using promoters known to provide optimum transposase expression for in vivo transposition. Using this genetic arrangement, we have co-delivered SB transposase mRNA with a transposon encoding the Luc gene (driven by the Caggs promoter) to the liver of transgenic mice in which transcription of Cre recombinase is under the control of an interferon response element. To distinguish between plasmid-mediated expression and expression mediated by transposition, we have engineered this transposon encoding plasmid with lox-P sites positioned such that Cre-mediated recombination excises a sequence which includes the Caggs promoter, unless it has been segregated away from the plasmid by the SB transposase (Score et al., Mol. Ther. 7: S9, 2003). We found that pIpC induction of Cre recombinase caused a 1600-fold reduction in luciferase expression in animals infused with the transposon alone. While co-delivery of transposase mRNA resulted in a moderate degree of inadvertent Cre induction, we observed only a 45-fold reduction in luciferase expression following analogous pIpC administration. This remaining luciferase expression is most likely attributable to SB mediated transposition. These results provide the first evidence for use of SB transposase mRNA to mediate transposition and long-term expression in an in vivo setting.

The development of efficient nonviral vectors that allow for long-term transgene expression may have significant impact on future gene therapy applications. Sleeping Beauty (SB) is a DNA-based transposon system derived from teleost fish sequences that has been shown to have activity within mammalian cells. In this two-part system, the transposase facilitates gene transfer by binding to specific sequences within the transposon (IR/DR elements) and catalyzing its integration into TA dinucleotide sequences within the target genome. The original SB transposase (SB10) was constructed based on consensus sequence alignments using several fish species. In attempt to further improve the transpositional activity of SB10, alignment analysis was conducted using transposase sequences across a wide range of species. Based upon this data, specific mutations were made within the SB10 protein and the functional activity of each mutant was evaluated.