Hybrid adeno-associated viral vectors utilizing transposase-mediated somatic integration for stable transgene expression in human cells

Abstract

Recombinant adeno-associated viral (AAV) vectors have been shown to be one of the most promising vectors for therapeutic gene delivery because they can induce efficient and long-term transduction in non-dividing cells with negligible side-effects. However, as AAV vectors mostly remain episomal, vector genomes and transgene expression are lost in dividing cells. Therefore, to stably transduce cells, we developed a novel AAV/transposase hybrid-vector. To facilitate SB-mediated transposition from the rAAV genome, we established a system in which one AAV vector contains the transposon with the gene of interest and the second vector delivers the hyperactive Sleeping Beauty (SB) transposase SB100X. Human cells were infected with the AAV-transposon vector and the transposase was provided in trans either by transient and stable plasmid transfection or by AAV vector transduction. We found that groups which received the hyperactive transposase SB100X showed significantly increased colony forming numbers indicating enhanced integration efficiencies. Furthermore, we found that transgene copy numbers in transduced cells were dose-dependent and that predominantly SB transposase-mediated transposition contributed to stabilization of the transgene. Based on a plasmid rescue strategy and a linear-amplification mediated PCR (LAM-PCR) protocol we analysed the SB100X-mediated integration profile after transposition from the AAV vector. A total of 1840 integration events were identified which revealed a close to random integration profile. In summary, we show for the first time that AAV vectors can serve as template for SB transposase mediated somatic integration. We developed the first prototype of this hybrid-vector system which with further improvements may be explored for treatment of diseases which originate from rapidly dividing cells.

Figure 1. Principle of the hybrid-vector system based on Sleeping Beauty (SB) transposase-mediated transposition from AAV vector genomes.

For somatic integration cells were simultaneously infected with the transposon-donor vector and the SB transposase encoding virus. After entering the cell, single-stranded AAV vector genomes form double-stranded DNA and generate several different molecular forms including circular monomers, dimers, and concatemers as well as linear monomers and concatemers [16]. Next, the transposon flanked by inverted repeats (IR, white horizontal arrow) and the AAV-derived inverted terminal repeats (ITR, black squares) needs to be mobilized from the various AAV vector genome forms. The transposon is then integrated into the host genome by a cut-and-paste mechanism mediated by the SB transposase protein which is delivered in trans and encoded by the second vector. Subsequently the transposon is integrated into chromosomal DNA (waved line) into the genomic target site (TA dinucleotide, grey square).

Figure 2. Transposition efficiencies in HEK293 cells stably expressing hyperactive Sleeping Beauty transposase SB100X.

(a) Plasmids pIRES-Puro-SB100X and pIRES-Puro-mSB were generated containing a bicistronic IRES construct expressing SB100X or mSB and the puromycin resistance gene under the control of the cytomegalovirus promoter (CMV). (b) Cells were stably transfected with plasmids pIRES-Puro-SB100X or pIRES-Puro-mSB and single cell clones were amplified under puromycin selection pressure. Total RNA was isolated from single cell clones and RT-PCR was performed to show expression of SB100X and mSB using primers SB100X-rev and SB100X-forw (Table S1). Clone 3 from the selected SB100X and mSB cell clones was chosen for further experiments. (+): Cell clones which stably express SB100X or mSB. (-): Cell clones which were negative for transposase expression. (c) The previously published plasmid pTnori [10] containing a neomycin encoding transposon was transfected into stably expressing SB100X and mSB (SB100X-HEK293 and mSB-HEK293) cells and kept under selection pressure using neomycin to select for transposition events. Obtained cell colonies were stained with methylene blue and counted. Error bars indicate standard deviation (n=3). *Significant difference between the SB100X and the mSB control groups (p-value < 0.05). (d) SB100X-HEK293 and mSB-HEK293 cells were infected with the recombinant vector AAV-neo at different MOIs (MOIs 1,000 and 10,000). AAV-neo represents the transposon-donor vector from which the transposon is mobilized. The transposon is flanked by Sleeping Beauty transposase derived inverted repeats (IR) and it expresses the neomycin resistance gene under control of the simian virus promoter (SV40) for eukaryotic expression and the Tn5 promoter for expression in bacteria. Additionally the transposon is flanked by the Flpe recognition sites FRT. Error bars indicate standard deviation (n=3). *Significant difference between the SB100X and mSB control groups (p-value < 0.05), “n.s.”: not significant, no significant difference between the SB100X and the mSB control groups (p-value > 0.05).

Figure 3. Transposition efficiency after infection with the transposon-donor vector AAV-neo co-transfected with transposase encoding plasmids.

(a) Colony forming assay to determine integration efficiencies from the AAV-neo vector co-transfected with transposase encoding plasmids. HeLa-cells were first transfected with 1µg of the respective transposase encoding plasmid (pCMV-mSB, pCMV-HSB5, or pCMV-SB100X) and one day post-transfection cells were infected with AAV-neo at MOI 10,000. Two days post-infection cells were diluted and kept under selection pressure for 14 days. Obtained cell colonies were either collected as cell pools for integration site analysis or stained with methylene blue to determine transposition rates. Transposase encoding plasmids contain expression cassettes for the hyperactive transposases HSB5 and SB100X and the inactive SB transposase version mSB expressed under the control of the cytomegalovirus promoter (CMV). (b) Result of the colony forming assay. The Y-axis shows the number of neomycin resistant colonies obtained from 4 x 10⁵ cells in different experimental settings. The lower panel shows examples of original tissue culture plates from the different groups after methylene blue staining. Error bars indicate standard deviation (n=3). *Significant difference between the hyperactive transposase groups (SB100X and HSB5) and the mSB control group (p-value < 0.05).

Figure 4. Transposition efficiency of the AAV/SB transposase hybrid-vector system.

Colony forming assay to determine integration efficiencies from the hybrid AAV/SB vector system in HeLa-cells. HeLa-cells were co-infected with AAV-neo and AAV-SB100X at increasing dosages (MOI 100, 1,000, 10,000 and 50,000). (a) The AAV vector AAV-SB100X contains a transgene expression cassette for the hyperactive Sleeping Beauty (SB) transposases SB100X expressed under the control of the cytomegalovirus promoter (CMV). The control vector AAV-mSB encodes the mutated and inactive version of SB (mSB). (b) After co-transduction and selection, obtained cell colonies were either collected as pools for integration site analyses or colonies were stained with methylene blue and counted to determine integration efficiencies. (c) Result of the colony forming assay. The Y-axis shows the number of neomycin-resistant colonies obtained in different experimental settings. Error bars indicate standard deviation (n=3). n.d.: not determined.

Figure 5. Genome copy numbers of each hybrid-vector element after AAV transduction.

Genomic DNA of neomycin-resistant cells (from Figure 4) was analysed by quantitative real-time PCR (qRT-PCR) using primers and probe detecting AAV-ITRs and primers detecting the cDNAs of the two main functional genes neo (neomycin resistance gene) and SB (Sleeping Beauty transposase encoding gene). As internal control a PCR detecting hB2M (Beta-2-microgloblin gene) was performed. (a) PCR setup and location of primers used for quantitative analysis of vector genome copy numbers. The neo-PCR detects a 337 bp region in the neomycin resistance gene, which include integrated transposon, non- integrated and integrated AAV genomes. The SB-PCR detects an 82 bp region contained in the SB transposase encoding gene. Primers and probe specifically detecting the AAV ITR region [44] were used for amplification and detection of a 62 bp PCR product. Red arrows depict real-time PCR primer binding sites. (b) Neomycin copy numbers per 1000 cells after AAV transduction. Error bars indicate standard deviation (n=3). (c) SB transposase copy numbers per 1000 cells. Error bars indicate standard deviation (n=3). (d) 2XITR copy numbers per 1000 cells in co-transduced cells. Error bars indicate standard deviation (n=3).

Figure 6. Integration sites from the hybrid AAV/SB vector system identified by plasmid rescue and LAM-PCR approaches.

(a) Integration events identified by plasmid rescue. Percentages of integration events within each individual chromosome in comparison to a random control set is depicted in the upper panel. A schematic overview of chromosomal distribution is shown in the lower panel. The relative position of all unique integration sites identified from AAV/SB hybrid vector system infected cells were mapped within the human genome. Respective triangles indicate the relative positions of the chromosomal integration site observed (red triangle for SB-mediated integration; blue triangles for AAV -mediated integration). (b) Sequence reads and chromosomal distribution identified after performing LAM-PCR in AAV/SB hybrid-vector infected HeLa-cells (upper panel). Integration sites were compared to a computer-simulated random integration profile [13]. The lower panel shows mapping of all 1716 unique integration sites identified by LAM-PCR. Respective triangles indicate the relative positions of the chromosomal transposon integration site observed from sample 1 (blue triangles, AAV-neo MOI 10,000 and AAV-SB100X MOI 10,000) and sample 2 (red triangles, AAV-neo MOI 10,000 and AAV-SB100X MOI 50,000).

Figure 7. Frequencies of insertions from hybrid AAV/SB vector system within or outside of genes.

(a) Characterization of AAV/transposase hybrid-vectors mediated integrations identified by plasmid rescue. Percentage of integration sites which were identified to be intragenic (in introns, or in exons) and intergenic after performing plasmid rescue (left panel). Random: computer predicted data [13]. The right panel shows the distance to the nearest genes of integration sites which hit intergenic regions. Distances of genes upstream and downstream of the transposition event are depicted. (b) Characterization of AAV/transposase hybrid-vectors mediated integrations identified by LAM-PCR method. Insertion frequencies compared to a random dataset are shown with respect to integration within and outside of RefSeq genes (left panel), and distances upstream and downstream of genes (right panel). The bars depict fold changes of integration frequencies compared to the random distribution profile. Abbreviation: sample 1 (AAV-neo MOI 10,000 and AAV-SB100X MOI 10,000) and sample 2 (AAV-neo MOI 10,000 and AAV-SB100X MOI 50,000).

Figure 8. Transposition efficiencies are dose-dependent.

(a) To calculate the ratio of SB transposase-mediated integration versus AAV-vector mediated integration, the percentage of SB-mediated integration events contained in all integration events (SB-mediated integration events + AAV-vector mediated integration events) was determined. Calculation: (neomycin gene copy number determined by qRT-PCR) – [(AAV-ITR copy number determined by qRT-PCR) – (SB copy number determined by qRT-PCR)] / (neomycin gene copy number determined by qRT-PCR) x 100. It is of note that the low percentage of potentially remaining non-integrated AAV vector genomes should also be detected by the neomycin-specific primers. (b) Schematic overview of dose-dependent transposition efficiencies. The X-axis indicates increased dosages of the transposase encoding vector AAV-SB100X while the Y-axis indicates the increase of the AAV-neo vector. The dashed red arrow represents the trend of AAV-vector mediated integrations and the blue curve displays the efficiency of SB-mediated integrations in a dose-dependent manner.