Adeno-Associated Virus Site-Specific Integration and AAVS1 Disruption
ABSTRACT
Adeno-associated virus (AAV) is a single-stranded DNA virus with a unique biphasic lifestyle consisting of both a productive and a latent phase. Typically, the productive phase requires coinfection with a helper virus, for instance adenovirus, while the latent phase dominates in healthy cells. In the latent state, AAV is found integrated site specifically into the host genome at chromosome 19q13.4 qtr (AAVS1), the only animal virus known to integrate in a defined location. In this study we investigated the latent phase of serotype 2 AAV, focusing on three areas: AAV infection, rescue, and integration efficiency as a function of viral multiplicity of infection (MOI); efficiency of site-specific integration; and disruption of the AAVS1 locus. As expected, increasing the AAV MOI resulted in an increase in the percentage of cells infected, with 80% of cells infected at an MOI of 10. Additional MOI only marginally effected a further increase in percentage of infected cells. In contrast to infection, we found very low levels of integration at MOIs of less than 10. At an MOI of 10, at which 80% of cells are infected, less than 5% of clonal cell lines contained integrated AAV DNA. At an MOI of 100 or greater, however, 35 to 40% of clonal cell lines contained integrated AAV DNA. Integration and the ability to rescue viral genomes were highly correlated. Analysis of integrated AAV indicated that essentially all integrants were AAVS1 site specific. Although maximal integration efficiency approached 40% of clonal cell lines (essentially 50% of infected cells), over 80% of cell lines contained a genomic disruption at the AAVS1 integration locus on chromosome 19 (≈100% of infected cells). Rep expression by itself and in the presence of a plasmid integration substrate was able to mediate this disruption of the AAVS1 site. We further characterized the disruption event and demonstrated that it resulted in amplification of the AAVS1 locus. The data are consistent with a revised model of AAV integration that includes preliminary expansion of a defined region in AAVS1.
Adeno-associated virus (AAV), a single-stranded DNA virus of the family Parvoviridae, genus Dependovirus, is among the smallest known viruses. The serotype 2 virus has a genome of 4,680 bases contained in an icosahedral capsid of ≈24 nm (31). The defining characteristic of this class of viruses is their dependence on coinfection with a helper virus (or certain perturbed cellular states) to provide the functions necessary for productive infection (5). Despite a recent study demonstrating low-level AAV virus production in a differentiated epithelial cell system, the latent state appears to predominate in healthy infected cells (4, 23).
AAV is unique among animal viruses in that its latent state involves insertion into the host genome at a preferred site (17, 28). This site, referred to as AAVS1 (27), is found at chromosome 19q13.4qter and has been mapped to the first exon of myosin binding subunit 85 of protein phosphatase 1 (33). Although the molecular mechanisms of AAV integration have not been fully elucidated, the sequence requirements necessary in both the virus and the host are known. In the host, a 34-bp sequence is both necessary and sufficient for efficient integration (11, 19). In the virus, sequences in the viral promoter at map unit 5 (p5) are required for efficient integration (26, 27). Furthermore, integration has been associated with gross disruptions in the AAVS1 locus, including both duplications and deletions (12). Disruption of a nearby gene has also been reported in one case (8).
The virus contains two open reading frames (ORFs) (5). The right ORF encodes the structural capsid proteins, and the left ORF encodes the nonstructural Rep proteins. The Rep proteins are required for all phases of the viral life cycle, including transcription, replication, encapsidation, integration, and rescue from the latent state (5). The larger Rep proteins, Rep 78 and 68, in tandem with the p5 integration sequence element, are able to mediate all of the required virus-based steps of integration (27, 32).
Rep activation of the AAVS1 site reflects many of the known replication activities of this protein. In vitro, the AAVS1 site was found to be functionally equivalent to a viral origin of replication (35, 37). Three cis elements are present, a Rep binding site, a terminal resolution site, and a spacer element between the Rep binding site and terminal resolution site, which has both sequence and length constraints (19, 22). Rep mediates activation of DNA synthesis at the AAVS1 site in vitro and in vivo and forms a putative integration complex between AAV and AAVS1 (35, 38).
Recent studies have also defined the viral sequences required for site-specific integration. In contrast to AAVS1, replication at the viral terminal resolution site does not appear necessary (40). In fact, an integration efficiency element (p5IEE) that is present in the AAV p5 promoter sequence is the sole cis requirement for high-efficiency integration of plasmid DNA into AAVS1 (26, 27). With the exception of an absolute requirement for Rep protein, no other viral factors, including the inverted terminal repeats, are required to mediate efficient AAV site-specific integration (27). While replication at the viral inverted terminal repeats appears to be unnecessary for integration, sequences in the p5IEE promoter can function as an origin of replication (36). Replication and/or nicking at this site during integration cannot be ruled out.
Although virus infection appears to be widespread, with over 80% of humans testing seropositive, AAV causes no known pathology (5). The infectivity, moderate immunogenicity, and apparent lack of pathogenicity have made AAV a promising candidate as a vector for human gene therapy (20). Recently, random integration and subsequent activation of cellular genes have created safety concerns about retrovirus vectors (13), and recombinant AAV vectors have also been shown to disrupt cellular genes (16, 25). Also, randomly integrating recombinant AAV vectors are severely compromised in integration efficiency (26). Site-specific integration as mediated by AAV may have both safety and efficiency advantages in certain gene therapy applications.
The mechanism of wild-type AAV integration is not well understood. Site-specific integration efficiency has been estimated with a PCR assay, but actual levels are unknown (15). Furthermore, the correlation between integrated and site-specifically integrated virus has not been carefully established. Finally, little is known about the initial events leading to virus-host recombination. To further develop our understanding of AAV site-specific integration and its use as a vector for gene transfer applications, we characterized the efficiency of wild-type virus infection and integration and further defined the initial steps required for site-specific recombination.
MATERIALS AND METHODS
Cells and viruses.
HeLa cells were obtained from the American Type Culture Collection (ATCC) and grown in Dulbecco's modified Eagle's medium (DMEM) (Gibco-BRL, Gaithersburg, Md.) supplemented with 5% fetal bovine serum and 5% bovine calf serum (HyClone, Logan, Utah) and grown at 37°C in 5% CO2. HeLa cells were converted to suspension by growth in suspension minimal essential medium (Gibco-BRL, Gaithersburg, Md.) supplemented with 10% horse serum (HyClone). Adenovirus type 5 was grown and purified by CsCl gradient centrifugation. Wild-type AAV was grown by infecting HeLa-spinner cells with AAV2 at a multiplicity of infection (MOI) of 5 and adenovirus type 5 at an MOI of 10. At full cytopathic effect, virus was purified by iodixanol step centrifugation followed by heparin-agarose column purification (41), with the following modifications: crude viral lysate was heated to 56°C for 30 min to inactivate contaminating adenovirus; 50 mg of DNase I per ml was used instead of benzonase treatment to reduce viscosity. Wild-type AAV was titered with an infectious center assay, and particle number was determined with a dot blot assay for genome content as previously described (14). The viral stock was 1012 particles/ml, with 50 particles/infectious unit.
Plasmid constructs.
Plasmid p78Rep was constructed as previously described (26). The inverted terminal repeat-deleted integration-substrate plasmid was generated by cloning a 138-bp PCR fragment of the AAV p5 promoter element (corresponding to AAV nucleotides 151 to 289) upstream of a chloramphenicol acetyltransferase (CAT) reporter gene in place of the cytomegalovirus promoter element in plasmid pAdCMVCAT (27).
Wild-type AAV infection efficiency.
Suspended HeLa cells were counted and resuspended at 106 cells/ml in DMEM; 1 ml of cells was infected with wild-type AAV at various MOIs for 1 h at 37°C and then placed in 10-cm plates with 10 ml of DMEM plus 10% fetal bovine serum. At 24 h, cells were superinfected with wild-type adenovirus type 5 at an MOI of 10 by removing the cell medium and replacing it with 1 ml of DMEM containing wild-type adenovirus type 5 for 1 h and then overlaying it with 9 ml of DMEM plus 10% fetal bovine serum. At 20 h post-adenovirus infection, single-cell suspensions were made by digestion with Accutase (Innovative Cell Technologies, San Diego, Calif.). These suspensions were diluted with 1× phosphate-buffered saline-0.1% bovine serum albumin to a final concentration of ≈10 cells/ml, and 5 ml of each dilution was filtered onto Hybond+ positively charged nylon membranes (Amersham, Piscataway, N.J.). The cells were lysed in situ by placing the filters cell side up for 5 min on Whatman paper soaked in 0.5 M NaOH-1.5 M NaCl and neutralized by the same method with 1 M Tris, pH 7.0. Filters were probed with a 32P-labeled Rep gene fragment, as described below, to score for productively infected cells. Membranes were then stripped and reprobed with HeLa genomic DNA (isolated according to protocols described in the Southern blot protocol) to score for total cells.
AAV rescue assay.
HeLa cells were infected with wild-type AAV at specific MOIs and plated in 96-well plates to make single-cell clones. Three weeks postinfection, the wells were microscopically examined for the presence of individual clones. Clonal cell lines were replated in 96-well dishes and infected with wild-type adenovirus type 5 at an MOI of 10. At full cytopathic effect (48 to 72 h), cells were lysed by addition of NaOH and EDTA to 0.4 M and 10 mM, respectively, and denatured for 15 min at 68°C. Lysates were filtered onto Hybond+ nylon membranes with a dot blot apparatus and probed for the presence of wild-type AAV Rep DNA as described above.
Southern blot analysis.
Whole-cell DNA was isolated from HeLa cell lines with a standard salting-out protocol (24). Designated restriction endonucleases were used to digest 10 μg of DNA from each clone, and digested DNA was separated on 1% agarose gels. After transferring DNA fragments to nylon membranes, hybridization was carried out with 32P-labeled probes at a concentration of 3 × 106 cpm per ml of Sigma PerfectHyb hybridization solution, according to the manufacturer's instructions.
The following DNA fragments were generated for DNA probes. The 800-bp Rep PCR fragment was generated from oligonucleotides GATCGAAGCTTCCGCGTCTGACGTCGATGG and GGACCAGGCCTCATACATCTCCTTCAATGC; the AAVS1 2-kb PCR product was obtained with oligonucleotides GCGCCGTGACGTCAGCACGC and CACCAGATAAGGAATCTGCC; the CAT DNA fragment of 700 bp was PCR amplified with oligonucleotides GCTAGCTTGAGGTGTGGCAGGC and GGCATGATGAACCTGAATCGC; the a 650-bp β-actin DNA fragment was PCR amplified with oligonucleotides TGACGGGGTCACCCACACTGTGCCCATCTA and CTAGAAGCATTTGCGGTGGACGATGGAGGG. DNA probes were 32P labeled with the Rediprime II kit (Amersham, Piscataway, N.J.) according to the manufacturer's instructions. Bands were visualized by autoradiography. Quantitation was performed with a Storm phosphorimager system with ImageQuant software as per the manufacturers' instructions (Sigma-Aldrich, St. Louis, Mo.).
Plasmid integration assay.
We washed 1.6 × 106 HeLa cells with phosphate-buffered saline and resuspended them in 200 μl of electroporation buffer (21 mM HEPES [pH 5.05], 137 mM NaCl, 0.7 mM KCl, 6 mM glucose). Cells were electroporated (280 V, 960 μF) with plasmid DNA (30 μg in 30 μl of Tris-EDTA, including 5 μg of pEGFP [Invitrogen]) in a 4-mm gap width electroporation cuvette. At 48 h posttransfection, green fluorescent protein-positive cells were sorted and isolated by flow cytometry with a Beckman-Coulter Altra cell sorter. Cells were plated at 1 cell per well into 96-well plates, and clonal cell lines were grown in Dulbecco's modified Eagle's medium containing 5% calf serum and 5% fetal calf serum. Whole-cell DNA was harvested at 6, 12, and 18 weeks posttransfection, digested with various restriction endonucleases, and separated on 1% agarose gels. The DNA was transferred to nylon membranes and hybridized to 32P-labeled probes (Southern blot protocol described above).
RESULTS
AAV infectivity.
Assays addressing wild-type virus integration efficiency are dependent on an accurate knowledge of the starting infectious virus. With standard virus purification protocols, as described in Materials and Methods, virus stocks were produced at a final titer of 1012 particles/ml, with 50 particles/infectious unit. Infectious units were determined with infectious center assays by serial dilutions with wild-type adenovirus type 5 coinfection (41). To determine the efficiency of site-specific integration with wild-type virus in our HeLa cell-based system, we first needed to know the percentage of cells that could be productively infected at various MOIs. We developed an assay derived from the infectious center assay to answer this question. HeLa cells were infected with wild-type AAV at MOIs ranging from 1 to 1,000. The infection was allowed to proceed for 24 h, sufficient time for AAV DNA to localize to the nucleus and uncoat (39). Adenovirus was then added as a helper at an MOI of 10 to initiate AAV replication in productively infected cells. The infection was allowed to proceed for 20 h, a time span that allows high-level AAV replication but is insufficient for the infection to spread to neighboring cells (14).
Single-cell suspensions were made and blotted so as to distribute approximately 50 cells per filter. The filters were hybridized to radiolabeled wild-type AAV DNA and exposed to film. Filters were stripped and then rehybridized to radiolabeled genomic HeLa DNA. The fraction of infected cells at the various MOIs was determined by dividing the number of AAV-positive cells by the total number of cells per filter. A representative filter showing both wild-type AAV and HeLa signals from an AAV MOI 100 infection is shown in Fig. 1A. By comparing the signals from each filter, the percentage of cells infected by AAV can be determined for each MOI. Figure 1B is a graph of these data, representing the averages of four filters for each MOI. As expected, increasing the MOI resulted in an increase in the percentage of cells infected, with maximal levels of infection occurring around an MOI of 10. We never observed 100% infected cells even at very high MOIs, indicating either that some HeLa cells are refractory to infection or that our assay is not sufficiently sensitive to detect every event.
We next assayed the ability of the virus to maintain a latent state. HeLa cells were infected with wild-type AAV at various MOIs as described above but with no subsequent adenovirus infection. After 48 h, each infected cell population was passaged and plated into 96-well dishes at an estimated single cell/well. Two weeks later, wells were microscopically inspected to identify wells containing individual colonies. We noted that the number of colonies per plate was similar for each MOI. This suggested that wild-type AAV infection MOIs of up to 1,000 have little effect on cell survival. The colonies were passaged into duplicate 96-well dishes and allowed to grow to near confluency. One of the duplicates was passaged for continued growth, and the second was used in an AAV rescue assay.
Virus rescue.
For the AAV rescue assay, which we performed approximately 3 weeks after the initial AAV infection, each clonal cell line was infected with wild-type adenovirus at an MOI of 10 to rescue and expand latent AAV. Only cells containing at least one complete copy of AAV, including intact inverted terminal repeats, will be subject to rescue and replication (6). The adenovirus infection was allowed to proceed to complete cytopathic effect (48 to 72 h), whereupon the cells were lysed and denatured as described in Materials and Methods. Cells were microscopically inspected for complete lysis and transferred to a nylon membrane with a dot blot apparatus. The membranes were hybridized to a radiolabeled wild-type AAV DNA probe and exposed to film. A representative blot showing clonal cell lines rescued from a wild-type AAV MOI 100 infection is shown in Fig. 2A. A graph showing the results from MOIs 1 to 1,000 is shown in Fig. 2B; each point represents at least 36 cell lines tested (see Table 1). At an MOI of 100, 37% of cell lines were positive for rescue. Despite having similar levels of infection at 24 h at an MOI of 100 (Fig. 1B), an MOI of 10 gave only 1 cell line positive for rescue among 90 tested. MOIs above 100 do not appear to significantly increase the percentage of AAV-rescuable cell lines. Interestingly, this maximal level of rescue is approximately half the number of cells infected.
Site-specific integration.
The cell lines used in each infection were expanded into 10-cm dishes, and 6 weeks postinfection, cells were harvested and total genomic DNA was isolated for characterization of AAV integration and AAVS1 disruption. Genomic DNA from each cell line was digested with EcoRI and subjected to Southern blot analysis with either a wild-type AAV-specific or an AAVS1-specific probe as previously described (26). Figure 3 represents the percentage of cell lines that contained an AAV integrant and the percentage of cell lines that indicated disruption of the AAVS1 site for each of the MOIs. Each point in the graph represents analysis of 16 to 48 individual cell lines (Table 1). Figure 4A presents a diagrammatic representation of the chromosome 19 site, showing the sites for restriction enzymes EcoRI, ApaLI, and BamHI, and the restriction fragment lengths of AAVS1-containing DNA fragments. Representative Southern blots of MOI 100 infections digested with EcoRI (Fig. 4B), BamHI (Fig. 4C), and ApaLI (Fig. 4D) are shown probed with the AAV probe or the AAVS1 probe. For each cell line that contained integrated AAV, in at least two of the digests, one of the AAV bands comigrated with an AAVS1 genomic fragment, indicating that site-specific integration efficiency approached 100%.
The characterization of the genomic DNA from each of the cell lines led us to several observations with respect to Rep-mediated integration. As was observed with the AAV rescue assay in Fig. 2, we found an integration efficiency plateau of approximately 35 to 40% of cell lines are reached at an MOI of 100. Nearly all of the rescuable cell lines characterized in Fig. 2 corresponded to an AAVS1 integrant, and 75% of the integrants were able to rescue, indicating a high correlation between site-specific integration and rescue.
The percentage of cell lines that demonstrated site-specific integration was consistently half the number of cell lines that contained AAVS1 disruptions. This ≈1:2 ratio of AAV integrants to AAVS1 disruptions was also seen at other MOIs (Fig. 3), as well as in previous studies with transient transfections of Rep (27). It is also interesting that digests with either EcoRI (which cuts wild-type AAV) or ApaL1 (which does not digest wild-type AAV DNA) both yielded a new AAVS1 band that had significantly increased in size irrespective of the presence of integrated AAV DNA. Based on a rough approximation, these bands increased over the wild-type restriction fragment length by between 5 to 10 kb, whereas BamHI digestion resulted in a more heterogeneous banding pattern in each cell line. We were also struck by the indication that in both disrupted cell lines and integrated cell lines, there appeared to be an increase in AAVS1 DNA.
AAVS1 amplification.
Randomly selected cell lines from the MOI 100 infections were characterized for differences in band intensity of the AAVS1 fragment as shown in Fig. 5. Three classes of clonal cell lines were selected: cell lines containing nondisrupted AAVS1 sites; cell lines with disrupted AAVS1 sites but no integrated virus and; cell lines with disrupted AAVS1 and integrated AAV. Genomic DNA from three samples of each class was digested with ApaLI and analyzed as described above. Briefly, Southern blots were first probed with a probe for AAVS1 and analyzed for band intensity with a phosphorimager. The blots were then stripped and rehybridized to a probe for β-actin as a loading control (Fig. 5A). The first lane contains HeLa genomic DNA from uninfected cells as a standard. Band intensity data are graphed in Fig. 5B. As expected, cell lines with no disruptions exhibited a negligible change in AAVS1 band intensity. However, cell lines with AAVS1 disruptions, whether in the presence or absence of integrated viral DNA, showed high variability in the amount of AAVS1 DNA present. Despite the wide range of band intensities, all disrupted cell lines contained amplified AAVS1 DNA, with the overall amplification being 1.8 ± 0.6%. Given that HeLa cells are on average triploid for ch19 (21) and AAV has never been observed to integrate into more than one locus (5), the overall amplification of an individual AAVS1 integration locus may be as high as threefold. The striking similarity between AAVS1 phenotype in both nonintegrated and integrated cell lines led us to question if Rep protein could directly act to induce an AAVS1 amplification or rearrangement event in the absence of a viral integration substrate.
p5IEE independent disruption.
HeLa cells were transfected with a plasmid, p78Rep, which expresses Rep but does not contain the p5IEE signal for site-specific integration (27). Clonal cell lines were established following transfection of either p78Rep or p78Rep and p5IEE-CAT as described in Materials and Methods. Cell lines were expanded, and genomic DNA was characterized 6 weeks posttransfection. AAVS1 disruptions were detected in 27% of pRep78-transfected cell lines (11 of 40) and in 27% of pRep78 plus p5IEE-CAT clonal cell lines (8 of 30) (Fig. 6). When genomic blots were probed for CAT, 13% of p78 plus p5IEE-CAT cell lines (4 of 30) retained the CAT sequence. The results of the transfection integration assay are consistent with Rep rearrangement of AAVS1 in the absence of a p5IEE integration substrate. Rep-mediated rearrangements of AAVS1 have been previously identified as abortive integration events (34). Also, when a p5IEE integration substrate is present, approximately half of the cell lines containing AAVS1 disruptions also contain an integrant. These data are also consistent with the data obtained with wild-type virus infection, where integration was also limited to approximately 50% of AAVS1-disrupted cell lines (Fig. 2 and Table 1).
DISCUSSION
In this study, we characterized the effect of MOI on the efficiency of Rep-mediated site-specific integration with AAV. The results obtained indicate several new insights into the biology of AAV Rep-mediated site-specific integration, provide important supporting data for aspects of AAV integration that have previously been implied but not directly demonstrated, and corroborate and extend early studies that characterized AAV rescue from latently infected cells.
What is an infectious unit of AAV? Standard protocol for determining AAV infectious units have relied on limiting dilutions of AAV and coinfection with wild-type adenovirus, which results in a high-yield AAV replication (14). This assay, although an accurate reflection of one aspect of the biology of AAV (i.e., that occurring in coincidence with helper virus infection), may not provide the most useful insight into the biology of latent wild-type AAV or recombinant AAV vectors. Helper virus has been shown to potentially influence the biology of AAV by facilitating virus transport to the nucleus (39) and by enhancing conversion of the single-stranded AAV genome into a transcriptionally active duplex DNA substrate (9, 10). In the absence of helper virus, AAV is dependent on interaction with the cell for internalization and genome transport to the nucleus. One of the most striking observations made in our studies was the observed disparity between the efficiency of infection as characterized by traditional techniques and our characterization of viral rescue, AAVS1 disruptions, and AAV integrants into the AAVS1 site. These studies were performed in actively dividing cells. There may be differences between dividing and nondividing cells with respect to infection, uncoating, and integration efficiencies that have not been addressed in this report. Our studies demonstrate that AAV integration into chromosome 19 may be a useful alternative measurement of the AAV infectious unit.
High MOI, then, may be a necessary requirement for AAV site-specific integration. Certainly, high MOI was correlated with both latency and integration in several studies prior to the discovery of site-specific integration (3, 7). Interestingly, in these early studies, low MOIs (0.25 and 2.5) failed to yield persistent antigen-producing cells after 40 passages, whereas 29% of the cell lines established after 40 passages postinfection with an MOI of 250 were able to produce viral antigen (4). Considering differences in techniques and cell lines, we find a remarkable consistency between our observations of integration efficiency and those of earlier studies. The data suggest that, for wild-type AAV, latency, integration, and site-specific integration may in fact be different ways of measuring the same event.
What are the steps that lead to AAV integration into AAVS1? Studies in dividing cells demonstrate that AAV uptake is efficient, and virions localize to the perinuclear region within 2 h postinternalization (1, 29, 39). The efficiency of capsid uncoating and transport of viral DNA to the nucleus is a potential rate-limiting step that may be influenced by viral MOI. The model presented in Fig. 7 indicates a proposed sequence of events that occur during the process of AAV integration. Following localization to the nucleus, cellular replication and/or DNA repair mechanisms are essential for second-strand synthesis of the AAV genome. Alternatively, in the absence of efficient second-strand synthesis, a second pathway for production of transcriptionally active duplex AAV genome would be through plus- and minus-strand hybridization. Subsequent to the formation of duplex genome production, the p5 promoter is activated to transcribe Rep (Fig. 7A.2). Transcription from the p5 promoter and translation of Rep mRNA yield the multifunctional Rep proteins (for the sake of simplicity, we will focus on Rep 68 and 78 because they are able to mediate integration in the absence of the smaller Rep proteins [42 and 50]). Production of the duplex genome is a necessary prerequisite for p5 transcription of the rep gene and is a likely substrate target DNA for integration, as duplex plasmid DNA integrates with high efficiency. We could not distinguish between duplex formation or Rep expression levels as a specific rate-limiting entity.
An additional limitation may simply be template copy number at low MOI. In a latent infection, Rep expression levels are presumably very low, and we were unable to detect either rep mRNA or protein in the absence of helper virus coinfection. Following Rep protein expression, there are three biologically active target DNAs that Rep can bind relevant to integration: Rep binds to the AAVS1 site on chromosome 19; Rep binds to the p5 Rep binding site of AAV; Rep binds to the inverted terminal repeat Rep binding site. Data obtained through transfection of plasmid DNA corroborate previous studies which show that the inverted terminal repeat elements are not required for efficient integration; therefore, we view the role of the inverted terminal repeat as essential for maintaining the fidelity of the integrating genome and for rescue from latency but not as playing a role in mediating a Rep-dependent integration event.
In our view, integration efficiency is defined by Rep interaction with the AAVS1 site (Fig. 7B) and by Rep interaction with the p5 promoter (Fig. 7A). We have illustrated Rep to function as a hexameric complex, but this has not been confirmed to be the complex that is acting on the AAVS1 site or on p5. We have demonstrated that Rep expression in the absence of helper virus or in the absence of integration substrate mediates a rearrangement of the AAVS1 site. The AAVS1 site rearrangements are consistent with a local amplification event (depicted in Fig. 7B.3 and 7C.2) and demonstrate a significant increase in restriction fragment length size (roughly 5 to 10 kb). The amplification event, as measured by phosphoimager quantitation of Southern analysis, indicates a ≈2-fold increase in AAVS1. Since HeLa cells are aneuploid and have been characterized as having three copies of chromosome 19, we would predict that, if localized to a single chromosome, the rearrangement would contain on average three new repeat elements of the AAVS1 region being probed (Fig. 7C). When an integration substrate (Fig. 7A.2) is present in sufficient quantities, localization of the p5IEE integration substrate to the AAVS1 integration site takes place, and integration occurs through what is predicted to be a strand switch mechanism (19) or perhaps ligation between cellular and viral DNA mediated by Rep (30). In our characterization of AAV integration, increasing the copy number of input virus above an MOI of 100 had no significant effect on the final yield of integration products. These observations indicate that the process of integration is at some level regulated autonomously.
One important level of autonomous regulation of AAV integration is through regulation of Rep expression. Rep binding to the p5 promoter (p5IEE) has been shown to bring about a stringent block to Rep transcription (2, 18). The repression of rep transcription therefore implies that there is a narrow window of opportunity for Rep-mediated site-specific integration to occur. Consistent with this hypothesis, Huser et al. demonstrated that maximal integration of AAV into chromosome 19 occurs within the first 4 days postinfection (15). Studies by Giraud et al. demonstrated that targeting to plasmid-based AAVS1 targets occurred by 24 h postinfection (11). Based on our measurements of integration yield as a function of input virus, at an input MOI of 100, 80% of cell lines demonstrate AAVS1 rearrangement and half of those cell lines contain integrants. These are the maximal number of events that occur regardless of input viral dose.
In our view, this result offers a critical insight into the integration pathway. Because increasing the dose of virus does not yield an increase in the occurrence of integrants, we would argue that there is an inherent limit to the number of discernible recombination events that can occur when a cell is infected, regardless of the number of available integration substrate targets. In this model, activation of AAVS1 establishes the kinetics of integration. Presentation of an integration substrate during the time of AAVS1 rearrangement creates the opportunity for a copy choice mechanism to integrate the AAV substrate DNA. Since we consistently found that half of the AAVS1 rearrangements contained integrants, perhaps recombination occurs during cellular DNA replication and /or cell division. In all cell lines examined, regardless of the viral dose used for infection, we always found at least one copy of an intact chromosome 19 AAVS1 locus.
In addition to providing insight into the biology of integration, the data presented in this study revealed a high correlation between rescuable replication-competent AAV in cell lines and cell lines that contain site-specific integrants (Table 1). Furthermore, careful analysis of selected MOI 100 cell lines revealed that virtually all integrations were site specific. These remarkable correlations provide strong support for a model in which the AAV genomes that persist in human cells do so primarily as AAVS1 integrants.
FIG. 1.
FIG. 2.
FIG. 3.
FIG. 4.
FIG. 5.
FIG. 6.
FIG. 7.
TABLE 1.
| MOI | % cells infected | % cells rescued (no. rescued/total) | % cells disrupted (no. disrupted/total) | % cells with integrant (no. integrated/total) |
|---|---|---|---|---|
| 1 | 36 ± 9 | 0 (0/82) | 0 (0/44) | 0 (0/44) |
| 5 | 67 ± 13 | 0 (0/73) | 0 (0/48) | 0 (0/48) |
| 10 | 77 ± 4 | 1.1 (1/90) | 28 (7/25) | 8 (2/25) |
| 100 | 84 ± 7 | 38 (34/89) | 81 (24/33) | 36 (12/33) |
| 500 | 84 ± 11 | 33 (29/88) | 76 (28/37) | 32 (12/37) |
| 1,000 | 91 ± 9 | 42 (15/36) | 81 (13/16) | 43 (7/16) |
Acknowledgments
We thank Jose Trevejo, Peter Ward, and Nicola Philpott for helpful comments and discussion.
This work was supported in part by NIH funding (RO1GM067102) to E.F.-P.
REFERENCES
1.
Bartlett, J. S., R. Wilcher, and R. J. Samulski. 2000. Infectious entry pathway of adeno-associated virus and adeno-associated virus vectors. J. Virol.74:2777-2785.
2.
Beaton, A., P. Palumbo, and K. I. Berns. 1989. Expression from the adeno-associated virus p5 and p19 promoters is negatively regulated in trans by the rep protein. J. Virol.63:4450-4454.
3.
Berns, K. I., T. C. Pinkerton, G. F. Thomas, and M. D. Hoggan. 1975. Detection of adeno-associated virus (AAV)-specific nucleotide sequences in DNA isolated from latently infected Detroit 6 cells. Virology68:556-560.
4.
Berns, K. I., and R. M. Linden. 1995. The cryptic life style of adeno-associated virus. Bioessays17:237-245.
5.
Berns, K. I. 1996. Parvoviridae: the viruses and their replication, p. 1017-1041. In B. N. Fields (ed.), Fundamental virology, 3rd ed. Lippincott-Raven Publishers, Philadelphia Pa.
6.
Bohenzky, R. A., R. B. LeFebvre, and K. I. Berns. 1988. Sequence and symmetry requirements within the internal palindromic sequences of the adeno-associated virus terminal repeat. Virology166:316-327.
7.
Cheung, A. K., M. D. Hoggan, W. W. Hauswirth, and K. I. Berns. 1980. Integration of the adeno-associated virus genome into cellular DNA in latently infected human Detroit 6 cells. J. Virol.33:739-748.
8.
Dutheil, N., F. Shi, T. Dupressoir, and R. M. Linden. 2000. Adeno-associated virus site-specifically integrates into a muscle-specific DNA region. Proc. Natl. Acad. Sci. USA97:4862-4866.
9.
Ferrari, F. K., T. Samulski, T. Shenk, and R. J. Samulski. 1996. Second-strand synthesis is a rate-limiting step for efficient transduction by recombinant adeno-associated virus vectors. J. Virol.70:3227-3234.
10.
Fisher, K. J., G. P. Gao, M. D. Weitzman, R. DeMatteo, J. F. Burda, and J. M. Wilson. 1996. Transduction with recombinant adeno-associated virus for gene therapy is limited by leading-strand synthesis. J. Virol.70:520-532.
11.
Giraud, C., E. Winocour, and K. I. Berns. 1994. Site-specific integration by adeno-associated virus is directed by a cellular DNA sequence. Proc. Natl. Acad. Sci. USA91:10039-10043.
12.
Giraud, C., E. Winocour, and K. I. Berns. 1995. Recombinant junctions formed by site-specific integration of adeno-associated virus into an episome. J. Virol.69:6917-6924.
13.
Hacein-Bey-Abina, S., C. von Kalle, M. Schmidt, F. Le Deist, N. Wulffraat, E. McIntyre, I. Radford, J. L. Villeval, C. C. Fraser, M. Cavazzana-Calvo, and A. Fischer. 2003. A serious adverse event after successful gene therapy for X-linked severe combined immunodeficiency. N. Engl. J. Med.348:255-256.
14.
Hauswirth, W. W., A. S. Lewin, S. Zolotukhin, and N. Muzyczka. 2000. Production and purification of recombinant adeno-associated virus. Methods Enzymol.316:743-761.
15.
Huser, D., S. Weger, and R. Heilbronn. 2002. Kinetics and frequency of adeno-associated virus site-specific integration into human chromosome 19 monitored by quantitative real-time PCR. J. Virol.76:7554-7559.
16.
Inoue, N., R. Dong, R. K. Hirata, and D. W. Russell. 2001. Introduction of single base substitutions at homologous chromosomal sequences by adeno-associated virus vectors. Mol. Ther.3:526-530.
17.
Kotin, R. M., M. Siniscalco, R. J. Samulski, X. D. Zhu, L. Hunter, C. A. Laughlin, S. McLaughlin, N. Muzyczka, M. Rocchi, and K. I. Berns. 1990. Site-specific integration by adeno-associated virus. Proc. Natl. Acad. Sci. USA87:2211-2215.
18.
Kyostio, S. R., R. S. Wonderling, and R. A. Owens. 1995. Negative regulation of the adeno-associated virus (AAV) P5 promoter involves both the P5 rep binding site and the consensus ATP-binding motif of the AAV Rep68 protein. J. Virol.69:6787-6796.
19.
Linden, R. M., E. Winocour, and K. I. Berns. 1996. The recombination signals for adeno-associated virus site-specific integration. Proc. Natl. Acad. Sci. USA93:7966-7972.
20.
Linden, R. M., and K. I. Berns. 1997. Site-specific integration by adeno-associated virus: a basis for a potential gene therapy vector. Gene Ther.4:4-5.
21.
Macville, M., E. Schrock, H. Padilla-Nash, C. Keck, B. M. Ghadimi, D. Zimonjic, N. Popescu, and T. Ried. 1999. Comprehensive and definitive molecular cytogenetic characterization of HeLa cells by spectral karyotyping. Cancer Res.59:141-150.
22.
Meneses, P., K. I. Berns, and E. Winocour. 2000. DNA sequence motifs which direct adeno-associated virus site-specific integration in a model system. J. Virol.74:6213-6216.
23.
Meyers, C., M. Mane, N. Kokorina, S. Alam, and P. L. Hermonat. 2000. Ubiquitous human adeno-associated virus type 2 autonomously replicates in differentiating keratinocytes of a normal skin model. Virology272:338-346.
24.
Miller, S. A., D. D. Dykes, and H. F. Polesky. 1988. A simple salting out procedure for extracting DNA from human nucleated cells. Nucleic Acids Res.16:1215.
25.
Nakai, H., E. Montini, S. Fuess, T. A. Storm, M. Grompe, and M. A. Kay. 2003. AAV serotype 2 vectors preferentially integrate into active genes in mice. Nat. Genet.34:297-302.
26.
Philpott, N. J., C. Giraud-Wali, C. Dupuis, J. Gomos, H. Hamilton, K. I. Berns, and E. Falck-Pedersen. 2002. Efficient integration of recombinant adeno-associated virus DNA vectors requires a p5-rep sequence in cis. J. Virol.76:5411-5421.
27.
Philpott, N. J., J. Gomos, K. I. Berns, and E. Falck-Pedersen. 2002. A p5 integration efficiency element mediates Rep-dependent integration into AAVS1 at chromosome 19. Proc. Natl. Acad. Sci. USA99:12381-12385.
28.
Samulski, R. J., X. Zhu, X. Xiao, J. D. Brook, D. E. Housman, N. Epstein, and L. A. Hunter. 1991. Targeted integration of adeno-associated virus (AAV) into human chromosome 19. EMBO J.10:3941-3950.
29.
Sanlioglu, S., P. K. Benson, J. Yang, E. M. Atkinson, T. Reynolds, and J. F. Engelhardt. 2000. Endocytosis and nuclear trafficking of adeno-associated virus type 2 are controlled by rac1 and phosphatidylinositol-3 kinase activation. J. Virol.74:9184-9196.
30.
Smith, R. H., and R. M. Kotin. 2000. An adeno-associated virus (AAV) initiator protein, Rep78, catalyzes the cleavage and ligation of single-stranded AAV ori DNA. J. Virol.74:3122-3129.
31.
Srivastava, A., E. W. Lusby, and K. I. Berns. 1983. Nucleotide sequence and organization of the adeno-associated virus 2 genome. J. Virol.45:555-564.
32.
Surosky, R. T., M. Urabe, S. G. Godwin, S. A. McQuiston, G. J. Kurtzman, K. Ozawa, and G. Natsoulis. 1997. Adeno-associated virus Rep proteins target DNA sequences to a unique locus in the human genome. J. Virol.71:7951-7959.
33.
Tan, I., C. H. Ng, L. A Lim, and T. A Leung. 2001. Phosphorylation of a novel myosin binding subunit of protein phosphatase 1 reveals a conserved mechanism in the regulation of actin cytoskeleton. J. Biol. Chem.276:21209-21216.
34.
Urabe, M., K. Kogure, A. Kume, Y. Sato, K. Tobita, and K. Ozawa. 2003. Positive and negative effects of adeno-associated virus Rep on AAVS1-targeted integration. J. Gen. Virol.84:2127-2132.
35.
Urcelay, E., P. Ward, S. M. Wiener, B. Safer, and R. M. Kotin. 1995. Asymmetric replication in vitro from a human sequence element is dependent on adeno-associated virus Rep protein. J. Virol.69:2038-2046.
36.
Wang, X. S., and A. Srivastava. 1997. A novel terminal resolution-like site in the adeno-associated virus type 2 genome. J. Virol.71:1140-1146.
37.
Ward, P., and K. I. Berns. 1995. Minimum origin requirements for linear duplex AAV DNA replication in vitro. Virology209:692-695.
38.
Weitzman, M. D., S. R. Kyostio, R. M. Kotin, and R. A. Owens. 1994. Adeno-associated virus (AAV) Rep proteins mediate complex formation between AAV DNA and its integration site in human DNA. Proc. Natl. Acad. Sci. USA91:5808-5812.
39.
Xiao, W., K. H. Warrington, Jr., P. Hearing, J. Hughes, and N. Muzyczka. 2002. Adenovirus-facilitated nuclear translocation of adeno-associated virus type 2. J. Virol.76:11505-11517.
40.
Young, S. M., Jr., and R. J. Samulski. 2001. Adeno-associated virus (AAV) site-specific recombination does not require a Rep-dependent origin of replication within the AAV terminal repeat. Proc. Natl. Acad. Sci. USA98:13525-13530.
41.
Zolotukhin, S., B. J. Byrne, E. Mason, I. Zolotukhin, M. Potter, K. Chesnut, C. Summerford, R. J. Samulski, and N. Muzyczka. 1999. Recombinant adeno-associated virus purification using novel methods improves infectious titer and yield. Gene Ther.6:973-985.
Information & Contributors
Information
Published In
Copyright
© 2004 American Society for Microbiology.
History
Received: 23 February 2004
Accepted: 6 April 2004
Published online: 1 August 2004
Contributors
Metrics & Citations
Metrics
Note:
- For recently published articles, the TOTAL download count will appear as zero until a new month starts.
- There is a 3- to 4-day delay in article usage, so article usage will not appear immediately after publication.
- Citation counts come from the Crossref Cited by service.
Citations
Citation
2004.
Adeno-Associated Virus Site-Specific Integration and AAVS1 Disruption. J Virol
78:.
If you have the appropriate software installed, you can download article citation data to the citation manager of your choice. For an editable text file, please select Medlars format which will download as a .txt file. Simply select your manager software from the list below and click Download.