Figure 1.
Linker-selection-mediated (LSM) PCR for cloning of chromosomal AAV integration sites.
(A) Genome structure of AAV-2 with the rep and cap genes and their promoters flanked by inverted terminal repeats (ITRs) at either end of the ssDNA genome. The hairpin-structured ITRs contain internal repeat elements and complements thereof, represented by small letters. The positions of the Rep-binding sites (RBS) are represented by asterisks. (B) Schematic representation of the LSM-PCR strategy for amplification of AAV-chromosomal junction fragments. Wavy lines indicate chromosomal DNA. Linkers are displayed by thick, grey lines, AAV-specific primers by small, horizontal arrows. For restriction enzyme digestion indicated by vertical arrows either one of the following enzymes were used: PvuII or EcoRV (non-cut for AAV-2) or DraI (single-cut in AAV-2).
Figure 2.
Chromosomal distribution of AAV-2 integration sites.
DNA of AAV-2 wildtype infected Hela cells was analyzed for viral integration by the LSM-PCR-method. DNA sequence data from 117 cloned AAV-chromosome junctions were assigned to unique loci. (A) Distributions of junctions on individual chromosomes are shown as percentage of the total 117. (B–D) AAV integration sites drawn to scale for chromosome 19 (B), 5 (C), and 3 (D). Shown are the chromosome ideograms and enlarged bands of hotspots found on chr. 19q13.42 (AAVS1) and chr. 5p13.3 (AAVS2). Solid arrows represent sites detected upon wildtype AAV-2 infection. Open arrows represent junctions stemming from cotransfection of AAV vector- and Rep-expression plasmids.
Table 1.
Summary of data sets analysed in this study.
Figure 3.
Sequence alignment of Rep-binding sites.
RBS elements present in the ITR and the p5 promoter of all known AAV serotypes are aligned and related to consensus RBS sites present at chromosomal integration hotspots. Rep binding element GAGC is displayed in bold letters. Both GAGC and GAGT elements are highlighted in grey.
Figure 4.
Binding of MBP-Rep78 to Rep-binding sites of AAV-2 and of chromosomal integration hotspots.
(A) to (C) Electrophoretic mobility shift assays (EMSA) were performed with 32P-labeled double-stranded oligonucleotides in the presence of increasing amounts of affinity-purified MBP-Rep78 as indicated above the autoradiograms. (D) Quantitative determination of the bound fractions of the different RBS and control oligonucleotide probes as a function of the amount of MBP-Rep78 protein in the binding reaction. EMSA gels shown in (A) to (C) were subjected to phosphorimager analysis to determine the relative amount of unbound and bound 32P-labeled oligonucleotides. The relative binding affinity was calculated as follows: The highest amount of Rep used in this assay (1000 ng) bound 31% of the random oligonucleotide. The amount of Rep that bound the same fraction of the other oligonucleotides was determined and normalized to the binding of the chr. 19 (AAVS1) oligonucleotide.
Table 2.
Genomic features and chromatin state associated with AAV-2 wildtype- and AAV vector-derived integration sites.
Figure 5.
Statistical analysis of distances of integration sites from human Rep-binding motifs.
AAV integration sites of different data sets were analyzed for the proximity to the next Rep binding sites. Calculations were performed with the following putative Rep binding sites: GAGC GAGC (A); GAGT GAGC (B); GAGY GAGC GAGC (C); GAGC GAGC GAGC, one mismatch (D), or two mismatches allowed (E), and GAGY GAGC GAGA (F). Average distances of integration sites from putative Rep binding sites are displayed as mean +/− S.D. The levels of significance are marked by asterisks. P-values >0.05 were not considered statistically significant. P-values <0.01 were considered highly significant. For the analysis of wildtype AAV integration data a PvuII-related control set was used in addition to the random control set.
Table 3.
Neighbourhood analysis of wildtype AAV-2 integration sites and RBS motifs found outside of the hotspots on chr. 19q13.42 (AAVS1) and on chr. 5p13.3 (AAVS2).