Evaluation of the loading capacity and patterns of packaged DNA in AAV genomes of different sizes using long-read sequencing

Abstract

The loading capacity of adeno-associated virus (AAV) vectors is reportedly 4.7-5.0 kb, which limits the size of genes that can be treated with gene therapy. However, the effects of oversized genomes on the integrity of packaged AAV genomes are poorly understood. Herein, nanopore long-read sequencing was used to evaluate genomic integrity in AAV vectors harboring genomes of various sizes. AAV had a reduced proportion of full-length genomes at a vector length of 4.9 kb, which declined rapidly between 4.9 and 5.0 kb. This was mainly attributable to defects in genome packaging rather than genome synthesis. Furthermore, the pattern of packaged DNA was unique to the arrangement of the components of the oversized genome. However, an 86.3% reduction in the proportion of full-length genomes (4.7 vs. 5.0 kb) was not consistent with the retained expression of the reporter gene in the mouse retina. This discrepancy might be partially attributable to the preferential inclusion of the region containing the reporter gene. These results highlight the utility of long-read sequencing in assessing the genomic integrity and design of AAV vectors, as the pattern of packaged genomes appears to be unique to each vector, particularly for oversized AAV genomes.

Keywords: AAV genome loading capacity; AAV genome packaging pattern; AAV vector; adeno-associated virus; full-length genome proportion; gene therapy; long-read sequencing; nanopore sequencing.

Graphical abstract

Figure 1

Vectors of different sizes Vectors of different sizes were constructed, and each plasmid-encoded vector was evaluated using nanopore sequencing. (A) The structure of the 3.5–5.5 kb AAV8 genome. The vectors contained a fixed-size EGFP expression cassette and lambda phage DNA of various lengths to alter the length of the vector. (B) The plasmid-encoded vectors of different sizes were quantified and confirmed using TapeStation. (C) The plasmids (3.5–5.5 kb) were analyzed using nanopore sequencing. The AAV genome components are displayed above the sequence alignment. Each representative read alignment is displayed in IGV; nucleotides that match the reference sequence are shown in gray. The graph below each read alignment shows the count of the start and end positions of the reads; the start positions are indicated in blue, and the end positions are indicated in orange. The percentages of start and end positions of each read inside the ITRs are shown in the graph.

Figure 2

Rapid decline in the proportion of full-length AAV genomes between 4.9 and 5.0 kb AAV vectors with various genome sizes were evaluated using (A) agarose gel electrophoresis, (B) TapeStation, and (C) nanopore sequencing. All evaluations showed a rapid decline in the proportion of full-length genomes between 4.9 and 5.0 kb. (D) The 4.7 kb had a significantly higher proportion of full-length genomes than the 5.0 kb using nanopore sequencing (N = 3 each, p = 0.01). Data are represented as mean (standard deviation). The p value was calculated using a t test (two-sided).

Figure 3

Agarose gel electrophoresis before and after DNase treatment with and without capsid gene Without DNase treatment, the 5.5 kb genome was clearly visible in cap (−) samples but not in cap (+) samples, whereas the 3.5 kb genome was visible in both cap (−) and cap (+) samples. With DNase treatment, only the 3.5 kb genome with cap (+) was detected.

Figure 4

Decreased proportion of full-length AAV genomes coupled with an increased proportion of short genomes retaining only EGFP and polyA at the 3′ end Nanopore sequencing analysis of the AAV genome with functional components. A schematic representation of the AAV vector constructs is shown at the top of the figure. (A) Analysis of the 10 major patterns of the packaged genome in AAVs of different sizes. The black circles indicate the fully mapped components. The bar graph above the map indicates the percentage of reads within each AAV vector. The bar graph in the lower left displays the percentage of reads for each genomic component. The pattern was similar for AAV genomes from 3.5 to 4.8 kb in length, but small changes were observed in AAV genomes from 4.9 kb and larger. The 5.0 and 5.5 kb genomes show a completely different pattern. (B) Changes in the patterns of the packaged genome for AAV vectors of different sizes. The ratio between the percentages of full reads from lamba phage to polyA and the percentage of reads from EGFP to polyA was calculated and plotted on a logarithmic graph. A change was observed in genomes above 4.9 kb, with the 5.0 and 5.5 kb showing markedly different patterns.

Figure 5

Reporter EGFP expression in mice using oversized AAV vectors with low proportions of full-length genomes EGFP expression was assessed in mouse retinas after subretinal injections with AAV vectors of various sizes. (A) A significant number of EGFP-positive photoreceptors were observed after the injection of AAV8-EGFP vectors with genome sizes ranging from 3.5 to 5.5 kb. A rapid decline in reporter protein expression was not detected between 4.7 and 5.0 kb genomes. Scale bars: 20 μm. (B) Western blotting of EGFP after injection of AAVs with genome sizes of 4.7, 5.0, and 5.5 kb. (C) Quantification of the western blotting revealed unchanged EGFP expression at 5.0 kb followed by a moderate decline at 5.5 kb relative to 4.7 kb AAV vector genome sizes. INL, inner nuclear layer; OPL, outer plexiform layer; ONL, outer nuclear layer; IS/OS, inner segment/outer segment; NoTx, no treatment.

Figure 6

Retained reporter gene and expression at the 3′ end in AAV with an oversized genome (A) An oversized AAV vector (5.3 kb) with Venus at the 5′ end and mKO1 at the 3′ end (AAV8.CAG.Venus-GRK1-702.mKO1) was analyzed using nanopore sequencing. The packaged genome was observed at both the 5′ and 3′ ends. (B) Analysis of the patterns of the packaged genome by functional genomic components in AAV8.CAG.Venus-GRK1-702.mKO1. (C) The western blotting of murine retinas injected with AAV.CAG.Venus-GRK1-702.mKO1. Marked Venus and mKO1 expression was confirmed. (D) Representative retinal histology showing Venus and mKO1 expression in the photoreceptors of mice treated with the AAV vector. Scale bars: 20 μm. INL, inner nuclear layer; OPL, outer plexiform layer; ONL, outer nuclear layer; IS/OS, inner segment/outer segment; mKO1, monomeric Kusabira-Orange 1; NoTx, no treatment.