Fig 1.
Generation and characterization of the AAV8 vector library. (a) Dot plot showing the production yield of each AAV8 capsid variant in a small-scale vector production assay packaging the same EGFP transgene cassette. Each dot represents a unique capsid variant. The production yield is normalized to the AAV8 vector level (defined as 100%) and ranked from the highest to the lowest. Data were based on one biological repeat. (b) Bar graph showing the count of Nanopore sequencing reads mapped to the unique vector transgene barcodes packaged in AAV8, AAV9, or AAV8 variants. The dashed line shows the mean value. Data were based on one biological repeat.
Fig 2.
Vector library screen in mice and ferrets.
(a, b) Scatter dot plots showing the relative vector genome abundance of each barcoded capsid vector in mouse (a) or ferret (b) livers. Data are normalized to the AAV8 vector level (defined as 1.0), and presented as mean ± SD. Each dot represents an individual animal. v1 to v37 denote the identifiers of AAV8 variants. The variants in bold indicate that they resulted in low or barely detectable levels of vector DNA. (c) Scatter plot showing the relationship between the relative vector DNA abundance of each capsid variant in mouse liver (x-axis) and ferret liver (y-axis). Data are presented as mean and standard deviation of three animals. In (a) and (b), statistical analysis is performed using two-tailed one-way ANOVA followed by Dunnett’s multiple comparisons test against AAV8. ***p < 0.001. In (c), the linear regression statistics are shown.
Fig 3.
AAV8.v5 shows liver detargeting phenotype in mice.
(a) Bar graph showing the relative vector RNA (cDNA) abundance of each barcoded capsid vector in mouse liver (blue), heart (orange), and tibialis anterior (TA) muscle (green). Data are normalized to the AAV8 vector level in respective tissues (defined as 1), and presented as mean ± SD. Each dot represents an individual animal. v1 to v37 denote the identifiers of AAV8 variants. The variants in bold indicate that they resulted in low or barely detectable levels of vector RNA expression in the liver, with the ones underlined indicating well-detectable levels of vector RNA expression in the heart and/or TA muscle. Statistical analysis is performed using two-tailed one-way ANOVA to compare vector RNA levels in the three tissues transduced by each vector. *p < 0.05; **p < 0.01; ***p < 0.001; ns: not significant. (b) Alignment of the amino acid sequences of multiple AAV capsids. Only variable region I (VR-1) and surrounding residues are shown as single-letter abbreviations. Dashes indicate gaps. The 265insT (AAV3B VP1 numbering) residue described in Cabanes-Creus M. et al. 2021 [22], the liver toggle residue described in Zinn E. et al. 2022 [23], and the N271D (AAV8 VP1 numbering) residue described in this study are highlighted with gray background. The residues that differ from parental capsids described in these studies are highlighted in bold. The Warischalk study [24] investigated a wide range of VR-1 mutants across multiple serotype capsids.
Fig 4.
Characterization of vector performance in mice following individual vector administration.
(a-c) Box plots showing the vector DNA abundance (left), EGFP mRNA levels (middle), and relative EGFP protein levels (right) in the liver, heart, and tibialis anterior (TA) muscle collected from the mice treated with AAV8 or AAV8.N271D vectors (a), AAV9 or AAV9.N270D vectors (b), and MyoAAV or MyoAAV.N270D vectors (c). Note that the protein levels are relative to the parental capsid in each tissue. Each dot represents an individual mouse. The box extends from the first to the third quartiles with the line inside denoting median. The whiskers end at minimum and maximum values. The fold changes of medians and p values are labeled. Statistical analysis is performed using two-tailed non-parametric Mann-Whitney U test.