Mapping administration route-dependent transduction profiles of commonly used AAV variants in mice by barcode amplicon sequencing

Figures

• 

  Figure 1. Production of a barcoded AAV library to identify administration route-dependent tropism profiles (A) To determine AAV biodistribution dependence on the route of administration, 34 AAV variants harboring a CMV-eGFP expression cassette and unique 15 nt BC sequences in the 3′UTR were individually produced, purified, and quality controlled. Equimolar pooling generated a BC-AAV library that was administered to mice via intravenous (i.v.) and intraperitoneal (i.p.) injection, respectively. After 14 days, 18 tissues were harvested, and viral BCs were amplified from isolated DNA and RNA. BCs from viral genomes and transcripts in each tissue were quantified by next-generation amplicon sequencing, and AAV biodistribution profiles were determined. (B) Production efficiency of the 34 selected AAV variants, shown in vg per cm2 culture area of HEK293H producer cells. The mean productivity over all generated AAV variants is indicated by the dotted line (7.28 × 109 vg/cm2). AAV genome integrity (expected size: 2,086 bp) was analyzed by agarose gel electrophoresis with 2.5 × 1010 vg per lane (top). AAV capsid composition and batch purity was assessed by SDS-PAGE and oriole fluorescent gel…

• 

  Figure 2. Administration route-dependent biodistribution of BC-AAV library (A) Distribution of total BC-AAV library over 18 murine tissues after i.v. (top) and i.p. (bottom) administration, determined by quantification of viral genomes (CMV promoter probe) per diploid genome (Gapdh housekeeper probe) (vg/dg) for each tissue. Mean (SD) for n = 8 animals is shown. (B and C) Distribution of all 34 individual BC-AAV variants over 18 murine tissues after i.v. administration (left) and i.p. administration (right) on (B) viral transcript level (RNA) and (C) viral genome level (DNA), shown as normalized Bαβ values, reflecting the normalized count ratio (P∗αβ) of each BC in each tissue after normalization with the viral genomes per cell in the respective tissues (as shown in A). Mean Bαβ values of 8 C57BL/6 mice per administration route are depicted. (D) Correlation of log transformed normalized count ratios (P∗αβ) of each AAV BC quantified on DNA level (viral genomes) and RNA level (viral transcripts) for both i.v. and i.p. administrations combined. Pearson correlation coefficient r = 0.76. A linear regression curve is depicted as green solid line, 95% confidence intervals as black dotted lines.

• 

  Figure 3. Examples of AAV tropism differences after i.v. and i.p. administration (A) Top 10 AAV variants for brain, diaphragm, liver, and pancreas targeting are depicted after i.v. and i.p. administration, as determined by the normalized transcriptional efficiency (Vαβ). Mean (SD) of n = 8 mice per administration group. (B) Differences in transcriptional specificity (Tαβ) of AAV-PHP.eB (blue), AAV6.2 (yellow), AAV-DJ (green), and AAV1P4 (violet) between i.v. and i.p. administration are depicted as normalized proportion per cell. Mean (SD) of n = 8 mice per administration group.

• 

  Figure 4. Biodistribution of AAV-PHP.eB and AAV6.2 after i.v. and i.p. administration Fourteen days after administration of 1.0 × 1012 vg/animal via i.v. or i.p. injection, biodistribution of AAV-PHP.eB (A and B) and AAV6.2 (C and D) was determined. (A and C) eGFP expression in brain, diaphragm, eWAT, heart, iWAT, liver, lung, pancreas, and spleen was assessed by qPCR. eGFP copies were calculated and normalized to the detected housekeeper gene copies (Polr2a), based on eGFP/Polr2a plasmid-based standard curves. The fold change (FC) of normalized eGFP expression after i.p. over i.v. administration for all tissues is shown (top). For selected tissues (brain, diaphragm, eWAT, liver, and pancreas), the normalized copy ratios are depicted (bottom). n = 5 animals per administration group and n = 4 animals (2 per administration group) for PBS, mean (SD). Group comparison by unpaired Student’s t test (AAV-PHP.eB: brain and AAV6.2: diaphragm), unpaired Student’s t test with Welch’s correction (AAV-PHP.eB: eWAT, liver and AAV6.2: eWAT, pancreas), and Mann-Whitney test (AAV-PHP.eB: pancreas and AAV6.2: liver), exact p values. (B and D) Representative images of eGFP IHC are shown for the selected…

• 

  Figure 5. Biodistribution of AAV-DJ and AAV1P4 after i.v. and i.p. administration Fourteen days after administration of 1.0 × 1012 vg/animal via i.v. or i.p. injection, biodistribution of AAV-DJ (A and B) and AAV1P4 (C and D) was determined. (A and C) eGFP expression in brain, diaphragm, eWAT, heart, ileum, iWAT, kidney, liver, lung, pancreas, and spleen was assessed by qPCR. eGFP copies were calculated and normalized to the detected housekeeper gene copies (Polr2a), based on eGFP/Polr2a plasmid-based standard curves. The fold change (FC) of normalized eGFP expression after i.p. over i.v. administration for all tissues is shown (top). For selected tissues (diaphragm, eWAT, liver, and pancreas), the normalized copy ratios are depicted (bottom). n = 5 animals per administration group and n = 4 animals (2 per administration group) for PBS, mean (SD). Group comparison by unpaired Student’s t test (AAV1P4: liver), unpaired Student’s t test with Welch’s correction (AAV-DJ: diaphragm, eWAT, and pancreas), and Mann-Whitney test (AAV-DJ: liver and AAV1P4: diaphragm, eWAT, and pancreas), exact p values. (B and D) Representative images of eGFP IHC are shown for the selected tissues from (A) and (C),…