A novel dual-plasmid platform provides scalable transfection yielding improved productivity and packaging across multiple AAV serotypes and genomes

Abstract

Transient transfection of mammalian cells using plasmid DNA is a standard method to produce adeno-associated virus (AAV) vectors allowing for flexible and scalable manufacture. Typically, three plasmids are used to encode the necessary components to facilitate vector production; however, a dual-plasmid system, termed pDG, was introduced over 2 decades ago demonstrating two components could be combined resulting in comparable productivity to triple transfection. We have developed a novel dual-plasmid system, pOXB, with an alternative arrangement of sequences that results in significantly increased AAV vector productivity and percentage of full capsids packaged in comparison to the pDG dual design and triple transfection. Here, we demonstrate the reproducibility of these findings across seven recombinant AAV genomes and multiple capsid serotypes as well as the scalability of the pOXB dual-plasmid transfection at 50-L bioreactor scale. Purified drug substance showed a consistent product quality profile in line with triple-transfected vectors, except for a substantial improvement in intact genomes packaged using the pOXB dual- transfection system. Furthermore, pOXB dual- and triple-transfection-based vectors performed consistently in vivo. The pOXB dual plasmid represents an innovation in AAV manufacturing resulting in significant process gains while maintaining the flexibility of a transient transfection platform.

Keywords: AAV; AAV manufacturing; VG productivity; dual-plasmid system; transient transfection; vector quality.

Graphical abstract

Figure 1

Schematics of dual- and triple-plasmid design configurations

Figure 2

Productivity assessment of three possible dual-plasmid configurations and triple transfection Shake flasks were transfected with pOXB, pLV, or pDG-like dual or triple plasmids using GOI-ITR genome G sequence at a 1:1 or 1:1:1 M ratio. Crude lysate samples were quantified for (A) VG productivity, (B) capsid productivity, and (C) the percentage of calculated full vectors. Western blot analysis was performed on crude lysates to detect (D) Rep and (E) Cap expression. Quantification of (F) Rep and (G) Cap expression from the western blot is shown. All conditions were completed in duplicate, and the data are displayed as mean ± SD. Statistical significance was calculated using a one-way ANOVA, ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001.

Figure 3

Productivity assessment of pOXB, pLV, and pDG-like dual plasmids compared with triple transfection Shake flasks were transfected with pOXB, pLV, or pDG-like dual plasmids at molar ratios of 2:1, 1:1, and 1:3 with the first plasmid containing the GOI-ITR sequence. Triple transfection is shown at a 1:2:1 and 1:1:1 M ratio GOI-ITR:RepCap:helper. Crude lysate samples were quantified for (A) VG production, (B) capsid production, and (C) calculated full vectors. All conditions were completed in duplicate, and the data are displayed as mean ± SD.

Figure 4

Productivity assessment of seven AAV genomes comparing pOXB dual to triple transfection Transfections were performed with pOXB dual or triple plasmids for each GOI-ITR genome sequence in 2-L bioreactors. Crude lysate samples were quantified for (A) VG production, (B) capsid production, and (C) calculated full vectors. All vectors were packaged with AAVHSC15 except for genome F, which was packaged with AAVHSC17. All conditions were completed at a minimum of n = 2, and the data are displayed as mean ± SD.

Figure 5

Productivity assessment of AAV2 with pOXB dual and triple transfection Shake flasks were transfected with pOXB dual or triple plasmids at a molar ratio of 1:1 or 1:1:1 using the GOI-ITR genome A sequence and AAV2 capsid sequence. Crude lysate samples were quantified for (A) VG production, (B) capsid production, and (C) calculated full vectors. All conditions were completed in duplicate, and the data are displayed as the mean ± SD. Statistical significance was determined using a Student's t test, ∗∗p < 0.01.

Figure 6

Increase in packaged intact vector genomes and reduction in partial packaged vectors and empty capsids observed for pOXB dual compared with triple transfection AUC was used to analyze purified vectors produced with pOXB dual- and triple-plasmid transfection to determine the proportion of capsids containing full-length intact genomes, partial packaged genomes, or empty capsids. All data were generated at the 2-L bioreactor scale except for genome D, which was produced in 50-L bioreactors. Data are displayed as the mean ± SD of (A) intact packaged genomes, (B) partial packaged genomes, and (C) empty capsids for pOXB dual and triple produced vectors for each genome. pOXB dual produced vectors were analyzed in the following replicates: genome B (n = 3), genome C (n = 3), genome D (n = 2), and genome F (n = 1). Triple produced vectors are all singlicate runs except genome D (n = 3).

Figure 7

Productivity and product quality assessment of pOXB dual and triple plasmid produced drug substance Genome D was produced at 50-L bioreactor scale using pOXB dual- (n = 2) and triple-plasmid (n = 3) transfection. Crude lysate samples were analyzed for (A) VG productivity, (B) capsid productivity, and (C) calculated full vectors, while purified vectors from these lots were analyzed for (D) purity, (E) aggregation, (F) residual host cell protein, (G) residual packaged host cell DNA, (H) residual packaged RepCap plasmid DNA, (I) residual packaged E1A DNA, and (J) residual packaged helper plasmid DNA. Data are displayed as mean ± SD. The dashed lines indicate the limit of detection for the assays where samples were determined to be below the limit of quantification (BLoQ). Statistical significance was determined using a Student's t test, ∗p < 0.05, ∗∗∗p < 0.001; ns, not significant.

Figure 8

Comparison of in vivo bioactivity of pOXB dual and triple transfection vectors BTBR Pah^enu2 mice were systemically injected with (A) a low dose of 1 x 10¹² VG/kg or (B) a high dose of 1 x 10¹⁴ VG/kg of vector produced by pOXB dual- or triple-plasmid transfection. A vehicle-only injection group was included as a negative control. Mice were monitored over 6 weeks including weekly serum Phe analysis. At endpoint, liver samples were collected to analyze (C) VGs and (D) mRNA expression. Statistical significance was determined using a Student's t test; ns, not significant.