Genome-wide activation screens to increase adeno-associated virus production

Abstract

We describe a genome-wide screening strategy to identify target genes whose modulation increases the capacity of a cell to produce recombinant adeno-associated viral (AAV) vector. Specifically, a single-guide RNA (sgRNA) library for a CRISPR-based genome-wide transcriptional activation screen was inserted into an AAV vector, and iterative rounds of viral infection and rescue in HEK293 producer cells enabled the enrichment of sgRNAs targeting genes whose upregulation increased AAV production. Numerous gain-of-function targets were identified, including spindle and kinetochore associated complex subunit 2 (SKA2) and inositol 1, 4, 5-trisphosphate receptor interacting protein (ITPRIP). Furthermore, individual or combinatorial modulation of these targets in stable producer cell lines increased vector genomic replication and loading into AAV virions, resulting in up to a 3.8-fold increase in AAV manufacturing capacity. Our study offers an efficient approach to engineer viral vector producer cell lines and enhances our understanding of the roles of SKA2 and ITPRIP in AAV packaging.

Keywords: adeno-associated virus; cell engineering; gene therapy; high-throughput screening; vector manufacturing.

Graphical abstract

Figure 1

Parallel genome-wide screens identified candidate genes that increased HEK293 AAV manufacturing capacity (A) Schematic of the iterative AAV-based selection with the Cas9-based SAM system to identify sgRNAs and corresponding candidate HEK293T gene targets, whose overexpression enhances AAV vector production. (B) Boxplots showing the distribution of SAM sgRNA frequencies before and after selection, showing that sgRNAs were enriched over eight rounds. Multiple sgRNAs were enriched after the selection rounds. Line: median; boxes, 25th to 75th percentiles; whiskers: 2.5th to 97.5th percentiles. (C) Compiled fold increase of the top 50 hits from each screen and their fold increase relative to the frequency in the pre-selected library. (D) mRNA induction levels of SKA2 and ITPRIP after triple plasmid transfection at indicated time points. Error bars, mean ± SD (E) Stable expression of SKA2 and ITPRIP sgRNA hits led to increases in AAV2 titer relative to WT HEK293Ts with no guide. HEK293T cells stably expression the SAM system were generated by lentivirus with doxycycline-inducible promoter and each sgRNA. Cell lines were treated with doxycycline for 24 h, followed by triple transfection for another 72 h. Line: median; boxes, 25th to 75th percentiles; whiskers: min/max.

Figure 2

Cells expressing SKA2, ITPRIP, or both SKA2 and ITPRIP increased viral and infectious titer relative to HEK293 cells Cells were seeded, and gene expression was induced 24 h before triple transfection with AAV packaging plasmids. Virus was harvested and titered by qPCR. (A) The fold increase in viral titer of AAV2 was determined in cell lines that expressed either SKA2 or ITPRIP in HEK293Ts (n = 21). (B) Increases in AAV2 infectious titer were also found in cells expressing either SKA2 or ITPRIP in HEK293Ts (n = 4). (C) Increases in AAV6 titer were found in both cells expressing either SKA2 or ITPRIP in HEK293Ts (n = 4). (D) Increase in AAV6 infectious titer was also found in both cells expressing either SKA2 or ITPRIP in HEK293Ts (n ≥ 3). (E) Fold increase of AAV2 packaged in a cell line that expresses both genes from the activation screen were able to increase titer by 3.8-fold compared to WT (n = 4). (F) Fold increase of AAV6 packaged in a cell line expressing both genes increased titer by 3.5-fold (n ≥ 4); all data shown ∗p ≤ 0.05, ∗∗p < 0.01, ∗∗∗p ≤ 0.005, ∗∗∗∗p ≤ 0.0005 compared to WT. Line: median; boxes, 25th to 75th percentiles; whiskers: min/max.

Figure 3

Overexpression of SKA2 and ITPRIP regulate AAV genome replication and cell cycle (A) Fold increase of transgene replication at 2 days after triple transfection in the stable HEK293T cell lines (n ≥ 5). (B and C) Cell-cycle analysis by flow cytometry with DAPI staining (B) or EdU incorporation intensity (C) in the stable cell lines at day 2. (D) Fold increase of AAV2 packaged with thymidine treatment for 24 or 48 h. HEK293T cells were transfected for AAV packaging and treated with thymidine at the concentration of 4 mM (n = 4). (E) Comparison of AAV2 packaging with thymidine treatment from the HEK293T cells cultured at 100% confluency. Cells were transfected with AAV2 packaging genes, followed by thymidine treatment at 4 and 8 mM (n = 4). (F) AAV2 packaging changes in the stable cell lines cultured at 100% confluency (n = 4). For all data shown: ∗p ≤ 0.05, ∗∗p < 0.01, ∗∗∗p ≤ 0.005, ∗∗∗∗p ≤ 0.0005 compared to WT. Line: median; boxes, 25th to 75th percentiles; whiskers: min/max.

Figure 4

Overexpression of SKA2 and ITPRIP increase AAV full/empty capsid ratio (A) HEK293T cell lysate was analyzed using western blot showing 1:1:10 ratio of VP1, VP2, and VP3 in different cell lines expressing SKA2, ITPRIP, or both SKA2 and ITPRIP, as well as WT HEK293Ts loaded with equivalent vgs (1 × 10⁹). (B) Western blot under the same conditions but instead loading equal protein concentrations (20 μg). (C) AAV2 capsid fold increase relative to WT HEK293Ts in each of the cell lines as quantified by ELISA (n ≥ 3). One-way ANOVA analysis indicated no statistical differences. (D) Ratio of AAV2 viral genomes to capsids in WT cell lines or cell lines expressing SKA2, ITPRIP, or both. One-way ANOVA analysis showed a significant difference between groups expressing SKA2 and ITPRIP compared to the WT (n ≥ 3, p ≤ 0.05). Line: median; boxes, 25th to 75th percentiles; whiskers: min/max.