Biolayer interferometry for adeno-associated virus capsid titer measurement and applications to upstream and downstream process development.

Faster and more accurate analytical methods are needed to support the advancement of recombinant adeno-associated virus (rAAV) production systems. Recently, biolayer interferometry (BLI) has been developed for high-throughput AAV capsid titer measurement by functionalizing the AAVX ligand onto biosensor probes (AAVX-BLI). In this work, an AAVX-BLI method was evaluated using Octet AAVX biosensors across four rAAV serotypes (rAAV2, -5, -8, and -9) and applied in an upstream and downstream processing context. AAVX-BLI measured the capsid titer across a wide concentration range (1 × 1010-1 × 1012 capsids/mL) for different rAAV serotypes and sample backgrounds with reduced measurement variance and error compared to an enzyme-linked immunosorbent assay (ELISA) method. Biosensors were regenerated for repeated use, with lysate samples showing reduced regeneration capacity compared to purified and supernatant samples. The AAVX-BLI method was applied in a transfection optimization study where direct capsid titer measurement of culture supernatants generated a representative response surface for the total vector genome (VG) titer. For rAAV purification, AAVX-BLI was used to measure dynamic binding capacity with POROS CaptureSelect AAVX affinity chromatography, showing resin breakthrough dependence on the operating flow rate. Measurement accuracy, serotype and sample background flexibility, and high sample throughput make AAVX-BLI an attractive alternative to other capsid titer measurement techniques.

Design space determination to optimize DNA complexation and full capsid formation in transient rAAV manufacturing.

Recombinant adeno-associated virus (rAAV) vectors are a promising platform for in vivo gene therapies. However, cost-effective, well-characterized processes necessary to manufacture rAAV therapeutics are challenging to develop without an understanding of how process parameters (PPs) affect rAAV product quality attributes (PQAs). In this work, a central composite orthogonal experimental design was employed to examine the influence of four PPs for transient transfection complex formation (polyethylenimine:DNA [PEI:DNA] ratio, total DNA/cell, cocktail volume, and incubation time) on three rAAV PQAs related to capsid content (vector genome titer, vector genome:capsid particle ratio, and two-dimensional vector genome titer ratio). A regression model was established for each PQA using partial least squares, and a design space (DS) was defined in which Monte Carlo simulations predicted < 1% probability of failure (POF) to meet predetermined PQA specifications. Of the three PQAs, viral genome titer was most strongly correlated with changes in complexation PPs. The DS and acceptable PP ranges were largest when incubation time and cocktail volume were kept at mid-high setpoints, and PEI:DNA ratio and total DNA/cell were at low-mid setpoints. Verification experiments confirmed model predictive capability, and this work establishes a framework for studying other rAAV PPs and their relationship to PQAs.

Comparison of vector elements and process conditions in transient and stable suspension HEK293 platforms using SARS-CoV-2 receptor binding domain as a model protein.

BACKGROUND: Mammalian cell lines are frequently used as protein expression hosts because of their ability to correctly fold and assemble complex proteins, produce them at high titers, and confer post-translational modifications (PTMs) critical to proper function. Increasing demand for proteins with human-like PTMs, particularly viral proteins and vectors, have made human embryonic kidney 293 (HEK293) cells an increasingly popular host. The need to engineer more productive HEK293 platforms and the ongoing nature of the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) pandemic presented an opportunity to study strategies to improve viral protein expression in transient and stable HEK293 platforms. RESULTS: Initial process development was done at 24 deep well plate (DWP) -scale to screen transient processes and stable clonal cell lines for recombinant SARS-CoV-2 receptor binding domain (rRBD) titer. Nine DNA vectors that drove rRBD production under different promoters and optionally contained Epstein-Barr virus (EBV) elements to promote episomal expression were screened for transient rRBD production at 37 °C or 32 °C. Use of the cytomegalovirus (CMV) promoter to drive expression at 32 °C led to the highest transient protein titers, but inclusion of episomal expression elements did not augment titer. In parallel, four clonal cell lines with titers higher than that of the selected stable pool were identified in a batch screen. Flask-scale transient transfection and stable fed-batch processes were then established that produced rRBD up to 100 mg/L and 140 mg/L, respectively. While a bio-layer interferometry (BLI) assay was crucial for efficiently screening DWP batch titers, an enzyme-linked immunosorbent assay (ELISA) was used to compare titers from the flask-scale batches due to varying matrix effects from different cell culture media compositions. CONCLUSION: Comparing yields from the flask-scale batches revealed that stable fed-batch cultures produced up to 2.1x more rRBD than transient processes. The stable cell lines developed in this work are the first reported clonal, HEK293-derived rRBD producers and have titers up to 140 mg/L. As stable production platforms are more economically favorable for long-term protein production at large scales, investigation of strategies to increase the efficiency of high-titer stable cell line generation in Expi293F or other HEK293 hosts is warranted.

Analytical methods to characterize recombinant adeno-associated virus vectors and the benefit of standardization and reference materials.

Recombinant adeno-associated virus (rAAV) is an increasingly important gene therapy vector, but its properties present unique challenges to critical quality attribute (CQA) identification and analytics development. Advances in, and ongoing hurdles to, characterizing rAAV proteins, nucleic acids, and vector potency are discussed in this review. For nucleic acids and vector potency, current analytical techniques for defined CQAs would benefit from further optimization, while for proteins, more complete characterization and mapping of properties to safety and efficacy is needed to finalize CQAs. The benefits of leveraging reference vectors to validate analytics and CQA ranges are also proposed. Once defined, CQA specifications can be used to establish target parameters for and inform the development of next generation rAAV processes.