Purification of Adeno-Associated Virus (AAV) Serotype 2 from Spodoptera frugiperda (Sf9) Lysate by Chromatographic Nonwoven Membranes

Abstract

The success of adeno-associated virus (AAV)-based therapeutics in gene therapy poses the need for rapid and efficient processes that can support the growing clinical demand. Nonwoven membranes represent an ideal tool for the future of virus purification: owing to their small fiber diameters and high porosity, they can operate at high flowrates while allowing full access to target viral particles without diffusional limitations. This study describes the development of nonwoven ion-exchange membrane adsorbents for the purification of AAV2 from an Sf9 cell lysate. A strong anion-exchange (AEX) membrane was developed by UV grafting glycidyl methacrylate on a polybutylene terephthalate nonwoven followed by functionalization with triethylamine (TEA), resulting in a quaternary amine ligand (AEX-TEA membrane). When operated in bind-and-elute mode at a pH higher than the pI of the capsids, this membrane exhibited a high AAV2 binding capacity (9.6 × 1013 vp·mL−1) at the residence time of 1 min, and outperformed commercial cast membranes by isolating AAV2 from an Sf9 lysate with high productivity (2.4 × 1013 capsids·mL−1·min−1) and logarithmic reduction value of host cell proteins (HCP LRV ~1.8). An iminodiacetic acid cation-exchange nonwoven (CEX-IDA membrane) was also prepared and utilized at a pH lower than the pI of capsids to purify AAV2 in a bind-and-elute mode, affording high capsid recovery and impurity removal by eluting with a salt gradient. To further increase purity, the CEX-IDA and AEX-TEA membranes were utilized in series to purify the AAV2 from the Sf9 cell lysate. This membrane-based chromatography process also achieved excellent DNA clearance and a recovery of infectivity higher that that reported using ion-exchange resin chromatography.

Rational design and experimental evaluation of peptide ligands for the purification of adeno-associated viruses via affinity chromatography

“…[59,60] Combining the cost of synthesis with the average values of number of residues (≈15-17) and molecular weight (≈1.5-1.7 kg mol −1 ) of the peptide ligands, their density on the resin surface (≈0.03 mol per liter), and the cost of the base resin (≈$2500 per liter) indicates that direct material cost of the peptide-functionalized adsorbent ranges between $7900 and $9500 per liter, when produced at the ≈100 L scale (note: direct labor and manufacturing overhead are not factored). These considerations, combined with the purification performance of peptide-functionalized adsorbents presented by our team and by several others in the literature, [9,11,[55][56][57][58][61][62][63][64][65] show the promise of this technology to transform the biomanufacturing of modern medicines and reduce their cost (note: the latter is of particular concern, given the price tag of gene therapies well above US$1M per patient). Under the light of these considerations,…”

“…The HEK293 cell lysate (AAV2 titer ≈1.9 × 10 12 vp mL −1 ; HCP titer ≈0.3 mg mL −1 ) and the Sf9 cell lysate (AAV2 titer ≈1.56 × 10 12 vp mL; HCP titer ≈1.1 mg mL −1 ) were prepared via triple transfection and baculovirus infection, respectively, following the protocols described in the literature. [11,54,[55][56][57][58] Similarly, a residence time (RT) of 3 min for both and washing steps was adopted based on industrial operating conditions. The chromatograms of AAV2 purification are presented in Figure S3, while the analysis of the collected fractions via size exclusion (SEC) and steric exclusion chromatography (SXC) are reported in Figures S4 and S5.…”

Rational design and experimental evaluation of peptide ligands for the purification of adeno‐associated viruses via affinity chromatography

“…[ 59,60 ] Combining the cost of synthesis with the average values of number of residues (≈15–17) and molecular weight (≈1.5–1.7 kg mol −1 ) of the peptide ligands, their density on the resin surface (≈0.03 mol per liter), and the cost of the base resin (≈$2500 per liter) indicates that direct material cost of the peptide‐functionalized adsorbent ranges between $7900 and $9500 per liter, when produced at the ≈100 L scale ( note : direct labor and manufacturing overhead are not factored). These considerations, combined with the purification performance of peptide‐functionalized adsorbents presented by our team and by several others in the literature, [ 9,11,55–58,61–65 ] show the promise of this technology to transform the biomanufacturing of modern medicines and reduce their cost ( note : the latter is of particular concern, given the price tag of gene therapies well above US$1M per patient). Under the light of these considerations, our team plans to demonstrate further the technology introduced in this study by purifying AAVs of different serotypes from a variety of HEK293 and Sf9 fluids.…”

“…The HEK293 cell lysate (AAV2 titer ≈1.9 × 10 12 vp mL −1 ; HCP titer ≈0.3 mg mL −1 ) and the Sf9 cell lysate (AAV2 titer ≈1.56 × 10 12 vp mL; HCP titer ≈1.1 mg mL −1 ) were prepared via triple transfection and baculovirus infection, respectively, following the protocols described in the literature. [11,54,[55][56][57][58] Similarly, a residence time (RT) of 3 min for both binding and washing steps was adopted based on industrial operating conditions. The chromatograms of AAV2 purification are presented in Figure S3, while the analysis of the collected fractions via size exclusion (SEC) and steric exclusion chromatography (SXC) are reported in Figures S4 and S5.…”

Advances in high‐throughput, high‐capacity nonwoven membranes for chromatography in downstream processing: A review

“…Several ligands have been successfully immobilized on the PBT base nonwoven matrix: diethylamine (DEA) weak anion and iminodiacetic acid (IDA) weak cation exchangers; triethylamine (TEA) strong anion and sulfonic acid (SO 3 ) strong cation exchangers; as well as salt tolerant ligands (under patent application). Salt tolerant ligands have been shown to exhibit high binding capacities at short residence times as listed in Table 1 (Fan et al, 2021(Fan et al, , 2022(Fan et al, , 2023Lemma et al, 2021) and used in laboratory scale experiments to purify real cell culture fluids.…”