Techniques and Advancements in Recombinant Adeno-Associated Virus Purification for Gene Therapies

Adeno-associated viruses (AAVs) have emerged in recent years as a crucial tool in clinical gene therapies, offering low pathogenicity and long-lived gene expression in target cells. AAVs are simple non-enveloped viruses with a single-stranded DNA (ssDNA) genome. Recombinant adeno-associated viruses (rAAVs) are structurally similar to wild-type AAVs; however, the Rep and Cap genes are removed and replaced with a transgene – the genetic cargo that can repair or replace a faulty gene.

rAAVs are increasingly important in viral-based therapies,^1,2 and at present, there are seven rAAV therapies approved by the European Medicines Agency and the US Food and Drug Administration for the treatment of genetic diseases ranging from congenital blindness to muscular dystrophy and hemophilia. Over 200 clinical trials have been undertaken to bring new rAAV therapies to the clinic; however, manufacturing innovation has lagged behind.³

Dr. Guangping Gao, professor of microbiology and physiological systems and Penelope Booth Rockwell Professor in Biomedical Research at UMass Chan Medical School, explained some of the issues we currently face: “As a field, we are experiencing several obstacles. First, we need more efficient and target-specific capsids, which are the essence of any therapy using AAVs. Secondly, manufacturing is a major hurdle, which is part of the reason why rAAV therapies are so expensive.”

This article explores advances in the purification of rAAVs to meet current clinical and research needs.

How are rAAVs made?

AAVs and rAAVs cannot replicate independently and require co-infection with a helper virus – typically adenoviruses (AdVs) or herpes simplex virus (HSV) – to amplify in a host. rAAVs offer an even safer profile, as they lack viral genes and cannot replicate even with a helper virus present. However, their inability to self-amplify restricts sustained rAAV production within a batch. 

rAAVs are typically produced in suitable host cells through the coordinated expression of Rep and Cap components, together with helper genes or a helper virus, and the desired transgene. There are several options for production, as noted by Dr. Jessica Whelan, lecturer and assistant professor at University College Dublin. “There is not a clearly preferred production system. Transient transfection, producer cell lines and baculovirus-based insect cell production systems are used – each with their advantages and disadvantages. Depending on the production system, the characteristics of the material to be purified vary, which impacts downstream process design.”

Traditional rAAV production has relied on transient transfection of Human Embryonic Kidney 293 cells with plasmid DNA encoding the transgene, AAV Rep/Cap genes and AdV helper genes.⁴ While this approach offers tunability, it remains inefficient and scaled-up suspension cultures can often produce ≥80% empty capsids, necessitating extensive downstream purification to separate empty capsids from full ones.

Optimize AAV Filtration for Faster, Reliable Results

Discover a streamlined approach to AAV serotype filtration.
View Application Note

Advertisement

An alternative is the Baculovirus expression vector system using Spodoptera frugiperda (Sf9) insect cells,^5,6, which was employed to manufacture Glybera®, the first approved rAAV therapy (now discontinued). This scalable platform is widely used in protein production but tends to generate higher levels of defective or non-transducing particles. Another strategy uses recombinant herpes simplex virus (rHSV), where one rHSV delivers AAV Rep and Cap genes to producer cells and another delivers the transgene.⁷ HSV’s natural role as a helper virus enables high-quality rAAV production at high titers. However, scalability is limited by the complexity of generating starting materials, and rigorous purification is required to eliminate residual rHSV, which is neurotropic and neurotoxic.

Stable producer cell lines offer further scalability and reproducibility. In this system, cells stably expressing Rep and Cap are stimulated with recombinant AdV to drive rAAV production.⁸ Although effective, this approach again requires stringent purification to remove residual AdV and oncogenic DNA derived from transformed cell lines.

Across all viral-based production systems, scalability is offset by the burden of downstream purification. As Gao emphasized, “When you use viral systems for production, the key challenge is removing the helper virus – at the virus, DNA and protein level. This requires more stringent purification, as mammalian helper viruses are much more immunogenic than AAV itself.”

Purification of rAAVs

The choice of upstream methods for rAAV production has a direct and significant impact on downstream processing requirements. Both Gao and Whelan emphasized this connection. “The upstream and downstream processes are highly interconnected,” explained Gao. “What you have in your upstream processes will determine the downstream output.” Whelan echoed this, noting, “depending on the production system, the characteristics of the material to be purified vary, which impacts downstream process design.”

Each production method introduces process-related contaminants and product-related impurities, such as empty or partially full capsids. These by-products must be removed, but each purification step can lead to product losses and sequential processing may significantly reduce the final yield.

“Purification remains challenging for a multiplicity of reasons,” Whelan explained. “These include low AAV concentrations, the difficulty of balancing yield and purity given the similarity between full capsids and empty or partially filled capsids, the lack of a standardized downstream platform comparable to that for monoclonal antibodies and the need for improved analytical tools and methods.” The main steps of purification are broadly broken down in the following sections.

Cell harvest and lysis

rAAVs must first be harvested from producer cells.⁹ For small research volumes (<10 mL), freeze-thaw is suitable, but larger scales require chemical or physical cell disruption. Microfluidic disruption is increasingly recognized as scalable and current good manufacturing practice-compliant, using high-pressure shear forces to lyse cells while simultaneously shearing DNA, reducing viscosity and aiding clarification.

Clarification

After lysis, clarification removes cellular debris. Initial low-speed centrifugation removes large particles, while enhanced depth filtration offers scalable, gentle separation of rAAVs from crude lysates. Tangential flow filtration is now widely used as a filtration method, concentrating rAAV lysates ~10-fold and removing soluble impurities without clogging.¹⁰

Capsid capture

Clarified lysates still contain impurities such as proteins, nucleic acids and empty capsids. Early methods like iodixanol or cesium chloride (CsCl) gradient ultracentrifugation exploited physical differences but were labor-intensive, low-throughput and unsuitable for scale, with CsCl particularly damaging to capsids. Affinity chromatography has largely replaced these methods, capturing rAAV via serotype-specific ligands or newer cross-serotype resins (e.g., AVB Sepharose, POROS™ CaptureSelect™) using nanobody-derived ligands.¹¹ These offer high yield and purity, but at high cost.

Emerging ligands include AAVX, which binds serotypes 1–9, and AAV receptor, a biologically relevant receptor candidate. Computationally designed peptide ligands also show promise, offering gentle elution and lower production costs, potentially enabling broad serotype-agnostic resin development. “Relative to commercial resins for antibodies, resins for AAVs have lower binding capacities, short lifespans and are often serotype-specific,” noted Whelan. “However, significant progress is being made.”

The What, How and Why of qPCR Analysis

This infographic highlights the principles, strengths and limitations of qPCR while showcasing its applications across fields including oncology, pathogen detection, gene expression, biomarker discovery and beyond.
View Infographic

Advertisement

Aqueous two-phase systems (or aqueous biphasic systems) partition molecules based on physicochemical properties, offering gentle separation without harsh solvents. While design complexity limits scalability, well-optimized systems show strong potential.¹² Steric exclusion chromatography (SXC) uses PEG-based molecular crowding and hydrophilic membranes to bind AAVs, providing up to 10-fold greater binding capacity than affinity resins while maintaining physiological conditions.

Polishing – ion exchange chromatography

The final stage of downstream purification involves polishing the final product to remove product-related impurities and any residual process-related ones. “Your downstream processing is the key,” emphasized Gao. “It doesn't matter which production method is used; one of the major issues is removing empty particles. This is a major challenge because you need product-specific downstream processing, particularly polishing, to remove empty particles.”

Ion (cation/anion) exchange chromatography is widely used, often sequentially, to enhance yields. Ultracentrifugation remains an option,¹³ but it is low-yield and unscalable.

“Sedimentation processes may or may not be perfect for clinical vector manufacturing or scaling up, and there is a push for scaling up of column chromatography,” said Gao. “Affinity columns are good for capture of serotype-specific particles but cannot differentiate between empty and full capsids. If you wish to remove gradient-based sedimentation, you need a polishing step that captures minor differences based on surface charge, ionic strength or anything that can distinguish empty from full or partial from full.”

Recent advances include monolithic columns,¹⁴ which speed processing by improving flow through a continuous porous phase, and membrane-based chromatography,¹⁵ where modified anion exchange membranes deliver high yields and preserve infectivity.

Outlook

While large-scale rAAV production is still developing, both experts expressed optimism. Gao highlighted recent progress: “From ~10% full particle recovery in the first clinical trials in 2007/2008, we can now reach 60% owing to chromatography improvements.” He added, “different technologies have emerged in recent years, but they remain small-scale. I’m watching their development and migration to larger scales, as well as improvements in reproducibility.”

Whelan emphasized lessons from therapeutic proteins: “Leveraging knowledge and experience from the therapeutic protein space, chromatography and membrane-based filtration are the current focus for scalable, commercial manufacturing.” She remained positive about the field’s outlook: “While there is a distance to go, the growing demand for AAVs is driving intense innovation, with regular advances in resins, analytical techniques and process intensification. Despite AAV-specific challenges, high innovation levels and lessons from other modalities point to a promising future.”