- Key takeaways
- The production platform landscape: HEK293 triple transfection and its alternatives
- Why does triple transfection create empty capsids, and what can be done about it?
- Optimizing HEK293 cell culture for AAV production
- How does suspension cell culture change the scale-up equation for AAV?
- Downstream purification: the yield and purity challenge
- When do stable producer cell lines become the answer?
Viral vector manufacturing is the infrastructure challenge that determines whether the gene therapy industry can deliver on its clinical promise. Over 500 viral vector-based programs are in clinical trials in the US alone, but the production systems capable of supplying those programs at clinical and commercial scale are still maturing. The bottlenecks are not conceptual: they are engineering and process development challenges concentrated in upstream transfection efficiency, plasmid quality, and downstream empty capsid removal.
Key takeaways |
|
|
|
|
|
For the industry-level context of the viral vector capacity crunch and the transition from small-scale artisanal production to industrial manufacturing, see the full analysis of the viral vector bottleneck. For the strategic overview of how viral vector manufacturing fits within the cell, gene, and mRNA therapy manufacturing landscape, see Scaling the unscalable: manufacturing cell, gene, and mRNA therapies.
The production platform landscape: HEK293 triple transfection and its alternatives
Recombinant AAV production by triple transfection of human embryonic kidney 293 (HEK293) cells is the dominant manufacturing platform for gene therapy programs. As confirmed in a 2025 study developing HEK293 suspension media and transfection workflow for batch rAAV manufacturing, the HEK293 transient transfection platform is adaptable to the manufacture of different therapies by changing plasmid DNA sequences encoding the capsid and therapeutic genome, making it the default starting point for most new gene therapy programs. Three plasmids are delivered simultaneously: one encoding the AAV rep and cap genes, one encoding adenovirus helper functions, and one containing the therapeutic transgene flanked by AAV inverted terminal repeats.
The scalability of transient transfection is constrained by the plasmid DNA required at each manufacturing run and by the inherent variability of a process where transfection efficiency differs from cell to cell and run to run. At clinical scale, these limitations are manageable. At commercial scale, the quantities of GMP-grade plasmid DNA required, and the regulatory challenge of maintaining batch-to-batch consistency from a transiently driven process, become significant operational and quality challenges.
Alternative production platforms exist for specific applications. The baculovirus expression vector system, which infects insect Sf9 cells to drive AAV production, offers high scalability and has been used for commercial manufacturing of approved gene therapy products but requires specific regulatory characterization for the insect cell-derived impurity profile and glycosylation differences from HEK293-derived vector. Helper virus systems using adenoviral or herpes simplex virus vectors to provide helper functions to stable AAV packaging cell lines are being developed as routes to eliminate transient transfection, but remain primarily in development-stage use. The selection between platforms involves trade-offs across every dimension relevant to commercial manufacturing.
Platform | Host cell | Serotype flexibility | Scalability | Primary limitation |
HEK293 triple transfection | Mammalian; HEK293 adherent or suspension | High; capsid plasmid changed for each serotype | Clinical to commercial; suspension bioreactors to 2,000 L reported | Plasmid DNA supply at scale; batch-to-batch transfection variability; high empty capsid ratio requiring downstream resolution |
Baculovirus / Sf9 insect cells | Insect (Sf9); animal-origin-free conditions | Good; serotype changed by swapping baculovirus recombinants | High; established at commercial scale for approved products | Insect cell-specific impurity profile; glycosylation differences from human cells; regulatory characterization overhead per product |
Stable producer cell line (HEK293-based) | Mammalian; HEK293 with stable genomic integration of packaging elements | Lower; separate cell line engineering required for each serotype | High once established; eliminates plasmid transfection from commercial runs | Long development timeline per serotype; regulatory demonstration of genetic stability across production; limited commercial examples to date |
Helper virus systems (HSV or AdV) | Mammalian; typically HEK293 or BHK with stable AAV elements | Moderate; rep/cap provided by helper; transgene construct changed per product | High throughput potential; used in commercial-scale production | Helper virus clearance validation required; downstream complexity from helper virus co-production; dedicated facility for containment |
Why does triple transfection create empty capsids, and what can be done about it?
Empty AAV capsids, which are structurally intact viral shells containing no therapeutic DNA, are generated in every triple transfection run as a byproduct of the production mechanism. A 2025 study developing a minicircle-based approach to reduce empty capsid formation identified the molecular mechanism: empty capsids are generated primarily due to unbalanced uptake of the three plasmids required for triple transfection at the individual cell level. Cells that receive the rep/cap plasmid and the helper plasmid but not the transgene plasmid will produce empty capsids because the capsid proteins are expressed and assembled in the absence of the transgene to package. The ratio of plasmids, their delivery efficiency, and their relative intracellular concentrations all influence how many cells fall into this empty-capsid-producing configuration.
The clinical consequences of high empty capsid ratios are significant. Empty capsids are not therapeutically active but are antigenically identical to full capsids and contribute to the patient's immune response to the vector. A product with a high empty-to-full ratio requires a larger total dose of capsid particles to deliver the effective dose of therapeutic vector genomes, increasing the immunogenic burden on the patient and potentially triggering higher neutralizing antibody titers that could limit re-dosing. Regulatory agencies have established expectations for empty:full capsid characterization and control in gene therapy product quality dossiers.
Process-level strategies for reducing empty capsid formation include optimizing the molar ratios of the three plasmids delivered per cell, using high-quality, minimally contaminated plasmid DNA preparations that reduce competition between plasmid species during transfection, and timing the harvest to capture the window of maximum full-capsid production before cell viability declines. Downstream strategies include analytical ultracentrifugation for high-resolution separation of full and empty capsids, which is technically reliable but difficult to scale, and density gradient techniques using cesium chloride or iodixanol that have been adapted for pilot-scale purification.
Optimizing HEK293 cell culture for AAV production
HEK293 cell culture optimization for rAAV production must address several interdependent variables that interact across the transfection event and the post-transfection production window. Research from 2025 on developing and optimizing HEK293 suspension media and transfection workflows for batch rAAV manufacturing demonstrated that media formulation, cell density at time of transfection, plasmid molar ratios among the three vectors, and transfection reagent selection cannot be optimized independently; each variable affects the output of the others, requiring design-of-experiment approaches across the combined parameter space rather than sequential one-factor-at-a-time optimization.
Media selection for suspension HEK293 AAV production must support two distinct cellular states: the rapid proliferative growth phase before transfection, where cell density must reach the target for transfection without clumping or viability loss, and the post-transfection production phase, where the cells must survive long enough to produce and release vector while the transfection machinery drives a metabolically demanding production process. Media formulations optimized for HEK293 growth may not provide optimal conditions for transfection efficiency or post-transfection viability, and commercially available suspension media must be evaluated empirically against both performance requirements.
Cell viability management after transfection is a specific challenge for AAV production that does not arise in conventional biologic manufacturing. Research from Johns Hopkins on cytotoxicity in HEK293 cells during rAAV production confirmed that rAAV production is cytotoxic to HEK293 cells, with apoptosis activation progressively reducing cell viability over the production window. The cytotoxicity is linked to the expression of AAV rep protein, which interacts with the host cell cycle machinery. Strategies that reduce apoptotic cell death during the production window, including caspase inhibition and media optimization for post-transfection viability, have been shown to improve the full-to-empty capsid ratio by maintaining the cellular machinery needed for therapeutic genome packaging.
How does suspension cell culture change the scale-up equation for AAV?
Adherent HEK293 cell culture on flat multilayer vessels was the original manufacturing format for laboratory-scale AAV production and remains in use at small clinical scales where the manufacturing complexity of suspension adaptation is not justified by volume requirements. The fundamental limitation of adherent culture is surface-area-dependent scaling: adding production capacity requires adding physical surface area, which at commercial scale means hundreds or thousands of individual cell culture vessels, with all the associated labor, contamination risk, and lot-to-lot variability of manual manipulation.
Suspension-adapted HEK293 cells grown in serum-free media in bioreactors provide the volumetric scalability that adherent culture cannot. The transition from adherent to suspension requires adaptation of an adherent cell line to grow without surface attachment, selection of a compatible serum-free suspension media, and confirmation that the adapted cells retain transfection efficiency and post-transfection AAV production performance equivalent to the adherent parent. This adaptation process takes time but is well-established, and suspension HEK293 lines adapted for AAV production are available through academic and commercial sources.
In controlled bioreactors, suspension HEK293 cells can be transfected at densities of 1 to 2 million cells per milliliter and harvested 48 to 72 hours post-transfection with volumetric yields reported above 10^13 to 10^14 vector genomes per liter, depending on the AAV serotype and the specific media and transfection conditions. These yields are achievable in bioreactors from 3 liters in process development through 200 liters and above in GMP production suites, providing a scalable production format that bridges clinical and early commercial supply.
Downstream purification: the yield and purity challenge
AAV downstream purification is more complex than the purification of monoclonal antibodies primarily because the target product, the full AAV capsid, is nearly physically identical to the primary impurity, the empty AAV capsid. Both are spherical protein shells approximately 25 nanometers in diameter with indistinguishable surface chemistry and electrostatic properties under most chromatographic conditions. This near-identity removes the selectivity that conventional ion exchange and affinity chromatography rely on to separate product from impurities.
Affinity chromatography using immobilized antibodies against specific AAV serotype epitopes, or newly developed capsid-specific small molecule and nanobody ligands, provides the capture selectivity and yield improvement that Protein A provides for monoclonal antibodies. However, unlike Protein A, which works broadly across IgG-class molecules, AAV affinity capture is serotype-specific: a resin developed for AAV2 does not capture AAV9. Each program therefore requires its own affinity capture solution or must rely on ion exchange chromatography at lower selectivity and yield.
The separation of full from empty capsids typically requires either analytical ultracentrifugation, which resolves the small density difference between full capsids containing double-stranded DNA and empty capsids containing only protein, or ion exchange chromatography under carefully optimized gradient conditions that can exploit subtle charge distribution differences between full and empty populations. Neither method achieves the throughput and yield of a standard chromatographic capture step, and both require specialized process development work for each new vector product.
The regulatory implications of empty capsid content in gene therapy products, including the analytical methods required to characterize the full-to-empty ratio and the specifications that have been applied in approved product submissions, are discussed in the related DDN coverage of why gene and cell therapies are stalling at the FDA.
When do stable producer cell lines become the answer?
Stable producer cell lines for AAV address the two most significant process limitations of transient triple transfection: the requirement for large quantities of GMP-grade plasmid DNA at every production run, and the batch-to-batch variability introduced by a process whose output depends on the efficiency of plasmid delivery to each individual cell in the culture. In a stable producer line, the genetic elements required for AAV production are permanently integrated into the host cell genome and are expressed in response to a defined inducible signal rather than a transiently delivered plasmid.
The development timeline for a stable producer cell line is substantially longer than for a transient transfection process: cell line construction, selection, characterization, and stability demonstration typically require 12 to 24 months before GMP banking is possible, compared to three to six months for a transient transfection process development program. This timeline means that most programs in early clinical development use transient transfection for speed, with stable producer cell line development pursued in parallel for commercial manufacturing readiness.
The broader strategic question of when to invest in internal manufacturing infrastructure versus outsourcing vector production to a CDMO, which interacts directly with the build-versus-buy decision for stable producer line development, is addressed from the industry perspective in the related article on the viral vector bottleneck and the industrialization of gene therapy manufacturing. A 2025 study on intensification of rAAV production from HEK293 transient transfection also demonstrates that significant productivity gains are still achievable within the transient platform through process intensification, suggesting that the decision to move to stable producer lines is not as urgent for programs that can access the full optimization potential of their transient process.
This article was produced under Drug Discovery News' AI Editorial Guidelines.
