Cell Culture Variables for Gene Therapy Vectors

Published on: 
BioPharm International, BioPharm International-02-01-2020, Volume 33, Issue 2
Pages: 20–21

The production of viral vectors for use in gene therapy benefits from being able to use similar cell-culture processes as mAbs, but it faces limitations under current cell-culture technologies.

Much attention is focused on the use of viral vectors as the delivery vehicle for gene therapy because of their ability to target specific cells, attach to the cell, and inject genetic material into the cell, which is a critical mechanism for a gene therapy. The structure of a viral vector, empty of its own DNA, makes for practical “housing” for the DNA, RNA, or other genetic material that is the therapy. Other forms of packaging for genetic material may work as a “housing” structure but lack the mechanism to deliver the therapy once in vivo. Several virus types have been investigated for use as gene therapy delivery vectors, including retroviruses, adenoviruses, adeno-associated viruses (AAV), poxviruses, rhabdoviruses, parvoruses, and alpha viruses (1).

Similarities to mAbs

Viral vectors are cultivated in a similar manner to monoclonal antibodies (mAbs) in that the vectors can be mass produced in cell culture. This offers the advantages of having technology, processes, equipment, know-how, and infrastructure that are already established and that have been standardized. “Both can share many similarities when based upon suspension cell cultures (Chinese hamster ovary [CHO] and adapted human embryonic kidney-293 [HEK-293] cells) where similar concerns are in play: cell growth, pH control, dissolved oxygen control, sufficient mixing, shear avoidance, etc.,” says Tony W. Hsiao, staff engineer, R&D, at Thermo Fisher Scientific. “They may differ, however, in that some viral vector processes have a need for adherent or even non-mammalian (Spodoptera frugiperda 9 [SF9]) cell cultures that are less common in standard mAb production.” 

“There are a number of similarities ranging from culture vessels (i.e., shake flasks, WAVE bioreactor, stirred tank reactors, etc.) and media formulations, but there are several important distinctions,” say experts at Novartis. “For example, the most commonly used mammalian cell lines for viral vector manufacturing are HEK-293 derived, whereas the overwhelming majority of antibody and recombinant protein production processes utilize stable CHO-derived cell lines.”

Furthermor, antibody and recombinant protein production processes are increasingly implementing continuous processing (i.e., perfusion) while many viral vector manufacturing processes still utilize fed-batch processes, focusing instead on the transfection step, the Novartis experts note.

Cell culture for AAV viral vectors is particularly similar to mAb cell culture, including the fact that the basis of the cell-culture process for both are mammalian cells, adds Diane Blumenthal, head, technical operations, Spark Therapeutics. “We use reactors that are the same, albeit much smaller, as the amount of product needed for the patient population is smaller. We are often tackling diseases that are rare. As we put gene therapy to use where patient populations are larger, the scale will be larger.”

To emphasize the difference in output of therapeutic antibodies versus viral vectors for gene therapy, Blumenthal points out that antibodies are produced in bioreactors of 10,000 L in size or larger (commercial stainless-steel bioreactors), whereas AAV viral vectors are produced in volumes of 200 L to 500 L. “There may be some [AAV viral vector production] even larger at 2000 L, but I am not aware of any being utilized for larger stage product development,” Blumenthal says. 

“What is different is that we are using mammalian cells to make a virus as opposed to a protein,” she continues. In antibody production, the mammalian cells are modified to express the antibody of interest, but in viral vector production, the mammalian cells are treated to make them porous. This then allows the cells to take up plasmid DNA, which encodes for the production of virus. “The mammalian cell becomes the factory for the production of virus containing the gene of interest to be delivered to patient cells,” Blumenthal explains. 

The similarities to mAb cell culture benefit viral vectors. “In many ways, techniques that are now well established for antibody production can be applied to viral vector manufacturing. These include: metabolic flux analysis, bioreactor optimization based on mass transfer scale-up, and media and feed development. In addition, best practices around cell bank safety testing and genetic stability for cell lines (established with antibody producing cell lines) can be leveraged for viral vector manufacturing,” the Novartis experts state.

“It would seem that viral-vector manufacturing has the fortune of being able to leap-frog some technology development steps and jump right in line with where existing biopharma currently is,” Hsiao contemplates. “There is a bevy of tools and technologies available that have been put through their paces and available for deployment in current good manufacturing practices (cGMP) processes for making approved commercial products. The essential steps of scaling up cells, traversing seed trains, and reaching production volume are well established,” he notes.



“The nuances of proper transfection, optimization, reproducibility, and maximizing ergonomics are all areas that will require some development and for viral producers to potentially blaze new trails,” Hsiao adds.

Where they differ

The differences in the cell culture process between mAbs and viral vectors may also vary (e.g., media, buffers, equipment, processing techniques, etc.). For instance, media formulations themselves may differ as there may be different goals in production (e.g., cell health, protein production, or transfection susceptibility), but an emphasis on details, such as defined compositions, animal-origin free (AOF) components, or others, are always in play, notes Hsiao. “Current viral processes are largely batch processes of shorter duration, so media nutrient loading may also differ,” Hsiao says. 

“As for the equipment, similar stirred tank reactor (STR) systems are a straightforward option for suspension processes (mammalian or other). For adherent cell systems, alternatives include microcarriers in STRs, packed bed reactors, or roller bottles/multi-layer planar systems. The decision of which system to use is largely driven by capabilities and volume demands,” he adds.  

Meanwhile, regarding techniques, overall aseptic concerns are similar, but a higher biosafety level for the facility is needed to support human cell lines and potentially infective viruses and particles, Hsiao points out. “Similarly, standard 0.2-micron filters used in many antibody processes may or may not be sufficient in applications involving viruses, particularly when considering cross-contamination and operator safety,” he says.

For AAV manufacture, Blumenthal emphasizes, the media, buffers, and equipment are similar. “The main difference is in the solutions needed for the transfection step. In addition, the sterile technique is similar when culturing cells in bioreactors unless you are using roller bottles, which are more of a manual process and require manipulations to be done under laminar flow,” she says.

The Novartis experts weigh in as well, saying that there are differences between mAb and viral vector cell culture, but that those differences can be thought of as pertaining to the different cell lines (HEK-293 for viral vectors, CHO for antibodies) with different growth characteristics and different metabolic profiles. “Generally speaking, CHO cells tend to be more robust when it comes to tolerating stresses (shear forces, pH, or temperature excursions, etc.), but viral vector manufacturing is typically carried out at a smaller scale than antibody production due to inefficiencies in transfection at large scale and instability in viral particles upon passaging. In addition, common bioprocessing steps, such as filtration, are more challenging with viral vector manufacturing because of the physical properties associated with viruses (lentivirus, AAV, etc.),” the company states.

Limitations for viral vectors

Despite the positives on the side of viral vector cell culture, viral vectors are also subject to limitations under current cell-culture technologies. Among these limitations is scale, according to Blumenthal. “The ability to perform the transient transfection process at larger scales is still a challenge,” she says. The transient transfection process requires the use of plasmids, and for some systems, three plasmids are required. In some cases, only two. Because of this, the quantities of DNA required can become large, and this is a critical raw material. Unfortunately, there is only a limited number of manufacturers that produce cGMP quality plasmids, Blumenthal explains. Many companies are moving to making their own plasmids, however, she says. 

Further, production demands in viral vector manufacturing seem lower than that for mAbs, in general, notes Hsiao. He attributes this lower demand to current targeted indications that have not required wide deployment, as is the case in some therapeutic antibodies. “Thus, the high cell densities, large terminal volumes, and maximal cellular titer that are the hallmarks for optimizing antibody processes may not be highest priorities in viral work. Other items, such as efficiency of gene delivery, proper reagent addition timing, and exposure durations, likely take precedent. In both cases, product quality and purity are still vital considerations,” Hsiao says.

“Current methods for the production of infectious viral particles rely on hard-to-produce, low yielding processes,” adds Novartis. The company points out that improvements in specific areas, such as electroporation, stable cell line development, filtration, and aggregation deterrence would dramatically improve the robustness and consistency of viral vector manufacturing. “While some companies have made the switch from adherent to suspension cell lines, not everyone has. Adherent processes are labor-intensive and costly,” the company’s experts state.


1. J.N. Warnock, et al., Cytotechnology 50, 141–162 (2006). 

Article Details

BioPharm International
Vol. 33, No. 2
February 2020
Pages: 20–21


When referring to this article, please cite it as F. Mirasol, “Cell Culture Variables for Gene Therapy Vectors,” BioPharm International 33 (2) 20–21 (2020).