Our two investigated cell strains, although derived from the same ancestral HEK-293 cell, have already been shown to exhibit a highly versatile transcriptome. Accordingly, their response to transient plasmid transfection is unique, reflecting their strength and weaknesses as a producer cell line (Grieger, Soltys et al. 2016, Srivastava, Mallela et al. 2021, Torabfam, Yetisgin et al. 2022, Pistek, Kahlig et al. 2023). In this study, we identified multiple bottlenecks for LP cell line along the whole transfection process, including reduced plasmid uptake and nucleus entry in addition to disadvantages in protein translation. The data shown here indicates that the ratio of vector genome-to-capsid can be improved for both cell lines to an individual extend by optimizing relative plasmid ratios. However, further optimization of the transfected DNA composition is required, and must be tailored to each production cell line.
Cell strain-specific transfectability.
The efficacy of nuclear transport of plasmids is a multifaceted issue in transient gene delivery, influenced both, by their intrinsic design and varying cellular environments (Vacik, Dean et al. 1999, Le Guen, Pichon et al. 2021, Haraguchi, Koujin et al. 2022). For example, plasmid size and supercoiled structures influence cellular and nuclear entry (Sousa, Prazeres et al. 2009, Ribeiro, Mairhofer et al. 2012), methylation patterns can affect their stability and susceptibility to enzymatic degradation (Hong, Sherley et al. 2001), and the choice of codon usage determines translational efficiency within the host cell system (Powell and Dion 2015).
Nevertheless, identical plasmids may show different transfection efficacy across different cell lines, indicating a strong interplay with cellular factors (Vacik, Dean et al. 1999, Maurisse, De Semir et al. 2010). For instance, elevated expression of cytosolic nucleases may lead to the degradation of higher amounts of plasmid DNA; to allow cytosolic movement along the cytoskeleton, an intricate interplay of numerous cellular genes is required; and the presence of importins has shown to be important for successful nuclear entry (James and Giorgio 2000, Pollard, Toumaniantz et al. 2001, Vaughan and Dean 2006, Yang and Musser 2006).
In our comprehensive study with HEK-293 cell lines, we demonstrated that transfection efficiency even varies across cell strains. Cellular uptake, nuclear accumulation, and the ability to express the genetic information has been shown to differ between HP and LP. Underlying mechanisms may be found in the cell’s distinct transcriptome, and thus, diverse responses to transfection stimuli (Pistek, Kahlig et al. 2023). Detailed investigation presented by Warga and colleagues (Warga, Anderson et al. 2023), sheds further light on molecular factors influencing gene delivery outcomes and provide a foundation to our current findings. However, deeper analysis of our transcriptome data will be required to unravel potential influencing host cell genes for future cell line optimization.
The importance of balancing the amount of exogenous viral genes during rAAV particle production.
Numerous efforts have been made to refine the production of recombinant adeno-associated virus particles, focusing on improving culture conditions and transfection techniques, such as plasmid ratio optimization (Durocher, Pham et al. 2007, Zhao, Lee et al. 2020). We found that the success of those approaches varies across literature, and that the effectiveness of these optimizations is influenced by variables like the specific producer cell strains used, the reagents and methods applied, and the produced AAV serotype. Consequently, the ideal conditions including the plasmid ratios, may vary depending on each unique experimental setup.
In our study, we reduced the amount of RepCap plasmid, which resulted in a substantial decrease in the number of produced particles exceeding our expectations. The multiplicity of infection therefore emerged as a critical factor for RepCap plasmids, determining overall yield. Concurrently, we significantly elevated the quantity of transgene plasmid with the expectation of remarkably enhancing transgene integration into our pre-formed empty capsids. Nevertheless, this was not observed to the expected extend, indicating that the availability of the transgene may not be the limiting factor. Instead, we found upon closer examination, the packaging mechanism might entail a more complex interplay among the gene products of all three plasmids: We discovered that the quantity of RepCap plasmid alone may not solely determine the number of produced particles. Reducing the amount of transgene and helper plasmids while keeping the RepCap plasmid constant resulted in increased particle production. Potential explanations consider the impact of varying amount of input DNA used or that the ratio between RepCap and helper plasmid plays a significant role in the outcome of overall production (Grieger, Soltys et al. 2016, Guan, Chen et al. 2022). Support for the latter hypothesis is evident in our comparison of the vector genome-to-particle ratio: a superior full-empty ratio of particles was observed under conditions with a higher relative ratio of helper to RepCap plasmid. This aligns with the findings of Nguyen and colleagues (Nguyen, Sha et al. 2021), suggesting that increasing the relative amount of helper to RepCap plasmids to a certain extend might yield beneficial outcomes in generating full particles. Nevertheless, helper genes have been shown to not only have beneficial, but also negative effects on rAAV production (Smith and Kotin 1998, Nayak and Pintel 2007), therefore, balancing the gene products in a triple plasmid transfection is crucial for resulting yield. We therefore propose for subsequent experiments to increase the helper:RepCap ratio of our standard condition T without significantly raising the overall DNA amount to obtain a better vector genome-to-capsid ratio in secreted particles.
Introducing another layer of complexity involves considering the individual transcriptional and translational capacities of the respective cell lines. It is established that rAAV production follows a sequential process, commencing with the generation of empty capsids and subsequently involving the replication of the transgene and its encapsulation into pre-formed capsids. Consequently, swift gene expression leads to the predominant production of empty capsids, as they are secreted before the incorporation of the desired transgene can occur (Nguyen, Sha et al. 2021). Conversely, maintaining a moderate level of protein expression may be advantageous, facilitating the concurrent production of both protein translation and transgene transcription. This could potentially lead to higher quantities of transgene encapsulation. The difference in protein expression dynamics may explain why our high producer (HP) cell line secretes a substantial number of particles, yet the low producer (LP) cell line exhibits superior full-empty ratios despite an overall lower yield. Hence, there exists the potential for optimization in each cell line not only in terms of quantities but also in the precise timing coordination of gene expression (Ohba, Sehara et al. 2023).
Taken together, rAAV production is a delicate interplay between cellular pathways and induced viral gene expression. Cellular burden stimulated by exogenous DNA may differ between cell strains and need to be considered when designing future gene therapy approaches. Further dissection of factors that influence viral vector production will allow to increase productivity, and moreover, patient safety. Therefore, understanding the unique capabilities of the utilized producer cell line and their reaction on viral vector production is crucial for the development of efficient production processes in biopharmaceutical manufacturing and will contribute to future advances in rAAV gene therapy applications.