Herein we present a novel computational algorithm based on Flux Balance Analysis (FBA), that can account for the metabolic burden associated with the integration of plasmids in host-cells as a function of plasmid copy number, plasmid size and recombinant gene expression efficiency. The aim is to provide a computational tool to assist the decision-making process during plasmid design and cell-line development. Initially, FASTA files for the plasmid and any included recombinant proteins are parsed to derive the resource costs in terms of required nucleotide and amino acid building blocks as well as in terms of energy (ATP/ADP) metabolites. Subsequently, bespoke stoichiometric reactions for the synthesis of the plasmid (if/when applicable), the selection marker protein (usually some form of Antibiotic Resistance) as well as any other recombinant protein are developed based on the required metabolic costs computed in the previous step. Finally, by employing the recently developed Design Space Identification (DSI)[3] method from Papathanasiou and coworkers, the effect of heterologous gene expression intensity and selected growth rate on the flexibility and robustness of the process operating window is quantified.
We compared the metabolic profiles of E. coli BL21 cells transformed with the pORI1 and pORI2 plasmids using FBApi. Since the expression level of the plasmid encoded genes and the heterologous protein fractional contribution to the biomass vary across different strains and organisms, we created multiple models that cover the range of these parameters.
The solution space of the resulting models was sampled using the Artificial Centering Hit and Run (ACHR) algorithm and the retrieved samples were anonymized and randomly scrambled to eliminate any potential bias. Principal Component Analysis (PCA) was then applied[4] to the mixed samples to identify whether any differences in metabolic configuration could be identified and whether said differences could be a factor of plasmid size. PCA was able to discriminate the metabolic states of the cells containing plasmids and to effectively select reactions/subsystems affected in the metabolic network.
The proposed algorithm was able to quantify the productivity of plasmids, mRNA, and recombinant proteins by considering their structural characteristics, such as size and nucleotide/amino acid sequences, along with the tradeoff between growth rate and plasmid-associated reactions. This makes it a robust and versatile tool well-suited for a wide range of biotechnological applications.
REFERENCES:
[1]: Heyland, J., Blank, L. M., & Schmid, A. (2011). Quantification of metabolic limitations during recombinant protein production in Escherichia coli. Journal of Biotechnology, 155(2), 178–184. https://doi.org/10.1016/j.jbiotec.2011.06.016
[2]: Gotsmy, M., Strobl, F., Weiß, F., Gruber, P., Kraus, B., Mairhofer, J., & Zanghellini, J. (2023). Sulfate limitation increases specific plasmid DNA yield and productivity in E. coli fed-batch processes. Microbial Cell Factories, 22(1). https://doi.org/10.1186/s12934-023-02248-2
[3]: Sachio, S., Likozar, B., Kontoravdi, C., & Papathanasiou, M. M. (2024). Computer-aided design space identification for screening of protein A affinity chromatography resins. Journal of Chromatography A, 1722. https://doi.org/10.1016/j.chroma.2024.464890
[4]: Luginsland, M., Kontoravdi, C., Racher, A., Jaques, C., & Kiparissides, A. (2024). Elucidating lactate metabolism in industrial CHO cultures through the combined use of flux balance and principal component analyses. Biochemical Engineering Journal, 202. https://doi.org/10.1016/j.bej.2023.109184