Utilising a genome scale metabolic model to design high-producing CHO cell lines through bi-level optimisation
File(s)
Author(s)
Antonakoudis, Athanasios
Type
Thesis or dissertation
Abstract
Monoclonal antibodies (mAbs) are a type of biopharmaceutical product, that binds to specific antigens in the body and are administrated mainly for the treatment of cancer or autoimmune disorders. Unfortunately, mAbs suffer from a high market cost, forbidding treatment to many patients. Chinese hamster ovary (CHO) cells are the dominant expression platform of the biopharmaceutical mAb. However, despite increasing global demand and persistent societal needs, mammalian cell-based production systems suffer from low product yields as specific productivity is inversely proportional to cellular growth rates. The low yields can carry a burden on the upstream production phase of mAb production in CHO cell cultures, increasing the cost of the final bio product. The work presented here aims to reduce the upstream production cost of mAbs in CHO cell systems by proposing an increase of their specific productivity. Specifically, the design of a high-producing CHO cell line is attempted herein using a combination of in silico and in vitro techniques. Initially, the most recently developed genome-scale metabolic model (GEM) iCHO1766, on the time of commencing the research, is manually curated by limiting the cellular uptake and secretion rates based on four different experimental exometabolomic datasets and applying elemental balance algorithms to obtain realistic bounds for the intracellular reactions. Then, bilevel optimization is applied to the GEM, aiming in detecting genetic regulations that would increase specific mAb productivity during the growth phase of the cell culture. The strain optimisation on the CHO-K1 GEM results in a plethora of scenarios, all of which can secrete antibody in a biomass maximising Flux Balance Analysis (FBA) optimisation. These scenarios contain a large number of regulations, making them unsuitable for further investigation. A sensitivity analysis framework is then developed to first reduce the number of regulations to the absolute necessary ones for antibody synthesis, and secondly, to detect the best scenarios based on their FBA-predicted growth rate and specific productivity. An in-depth analysis of the proposed metabolic changes suggests that the regulations create an abundance in one of the amino acids that the cell uptakes, which must be consumed through the mAb synthesis pathway. Finally, two genetic engineering scenarios are screened experimentally. The first contains a knockout of Aldh4a1 and an overexpression of Aldh18a1. However, the protein ALDH18A1 could not be detected, so it was thought that the gene is not being translated in the CHO cell line used for genetic modifications. The second scenario included just a knockdown of Bcat1. The cells are then cultured in mild hypothermic fed-batch conditions for 14 days. This modification displayed a 35% increase in the final mAb titer compared to the unmodified CHO cell line. The scenario requires further future testing to determine the underlying reason for the 35% increase in titer; however, the methodology followed shows promise for future use for CHO cell and, more generally, mammalian cell metabolic network engineering.
Version
Open Access
Date Issued
2022-07
Date Awarded
2022-12
Copyright Statement
Creative Commons Attribution NonCommercial Licence
Advisor
Kontoravdi, Kleio
Κiparissides, Alexandros
Sponsor
Engineering and Physical Sciences Research Council
Grant Number
EP/IO33270/1
Publisher Department
Chemical Engineering
Publisher Institution
Imperial College London
Qualification Level
Doctoral
Qualification Name
Doctor of Philosophy (PhD)