The accumulation of lactate during cultivation of mammalian cells for biopharmaceutical production is a longstanding issue affecting glycosylation quality and productivity. Many approaches exist to mitigate its impact, either through the replacement of glucose with slowly metabolised sugars, dynamic feeding strategies, or host cell engineering. The manipulation of genes in this latter approach is constitutive and may suboptimally respond to cellular needs.
The LldR proteins from Corynebacterium glutamicum and Pseudomonas aeruginosa have been used in this project to create a lactate-inducible transgene expression system, which can be used subsequently to dynamically drive expression of proteins previously targeted to mitigate the accumulation of lactate. Expression and purification of these LldR proteins, fused to transcriptional effector domains in various orientations and with fusion linkers in certain cases, allowed in vitro characterisation and optimisation of the constituent parts of the inducible system. This provided crucial information in some cases about the need to use a flexible linker between LldR and a VP64 transactivation domain. In vivo experimentation of these optimised systems showed significant levels of induction in response to 20 mM lactate, with a 3.46-fold decrease in expression seen for one construct. Some preliminary work was also carried out with Cas9-VPR, which was shown to be able to upregulate transiently transfected genes up to 1.6-fold. In the future, this will be a useful tool for upregulating multiple previously identified targets, as well as helping to find new beneficial targets.
The general approach outlined here for the development of this lactate-inducible transgene expression system will be appropriate for any other such project where the ligand of interest is a central metabolite. Inherently weaker induction might be a feature of such a system, given the presence of the inducer at low and relatively benign or neutral concentrations throughout a period of interest; testing an unoptimised system in mammalian cells may return little or no detectable induction signal and therefore it will not be straightforward to optimise such a system solely through the use of in vivo experimentation. In vitro characterisation of the inducible system components, as performed here, can provide essential feedback regarding the impact of effector domain fusion and operator design on the DNA-binding affinity of the biosensor prior to in vivo testing.
The work in this thesis will allow the future exploration of dynamically regulated host cell engineering designed to combat the lactate accumulation phenotype.