Development of a Tunable Mammalian Protein Expression System and an Investigation of Promoter Interference in Three Promoters Often Utilized in the Production of Biopharmaceuticals

Cell line engineering strategies for improved biopharmaceutical production in mammalian cells often involve the expression of one/multiple genes to try and improve the cellular processes involved in recombinant protein production. Most strategies are relatively simple, involving the use of a strong constitutive promoter for expression of one or more proteins to help increase production. Results often vary and can be cell line and product specific and mean a generic strategy is unlikely to be found. There is a need for more sophisticated expression systems which can express multiple genes but in a controlled fashion and tuned to meet the needs of a specific product. This thesis can be split into two distinct parts but both concern the expression of multiple genes in mammalian cells and recombinant protein production. A tunable mammalian expression system for multi-gene engineering composed of elements of the mammalian unfolded protein response has been developed. ATF6 (activating transcription factor 6) and its binding element (ERSE – ER stress response element) were used to control the expression of the reporter proteins SEAP (secreted alkaline phosphatase) and GFP (green fluorescent protein). By expressing different amounts of ATF6 and by inserting different numbers of ERSEs upstream of a SV40 (Simian virus 40) promoter, driving SEAP/GFP gene transcription, the level of reporter protein expression could be manipulated in a controlled fashion. The system was capable of controlled/tunable expression of both reporter proteins when expressed alone and when they were co-expressed. This a novel use for ATF6 and ERSE and the first step towards the development of a tunable mammalian expression system for multi-gene engineering. This system could also be easily modified to include or use different transcription factors and binding sites as well as having the potential to use completely synthetic components. This work also showed that the presence of ‘promoter interference’ (the negative influence of one promoter on another) could be used to our advantage to increase the range of expression. The SV40 early, human CMV (cytomegalovirus) major immediate-early and human EF1α (elongation factor 1 alpha) are constitutive promoters frequently used in recombinant protein production. The former being used mainly for expression of selection genes and the latter two for strong expression of recombinant proteins. The differences in the strengths of the promoters was demonstrated in CHO (Chinese hamster ovary) cells (CMV > EF1α > SV40) and also their abilities to negatively affect the expression from a co-expressed promoter. The negative influence of one promoter on another is termed ‘promoter interference’. The CMV promoter was shown to have the greatest negative effect on expression from another promoter, decreasing both SEAP mRNA and protein expression, while the SV40 had the least. SEAP expression from the SV40 was reduced the most by the presence of a competing promoter. The level of interference inflicted by a competing promoter (CMV > EF1α > SV40) seemed to be relative to its strength. This is the first time these three important promoters have been compared in a way which not only demonstrates their relative strengths but also their ability to interfere with another promoter when present in the same transient expression system. This also has implications for their use in multi-gene engineering strategies if there is a need for controlled/tunable expression of multiple genes. The work with ATF6 and ERSE showed how ‘promoter interference’ could be used to our advantage and not necessarily be just a negative occurrence. One hypothesis for why promoter interference occurs is there is competition between promoters for shared transcription factors (TFs). The promoters were analysed for potential transcription factor binding sites (TFBSs) using the programs MatInspector and ModelInspector. The analysis showed that the SV40 promoter had the least number and variety of potential TFBSs. Both the CMV and EF1α had greater numbers and variety of potential TFBSs. All three promoters had common potential TFBSs but the SV40 promoter shared a greater proportion of its sites with the other two promoters. The number of potential TFBSs and the proportion which were shared reflected both the strength and the ability of a promoter to interfere with another. All three contained TFBSs for the SP1 (specificity protein 1) family of TFs and over-expression of SP1 counter-acted the effects of promoter interference showing that it can affect the expression of all three promoters. However, promoter interference will involve more than just a single TF and also more than just competition for transcriptional activators. This is the first time these three promoters have been compared in terms of the potential TFBSs they contain. The TFBS analysis highlighted the complexity in the control of these promoters and with the effects of promoter interference means that they will be ill suited for the controlled expression of multiple genes without modification. The work in this thesis was directed towards the controlled expression of multiple genes in mammalian cells for recombinant protein production. We have presented one novel way of controlling the expression of one/two genes with the rest of the thesis looking at the phenomenon of ‘promoter interference’ between three commonly used promoters. This thesis tries to highlight the importance of multi-gene expression systems as well showing that these three promoters may not be suitable without further modification and also the importance of considering promoter interactions when more than one is present in the same system. The switch to completely synthetic multi-gene control systems is something we envisage happening in the future as the complexities and capabilities of these systems grow.