Application of metabolomics profiling of recombinant mammalian cells to bioprocess design

The diagnosis and treatment of many clinical conditions is dependent on developments of new therapeutic drugs / complex protein molecules that require production by mammalian cells in culture. Advances in genetic engineering and understanding in gene expression has made it possible to turn mammalian cells in defined culture conditions into 'factories' to permit harvest of the valuable therapeutic protein drugs. Hence it is possible to introduce into mammalian cells in culture a gene that will direct synthesise the desired drug and allow the mammalian cells to perform all the necessary reactions required to generate the therapeutic protein. Some of the drugs generated by these approaches are household names (eg insulin and clot-buster drugs) and others are not (eg antibodies), however all are critical and essential components in treatments for many life-threatening clinical conditions. A key problem is that the development and generation of sufficient amounts of therapeutic protein drugs is limited by the capacity to generate sufficient production from the cell cultures. Consequently, the investment of development and process time to harvest sufficient amounts of therapeutic proteins makes the process costly and these are expensive drugs. Despite significant effort by industrial and academic researchers we still do not fully understand the factors that limit production by the cell culture system and such information will be essential if we are to devise means to increase production. The unit cost for certain of these therapeutic protein drugs has implications for availability for treatment and for certain 'difficult-to-make' drugs financial implications may prevent the drug being developed for the market. This proposal addresses this problem and seeks to use novel approaches to ask how we can improve the process of therapeutic protein formation by mammalian cells in culture. Our approach is to take a wide vision of the events within cells that limit formation of the desired drug. Too frequently experimental approaches to study the functions of cells focus on one specific aspect or examine functions in the cell for which there is no direct certainty of involvement in the processes under study. We are building complex models that will determine how the many factors in the cell interact to define how production of the therapeutic protein is controlled. We will then test the model in response to conditions that we impose in cells and, from this, determine if there is a means to identify those cells in a population with desirable characteristics that will give excellent production or if we can engineer these functions into cells. The research team in this programme have come together from distinctive areas of research, allowing a fusion of technologies to emerge from the interactions. The fusion of analytical scientists, cell biologists, chemical engineers and mathematical modellers generates a novel team-based approach to an industrially- (and, ultimately) clinically-relevant issue. The team will be part of a wider network of scientists (in academic and industrial groups) who are part of the Bioprocessing Research for Industry Club, the government/industrial initiative to maintain the UK research strength in this area.

CHO and NS0 myeloma cells can be engineered for directed expression of desirable biopharmaceutical proteins. Optimisation of the overall bioprocess is critical for the commercial viability for the potential biotherapeutic. As an overall process, we do not understand how the complexity of the cell and interactions within the cell and with the environment operates to determine a 'good' or 'bad' bioprocess. Recombinant mammalian cells in culture present an ideal subject for analysis by a Systems Biology approach to explain the higher order behaviour and function of this complex systems by analysing and modelling genetic, macromolecular and metabolic networks. We will identify parameters that characterise good recombinant CHO and NS0 myeloma cells in relation to expression. Initial metabolic fingerprints will be obtained by FT-IR screening and more detailed integrative data will be obtained by application of metabolome analysis (GC-MS and LC-MS), proteome analysis (2-D gels, peptide mass fingerprinting, ESI-MS-MS) and transcriptome analysis (Affymetrix technology). The primary focus will be at the level of the metabolome, with data at protein and transcript level being used to aid in the generation of a quantitative understanding of the interactions between elements in the system. Our data analysis will interlink with the development of predictive and explanatory models of the behaviour of recombinant cells in relation to recombinant protein expression. Modelling of central metabolism and glycosylation will be used to provide a quantitative perspective to our experimentally-obtained data and we will use these complementary approaches to predict which experimental manipulations may provide a route towards engineering improved bioprocessing. The 'hallmarks' of systems biology are that it seeks to be quantitative, integrative, and synthetic - approaches that complement the more qualitative analytical strategies of pre-genomic biology.