Lipidomics for robust high performance process development

Background and novelty: CHO cell lines are common hosts for the production of biopharmaceuticals proteins. So far, considerable progress has been made increasing productivity of cell culture to meet the rapidly growing demand for antibody biopharmaceuticals through increased cell densities and longer cultured time. These major improvements have nevertheless lead to an important drawback : the increase of the process related impurities, bringing new challenges for process and harvest development. Among the process related impurities such as HCPs or DNA the potential impact of lipids production and release during a cell culture is still poorly understood due to the complex nature and diversity of this class of molecules. Thanks to recent advances in analytical tools especially mass spectrometry, the advent of lipidomics offers now the feasibility to study several thousands of lipid species, unravelling the possibility to understand and potentially control the interactions between high performance bioreactor processes, harvest conditions and purification.
Experimental approach: In order to analyze and quantify lipids, we developed a three steps method. In a first step, lipids were extracted with Methyl tert-butyl ether (MTBE) according to Matyash method. Lipids were then separated by liquid chromatography using either HILIC of reverse phase column prior to detection and quantification by mass spectrometry. All lipid classes were detected by ESI-MS/MS excepted cholesterol (APCI-MS/MS). Finally we applied this method to analyze the lipid content of four different cell lines each expressing a different recombinant protein, during a 14 days fed batch process.
Results and discussion: Lipid from CHO cells were successfully extracted with a yield between 80% and 95% depending on the different lipid classes. Stable isotope labelled lipids were used as internal standard in order to have comparable results between batches. The lipid profiles were not only different for the 4 tested cell lines, it was also different for a given cell line cultivated under various experimental conditions. Interestingly, in some cell lines/experimental conditions, we highlighted an overproduction of triglycerides and cholesterol leading to the accumulation of lipid droplets known as energy storage sink. At the metabolic level, these finding suggest a relative overflow of the carbon metabolism. From a process development perspective these findings can be considered on the one hand as a resource waste since the stored energy is not used for protein/biomass biosynthesis and on second hand to additional process problems during the harvest and the first capture steps given the hydrophobic nature of these molecules. Additional analysis will required to fully understand the link existing between lipids metabolism, cell line, and process development conditions. We believe that these investigation will open the door to many applications such as clone selection, process, and harvest development.