Development of a High-Throughput Analytical Platform for the Characterisation of Biopharmaceuticals Generated from Multi-Gene Engineered CHO Cells

The biopharmaceuticals market was valued $325 billion in 2020 with revenues forecasted to rise in the upcoming decade. High-throughput process development (HTPD) tools characterised by miniaturisation, parallelisation and automation of industrial bioprocessing workflows have emerged as the key enabler for leveraging costs and rapidly launching new products to market. However, the small sample generated from HTPD experiments and the fast timelines introduce analytical bottlenecks. Hence, more efficient high-throughput (HT) analytical tools are required to ensure the quality necessary to meet regulatory approval. This study aims to address some of the gaps in the current methodologies for the characterisation of protein-based biopharmaceuticals and implement new HT analytical solutions for driving product development in the industry. The first part of this work presents the generation of a flexible HT platform for the analysis of proteins expressed in microscale cell culture systems. Chemometric tools such as design-of-experiment (DoE), multivariate analysis (MVA) and predictive modelling were used to leverage method development and support product quality (PQ) analysis with statistical data, in line with the quality-by-design (QbD) approach. Fully automated pipette-tip based methods were optimised for the purification and preparation of low volume samples. Rapid HT size-exclusion (SEC) and cation-exchange chromatography (CEX) methods enabled the profiling of size and charge variants with sub-2 g sensitivity. Online coupling of SEC with CEX in a two-dimensional liquid chromatography (2D-LC) set up with mass spectrometry (MS) detection (2D SEC-CEX-MS) combined aggregation analysis and charge variants peak identification in a single method maximising the data acquired from minimal sample requirements. Targeted MS analysis of cysteine modifications by multiple-reaction monitoring (MRM) was used to quantitate aggregates and fused into multi-attribute method (MAM) for supporting the analysis of cysteine modifications, glycosylation, deamidation and C-terminal Lys processing throughout the development of a bispecific antibody product. Finally, the methods were integrated into a HT analytical platform supporting HTPD experiments for cell line development (CLD). The HT analytical platform processed several hundreds of small volume samples providing PQ data necessary for clone selection. This demonstrated the benefits of hyphenating HT analytics as early as possible to drive CLD activities and gather a deeper understanding of the scientific and engineering principles that govern protein expression in cells. The HT analytics developed in this research will be implemented in the industrial analytical workstream to support AstraZeneca pipeline.