Functional Characterisation of the Cellular Proteome

Engineered mammalian (CHO, NS0 and SP2/0) cells are commonly used for large-scale production of recombinant therapeutic antibodies. Increased antibody production in these commercial manufacturing systems is often achieved through systematic empirical optimisation of culture environment and improved expression vectors. To date, our a priori knowledge of the molecular mechanisms of antibody manufacture in these expression systems is very limited. As a result, genetic engineering of mammalian host cells to improve product yields have often produced mixed results, further indicating the need for a more exploratory approach to understanding glycoprotein production. Identification of coordinated changes specifically related to one particular phenotype (i.e. cell-specific productivity; qP) at a clonal level will improve our understanding of the rate limiting steps or intracellular processes related to that phenotype. Using quantitative two-dimensional polyacrylamide gel electrophoresis, differences in the host cell proteome were assessed between murine myeloma NS0 cells producing a recombinant monoclonal antibody at varying cell-specific production rates. Furthermore the variation in synthesis and abundance of these proteins were evaluated during fed-batch culture. Conserved and coordinated changes in specific proteins known to interact with antibody folding and assembly were identified, as were changes in the intracellular abundance of recombinant heavy and light chain polypeptides. Conserved changes in the abundance of proteins as functional categories (chaperones, metabolic and mitochondrial, cytoskeletal and cell signaling) also demonstrated correlation with enhanced cell-specific productivity between clonal cell lines. However, variation in the synthesis and abundance of these proteins (individually and in categories) during a fed-batch process did not demonstrate the same coordinated changes with qP observed between clonal cell lines. In conclusion these data suggest that cell performance in production processes is de- fined primarily by the basal phenotype selected after transfection. Bioprocess manipulations modulate management of host cellular processes altering metabolism and recombinant gene transcription or translation, but only within the confines of the host cell machinery selected post-transfection. Pre-screening of cells prior to selection for coordinated changes associated with improved qP will reduce the variation in transfected cell phenotypes, reducing the time to select cells with good performance characteristics