(175be) Integrated Omic Analysis Elucidates Phenotypic Plurality and Metabolic Bottlenecks in Industrial Antibody-Producing Chinese Hamster Ovary (CHO) Cell Lines

Bioprocess optimization and cell line development benefits from host cell line characterization to understand their metabolic capabilities, responses to the reactor conditions, and product-specific demands. While process data reveals nutrient limitations arising from depletion, it does not provide any insights into bottlenecks in cellular metabolism that currently affect antibody productivity. Here, we present a multi-omic characterization of six independently selected antibody-producing Chinese Hamster Ovary (CHO) cell lines to reveal the key metabolic reprogramming correlated with increased antibody production as well as the metabolic bottlenecks limiting antibody production in high-producer clones for each cell line. Analysis of spent media and product formation in a fed-batch process reveals that high-producer clones exhibit diverse phenotypes with increased antibody titer and no clear correlation between final cell density and titer. exo-metabolomic analysis also revealed the critical role played by reprogramming of nitrogen metabolism in improving antibody production by high-producing clones. High-producing clones showed changes in overall glycine and serine consumption and overall lactate, alanine, aspartate, and glutamate secretion over the four identified phases of the bioprocess. To further investigate the metabolic differences among the six cell lines, we constructed context-specific models by integrating gene expression data with iCHO1766, the genome-scale metabolic model for CHO using the mCADRE method. Context-specific-specific metabolic models revealed differences in lipid and cofactor metabolism, glycosylation, and transporters; meanwhile, central, amino acid, nucleotide, and energy metabolism were largely conserved across cell lines and process phases. To further elucidate pathway usage differences in the different phases and cell lines, we performed Monte-Carlo flux sampling. Sampled metabolic fluxes demonstrated substantial changes in the pentose phosphate pathway and lactate secretion, especially in high-producing clones. Generally, glycolytic metabolism was fueled by glucose uptake and TCA was fueled by the degradation of asparagine, glutamine, and other essential amino acids. High-producing clones exhibited changes in glycine and serine consumption and lactate, alanine, aspartate, and glutamate secretion, and on average, degraded 65% of the consumed amino acids. The uptake of arginine, proline, asparagine, glutamine, phenylalanine, valine, and isoleucine limited cell growth and antibody production in different cell lines. Thus, a wide range of metabolic activities can exist, even with similar cell line development strategies, and a multi-omic analysis of high-producer clones can guide cell engineering and bioprocess optimization.