Back

Exhaustive Isotope Tracing Reveals the Butterfly Effect of Mammalian Producer Cell Line Selection for Biomanufacturing

Dhara, V. G.; Gonzalez, J. E.; Naik, H. M.; McConnell, B. O.; Khare, P. A.; Sargunas, J.; Antoniewicz, M. R.; Betenbaugh, M. J.

2026-08-05 cell biology
10.64898/2026.08.04.742784 bioRxiv
Show abstract

Chinese hamster ovary (CHO) cells are the dominant platform for recombinant biotherapeutics, yet the impact of producer cell line selection on mammalian cell metabolism remains poorly understood. Here, we performed 41 independent 13C-tracer experiments using uniformly labeled glucose or individual amino acids to comprehensively map carbon utilization in the two principal CHO production platforms: methotrexate-selected CHO-K1 and glutamine synthetase-selected CHO-GS cells. Time-resolved GC-MS analysis revealed distinct metabolic phenotypes spanning central carbon metabolism, amino acid interconversion, lipid biosynthesis, and one-carbon metabolism. CHO-K1 cells exhibited extensive reductive carboxylation and pyruvate carboxylase-mediated anaplerosis, whereas CHO-GS cells redirected glutamate toward glutamine synthesis and relied on asparagine and aspartate to support TCA cycle activity. Isotopomer analysis uncovered substantial intracellular-extracellular cycling of alanine, glycine, glutamate, and serine despite contrasting uptake profiles and quantified differential amino acid contributions to fatty acids and cholesterol. Serine, glycine, and methionine labeling revealed active folate-cycle interconversion in CHO-K1 and enhanced methionine-cycle activity in CHO-GS. Aspartate is identified as a key redox exchange factor and uniquely informative tracer for pathway characterization following glutamine depletion. Together, this exhaustive isotope-tracing framework establishes how producer cell line selection rewires mammalian metabolism and provides a foundation for cell engineering, media optimization, metabolic modeling, and next-generation biomanufacturing.

Matching journals

The top 6 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.