At-line quantitative profiling of monoclonal antibody products during bioprocessing using HPLC-MS

(1) Background: The N-glycosylation profile as well as the subunit assembly of monoclonal antibodies (mAbs) are strongly dependent on manufacturing conditions and thus need to be monitored during the bioprocess. Commonly, mAbs are characterized downstream of the fermentation process applying different analytical techniques like released glycan analysis, peptide mapping, or subunit profiling. However, these procedures are time-consuming and difficult to perform in real-time. (2) Methods: We applied a simple HPLC-MS workflow with minimal sample preparation to characterize mAb product quality at the intact protein level at different time points during fermentation. After harvest, the cell culture medium was centrifuged briefly. The supernatant containing the fermentation product was diluted and immediately subjected to HPLC-MS analysis. (3) Results: Besides the product of interest (mAb), the fermentation broth contained misassembled variants, mostly light chain and light chain dimer. The mAb's glycosylation profile changed over time showing an increase in galactosylated variants with G0F/G1F being the most abundant glycoform at all time points of fermentation. Furthermore, expressed protein species were relatively and absolutely quantified. The workflow was very robust despite analyzing a highly complex matrix. Relative standard deviations for retention times were below 0.5% for both intra and inter-day comparison, whereas relative procedural standard deviations for quantification of the different protein species ranged between 7 and 13%. (4) Conclusions: This approach

Automated summary

The study used a simple HPLC-MS workflow to analyze mAb product quality in real-time at different fermentation time points. The results revealed the presence of misassembled variants in the fermentation broth, alongside changes in mAb glycosylation profile over time, with galactosylated variants being the most abundant. The workflow proved to be robust for analyzing the complex matrix with low standard deviations.