Online multimethod platform for comprehensive characterization of monoclonal antibodies in cell culture fluid from a single sample injection - Intact protein workflow

Background: Therapeutic monoclonal antibodies (mAbs) comprise a large structural variability with respect to charge, size and post-translational modifications. These critical quality attributes (CQAs) need to be assessed during and after the production of mAbs. This normally requires off-line purification and sample preparation as well as several chromatographic selectivities, which makes the whole process time-consuming and error-prone. To improve on this, we developed an integrated and automated multi-dimensional analytical platform for the simultaneous assessment of multiple CQAs of mAbs in cell culture fluid (CCF) from upstream processes.

Results: The on-line system allows mAb characterization at the intact level, combining protein A affinity chromatography (ProtA) with size-exclusion, ion-exchange, and reversed-phase liquid chromatographic modes with UV and mass spectrometric detection. Multiple heart cuts of a single mAb elution band from ProtA are stored in 20-μL loops and successively sent to the multimethod options in the second dimension. ProtA loading and elution conditions and their compatibility with second-dimension LC modes were studied and optimized. Subsequently, heart-cutting and valve-switching schemes were investigated to achieve effective and reproducible analyses. The applicability of the developed workflow was demonstrated by the direct analysis (i.e. not requiring off-line sample preparation) of a therapeutic mAb in CCF, obtaining useful information on accurate molecular mass, glycosylation, and charge and size variants of the mAb product at the same time and in just over 1 h.

Significance: The developed multidimensional platform is the first system that allows for multiple fractions from a single ProtA band to be characterized using different chromatographic selectivities in a single run allowing direct correlation between CQAs. The performance of the system is comparable to established off-line methods, fully compatible with upstream process samples, and provides a significant time-reduction of the characterization procedure.