The glycan moiety of human recombinant gonadotrophins (r-hFSH, r-hLH, and r-hCG) produced in CHO cell lines has been characterized by a combination of chromatographic and mass spectrometric techniques, including both matrix-assisted laser desorption ionization and electrospray. Two glycan mapping methods have been developed for the three gonadotrophins that allow separation of the glycans according to either their charge or sialylation level or their antennarity. A method was also developed for r-hCG that permits the complete resolution of the N-glycan from the O-glycan species. Whereas the structure found for the N-glycans of the gonadotrophins was in agreement with the complex type model, the structure for an O-glycan of r-hCG, not yet described, has been unambiguously determined using nanoelectrospray ion trap mass spectrometry. Using these two glycan mapping methods, the high level of batch-to-batch consistency achieved for the glycosylation of the three recombinant gonadotrophins in commercial production has been shown. These data demonstrate the tight control that can be achieved in the manufacturing of complex recombinant therapeutic glycoproteins, which is a prerequisite to the delivering of a guaranteed dose of drug from vial to vial, and in turn to ensuring the clinical efficacy of the product.

No abstract

Host cell proteins (HCP) are a major class of impurities derived from recombinant protein production processes. While HCP are usually monitored by ELISA, mass spectrometry (MS)-based approaches are emerging as promising orthogonal methods. Here, we developed an original method relying on data-independent acquisition (DIA) coupling global HCP amount estimation (Top 3) and absolute quantification with isotope dilution (ID). The method named Top 3-ID-DIA was benchmarked against ELISA and a gold-standard selected reaction monitoring assay (ID-SRM). Various samples generated at different steps and conditions of the purification process, including different culture durations, harvest procedures, and purification protocols were used to compare the methods. Overall, HCP were quantified over 5 orders of magnitude and down to the sub-ppm level. The Top 3-ID-DIA strategy proved to be equivalent to the gold-standard ID-SRM in terms of sensitivity (1-10 ppm), accuracy, and precision. Moreover, 81% of the Top 3 estimations were accurate within a factor of 2 when compared to ID-SRM. Thus, our approach aggregates global HCP profiling for comprehensive process understanding with absolute quantification of key HCP within a single analysis and provides an improved support for bioprocess development and product purity assessment.

No abstract

No abstract