Mar 11, 2014 – MSDG Meeting

Abstract for Talk II:

Therapeutic proteins are characterized methodically with respect to primary structure, posttranslational modifications, charge and size heterogeneity, higher order structure, and biological activity, in concert with process and product development, to achieve a comprehensive understanding of the active substance prior to commercialization. The ESI-quadrupole-time-of-flight (QTOF) mass spectrometer is well-established in the heightened characterization of protein therapeutics, and is highly suited for antibodies, antibody-drug conjugates, fusion proteins, novel constructs, and vaccines. This single instrument platform provides detailed structural information for the intact protein molecule, as well as for the subunits/domains, proteolytic peptides, and released N-glycans (if present) in orthogonal analyses. The newest generation of ultrahigh-resolution (UHR) ESI-QTOF mass spectrometers offers significant improvements in critical performance parameters such as resolution, mass accuracy, sensitivity and dynamic range. These instruments, in conjunction with new approaches to protein characterization, provide accurate measurements to within 2 ppm of theoretical values for isotopically-resolved proteins ≤35 kDa, which is ideal for the detailed analysis of whole antibody subunits/domains and smaller single-chain proteins.

In this presentation, a comparison of newer LC/MS-subunit analysis methods for antibody characterization will be discussed that improve on traditional reduction/alkylation, and light and heavy chain subunit mass analysis. Through hinge region-proteolysis and disulfide bond reduction, heavy chain (~50 kDa) essentially can be cleaved homogeneously into two domain fragments (~25 kDa each) with molecular masses more similar to light chain (~24 kDa). Since UHR QTOF MS excels in this 10-35 kDa mass range, these newer approaches to subunit analysis afford more sensitive detection and definitive identification of both expected and unexpected antibody modifications at the whole subunit/domain level, all in a rapid manner – in contrast to traditional, data-intensive proteolytic mapping at the peptide level. In addition, rapid confirmation of the amino acid sequence at the subunit/domain level and detection of potential sequence variants and mis-incorporations during the early stages of process development will be demonstrated with these new methods. Finally, current endeavors regarding automated data analysis for all LC/MS methods will be presented.