ABSTRACT
Antibody-drug conjugates (ADCs) are a heterogeneous mixture of conjugated species with varied drug loadings. Depending on conjugation sites, linkers and drugs can exhibit different stability as influenced by the solvent-accessibility and local charge, resulting in different ADC efficacy, pharmacokinetics, and toxicity. Conjugation site analysis is critical for ADC structural characterization to assure product quality and consistency. It enables early conjugation studies at site-specific levels, confirms the absence of unexpected products to support conjugation process development, and aids in ensuring lot-to-lot consistency for comparability studies. Peptide mapping using liquid chromatography-tandem mass spectrometry is the industry standard method for analyzing conjugation sites. However, some concerns remain for this approach as the large and hydrophobic drug moieties often result in poor MS/MS fragmentation quality and impede the identification of conjugation sites. Additionally, the ionization discrepancy between conjugated and unconjugated peptides can lead to a relatively large bias for site occupancy calculation. In this work, we present a simple drug deconjugation-assisted peptide mapping method to identify and quantify the drug conjugation for ADCs with protease-cleavable linkers. Papain-based drug deconjugation was used to remove the highly hydrophobic drug moiety, which significantly improved the quantitation accuracy of conjugation level and the fragmentation quality. Sample preparation conditions were optimized to avoid introducing artificial modifications, allowing the tracking of initial sample status and subsequent changes of quality attributes during process development and stability assessment. This method was applied to analyze thermally-stressed ADC samples to monitor changes of site-specific conjugation levels, DAR, succinimide hydrolysis of the linker, and various PTMs. We believe this is an effective and straightforward tool for conjugation site analysis while simultaneously monitoring multiple quality attributes for ADCs with protease-cleavable linkers.
