Analysis of lysine clipping of a humanized Lewis-Y specific IgG antibody and its relation to Fc-mediated effector function.

During the analytical characterization of the humanized Lewis-Y specific monoclonal antibody IGN311 (IgG1/kappa) used for passive anti-cancer therapy in humans, isoelectric focusing (IEF) experiments revealed that IGN311 batches produced in serum-containing and serum-free medium, respectively, displayed different banding patterns. The additional bands in the IEF pattern correlated with additional peaks observed by subsequent cation exchange (CEX)-HPLC analysis. Since the IEF pattern is one of the specification criteria in the quality control of monoclonal antibodies and a non-matching pattern may be indicative for lot-to-lot inconsistency, this phenomenon was investigated in detail. First, we investigated whether a difference in antibody glycosylation was the cause for the observed charge heterogeneity. De-N-glycosylation experiments demonstrated that charge heterogeneity observed in the IEF pattern is not a consequence of glycosylation. In contrast, sample treatment by carboxypeptidase B, removing the carboxy-terminal lysine residues from the two heavy chains of the antibody, resulted in reduced charge heterogeneity eliminating the two most basic bands observed in IEF. These data were supported by reversed phase HPLC-MALDI-TOF-MS analysis of enzymatically cleaved peptides of the antibody as well as by carboxy-terminal sequencing of the heavy chains. It was demonstrated that the differences in the IEF banding pattern were due to lysine clipping occurring during the production of the antibody. The antibody batch produced under serum-free conditions was less affected by lysine clipping. Both antibody variants--clipped and unclipped--elicited the same potency in a complement dependent cytotoxicity (CDC) assay demonstrating that lysine clipping of IGN311 does not impair Fc-mediated effector functions.