In-Membrane Enrichment and Peptic Digestion to Facilitate Analysis of Monoclonal Antibody Glycosylation

Abstract

The number of therapeutic monoclonal antibodies (mAbs) is growing rapidly due to their widespread use for treating various diseases and health conditions. Assessing the glycosylation profile of mAbs during production is essential to ensuring their safety and efficacy. This research aims to rapidly isolate and digest mAbs for liquid chromatography-tandem mass spectrometry (LC-MS/MS) identification of glycans and monitoring of glycosylation patterns, potentially during manufacturing. Immobilization of an Fc region-specific ligand, oFc20, in a porous membrane enables the enrichment of mAbs from cell culture supernatant and efficient elution with an acidic solution. Subsequent digestion of the mAb eluate occurred in a pepsin-modified membrane within 5 min. The procedure does not require alkylation and desalting, greatly shortening the sample preparation time. Subsequent LC-MS/MS analysis identified 11 major mAb N-glycan proteoforms and assessed the relative peak areas of the glycosylated peptides. This approach is suitable for the glycosylation profiling of various human IgG mAbs, including biosimilars and different IgG subclasses. The total time required for this workflow is less than 2 h, whereas the conventional enzymatic release and labeling of glycans can take much longer. Thus, the integrated membranes are suitable for facilitating the analysis of mAb glycosylation patterns.

Figure 1.

Scheme of the experimental workflow for analyzing the glycan composition of mAbs in cell culture supernatant. The procedure includes selective enrichment of mAb from cell culture supernatant, elution of captured mAb, reduction and denaturation of eluted mAb, in-membrane digestion of eluate, drying (not shown) and analysis of peptides using LC-MS/MS.

Figure 2.

Comparison of Trastuzumab binding and elution in Protein A- or oFc20-modified membranes. (A) Binding capacity of Protein A- or oFc20-modified membranes. The loading solution was 7 mL of 0.1 mg/mL Trastuzumab in Buffer C. (B, C) Efficiency of the elution of captured Trastuzumab in (B) Protein A- or (C) oFc20-modified membranes using different eluents. Error bars show the standard deviations from experiments with three different membranes.

Figure 3.

Glycan relative MS1 peak areas for Trastuzumab digested with different proteases. The white columns are results of a Trastuzumab standard (in DI water) digested through trypsin in-membrane proteolysis. We observed two tryptic glycopeptides EEQYNSTYR and TKPREEQYNSTYR with 3+ and 4+ charge states. Grey columns represent Trastuzumab in the first eluate aliquot from an oFc20 membrane and processing with pepsin in-membrane digestion. The error bars are standard deviations of analyses of three digestion replicates either from the same dialyzed Trastuzumab standard in DI water or the first eluate aliquots from three different membranes. One and two *’s denote a significant difference at p < 0.05 or 0.01, respectively. The experimental section defines the relative peak area.

Figure 4.

Glycan compositions of Trastuzumab from three different manufacturing batches. Batch 2 is the Trastuzumab used for all Trastuzumab-related experiments here unless otherwise noted. * is p < 0.05, whereas ** is p < 0.01 in an unpaired T test. Error bars represent the standard deviations of analyses of the first eluate aliquot from three different membranes. The experimental section defines the relative peak area.

Figure 5.

Glycan compositions of three different subclasses of human IgGs. Trastuzumab is IgG1, whereas Panitumumab is IgG2 and Nivolumab is IgG4. The Trastuzumab result is the same as Batch 2 in Figure 4. Error bars represent the standard deviations of analyses of the first eluate aliquots from three different membranes. *** is p < 0.001 in unpaired T test. The experimental section defines the relative peak area.

Figure 6.

Glycan composition of Ab3022, an in-house expressed human IgG1 antibody from ExpiCHO-S cells. The cells were cultured 7 days before harvest. The supernatant was clarified, and filter sterilized prior to analysis. Error bars are the standard deviations of measurements of replicate first eluate aliquots from three different membranes. The experimental section defines the relative peak area.