关键词: 超高效液相色谱, 高分辨质谱, 糖基化, 糖苷酶, 西妥昔单抗

Abstract:

The N-glycosylation of proteins is a typical post-translational modification. Compared with other monoclonal antibodies, N-glycosylation modification in cetuximab is more complicated. Because cetuximab contains two N-glycosylation sites, one is located on the antigen-binding fragment (Fab) and the other is on the crystallizable fragment (Fc) of the heavy chain (HC). Among the two, the glycosylation of the Fab segment is more complicated. As this segment is located in the hypervariable region (VH), it may affect the affinity of the antibody antigen and cause other issues. Therefore, it is necessary to study glycosylation modification at this site. This modification is particularly challenging, necessitating the development of specific glycan cutting technology and a stable glycan ratio analysis method.
In this study, cetuximab expressed in Chinese hamster ovary (CHO) cell was used as the experimental research object. Based on the digestion with endo-β-N-acetylglucosaminidase F2 (Endo F2), an experimental method was developed that can quickly release Fab glycans. Qualitative and glycan ratio analyses were carried out by ultra performance liquid chromatography-high resolution mass spectrometry (UPLC-HRMS). The test was divided into two steps: in the first step, a non-denaturing (native state) glycosidase excision test was performed on the CHO-cetuximab drug substance. The drug substance was diluted to 1.0 mg/mL by adding ultrapure water, following which 1.0 μL of Endo F2 was directly added to 100 μL of the drug substance for enzyme digestion at 37 ℃. Through HRMS, the data were deconvoluted to obtain the accurate mass of the drug substance. The results showed that when the digestion time of Endo F2 was 5 min, the glycans in the Fab segment could be completely removed, whereas those in the Fc segment were not affected. Rapid enzyme cutting of the Fab glycans was realized; simultaneously, it was concluded that this method was also very specific for the removal of Fab glycans. In the second step, an accurate ratio analysis test was performed on Fab glycans excised from CHO-cetuximab. The released Fab glycans were precipitated with ice ethanol, the supernatant was centrifuged and spin-dried, and then labeled with para-aminobenzyl (2-AB). 2-AB labeling could make glycans have fluorescent detectable signals, and after reconstitution in 70% acetonitrile aqueous solution, was detected by UPLC coupled with a fluorescence detector (FLR). Good chromatographic peak separation was obtained using a hydrophilic interaction chromatography (HILIC) column. Thus, the test enabled stable glycan ratio analysis.
The molecular weight results for three independent Endo F2 digestion cycles for 5 min showed that the masses after digestion were similar; subsequently, glycan ratio analysis was performed based on HILIC. The results of three independent glycan ratio analysis experiments were also similar, indicating that the rapid enzyme digestion of Endo F2 followed by glycan ratio analysis after 5 min of digestion yielded good stability and reliability. Data obtained by measuring the samples produced using two different processes employed by our company showed that there were distinct differences in the glycan profiles of the two processes, especially in terms of the sialic acid glycoforms. These results prove that the method developed in this study can accurately analyze the ratio of glycans. Monitoring the antibody production process is important and meaningful for the evaluation of the process.

Key words: ultra performance liquid chromatography (UPLC), high resolution mass spectrometry (HRMS), glycosylation, glycosidase, cetuximab