Comprehensive Glycoproteomic Analysis of Chinese Hamster Ovary Cells

Abstract

The Chinese hamster ovary (CHO) cell line is a major expression system for the production of therapeutic proteins, the majority of which are glycoproteins, such as antibodies and erythropoietin (EPO). The characterization glycosylation profile of therapeutic proteins produced from engineered CHO cells and therapeutic functions, as well as side effects, are critical to understand the important roles of glycosylation. In this study, a large scale glycoproteomic workflow was established and applied to CHO-K1 cells expressing EPO. The workflow includes enrichment of intact glycopeptides from CHO-K1 cell lysate and medium using hydrophilic enrichment, fractionation of the obtained intact glycopeptides (IGPs) by basic reversed phase liquid chromatography (bRPLC), analyzing the glycopeptides using LC-MS/MS, and annotating the results by GPQuest 2.0. A total of 10 338 N-linked glycosite-containing IGPs were identified, representing 1162 unique glycosites in 530 glycoproteins, including 71 unique atypical N-linked IGPs on 18 atypical N-glycosylation sequons with an overrepresentation of the N-X-C motifs. Moreover, we compared the glycoproteins from CHO cell lysate with those from medium using the in-depth N-linked glycoproteome data. The obtained large scale glycoproteomic data from intact N-linked glycopeptides in this study is complementary to the genomic, proteomic, and N-linked glycomic data previously reported for CHO cells. Our method has the potential to monitor the production of recombinant therapeutic glycoproteins.

Figure 1.

Workflows of intact glycopeptide analysis strategies for lysate and medium from human EPO-expressing CHO-K1 cells. (a) The workflow of large scale deglycosylated peptide analysis enriched by MAX extraction cartridges. The proteins were digested, fractionated and MAX enriched. The glycopeptides were then deglycosylated and analyzed by LC-MS/MS. (b) Different glycopeptide analysis strategies using MAX cartridge enrichment followed by fractionation or fractionation of global peptides, followed by enrichment of IGPs using MAX cartridges for large-scale IGPs analysis.

Figure 2.

Depth of the identified IGPs in cell lysate and medium from human EPO-expressing CHO-K1 cells. (a) Identification and distribution of total and subtype IGPs identified from cells and medium. (b) Distribution of identified glycosites and glycoproteins in cell lysate and medium, showing that most glycoproteins are present in CHO-K1 cell lysate and medium. (c) Identification and distribution of glycans and the composition and distribution of fucosylated and sialylated N-linked glycans in cell lysate and medium.

Figure 3.

Heterogeneity of detected glycoproteins in CHO cell lysate and medium. (a) Distribution of glycosites per protein. (b) Distribution of glycans per glycosite. (c) Heat map of the differences in abundance of the subtypes of N-linked glycans between CHO cell lysate and medium on glycopeptide VPFIFNINPSTTN^#FTGSCHPQTAQLR and THLTCSLN^#SSGVDIIGHR. ^# indicates an N-linked glycosite.

Figure 4.

Identification of atypical N-linked glycopeptides and preference of N-glycosylation peptide consensus sequence. (a) Representative MS/MS spectra of the atypical N-glycopeptides of TYN^#IVLGSGQVVLGMDMGLR + N2H6F0S0G0 from peptidyl-prolyl cis–trans isomerase FKBP9 and LCN^#ECSEGSFHLSK + N2H6F0S0G0 from basement membrane-specific heparin sulfate proteoglycan core protein. ^# indicates an N-linked glycosite. (b) Distribution and preference of typical and atypical glycosite consensus sequence derived using pLogo.

Figure 5.

Relative abundance analysis of N-linked intact glycopeptides (IGPs). (a) Relative abundance profiling of IGPs between CHO cell lysate and medium, showing a strong correlation between the IGP abundance in cell lysate and medium. (b) The N-glycosylation of pro-low-density lipoprotein receptor-related protein 1 in cell lysate and medium, including the structure and relative abundance (PSM) of glycans on each glycosite. The inner circle shows the relative abundance in CHO cell lysate and the outer circle shows the relative abundance in medium.