Expression of recombinant human coagulation factors VII (rFVII) and IX (rFIX) in various cell types, glycosylation analysis, and pharmacokinetic comparison
© Böhm et al; licensee BioMed Central Ltd. 2011
Published: 22 November 2011
Clearance mechanisms for rFVII (or the active enzyme rFVIIa) and rFIX are influenced by post-translational modifications, especially N-glycosylation. This should be considered when choosing a recombinant expression system in view of the varying ability of frequently used cell lines to perform modifications similar to human proteins.
Differences in the pharmacokinetic properties of recombinant FVIIa versus plasma-derived (pd)FVII or desialylated rFVIIa are known for human FVII(a). Asialo rFVIIa clears quickest, whereas pdFVII, having a higher degree of sialylation, is cleared to a lesser extent . In the case of FIX, the degrees of serine phosphorylation and tyrosine sulfation in the activation peptide have been postulated to influence pharmacokinetic behavior, especially in vivo recovery .
We chose CHO, BHK and HEK293 cells for expression of rFVII to compare post-translational protein modifications, and HEK293-derived cell lines to generate highly phosphorylated and sulfated rFIX for in vivo studies. rFIX from the same clone and production run was purified using two different down-stream processes: The first to enrich high phosphorylated and sulfated protein, the second to purify total rFIX at high yield. These HEK293-derived rFIX isoforms were compared with CHOrFIX and pdFIX in a pharmacokinetic study in FIX knock-out mice.
Materials and methods
pdFVII and pdFIX were from HTI, recombinant, CHO-derived FIX was from Wyeth.
BHK (BHK-21; ATCC#CCL-10™) and HEK293 cells (ATCC#1531) were from American Type Culture Collection, CHO DXB11 from University of Columbia. All were cultivated in DMEM/Ham’s F12 medium containing fetal bovine serum (FBS). BHK and HEK293 cell-derived rFVII producer clones were selected by antibiotic resistance. CHO DXB11-derived producer clones were generated by methotrexate gene co-amplification. rFVII was produced in vitamin K-containing medium without FBS.
HEK293 cells producing human FIX were selected by antibiotic resistance. Clones were adapted to serum-free suspension culture in Excell293 medium containing vitamin K, and cultivated in repeated batch mode fermentation runs.
rFVII was purified using Q-Sepharose FF for capture, and a tandem step containing a cellufine sulfate column connected to Q-Sepharose FF.
HEK293-derived rFIX from the same clone from one fermentation run was purified from culture supernatants with two different purification schemas: for high yield, rFIX was loaded in the presence of EDTA on Q-Sepharose FF, the column was washed with EDTA at high salt concentration, FIX was eluted at low salt CaCl2. For enrichment of the highly phosphorylated and sulfated protein fraction, rFIX was loaded in the presence of EDTA on fractogel TMAE, washed with EDTA at high salt concentration, and eluted with a salt gradient in the presence of CaCl2.
Purified FVII forms were analyzed using reversed phase HPLC, LC-MS for intact protein or peptide mapping after trypsin digestion to confirm sequences and post-translational modifications, and anion exchange HPLC for monosaccharide composition, sialic acid quantification, and comparison of N-glycans.
Degrees of FIX-phosphorylation/sulfation were monitored by LC-MS. N-glycosylation was characterized by oligosaccharide mapping (HPLC method).
In vivo pharmacokinetic study
FIX preparations were administered to FIX-knock-out mice intravenpusly at a nominal dose of 75 U/kg. Citrate plasmas were obtained by heart puncture. FIX levels in plasmas were determined by antigen enzyme-linked immunoassay and by activated partial thromboplastine time (APTT) clotting assay.
Results and conclusions
Identical protein structures and minor differences in post-translational modifications between rFVII from BHK, CHO, HEK293 cells, and pdFVII were found, except for N-glycosylation. Protein activities of rFVII determined by antigen and activity assays were similar for each cell type. Most N-glycans of pdFVII are bi- and triantennary, non-fucosylated structures with almost complete sialylation . In contrast, all rFVII forms were sialylated only in part. Approximately 50 % N-glycans of BHKrFVII were composed of a core-fucosylated biantennary complex-type with two terminal sialic acids. Another 20 % contained terminal GalNAc, a sugar moiety with a high affinity for the asialoglycoprotein receptor involved in the in vivo clearance of glycoproteins from circulation. GalNAc were not found on N-glycans on CHO-derived rFVII. Most oligosaccharides on CHOrFVII were core fucosylated, biantennary structures carrying two sialic acid residues. HEK293rFVII showed major differences from all other rFVII forms, and, unexpectedly, from pdFVII. Most oligosaccharide structures were not comparable to those published for pdFVII , or found on rFVII from the other cell lines. A high content of fucosylation and terminal GalNAc, and a lower degree of terminal sialic acids were measured, and tri- or tetra-antennary structures were not found. Some oligosaccharides were of a hybrid type with high-mannose structures. These N-glycan structures do not favor the use of HEK293 cells as a production platform for human biotherapeutics, when protein recovery and half-life in the circulation are of importance, and influenced by N-glycosylation.
AUCs of pdFIX, CHOrFIX, HEK293rFIX, and high-phospho/sulfo HEK293rFIX in a FIX-knock-out mouse pharmacokinetic study. Only relative AUC values are shown. Statistically significant different ratios to CHOrFIX (at the 5 % level) are marked with an asterisk*. The ratio of HEK293rFIX high phospho/sulfo to HEK293- total rFIX was also statistically significantly different for both antigen and activity.
Relative value of AUC: antigen
Relative value of AUC: clotting activity
HEK293rFIX high phospho/sulfo
In conclusion, these data showed that HEK293 cells were not adequate for rFIX or rFVII production due to improper N-glycosylation compared with the material from other cell lines, or from human plasma. Consequently, an advantage of this human cell line for the production of “more-human-like” biotherapeutics could not be observed.
- Appa RS, Theill C, Hansen L, Møss J, Behrens C, Nicolaisen EM, Klausen NK, Christensen MS: Investigating clearance mechanisms for recombinant activated factor VII in a perfused liver model. Thromb Haemost. 2010, 104 (2): 243-251. 10.1160/TH09-10-0723.View ArticlePubMedGoogle Scholar
- Ewenstein BM, Joist JH, Shapiro AD, Hofstra TC, Leissinger CA, Seremetis SV, Broder M, Mueller-Velten G, Schwartz BA, Mononine Comparison Study Group: Pharmacokinetic analysis of plasma-derived and recombinant F IX concentrates in previously treated patients with moderate or severe hemophilia B. Transfusion. 2002, 42 (2): 190-197. 10.1046/j.1537-2995.2002.00039.x.View ArticlePubMedGoogle Scholar
- Fenaille F, Groseil C, Ramon C, Riandé S, Siret L, Chtourou S, Bihoreau N: Mass spectrometric characterization of N- and O-glycans of plasma-derived coagulation factor VII. Glycoconj J. 2008, 25 (9): 827-842. 10.1007/s10719-008-9143-7.View ArticlePubMedGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.