Many proteins are expressed recombinantly in mammalian cell lines such as Chinese hamster ovary cells (CHO) or Human Embryonic Kidney 293 cells (HEK 293). Production of proteins in these cell systems has the advantage of producing mammalian type glycosylation. However, the pattern of glycosylation may not be fully human. For example the Gal α1-3 Gal epitope and the N-glycolylneuraminic acid (NGNA) have both been found on glycoproteins produced by CHO cells but are not found in normal human glycans (1). Both of these modifications can be immunogenic in humans. Some mammalian cell lines have been mutated to produce humanized glycoproteins. However challenges such as lowered viability and reversion can be encountered with these modifications of the culture cell lines.
Insect cell lines using a Baculovirus expression system are also often used to produce glycoproteins. The advantage of this system is that heterologous genes are well expressed and folded. Post-translational modification and oligomerization are often identical to that which occurs in mammalian cells. However, issues of non-cognate glycosylation can also be a problem with this production system (2). Glycoproteins produced in insect cells can have N-linked glycans with simpler side chains called paucimannose and they are often modified at the GlcNAc core by an α1-3 Fucose (Fig. 1). This core α1-3 fucosylation is know to be allergenic in humans and can cause difficulties in glycan analysis of the produced glycoprotein since this structure is resistant to PNGase F (NEB #P0704, NEB #P0705), the most commonly used enzyme for N-glycan removal. Research is underway to engineer insect N-glycan pathways to make them able to produce glycans with human-like modifications (1).
Yeast such as Saccharomyces cerevisiae, Pichia pastoris and Kluyveromyces lactis have also been used to produce glycoproteins. The advantages to these systems are that they are relatively easy to grow in shake flasks or fermentors and their biosynthetic pathways resemble higher eukaryotic cells. They are however limited in the production of human-like glycoproteins by their inability to produce complex N-linked glycans (Fig. 1). In addition, both S. cerevisiae and K. lactis produce hypermannosylated N-linked glycans with as many as several hundred mannose residues being attached to the chitobiose core (Fig. 1). A great deal of effort has been put into altering the glycosylation pathways in P. pastoris to produce some Pichia strains that have nearly human N-glycosylation (3).
Several plant based expression systems have been developed. They can produce high yields of proteins and have been reported to express a variety of different proteins including membrane proteins, multiple sub-unit domains as well as monoclonal antibodies. Plants typically modify their N-linked glycans with the same α1-3 Fucose as seen on some insect produced proteins as well as having a xylose residue on the β-mannose of the glycan core (Fig. 1). However work is ongoing to modify the glycosylation pathways to avoid these nonhuman modifications (4).
In addition to problems with nonhuman glycan modifications, glycosylation patterns of recombinant proteins are greatly affected by conditions such as culture medium, pH, temperature and cell concentration. These variations can often affect the biological properties of biopharmaceuticals and so careful monitoring of the glycans on these proteins during the production process is essential. Even with these challenges there is a large and growing population of recombinant glycoproteins being produced for the biopharmaceutical market. Erythropoietin, monoclonal antibodies and Interferon-β are but a few examples of recombinant glycoproteins produced to improve human health. These proteins are produced in a variety of expression systems; it is likely that no single system will be optimal for the production of all biotherapeutics that will emerge in the future.
- Galili U., et al (1988) J Biol Chem. 25;263(33):17755-62. PMID: 2460463
- Shi, X., Jarvis, D.L. (2007) Current Drug Targets. 8(10):1116-25. PMID: 17979671
- Hamilton, S.R., Gerngross, T.U. (2007) Curr Opin Biotechnol. 18(5):387-92. PMID: 17951046
- Gomord, V., et al (2010) Plant Biotech J. 8(5):564-87. PMID: 20233335