Literature Review: CHO versus HEK Cell Glycosylation

FLOW TRANSFEC TION SY STEMS™
™
Literature Review: CHO versus HEK Cell Glycosylation
Contributing Authors: Krista Steger, Ph.D., James Brady, Ph.D., Meg Duskin, Karen Donato, Ph.D.
Abstract
Choosing a host system for the expression of recombinant proteins should be carefully evaluated prior to the initiation of any
biotherapetuic development programs. Since different hosts express proteins with different efficiencies and with different posttranslational modifications, changing hosts during the course of development may impact the expected activity of the protein
and may require repeated studies or additional process development, delaying the time to market. Scientists must also consider
the expression system’s productivity, consistency, and the current regulatory environment. This paper includes a commentary on
maintaining the same host system throughout the development process, lays out a brief review of key references examining the
differences in glycosylation in HEK versus CHO cells, and provides a more in-depth list of related literature.
Introduction
One of the most common forms of post-translational
modifications is glycosylation. Half of all human proteins,
and 40% of all approved therapeutic proteins, are estimated
to be glycoproteins.1,2 Glycosylation is a complex, enzymedirected process that links oligosaccharides to specific
amino acid sites. Each potential glycosylation site may be
occupied or unoccupied (macroheterogeneity) and, if
occupied, may contain one of a variety of glycan structures
(microheterogeneity). Thus, glycosylation produces a mixed
array of different protein glycoforms rather than a single
homogeneously glycosylated protein.
Glycosylation impacts a large number of protein attributes
ranging from folding, stability, solubility, and proteinprotein interactions to more physiological properties such
as protein activity, in vivo bioavailability, biodistribution,
pharmacokinetics, and immunogenicity.3-9 Accordingly,
changes to the pattern of glycosylation can alter key
therapeutic properties of a candidate molecule such as its
efficacy, safety, half-life, and manufacturability.10-14 This
suggests that data generated during preclinical development
may not be indicative of the final therapeutic product if
glycosylation is not considered. This in turn may translate
to the progression of irrelevant candidates or overlooking
promising candidates, both of which negatively impact the
identification of the right drug candidate.
Antibodies, the most prevalent category of biotherapeutic
under development, are glycosylated on the Fc region and to
a lesser degree on the Fab region.15,16 Glycosylation is known
to be a significant source of monoclonal antibody (MAb)
heterogeneity. Numerous studies link this heterogeneity
www.maxcyte.com
to changes in the biological, pharmacological, and
physiochemical properties of antibodies.17-20
Influences on Glycosylation Patterns
A variety of factors are known to influence protein
glycosylation including the protein amino acid sequence
itself, the host cell, and cellular growth.21-25 Additionally,
changes to cell culture conditions ranging from culture
stirring speed to nutrient availability can impact protein
glycosylation.16
With regard to host cell-specific effects on glycosylation, it
is generally accepted that bacteria, yeast, and mammalian
cells differ in the level and type of post-translational
modifications expressed. Unfortunately, many researchers
assume all mammalian cells produce proteins with similar
human-like post-translational modifications. In a key study,
species-specific differences in glycosylation were observed
when a MAb was expressed in cells from 13 different species,
including humans. 26 All 13 species expressed the protein of
interest with a heterogeneous array of biantennary complex
type oligosaccharides. Variations in core fucosylation
and terminal galactosylation between species were noted
in addition to significant variations in oligosaccharides
sialylation.
A large number of mammalian cells derived from different
species and different tissues have been successfully used
to express recombinant proteins for clinical applications,
including myeloma (NS0 and SP2/0) cells, Chinese hamster
ovary (CHO) cells, human retina-derived cells (PerC6) cells,
human embryonic kidney (HEK) cells, and baby hamster
kidney (BHK) cells.15,21 CHO cells have well established
info@maxcyte.com
advantages for biomanufacturing including the capacity for
high-level production of heterologous proteins; the ability
to be cultured at high cell densities in chemically defined,
protein-free media; and a strong regulatory track record.
They continue to be the dominant system of choice for
biomanufacturing of clinical-grade biotherapeutics. As of
2012, 70% of all recombinant therapeutics are produced
using CHO cell lines.31
CHO versus HEK Therapeutic Protein
Production
Initial efforts using CHO-based transient production of
proteins were limited by poor transfection efficiencies,
viabilities, and production of insufficient quantities of
recombinant proteins. This has led to the use of HEKbased transient systems, which historically have higher
transfection efficiencies and protein production capabilities.
Multiple studies, however, have reported differences in the
glycosylation patterns of proteins and antibodies when
produced in CHO versus HEK cells.32-38
Five key studies have demonstrated that CHO cells
produce proteins with higher molecular weights that can be
attributed directly to differences in glycosylation (both site
occupancy and associated glycostructures) and consistently
contain higher sialic acid contents. This suggests that
preclinical development in HEK cells increases the risk of
late-stage developmental failures due to potential alterations
in biophysical properties upon transition to CHO-based
stable expression. Thus, if CHO cells are the intended
means of biomanufacturing, the most relevant candidates
will be identified using CHO cell-based protein production
during early development.
Future of CHO Cell Protein Production
CHO cells lack several glycosylation enzymes present
in human cells. Does this mean CHO cells produce
therapeutics with suboptimal efficacy? To date, there is
not a consensus on the ‘ideal’ pattern of glycosylation
for optimal in vivo efficacy. Research has linked specific
glycopatterns to therapeutic characteristics of antibodies
including in vivo effector functions such as more effective
antibody-dependent cell-mediated cytotoxicity (ADCC)
or reduced complement-dependent cellular cytotoxicity
(CDC).39-41 While some of the identified glycopatterns may
be universally beneficial, for example increased protein halflife due to increased sialylation, the optimal glycosylation
state will most likely be protein-dependent, as each
therapeutic protein exhibits different mechanisms of action
www.maxcyte.com
and bioavailability. Consequently, care must be taken during
candidate identification and lead optimization not only to
assess protein glycosylation, but also to systematically
examine the effects of specific glycosylation patterns on
therapeutic efficacy.
The CHO genome project has vastly increased our
understanding of CHO glycosylation pathways. Researchers
have applied this knowledge to genetically engineer CHO
cells to inhibit and/or express specific genes involved in
glycosylation, enabling production of proteins with the
desired glycopattern.42,43 Advances in glycoengineering
such as these, and the strong regulatory track record of
CHO usage, support the continued use of CHO cells for
the biomanufacturing of therapeutic proteins. Given the
evidence that the type(s) and pattern of glycosylation are
cell-type dependent and the pressures biopharmaceutical
companies are under to reduce late-stage attrition rates,
scientists must carefully consider the implications of
switching host cell backgrounds during development.
Key Reference Review
Five recent studies examining the glycosylation of proteins
expressed side-by-side in CHO and HEK cells are
summarized below. While the proteins expressed and the
specific type of glycoanalysis differed in each paper, the
authors of each article concluded that protein glycosylation
was host cell-type dependent.
Differences in the Glycosylation of Recombinant
Proteins Expressed in HEK and CHO Cells. J. Biotech.
2012, 161: 336-348.
The authors of this paper concluded that proteins expressed
in CHO and HEK cells had significant differences in
glycostructures, including a consistently increased number
acidic isoforms in CHO cells due to a higher level of
sialylation.
Glycoanalysis of 12 different proteins transiently expressed
in CHO-S, HEK293 EBNA, and HEK 6E cell lines was
conducted. The expressed proteins differed in size (9.552 kDa), the number of potential N-linked and O-linked
glycosylation sites, and expression levels. Secreted proteins
were purified and extensively characterized using SDSPAGE, isoelectric focusing (IEF), mass spectrometry
(MS), and capillary gel electrophoresis of released N-linked
glycans.
For all proteins expressed, clear differences were detected
in both the size and the number of glycans as well as the
amount of sialic acid. The differences were so significant
that they were detectable as changes in molecular weight
observed by SDS-PAGE analysis. Changes to the molecular
weight could be attributed to altered glycosylation using
deglycosylation enzymes.
The most significant differences in glycosylation patterns
were seen between CHO and HEK cells. Differences
between two HEK cell lines (EBNA vs. 6E) as well as those
due to changes in the transfection procedure or culture
conditions were minor. Neither protein size nor the level of
protein expression correlated with specific glycopatterns.
Characterization of
Host-Cell Line Specific
Glycosylation Profiles of Early Transmitted/Founder
HIV-1 gp120 Envelope Proteins. J. Proteome Res. 2013,
12(3): 1223-1234.
This article reports the detailed glycoanalysis of HIV-1
envelope protein, gp120, transiently expressed in CHO and
HEK293T cells. The authors concluded that the general
degree of fucosylation and sialylation for gp120 expressed
in CHO cells was higher than compared to HEK-derived
gp120 and went on to detail site-specific glycan differences.
gp120 is highly glycosylated with 23 potential N-linked
glycosylation sites. High-resolution liquid chromatography/
mass spectrometry, electron transfer dissociation, and
collision-induced dissociation were used for site-specific
analysis of glycosylation. gp120 purified from both CHO
and HEK cells consisted of a broad spectrum of glycan
structures across all the potential N-glycosylation sites.
The glycan structures included high mannose, hybrid, and
complex glycans containing multi-antennary structures with
varying levels of core fucosylation and sialylation.
Occupancy of individual glycosylation sites showed a high
degree of similarity between CHO- and HEK-derived
gp120, but distinct differences in both O-linked and
N-linked glycans were identified. Specifically, two sites,
N386 and N392, in the V4 region of CHO cell-derived
gp120 were populated with high mannose glycans, while
they were a mixture of high mannose and processed glycans
in the HEK-derived gp120.
Cell Type-specific and Site Directed N-glycosylation
Pattern of FcΥRIIIa. J. Proteome Res. 2011, 10(7): 30313039.
The dominant factor contributing to differences in
FcΥRIIIa glycopatterns in these studies was the host cell
used. Previous studies determined that the carbohydrate
moieties at Asn-162 are important for binding of human
leukocyte receptor IIIa (hFcΥRIIIa) to the Fc region of
antibodies. This paper expressed hFcΥRIIIa in HEK
and CHO cells and performed site-directed carbohydrate
analysis via mass spectrometry and a multienzyme protein
digest. The reported studies found that the proteins had
similar sites of glycosylation but the glycostructures at a
number of sites were cell-type specific, including Asn-162.
Furthermore, the glycosylation pattern influenced antibody
binding when accessed via surface plasmon resonance.
Transient Expression of an IL-23R Extracellular
Domain Fc Protein in CHO vs. HEK Cells Results in
Improved Plasma Exposure. Protein Expr. Purif. 2010, 71:
96-102.
These studies transiently expressed the IL-23R extracellular
domain Fc fusion protein (IL23R-Fc) in both CHO-S and
HEK293 cells and examined the resulting glycosylation
patterns. Significant host cell-specific glycosylation was
reported and shown to alter in vivo pharmacokinetics.
IL23R-Fc contains 18 potential N-linked glycosylation
sites within a single dimer. Following expression, IL-23R
glycopatterns were examined via SDS-PAGE, IEF, and
lectin microarray binding. Additionally, the purified IL23RFc proteins were administered to mice and the in vivo
pharmacokinetics analyzed. Differences in the molecular
weights seen by SDS-PAGE were confirmed to be due
to differences in N-linked glycosylation. Additionally, the
IL23R-Fc produced by CHO cells exhibited a lower pI,
indicating an increased number of acidic groups. Sialidase
treatment and IEF analysis confirmed the decrease in pI
was due to differences in sialic acid content. Compared with
CHO-derived protein, that from HEK cells had increased
binding to lectins specific for terminal galactose (Gal) or
N-acetyl-glactosamine (GalNAc) residues. Overall, the
authors concluded that CHO-derived protein had a higher
total sialic acid content while HEK-derived protein had a
higher terminal Gal content.
Despite clear differences in glycosylation, CHO- and HEKderived proteins had comparable in vitro activity as measured
by binding to IL-23 and stimulation of IL-18 production
upon exposure to human peripheral blood mononuclear
cells (PBMCs). In vivo, however, CHO-derived IL23R-Fc
had approximately 30-fold higher plasma exposure and a
2-fold longer plasma half-life. The major difference was
the very high clearance of HEK-derived IL23R-Fc from
plasma within the first hour. This finding is consistent with
glycoprotein receptor-mediated clearance, as carbohydrates
lacking terminal sialic acid are potential substrates for
info@maxcyte.com
glycoprotein receptor binding.
High-level Protein Expression in Scalable CHO
Transient Transfection. Biotechnol. Bioeng. 2009, 103(3):
542-551.
This article reports the glycoanalysis of a monoclonal
antibody purified from three sources: 1. transiently
expressed in HEK293 EBNA cells, 2. transiently expressed
in CHO-K1SV cells, and 3. purified from a CHO-derived
stable cell line. The authors concluded that the glycan
distribution of HEK-derived antibody was dramatically
different than CHO-derived antibodies when examined via
reverse phase liquid coupled with electrospray time-of-flight
chromatography.
The distribution of major CHO glycoforms was comparable
between transiently and stably expressed antibodies. These
data support the use of transient expression as a means of
protein production during therapeutic development, even
for proteins that will ultimately be manufactured using
stable cell lines, if the cell background is maintained.
Further Reading
Relevant Review Articles
The Choice of Mammalian Cell Host and Possibilities for
Glycosylation Engineering. Current Opin. Biotechnol. 2014,
30: 107-112.
Application of Quality by Design Paradigm to the
Manufacture of Protein Therapeutics. Glycosylation. Dr.
Stefana (Ed.). 2012, DOI: 10.5772/50261. Available
from: http://www.intechopen.com/books/glycosylation/
application-of-quality-by-design-paradigm-to-themanufacture-of-protein-therapeutics
Towards the Implementation of Quality by Design to the
Production of Therapeutic Monoclonal Antibodies with
Desired Glycosylation Patterns. Biotechnol. Prog. 2010, 26:
1501-1527.
Glycosylation of Therapeutic Proteins in Different
Production Systems. Acta Paediatr. Suppl. 2007, 96(455): 1722.
Additional Comparison of CHO versus Human-Cell Line
Glycosylation
Erythropoietin Produced in a Human Cell Line (Dynepo)
Has Significant Differences in Glycosylation Compared
with Erythropoietins Produced in CHO Cells. Mol. Pharm.
2011, 8: 286-296.
www.maxcyte.com
Purification of the Extracellular Domain of the Membrane
Protein GlialCAM Expressed in HEK and CHO Cells and
Comparison of the Glycosylation. Protein Expr. Purif. 2008,
58(1): 94-102.
Recombinant Glycodelin Carrying the Same Type of Glycan
Structures as Contraceptive Glycodelin-A Can Be Produced
in Human Kidney 293 Cells But Not in Chinese Hamster
Ovary Cells. Eur. J. Biochem. 2000, 267(15): 4753-4762.
Protein Glycosylation in CHO Stable versus Transient
Expression
A High-yielding CHO Transient System: Co-expression
of Genes Encoding EBNA-1 and GS Enhances Transient
Protein Expression. Biotechnol. Prog. 2014, 30(1): 132-141.
Control of Culture Environment for Improved
Polyethylenimine-Mediated Transient Production of
Recombinant Monoclonal Antibodies by CHO Cells.
Biotechnol. Prog. 2006, 22(3): 753-762.
Scalable Transient Gene Expression in Chinese Hamster
Ovary Cells in Instrumented and Non-instrumented
Cultivation Systems. Biotechnol. Lett. 2007, 29(5): 703-711.
Rapid Protein Production Using CHO Stable Transfection
Pools. Biotechnol. Prog. 2010, 26(5): 1431-1437.
References Sited within Commentary
1. Biopharmaceutical Benchmarks 2010. Nature Biotechnol.
2010, 28(9): 917-924
2. Protein Glycosylation: New Challenges and
Opportunities. J. Organic Chem. 2005, 70: 4219-4225.
3. Role of Glycosylation in Structure and Stability of
Erythrina Corallodendron Lectin (EcorL): A Molecular
Dynamic Study. Protein Sci. 2011, 20: 465-481.
4. Glycosylation Influences on the Aggregation Propensity
of Therapeutic Monoclonal Antibodies. Biotechnol. J.
2011, 6: 38-44.
5. Glycosylation Increases the Thermostability of Human
Aquaporin 10. J. Biological Chem. 2011, 286: 3191531923.
6. In vivo Recognition of Mannosylated Proteins by
Hepatic Mannose Receptors and Mannan-Binding
Protein. Am. J. Physiol. Gastrointest. Liver Physiol. 2001,
280: G879-G889.
7. Glycosylation Regulates Prestin Cellullar Activity. J
Assoc. Res. Otolaryngol. 2010, 11: 39-51.
8. Effects of N-linked Glycosylation on the Creatine
Transporter. Biochem. J. 2006, 393: 459-469.
9. Alternative Glycosylation Modulates Function of IgG
and Other Proteins – Implications on Evolution and
Disease. Biochem. Biophys. Acta 2012, 1820: 1318-1326.
10. The Significance of Glycosylation Analysis in
Development of Biopharmaceutical. Biol. Pharm. Bull.
2009, 32(5): 796-800
11. Glycoengineering: the Effect of Glycosylation on the
Properties of Therapeutic Proteins. J. Pharm. Sci. 2005,
94: 1626-1635.
12. Pharmacological Significance of Glycosylation in
Therapeutic Proteins. Curr. Opin. Biotechnol. 2009, 20:
678-684.
13. Glycosylation-Modified Erythropoietin With Improved
Half-Life and Biological Activity. Int. J. Hematol. 2010,
91: 238-244.
14. Impact of Variable Domain Glycosylation on Antibody
Clearance: An LC/MS Characterization. Analytical
Biochem. 2006, 349(2): 197-207.
15. Application of Quality by Design Paradigm to the
Manufacture of Protein Therapeutics. Glycosylation.
Dr. Stefana (Ed.).
2012, DOI: 10.5772/50261.
Available from: http://www.intechopen.com/books/
glycosylation/application-of-quality-by-designparadigm-to-the-manufacture-of-protein-therapeutics
16. Towards the Implementation of Quality by Design to
the Production of Therapeutic Monoclonal Antibodies
with Desired Glycosylation Patterns. Biotechnol. Prog.
2010, 26: 1501-1527.
17. Structural analysis of Human IgG-Fc Glycoforms
Reveals a Correlation between Glycosylation and
Structural Integrity. J. Mol. Biol. 2003, 325: 979-989.
18. Control of Recombinant Monoclonal Antibody
Effector Functions by Fc N-glycan Remodeling In Vitro
Biotechnol. Prog. 2005, 21: 1644-1652.
19. Terminal Sugars of the Fc Glycans Influence Antibody
Effector Functions of IgGs. Curr. Opin. Immunol. 2008,
20: 471-478.
20. Glycosylation of a VH Residue of a Monoclonal
Antibody Against Alpha (1-6) Dextran Increases its
Affinity for Antigen. J. Exp. Med. 1988, 168: 10991109.
21. The Choice of Mammalian Cell Host and Possibilities
for Glycosylation Engineering. Curr. Opin. Biotechnol.
2014, 30: 107-112.
22. The Availability of Glucose to CHO Cells Affects the
Intracellular Lipid-Linked Oligosaccharide Distribution,
Site-Occupancy and the N-Glycosylation Profile of a
Monoclonal Antibody. J. Biotechnol. 2014, 170: 17-27.
23. Optimization of the Cellular Metabolism of
Glycosylation for Recombinant Proteins Produced by
Mammalian Cell Systems. Cytotechnology 2006, 50(1-3):
57-76.
24. Different Culture Methods Lead to Differences in
Glycosylation of a Murine IgG Monoclonal Antibody.
Biochem. J. 1992, 285: 839-845.
25. Glycosylation of Therapeutic Proteins in Different
Production Systems. Acta Paediatr. Suppl. 2007, 96(455):
17-22.
26. Species-specific Variation in Glycosylation of IgG:
Evidence for the Species-specific Sialylation and
Branch-specific Galactosylation and Importance for
Engineering Recombinant Glycoprotein Therapeutics.
Glycobiology. 2000, 10(5): 477-486.
27. Analysis of Site-Specific Glycosylation of Renal and
Hepatic Gamma-Glutamyl Transpeptidase from
Normal Human Tissue. J. Biolog. Chem. 2010, 285:
29511-29524.
28. Glycosylation of Serum Proteins in Inflammatory
Diseases. Dis. Markers 2008, 25: 267-278.
29. Aberrant Glycosylation Associated with Enzymes as
Cancer Biomarkers. Clin. Proteom. 2011, 8(1): 7-21.
30. CHO Cells in Biotechnology for Production of
Recombinant Proteins: Current State and Further
Potential. Appl. Microbiol. Biotechnol. 2012, 93: 917-930.
31. Differences in the Glycosylation of Recombinant
Proteins Expressed in HEK and CHO Cells. J. Biotech.
2012, 161: 336-348.
32. Characterization of
Host-Cell Line Specific
Glycosylation Profiles of Early Transmitted/Founder
HIV-1 gp120 Envelope Proteins. J. Proteome Res. 2013,
12(3): 1223-1234.
33. Cell Type-specific and Site Directed N-glycosylation
Pattern of FcΥRIIIa. J. Proteome Res. 2011, 10(7): 30313039.
34. Transient Expression of an IL-23R Extracellular
Domain Fc Protein in CHO vs. HEK Cells Results in
Improved Plasma Exposure. Protein Expr. Purif. 2010,
71: 96-102.
35. High-level Protein Expression in Scalable CHO
Transient Transfection. Biotechnol. Bioeng. 2009, 103(3):
542-551.
36. Purification of the Extracellular Domain of the
Membrane Protein GlialCAM Expressed in HEK and
CHO Cells and Comparison of the Glycosylation.
Protein Expr. Purif. 2008, 58(1): 94-102.
37. Recombinant Glycodelin Carrying the Same Type of
Glycan Structures as Contraceptive Glycodelin-A Can
Be Produced in Human Kidney 293 Cells But Not in
Chinese Hamster Ovary Cells. Eur. J. Biochem. 2000,
267(15): 4753-4762.
38. The Absence of Fucose but Not the Presence of
Galactose or Bisecting N-acetylglucoasmine of the
Human IgG1 Complex-Type Oligosaccharides Shows
the Critical Role of Enhancing Antibody-Dependent
Cellular Cytotoxicity. J. Biol. Chem. 2003, 278: 3466-
info@maxcyte.com
3473.
39. Two Mechanisms of the Enhanced AntibodyDependent Cellular Cytotoxicity (ADCC) Efficacy of
Non-Fucosylated Therapeutic Antibodies in Human
Blood. BMC Cancer 2009, 9: 58.
40. The Influence of Glycosylation on the Thermal Stability
and Effector Function Expression of Human IgG1-Fc:
Properties of a Series of Truncated Glycoforms. Mol.
Immunol. 2000, 37: 697-706.
41. CHO Glycosylation Mutants as Potential Host Cells to
Produce Therapeutic Proteins with Enhanced Efficacy.
Adv. Biochem. Eng. Biotechnol. 2013, 131: 63-87
42. Highly Sialylated Recombinant Human Erythropoietin
Production in Large-Scale Perfusion Bioreactor
Utilizing CHO-gmt4 (JW152) with Restored GnT I
Function. Biotechnol. J. 2014, 9(1): 100-109.
www.maxcyte.com
info@maxcyte.com