Monoclonal antibody production relies primarily on mammalian expression systems to ensure correct folding, assembly, and post-translational modification. Among available hosts, Chinese hamster ovary (CHO) cells and human embryonic kidney 293 (HEK293) cells represent the two most widely used platforms. While CHO dominates commercial biomanufacturing, HEK293 systems are extensively used in discovery, screening, structural studies, and certain clinical applications.
The choice between HEK293 and CHO involves differences in cell biology, protein processing capacity, glycosylation machinery, scalability, and regulatory precedent. In some cases, host selection can directly determine whether a construct expresses efficiently or exhibits product quality challenges.
This article examines HEK293 and CHO across biological characteristics, productivity, glycosylation, product integrity, and application-specific strategy, integrating findings from recent comparative studies.
HEK293 cells originate from human embryonic kidney tissue transformed with adenovirus DNA. Numerous derivatives exist, including adherent and suspension-adapted lines such as HEK293F, Expi293F, and HEK293S.
HEK293 cells are widely used for transient transfection due to high transfection efficiency using polyethyleneimine (PEI) or lipid-based reagents. Their human origin provides a human glycosylation enzyme repertoire, including α2,6-sialyltransferase activity.
Several reports indicate that HEK293 cells exhibit broader post-translational modification capacity for specific modifications, including tyrosine sulfation and γ-carboxylation, compared to CHO.
Beyond enzymatic capacity, transcriptomic analysis has demonstrated differences in expression of secretory pathway genes between HEK293 and CHO cells. In a comparative study of 24 difficult-to-express human secreted proteins expressed in episomal systems (293ALL vs CHOEBNALT85-1E9), 9 proteins showed greater than 2-fold higher secretion in HEK293.
These findings indicate that HEK293 cells may provide intracellular processing conditions favorable for certain human proteins.
CHO cells originate from Chinese hamster ovary tissue and have been extensively adapted for suspension growth and industrial-scale bioprocessing. Subclones such as CHO-K1, CHO-S, and DG44 are widely used in stable cell line development.
CHO cells are optimized for large-scale fed-batch and perfusion processes and have a well-established record in commercial monoclonal antibody manufacturing. Gene amplification systems such as DHFR and glutamine synthetase (GS) are commonly used to generate high-producing stable clones.
CHO glycosylation differs from human systems. CHO cells lack α2,6-sialyltransferase activity and can generate Neu5Gc and α-gal epitopes, depending on culture conditions and cell line characteristics.
Despite these differences, CHO-derived monoclonal antibodies have demonstrated clinical safety and efficacy across hundreds of approved products.
Recombinant Protein Production →
HEK293 derivatives are widely used for rapid antibody production in research and preclinical workflows. High transfection efficiency enables fast turnaround from plasmid to purified antibody.
Transient HEK293 systems are particularly suited for:
Early-stage candidate screening
Epitope validation
Small-scale in vivo studies
Structural biology workflows
Stable HEK293 pools can also be generated. In Expi293F systems, stable pools were reported to produce approximately 3-fold higher titers than transient expression for selected constructs.
CHO cells are primarily used in stable expression workflows. Gene amplification systems (e.g., DHFR, GS) enable selection of high-producing clones capable of multi-gram per liter titers in optimized fed-batch processes.
CHO systems are compatible with large-scale bioreactors and long-term production campaigns. Stable CHO clones demonstrate genetic stability and reproducibility across extended culture durations.
A mechanistic dimension to host selection was illustrated in a study of 24 difficult-to-express proteins. Nine proteins showed greater than 2-fold higher secretion in HEK293 compared to CHO in episomal systems.
Transcriptomic analysis identified differences in secretory pathway gene expression between hosts. Functional testing demonstrated that co-expression of selected genes in CHO could increase secretion for specific proteins. For example:
ATF4 or SRP9 increased THBS4 secretion by more than 2-fold in CHO
ATF4, PDIA3, and HSPA8 increased ARTN secretion
However, not all HEK293-enriched genes improved secretion in CHO, and some affected cell growth. These findings demonstrate that secretion limitations may be host-dependent and, in certain cases, correctable through targeted co-expression strategies.
HEK293 cells provide α2,6-sialyltransferase activity and have been reported to exhibit greater capacity for tyrosine sulfation and γ-carboxylation compared to CHO.
Human-derived glycosylation machinery may be advantageous for certain proteins sensitive to glycan structure or for constructs requiring specific enzymatic modifications.
GnTI− HEK293 derivatives (HEK293S GnTI− and Expi293F GnTI−) have been used to reduce glycan complexity. Proteins expressed in these systems exhibited faster SDS-PAGE migration and later size exclusion chromatography elution, consistent with reduced glycan modification.
Expi293F GnTI− stable pools produced higher titers than HEK293S GnTI− stable pools (1.3- to 10-fold difference).
CHO cells lack α2,6-sialyltransferase activity and can generate Neu5Gc and α-gal epitopes. Despite these differences, CHO glycan profiles are well characterized and have supported widespread therapeutic use.
Related: Antibody Glycoengineering
Product integrity differences between hosts were observed in a comparative analysis of CHO-K1 and Expi293F systems.
Several recombinant proteins expressed in CHO-K1 stable pools exhibited clipping. Expression of the same constructs in Expi293F reduced or eliminated clipping.
In one single-chain complex:
Expi293F stable pools produced ~3-fold higher titers than transient expression
~3-fold higher titers than CHO-K1 stable pools
After purification, intact yield was ~10-fold higher due to reduced clipping
PNGase F analysis demonstrated that certain electrophoretic differences in Expi293F-derived proteins were attributable to glycosylation rather than proteolysis. These findings indicate that host background can influence proteolytic processing and intact product recovery.
Commercial License for CHOK1BN →
Antibody format influences expression outcomes. In a comparative dataset:
IgG expression exceeded IgM expression
IgM assembly requires coordination of 21 polypeptide chains
Light chain choice (κ vs λ) had minimal impact on overall yield
Passage number influenced productivity independently of cell density
These findings demonstrate that antibody structural complexity and cellular condition contribute to expression performance.
HEK293 transient systems enable rapid material generation for:
Screening
Functional assays
Structural studies
Host-dependent secretion differences may be relevant when evaluating difficult-to-express constructs.
CHO remains the dominant platform for clinical-stage and commercial monoclonal antibody manufacturing due to:
Established regulatory precedent
Stable high-producing clones
Compatibility with industrial-scale processes
HEK293 systems demonstrated:
Higher secretion for selected difficult-to-express proteins
Reduced clipping for certain constructs
Broader enzymatic capacity for specific post-translational modifications
Glycan simplification capability via GnTI− derivatives
CHO systems provide:
Mature stable expression workflows
Established gene amplification strategies
Industrial-scale productivity
Extensive clinical track record
HEK293 and CHO differ in secretory pathway gene expression, enzymatic modification capacity, and protein processing outcomes. Experimental studies demonstrate that certain proteins exhibit higher secretion in HEK293, that clipping observed in CHO can be reduced in HEK293 for specific constructs, and that glycosylation capabilities differ between hosts.
At the same time, CHO remains the dominant platform for commercial monoclonal antibody production due to established scalability and regulatory history.
Host selection therefore depends on construct characteristics, development stage, and manufacturing objectives. Integrating mechanistic understanding with platform experience enables informed decision-making in antibody development programs.
References:
Tan, E., Chin, C. S. H., Lim, Z. F. S., & Ng, S. K. (2021). HEK293 Cell Line as a Platform to Produce Recombinant Proteins and Viral Vectors. Frontiers in Bioengineering and Biotechnology, 9. https://doi.org/10.3389/fbioe.2021.796991
Lalonde, M., & Durocher, Y. (2017). Therapeutic glycoprotein production in mammalian cells. Journal of Biotechnology, 251, 128-140. https://doi.org/10.1016/j.jbiotec.2017.04.028
Büssow, K. (2015). Stable mammalian producer cell lines for structural biology. Current Opinion in Structural Biology, 32, 81-90. https://doi.org/10.1016/j.sbi.2015.03.002
Veber, A., Lenau, D., Gkragkopoulou, P., Bauer, D. K., Focken, I., Leuschner, W. D., Beil, C., Weil, S., Rao, E., & Langer, T. (2025). Impact of Light-Chain Variants on the Expression of Therapeutic Monoclonal Antibodies in HEK293 and CHO Cells. Antibodies, 14(3). https://doi.org/10.3390/antib14030053
Malm, M., Kuo, C., Barzadd, M. M., Mebrahtu, A., Wistbacka, N., Razavi, R., Volk, A., Lundqvist, M., Kotol, D., Tegel, H., Hober, S., Edfors, F., Gräslund, T., Chotteau, V., Field, R., Varley, P. G., Roth, R. G., Lewis, N. E., Hatton, D., . . . Rockberg, J. (2022). Harnessing secretory pathway differences between HEK293 and CHO to rescue production of difficult to express proteins. Metabolic Engineering, 72, 171-187. https://doi.org/10.1016/j.ymben.2022.03.009
Sun, H., Wang, S., Lu, M., Tinberg, C. E., & Alba, B. M. (2023). Protein production from HEK293 cell line-derived stable pools with high protein quality and quantity to support discovery research. PLOS ONE, 18(6), e0285971. https://doi.org/10.1371/journal.pone.0285971
January 2026 deal activity leaned heavily toward bispecific, trispecific, and AD……
ADCs to watch in 2026. Clinical Trials, Primary Endpoints.
PepTalk 2026 highlights and event recap. Biotech antibody conference!
ScientistとScience Exchangeとのご利用可能です。
