Skip to main content

Cell culture efforts to reduce glycation in recombinant humanized antibody


Glycation is a common post-translational modification of proteins, resulting from the chemical reaction between reducing sugars such as glucose and the primary amino groups on protein [1]. This non-enzymatic glycosylation reaction generates structural heterogeneity in recombinant IgG1 antibodies produced by cell culture processes [2]. Recent analytical characterization of a full-length humanized antibody secreted by Chinese Hamster Ovary (CHO) cells revealed that glycation of this protein occurs predominantly at lysine 49 on the light chain of the antibody [3]. This finding contrasts with historical data that have suggested that glycation sites are typically located randomly at all accessible lysine residues distributed over the entire molecule [2, 3].

The glycated species accounted for 40–50% of the total antibody produced by transient CHO cell transfections in bioreactors. By contrast, other recombinant antibody molecules produced by CHO cultures generally showed only ~5% glycation [3]. This work documents cell culture process development efforts taken to reduce glycation of this antibody in stably expressing CHO cell lines.


Early stable antibody expression by different CHO clones in 60 mm plates demonstrated no significant differences in glycation (see Table 1). This supports the expectation that glycation of this antibody is an extracellular event such that glycation levels should depend on cell culture conditions and should not vary from clone to clone.

Table 1 Glycation of antibody produced by different CHO clones in 60 mm plates.

Antibody glycation was ~40–60% in 40L bioreactors and ~10–15% in 1L spinners at the time of harvests (see Table 2). Since the extent of this chemical modification should increase with reaction time and substrate availability, the considerable disparity in glycation between bioreactor and spinner samples is attributed partly to the differences in cultivation time and glucose concentrations.

Table 2 Glycation of antibody produced in 40L bioreactors and 1L spinners.

The results from plate and spinner experiments show that the glycation of this antibody can be reduced to below 40%. Hence, 2L bioreactor experiments were conducted to test the feasibility of lowering glycation by reducing the glucose concentration in the culture medium. The pH, dissolved oxygen, and temperature profiles were controlled identically in all the 2L bioreactors. The extent of glycation in the bioreactor samples collected at the end of culture (day 14) was determined using a boronate affinity chromatographic method previously described [3].

In the first series of bioreactor experiments, the glucose concentration in the batch feed was lowered by ~67%, and the amount of supplemental glucose added was reduced by ~50%. These modifications lowered glycation to 14–20% for each of the three antibody stably-expressing clones tested (see Table 3). Subsequent bioreactor experiments employed clone 2 exclusively, and this reduced glucose feed process was used as the control.

Table 3 Glycation of antibody produced in the first, second, and third series of 2L bioreactor experiments. Each condition was evaluated in duplicate bioreactors. Data represents average ± standard deviation obtained from duplicate cultures.

In the second set of bioreactor experiments, antibody glycation was further reduced to ~10% (see Table 3) by eliminating glucose from the batch feed and replacing it with a continuous glucose feed. In the final round of bioreactor experiments, by using glucose-free inoculation medium and batch feed, and by implementing a continuous glucose feed strategy to maintain glucose at even lower concentrations throughout the culture, the glycated species was minimized to 6% (see Table 3). Despite the variation in antibody glycation, product titers and cell-specific productivities were comparable in all the 2L bioreactor experiments.


The percentage of glycated antibody was reduced by lowering the glucose concentration in the culture medium. The extent to which glucose concentration in the cultures was controlled directly impacted the antibody glycation level.


  1. 1.

    Baynes JW: The Maillard hypothesis on aging: time to focus on DNA. Ann N Y Acad Sci. 2002, 959: 360-367.

    CAS  Article  Google Scholar 

  2. 2.

    Harris RJ: Heterogeneity of recombinant antibodies: linking structure to function. Dev Biol (Basel). 2005, 122: 117-127.

    CAS  Google Scholar 

  3. 3.

    Zhang B, Yang Y, Champion K: Revealing an unusual pattern of antibody glycation. WCBP. 2006, Jan 24–27, 2006, San Francisco, CA, USA, : 10th Symposium on the interface of regulatory and analytical sciences for biotechnology health products

    Google Scholar 

Download references


We thank AORS and Media Preparation department for providing assay and media preparation support, respectively. We also thank Lily Chu, John Joly, Cindy Quan, Stacey Ma, Ron Taticek, Bob Kiss, Dana Andersen, Martin Gawlitzek, Robb Shawley and Reed Harris for helpful discussions.

Author information



Corresponding author

Correspondence to Inn H Yuk.

Rights and permissions

Open Access This article is published under license to BioMed Central Ltd. This is an Open Access article is distributed under the terms of the Creative Commons Attribution 2.0 International License (, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Reprints and Permissions

About this article

Cite this article

Yuk, I.H., Huynh, H., Leach, K. et al. Cell culture efforts to reduce glycation in recombinant humanized antibody. Microb Cell Fact 5, P57 (2006).

Download citation


  • Chinese Hamster Ovary Cell
  • Chinese Hamster Ovary
  • Chinese Hamster Ovary Cell Line
  • Cell Culture Process
  • Bioreactor Experiment