Skip to main content
NCBI home page
Cytotechnology logoLink to Cytotechnology
. 2020 Mar 3;72(3):343–355. doi: 10.1007/s10616-020-00381-z

Enhancement of sialylation in rIgG in glyco-engineered Chinese hamster ovary cells

Thi Sam Nguyen 1,#, Ryo Misaki 1,#, Takao Ohashi 1, Kazuhito Fujiyama 1,2,✉
PMCID: PMC7225233  PMID: 32125558

Abstract

Since about 70% of commercial biopharmaceutical products have been produced in Chinese hamster ovary (CHO) cells, this cell line is undeniably a workhorse for biopharmaceuticals production. Meanwhile, sialic acid terminals were reported to affect anti-inflammatory activity, antibody-dependent cellular cytotoxicity efficacy of IgG antibodies. Taking these findings together, we aimed to establish CHO cell lines that highly produce sialic acid terminals by overexpressing two N-acetylneuraminic acid-based key enzymes, α(2,6)-sialyltransferase and UDP-N-acetylglucosamine 2-epimerase/N-acetylmannosamine kinase using dihydrofolate reductase/methotrexate gene amplification method. Indeed, the number of total sialic acid terminal glycan structures increased tremendously, by 12-fold compared to the wild type in total protein extracts. With the methotrexate supplementation, a targeted cell line, CHOmt17-100, showed up to 1.4 times more sialylated structures of glycoforms in total proteins. Interestingly, immunoglobulin G, used as the model protein in CHOmt17-100, showed about 53% sialylated structures in its glycoforms. These resultant sialylated glycans exhibited more than approximately 14.5 times increase as compared to that of the wild type. Moreover, the resultant glycan structures mostly had N-acetylneuraminic acid terminals, while N-glycolylneuraminic acid terminal composition remained less than 5% as compared to the wild type. Engineered antibodies derived from CHO cell lines that produce high levels of sialic acid will contribute to the examination of glycoforms’ efficacy and usefulness toward bio-better products.

Keywords: Chinese hamster ovary (CHO) cell, IgG, Antibody, Glycosylation, Sialyltransferase, UDP-GlcNAc 2-epimerase/ManNAc kinase

Introduction

Numerous biopharmaceutical products produced in Chinese hamster ovary (CHO) cells have been recognized for clinical uses. Hence, CHO cells have been certified as a safe workhorse for glycoprotein products (Jayapal et al. 2007). Moreover, CHO cells are resistant to many human pathogenic viruses such as HIV, influenza, polio, and herpes etc (Jayapal et al. 2007; Shukla and Spear 1999; Dechecchi et al. 2001). On the other hand, CHO cells are easy to scale up because of their adaptability and ease of genetic manipulation (Jayapal et al. 2007). Thus, CHO cells have been used on an industrial scale with high accretion in cell cultures. With so many benefits, CHO cells may be considered the premier host cell line for biopharmaceutical production, at least in the near future. In our previous research in Thi Sam et al. (2018), we attempted to establish host cell lines that could better improve sialylation patterns in total proteins. The enhancement of terminal sialylation of N-glycan is one of the most important goals in improving the quality of therapeutics, which should be elucidated in recombinant protein production (Lipscomb et al. 2005). N-Acetylneuraminic acid (NeuAc or Neu5Ac), commonly known as sialic acid, is well known as an important moiety because of its biological functions. Due to its strong electronegativity, Neu5Ac enhances the stabilization and modulation of the interaction of molecules and membranes in organisms. Sialic acid helps to enhance the stability of protein conformation as well as to prevent cells from proteases and glycosidase digestion. Neu5Ac is also believed to be able to modulate signaling in cell transmembrane, differentiation, and growth. Moreover, Neu5Ac is reported to have antioxidant and anti-apoptotic effects (Varki et al. 2009). With so many important effects on cells, Neu5Ac-terminal glycoproteins are highly desirable for biopharmaceutical production. In our previous research, we established a transfected CHO cell line called CHOmt17-0, carrying two Neu5Ac-based key enzymes, α(2,6)-sialyltransferase (ST6GAL-I) and UDP-N-acetylglucosamine (GlcNAc) 2-epimerase/N-acetylmannosamine (ManNAc) kinase (GNE), using the dihydrofolate reductase/methotrexate (DHFR/MTX) gene amplification method. Here, we aim to select a host cell line with enhanced production of sialylation in both total proteins and model proteins, specifically antibody immunoglobulins.

Immunoglobulin Gs (IgGs), which are glycoproteins in serum, protect vertebrates from pathogens and unknown substances (Beale and Feinstein 1976; Raju 2003). IgG is the most abundant component in serum, making up more than about 75% of it. “Fragment antigen binding” (Fab) regions of IgG can help the body to recognize antigenic substances and contribute to the diversity of the antibodies. Moreover, the antibodies have “fragment crystallizable” (Fc) regions that enable antibodies to interact with Fc gamma receptors on phagocytes. In addition, the Fc fragment and the hinge are more sensitive toward proteolysis, while Fab regions can resist proteases (Ryan et al. 2008; Raju and Scallon 2006, 2007). IgGs depend on their Fc regions for the initiation of responses to binding antigens. The domain of Fc fragment CH2, composed of glycan structures in the heavy chain, is attached to asparagine 297 (Asn297) (Lund et al. 1996; Krapp et al. 2003). The glycan structures can help IgG to activate the complement and binding functions to FcγR (Krapp et al. 2003). IgG binds to the foreign epitope on an antigen, whereas the Fc region interacts with Fcγ receptors of immune cells. These immune cells release a cytotoxic substance that causes cell death (Strome et al. 2007). More than 30 glyco-patterns have been linked to the Fc regions of IgG (Burton and Dwek 2006). The glycan structures in the antibodies contain variable monosaccharides, though their effects on the biological functions of IgG are not fully understood. Thus, there is increasing interest in altering glycosylation to elucidate the effects of glycan structures on the functions of antibodies. As an example, the Fc regions of antibodies without core fucose were reported to be able to increase antibody-dependent cellular cytotoxicity (ADCC) (Beale and Feinstein 1976; Shields et al. 2002; Shinkawa et al. 2003). Recently, the enhancement of α(2,6) sialylation of IgGs was reported to result in superior ADCC (Kurogochi et al. 2015). Moreover, the Fc region carrying the Neu5Ac-terminal glycan was demonstrated to show anti-inflammatory ability. Interestingly, the study revealed that intravenous immunoglobulin is most effective for anti-inflammation when it contains a large amount of α(2,6) sialylation (Kaneko et al. 2006; Anthony et al. 2008).

In conclusion, to prove the concept of using gene amplification to enhance sialylation in glycoprotein by overexpressing two Neu5Ac-based key enzymes, we attempted to engineer a CHO cell line that exhibits superior glycan profiles in total proteins and in IgG antibodies.

Materials and methods

Cells and culture media

In our previous study, we established the cell line CHOmt17-0 by co-overexpressing two related human enzymes as described in Thi Sam et al. (2018), which we then cultured in a (–) Minimum Essential Medium Eagle Alpha Modification (α MEM) (Invitrogen, Massachusetts, USA). The medium was supplemented with 10% dialyzed fetal bovine serum (FBS) and 100 µg/mL neomycin without nucleosides or l-glutamine. Antibody IgG transfectants were cultured in an (–)α MEM (Invitrogen) without nucleosides, supplemented with 10% dialyzed FBS, 100 µg/mL neomycin, 100 µg/mL hygromycin, and 1 µg/mL puromycin. High-resistance MTX cell lines were also cultured in an (–)α MEM (Invitrogen) without nucleosides, supplemented with 10% dialyzed FBS, 100 µg/mL neomycin, and an MTX concentration of 100, 200, or 300 nM. All established cells were then cultured in a cell culture incubator monitored at 37 °C and 5% CO2.

Quantitative real-time PCR (qRT-PCR)

The procedure to prepare a sample was described in Thi Sam et al. (2018). In brief, phosphate-buffered saline (PBS) was used to wash harvested cultured cells (1 × 107). Then, the total RNA from the collected cells was extracted using an RNeasy Mini Kit (Qiagen, Hilden, Germany) according to the manufacturer’s instructions. Complementary DNA (cDNA) was next synthesized using a SuperScript® VILO™ cDNA Synthesis Kit from the RNA following the manufacturer’s instructions. In order to elucidate the expression of the investigated genes in the cell lines, qRT-PCR was used in the StepOnePlus™ real-time PCR system (Applied Biosystems, Illinios, USA). The following primers were employed for the PCR reaction: 5′ TGTGTGGAAGGAAAAGAAGAAAGG 3′ (sense primer for ST6GAL-I), 5′ GGGTGCTGCTTGAGGATACA 3′ (antisense primer for ST6GAL-I), 5′ GTGGAAGGGATGTCAGTACCAAA 3′ (sense primer for Gne), and 5′ CCAAGTCCCAAAGCAGTTCC 3′ (antisense primer for Gne) as described in Thi Sam et al. (2018).

Generation of highly methotrexate-resistant cell lines

To generate highly MTX-resistant cell lines to upregulate genes of interest in CHOmt17, the glyco-engineered cell lines (1  ×  107) were supplemented with various concentrations of MTX into culture medium containing 100, 200, or 300 nM of MTX. Stable highly MTX-resistant cell lines were generated after 3 weeks of culture in an incubator at 37 °C and 5% CO2. After selection, the stable cell lines CHOmt17-100, CHOmt17-200, and CHOmt17-300, exhibiting high resistance to 100, 200, and 300 nM of MTX, respectively, were established.

Purification using Protein G Sepharose

Due to its specific binding to the antibodies, Protein G Sepharose™ (GE, Illinios, USA) was employed to purify the antibodies from the crude sample. The IgG1-producing cells (2 × 108 ) derived from highly MTX-resistant transfected cell lines were cultured in an α MEM without FBS supplementation. After 5 days in culture, it was then purified using Protein G Sepharose™. The sample was next subjected to centrifugation at 5000×g for 5 min at 4 °C. Four milliliters of gel column was then loaded by 10 mL binding buffer composed of 0.02 M sodium phosphate pH 7.0. After running the medium sample, the gel column was next washed with binding buffer and eluted with 0.1 M glycine–HCl pH 2.7. The elution was neutralized using 1 M Tris–HCl pH 9.0. Purification was performed at 4 °C.

Viable cell density

To elucidate viable cell density, the cell lines were washed twice with PBS. The harvested cells were then detached by using 1 mL 0.25% trypsin (Life technologies, California, USA) before adding 9 mL of fresh α MEM into each dish. Next, 50 µL of the sample containing cells was gently mixed with 50 µL of 0.4% trypan blue (Sigma, St. Louis, MO, USA). Then, 15 µL of the mixture was then transferred into each chamber in a hemocytometer. Counts were performed in three experiments on days 1, 3, 5, 7, 9, and 11, followed by analysis under a 40× microscope objective as described previously (Louis et al. 2011).

Enzyme-linked immunosorbent assay (ELISA)

Ninety-six-well ELISA plates were each coated with 2.5 µg/mL of goat anti-human IgG (H + L) (MBL, Aichi, Japan) was blocked by 50 µL of 5% bovine serum albumin (BSA) diluted in PBS. The plates were then each incubated with 50 µL of crude/purified sample. Then, 150 µL of PBS containing 0.05% Tween 20 was used to wash each plate three times before incubation with 50 µL of goat anti-human IgG (H + L) Fab_HRP. The amount of bound IgG1 antibody was then measured by absorbance at 450 nm. Before the addition of 3M HCl, the solution was mixed with SIGMAFAST™ o-phenylenediamine dihydrochloride (Sigma).

Silver staining

The recovered antibodies in the elution from Protein G Sepharose™ were subjected to SDS-PAGE and silver staining. After heating at 100 °C for 5 min, 20 µL of purified sample was mixed with 2× sample buffer (60 mM Tris–HCl pH 6.8, 2% SDS, 10% glycerol, 5% ß-mercaptoethanol, 0.01% bromophenol blue). The mixture was then loaded into 10% polyacrylamide gel, which were then run. The protein bands were analyzed using a silver staining kit, Sil-Best Stain One (Nacalai Tesque, Kyoto, Japan), according to the manufacturer’s instructions.

Glycan structure analysis total proteins in highly MTX-resistant cells

Cultured cells (1 × 108) were washed with PBS and lysed with lysis buffer [50 mM Tris-HCl pH 7.4, 150 mM NaCl, 1% NP-40, 0.5% sodium deoxycholate, 0.1% SDS, and 1 mM ethylenediaminetetraacetic acid (EDTA)] (Thi Sam et al. 2018). N-glycans were released from lyophilized lysate by peptide-N-glycosidase F (Takara, Shiga, Japan) according to the manufacturer’s instructions. The N-glycans were then labeled with pyridylamine (PA) as previously described (Kondo et al. 1990) and subjected to phenol/chloroform extraction to remove unreacted PA. An RP-HPLC apparatus (Hitachi 7000 HPLC system, Tokyo, Japan) was then employed to monitor PA-sugar chains using a Cosmosil C18 column (4.6 × 250 mm) (Nacalai Tesque). The excitation and emission wavelengths were 310 nm and 380 nm, respectively. With a flow rate of 1.2 mL/min, the elution was collected by linearly increasing the acetonitrile concentration from 0–35% in 0.02% trifluoroacetic acid (TFA) at 30 °C for 50 min as described in Thi Sam et al. The fractions eluted from RP-HPLC were then analyzed using an LC–MS system (Agilent Technologies, CA, USA) with HTC Plus software (Bruker Daltonics, MA, USA) as previously described (Limkul et al. 2016). In brief, the mobile phase of the LC was consisted of acetonitrile/acetic acid used as solvent A (98/2, v/v). The solvent B was composed of water/acetic acid/trimethylamine (92/5/3, v/v/v). The column Shodex Asahipak NH2P-50 2D (2.0 mm ID × 150 mm, Showa Denko, Tokyo, Japan) was used for separation of PA-sugar chain by linearly adding concentration of solvent B from 20% to 55% at 0.2 mL/min flowrate over 35 min. The MS/MS parameters conditions were following as: nebulizer flow at 5.0 psi; scan range m/z at 350–2750. Hence, the peak area of the LC was used for evaluation the N-glycan relative amount.

Glycan structure analysis of recombinant antibodies

The recovered antibodies from purification using Protein G Sepharose™ were then subjected to glycan pattern analysis, which was performed in a manner similar to that described above.

Results

Establishment of cell lines highly expressing both ST6GAL-I and GNE with increased MTX concentrations

In our previous studies, we established a tri-cistronic expression plasmid, which carries three genes (ST6GAL-I, Gne, and a Dhfr), for gene amplification. Those genes of interest were driven by a single cytomegalovirus promoter. Furthermore, these DNA fragments were linked to each other by an internal ribosome entry site (IRES) wild type and attenuated IRES as described in Thi Sam et al. (2018). Transfection of the expression vector pcDNA-SG2 into CHO-DG44 cells resulted in the establishment of clone CHOmt17-0 derived cell lines stably expressing ST6GAL-I and GNE (Thi Sam et al. 2018). qRT-PCR was employed to evaluate the expression levels of the ST6GAL-I and Gne genes in the highly MTX-resistant transfectants CHOmt17-100, CHOmt17-200, and CHOmt17-300. Indeed, as the MTX concentration increased, the genes of interest showed higher expression levels than the wild type and CHOmt17-0 (Fig. 1a). Interestingly, the expression levels of ST6GAL-I and Gne in CHOmt17-100, CHOmt17-200, and CHOmt17-300 showed approximately three-fold increases compared to those without MTX treated CHOmt17-0. In addition, the higher MTX concentrations of 200 nM and 300 nM did not increase the mRNA expression level compared to the lower concentration of 100 nM (Fig. 1a). To confirm the qRT-PCR results, we amplified the full-length sequences of ST6GAL-I and Gne using cDNA as a template, and found high expression levels of the genes of interest in the MTX-treated clones, as shown in Fig. 1b. We were thus able to confirm success in increasing the expression of both ST6GAL-I and Gne in the cell lines in the presence of MTX.

Fig. 1.

Fig. 1

Open in a new tab

a The expression levels of the investigated genes were evaluated using quantitative RT-PCR in both wild-type and stable transfected cell lines. Left column: ST6GAL-I; right column: GNE; P < 0.05. b Gel electrophoresis of the investigated genes’ PCR products using specific primers in the WT, transfected CHO cells. Control: plasmid pcDNA3.1-SG2 (Thi Sam et al. 2018)

Highly sialylated antennary structures generated in highly MTX-resistant cell lines

To evaluate the effects of MTX on the amplification of sialylation in CHOmt17, the detailed glycoforms of the cell lines with the addition of various MTX concentrations were further analyzed. CHOmt17-0 cells were incubated at 37 °C for 7 to 9 days in culture medium supplemented with 100 µg/mL neomycin and 10% dialyzed FBS in MTX concentrations of 100, 200, and 300 nM. Viable cell density was about 3 × 106 cells/mL at the time of harvesting and cell growth (Fig. 2). As described above, the PA-labeled N-glycans were fractionated by RP-HPLC and their structures were then analyzed by LC–MS. The RP-HPLC profiles were noticeably distinguishable among the wild type, CHOmt17-0, CHOmt17-100, CHOmt17-200, and CHOmt17-300, as shown in Fig. 3. In particular, the late-eluted peak exhibited a dramatic increase in the transfected cell lines. Hence this change resulted in a jump in the relative amounts of highly galactosylated and sialylated structures (Table 1). Interestingly, sialylation in CHOmt17-100 showed a great jump of 12-fold compared to the wild type. Moreover, the cell line CHOmt17-100 illustrated about 1.4 times higher sialylation compared to MTX-free CHOmt17-0. The cell lines CHOmt17-200 and CHOmt17-300 did not show a big increase in the level of sialylation compared to CHOmt17-100 (Table 1). This was thus in agreement with the mRNA expression level obtained previously (Fig. 1). These results suggested that the overexpression of ST6GAL-I and GNE with MTX supplementation considerably enriched the antennary sialylated glycan structures in this stable transfected cell line (Table 1). Although the terminal N-glycolylneuraminic acid (NeuGc) was detected in cell lines except for CHOmt17-0 and showed higher ratio of increase by MTX-addition compared to that of the terminal NeuAc, these amounts were still kept at low level. This clarified that the increased sialic acid residues in the engineered cell lines were mainly Neu5Ac. Collectively, these results indicated that the overexpression of both GNE and ST6GAL-I due to gene amplification using MTX resulted in a remarkable accretion of sialic acid terminals in total protein glycans. The sialylation of CHOmt17-100 was at a level similar to that of the cell lines with higher amounts of MTX (Table 1). Moreover, due to its better viable cell density, CHOmt17-100 was selected as the host cell line for the transfection of recombinant IgG as a proof of concept using the gene amplification method described in this study.

Fig. 2.

Fig. 2

Open in a new tab

Growth profiles of WT and engineered CHO cells cultured in α MEM. Error bars represent the standard deviation calculated from data obtained in three independent experiments (n = 3)

Fig. 3.

Fig. 3

Open in a new tab

Glycan profiles of CHOmt17-0, CHOmt17-100, CHOmt17-200, and CHOmt17-300 cells. The total N-glycans extracted from glycoproteins were labeled with PA. The PA-glycans were fractionated using RP-HPLC. Then, the collected peaks were used to perform LC–MS. Glucose units: RP-HPLC profile for identification of glucose oligomer 3–15

Table 1.

Composition of predicted terminal sialic acid residue and predicted N-glycan structures

Structures Relative amount (%) in total N-glycan
Wild-type CHOmt17-0 CHOmt17-100 CHOmt17-200 CHOmt17-300
Neu5Ac terminal 1.0 11.3 14.4 14.6 14.4
NeuGc terminal 0.3 – 1.1 1.2 1.5
High-mannose type 68.1 46.1 52.6 62.7 67.1
Galactosylated glycans 11.7 8.2 9.9 1.4 3.4
Sialylated glycans 1.3 11.3 15.5 15.8 15.9
Fucosylated glycans 13.4 30.4 40.5 38.1 18.9
Open in a new tab

Productivity of recombinant IgG in cell pools

To prove the concept of cell lines greatly expressing sialylation, we transfected plasmid carrying heavy- and light-chain IgG1 into CHOmt17-0 and CHOmt17-100. Next, we created stable cell lines that produced higher levels of antibodies. The crude samples carrying antibodies were then analyzed using ELISA to check their productivity. ELISA from equivalent sample amounts for each cell pool in both CHOmt17-0 and CHOmt17-100 was performed as shown in Fig. 4a, b. The productivity of each antibody was tested on days 1, 3, 5, 7, 9, and 11 in three independent experiments. Interestingly, the cell pools derived from both CHOmt17-0 and CHOmt17-100 demonstrated comparable capabilities in producing antibodies: about 1 µg/mL on day 5. Then the clones that produced the highest levels of antibodies were chosen for purification. The antibodies collected from CHOmt17-0 and CHOmt17-100 were purified using Protein G Sepharose. The recovered antibodies were then confirmed by ELISA and silver staining. Figure 4c shows confirmation of the variant heavy chain and variant light chain at 50 kDa and 25 kDa, respectively, in reduced-condition SDS-PAGE. In conclusion, stable cell lines that highly produce IgG1 have been established using CHOmt17-0 and CHOmt17-100. To elucidate the glycan patterns in these antibodies, the recovered IgG were then subjected to glycan analysis using HPLC and LC–MS.

Fig. 4.

Fig. 4

Open in a new tab

a Productivity of CHOmt17-0 b Productivity of CHOmt17-100 cell pools, producing IgG in α MEM on days 1, 3, 5, 7, 9, and 11 using ELISA. Error bars represent the standard deviation calculated from data obtained in three independent experiments (n = 3). c Analysis of 100 ng IgG-purified proteins by silver staining of SDS-PAGE gels in wild-type, CHOmt17-0, CHOmt17-100 cell lines. P: human plasma IgG (50 ng)

Highly sialylated antennary structures generated in the recombinant antibodies IgG

RP-HPLC was employed to examine the N-glycan structures in the recombinant antibodies IgG, followed by LC–MS in the investigated cell lines, the wild type, CHOmt17-0, and CHOmt17-100. The recovered antibodies from purification with Protein G Sepharose were then released N-glycans by using peptide-N-glycosidase F. The PA-labeled N-glycans were fractionated by RP-HPLC, and the structures were then analyzed by LC–MS. Indeed, the glycan profiles were significantly distinguishable among antibodies derived from the wild type, CHOmt17-0, and CHOmt17-100 (Fig. 5). Specifically, the late-eluted peak in the RP-HPLC profile resulted in a notable increase in the antibodies’ glycan profiles derived from CHOmt17-100 compared to those of the CHOmt17-0 and wild type (Fig. 5). These results are in agreement with the total protein glycan patterns in Fig. 3. The collected peaks in each clone were then transferred for glycan structural analysis using LC–MS. Table 2 shows the proportions of glycan structures in IgG from engineered CHOmt17-0, CHOmt17-100, and wild type. Indeed, the sialylated structures in IgG from CHOmt17-0 exponentially increased ten-fold, with more than 37% of glycans being sialylated compared to 3.7% of the wild type. Similarly, with supplementation by MTX, the IgG produced in CHOmt17-100 resulted in more than 53% sialylated glycoforms, approximately 14.5 times the level of the wild type. This indicates strong agreement with the glycan analysis of total protein extracts in cell host lines CHOmt17-0 and CHOmt17-100. Interestingly, Neu5Ac-terminal glycan patterns accounted for most of the increase in sialylated structures (Table 2). Despite the tremendous jump in sialylated structures, the composition of the NeuGc terminal in antibodies produced in engineered CHO remained unchanged at less than 5% compared to the wild type. These results demonstrated that the overexpression of two Neu5Ac-based key enzymes using DHFR/MTX efficiently improved Neu5Ac-terminal structures in the investigated antibodies.

Fig. 5.

Fig. 5

Open in a new tab

Glycan profiles of IgG antibodies produced in WT, CHOmt17-0, and CHOmt17-100 cells. The N-glycans extracted from the IgG antibodies were labeled with PA. The PA-glycans were fractionated using RP-HPLC. Then, the collected peaks were used to perform LC–MS. Glucose units: RP-HPLC profile for identification of glucose oligomer 3–15.

Table 2.

Composition of predicted terminal sialic acid residue and predicted N-glycan structures in antibodies produced in engineered cells

Structures Relative amount (%) in total N-glycan in IgG from engineered cells
Wild-type IgG CHOmt17-0 IgG CHOmt17-100 IgG
Neu5Ac terminal – 35 53.5
NeuGc terminal 3.7 2.4 –
Sialylated glycans 3.7 37.4 53.5
High-mannose type 12.4 – –
Galactosylated glycans 51.8 13.7 13.6
Fucosylated glycans 64.6 60.6 80.2
Open in a new tab

Discussion

For 20 years, CHO cells have been known as the predominant host cell line for biopharmaceutical production. Due to its contribution to the efficiency of biopharmaceuticals, many studies have aimed to modify the glycan patterns in CHO cell lines. Such have included altering glycosylation by process optimization, culture media supplement, or genetic modification. Furthermore, alteration of glycan profiles by modifying glycosylation-related genes is of the most widespread interest.

Numerous studies have attempted to optimize the conditions of the process affecting glycosylated proteins. Despite the various studies that have shown the effects of bioprocess parameters such as pH, dissolved oxygen (DO), temperature, and time on glycan profiles, the results obtained have not been definitive, since the effects depend on the cell types and targeted proteins. In terms of sialylation, Trummer et al. (2006) proved that the decrease in sialylated patterns resulted from the reduction of culture temperature. Lowering the temperature after the exponential growth phase causes a big decrease in Neu5Gc (Borys and Abu-Absi 2010). DO is also considered a parameter that has been proved to increase the sialic acid content of human follicle-stimulating hormone expression (Chotigeat and Gray 1994). According to Yang and Bulter (2000a), ammonium supplements recreated high pH, resulting in a decrease in the terminal sialylation of glycoproteins. Even though the modulation of glycosylation by altering process parameters is considered attractive due to its easy manipulation, the results obtained vary and are not so effective.

Medium component supplementation has been introduced as an additional method to control the glycosylation of glycoproteins. The components commonly used are sugars, nucleotide sugars, or metals. By reducing the glucose concentration in interferon (IFN)-γ expression culture, Chee and Gek (2005) proved that the percentage of sialylated glycans was reduced along with increases in the type of hybrid and in high mannose. Similarly, the decline of glutamine corresponded to decreases in sialylation and fucosylation patterns (Burleigh and Stroop 2011). Nucleotide sugars, known as building blocks for the biosynthesis of glycoproteins, are also utilized to modulate the glycosylation pathway. Adding glucosamine lowers the human tissue inhibitor of metalloproteinase 1 sialylation profiles expressed in NS0 and CHO cells (Baker and James 2001). The supplementation of ManNAc slightly increases intracellular CMP-sialic acid (Baker and James 2001). Studies on the effects of supplementation components on erythropoietin (EPO) glycan patterns in CHO cells revealed that the addition of glucosamine reduced tetra-sialylated glycan compositions (Yang and Bulter 2000b). Nevertheless, many studies have shown successful achievement solely through medium supplementation. In a study in which galactose and ManNAc were added to cell culture, the expressed IFN-γ showed an increase in sialic acid content due to higher concentrations of intracellular CMP-sialic acid and UDP-N-acetylhexosamine (Wong et al. 2010). Galactose solely supplemented in the CHO-GS cell line resulted in 44% enhancement of total sialic acid concentration and a 20% increase in sialylated glycan patterns in Fc-fusion protein expression (Liu and Fan 2015). Due to its simplicity and time efficiency, the supplementation component method is considered attractive, though the results still vary and depend on protein expression levels. In addition, this method is expensive and thus demands a better way, such as platform engineering, to modulate glycoforms of targeted proteins in CHO cells.

Even though there have been many efforts to alter the glycosylation pathway, the results obtained have not been dramatic improvements, nor have they been cost effective. A more stable, direct, and cost-effective method is to engineer a host cell line by deleting or inserting targeted genes involved in the glycosylation pathways of host cells or in the up- or down-regulation of enzyme expression is. Dozens of engineering methods can be used to establish cell lines that produce high levels of sialylation, such as glycosylated enzyme overexpression, small interfering RNA (siRNA), microRNAs, and zinc finger nucleases. Sialidase, a glycosidase that can cleave sialic acid from galactose terminal structures, has been utilized as a target by siRNA. An upgrade in sialic acid content and about a 60% decrease in sialidase activity of recombinant IFN-γ in CHO cell lines were obtained by using the siRNA method (Ngantung and Wang 2006). In another study about silencing sialidase activity, Zhang and team demonstrated that knocking out a sialidase, neuraminidase 3, greatly decreased sialidase activity, by about 98%, while inducing a 33% increase in sialic acid content in IFN-γ (Zhang et al. 2010). Sialylation can also be modified by the overexpression of sialylation-related genes. For example, ST3GAL-4 and β(1,4)-galactosyltransferase (β1,4-GT) have been co-expressed in CHO cells, which enriched sialic acid content and increased the number of tri-sialylation glycan structures compared to only ST3GAL-4 expression in EPO (Jeong and Kim 2008). Another way to enhance the sialylation of recombinant IFN-γ, namely introduction of CMP-sialic acid transporter (CMP-SAT), has increased CMP-SAT expression up to 2.8-fold and increased sialylation of IFN-γ products by 4–16% (Wong et al. 2006). This gene modification method seems to be more effective and attractive due to its permanent modulation of glycosylation. The method was also more cost-effective than the previous methods. Thus, there is great interest in improving the sialylation of biopharmaceutical products in CHO cells by gene manipulation.

Gene amplification defined as “the selective, repeated replication of a certain gene or genes without a proportional increase in other genes in the genome” is a well-known method of enhancing protein production (Smith et al. 1997). Meanwhile, the DHFR/MTX method is known as an amplification concept first used in mammalian cells (Omasa 2002). This method is by far the most widely used in CHO cells to improve biopharmaceutical production (Takagi et al. 2017). Thus, DHFR/MTX systems used for altering glycosylation remain unfamiliar. Here, we believe that this gene amplification system has high potential to be a great tool to boost sialylation profiles by greatly increasing Neu5Ac-based key enzymes. To our knowledge, the present study is the first to establish host cell lines that overexpress two Neu5Ac-based key genes in the same vector using a gene amplification method. In this study, we utilized the effect of MTX on increasing the expression levels of genes of interest. By supplementation with MTX to overexpress ST6GAL-I and GNE, the total proteins of the highly MTX-resistant cell lines CHOmt17-100, CHOmt17-100, and CHOmt17-300 showed considerably greater sialylation patterns, up to 1.5-fold compared to that of MTX-free CHOmt17-0. This result suggested that CHOmt17-100 might be a good choice for protein production. Gene amplification of DHFR/MTX has been demonstrated an efficient method in biopharmaceutical production since the 1980s (Kaufman and Sharp 1982). Goh and colleagues (2014) used DHFR/MTX to restore GnT-I in GnT-I-deficient CHO cells. The sialylated structures of EPO glycans produced in the obtained CHO cell line were 25% more abundant than in wild-type cells in laboratory conditions. With industrial-scale conditions, EPO sialylation was 2.5-fold higher than in laboratory conditions. In another study, with a similar approach of overexpression of ST6GAL-I to obtain a CHO host cell line, the IgG produced contained twice the total sialic acid content compared to the control (Lin et al. 2015). To our knowledge, the present study is the first to overexpress two Neu5Ac-based key enzymes, ST6GAL-I and GNE, using DHFR/MTX for gene amplification in CHO cell lines. The results obtained included noticeably escalated sialylation patterns of antibodies from modified CHOmt17-100 and more than a 14.5 times increase, with 53% of glycoforms being sialylated, compared to the wild type. Furthermore, the Neu5Ac terminal was mostly found in the glycan patterns while the NeuGc terminal remained unchanged compared to the wild type. This method may be able to confirm both the roles of ST6GAL-I and GNE in Neu5Ac production. These contributions might help engineered CHOmt17-100 cells more favorably host cell lines, since NeuGc is considered an immunogenic factor in human (Higashi et al. 1977; Merrick et al. 1978). Moreover, as shown in Fig. 2, CHOmt17-100 viable cell density was not substantially changed compared to CHOmt17-0. This indicates that 100 nM MTX might not reduce the viable cell density of this engineered CHO cell line. The antibodies produced in CHOmt17-0 and CHOmt17-100 both peaked on day 5, at approximately 1 g/L (Fig. 4). Thus, MTX supplementation might not alter the production of antibodies in CHOmt17-100. However, when we increased the concentration of MTX, the expression level of genes of interest did not show a big upregulation in CHOmt17-200 and CHOmt17-300. Similarly, this might result in the unchanged in the composition of the total sialylation structure of total proteins in CHOmt17-200, CHOmt17-300 as compared to CHOmt17-100. The growth rate of CHOmt17-200 and CHOmt17-300 were slower as compared to the CHOmt17-100. Thus, CHOmt17-100 stands as better candidate for further investigation. We expressed IgG antibodies which has one N-glycosylation at the position Asn297 in the Fc region to see its potential improvement in sialylation in CHOmt17-100 as compared to one expressed in CHOmt17-0. Because glycan is very dynamic in the Fc region, glycan can be accessible to glycosylation enzymes (Barb et al. 2012, Ramasamy et al. 2005). As mentioned in our previous studies, the wild type CHO cells have α(2,3)-sialyltransferase enzymes not α(2,6)-sialyltransferase enzymes. The upregulation of sialic acid in engineered cell lines mainly caused by ST6GAL-I and GNE. Thus, the linkage derived from the increased sialylation of the engineered cells are α(2,6) linkage sialic acid. These studies might possibly explain the upregulation of sialylation due to the increase of Neu5Ac-based key enzymes expressions. In conclusion, CHOmt17-100 could be an attractive host cell line for biopharmaceutical production due to its superior characteristics. Also, the DHFR/MTX method was proved effective for stably amplifying the glycosylation genes in CHO cell lines.

Conclusion

With the present and previous results taken together, we believe that this gene amplification system has high potential to be a great tool to boost superior glycosylation patterns by greatly increasing levels of Neu5Ac-based key enzymes. We therefore believe that this research might be useful to strengthen the position of bio-better products.

Acknowledgements

This study was performed under the research project of Manufacturing Technology Association of Biologics and supported by Japan Agency for Medical Research and Development.

Abbreviations

ST6GAL-I

α(2,6)-Sialyltransferase

ADCC

Antibody-dependent cellular cytotoxicity

Asn

Asparagine

BSA

Bovine serum albumin

CHO

Chinese hamster ovary

DHFR

Dihydrofolate reductase

Fc

Fragment crystallizable

IgG

Immunoglobulin G

IRES

Internal ribosome entry site

MTX

Methotrexate

α MEM

Minimum essential medium eagle–alpha modification medium

LC–MS

Liquid chromatography–mass spectrometry

Neu5Ac

N-acetylneuraminic acid

NeuGc

N-glycolylneuraminic acid

PBS

Phosphate buffered saline

PA

Pyridylamine

RP-HPLC

Reverse-phase high performance liquid chromatography

GNE

UDP-GlcNAc 2-epimerase/ManNAc

Compliance with ethical standards

Conflict of interest

The authors declare that they have no competing interests.

Footnotes

This research was conducted by Thi Sam Nguyen in partial fulfilment of the requirements for a Ph. D.

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Thi Sam Nguyen and Ryo Misaki contributed equally to this work.

References

  1. Anthony RM, Nimmerjahn F, Ashline DJ, Reinhold VN, et al. Recapitulation of IVIG anti-inflammatory activity with a recombinant IgG Fc. Science. 2008;320:373–376. doi: 10.1126/science.1154315. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Baker KN, James DC. Metabolic control of recombinant protein N-glycan processing in NS0 and CHO cells. Biotechnol Prog. 2001;30:1419–1431. doi: 10.1126/science.1154315. [DOI] [PubMed] [Google Scholar]
  3. Barb AW, Meng L, Gao Z, Johnson RW, Moremen KW, Prestegard JH. NMR characterization of immunoglobulin G Fc glycan motion on enzymatic sialylation. Biochemistry. 2012;51:4618–4626. doi: 10.1021/bi300319q. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Beale D, Feinstein A. Structure and function of the constant regions of immunoglobulins. Q Rev Biophys. 1976;9:135–180. doi: 10.1016/j.jaci.2009.09.046. [DOI] [PubMed] [Google Scholar]
  5. Borys MC, Abu-Absi NR. Effects of culture conditions on N-glycolylneuraminic acid (Neu5Gc) content of a recombinant fusion protein produced in CHO cells. Biotechnol Bioeng. 2010;105:1048–1057. doi: 10.1002/bit.22644. [DOI] [PubMed] [Google Scholar]
  6. Burleigh SC, Stroop CJM. Synergizing metabolic flux analysis and nucleotide sugar metabolism to understand the control of glycosylation of recombinant protein in CHO cells. BMC Biotechnol. 2011;11:95. doi: 10.1186/1472-6750-11-95. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Burton DR, Dwek RA. Immunology. Sugar determines antibody activity. Science. 2006;313:627–628. doi: 10.1126/science.1131712. [DOI] [PubMed] [Google Scholar]
  8. Chee D, Gek M. Impact of dynamic online fed-batch strategies on metabolism, productivity and N-glycosylation quality in CHO cell cultures. Biotechnol Bioeng. 2005;89:164–177. doi: 10.1002/bit.20317. [DOI] [PubMed] [Google Scholar]
  9. Chotigeat W, Gray RE. Role of environmental conditions on the expression levels, glycoform pattern and levels of sialyltransferase for hFSH produced by recombinant CHO cells. Cytotechnology. 1994;15:217–221. doi: 10.1007/bf00762396. [DOI] [PubMed] [Google Scholar]
  10. Dechecchi MC, Melotti P, Bonizzato A, Santacatterina M. Heparan sulfate glycosaminoglycans are receptors sufficient to mediate the initial binding of adenovirus types 2 and 5. J Virol. 2001;75:8772–8780. doi: 10.1128/jvi.75.18.8772-8780.2001. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Goh JS, Liu Y, Liu H, Chan KF, Wan C, et al. Highly sialylated recombinant human erythropoietin production in large-scale perfusion bioreactor utilizing CHO-gmt4 (JW152) with restored GnT I function. Biotechnol J. 2014;9:100–109. doi: 10.1002/biot.201300301. [DOI] [PubMed] [Google Scholar]
  12. Higashi H, Naiki M, Matuo S, Okouchi K. Antigen of "serum sickness" type of heterophile antibodies in human sera: indentification as gangliosides with N-glycolylneuraminic acid. Biochem Biophys Res Commun. 1977;79:388–395. doi: 10.1016/0006-291x(77)90169-3. [DOI] [PubMed] [Google Scholar]
  13. Jayapal KP, Wlaschin KF, Hu W, Yap MG. Recombinant protein therapeutics from CHO cells—20 years and counting. Chem Eng Prog. 2007;103:40–47. doi: 10.1002/9780470571224.pse506. [DOI] [Google Scholar]
  14. Jeong YT, Kim JH. Enhanced sialylation of recombinant erythropoietin in CHO cells by human glycosyltransferase expression. J Microbiol Biotechnol. 2008;18:1945–1952. doi: 10.4014/jmb.0800.546. [DOI] [PubMed] [Google Scholar]
  15. Kaneko Y, Nimmerjahn F, Ravetch JV. Anti-inflammatory activity of immunoglobulin G resulting from Fc sialylation. Science. 2006;313:670–673. doi: 10.1126/science.1129594. [DOI] [PubMed] [Google Scholar]
  16. Kaufman RJ, Sharp PA. Amplification and expression of sequences cotransfected with a modular dihydrofolate reductase complementary dna gene. J Mol Biol. 1982;159:601–621. doi: 10.1016/0022-2836(82)90103-6. [DOI] [PubMed] [Google Scholar]
  17. Kondo A, Suzuki J, Kuraya N, Hase S, Kato I, Ikenaka T. Improved method for fluorescence labeling of sugar chains with sialic acid residues. Agric Biol Chem. 1990;54:2169–2170. doi: 10.1080/00021369.1990.10870232. [DOI] [PubMed] [Google Scholar]
  18. Krapp S, Mimura Y, Jefferis R, Huber R, Sondermann P. Structural analysis of human IgG-Fc glycoforms reveals a correlation between glycosylation and structural integrity. J Mol Biol. 2003;325:979–989. doi: 10.1016/s0022-2836(02)01250-0. [DOI] [PubMed] [Google Scholar]
  19. Kurogochi M, Mori M, Osumi K, Tojino M, et al. GlyGlycoengineered monoclonal antibodies with homogeneous glycan (M3, G0, G2, and A2) using a chemoenzymatic approach have different affinities for FcγRIIIa and variable antibody-dependent cellular cytotoxicity activities. PLoS ONE. 2015;10:e0132848. doi: 10.1371/journal.pone.0132848. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Limkul J, Iizuka S, Sato Y, Misaki R, Ohashi T, Ohashi T, Fujiyama K. The production of human glucocerebrosidase in glyco-engineered Nicotiana benthamiana plants. Plant Biotechnol J. 2016;14:1682–1694. doi: 10.1111/pbi.12529. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Lin N, Mascarenhas J, Sealover NR, George HJ, Brooks J, et al. Chinese hamster ovary (CHO) host cell engineering to increase sialylation of recombinant therapeutic proteins by modulating sialyltransferase expression. Biotechnol Prog. 2015;31:334–346. doi: 10.1002/btpr.2038. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Lipscomb ML, Palomares LA, Hernández V, Ramírez OT, Kompala DS. Effect of production method and gene amplification on the glycosylation pattern of a secreted reporter protein in CHO cells. Biotechnol Prog. 2005;21:40–49. doi: 10.1021/bp049761m. [DOI] [PubMed] [Google Scholar]
  23. Liu J, Fan L. Galactose supplementation enhance sialylation of recombinant Fc-fusion protein in CHO cell: an insight into the role of galactosylation in sialylation. World J Microbiol Biotechnol. 2015;31:1147–1156. doi: 10.1007/s11274-015-1864-8. [DOI] [PubMed] [Google Scholar]
  24. Louis KS, Siegel AC, Levy GA. Comparison of manual versus automated trypan blue dye exclusion method for cell counting. In: Stoddart MJ, editor. Mammalian cell viability: methods and protocols. New York: Springer Protocols; 2011. pp. 7–12. [Google Scholar]
  25. Lund J, Takahashi N, Pound JD, Goodall M, Jefferis R. Multiple interactions of IgG with its core oligosaccharide can modulate recognition by complement and human Fc gamma receptor I and influence the synthesis of its oligosaccharide chains. J Immunol. 1996;157:4963–4969. [PubMed] [Google Scholar]
  26. Merrick JM, Zadarlik K, Milgrom F. Characterization of the Hanganutziu-Deicher (serum-sickness) antigen as gangliosides containing n-glycolylneuraminic acid. Int Arch Allergy Appl Immunol. 1978;57:477–480. doi: 10.1159/000232140. [DOI] [PubMed] [Google Scholar]
  27. Ngantung FA, Wang DIC. RNA interference of sialidase improves glycoprotein sialic acid content consistency. Biotechnol Bioeng. 2006;95:106–119. doi: 10.1002/bit.20997. [DOI] [PubMed] [Google Scholar]
  28. Omasa T. Gene amplification and its application in cell and tissue engineering. J Biosci Bioeng. 2002;94:600–605. doi: 10.1016/S1389-1723(02)80201-8. [DOI] [PubMed] [Google Scholar]
  29. Raju TS. Glycosylation variations with expression systems and their impact on biological activity of therapeutic immunoglobulins. Bioprocess Int. 2003;1:44–53. [Google Scholar]
  30. Raju TS, Scallon B. Glycosylation in the Fc domain of IgG increases resistance to proteolytic cleavage by papain. Biochem Biophys Res Commun. 2006;341:797–803. doi: 10.1016/j.bbrc.2006.01.030. [DOI] [PubMed] [Google Scholar]
  31. Raju TS, Scallon B. Fc glycans terminated with N-acetylglucosamine residues increase antibody resistance to papain. Biotechnol Prog. 2007;23:964–971. doi: 10.1021/bp070118k. [DOI] [PubMed] [Google Scholar]
  32. Ramasamy V, Ramakrishnan B, Boeggeman E, Ratner DM, Seeberger PH, Qasba PK. Oligosaccharide preferences of beta1,4-galactosyltransferase-I: crystal structures of Met340His mutant of human beta1,4-galactosyltransferase-I with a pentasaccharide and trisaccharides of the N-glycan moiety. J Mol Biol. 2005;353:53–67. doi: 10.1016/j.jmb.2005.07.050. [DOI] [PubMed] [Google Scholar]
  33. Ryan MH, Petrone D, Nemeth JF, Barnathan E, Björck L, Jordan RE. Proteolysis of purified IgGs by human and bacterial enzymes in vitro and the detection of specific proteolytic fragments of endogenous IgG in rheumatoid synovial fluid. Mol Immunol. 2008;45:1837–1846. doi: 10.1016/j.molimm.2007.10.043. [DOI] [PubMed] [Google Scholar]
  34. Shields RL, Lai J, Keck R, O'Connell LY, Hong K, Meng YG, Weikert SH, Presta LG. Lack of fucose on human IgG1 N-linked oligosaccharide improves binding to human Fc gamma RIII and antibody-dependent cellular toxicity. J Biol Chem. 2002;277:26733–26740. doi: 10.1074/jbc.M202069200. [DOI] [PubMed] [Google Scholar]
  35. Shinkawa T, Nakamura K, Yamane N, Shoji-Hosaka E, et al. The absence of fucose but not the presence of galactose or bisecting N-acetylglucosamine of human IgG1 complex-type oligosaccharides shows the critical role of enhancing antibody-dependent cellular cytotoxicity. J Biol Chem. 2003;278:3466–3473. doi: 10.1074/jbc.M210665200. [DOI] [PubMed] [Google Scholar]
  36. Shukla D, Spear PG. A novel role for 3-O-sulfated heparan sulfate in herpes simplex virus 1 entry. Cell Press. 1999;99:13–22. doi: 10.1016/s0092-8674(00)80058-6. [DOI] [PubMed] [Google Scholar]
  37. Smith AD, Datta SP, Smith GH, Campbell PN, Bentley R, McKenzie HA. Oxford dictionary of biochemistry and molecular biology. New York: Oxford University Press; 1997. [Google Scholar]
  38. Strome SE, Sausville EA, Mann D. A mechanistic perspective of monoclonal antibodies in cancer therapy beyond target-related effects. Oncologist. 2007;12:1084–1095. doi: 10.1634/theoncologist.12-9-1084. [DOI] [PubMed] [Google Scholar]
  39. Takagi Y, Masuda K, Masuda K, Nishii S, Kawakami B, Omasa T. Identification of regulatory motifs in the CHO genome for stable monoclonal antibody production. Cytotechnology. 2017;69:451–460. doi: 10.1007/s10616-016-0017-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Thi Sam N, Misaki R, Ohashi T, Fujiyama K. Enhancement of glycosylation by stable co-expression of two sialylation enzymes on Chinese hamster ovary cells. J Biosci Bioeng. 2018;126:102–110. doi: 10.1016/j.jbiosc.2018.01.010. [DOI] [PubMed] [Google Scholar]
  41. Trummer E, Fauland K, Seidinger S, Schriebl K, Lattenmayer C, et al. Process parameter shifting: part I. Effect of DOT, pH, and temperature on the performance of Epo-Fc expressing CHO cells cultivated in controlled batch bioreactors. Biotechnol Bioeng. 2006;94:1033–1044. doi: 10.1002/bit.21013. [DOI] [PubMed] [Google Scholar]
  42. Varki A, Cummings RD, Esko JD, Etzler ME. Structural basis of glycan diversity. Essentials of glycobiology. 2. Cold Spring Harbor (NY): Cold Spring Harbor Laboratory Press; 2009. [PubMed] [Google Scholar]
  43. Wong NS, Yap MG, Wang DI. Enhancing recombinant glycoprotein sialylation through CMP-sialic acid transporter over expression in Chinese hamster ovary cells. Biotechnol Bioeng. 2006;93:1005–1016. doi: 10.1002/bit.20815. [DOI] [PubMed] [Google Scholar]
  44. Wong NS, Wati L, Nissom PM, Feng HT, Lee MM, Yap MG. An investigation of intracellular glycosylation activities in CHO cells: effects of nucleotide sugar precursor feeding. Biotechnol Bioeng. 2010;107:321–336. doi: 10.1002/bit.22812. [DOI] [PubMed] [Google Scholar]
  45. Yang M, Bulter M. Effects of ammonia on CHO cell growth erythropoietin production, and glycosylation. Biotechnol Bioeng. 2000;68:370–380. doi: 10.1002/(SICI)10970290(20000520)68:4&#x0003c;370::AIDBIT2&#x0003e;3.0.CO;2-K. [DOI] [PubMed] [Google Scholar]
  46. Yang M, Butler M. Effects of ammonia and glucosamine on the heterogeneity of erythropoietin glycoforms. Biotechnol Prog. 2000;18:129–138. doi: 10.1021/bp0101334. [DOI] [PubMed] [Google Scholar]
  47. Zhang M, Koskie K, Ross JS, Kayser KJ, Caple MV. Enhancing glycoprotein sialylation by targeted gene silencing in mammalian cells. Biotechnol Bioeng. 2010;105:1094–1105. doi: 10.1002/bit.22633. [DOI] [PubMed] [Google Scholar]

Articles from Cytotechnology are provided here courtesy of Springer Science+Business Media B.V.

RESOURCES