A streamlined implementation of the glutamine synthetase-based protein expression system
© Knox et al.; licensee BioMed Central Ltd. 2013
Received: 3 July 2013
Accepted: 10 September 2013
Published: 24 September 2013
The glutamine synthetase-based protein expression system is widely used in industry and academia for producing recombinant proteins but relies on the cloning of transfected cells, necessitating substantial investments in time and handling. We streamlined the production of protein-producing cultures of Chinese hamster ovary cells using this system by co-expressing green fluorescent protein from an internal ribosomal entry site and selecting for high green fluorescent protein-expressing cells using fluorescence-activated cell sorting.
Whereas other expression systems utilizing green fluorescent protein and fluorescence-activated cell sorting-based selection have relied on two or more sorting steps, we obtained stable expression of a test protein at levels >50% of that of an “average” clone and ~40% that of the “best” clone following a single sorting step. Versus clone-based selection, the principal savings are in the number of handling steps (reduced by a third), handling time (reduced by 70%), and the time needed to produce protein-expressing cultures (reduced by ~3 weeks). Coupling the glutamine synthetase-based expression system with product-independent selection in this way also facilitated the production of a hard-to-assay protein.
Utilizing just a single fluorescence-activated cell sorting-based selection step, the new streamlined implementation of the glutamine synthetase-based protein expression system offers protein yields sufficient for most research purposes, where <10 mg/L of protein expression is often required but relatively large numbers of constructs frequently need to be trialed.
KeywordsProtein expression Glutamine synthetase Chinese hamster ovary cells IRES HEK 293S
Mammalian cells are useful for stably expressing recombinant proteins for use in structural and functional studies, as well as for the industrial production of, e.g., therapeutic antibodies and cytokines [1–3]. Mammalian cell-based expression systems are essential when the protein of interest cannot be expressed in bacterial- or yeast-based systems and/or when conventional glycosylation is needed for the folding or stability of the protein (reviewed in ). Establishing stable cell lines expressing a given protein typically involves transfection with plasmid vectors carrying the gene of interest and a selection marker [5–7]. Large numbers of resistant clones, often isolated in a multi-well plate format, are then screened to identify high expressers. This process is labor-intensive, time-consuming and limited by the number of clones that can feasibly be screened.
Considerable effort has therefore gone into developing selection strategies requiring reduced screening effort [8–10]. In particular, the development of fluorescence-activated cell sorting (FACS) protocols has significantly increased the throughput of selection using co-expressed fluorescent reporter proteins, e.g. green fluorescent protein (GFP), as second selectable markers . Previously, implementation of this approach has involved either two  or more (up to five)  rounds of FACS selection of the GFP-expressing cells, resulting in these methods still being labor-intensive and taking six months or longer. This effort is justified in the context of the industrial expression of therapeutic proteins, where production can be scaled and repeated indefinitely. For research purposes, however, where milligram quantities of protein may only be required on a one-off basis, faster and less labor-intensive solutions are needed.
We are long-term users of the glutamine synthetase (GS)-based protein expression system, developed by Lonza Biologics, which utilizes a robust viral promoter and selection via glutamine metabolism to allow the generation of high-yielding and stable cell lines derived from Chinese hamster ovary (CHO) cells, the major mammalian host for recombinant protein production [6, 14]. We previously established cell lines producing ~400 mg/L of a soluble form of the T-cell surface protein, CD4 , and yields as high as 5 g/L of antibody have been reported by others in commercial settings . The GS system utilizes the plasmid vector pEE14, which carries the gene of interest and encodes a GS mini-gene. Transfected cells are selected in the presence of graded amounts of the competitive GS inhibitor methionine sulphoximine (MSX), which allows the isolation of cells with very high plasmid copy numbers (>2000/cell ). However, CHO cells also readily amplify their own GS gene, necessitating the isolation and screening of single clones, adding 1–2 months to the generation of a high-expressing cell line.
We previously noticed that the expression levels of the top ~50% of protein-expressing clones are generally relatively uniform, which suggested that if weakly expressing clones could be removed along with untransfected resistant cells that had amplified their endogenous GS gene, clone selection might be unnecessary. Here, using both MSX selection and single-step fluorescence-activated cell sorting (FACS) for high co-expression of a green fluorescent protein marker, we establish a streamlined protocol in which cloning is eliminated. With the new method, the transfection-to-protein-purification stages can be completed in just two months. We also show that coupling the GS-based expression system with product-independent selection facilitates the high-level production of hard-to-assay proteins.
In T2A-eGFP-GS-pOPINEE12G (Figure 1C), expression of the eGFP is linked to the gene of interest via a self-cleaving 2A peptide from the insect virus Thosea asigna (T2A) . This 18 amino acid sequence was included in the 5’ oligonucleotide used to amplify the eGFP cDNA from pHR-IRES-eGFP. The stop codon between the gene of interest and T2A was omitted. The Dual promoter-GS-pOPINEE12G constructs (Figure 1B) were constructed by cloning the FcERα gene into the KpnI/PmeI sites of pOPINEE12G. eGFP under control of a second hCMV promoter was then added downstream of the FcERα gene by first inserting eGFP into the HindIII/EcoRI sites of pEE6  and then PCR cloning the resulting hCMV promoter–eGFP cassette into the NotI/SalI sites of pOPINEE12G. FcERα was also cloned via XbaI into pEE14 , the traditional GS system vector.
For making stable FcERα-expressing HEK 293S cells, the GS cDNA in FcERα-IRES-GS-pOPINEE12G was replaced with a bacterial aminoglycoside phosphotransferase 3’ II (Neo) gene, which confers resistance to aminoglycoside antibiotics. A transcription unit comprising the Neo gene under control of the SV40 early promoter was amplified from pcDNA-DEST40 (Invitrogen, Paisley, UK) and cloned into the NheI/BglII sites of FcERα-IRES-GS-pOPINEE12G creating FcERα-IRES-eGFP-Neo-pOPINEE12G, which enables the selection of transfectants with Geneticin (G418; Sigma Aldrich Company Ltd., Gillingham, UK).
Cell culture and transfection
CHO-K1 cells were grown at 37°C, 5% CO2 in high-glucose Dulbecco's modified Eagle's medium (DMEM; Gibco, Invitrogen) supplemented with 10% fetal calf serum (Sigma Aldrich Company Ltd.), 1% L-glutamine (Sigma Aldrich Company Ltd.), 1% sodium pyruvate (Invitrogen) and an amino acid supplement. For stable transfection, 106 cells (unless otherwise stated) were seeded in a 75 cm2 flask. The following day the medium was changed to DMEM supplemented with 10% FCS dialysed against PBS (First Link Ltd., Wolverhampton, UK), 1% sodium pyruvate and amino acids, before the cells were transfected with 10 μg DNA using Genejuice (Novagen, Merck Chemicals Ltd., Hoddesdon, UK) according to the manufacturer’s protocol. The following day, L-methionine sulfoximine (MSX; Sigma Aldrich Company Ltd.) was added to the medium at concentrations ranging from 20–50 μM. Medium was refreshed 5 days after transfection, and again after a further week. Upon the appearance of substantial numbers of clones, cells were removed with Accutase Cell Dissocation Reagent (Gibco, Invitrogen). Cell numbers and viability were assayed by trypan blue exclusion. Cells were spun down and resuspended in PBS, ready for FACS. The FACS-sorted cells were then expanded in MSX-containing medium as above. Once cells were confluent in cell factories or final-stage 175 cm2 flasks, sodium butyrate was added at a concentration of 2 mM.
HEK 293S cells were grown in high-glucose DMEM supplemented with 10% FCS and 1% L-glutamine. Transfection was carried out as above, using G418 disulphide salt (Invitrogen) at a concentration of 0.8 mg/mL to select for Geneticin-resistant clones.
Flow cytometry and preparative FACS
GFP expression of transfected CHO and HEK 293S cells was monitored via flow cytometry on a CyAn ADP Analyzer (Beckman Coulter, Krefeld, Germany). Preparative FACS was performed on a MoFlo high-speed cell sorter (Beckman Coulter). The argon-ion laser was tuned to 488 nm with 100 mW of power, and eGFP fluorescence detected in FL1 through a 530/40-nm bandpass filter. The top 30% (unless otherwise stated) of live eGFP-expressing cells were sorted into a single tube, stored on ice. Data analysis was performed using FlowJo software (Tree Star Inc., Ashland, OR, USA).
FcERα receptor yields were determined by competition ELISA. Supernatant was sampled from confluent 175 cm2 flask cultures of FACS-sorted cells, three weeks after the addition of sodium butyrate. ELISA plates (Costar, Corning Incorportated, New York) were coated with 50 μL purified FcERα at 10 μg/mL and incubated at 4°C overnight. The next day, this was removed and the plate washed three times with PBS 0.05% Tween 20 (Sigma Aldrich Company Ltd.), before blocking with 100 μL PBS 1% casein (VWR, Lutterworth, UK) for 30 minutes at room temperature. Meanwhile, competition mixtures consisting of 55 μL of serially titrated sample or standard (purified FcERα), and 55 μL mouse anti-FcERα antibody (AbCam, Cambridge, UK) at 3.3 mg/L were prepared, and incubated at 37°C for 30 minutes. After washing the plate as before, 50 μL competition mixture was then added to each well, and the plate was incubated at 4°C for 1 hour. After another wash, 50 μL hydrogen peroxidase-coupled goat anti-mouse IgG Fc (Sigma Aldrich Company Ltd., diluted 1 in 2000 in DMEM) was added, and the plate incubated at 4°C for a further hour. Peroxidase detection was via TMB substrate (Thermo Scientific, Hemel Hempstead, UK) according to the manufacturer’s protocol. FcERα titre was determined by plotting the absorbance of titrated samples, and reading off the dilution factor at 50% inhibition, compared to the standard. Where necessary the PD-1 yield was similarly determined by competition ELISA, using mouse anti-human-PD-1 antibody clone 2 (unpublished data, S. Morgan et. al.) at 2.5 mg/L, and HRP-coupled anti-mouse IgG Fc as above.
For identifying peak fractions containing CCL18, the sandwich ELISA method embodied in the Human CCL18/PARC Quantikine ELISA kit (R&D systems, Inc., Minneapolis) was employed. The kit was used according to the manufacturer’s instructions.
Purification with Ni-NTA column
The soluble His-tagged proteins were purified from the supernatant on a Ni-NTA column followed by gel filtration as previously described .
Results and discussion
Optimization of FACS-based selection
We set out to establish a shortened protocol for generating stable CHO cell cultures expressing proteins of interest using the glutamine synthetase-based gene expression system. Our principal test protein was the Type I IgE-specific Fc receptor (FcERα), which is involved in the control of allergic responses [23, 24]. We compared three approaches for co-expressing the eGFP selectable marker (see below), but initial optimization of the method was based on a vector that expressed eGFP from an internal ribosome entry site (IRES). The gene encoding a soluble form of FcERα (residues 26–201)  was cloned into the pOPINEE12G expression vector downstream of a start codon and sequence encoding a heterologous signal peptide, and upstream of sequence encoding an IRES and Emerald GFP (eGFP) reporter (residues 1–240), giving FcERα-IRES-eGFP-GS-pOPINEE12G (Figure 1A). Expression of both genes is in this way controlled by the human cytomegalovirus promoter, with the downstream position of the eGFP reporter ensuring that its expression will be lower than that of FcERα. Following transfection of CHO-K1 cells in a 75 cm2 flask with this vector, the cells were left undisturbed for 3 weeks, after which substantial numbers of clones were visible by eye. In preliminary experiments using 40 μM MSX, a low seeding density of 106 cells/flask produced the largest numbers of resistant clones (>200) (Additional file 2: Figure S1); cells at higher initial densities overgrew in the first week when MSX selection was presumably beginning to take effect.
Comparison of methods for co-expression of eGFP
Productivity of FACS-selected lines versus clones
The question arises of how the yields from the FACS-selected lines compare with those of clones generated with FcERα-IRES-eGFP-GS-pOPINEE12G. We also wanted to make comparisons with clones obtained using pEE14, the traditional vector used with CHO cells, which utilizes a GS mini-gene  rather than a GS cDNA. The FcERα gene was cloned into pEE14 and stable clones generated with this vector and FcERα-IRES-eGFP-GS-pOPINEE12G. The six best-expressing clones in each case, based on initial dot-blot analysis as described in , were expanded to 175 cm2 flask cultures, treated with sodium butyrate and left for 3 weeks. The average expression of the pEE14-derived but not the pOPINEE12G-derived clones was significantly higher (p = 0.002) than that of the FACS-selected lines, but this amounted to a less than 1.7-fold difference in expression (Figure 5B). Similarly, the best of the clones expressed the protein only ~1.6 fold better than the best of the sorted lines. However, the most important comparison is between the best of the clones and the average expression of the FACS-selected lines, assuming that multiple transfections will be pooled prior to sorting in order to reduce the effects of transfection variability. In this case the best clone performs 2.5-fold better than the pOPINEE12G-derived FACS-selected line (taking the average of the three replicates). It needs to be borne in mind, however, that if needed, better expressing lines could be obtained by FACS-selecting the top 10% of expressers. It is possible that the synthesis of an additional protein (eGFP) may burden the cell machinery in the case of the FcERα-IRES-eGFP-GS-pOPINEE12G transfected clones and lines, accounting for the reduced expression .
Utility of the method in other selection systems
An example: CCL18
In using the new method the trade-off between speed and expression levels could nevertheless be further considered. When low-expressing proteins are being studied, higher percentage sorts may be suitable, and in extreme cases, single-cell sorting could be applied to generate a highly expressing clone. One slight drawback with the new method is its dependence on the efficiency of transfection, which we generally find to be variable on a flask-to-flask basis over extended culture periods. However, combining replicate transfected flasks easily circumvents this issue. A final matter is that sorting for high expression is only worthwhile when there is large variability of expression among the selected cells. In the case of G418 selection, for which the levels of expression among GFP-positive cells was very uniform, almost no advantage was gained by sorting the GFP expressing cells, whereas substantially increased expression was obtainable by conventional cloning. Although unsuitable, of course, for prolonged expression of therapeutic proteins on an industrial scale, this method provides yields more than sufficient for research purposes, where less than 10 mg/L of protein is generally required and often large numbers of constructs have to be trialed.
The authors wish to thank Craig Waugh for performing the FACS sorts, Sara Morgan and Heather Brouwer for help with ELISA design, and Lonza Biologics for the licensing of the GS-CHO system. The OPPF-UK is funded by the Medical Research Council and the Biotechnology and Biological Sciences Research Council. The Wellcome Trust Centre for Human Genetics is supported by the Wellcome Trust (Grant no. 075491/Z/04).
- Davies A, Greene A, Lullau E, Abbott WM: Optimisation and evaluation of a high-throughput mammalian protein expression system. Protein Expr Purif. 2005, 42: 111-21. 10.1016/j.pep.2005.03.012.View ArticleGoogle Scholar
- Aricescu AR, Lu W, Jones EY: A time- and cost-efficient system for high-level protein production in mammalian cells. Acta Crystallogr D Biol Crystallogr. 2006, 62: 1243-50.View ArticleGoogle Scholar
- Matentzoglu K, Scheffner M: Ubiquitin-fusion protein system: a powerful tool for ectopic protein expression in mammalian cells. Biotechniques. 2009, 46: 21-2. 10.2144/000113023.View ArticleGoogle Scholar
- Davis SJ, Crispin MD: Solutions to the glycosylation problem for low- and high-throughput structural glycoproteomics. Functional and Structural Proteomics of Glycoproteins. Edited by: Owens RJ, Nettleship JE. 2011, Netherlands: Springer VerlagGoogle Scholar
- Deen KC, McDougal JS, Inacker R, Folena-Wasserman G, Arthos J, Rosenberg J, Maddon PJ, Axel R, Sweet RW: A soluble form of CD4 (T4) protein inhibits AIDS virus infection. Nature. 1988, 331: 82-4. 10.1038/331082a0.View ArticleGoogle Scholar
- Davis SJ, Ward HA, Puklavec MJ, Willis AC, Williams AF, Barclay AN: High level expression in Chinese hamster ovary cells of soluble forms of CD4 T lymphocyte glycoprotein including glycosylation variants. J Biol Chem. 1990, 265 (18): 10410-18.Google Scholar
- Newman-Tancredi A, Wootton R, Strange PG: High-level stable expression of recombinant 5-HT1A 5-hydroxytryptamine receptors in Chinese hamster ovary cells. Biochem J. 1992, 285: 933-8.View ArticleGoogle Scholar
- Browne SM, Al-Rubeai M: Selection methods for high-producing mammalian cell lines. Trends Biotechnol. 2007, 25 (9): 425-432. 10.1016/j.tibtech.2007.07.002.View ArticleGoogle Scholar
- Matasci M, Baldi L, Hacker DL, Wurm FM: The PiggyBac transposon enhances the frequency of CHO stable cell line generation and yields recombinant lines with superior productivity and stability. Biotechnol Bioeng. 2011, 108: 2141-2150. 10.1002/bit.23167.View ArticleGoogle Scholar
- Oberbek A, Matasci M, Hacker DL, Wurm FM: Generation of stable, high-producing CHO cell lines by lentiviral vector-mediated gene transfer in serum-free suspension culture. Biotechnol Bioeng. 2011, 108: 600-610. 10.1002/bit.22968.View ArticleGoogle Scholar
- Meng YG, Liang J, Wong WL, Chisholm V: Green fluorescent protein as a second selectable marker for selection of high producing clones from transfected CHO cells. Gene. 2000, 242: 201-207. 10.1016/S0378-1119(99)00524-7.View ArticleGoogle Scholar
- Mancia F, Patel SD, Rajala MW, Scherer PE, Nemes A, Schieren I, Hendrickson WA, Shapiro L: Optimisation of protein production in mammalian cells with a co-expressed fluorescent marker. Structure. 2004, 12: 1355-1360. 10.1016/j.str.2004.06.012.View ArticleGoogle Scholar
- Wilke S, Krause J, Gossen M, Groebe L, Jäger V, Gherardi E, van den Heuvel J, Büssow K: Glycoprotein production for structure analysis with stable, glycosylation mutant CHO cell lines established by fluorescence-activated cell sorting. Protein Sci. 2010, 19: 1264-1271. 10.1002/pro.390.View ArticleGoogle Scholar
- Cockett MI, Bebbington CR, Yarranton GT: High level expression of tissue inhibitor of metalloproteinases in Chinese hamster ovary cells using glutamine synthetase gene amplification. Biotechnology (NY). 1990, 8 (7): 662-667. 10.1038/nbt0790-662.View ArticleGoogle Scholar
- Lindgren K, Salmén A, Lundgren M, Bylund L, Ebler A, Fäldt E, Sörvik L, Fenge C, Skoging-Nyberg U: Automation of cell line development. Cytotechnology. 2009, 59: 1-10. 10.1007/s10616-009-9187-y.View ArticleGoogle Scholar
- Bebbington CR, Hentschel CCG: The use of vectors based on gene amplification for the expression of cloned genes in mammalian cells. DNA Cloning, Volume III. Edited by: Glover D. 1987, San Diego: Academic Press, 163-188.Google Scholar
- Bebbington CR, Renner G, Thomson S, King D, Abrams D, Yarranton GT: High-level expression of a recombinant antibody from myeloma cells using a glutamine synthetase gene as an amplifiable selectable marker. Biotechnology (NY). 1992, 2: 169-175.View ArticleGoogle Scholar
- Berrow NS, Alderton D, Sainsbury S, Nettleship J, Assenberg R, Rahman N, Stuart DI, Owens RJ: A versatile ligation-independent cloning method suitable for high-throughput expression screening applications. Nucleic Acids Res. 2007, 35 (6): 45-10.1093/nar/gkm047.View ArticleGoogle Scholar
- Bainbridge JW, Stephens C, Parsley K, Demaison C, Halfyard A, Thrasher AJ, Ali RR: In vivo gene transfer to the mouse eye using an HIV-based lentiviral vector; efficient long-term transduction of corneal endothelium and retinal pigment epithelium. Gene Ther. 2001, 8: 1665-1668. 10.1038/sj.gt.3301574.View ArticleGoogle Scholar
- Szymczak AL, Workman CJ, Wang Y, Vignali KM, Dilioglou S, Vanin EF, Vignali DAA: Correction of multi-gene deficiency in vivo using a single 'self-cleaving' 2A peptide–based retroviral vector. Nat Biotechnol. 2004, 22: 589-594. 10.1038/nbt957.View ArticleGoogle Scholar
- Stephens PE, Cockett MI: The construction of a highly efficient and versatile set of mammalian expression vectors. Nucleic Acids Res. 1989, 17 (17): 7110-10.1093/nar/17.17.7110.View ArticleGoogle Scholar
- Ikemizu S, Gilbert RJ, Fennelly JA, Collins AV, Harlos K, Jones EY, Stuart DI, Davis SJ: Structure and dimerization of a soluble form of B7-1. Immunity. 2000, 12: 51-60. 10.1016/S1074-7613(00)80158-2.View ArticleGoogle Scholar
- Abramson J, Pecht I: Regulation of the mast cell response to the type 1 Fc epsilon receptor. Immunol Rev. 2007, 217: 231-54. 10.1111/j.1600-065X.2007.00518.x.View ArticleGoogle Scholar
- Gould HJ, Sutton BJ: IgE in allergy and asthma today. Nat Rev Immunol. 2008, 8: 205-17. 10.1038/nri2273.View ArticleGoogle Scholar
- Holdom MD, Davies AM, Nettleship JE, Bagby SC, Dhaliwal B, Girardi E, Hunt J, Gould HJ, Beavil AJ, McDonnell JM, Owens RJ, Sutton BJ: Conformational changes in IgE contribute to its uniquely slow dissociation rate from receptor FcεRI. Nat Struct Mol Biol. 2011, 18 (5): 571-6. 10.1038/nsmb.2044.View ArticleGoogle Scholar
- Gorman CM, Howard BH, Reeves R: Expression of recombinant plasmids in mammalian cells is enhanced by sodium butyrate. Nucleic Acids Res. 1983, 11: 7631-7648. 10.1093/nar/11.21.7631.View ArticleGoogle Scholar
- Shinohara T, Taniwaki M, Ishida Y, Kawaichi M, Honjo T: Structure and chromosomal localization of the human PD-1gene (PDCD1). Genomics. 1994, 23: 704-706. 10.1006/geno.1994.1562.View ArticleGoogle Scholar
- Colbère-Garapin F, Horodniceanu F, Kourilsky P, Garapin AC: A new dominant hybrid selective marker for higher eukaryotic cells. J Mol Biol. 1981, 150: 1-14. 10.1016/0022-2836(81)90321-1.View ArticleGoogle Scholar
- Reeves PJ, Callewaert N, Contreras R, Khorana HG: Structure and function of rhodopsin with restricted and homogeneous N-glycosylation by a tetracycline-inducible N-acetylglucosaminyltransferase I-negative HEK293S stable mammalian cell line. Proc Natl Acad Sci USA. 2002, 99: 13419-13424. 10.1073/pnas.212519299.View ArticleGoogle Scholar
- Schutyser E, Richmond A, Van Damme J: Involvement of CC chemokine ligand 18 (CCL18) in normal and pathological processes. J Leukoc Biol. 2005, 78 (1): 14-26. 10.1189/jlb.1204712.View ArticleGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.