Publisher’s version / Version de l'éditeur: Journal of Biotechnology, 2018-06-07
READ THESE TERMS AND CONDITIONS CAREFULLY BEFORE USING THIS WEBSITE.
https://nrc-publications.canada.ca/eng/copyright
Vous avez des questions? Nous pouvons vous aider. Pour communiquer directement avec un auteur, consultez la première page de la revue dans laquelle son article a été publié afin de trouver ses coordonnées. Si vous n’arrivez pas à les repérer, communiquez avec nous à PublicationsArchive-ArchivesPublications@nrc-cnrc.gc.ca.
Questions? Contact the NRC Publications Archive team at
PublicationsArchive-ArchivesPublications@nrc-cnrc.gc.ca. If you wish to email the authors directly, please see the first page of the publication for their contact information.
NRC Publications Archive
Archives des publications du CNRC
This publication could be one of several versions: author’s original, accepted manuscript or the publisher’s version. / La version de cette publication peut être l’une des suivantes : la version prépublication de l’auteur, la version acceptée du manuscrit ou la version de l’éditeur.
For the publisher’s version, please access the DOI link below./ Pour consulter la version de l’éditeur, utilisez le lien DOI ci-dessous.
https://doi.org/10.1016/j.jbiotec.2018.06.307
Access and use of this website and the material on it are subject to the Terms and Conditions set forth at
Optimization of a high-cell-density Polyethylenimine transfection
method for rapid protein production in cho-ebna1 cells
Stuible, Matthew; Burlacu, Alina; Perret, Sylvie; Brochu, Denis; Paul-Roc,
Béatrice; Baardsnes, Jason; Loignon, Martin; Grazzini, Eric; Durocher, Yves
https://publications-cnrc.canada.ca/fra/droits
L’accès à ce site Web et l’utilisation de son contenu sont assujettis aux conditions présentées dans le site LISEZ CES CONDITIONS ATTENTIVEMENT AVANT D’UTILISER CE SITE WEB.
NRC Publications Record / Notice d'Archives des publications de CNRC: https://nrc-publications.canada.ca/eng/view/object/?id=2bedaef1-0263-4c6b-afb8-a90b402dd4b7 https://publications-cnrc.canada.ca/fra/voir/objet/?id=2bedaef1-0263-4c6b-afb8-a90b402dd4b7
Accepted Manuscript
Title: OPTIMIZATION OF A HIGH-CELL-DENSITY POLYETHYLENIMINE TRANSFECTION METHOD FOR RAPID PROTEIN PRODUCTION IN CHO-EBNA1 CELLS Authors: Matthew Stuible, Alina Burlacu, Sylvie Perret, Denis Brochu, B´eatrice Paul-Roc, Jason Baardsnes, Martin Loignon, Eric Grazzini, Yves Durocher
PII: S0168-1656(18)30484-X
DOI: https://doi.org/10.1016/j.jbiotec.2018.06.307 Reference: BIOTEC 8191
To appear in: Journal of Biotechnology Received date: 28-3-2018
Revised date: 23-5-2018 Accepted date: 6-6-2018
Please cite this article as: Stuible M, Burlacu A, Perret S, Brochu D, Paul-Roc B, Baardsnes J, Loignon M, Grazzini E, Durocher Y, OPTIMIZATION OF A HIGH-CELL-DENSITY POLYETHYLENIMINE TRANSFECTION METHOD FOR RAPID PROTEIN PRODUCTION IN CHO-EBNA1 CELLS, Journal of Biotechnology (2018), https://doi.org/10.1016/j.jbiotec.2018.06.307
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
1
OPTIMIZATION OF A HIGH-CELL-DENSITY POLYETHYLENIMINE TRANSFECTION METHOD FOR RAPID PROTEIN PRODUCTION IN CHO-EBNA1 CELLS
Matthew Stuible, Alina Burlacu, Sylvie Perret, Denis Brochu, Béatrice Paul-Roc, Jason Baardsnes, Martin Loignon, Eric Grazzini and Yves Durocher*
Human Health Therapeutics Research Centre, National Research Council Canada, 6100 Royalmount, Montreal, QC H4P 2R2
* Corresponding author. Email address: yves.durocher@nrc.gc.ca
Highlights
EBNA1 expression in CHO cells improves transient gene expression from OriP-containing plasmid vectors.
A new, transfection-compatible media formulation supports high-density culture of CHO-EBNA1 cells.
With optimized transfection conditions, protein yield is among the highest reported to date for CHO TGE.
Abstract
2 For pre-clinical evaluation of biotherapeutic candidates, protein production by transient gene expression (TGE) in Chinese Hamster Ovary (CHO) cells offers important advantages, including the capability of rapidly and cost-effectively generating recombinant proteins that are highly similar to those produced in stable CHO clones. We have established a novel CHO clone (CHO-3E7) expressing a form of the Epstein-Barr virus nuclear antigen-1 (EBNA-1) with improved TGE productivity relative to parental CHO cells. Taking advantage of a new transfection-compatible media formulation that permits prolonged, high-density culture, we optimized transfection parameters (cell high-density, plasmid vector and
polyethylenimine concentrations) and post-transfection culture conditions to establish a new, high-performing process for rapid protein production. The growth media is chemically defined, and a single hydrolysate feed is added post-transfection, followed by periodic glucose supplementation. This method gave significantly higher yields than our standard low-cell density, F17-based CHO-3E7 TGE method, averaging several hundred mg/L for a panel of recombinant proteins and antibodies. Purified antibodies produced using the two methods had distinct glycosylation profiles but showed identical target binding kinetics by SPR. Key advantages of this new protein production platform include the cost-effectiveness of the transfection reagent, the commercial availability of the culture media and the ability to perform high-cell-density transfection without media change.
Abbreviations
r-protein, recombinant protein; PEI, polyethylenimine; CHO, Chinese Hamster Ovary; TGE, transient gene expression; EBNA1, Epstein Barr virus Nuclear Antigen-1; BCDT, BalanCD Transfectory CHO
3
1. Introduction
Recombinant proteins (r-proteins) represent the largest class of new therapeutic products being developed by the biopharmaceutical industry. Because many of them are highly complex entities possessing extensive post-translational modifications that are necessary for their biological activity, mammalian cell expression systems are preferred for their manufacturing (Durocher and Butler, 2009). As an alternative to the long and tedious process of generating stable clones, transient transfection of mammalian cells using the cationic polymer polyethylenimine (PEI) has become a very popular
technology allowing the rapid production of milligram to gram quantities of biologically active r-proteins (Baldi et al., 2007; Geisse, 2009; Pham et al., 2006).
Chinese Hamster Ovary (CHO) cells are the predominant host cell line used for r-protein manufacturing and have therefore also become a desirable host for developing platforms allowing the rapid production of potential therapeutic candidates for pre-clinical R&D. HEK293 cells are another popular transfection host that can give transient r-protein yields up to 1 g/L (Backliwal et al., 2008); however, compared to HEK293, r-proteins produced in transiently transfected CHO cells show biophysical characteristics (e.g. glycosylation and SEC profiles) more similar to those produced by stable CHO clones (Rajendra et al., 2015b; Ye et al., 2009). In addition, successful transient production is often indicative of high-yield production by stable clones (Diepenbruck et al., 2013; Jordan et al., 1996; Mason et al., 2012), de-risking subsequent resource-intensive steps of CHO clone generation. However, the productivity of CHO
transient production processes has remained well below that of the best stable CHO platforms.
Recently, there has been renewed interest in CHO TGE as a key component in the biologics research and development pipeline, with several large pharmaceutical companies reporting development of robust in-house methods. Significant peer-reviewed contributions have come from industry research groups, who have recently described CHO TGE platforms optimized using Design of Experiments (DoE)
4 methodology (Daramola et al., 2014; Rajendra et al., 2015b). Interestingly, their finalized fed-batch protocols are remarkably distinct: important differences include cell density at transfection (1 or 4x106 cells/ml), the concentration of expression plasmid used for transfection (0.5 or 3.2 ug/ml) and cell type (glutamine synthetase (GS) KO or GS-overexpressing CHO cells). Both groups were able to achieve yields of 200-300 mg/L for different monoclonal antibodies, and notably, they also reported that these
processes could be scaled up and extended using an undisclosed protocol and proprietary feed to achieve 1-2 g/L yields after 16-21 days (Daramola et al., 2014; Rajendra et al., 2015b).
Expression of the Epstein Barr virus (EBV) Nuclear Antigen-1 (EBNA1) is a widely employed strategy to improve r-protein expression in transiently transfected mammalian cell lines, including CHO (Abbott et al., 2015) and HEK293 (Baldi et al., 2007; Durocher et al., 2002; Pham et al., 2006). In human cells, EBNA1 enhances TGE from vectors containing the EBV OriP sequence by various mechanisms, including promotion of episomal replication, plasmid nuclear import, plasmid segregation in sister cells during division and DNA transcription (Kennedy and Sugden, 2003; Kishida et al., 2008). Here, we describe the generation of a CHO-derived cell line stably expressing a truncated form of EBNA1 (CHO-3E7), which exhibits a 3-fold improvement in transient gene expression when using oriP-based vectors. We have proceeded to optimize a high cell density transfection method using a commercially-available, chemically-defined media. This media supports the maintenance of CHO-3E7 cells for up to 2 weeks post-transfection, following a simple fed-batch protocol with a single feed and periodic glucose supplementation. The process generated yields averaging 460 mg/L (range: 100-900 mg/L) for several recombinant proteins and antibodies and displayed scalability up to a 1.6-L culture volume in 5-L shake flasks.
2. Materials and Methods
2.1 Cell culture
5 During development of the 3E7 clone (experiments described in Figures 1 and 2), cells were maintained in Freestyle CHO media (Gibco), supplemented with 8 mM glutamine, in vented polycarbonate
Erlenmeyer flasks under constant agitation (120 rpm) at 37°C, 5% CO2 and standard humidified
conditions. For TGE experiments described in remaining figures, CHO-3E7 cells were grown in BalanCD Transfectory CHO media (BCDT, Irvine Scientific) supplemented with 4 mM glutamine, or FreeStyle-17 (F17) media (Gibco), supplemented with 4 mM glutamine and 0.1% Kolliphor P-188. For routine culture, cells were diluted every 2-3 days to maintain cell densities below 3x106/ml.
2.2 Western blotting
Western blots for detection of EBNA1 protein were performed using the rat monoclonal antibody 1H4, generously provided by Dr F. Grasser (Grasser et al., 1994).
2.3 SEAP activity assay
Secreted alkaline phosphatase was expressed using the pTT-SEAP plasmid and SEAP activity was measured in cell culture supernatant as described previously (Delafosse et al., 2016).
2.4 Plasmids
For stable expression of EBNA1, a truncated form (EBNA1c) encoding the linking region 2 and DNA binding domain was cloned in the plasmid pYD7 vector that also contains a blasticidin selection marker (Loignon et al., 2008). The pTT5 expression vector, described previously (Shi et al., 2005), was used for expression of all recombinant proteins and mAbs with the exception of syndecan-4 ectodomain-mRFP (SCD4-mRFP), which was cloned in the pTT40 vector (Delafosse et al., 2016). For mAbs, heavy and light chain coding sequences were cloned into separate pTT5 plasmids. Coding sequences for His-fusion proteins include 6, 8 or 10 consecutive histidine residues at the C-terminus. Vectors pTT-GFP and
6 hAktDD (constitutively active Akt mutant) were co-transfected as described previously
(Dorion-Thibaudeau et al., 2014; Dorion-(Dorion-Thibaudeau et al., 2016; Durocher et al., 2002).
2.5 CHO-3E7 transfection in F17 media (reference method)
For protein production in F17 media, we used a modified version of a protocol described previously (Delafosse et al., 2016; Raymond et al., 2015). Briefly, CHO-3E7 cells, maintained in F17 media
supplemented with 4 mM L-Glutamine and 0.1% Kolliphor P-188, were seeded 24 h before transfection to reach a density of 2x106 cells/ml at the time of transfection. The transfected DNA was a mix of plasmids encoding the r-protein of interest (in pTT5 or similar vector, 80% [w:w]), constitutively active Akt (in pTT22, 15% [w:w]) and GFP (in pTT, 5% [w:w]); for mAbs, the r-protein-encoding DNA was a 1:1 mix (w:w) of pTT5 containing heavy and light chain coding sequences. Plasmid DNA and PEI (PEI MAX, PolySciences) were diluted separately in 1/20th of the final culture volume of F17 (final concentrations of DNA and PEI were 1.0 µg/ml and 5.0 µg/ml, respectively). Diluted PEI was added to diluted DNA and incubated for 7 minutes at room temperature. The PEI/DNA mixture was then added to cells and returned to incubator, shaking, at 37°C. At 24 h post-transfection (hpt), cultures were supplemented with Tryptone N1 (1% [w/v], Organotechnie) and valproic acid (1 mM) and moved to a 32°C incubator. Supernatants were harvested at 6 days post-transfection.
2.6 CHO-3E7 transfection in BCDT media
CHO-3E7 cells were seeded at 48 h prior to transfection to give desired cell density (7x106 cells/ml for optimized protocol) at the time of transfection. Just before transfection, cells were diluted with fresh media (from 7x106 cells/ml to 5x106 cells/ml for optimized protocol) and dimethylacetamide was added to 0.075% (v/v). Plasmid DNA and PEI were diluted separately in 1/20th of the final culture volume of BCDT (optimized final concentrations of DNA and PEI were 1.3 µg/ml and 7.5 µg/ml, respectively). Diluted PEI was added to diluted DNA and incubated for 7 minutes at room temperature. The PEI/DNA
7 mixture was then added to cells and returned to incubator, shaking, at 37°C. At 24 h post-transfection, cultures were supplemented with Anti-Clumping Supplement (0.2% [v:v], Irvine Scientific), Transfectory Supplement (10% [v:v], Irvine Scientific) and glucose (10 mM) and moved to a 32°C incubator. Starting at 48 hpt and subsequently every 2-3 days, glucose and lactate concentrations in the culture supernatant were monitored using a Vitros 350 instrument (Ortho Clinical Diagnostics); additional glucose was added to 35-40 mM such that concentration was continuously maintained >10 mM. All transfections were performed in a final volume of 25 ml (prior to addition of feed) in 125-ml shake flasks, with the exception of scale-up experiments (Figure 6) which were performed in volumes of 100 ml (500-ml shake flasks), 200 ml (1-L shake flasks) or 1.6 L (5-L Optimum Growth Flasks, Thomson Instrument Company). For experiments using the alternative feed regime (Figure 6), the Transfectory Supplement was replaced with feed F12.7 (Irvine Scientific), added on Days 1, 4, 6, 8, 11 and 13 post-transfection at volumes of 5%, 5%, 10%, 15%, 10% and 10% (v/v), respectively, of the initial transfection volume. For all
experiments, supernatants were harvested at 14 days post-transfection.
2.7 Purification of r-proteins from cell culture supernatants
Cell cultures were centrifuged 20 min at 3000 xg and supernatants were filter-sterilized using a 0.22 µm membrane vacuum filter (Express PLUS, Millipore). Filtered supernatants were loaded on MabSelect SuRe (for mAbs, GE Healthcare) or Ni-Sepharose Excel (for His-tagged proteins, GE Healthcare) columns equilibrated in PBS. The columns were washed with PBS and proteins were eluted with 100 mM citrate buffer pH 3.6 (mAbs) or 300 mM imidazole in 50 mM sodium phosphate, pH 7.0, containing 300 mM NaCl (His-tagged proteins). The fractions containing eluted proteins were pooled and elution buffer was exchanged for PBS using NAP-25 columns (GE Healthcare).
2.8 Protein Quantification
8 mAb titers in culture supernatants were determined by protein-A HPLC using a 800 µL POROS 20 micron Protein A ID Cartridge (Applied Biosystems) accordi g to the a ufacturer’s recommendations. Purified proteins in PBS were quantified by absorbance at 280 nm using a Nanodrop spectrophotometer
(Thermo Fisher Scientific) and the calculated extinction coefficient for each protein.
2.9 Analysis of Fc-associated N-Glycosylation by HILIC
N-Glycan analysis was performed by PNGase-F digestion, 2-aminobenzamide (2-AB) labelling and
HILIC-HPLC based on a method reported previously (Melmer et al., 2010), with some modifications. PNGase digests were performed using 48 µg of each purified antibody. For mAb MEDI-573 which contains both Fc and Fab N-Glycans, Fc glycans were released specifically by treating purified mAb with PNGase-F (New England Biolabs) under non-denaturing conditions. For the other mAbs (with only Fc glycans), glycans were released by treatment with Rapid PNGase F (NEB) under denaturing conditions. PNGase-released glycans were cleaned up on Discovery SPE 50 mg cartridges (Sigma) according to the
a ufacturer’s i structio s. The 2-AB labelling reaction was performed at 60°C for 2 h, and the labeled glycans were dissolved in 200 µl of 70% acetonitrile. 25 µl were injected on a 4.6 x 150 mm TSKamide-80 column. Elution of 2-AB-labelled glycans was performed using a 25 to 50% gradient of solvent A (100 mM ammonium formate, pH 4.5) in solvent B (100% acetonitrile) at 0.5 ml/min over 50 minutes.
2.10 SPR binding assays
All surface plasmon resonance assays were carried out using a Biacore T200 Surface Plasmon Resonance instrument (GE Healthcare Life Science, (Mississauga, ON)) with PBST running buffer (PBS with 0.05% Tween-20 and 3.4 mM EDTA) at a temperature of 25°C. The anti-human Fc capture surface was generated using a Series S Sensor Chip CM5 using the default parameters under the Immobilization Wizard in the Biacore T200 control software which was set to target 2000 resonance units (RUs). Her2 ectodomain (Her2ED) was purchased from Thermo Fisher Scientific (Waltham, MA) and the EGFR
9 ectodomain (EGFRED) was produced internally using a baculovirus expression system (Brown et al., 1994). hIGF-1 was purchased from Sigma Aldrich (Oakville, ON). Her2ED, EGFRED and all mAbs were purified by size exclusion chromatography prior to SPR to eliminate aggregates and ensure
homogeneity. Analysis of binding of mAbs to Her2ED, EGFRED or IGF-1 antigen targets occurred in two steps: an indirect capture of the antibodies onto the anti-human Fc antibody flow cell surface followed by the injection of 5 concentrations of purified antigen for kinetic analysis using the single cycle kinetics methodology. Antibodies for capture were injected at 0.5 to 5 µg/mL over individual flow cells for 60 s at flow μL/ i . I ge eral, this resulted i a capture of appro i atel 5 to RUs o to the a ti-human Fc surface. The first flow cell was left empty to use as a blank control. This capture step was immediately followed by five concentrations of antigen (either 5.0 nM, 2.5 nM, 1.25 nM, 0.63 nM and 0.31 nM for EGFRED; 40 nM, 20 nM, 10 nM, 5 nM, and 2.5 nM for Her2ED antigen; or 10 nM, 5.0 nM, 2.5 nM, 1.25 nM and 0.63 nM for hIGF-I) were seque tiall i jected at μL/ i for 8 s with a
dissociation phase of 1200 s for EGFR ECD and hIGF-I, and 1800 s for Her2ED. The captured antibody surfaces were regenerated by 10 mM glycine, pH 1.5, for s at μL/ i to prepare for the e t injection cycle. At least two mock-buffer injections were performed for each analyte injection to be used for referencing. The resulting single cycle kinetics sensorgrams were analyzed using Biacore T200 BiaEvaluation software fit to the 1:1 binding model.
3. RESULTS
3.1 Development of a stable CHO-EBNA1 cell line
In order to increase TGE in CHO cells when using EBV oriP-bearing pTT plasmids, we first generated a stable clone expressing a codon-optimized, truncated EBNA1 protein. CHO cells were transfected with the pYD7 expression plasmid (Loignon et al., 2008) encoding EBNA1c, a truncated version of EBNA1 lacking the N-terminal DNA linking region 1 (LR1), TAD domain and Gly-Ala repeats. Cells stably
10 expressing EBNA1c were selected with blasticidin. The blasticidin resistant cells were then cloned by limiting dilution in 96 well plates in the absence of selection. Clones were tested for EBNA1c expression by western blotting; clone 3E7 showed expression of EBNA1 protein which is absent in parental CHO cells (Figure 1A). CHO-3E7 cells were maintained in culture for several weeks without blasticidin and samples were taken periodically. Western blot analysis showed that EBNA1c expression in clone 3E7 was stable for at least 132 days without selection (Figure 1B).
In initial experiments, CHO-3E7 cells were maintained and transfected in FS-CHO media. In this media, we determined that optimal conditions for transfection were at a density of 1.5-2.0x106 cells/ml using 1 µg/ml of DNA and 8 µg/ml of linear, 25 kDa PEI (data not shown). We compared the transfection efficiency of CHO-3E7 and parental CHO cells using the EBV oriP-containing pTT-GFP vector and flow cytometry analysis, with mock-transfected CHO-3E7 cells used to establish gating parameters. The percentage of GFP-positive cells / mean fluorescence intensities were 42.9% / 350 and 65.7% / 1110 for CHO and CHO-3E7 cells, respectively, showing that CHO-3E7 cells expressed higher levels of GFP in a greater proportion of cells.
Recombinant protein productivity was also evaluated by monitoring secreted alkaline phosphatase (SEAP) activity following transfection with the pTT-SEAP plasmid. Consistent with the higher level of GFP expressed in CHO-3E7 cells, SEAP activity was 3-fold higher in CHO-3E7 vs CHO cells, reaching 50 mg/L and 17 mg/L, respectively, six days post-transfection (Figure 2A). This higher productivity was likely due to EBNA1 expression as SEAP activity was increased by 3.7-fold in CHO-3E7 transfected with the oriP-encoding plasmid compared to the same plasmid devoid of the oriP (Figure 2B). SEAP expression was identical in CHO and CHO-3E7 cells when using the oriP-deleted pTT vector or when CHO cells were transfected with oriP-positive or -negative plasmid, suggesting that cell transfectability was not altered by stable EBNA1c expression.
11
3.2 High cell-density transfection in BalanCD Transfectory CHO media
The process we initially developed for transient protein expression in CHO-3E7 cells using FreeStyle CHO (FS-CHO) or F17 media has certain limitations. Specifically, FS-CHO and F17 media do not permit
prolonged high-density growth of CHO-3E7 cells, the cells are best transfected at low density
(~2x106/ml), and transfected cultures cannot normally be maintained longer than 7-8 days. Together, these factors limit the potential productivity of the method.
A novel media formulation for CHO transient protein production, BalanCD Transfectory CHO (BCDT), was recently developed and commercialized by Irvine Scientific. This chemically-defined media is compatible with PEI-mediated transfection, supports high cell-density growth (>8x106 viable cells/ml), and allows maintenance of productive cultures for more than 14 days post-transfection with a single addition of animal-component-free hydrolysate feed (data not shown). These favourable features led us to re-optimize our transfection and culture conditions for the CHO-3E7/OriP system using the BCDT media. For subsequent optimization steps, we expressed the monoclonal antibody Trastuzumab, transfecting a mix of OriP-encoding pTT vectors expressing the mAb heavy chain (40%), light chain (40%), constitutively active Akt (15%) and GFP (5%). The transfection agent was PEI MAX, a fully-deacylated 40 kDa linear PEI (HCl salt of a 22 kDa free base).
The BCDT a ufacturer’s protocol suggests seedi g cells 3 days prior to transfection to achieve a cell density of 2.5-3x106 cells/ml at the time of transfection. In initial experiments, we noted that higher transfection cell densities as well as dilution of cultures with fresh media just before transfection tended to give better productivity, particularly in the first 10 days post-transfection. To examine these factors in parallel, we seeded CHO-3E7 cells at various densities in three flasks at 2 days before transfection, aiming for cell densities of 3, 5 and 7x106 cells /ml. The cells were then transfected without dilution at 3
12 and 5x106 cells/ml or diluted to these densities from the 7x106 cells/ml culture with fresh media before transfection.
As shown in Figure 3A, the dilution of cultures before transfection had a consistent, positive effect on mAb productivity: cells transfected at 3 or 5x106 cells /ml after dilution from higher cell densities were more productive than those transfected without dilution at the same densities. Among the conditions tested, the most productive were the cells transfected at 5x106 cells/ml after dilution from 7x106 cells/ml, giving a Trastuzumab yield >450 µg/ml at 13 days post-transfection (dpt). The addition of fresh media before transfection correlates with an increase in the percentage of GFP-positive cells at 24 hpt (Figure 3B), indicating that the improved productivity is a result of increased transfection efficiency. Notably, there is an inverse, but small, effect on cell viability (shown in Figure 3C for Day 8 post-transfection), likely due to toxicity associated with transfection.
The organic solvent dimethylacetamide (DMA) has been demonstrated to positively affect CHO cell productivity in transient transfection processes (Rajendra et al., 2015a). A concentration of 0.125% (v/v), added at the time of transfection, was reported to improve yields by 50-80% for a GS-knockout CHO transfection protocol (Rajendra et al., 2015b). For CHO-3E7 cells in BCDT media, DMA had a consistent, but smaller, positive effect on productivity, but we found that a lower concentration (0.075%) was better. The addition of DMA at this concentration resulted in a 20-25% increase in yields under a variety of transfection conditions, with minimal effect on cell viability (data not shown).
3.3 Design-of-experiments optimization of transfection conditions
We next implemented a Design of Experiments (DoE) method to optimize DNA and PEI concentrations for CHO-3E7 transfection using BCDT media. DoE approaches have been successfully applied by several groups to optimize transfection parameters for various transient gene expression processes (Abbott et al., 2015; Bollin et al., 2011; Rajendra et al., 2015b; Thompson et al., 2012). We generated a 2-variable,
13 central composite design with ranges for DNA and PEI concentrations set at 0.5-2 µg/ml and 5-9 µg/ml, respectively, using Reliasoft DOE++ software. Other parameters corresponded to default software settings for this design. The center point in the design (1.25 µg/ml DNA and 7 µg/ml PEI) was repeated 5 times to account for experimental variability.
As shown in Figure 4A, the best transfection efficiency, as assessed by the percentage of GFP+ cells at 24 hpt), was at the highest DNA and PEI concentrations. However, these conditions were associated with reduced cell viability post-transfection (shown in Figure 4B for 8 dpt). The best antibody yields were achieved within the ranges of the two factors tested: the predicted optimal DNA and PEI MAX concentrations were 1.3 µg/ml and 7.5 µg/ml, respectively, estimated to give Trastuzumab
concentrations >800 µg/ml in the culture supernatant at Day 13. The same design was repeated in the absence of DMA; optimal DNA and PEI concentrations were similar, but yields were reduced by ~25%, as expected (not shown).
Following DoE optimization, we performed Trastuzumab productions using the predicted optimal DNA and PEI MAX concentrations. As shown in Figure 5A, based on three independent experiments, the protocol achieves a relatively constant viable cell density of 6-7x106/ml, with cell viability of ~80%, throughout the course of the 14 day process. Glucose consumption remained relatively stable at 5-7 mM/day through Day 12 (not shown). Yields of Trastuzumab were similar in the three experiments, averaging 787 µg/ml (standard deviation 34 ug/ml) at Day 14 (Figure 5B). Lactate concentrations in culture supernatant were routinely monitored to determine whether accumulation of this metabolite, and its effect on culture pH, could have a limiting effect on productivity. Interestingly, the cultures display marked lactate consumption, with concentrations dropping from >10 mM on Day 5 to undetectably low (< 1 mM) concentrations by Day 12 (Figure 5B). Only a few previous reports of CHO TGE processes have monitored lactate, and in these few cases, concentrations were relatively constant
14 at 5-20 mM over 1-2 weeks post-transfection (Codamo et al., 2011; Wulhfard et al., 2008); notably, a metabolic switch to lactate consumption is regarded as a positive feature for r-protein production using stable CHO clones (Luo et al., 2012).
3.4 Scalability of the optimized CHO-3E7/BCDT method
The optimized method was tested in different culture vessels, including 6- deep-well plates (6DWP) and various-sized shake flasks. As shown in Figure 6, the protocol, which had been developed using 125-ml flasks, could be applied to transfections in 6DWP, 250-ml flasks and 1-l flasks with less than a 15% decrease in productivity. In contrast, productions in 5-L Optimum Growth shake flasks consistently gave yields that were >30% lower than in the smaller flasks. Previously, we had experienced similar issues scaling up protein production using stable CHO pools; the use of an alternative feed (F12.7, Irvine Scientific) and an increased poloxamer concentration (0.2% Kolliphor P-188 rather than the 0.1% contained in BCDT) improved productivity in the large flasks. We repeated the 125-ml and 5-L flask productions using F12.7 and 0.2% Kolliphor; interestingly, these changes did not alter Trastuzumab yield in the 125-ml flask, but the performance of the cells in the 5-L flask improved dramatically, matching the productivity of the smaller flasks (Figure 6).
3.5 Production of a panel of recombinant antibodies and other proteins using optimized protocol
To evaluate the applicability of this method to other antibodies and recombinant proteins, we produced a panel of 6 monoclonal antibodies and 5 recombinant His-tagged proteins using the optimized method. In parallel, the same set of proteins were produced following an established, reference protocol based on CHO-3E7 cells cultured in FreeStyle F17 media (Delafosse et al., 2016; Durocher and Loignon, 2014; Raymond et al., 2015). All coding sequences were optimized with CHO-codon bias (gene optimization and synthesis was performed by Genscript, Geneart, or DNA2.0) and cloned in pTT5 (or similar) vectors containing a modified EBNA OriP sequence. As shown in Table 1, for the BCDT method, the highest
15 yields were close to 900 µg/ml for the chimeric (mouse/human IgG4) mAb B72-3. Palivizumab,
Rituximab and MEDI-573 gave approximately 500-600 µg/ml while Cetuximab gave the lowest yield, 172 µg/ml. Productions of the His-tagged proteins yielded approximately 300-500 µg/ml, with the exception of erythropoietin (106 µg/ml). For all of the proteins tested, the BCDT method gave better yields, on average 3-fold higher, than the F17 method.
As shown in Figure 7, after a single Protein A purification step, the antibodies gave predominantly single species by SDS-PAGE. For the His-tagged proteins, alpha-1 antitrypsin (A1AT) and syndecan 4 (SCD4)-mRFP fusion gave more than one strong band by SDS-PAGE, while erythropoietin (EPO) gave a single diffuse band; these results are likely the result of glycosylation heterogeneity and are consistent with previous productions of these proteins. SEAP and the single-domain antibody (VHH) gave predominantly single species by SDS-PAGE.
The binding of three mAbs (Trastuzumab, Cetuximab and MEDI-573) to their respective target proteins (Her2, EGFR and IGF1) was evaluated by SPR. As shown in Figure 8, there were no significant differences in target binding for mAbs produced using the BCDT vs. F17 protocols. Finally, we profiled the
Fc-associated N-linked glycans present in the six purified mAbs by HILIC-HPLC. As shown in Figure 9, compared to those produced using the F17-based reference method, mAbs produced using the optimized BCDT method show increased percentages of galactosylated glycans (G1F and G2F) and a corresponding decrease in the non-galactosylated form (G0F). The mAbs produced in BCDT showed higher proportions of high-mannose (M5) glycans (6-13%) compared to the mAbs produced in F17 (1-2%).
4. DISCUSSION
Following the approval of the first CHO-derived recombinant protein therapeutic in 1986, CHO cells rapidly became the preferred host for biologics manufacturing (Lalonde and Durocher, 2017; Wurm and
16 Petropoulos, 1994). Just over a decade ago, while stable CHO clones expressing recombinant proteins at >1 g/L had been developed routinely, the productivity of CHO TGE processes was generally <10 mg/L (Pham et al., 2006). Despite this poor productivity, the potential of CHO TGE for the rapid production of material for research and preclinical studies had already been recognized (Wurm, 2004). Over the last few years, the performance of transient CHO production processes has improved markedly, with several published reports of yields of hundreds of mg/L (Daramola et al., 2014; Jain et al., 2017; Rajendra et al., 2015b; Steger et al., 2015). The advancement of yields towards the g/L range has reinforced the advantages, particularly the cost-effectiveness, of producing research material and preclinical candidates by CHO TGE, and has even raised the possibility that certain biologics, especially those required rapidly and in smaller quantities (e.g. for clinical phase 1 studies or a rapid response to pandemics), could eventually be manufactured in this manner (Gutierrez-Granados et al., 2018; Wurm and De Jesus, 2016).
Here, we have described a novel CHO-based transient expression platform, with optimized conditions summarized in Table 2, that enables r-protein yields among the highest described in the literature (Daramola et al., 2014; Jain et al., 2017; Rajendra et al., 2015b; Steger et al., 2015) but with several key advantages. In particular, the method we developed uses the cost-effective linear PEI, PEI MAX, for DNA transfection rather than more expensive forms of PEI (eg. PEIPro from Polyplus), lipid-based transfection agents (eg. Expifectamine developed for the Thermo Fisher ExpiCHO kit (Jain et al., 2017)), or
specialized/expensive electroporation equipment (Steger et al., 2015). In addition, in contrast to
recently-reported methods that require complete medium exchange by centrifugation and resuspension of cells in order to achieve a high cell density in transfection-compatible media (Rajendra et al., 2015b; Steger et al., 2015), the BCDT media used in the current method permits high-density growth and transfection without media change. Finally, the BCDT media and feed are commercially available,
17 whereas many other transient CHO processes rely on proprietary media and/or feed components to achieve reported yields (Daramola et al., 2014; Rajendra et al., 2015b).
In our group, the current, routine method for transient protein production, based on CHO-3E7cultured in F17 media, is elaborated from a method described in a 2009 patent (Durocher and Loignon, 2014). For all of the mAbs/r-proteins listed in Table 1, the optimized Transfectory CHO-based method described in this manuscript gives superior yields, with an average 3-fold difference. A recent paper from our group described the production of a subset of these mAbs (Trastuzumab, Palivizumab and B72-3) in CHO (non-optimal conditions) and HEK293-EBNA1 cells transfected in F17 (Delafosse et al., 2016). The method described in the current manuscript performed substantially better in all cases, with the exception of Palivizumab, which gave slightly higher yields with the F17-based HEK293-EBNA1 method.
The antibodies produced in BCDT and F17 showed different proportions of N-linked Fc glycans: BCDT-derived antibodies had higher levels of galactosylated (G1F and G2F) vs non-galactosylated (G0F) glycans but had greater proportions of high-mannose structures. It has been observed previously that the type of media and feed used for antibody production in CHO cells can significantly affect N-glycosylation patterns (Kildegaard et al., 2016; Reinhart et al., 2015; Shi and Goudar, 2014). Furthermore, increasing CHO culture duration can markedly increase high-mannose glycan (HMG) levels (Robinson et al., 1994), and increased HMG levels have been reported previously for CHO TGE methods lasting 14 days (Jain et al., 2017; Ye et al., 2009). Therefore, the differences in glycan structures were not unexpected.
Galactosylated glycans are generally considered a positive feature (Reinhart et al., 2015); galactosylation is particularly sensitive to glucose levels, and the increased rates of glucose consumption in BCDT- vs F-17-based methods (and the increased glucose supplementation required) may contribute to the differences observed (Liu et al., 2014). HMG can augment antibody-dependent cellular cytotoxicity (ADCC) (Yu et al., 2012) but can negatively impact pharmacokinetics by promoting antibody clearance in
18
vivo (Goetze et al., 2011; Pacis et al., 2011). A HMG content of 17% was reported to result in a 6%
decrease in antibody exposure over a 14-day treatment period (Goetze et al., 2011). Previously reported methods to reduce HMG levels using media additives or by reducing culture duration (Pacis et al., 2011) could be considered if HMG levels in BCDT-derived mAbs (6-13%) prove problematic.
It is not straightforward to compare the performance of the method we have described with other transient CHO protocols reported in the literature or offered as a kit or service. Certain factors that can have important impacts on r-protein expression, including the sequences of signal peptides and codon-optimized protein-coding regions are not regularly disclosed in publications. In addition, the
recombinant antibodies/proteins used to test expression methods are often not identified, with
antibodies being frequently defined by subclass only (e.g. (Bollin et al., 2011; Daramola et al., 2014; Jain et al., 2017; Rajendra et al., 2015b)). In our experience, different antibodies of the same subclass can display widely variable expression levels; this is also evident from the yields obtained for the four mAbs of human IgG1 isotype (humanized and chimeric) in Table 1. Nonetheless, among the small set of antibodies we have produced using our optimized method, the highest yielding, B72.3, Trastuzumab, and MEDI-573, were of three different subtypes (human IgG4, IgG1 and IgG2, respectively), suggesting that the method will be effective for production of a variety of recombinant antibodies.
ACKNOWLEDGEMENTS
We thank Michel Gilbert for helpful discussions regarding N-glycan analysis by HILIC, and Martine Pagé and Louis Bisson for technical contributions. This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
REFERENCES
19 Abbott, W.M., Middleton, B., Kartberg, F., Claesson, J., Roth, R., Fisher, D., (2015) Optimisation of a
simple method to transiently transfect a CHO cell line in high-throughput and at large scale. Protein Expr Purif 116, 113-119.
Backliwal, G., Hildinger, M., Chenuet, S., Dejesus, M., Wurm, F.M., (2008) Coexpression of acidic
fibroblast growth factor enhances specific productivity and antibody titers in transiently transfected HEK293 cells. N.Biotechnol.
Baldi, L., Hacker, D.L., Adam, M., Wurm, F.M., (2007) Recombinant protein production by large-scale transient gene expression in mammalian cells: state of the art and future perspectives.
Biotechnol.Lett. 29, 677-684.
Bollin, F., Dechavanne, V., Chevalet, L., (2011) Design of Experiment in CHO and HEK transient transfection condition optimization. Protein Expression and Purification 78, 61-68.
Brown, P.M., Debanne, M.T., Grothe, S., Bergsma, D., Caron, M., Kay, C., O'Connor-McCourt, M.D., (1994) The extracellular domain of the epidermal growth factor receptor. Studies on the affinity and stoichiometry of binding, receptor dimerization and a binding-domain mutant. European journal of biochemistry 225, 223-233.
Codamo, J., Hou, J.J.C., Hughes, B.S., Gray, P.P., Munro, T.P., (2011) Efficient mAb production in CHO cells i corporati g PEI‐ ediated tra sfectio , ild h pother ia a d the co‐e pressio of XBP‐ . Journal of Chemical Technology and Biotechnology 86, 923-934.
Daramola, O., Stevenson, J., Dean, G., Hatton, D., Pettman, G., Holmes, W., Field, R., (2014) A high-yielding CHO transient system: coexpression of genes encoding EBNA-1 and GS enhances transient protein expression. Biotechnol Prog 30, 132-141.
Delafosse, L., Xu, P., Durocher, Y., (2016) Comparative study of polyethylenimines for transient gene expression in mammalian HEK293 and CHO cells. J Biotechnol 227, 103-111.
Diepenbruck, C., Klinger, M., Urbig, T., Baeuerle, P., Neef, R., (2013) Productivity and quality of
recombinant proteins produced by stable CHO cell clones can be predicted by transient expression in HEK cells. Mol Biotechnol 54, 497-503.
Dorion-Thibaudeau, J., Raymond, C., Lattova, E., Perreault, H., Durocher, Y., De Crescenzo, G., (2014) Towards the development of a surface plasmon resonance assay to evaluate the glycosylation pattern of monoclonal antibodies using the extracellular domains of CD16a and CD64. J Immunol Methods 408, 24-34.
Dorion-Thibaudeau, J., St-Laurent, G., Raymond, C., De Crescenzo, G., Durocher, Y., (2016) Biotinylation of the Fcgamma receptor ectodomains by mammalian cell co-transfection: application to the development of a surface plasmon resonance-based assay. J Mol Recognit 29, 60-69.
Durocher, Y., Butler, M., (2009) Expression systems for therapeutic glycoprotein production. Curr Opin Biotechnol 20, 700-707.
Durocher, Y., Loignon, M., (2014) Patent US8637315 - Process, vectors and engineered cell lines for enhanced large-scale transfection. National Research Council Canada.
Durocher, Y., Perret, S., Kamen, A., (2002) High-level and high-throughput recombinant protein production by transient transfection of suspension-growing human 293-EBNA1 cells. 30, 1-9. Geisse, S., (2009) Reflections on more than 10 years of TGE approaches. Protein Expr Purif 64, 99-107. Goetze, A.M., Liu, Y.D., Zhang, Z., Shah, B., Lee, E., Bondarenko, P.V., Flynn, G.C., (2011) High-mannose
glycans on the Fc region of therapeutic IgG antibodies increase serum clearance in humans. Glycobiology 21, 949-959.
Grasser, F.A., Murray, P.G., Kremmer, E., Klein, K., Remberger, K., Feiden, W., Reynolds, G., Niedobitek, G., Young, L.S., Mueller-Lantzsch, N., (1994) Monoclonal antibodies directed against the Epstein-Barr virus-encoded nuclear antigen 1 (EBNA1): immunohistologic detection of EBNA1 in the malignant cells of Hodgkin's disease. Blood 84, 3792-3798.
20 Gutierrez-Granados, S., Cervera, L., Kamen, A.A., Godia, F., (2018) Advancements in mammalian cell
transient gene expression (TGE) technology for accelerated production of biologics. Critical reviews in biotechnology, 1-23.
Jain, N.K., Barkowski-Clark, S., Altman, R., Johnson, K., Sun, F., Zmuda, J., Liu, C.Y., Kita, A., Schulz, R., Neill, A., Ballinger, R., Patel, R., Liu, J., Mpanda, A., Huta, B., Chiou, H., Voegtli, W., Panavas, T., (2017) A high density CHO-S transient transfection system: Comparison of ExpiCHO and Expi293. Protein expression and purification 134, 38-46.
Jordan, M., Schallhorn, A., Wurm, F.M., (1996) Transfecting mammalian cells: optimization of critical parameters affecting calcium-phosphate precipitate formation. Nucleic Acids Res. 24, 596-601. Kennedy, G., Sugden, B., (2003) EBNA-1, a bifunctional transcriptional activator. Mol.Cell Biol. 23,
6901-6908.
Kildegaard, H.F., Fan, Y., Sen, J.W., Larsen, B., Andersen, M.R., (2016) Glycoprofiling effects of media additives on IgG produced by CHO cells in fed-batch bioreactors. Biotechnol Bioeng 113, 359-366. Kishida, T., Asada, H., Kubo, K., Sato, Y.T., Shin-Ya, M., Imanishi, J., Yoshikawa, K., Mazda, O., (2008)
Pleiotrophic functions of Epstein-Barr virus nuclear antigen-1 (EBNA-1) and oriP differentially contribute to the efficiency of transfection/expression of exogenous gene in mammalian cells. Journal of biotechnology 133, 201-207.
Lalonde, M.E., Durocher, Y., (2017) Therapeutic glycoprotein production in mammalian cells. J Biotechnol 251, 128-140.
Liu, B., Spearman, M., Doering, J., Lattová, E., Perreault, H., Butler, M., (2014) 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. Journal of Biotechnology 170, 17-27.
Loignon, M., Perret, S., Kelly, J., Boulais, D., Cass, B., Bisson, L., Afkhamizarreh, F., Durocher, Y., (2008) Stable high volumetric production of glycosylated human recombinant IFNalpha2b in HEK293 cells. BMC Biotechnol 8, 65.
Luo, J., Vijayasankaran, N., Autsen, J., Santuray, R., Hudson, T., Amanullah, A., Li, F., (2012) Comparative metabolite analysis to understand lactate metabolism shift in Chinese hamster ovary cell culture process. Biotechnology and bioengineering 109, 146-156.
Mason, M., Sweeney, B., Cain, K., Stephens, P., Sharfstein, S.T., (2012) Identifying bottlenecks in transient and stable production of recombinant monoclonal-antibody sequence variants in Chinese hamster ovary cells. Biotechnol Prog 28, 846-855.
Melmer, M., Stangler, T., Schiefermeier, M., Brunner, W., Toll, H., Rupprechter, A., Lindner, W., Premstaller, A., (2010) HILIC analysis of fluorescence-labeled N-glycans from recombinant biopharmaceuticals. Anal Bioanal Chem 398, 905-914.
Pacis, E., Yu, M., Autsen, J., Bayer, R., Li, F., (2011) Effects of cell culture conditions on antibody N-linked glycosylation--what affects high mannose 5 glycoform. Biotechnology and bioengineering 108, 2348-2358.
Pham, P.L., Kamen, A., Durocher, Y., (2006) Large-Scale Transfection of Mammalian Cells for the Fast Production of Recombinant Protein. Molecular Biotechnology 34, 225-237.
Rajendra, Y., Balasubramanian, S., Kiseljak, D., Baldi, L., Wurm, F.M., Hacker, D.L., (2015a) Enhanced plasmid DNA utilization in transiently transfected CHO-DG44 cells in the presence of polar solvents. Biotechnol Prog 31, 1571-1578.
Rajendra, Y., Hougland, M.D., Alam, R., Morehead, T.A., Barnard, G.C., (2015b) A high cell density transient transfection system for therapeutic protein expression based on a CHO GS-knockout cell line: process development and product quality assessment. Biotechnol Bioeng 112, 977-986. Raymond, C., Robotham, A., Spearman, M., Butler, M., Kelly, J., Durocher, Y., (2015) Production of
alpha2,6-sialylated IgG1 in CHO cells. MAbs 7, 571-583.
21 Reinhart, D., Damjanovic, L., Kaisermayer, C., Kunert, R., (2015) Benchmarking of commercially available
CHO cell culture media for antibody production. Applied microbiology and biotechnology 99, 4645-4657.
Robinson, D.K., Chan, C.P., Yu Lp, C., Tsai, P.K., Tung, J., Seamans, T.C., Lenny, A.B., Lee, D.K., Irwin, J., Silberklang, M., (1994) Characterization of a recombinant antibody produced in the course of a high yield fed-batch process. Biotechnology and bioengineering 44, 727-735.
Shi, C., Shin, Y.O., Hanson, J., Cass, B., Loewen, M.C., Durocher, Y., (2005) Purification and
characterization of a recombinant G-protein-coupled receptor, Saccharomyces cerevisiae Ste2p, transiently expressed in HEK293 EBNA1 cells. Biochemistry 44, 15705-15714.
Shi, H.H., Goudar, C.T., (2014) Recent advances in the understanding of biological implications and modulation methodologies of monoclonal antibody N-linked high mannose glycans. Biotechnol Bioeng 111, 1907-1919.
Steger, K., Brady, J., Wang, W., Duskin, M., Donato, K., Peshwa, M., (2015) CHO-S antibody titers >1 gram/liter using flow electroporation-mediated transient gene expression followed by rapid migration to high-yield stable cell lines. Journal of biomolecular screening 20, 545-551.
Thompson, B.C., Segarra, C.R., Mozley, O.L., Daramola, O., Field, R., Levison, P.R., James, D.C., (2012) Cell line specific control of polyethylenimine-mediated transient transfection optimized with "Design of experiments" methodology. Biotechnol Prog 28, 179-187.
Wulhfard, S., Tissot, S., Bouchet, S., Cevey, J., De, J.M., Hacker, D.L., Wurm, F.M., (2008) Mild
hypothermia improves transient gene expression yields several fold in Chinese hamster ovary cells. Biotechnol.Prog. 24, 458-465.
Wurm, F.M., (2004) Production of recombinant protein therapeutics in cultivated mammalian cells. Nature biotechnology 22, 1393-1398.
Wurm, F.M., De Jesus, M., (2016) Manufacture of Recombinant Therapeutic Proteins Using Chinese Hamster Ovary Cells in Large-Scale Bioreactors: History, Methods, and Perspectives. In: Liu, C., Morrow, J. (Eds.), Biosimilars of monoclonal antibodies : a practical guide to manufacturing, preclinical, and clinical development, pp. 327-353.
Wurm, F.M., Petropoulos, C.J., (1994) Plasmid integration, amplification and cytogenetics in CHO cells: questions and comments. Biologicals 22, 95-102.
Ye, J., Kober, V., Tellers, M., Naji, Z., Salmon, P., Markusen, J.F., (2009) High-level protein expression in scalable CHO transient transfection. Biotechnology and Bioengineering 103, 542-551.
Yu, M., Brown, D., Reed, C., Chung, S., Lutman, J., Stefanich, E., Wong, A., Stephan, J.P., Bayer, R., (2012) Production, characterization, and pharmacokinetic properties of antibodies with N-linked mannose-5 glycans. mAbs 4, 47mannose-5-487.
FIGURE CAPTIONS
22
Figure 1: EBNA1c protein is expressed stably in CHO-3E7. (A) Western blotting for EBNA1 using 20 µg of
proteins from total cell extracts from CHO-3E7 and parental CHO cells. (B) CHO-3E7 cells were grown without blasticidin for more than 4 months. The expression of EBNA1 was determined by western blotting at the indicated days. Positions of molecular weight markers (kDa) are indicated.
Figure 2: CHO-3E7 cells display improved transient recombinant protein productivity dependent on
EBNA OriP. (A) Quantification of SEAP in cell supernatants at time points following transfection of CHO-3E7 or parental CHO cells with OriP-bearing SEAP expression vector. (B) Normalized SEAP activity of CHO-3E7 vs. CHO cells transfected with plasmids containing OriP or not.
Figure 3: The productivity of CHO-3E7 cells in BCDT media is enhanced by dilution of cultures with fresh
media prior to transfection. Cells were grown to pre-transfection densities of 3, 5 or 7x106 cells/ml, then transfected at 3 or 5x106/ml without dilution (direct), or diluted to 3x106 cells/ml from the higher-density cultures with fresh media. Trastuzumab concentrations in cell supernatants were measured at Day 14 transfection (A). Transfection efficiency (GFP+ percentage) was evaluated at 24 h post-transfection (B). Cell viability at Day 8 post-post-transfection is shown in (C).
Figure 4: Design of Experiments (DoE) optimization of DNA and PEI concentrations. A central composite
DoE design was generated based on a PEI concentration range of 5-9 µg/ml and a plasmid DNA concentration range of 0.5-2.0 µg/ml. Models of transfection efficiency (GFP+ percentage at 24 h, A), cell viability on Day 8 (B) and Trastuzumab yield on Day 12 (C) were generated from experimental data.
Figure 5: Cell culture parameters for production of Trastuzumab using optimized CHO-3E7/Transfectory
CHO method. Errors bars are +/- one standard deviation for three independent experiments.
Figure 6: The BDCT/CHO-3E7 method can be applied to different culture volumes and culture vessel
formats. Transient Trastuzumab productions were performed in the indicated culture vessels; the
23 culture volumes at the time of transfection are shown in brackets. Cultures were fed with Transfectory Supplement (dark grey bars) or F12.7 (light grey bars), and mAb yields were measured at 14 days post-transfection.
Figure 7: SDS-PAGE analysis of a panel of recombinant antibodies (A) and His-tagged proteins (B)
produced using the optimized CHO-3E7/Transfectory CHO method. 3 µg of each purified protein was prepared in denaturing or non-denaturing sample buffer and separated on a 4-12% polyacrylamide gel before staining with Coomassie Blue. Panel A: lane 1, B72-3; lane 2, Palivizumab; lane 3, Rituximab; lane 4, Cetuximab; lane 5, MEDI-573; lane 6, Trastuzumab. Panel B: lane 1, SEAP; lane2, A1AT; lane 3, EPO; lane 4, SDC4-mRFP; lane 5, VHH.
Figure 8: The type of media used for transient production does not significantly impact mAb-target
binding, as determined by SPR. Trastuzumab, Cetuximab and MEDI-573, produced using BCDT- or F17-based methods were captured on the SPR flow cell surface. This immobilization step was followed by sequential injections of multiple, increasing concentrations of ligands (Her2ED, EGFRED or IGF1). Binding constants were calculated from single cycle kinetics sensorgrams based on a 1:1 binding model. The average and standard deviation from three independent injections are shown.
Figure 9: HILIC analysis of glycan structures present in mAbs produced using BCDT- and F17-based
methods. Fc glycans were released from Protein A-purified mAbs by PNGase digest and identified by HPLC following 2-aminobenzimide labelling.
24
25
26
27
28
29
30
31
32
33
Table 1: Comparison of yields for a panel of recombinant mAbs and proteins produced using optimized
CHO-3E7/BalanCD Transfectory CHO protocol (BCDT) or a reference FreeStyle F17-based method (F17).
Protein Type BCDT Yield (mg/l)a
F17 Yield (mg/l)a
B72-3 Chimeric hIgG4 894 224
Palivizumab Humanized IgG1 484 173
Rituximab Chimeric hIgG1 523 145
Cetuximab Chimeric hIgG1 172 68
MEDI-573 Human IgG2 616 150
Trastuzumab Humanized IgG1 759 183
SEAP His-fusion 383 92
A1AT His-fusion 496 238
EPO His-fusion 106 85
SDC4-mRFP His-fusion 292 185
VHH His-fusion 319 108
amAb yields were measured in unpurified supernatants by Protein A HPLC and His-fusion yields were
estimated after purification on Nickel Sepharose Excel.
34
Table 2: Summary of optimized conditions used for the CHO-3E7/BCDT method.
Condition Optimal value
Cell density at transfection 5x106 cells/ml Plasmid DNA concentration 1.3 µg/ml
PEI MAX concentration 7.5 µg/ml
Additives (pre-transfection) Dimethylacetamide (0.075% [v/v]) Additives (post-transfection) Transfectory Supplement (1:10)
Anti-Clumping (1:500) Glucose, maintain >10 mM
Culture duration 14 days
