[go: up one dir, main page]
More Web Proxy on the site http://driver.im/
You seem to have javascript disabled. Please note that many of the page functionalities won't work as expected without javascript enabled.
 
 
Sign in to use this feature.

Years

Between: -

Subjects

remove_circle_outline
remove_circle_outline
remove_circle_outline
remove_circle_outline
remove_circle_outline
remove_circle_outline

Journals

Article Types

Countries / Regions

Search Results (23)

Search Parameters:
Keywords = CHO cell cultivation

Order results
Result details
Results per page
Select all
Export citation of selected articles as:
5 pages, 6695 KiB  
Proceeding Paper
Investigation the Optical Contrast Between Nanofiber Mats and Mammalian Cells Dyed with Fluorescent and Other Dyes
by Nora Dassmann, Bennet Brockhagen and Andrea Ehrmann
Phys. Sci. Forum 2024, 10(1), 5; https://doi.org/10.3390/psf2024010005 - 26 Dec 2024
Viewed by 346
Abstract
Electrospinning can be used to prepare nanofiber mats from diverse polymers and polymer blends. A large area of research is the application of nanofibrous membranes for tissue engineering. Typically, cell adhesion and proliferation as well as the viability of mammalian cells are tested [...] Read more.
Electrospinning can be used to prepare nanofiber mats from diverse polymers and polymer blends. A large area of research is the application of nanofibrous membranes for tissue engineering. Typically, cell adhesion and proliferation as well as the viability of mammalian cells are tested by seeding the cells on substrates, cultivating them for a defined time and finally dyeing them to enable differentiation between cells and substrates under a white light or fluorescence microscope. While this procedure works well for cells cultivated in well plates or petri dishes, other substrates may undesirably also be colored by the dye. Here we show investigations of the optical contrast between dyed CHO DP-12 (Chinese hamster ovary) cells and different electrospun nanofiber mats, dyed with haematoxylin-eosin (H&E), PromoFluor 488 premium, 4,6-diamidino-2-phenylindole (DAPI) or Hoechst 33342, and give the optimum dyeing parameters for maximum optical contrast between cells and nanofibrous substrates. Full article
(This article belongs to the Proceedings of The 1st International Online Conference on Photonics)
Show Figures

Figure 1

Figure 1
<p>Cells grown on the well ground, dyed with (<b>a</b>) H&amp;E; (<b>b</b>) PromoFluor; (<b>c</b>) DAPI; and (<b>d</b>) Hoechst.</p>
Full article ">Figure 2
<p>Cells grown on a PAN/gelatin nanofiber mat, dyed with (<b>a</b>) H&amp;E; (<b>b</b>) PromoFluor; (<b>c</b>) DAPI; and (<b>d</b>) Hoechst.</p>
Full article ">
11 pages, 1357 KiB  
Article
Application of a Novel Disposable Flow Cell for Spectroscopic Bioprocess Monitoring
by Tobias Steinwedel, Philipp Raithel, Jana Schellenberg, Carlotta Kortmann, Pia Gellermann, Mathias Belz and Dörte Solle
Chemosensors 2024, 12(10), 202; https://doi.org/10.3390/chemosensors12100202 - 1 Oct 2024
Cited by 1 | Viewed by 1134
Abstract
The evaluation of the analytical capabilities of a novel disposable flow cell for spectroscopic bioprocess monitoring is presented. The flow cell is presterilized and can be connected to any kind of bioreactor by weldable tube connections. It is clamped into a reusable holder, [...] Read more.
The evaluation of the analytical capabilities of a novel disposable flow cell for spectroscopic bioprocess monitoring is presented. The flow cell is presterilized and can be connected to any kind of bioreactor by weldable tube connections. It is clamped into a reusable holder, which is equipped with SMA-terminated optical fibers or an integrated light source and detection unit. This modular construction enables spectroscopic techniques like UV-Vis spectroscopy or turbidity measurements by scattered light for modern disposable bioreactors. A NIR scattering module was used for biomass monitoring in different cultivations. A high-cell-density fed-batch cultivation with Komagataella phaffii and a continuous perfusion cultivation with a CHO DG44 cell line were conducted. A high correlation between the sensor signal and biomass or viable cell count was observed. Furthermore, the sensor shows high sensitivity during low turbidity states, as well as a high dynamic range to monitor high turbidity values without saturation effects. In addition to upstream processing, the sensor system was used to monitor the purification process of a monoclonal antibody. The absorption module enables simple and cost-efficient monitoring of downstream processing and quality control measurements. Recorded absorption spectra can be used for antibody aggregate detection, due to an increase in overall optical density. Full article
Show Figures

Figure 1

Figure 1
<p>Monitoring of a fed-batch cultivation with <span class="html-italic">Komagataella phaffii</span> [<a href="#B20-chemosensors-12-00202" class="html-bibr">20</a>]. (<b>A</b>) Scattering in various angles with offline measurements of optical density (OD) and cell dry weight (DCW). (<b>B</b>) the opacity is compared to the 90°-scattered light for optical densities (OD) up to 10.</p>
Full article ">Figure 2
<p>Opacity and viable cell density of a continuous perfusion cultivation of CHO DG44 cells [<a href="#B19-chemosensors-12-00202" class="html-bibr">19</a>]. The non-continuous course of the cell density is due to the process control [<a href="#B19-chemosensors-12-00202" class="html-bibr">19</a>], the online monitoring with the developed sensor shows changes quickly and reliably.</p>
Full article ">Figure 3
<p>Monitoring of an FPLC run of Protein A capture, comparison of different UV detectors.</p>
Full article ">Figure 4
<p>SEC-HPLC results to detect different rates of antibody aggregates by freeze–thaw cycles (FT) during downstream processing. (<b>A</b>) Complete HPLC chromatogram; (<b>B</b>) Focus on aggregate peak at 7.5 min.</p>
Full article ">Figure 5
<p>Absorption spectra for different aggregate rates by freeze–thaw cycles (FT). (<b>A</b>) FPLC chromatogram in normal scale; (<b>B</b>) FPLC chromatogram in logarithmic scale.</p>
Full article ">
18 pages, 2735 KiB  
Article
Stable Production of a Recombinant Single-Chain Eel Follicle-Stimulating Hormone Analog in CHO DG44 Cells
by Munkhzaya Byambaragchaa, Sei Hyen Park, Sang-Gwon Kim, Min Gyu Shin, Shin-Kwon Kim, Myung-Hum Park, Myung-Hwa Kang and Kwan-Sik Min
Int. J. Mol. Sci. 2024, 25(13), 7282; https://doi.org/10.3390/ijms25137282 - 2 Jul 2024
Cited by 1 | Viewed by 1289
Abstract
This study aimed to produce single-chain recombinant Anguillid eel follicle-stimulating hormone (rec-eel FSH) analogs with high activity in Cricetulus griseus ovary DG44 (CHO DG44) cells. We recently reported that an O-linked glycosylated carboxyl-terminal peptide (CTP) of the equine chorionic gonadotropin (eCG) β-subunit contributes [...] Read more.
This study aimed to produce single-chain recombinant Anguillid eel follicle-stimulating hormone (rec-eel FSH) analogs with high activity in Cricetulus griseus ovary DG44 (CHO DG44) cells. We recently reported that an O-linked glycosylated carboxyl-terminal peptide (CTP) of the equine chorionic gonadotropin (eCG) β-subunit contributes to high activity and time-dependent secretion in mammalian cells. We constructed a mutant (FSH-M), in which a linker including the eCG β-subunit CTP region (amino acids 115–149) was inserted between the β-subunit and α-subunit of wild-type single-chain eel FSH (FSH-wt). Plasmids containing eel FSH-wt and eel FSH-M were transfected into CHO DG44 cells, and single cells expressing each protein were isolated from 10 and 7 clones. Secretion increased gradually during the cultivation period and peaked at 4000–5000 ng/mL on day 9. The molecular weight of eel FSH-wt was 34–40 kDa, whereas that of eel FSH-M increased substantially, with two bands at 39–46 kDa. Treatment with PNGase F to remove the N glycosylation sites decreased the molecular weight remarkably to approximately 8 kDa. The EC50 value and maximal responsiveness of eel FSH-M were approximately 1.23- and 1.06-fold higher than those of eel FSH-wt, indicating that the mutant showed slightly higher biological activity. Phosphorylated extracellular-regulated kinase (pERK1/2) activation exhibited a sharp peak at 5 min, followed by a rapid decline. These findings indicate that the new rec-eel FSH molecule with the eCG β-subunit CTP linker shows potent activity and could be produced in massive quantities using the stable CHO DG44 cell system. Full article
(This article belongs to the Special Issue New Sights into Bioinformatics of Gene Regulations and Structure)
Show Figures

Figure 1

Figure 1
<p>Shape of the colony before isolation from the 96-well plate. After 10 d of incubation, the colonies were visually examined under a microscope for monoclonal colony growth. Images of representative colonies were obtained from eel FSH-wt samples approximately 3 weeks post-plating with complete cloning medium in a non-shaking incubator. Colonies were selected and transferred into 24-well plates. Next, they were transferred into 6-well, T-25 flasks and 125 mL shaker flasks. The scale bars represent 50 μm.</p>
Full article ">Figure 2
<p>Western blotting analysis of rec-eel FSH-wt proteins produced from single cells. Supernatants from 10 colonies were collected on the day of cultivation in a shaking incubator. The supernatants were resolved by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and blotted onto membranes. Proteins were detected using a monoclonal antibody (anti-eel FSH5A14) and horseradish peroxidase-conjugated goat anti-mouse IgG antibodies. (<b>A</b>) In total, 20 µL of the supernatant on day 9 was loaded in the wells. Numbers denote isolated clone counts. (<b>B</b>) Colonies with substantial secretion were selected for Western blot analyses. We choose four colonies (eel FSH-wt 3, FSH-wt 5, FSH-wt 8, and FSH-wt 9) and 20 µL of supernatant was evaluated by Western blotting on the day of culture. FSH, follicle-stimulating hormone.</p>
Full article ">Figure 3
<p>Concentrations of rec-eel FSH-wt proteins secreted from CHO-DG44 cells on the day of culture. The supernatant was collected on days 0, 1, 3, 5, 9, and 11 of culture. The expression levels of rec-eel FSH-wt protein in monoclonal cells were analyzed using a sandwich enzyme-linked immunosorbent assay. Values are expressed as the mean ± standard error of mean from at least three independent experiments. FSH, follicle-stimulating hormone.</p>
Full article ">Figure 4
<p>Western blotting analysis of rec-eel FSH-M proteins produced by monoclonal cells. Supernatants were collected from seven colonies on the day of cultivation. The samples were prepared for SDS-PAGE. Membranes were detected using specific monoclonal antibodies (anti-eel FSH5A14). (<b>A</b>) In total, 20 µL collected on day 9 was loaded in the wells. Positive controls produced from the CHO-S cells were concentrated by 20 times and 20 µg was loaded in the wells. Two specific bands were detected for all samples. (<b>B</b>) Colonies judged to secrete large amounts were selected for Western blot analyses. We choose two colonies (eel FSH-M 3 and FSH-M 8) and 20 µL of the supernatant was used for Western blotting on the day of culture. Faint bands were first detected on day 3 and band intensity increased gradually, reaching a maximum on day 9.</p>
Full article ">Figure 5
<p>Deglycosylation results for eel FSH-wt and FSH-M proteins. The proteins collected from eel FSH-wt 3, FSH-M 3, and FSH-M 8 were treated with peptide-N-glycanase F to remove <span class="html-italic">N</span>-linked oligosaccharides, followed by Western blotting. The molecular weights of rec-eel FSH-wt and FSH-M decreased significantly to approximately 8–10 kDa.</p>
Full article ">Figure 6
<p>Concentrations of rec-eel FSH-M proteins secreted from CHO-DG44 cells on the day of culture. The supernatants were collected on days 0, 1, 3, 5, 9, and 11 of culture. The expression levels of rec-eel FSH-M in monoclonal cells were analyzed using a sandwich enzyme-linked immunosorbent assay. Values are expressed as the mean ± standard error of mean from at least three independent experiments.</p>
Full article ">Figure 7
<p>Effects of rec-eel FSH-wt and FSH-M on cyclic adenine monophosphate (cAMP) production in cells expressing eel follicle-stimulating hormone receptor. CHO-K1 cells transiently transfected with eel FSHR were seeded in 384-well plates (10,000 cells per well) at 24 h post-transfection. Cells were incubated with rec-eel FSH-wt or FSH-M for 30 min at room temperature. cAMP production was detected using a homogeneous time-resolved fluorescence assay and results are represented as Delta F%. Each data point represents the mean ± standard error of mean from triplicate experiments. The mean values were fitted to an equation to obtain a single-phase exponential decay curve.</p>
Full article ">Figure 8
<p>Time course for pERK1/2 activation in rec-eel FSH-wt and FSH-M. HEK293 cells were transiently transfected with eel FSH receptor, stimulated with 400 ng/mL agonist and normalized to the basal response. (<b>A</b>) Ratios are shown as delta F%. (<b>B</b>) The folds change values are shown, with 0 time set to 1-fold.</p>
Full article ">Figure 9
<p>pERK1/2 activation stimulated by eel FSH receptor. The eel FSH receptor was transiently transfected into HEK293 cells, and the cells were starved for 4–6 h and stimulated with a 400 ng/mL agonist for the indicated times. Whole-cell lysates were analyzed for pERK1/2 and total ERK levels. Twenty micrograms of protein were used in each sample lane. (<b>A</b>) Phosphorylation of ERK1/2 by western blotting. (<b>B</b>) The pERK and total ERK bands were quantified by densitometry, and pERK was normalized to total ERK levels. Representative data are shown, and graphs represent the mean and SE values from three independent experiments. No significant differences were observed between the curves representing eel FSH-wt- and FSH-M-treated samples.</p>
Full article ">Figure 10
<p>Schematic diagram of rec-eel FSH-wt and eel FSH-M. The tethered form of eel FSH β/α-wt containing the β-subunit and common α-subunit sequences was engineered. In the eel FSH-M mutant, the eCG β-subunit carboxyl-terminal peptide linker was inserted between the β-subunit and α-subunit using polymerase chain reaction. The eCG β-subunit CTP linker contained 35 amino acids sequence corresponding to the carboxyl-terminal peptide of the eCG β-subunit with approximately 12 <span class="html-italic">O</span>-linked oligosaccharide sites. The numbers in FSH-wt and FSH-M indicate the amino acids of the mature protein, except for the signal sequences. “N” denotes <span class="html-italic">N</span>-linked glycosylation sites at the eel FSH β-subunit and FSH α-subunit. Yellow indicates FSH β, the α-subunit is shown in blue, and the light blue shows the eCG CTP linker. eCTPβ (115–149) is the amino acid sequences of the eCG β-subunit CTP linker. Red in eCTPβ (115–149) denotes potential <span class="html-italic">O</span>-linked oligosaccharide sites.</p>
Full article ">
16 pages, 2619 KiB  
Article
Deciphering Metabolic Pathways in High-Seeding-Density Fed-Batch Processes for Monoclonal Antibody Production: A Computational Modeling Perspective
by Carolin Bokelmann, Alireza Ehsani, Jochen Schaub and Fabian Stiefel
Bioengineering 2024, 11(4), 331; https://doi.org/10.3390/bioengineering11040331 - 28 Mar 2024
Viewed by 2012
Abstract
Due to their high specificity, monoclonal antibodies (mAbs) have garnered significant attention in recent decades, with advancements in production processes, such as high-seeding-density (HSD) strategies, contributing to improved titers. This study provides a thorough investigation of high seeding processes for mAb production in [...] Read more.
Due to their high specificity, monoclonal antibodies (mAbs) have garnered significant attention in recent decades, with advancements in production processes, such as high-seeding-density (HSD) strategies, contributing to improved titers. This study provides a thorough investigation of high seeding processes for mAb production in Chinese hamster ovary (CHO) cells, focused on identifying significant metabolites and their interactions. We observed high glycolytic fluxes, the depletion of asparagine, and a shift from lactate production to consumption. Using a metabolic network and flux analysis, we compared the standard fed-batch (STD FB) with HSD cultivations, exploring supplementary lactate and cysteine, and a bolus medium enriched with amino acids. We reconstructed a metabolic network and kinetic models based on the observations and explored the effects of different feeding strategies on CHO cell metabolism. Our findings revealed that the addition of a bolus medium (BM) containing asparagine improved final titers. However, increasing the asparagine concentration in the feed further prevented the lactate shift, indicating a need to find a balance between increased asparagine to counteract limitations and lower asparagine to preserve the shift in lactate metabolism. Full article
(This article belongs to the Special Issue Metabolic Modeling and Engineering)
Show Figures

Graphical abstract

Graphical abstract
Full article ">Figure 1
<p>Visualization of the mechanistic model of central carbon metabolism of CHO cells. This figure shows the reactions and metabolites included in the mechanistic metabolic model of CHO cells in the HSD process derived from the metabolic network. Reactions that differ in the different model structures are shown in grey.</p>
Full article ">Figure 2
<p>Comparison of viable cell density, viability, titer, and extracellular fluxes between different cultivation conditions. (<b>a</b>) Normalized viable cell density (VCD), viability, and titer, as well as (<b>b</b>) fluxes derived from concentration measurements for the biomass (growth), the mAb (productivity), glucose (Glc), lactate (Lac), asparagine (Asn), glutamine (Gln), aspartate (Asp), and isoleucine (Ile) formation of the standard fed-batch (STD FB) and high-seeding-density (HSD) processes without additions (control) and with additional lactate and cysteine feed (LAC + CYS) or bolus medium addition (BM) over 12 to 14 days are shown. Negative fluxes indicate consumption while positive fluxes signify secretion.</p>
Full article ">Figure 3
<p>Carbon source and effect on productivity: (<b>a</b>) ratio of lactate to glucose determined as the ratio of the uptake fluxes, and (<b>b</b>) the mAb productivity accordingly for the standard fed-batch (STD FB) and the three high-seeding-density (HSD) processes. A positive ratio suggests that the cells take up glucose and lactate in parallel, while a negative ratio suggests that glucose is consumed while lactate is produced. The background colors indicate two different metabolic phases.</p>
Full article ">Figure 4
<p>Prediction of the ensemble of models for the control run. The combined predictions are given as the means of the individual model predictions. The whole range of predictions of the models in the ensemble is also visualized for all metabolites and states included in the models.</p>
Full article ">Figure 5
<p>Comparison of the concentration curves for cultivations with different asparagine ratios in the feed media. The simulated development of the concentrations for the original feeding medium, a feeding medium with reduced asparagine concentration, and a feeding medium with elevated asparagine concentration compared to the original medium are shown.</p>
Full article ">Figure 6
<p>Comparison of enhanced and control asparagine feed in the experiments and simulations based on (<b>a</b>) lactate production and (<b>b</b>) lactate concentration ratio in the HSD setting.</p>
Full article ">
12 pages, 1344 KiB  
Technical Note
Automated Production at Scale of Induced Pluripotent Stem Cell-Derived Mesenchymal Stromal Cells, Chondrocytes and Extracellular Vehicles: Towards Real-Time Release
by Laura Herbst, Ferdinand Groten, Mary Murphy, Georgina Shaw, Bastian Nießing and Robert H. Schmitt
Processes 2023, 11(10), 2938; https://doi.org/10.3390/pr11102938 - 10 Oct 2023
Cited by 4 | Viewed by 2321
Abstract
Induced pluripotent stem cell (iPSC)-derived mesenchymal stem cells (iMSCs) are amenable for use in a clinical setting for treatment of osteoarthritis (OA), which remains one of the major illnesses worldwide. Aside from iPSC-derived iMSCs, chondrocytes (iCHO) and extracellular vesicles (EV) are also promising [...] Read more.
Induced pluripotent stem cell (iPSC)-derived mesenchymal stem cells (iMSCs) are amenable for use in a clinical setting for treatment of osteoarthritis (OA), which remains one of the major illnesses worldwide. Aside from iPSC-derived iMSCs, chondrocytes (iCHO) and extracellular vesicles (EV) are also promising candidates for treatment of OA. Manufacturing and quality control of iPSC-derived therapies is mainly manual and thus highly time consuming and susceptible to human error. A major challenge in translating iPSC-based treatments more widely is the lack of sufficiently scaled production technologies from seeding to fill-and-finish. Formerly, the Autostem platform was developed for the expansion of tissue-derived MSCs at scale in stirred tank bioreactors and subsequent fill-and-finish. Additionally, the StemCellDiscovery platform was developed to handle plate-based cultivation of adherent cells including their microscopic analysis. By combining the existing automation technology of both platforms, all required procedures can be integrated in the AutoCRAT system, designed to handle iPSC expansion, differentiation to iMSCs and iCHOs, pilot scale expansion, and formulation of iMSCs as well as extracellular vesicles and their purification. Furthermore, the platform is equipped with several in-line and at-line assays to determine product quality, purity, and safety. This paper highlights the need for adaptable and modular automation concepts. It also stresses the importance of ensuring safety of generated therapies by incorporating automated release testing and cleaning solutions in automated systems. The adapted platform concepts presented here will help translate these technologies for clinical production at the necessary scale. Full article
(This article belongs to the Special Issue Application of Deep Learning in Pharmaceutical Manufacturing)
Show Figures

Figure 1

Figure 1
<p>Scheme of the AutoCRAT production processes and quality controls.</p>
Full article ">Figure 2
<p>Visual representation of the AutoCRAT platforms with dimensions: StemCellDiscovery with (1) robot, (2) cooled tube storage at 4 °C, (3) freezer tube storage at −20 °C, (4) plate storage, (5) pipette tip storage, (6) endotoxin tester cassette storage, (7) plate sealer, (8) incubator, (9) PCR, (10) endotoxin tester, (11) liquid handling unit, (12) decapper for 50 mL tubes, (13) decapper for 5 mL and 1 mL tubes, (14) high-speed microscope, (15) plate reader, (16) liquid waste, (17) solid waste, (18) centrifuge; Autostem with (19) and (20) robots, (21) Eppendorf bioreactor, (22) and (23) Applikon Bioreactors, (24) sampling station for cell counting, (25) freezer storage at −80 °C, (26) solid waste, (27) Nucleocounter cassette, cool container and 1 mL tube storage, (28) Nucleocounter, (29) preheaters, (30) cooled media and liquid waste storage at 4 °C, (31) decapper for centrifuge bottles, 5 mL and 1 mL tubes, (32) centrifuge, (33) pipette, (34) solid waste, (35) pipette tip storage, (36) centrifuge bottle and 5 mL tube storage, (37) hatch, (38) fluid transfer device; EV-module with (39) fraction collector and isolator, (40) fast protein liquid chromatography. Stations to be changed are highlighted in green.</p>
Full article ">
17 pages, 5488 KiB  
Article
Exploring Parametric and Mechanistic Differences between Expi293FTM and ExpiCHO-STM Cells for Transient Antibody Production Optimization
by Jing Zhou, Guoying Grace Yan, David Cluckey, Caryl Meade, Margaret Ruth, Rhady Sorm, Amy S. Tam, Sean Lim, Constantine Petridis, Laura Lin, Aaron M. D’Antona and Xiaotian Zhong
Antibodies 2023, 12(3), 53; https://doi.org/10.3390/antib12030053 - 10 Aug 2023
Cited by 3 | Viewed by 4765
Abstract
Rapidly producing drug-like antibody therapeutics for lead molecule discovery and candidate optimization is typically accomplished by large-scale transient gene expression technologies (TGE) with cultivated mammalian cells. The TGE methodologies have been extensively developed over the past three decades, yet produce significantly lower yields [...] Read more.
Rapidly producing drug-like antibody therapeutics for lead molecule discovery and candidate optimization is typically accomplished by large-scale transient gene expression technologies (TGE) with cultivated mammalian cells. The TGE methodologies have been extensively developed over the past three decades, yet produce significantly lower yields than the stable cell line approach, facing the technical challenge of achieving universal high expression titers for a broad range of antibodies and therapeutics modalities. In this study, we explored various parameters for antibody production in the TGE cell host Expi293FTM and ExpiCHO-STM with the transfection reagents ExpiFectamineTM and polyethylenimine. We discovered that there are significant differences between Expi293FTM and ExpiCHO-STM cells with regards to DNA complex formation time and ratio, complex formation buffers, DNA complex uptake trafficking routes, responses to dimethyl sulfoxide and cell cycle inhibitors, as well as light-chain isotype expression preferences. This investigation mechanistically dissected the TGE processes and provided a new direction for future transient antibody production optimization. Full article
(This article belongs to the Section Antibody Discovery and Engineering)
Show Figures

Figure 1

Figure 1
<p>Optimizing the dilution buffer processes for the DNA complexation with ExpiFectamine™ and PEI for Expi293F<sup>TM</sup> and ExpiCHO-S<sup>TM</sup>. As described in <a href="#sec2dot4-antibodies-12-00053" class="html-sec">Section 2.4</a>, the target DNAs and the transfection reagents were diluted separately in different volumes of Opti-MEM<sup>TM</sup> (mL/L cell culture volume per tube); prior to the mixing of the two components for the complexation formation. For Expi293F<sup>TM</sup>, the expression titers were determined five days post-transfection. For ExpiCHO-S<sup>TM</sup>, the expression titers were measured seven days post-transfection. (<b>A</b>) The Opti-MEM<sup>TM</sup> medium dilution conditions for the DNA complexation with ExpiFectamine™ 293 and PEI (*, <span class="html-italic">p</span> &lt; 0.05) in Expi293F<sup>TM</sup> for antibody-A. The titers obtained under the conditions of 10 mL/L per tube were set as 100%. (<b>B</b>) The Opti-MEM<sup>TM</sup> medium dilution conditions for the DNA complexation with ExpiFectamine™ CHO (**, <span class="html-italic">p</span> &lt; 0.05) and PEI in ExpiCHO-S<sup>TM</sup> for antibody-A. The titers obtained under the conditions of 10 mL/L per tube were set as 100%. (<b>C</b>) The pH effects of the dilution buffers for the DNA complexation with ExpiFectamine™ 293 (#, <span class="html-italic">p</span> &lt; 0.05) and PEI (#, <span class="html-italic">p</span> &lt; 0.05) in Expi293F<sup>TM</sup> for antibody-A. The Opti-MEM<sup>TM</sup> medium and other indicated dilution buffers in different pHs were used for diluting the DNAs and the transfection agents in 100 mL/L per tube. The titers obtained under the conditions of Opti-MEM<sup>TM</sup> medium were set as 100%. (<b>D</b>) The pH effects of the dilution buffers for the DNA complexation with ExpiFectamine™-CHO (&amp;, <span class="html-italic">p</span> &lt; 0.05) and PEI (&amp;, <span class="html-italic">p</span> &lt; 0.05) in ExpiCHO-S<sup>TM</sup> for antibody-A. The titers obtained under the conditions of Opti-MEM<sup>TM</sup> medium were set as 100%.</p>
Full article ">Figure 2
<p>Optimizing the ratios between PEI and DNA in Expi293F<sup>TM</sup> and ExpiCHO-S<sup>TM</sup>. As described in Materials &amp; Methods, the target DNAs encoding antibody-A were incubated with PEI in various ratios and transfected into Expi293F<sup>TM</sup> and ExpiCHO-S<sup>TM</sup> cells. For Expi293F<sup>TM</sup>, the expression titers were determined five days post-transfection. For ExpiCHO-S<sup>TM</sup>, the expression titers were measured seven days post-transfection. The titers obtained for those with ExpiFectamine™ were set as 100% (<span class="html-italic">n</span> = 4 ± S.D., ##, <span class="html-italic">p</span> &lt; 0.05)).</p>
Full article ">Figure 3
<p>Optimizing the DNA complex formation time in Expi293F<sup>TM</sup> and ExpiCHO-S<sup>TM</sup> with ExpiFectamine™ and PEI. As described in <a href="#sec2dot4-antibodies-12-00053" class="html-sec">Section 2.4</a>, the target DNAs encoding antibody-A were incubated with either ExpiFectamine™ (Panel <b>A</b>) or PEI (Panel <b>B</b>) for various time points (−0.5 min: the transfection reagents and the DNAs were directly added into the cell culture without mixing; 0 min: the transfection reagents and the DNAs were mixed, but without incubation prior to the addition to cell culture; incubation time after mixing: 0.5 min, 1 min, 2.5 min, 5 min, 10 min, 15 min, and 30 min), and transfected into Expi293F<sup>TM</sup> and ExpiCHO-S<sup>TM</sup> cells. The highest titers with each transfection reagent in each cell host were set as 100% (<span class="html-italic">n</span> = 3 ± S.D., ***, <span class="html-italic">p</span> &lt; 0.05).</p>
Full article ">Figure 4
<p>Size measurement of the DNA: ExpiFectamine™CHO or PEI complex with DLS. DLS measurements for size determination were described in <a href="#sec2dot8-antibodies-12-00053" class="html-sec">Section 2.8</a>. ExpiFectamine™CHO alone, or PEI alone, or in complex with the target DNAs encoding antibody-A (DNA:PEI = 1:3.5; DNA: ExpiFectamine™CHO = 1 µg:3.2 µL) performed in Opti-MEM<sup>TM</sup> medium (#, <span class="html-italic">p</span> &lt; 0.05, ##, <span class="html-italic">p</span> &lt; 0.05).</p>
Full article ">Figure 5
<p>Expi293F<sup>TM</sup> and ExpiCHO-S<sup>TM</sup> cells responded differently to the endocytosis blockers. As described in <a href="#sec2dot4-antibodies-12-00053" class="html-sec">Section 2.4</a>, various endocytosis blockers were added to the cell culture of Expi293F<sup>TM</sup> and ExpiCHO-S<sup>TM</sup> prior to the transfection of the target DNAs encoding antibody-A complexed with ExpiFectamine™. Cell viability in ExpiCHO-S<sup>TM</sup> (pane <b>A</b>), and Expi293F<sup>TM</sup> (panel <b>B</b>) as well as expression titers (panel <b>C</b>) were determined. The titers for the mock-treated control in each cell host were set as 100% (*, <span class="html-italic">p</span> &lt; 0.05, **, <span class="html-italic">p</span> &lt; 0.05).</p>
Full article ">Figure 6
<p>Co-transfecting the genes encoding cell cycle inhibitor p21 and p27 enhanced transient expression in Expi293F<sup>TM</sup> and ExpiCHO-S<sup>TM</sup> cells, with a bigger effect in ExpiCHO-S<sup>TM</sup>. As described in <a href="#sec2dot4-antibodies-12-00053" class="html-sec">Section 2.4</a>, the target DNAs encoding antibody-A, antibody-B, and antibody C complexed with ExpiFectamine™ along with different amounts of plasmid DNAs encoding p21/p27 were transfected into Expi293F<sup>TM</sup> (Panel <b>A</b>) and ExpiCHO-S<sup>TM</sup> (Panel <b>B</b>) cells. The control titers without p21/p27 DNAs were set as 100% (<span class="html-italic">n</span> = 3 ± S.D, #, <span class="html-italic">p</span> &lt; 0.05).</p>
Full article ">Figure 7
<p>DMSO with increased concentrations could enhance more transient expression in ExpiCHO-S<sup>TM</sup> cells than in Expi293F<sup>TM</sup> cells. As described in <a href="#sec2dot4-antibodies-12-00053" class="html-sec">Section 2.4</a>, Expi293F<sup>TM</sup> (Panel <b>A</b>) and ExpiCHO-S<sup>TM</sup> (Panel <b>B</b>) cells were pretreated with different concentrations of DMSO prior to the DNA transfection for antibody-A. The control titers without DMSO treatment were set as 100% (<span class="html-italic">n</span> = 3 ± S.D., *, <span class="html-italic">p</span> &lt; 0.05, #, <span class="html-italic">p</span> &lt; 0.05).</p>
Full article ">Figure 8
<p>Expression preferences for kappa and lambda light chain isotypes were detected in Expi293F<sup>TM</sup> and ExpiCHO-S<sup>TM</sup>. As described in <a href="#sec2dot4-antibodies-12-00053" class="html-sec">Section 2.4</a>, the target DNAs encoding common light chain (CLC) antibodies A, B, and C with a kappa light chain or CLC antibody-D, E, and F with a lambda light chain were transfected into either Expi293F<sup>TM</sup> or ExpiCHO-S<sup>TM</sup> cells. (<b>A</b>) Expi293F<sup>TM</sup> produced high titers for CLC antibodies A-C with kappa light chain and for CLC antibodies D-F with a lambda light chain, yet ExpiCHO-S<sup>TM</sup> only expressed well for CLC antibodies A-C with kappa light chain (#, <span class="html-italic">p</span> &lt; 0.05). (<b>B</b>) Co-transfecting heavy chain (HC) of CLC antibody E with both kappa (LCK) and lambda (LCL) light chain into either Expi293F<sup>TM</sup> or ExpiCHO-S<sup>TM</sup> cells resulted in 1:1 kappa/lambda expression ratio in Expi293F<sup>TM</sup> cells but 99:1 expression ratio in ExpiCHO-S<sup>TM</sup> cells. (<b>C</b>) Mass spectrometry analysis of the kappa and lambda co-transfection in Expi293F<sup>TM</sup> cells. (<b>D</b>) Mass spectrometry analysis of the kappa and lambda co-transfection in ExpiCHO-S<sup>TM</sup> cells.</p>
Full article ">
13 pages, 1995 KiB  
Article
Cost-Effective Protein Production in CHO Cells Following Polyethylenimine-Mediated Gene Delivery Showcased by the Production and Crystallization of Antibody Fabs
by Klaudia Meskova, Katarina Martonova, Patricia Hrasnova, Kristina Sinska, Michaela Skrabanova, Lubica Fialova, Stefana Njemoga, Ondrej Cehlar, Olga Parmar, Petr Kolenko, Vladimir Pevala and Rostislav Skrabana
Antibodies 2023, 12(3), 51; https://doi.org/10.3390/antib12030051 - 4 Aug 2023
Viewed by 4983
Abstract
Laboratory production of recombinant mammalian proteins, particularly antibodies, requires an expression pipeline assuring sufficient yield and correct folding with appropriate posttranslational modifications. Transient gene expression (TGE) in the suspension-adapted Chinese Hamster Ovary (CHO) cell lines has become the method of choice for this [...] Read more.
Laboratory production of recombinant mammalian proteins, particularly antibodies, requires an expression pipeline assuring sufficient yield and correct folding with appropriate posttranslational modifications. Transient gene expression (TGE) in the suspension-adapted Chinese Hamster Ovary (CHO) cell lines has become the method of choice for this task. The antibodies can be secreted into the media, which facilitates subsequent purification, and can be glycosylated. However, in general, protein production in CHO cells is expensive and may provide variable outcomes, namely in laboratories without previous experience. While achievable yields may be influenced by the nucleotide sequence, there are other aspects of the process which offer space for optimization, like gene delivery method, cultivation process or expression plasmid design. Polyethylenimine (PEI)-mediated gene delivery is frequently employed as a low-cost alternative to liposome-based methods. In this work, we are proposing a TGE platform for universal medium-scale production of antibodies and other proteins in CHO cells, with a novel expression vector allowing fast and flexible cloning of new genes and secretion of translated proteins. The production cost has been further reduced using recyclable labware. Nine days after transfection, we routinely obtain milligrams of antibody Fabs or human lactoferrin in a 25 mL culture volume. Potential of the platform is established based on the production and crystallization of antibody Fabs and their complexes. Full article
Show Figures

Figure 1

Figure 1
<p>Development of expression platform. (<b>A</b>) Map of expression vector pCMV-3UTR, highlighting its individual components. The expression cassette contains Kozak sequence in blue, signal peptide in pink and 3′ UTR in red. NheI and PvuI sites were used for the cloning of the cassette into the parental pCMV-Script vector. Positions of individual cloning sites in the cassette are indicated as well. Designed in SnapGene Viewer v. 7.0.1. (<b>B</b>) Expression cassette of pCMV-3UTR. Colour coding as in panel A. Translated sequence of signal peptide is shadowed yellow. (<b>C</b>) The level of DC25 Fab expression in Chinese Hamster Ovary (CHO) cells from the plasmid with (pCMV-3UTR) or without (pCMV-MCS) the 3′ UTR. (<b>D</b>) Comparison of doubling times and viability of CHO cells cultivated in flasks made of glass or polycarbonate. In (<b>C</b>,<b>D</b>) the mean and standard deviation are indicated.</p>
Full article ">Figure 2
<p>Validation of transient gene expression from pCMV-3UTR plasmid in CHO cells after polyethylenimine (PEI) assisted gene delivery. (<b>A</b>) Coomassie-stained 12% SDS polyacrylamide gels after electrophoresis of cultivation media samples and/or purified proteins. The post-transfection day is indicated above the gels. All samples were analysed in non-reducing conditions, except the purified full-length DC25 and MN423, where the reduced samples were included as well, showing the positions of light chains (~25 kDa) and variably glycosylated heavy chains (~55 kDa). NR—non-reduced; R—reduced. (<b>B</b>) The yield in protein production from pCMV-3UTR plasmid in CHO cells after transfection through PEI and electroporation. Lactoferrin was expressed only through PEI transfection.</p>
Full article ">Figure 3
<p>Crystallization of DC11 Fab and complexes of antibodies with tau dGAE. (<b>A</b>) Isolation of dGAE-Fab complexes via size exclusion chromatography. Retention volumes are indicated above the peaks. Retention volume of monomeric Fabs and dGAE are in the range of the grey box. (<b>B</b>) Crystals of the DC11Fab; (<b>C</b>) crystals of the binary Complex I of MN423 Fab and dGAE; (<b>D</b>) crystals of the ternary Complex II of DC25 Fab, DC11 Fab and dGAE. Scale bar in (<b>B</b>–<b>D</b>) represents 100 μm.</p>
Full article ">
11 pages, 2230 KiB  
Article
Generic Chemometric Models for Metabolite Concentration Prediction Based on Raman Spectra
by Abdolrahim Yousefi-Darani, Olivier Paquet-Durand, Almut von Wrochem, Jens Classen, Jens Tränkle, Mario Mertens, Jeroen Snelders, Veronique Chotteau, Meeri Mäkinen, Alina Handl, Marvin Kadisch, Dietmar Lang, Patrick Dumas and Bernd Hitzmann
Sensors 2022, 22(15), 5581; https://doi.org/10.3390/s22155581 - 26 Jul 2022
Cited by 11 | Viewed by 2828
Abstract
Chemometric models for on-line process monitoring have become well established in pharmaceutical bioprocesses. The main drawback is the required calibration effort and the inflexibility regarding system or process changes. So, a recalibration is necessary whenever the process or the setup changes even slightly. [...] Read more.
Chemometric models for on-line process monitoring have become well established in pharmaceutical bioprocesses. The main drawback is the required calibration effort and the inflexibility regarding system or process changes. So, a recalibration is necessary whenever the process or the setup changes even slightly. With a large and diverse Raman dataset, however, it was possible to generate generic partial least squares regression models to reliably predict the concentrations of important metabolic compounds, such as glucose-, lactate-, and glutamine-indifferent CHO cell cultivations. The data for calibration were collected from various cell cultures from different sites in different companies using different Raman spectrophotometers. In testing, the developed “generic” models were capable of predicting the concentrations of said compounds from a dilution series in FMX-8 mod medium, as well as from an independent CHO cell culture. These spectra were taken with a completely different setup and with different Raman spectrometers, demonstrating the model flexibility. The prediction errors for the tests were mostly in an acceptable range (<10% relative error). This demonstrates that, under the right circumstances and by choosing the calibration data carefully, it is possible to create generic and reliable chemometric models that are transferrable from one process to another without recalibration. Full article
(This article belongs to the Section Biosensors)
Show Figures

Figure 1

Figure 1
<p>(<b>A</b>) Raman spectra acquired from a single cell culture. (<b>B</b>) Spectra after preprocessing revealing a baseline correction and providing relevant Raman contribution.</p>
Full article ">Figure 2
<p>Normalized prediction results of glucose (<b>A</b>), lactate (<b>B</b>), glutamine (<b>C</b>), and glutamate (<b>D</b>) from the training set. Dashed red lines are the ideal model fit (1:1).</p>
Full article ">Figure 3
<p>Normalized prediction results of glucose (<b>A</b>), lactate (<b>B</b>), glutamine (<b>C</b>), and glutamate (<b>D</b>) from the test set. Dashed lines are the ideal model fit (1:1).</p>
Full article ">Figure 4
<p>Predicted vs. off-line values obtained by the generic models (glucose (<b>A</b>), lactate (<b>B</b>), glutamine (<b>C</b>), and glutamate (<b>D</b>)) on spectra obtained from a dilution series of the compounds in FMX-8 mod medium. Dashed lines are the ideal model fit (1:1).</p>
Full article ">Figure 5
<p>Prediction obtained by the glucose generic model (<b>A</b>) and the lactate generic model (<b>B</b>) on spectra acquired from an independent CHO cell fed-batch cultivation performed with BM/FM. Dashed lines are presented to make it easier to follow the general trend. Similar predictions for glutamine and glutamate are not shown, as their concentrations were too low (&lt;5 mmol/L) to be predicted accurately.</p>
Full article ">
11 pages, 4079 KiB  
Communication
Impact of Doxorubicin on Cell-Substrate Topology
by Andreas Krecsir, Verena Richter, Michael Wagner and Herbert Schneckenburger
Int. J. Mol. Sci. 2022, 23(11), 6277; https://doi.org/10.3390/ijms23116277 - 3 Jun 2022
Cited by 2 | Viewed by 2079
Abstract
Variable-Angle Total Internal Reflection Fluorescence Microscopy (VA-TIRFM) is applied in view of early detection of cellular responses to the cytostatic drug doxorubicin. Therefore, we determined cell-substrate topology of cultivated CHO cells transfected with a membrane-associated Green Fluorescent Protein (GFP) in the nanometer range [...] Read more.
Variable-Angle Total Internal Reflection Fluorescence Microscopy (VA-TIRFM) is applied in view of early detection of cellular responses to the cytostatic drug doxorubicin. Therefore, we determined cell-substrate topology of cultivated CHO cells transfected with a membrane-associated Green Fluorescent Protein (GFP) in the nanometer range prior to and subsequent to the application of doxorubicin. Cell-substrate distances increased up to a factor of 2 after 24 h of application. A reduction of these distances by again a factor 2 was observed upon cell aging, and an influence of the cultivation time is presently discussed. Applicability of VA-TIRFM was supported by measurements of MCF-7 breast cancer cells after membrane staining and incubation with doxorubicin, when cell-substrate distances increased again by a factor ≥ 2. So far, our method needs well-defined cell ages and staining of cell membranes or transfection with GFP or related molecules. Use of intrinsic fluorescence or even light-scattering methods to various cancer cell lines could make this method more universal in the future, e.g., in the context of early detection of apoptosis. Full article
(This article belongs to the Special Issue Biomedical Optics in Cell Biology)
Show Figures

Figure 1

Figure 1
<p>Fluorescence spectra of CHO-pAcGFP1-Mem cells at varying TIR angles after 0 h, 2 h, and 24 h incubation with doxorubicin (2 µM) in culture medium. Subcultures with 56–64 cell splittings.</p>
Full article ">Figure 2
<p>Fluorescence images at Θ = 66° excitation (upper part) and cell-substrate topology (lower part) for 0 h, 2 h, and 24 h incubation with doxorubicin (2 µM) displayed in a color code. CHO-pAcGFP1-Mem cells of subcultures 56–64 after 72 h (left, middle) or 96 h (right) growth in culture medium.</p>
Full article ">Figure 3
<p>Histograms of cell-substrate distances for representative TIRFM images of CHO-pAcGFP1-Mem cells after 0 h, 2 h, and 24 h incubation with doxorubicin (2 µM) in cultivation medium. The value at 0 nm (resulting from outside the cells) has been omitted. Inset: most frequent distances evaluated from all histograms as mean value ± standard deviation and <span class="html-italic">p</span>-values for statistical significance obtained from a <span class="html-italic">t</span>-test for 2 samples assuming unequal variances (<span class="html-italic">p</span> ≤ 0.05: statistically significant). Subcultures with 56–64 cell splittings.</p>
Full article ">Figure 4
<p>Most frequent cell-substrate distances evaluated from all histograms of the subcultures with 28–35, 48–50, and 56–64 cell splittings as mean value ± standard deviation including the <span class="html-italic">p</span>-values for statistical significance (<span class="html-italic">p</span> ≤ 0.05: statistically significant; n.s.: nonsignificant). Results are shown for 0 h and 2 h incubation with doxorubicin (2 µM).</p>
Full article ">Figure 5
<p>Fluorescence images (<b>a</b>,<b>b</b>) and fluorescence spectra (<b>c</b>) of CHO-pAcGFP1-Mem cells 24 h after incubation with doxorubicin for Θ = 62° (whole cell excitation) and Θ = 66° (TIR excitation). Fluorescence images excited at Θ = 62° show some additional red fluorescence due to doxorubicin, which almost disappears at Θ = 66°. Fluorescence spectra excited at Θ = 62° exhibit two emission maxima around 530 nm (GFP) and 543 nm with a long-wave tail, possibly related to doxorubicin or its degradation product. The long-wave part of the spectrum is less pronounced in the TIRFM experiments.</p>
Full article ">Figure 6
<p>Cell-substrate distances of MCF-7 breast cancer cells incubated with the fluorescence membrane marker laurdan (8 µM, 1 h) prior to and after incubation with doxorubicin (2 µM, 2 h), as evaluated in the spectral maxima at 500–520 nm of VA-TIRFM experiments of <span class="html-italic">n</span> = 11 samples in each case. Mean values ± standard deviations including <span class="html-italic">p</span>-value for statistical significance (<span class="html-italic">p</span> ≤ 0.05: statistically significant). TIRFM images of MCF-7 cells prior to (0 h) and after (2 h) incubation with doxorubicin.</p>
Full article ">Figure 7
<p>Flow chart of the MATLAB script used for automated cell-substrate calculations (GUI = Graphical User Interface; T(Θ) = transmission factor; d(θ) = penetration depth of the evanescent field). Once the parameters, e.g., the corresponding metric pixel size and the refractive indices, are set and the recorded images are imported into the MATLAB environment, the code gradually starts its calculations. Images are implemented as a virtual stack with Ni corresponding to the maximum number of images for each set. Beginning with a loop covering the image stack, an algorithm is used to correct any possible shifts in the x and y directions, which may occur in individual experiments. Thus, it is ensured that every pixel addresses the same field of view of the recorded cell at various angles. Next, the angle of the recorded image is extracted from the image file name in order to calculate the transmission factor and the penetration depth, as reported above.</p>
Full article ">
25 pages, 6518 KiB  
Article
Single-Cell Analysis of CHO Cells Reveals Clonal Heterogeneity in Hyperosmolality-Induced Stress Response
by Nadiya Romanova, Julian Schmitz, Marie Strakeljahn, Alexander Grünberger, Janina Bahnemann and Thomas Noll
Cells 2022, 11(11), 1763; https://doi.org/10.3390/cells11111763 - 27 May 2022
Cited by 5 | Viewed by 4601
Abstract
Hyperosmolality can occur during industrial fed-batch cultivation processes of Chinese hamster ovary (CHO) cells as highly concentrated feed and base solutions are added to replenish nutrients and regulate pH values. Some effects of hyperosmolality, such as increased cell size and growth inhibition, have [...] Read more.
Hyperosmolality can occur during industrial fed-batch cultivation processes of Chinese hamster ovary (CHO) cells as highly concentrated feed and base solutions are added to replenish nutrients and regulate pH values. Some effects of hyperosmolality, such as increased cell size and growth inhibition, have been elucidated by previous research, but the impact of hyperosmolality and the specific effects of the added osmotic-active reagents have rarely been disentangled. In this study, CHO cells were exposed to four osmotic conditions between 300 mOsm/kg (physiologic condition) and 530 mOsm/kg (extreme hyperosmolality) caused by the addition of either high-glucose-supplemented industrial feed or mannitol as an osmotic control. We present novel single-cell cultivation data revealing heterogeneity in mass gain and cell division in response to these treatments. Exposure to extreme mannitol-induced hyperosmolality and to high-glucose-oversupplemented feed causes cell cycle termination, mtDNA damage, and mitochondrial membrane depolarization, which hints at the onset of premature stress-induced senescence. Thus, this study shows that both mannitol-induced hyperosmolality (530 mOsm/kg) and glucose overfeeding induce severe negative effects on cell growth and mitochondrial activity; therefore, they need to be considered during process development for commercial production. Full article
Show Figures

Graphical abstract

Graphical abstract
Full article ">Figure 1
<p>Cultivation data for stage cultivations of CHO DP-12 and DP-12 FUCCI cells. (<b>A</b>) Relative viable cell density (VCD) of CHO DP-12 and DP-12 FUCCI cells during stage cultivation under 300, 370, 460 and 530 mOsm/kg controlled either by addition of a highly supplemented feed (Glc) or the same feed, where only mannitol was added as control (Man). The cell densities were divided over the initial cell density to avoid the influence of small discrepancies in initial cell density. (<b>B</b>) Cell diameter of CHO DP-12 cells during stage cultivation with added high-supplemented feed (Glc) or high-mannitol feed (Man). (<b>C</b>) Cell diameter of DP-12 FUCCI cells during stage cultivation with added high-supplemented feed (Glc) or high-mannitol feed (Man). (<b>D</b>) Relative cell volume of CHO DP-12 and DP-12 FUCCI cells during stage cultivation with high-supplemented feed (Glc) or high-mannitol feed (Man).</p>
Full article ">Figure 2
<p>Cellular stress response, observed on a single cell or on a population level. (<b>A</b>) During a moderate osmolality increase, population-level observations register a uniform dose-dependent increase in cell size caused by hyperosmotic stress (left). However, the underlying heterogeneous increase in cell size can be observed only in the single-cell level (right). (<b>B</b>) During an intense osmolality increase, a more pronounced homogenous increase in cell size can be registered on a population level (left) or a heterogeneous cellular response can be seen on a single-cell level (right).</p>
Full article ">Figure 3
<p>Cell diameter dynamics for 0–4 days (0–96 h) of the single-cell cultivation of CHO DP-12 cells exposed to high-glucose (panel (<b>A</b>)) and high-mannitol (panel (<b>B</b>)) feed. The stages are ordered from left to right: 300 mOsm/kg (left, green); 370 mOsm/kg (light green, second from left); 460 mOsm/kg (orange, third from left); 530 mOsm/kg (red, fourth from left). The window was chosen between either 10–26 µm or 14–30 µm to best fit the data. Cellular division is seen as a vertical drop in the diameter curve.</p>
Full article ">Figure 4
<p>Data on single-cell cultivation of CHO DP-12 cells exposed to high-glucose and high-mannitol feed over the cultivation duration [h]. (<b>A</b>) Representative microscopic photographs of the CHO-DP-12 cells exposed for 96 h to different osmotic treatments (high-mannitol feed). (<b>B</b>) Relative diameter gain based on one randomly picked cell taken for one division period after at least 24 h of hyperosmolality exposure. Relative diameter [–], without dimension.</p>
Full article ">Figure 5
<p>Mitochondrial parameters measured on day 2 and day 4 of the stage cultivations. Here, 300, 370, 460, and 530 stand for 300, 370, 460 and 530 mOsm/kg ambient osmolalities. (<b>A</b>) Median fluorescence intensities of the cell population previously gated for live and single cells of CHO DP-12 cells stained with MitoTracker<sup>®</sup> Green FM, (ThermoFisher Scientific, Waltham, MA, USA) and MitoTracker<sup>®</sup> Red CMXRos (Cell Signalling Technology Inc., Danvers, MA, USA) exposed to high-mannitol and high-glucose feeds determined by the flow cytometry analysis. Relative mtDNA copy numbers (RQ Target/Reference) and calibrated normalized relative quantities (CNRQ normTar/Ref) of the CHO DP-12 cells exposed to (<b>B</b>) high-glucose feed and to (<b>C</b>) high-mannitol feed. (<b>D</b>) Volume, mtDNA copy number, and median fluorescence intensities of MitoTracker<sup>®</sup> Green FM, (ThermoFisher Scientific, Waltham, MA, USA) and MitoTracker<sup>®</sup> Red FM-stained (Cell Signalling Technology Inc., Danvers, MA, USA) populations relative to the parameters measured in the untreated populations (300 mOsm/kg condition). Statistical significance by a two-tailed Student’s t-test with the statistically significant threshold of <span class="html-italic">p</span> &lt; 0.05; the notations of * (<span class="html-italic">p</span> &lt; 0.05), ** (<span class="html-italic">p</span> ≤ 0.01), and *** (<span class="html-italic">p</span> ≤ 0.001) were used.</p>
Full article ">Figure 6
<p>Cell cycle distribution [%] of DP-12 FUCCI cells exposed to high-glucose or high-mannitol feed. Dead cells and cell doublets were gated out based on the FSC/SSC and the diagonal FCS-A vs. FCS-H plots, described in the Materials and Methods section. The gating of the phases was performed uniformly throughout all analyses based on [<a href="#B44-cells-11-01763" class="html-bibr">44</a>]. Cell cycle phase distribution [%] of the DP-12 FUCCI cells exposed to (<b>A</b>) high-glucose feed and (<b>B</b>) high-mannitol feed.</p>
Full article ">Figure A1
<p>Viable cell density (VCD) [10<sup>5</sup> cells/mL] and viability [%] of CHO DP-12 and DP-12 FUCCI cells exposed to high-mannitol and high-glucose feed against cultivation time [d]. (<b>A</b>) Viable cell density (VCD) [10<sup>5</sup> cells/mL] and viability [%] of CHO DP-12 cells exposed to high-glucose feed (left) and to high-mannitol feed (right). (<b>B</b>) Viable cell density (VCD) [10<sup>5</sup> cells/mL] and viability [%] of DP-12 FUCCI cells exposed to high-glucose feed (left) and to high-mannitol feed (right).</p>
Full article ">Figure A2
<p>Graphic representation of the doubling times of CHO DP-12 and DP-12 FUCCI cells relative to the doubling time in the reference culture (300 mOsm/kg) in response to high-mannitol and high-glucose feed exposure.</p>
Full article ">Figure A3
<p>Osmolality [mOsm/kg], lactate, and glucose concentrations [mMol/L] during CHO DP-12 and DP-12 FUCCI cultivations in four stages exposed to high-glucose (Glc) and high-mannitol (Man) feeds. (<b>A</b>) Osmolality [mOsm/kg], lactate, and glucose concentrations [mMol/L] during a batch cultivation of CHO DP-12 cells exposed to high-glucose (line) and to high-mannitol feed (dashed line). (<b>B</b>) Osmolality [mOsm/kg], lactate, and glucose concentrations [mMol/L] during a batch cultivation of DP-12 FUCCI cells exposed to high-glucose (line) and to high-mannitol feed (dashed line). The numbers in the legend stand for an osmotic condition and the added reagent, e.g., 370 Man is the 370 mOsm/kg stage in the high-mannitol cultivation.</p>
Full article ">Figure A4
<p>Viable cell density (VCD) and viability of CHO DP-12 (<b>left</b>) and DP-12 FUCCI (<b>right</b>) cells typical for cultivation under physiological osmolality (300 mOsm/kg).</p>
Full article ">
19 pages, 6552 KiB  
Article
Seed Train Intensification Using an Ultra-High Cell Density Cell Banking Process
by Jan Müller, Vivian Ott, Dieter Eibl and Regine Eibl
Processes 2022, 10(5), 911; https://doi.org/10.3390/pr10050911 - 5 May 2022
Cited by 12 | Viewed by 5318
Abstract
A current focus of biopharmaceutical research and production is seed train process intensification. This allows for intermediate cultivation steps to be avoided or even for the direct inoculation of a production bioreactor with cells from cryovials or cryobags. Based on preliminary investigations regarding [...] Read more.
A current focus of biopharmaceutical research and production is seed train process intensification. This allows for intermediate cultivation steps to be avoided or even for the direct inoculation of a production bioreactor with cells from cryovials or cryobags. Based on preliminary investigations regarding the suitability of high cell densities for cryopreservation and the suitability of cells from perfusion cultivations as inoculum for further cultivations, an ultra-high cell density working cell bank (UHCD-WCB) was established for an immunoglobulin G (IgG)-producing Chinese hamster ovary (CHO) cell line. The cells were previously expanded in a wave-mixed bioreactor with internal filter-based perfusion and a 1 L working volume. This procedure allows for cryovial freezing at 260 × 106 cells mL−1 for the first time. The cryovials are suitable for the direct inoculation of N−1 bioreactors in the perfusion mode. These in turn can be used to inoculate subsequent IgG productions in the fed-batch mode (low-seed fed-batch or high-seed fed-batch) or the continuous mode. A comparison with the standard approach shows that cell growth and antibody production are comparable, but time savings of greater than 35% are possible for inoculum production. Full article
Show Figures

Figure 1

Figure 1
<p>Schematic representation of the experiment design: (<b>1a</b>) Direct inoculation out of cryovials (B-CV-15) vs. standard inoculum production, freezing of different VCDs; (<b>1b</b>) Direct inoculation from standard cryovials vs. cryovials with higher cell densities (B-90–B250; 90–250 × 10<sup>6</sup> cells mL<sup>−1</sup>); (<b>2</b>) Perfusion processes (P01–P04) with standard inoculum production, daily inoculation of batch experiments (B-P03-d3–B-P03-d9), and freezing of the UHCD-WCB; (<b>3</b>) Performance test of the UHCD-WCB in batch (B-UHCD) and perfusion (P05) experiments.</p>
Full article ">Figure 2
<p>Scheme of the preparation of the cell suspension for freezing of the UHCD-WCB.</p>
Full article ">Figure 3
<p>Course of (<b>a</b>) VCD, (<b>b</b>) viability, (<b>c</b>) glucose (Glc), and (<b>d</b>) IgG concentration during the batch experiments inoculated directly from cryovials (nomenclature: B = Batch, CV = inoculated out of cryovial, 15 = VCD in cryovial (×10<sup>6</sup> cells mL<sup>−1</sup>)). Control: batch experiments with standard inoculum production (<span class="html-italic">n</span> = 3).</p>
Full article ">Figure 4
<p>Course of (<b>a</b>) VCD, (<b>b</b>) viability, (<b>c</b>) glucose (Glc), and (<b>d</b>) IgG concentration during the batch experiments inoculated directly from cryovials containing high VCDs from 90 to 250 × 10<sup>6</sup> cells mL<sup>−1</sup> (nomenclature: B = Batch, 90/115/150/180/250 = VCD in cryovial (×10<sup>6</sup> cells mL<sup>−1</sup>)). Control: batch experiments inoculated directly from cryovials containing lower VCDs (15–40 × 10<sup>6</sup> cells mL<sup>−1</sup>, <span class="html-italic">n</span> = 3).</p>
Full article ">Figure 5
<p>(<b>a</b>) Course of the VCD and viability of P01, P02, and P03, and (<b>b</b>) course of perfusion rate D and the CSPR of P01, P02, and P03.</p>
Full article ">Figure 6
<p>Course of (<b>a</b>) VCD, (<b>b</b>) viability, (<b>c</b>) glucose (Glc), and (<b>d</b>) IgG concentration during the batch experiments inoculated from perfusion cultivation (nomenclature: B = Batch, P03 = Source of inoculum, d5–d9 = day of cell harvest from P03). Control: batch experiments inoculated from P03 on days 3 and 4 (<span class="html-italic">n</span> = 4).</p>
Full article ">Figure 6 Cont.
<p>Course of (<b>a</b>) VCD, (<b>b</b>) viability, (<b>c</b>) glucose (Glc), and (<b>d</b>) IgG concentration during the batch experiments inoculated from perfusion cultivation (nomenclature: B = Batch, P03 = Source of inoculum, d5–d9 = day of cell harvest from P03). Control: batch experiments inoculated from P03 on days 3 and 4 (<span class="html-italic">n</span> = 4).</p>
Full article ">Figure 7
<p>Course of (<b>a</b>) VCD, (<b>b</b>) viability (<b>c</b>) glucose (Glc), and (<b>d</b>) IgG concentration during the batch experiments inoculated directly from the UHCD-WCB (nomenclature: B = Batch, UHCD = inoculated from UHCD-WCB). Control: batch experiments inoculated directly from cryovials containing lower VCDs (15 × 10<sup>6</sup> cells mL<sup>−1</sup>, <span class="html-italic">n</span> = 3).</p>
Full article ">Figure 7 Cont.
<p>Course of (<b>a</b>) VCD, (<b>b</b>) viability (<b>c</b>) glucose (Glc), and (<b>d</b>) IgG concentration during the batch experiments inoculated directly from the UHCD-WCB (nomenclature: B = Batch, UHCD = inoculated from UHCD-WCB). Control: batch experiments inoculated directly from cryovials containing lower VCDs (15 × 10<sup>6</sup> cells mL<sup>−1</sup>, <span class="html-italic">n</span> = 3).</p>
Full article ">Figure 8
<p>(<b>a</b>) Course of the VCD and viability of P05, and (<b>b</b>) course of D and the CSPR of P05, inoculated directly from the UHCD-WCB.</p>
Full article ">Figure A1
<p>Course of (<b>a</b>) glucose consumption rate, and (<b>b</b>) IgG production rate during the batch experiments inoculated directly from cryovials (nomenclature: B = Batch, CV = inoculated out of cryovial, 15 = VCD in cryovial (×10<sup>6</sup> cells mL<sup>−1</sup>)). Control: batch experiments with standard inoculum production (<span class="html-italic">n</span> = 3).</p>
Full article ">Figure A2
<p>Course of (<b>a</b>) glucose consumption rate, and (<b>b</b>) IgG production rate during the batch experiments inoculated directly from cryovials containing high VCDs from 90 to 250 × 10<sup>6</sup> cells mL<sup>−1</sup> (nomenclature: B = Batch, 90/115/150/180/250 = VCD in cryovial (×10<sup>6</sup> cells mL<sup>−1</sup>)). Control: batch experiments inoculated directly from cryovials containing lower VCDs (15–40 × 10<sup>6</sup> cells mL<sup>−1</sup>, <span class="html-italic">n</span> = 3).</p>
Full article ">Figure A3
<p>Course of (<b>a</b>) glucose consumption rate, and (<b>b</b>) IgG production rate during the batch experiments inoculated from perfusion cultivation (nomenclature: B = Batch, P03 = Source of inoculum, d5–d9 = day of cell harvest from P03). Control: batch experiments inoculated from P03 on days 3 and 4 (<span class="html-italic">n</span> = 4).</p>
Full article ">Figure A4
<p>Course of (<b>a</b>) glucose consumption rate, and (<b>b</b>) IgG production rate during the batch experiments inoculated directly from the UHCD-WCB (nomenclature: B = Batch, UHCD = inoculated from UHCD-WCB). Control: batch experiments inoculated directly from cryovials containing lower VCDs (15 × 10<sup>6</sup> cells mL<sup>−1</sup>, <span class="html-italic">n</span> = 3).</p>
Full article ">Figure A5
<p>Microscopic images (200×) of (<b>a</b>,<b>b</b>) cells after one-week standard inoculum production, (<b>c</b>,<b>d</b>) cells from a cryovial containing 15 × 10<sup>6</sup> cells mL<sup>−1</sup>, 1 d after thawing, and (<b>e</b>,<b>f</b>) cells from a cryovial containing 260 × 10<sup>6</sup> cells mL<sup>−1</sup>, 1 d after thawing. (<b>a</b>,<b>c</b>,<b>e</b>) phase contrast images, (<b>b</b>,<b>d</b>,<b>f</b>) phase contrast and fluorescence images (fluorescein diacetate staining for cells with intact cell membrane).</p>
Full article ">Figure A5 Cont.
<p>Microscopic images (200×) of (<b>a</b>,<b>b</b>) cells after one-week standard inoculum production, (<b>c</b>,<b>d</b>) cells from a cryovial containing 15 × 10<sup>6</sup> cells mL<sup>−1</sup>, 1 d after thawing, and (<b>e</b>,<b>f</b>) cells from a cryovial containing 260 × 10<sup>6</sup> cells mL<sup>−1</sup>, 1 d after thawing. (<b>a</b>,<b>c</b>,<b>e</b>) phase contrast images, (<b>b</b>,<b>d</b>,<b>f</b>) phase contrast and fluorescence images (fluorescein diacetate staining for cells with intact cell membrane).</p>
Full article ">
19 pages, 39555 KiB  
Article
New Insights from Locally Resolved Hydrodynamics in Stirred Cell Culture Reactors
by Fabian Freiberger, Jens Budde, Eda Ateş, Michael Schlüter, Ralf Pörtner and Johannes Möller
Processes 2022, 10(1), 107; https://doi.org/10.3390/pr10010107 - 5 Jan 2022
Cited by 13 | Viewed by 3524
Abstract
The link between hydrodynamics and biological process behavior of antibody-producing mammalian cell cultures is still not fully understood. Common methods to describe dependencies refer mostly to averaged hydrodynamic parameters obtained for individual cultivation systems. In this study, cellular effects and locally resolved hydrodynamics [...] Read more.
The link between hydrodynamics and biological process behavior of antibody-producing mammalian cell cultures is still not fully understood. Common methods to describe dependencies refer mostly to averaged hydrodynamic parameters obtained for individual cultivation systems. In this study, cellular effects and locally resolved hydrodynamics were investigated for impellers with different spatial hydrodynamics. Therefore, the hydrodynamics, mainly flow velocity, shear rate and power input, in a single- and a three-impeller bioreactor setup were analyzed by means of CFD simulations, and cultivation experiments with antibody-producing Chinese hamster ovary (CHO) cells were performed at various agitation rates in both reactor setups. Within the three-impeller bioreactor setup, cells could be cultivated successfully at much higher agitation rates as in the single-impeller bioreactor, probably due to a more uniform flow pattern. It could be shown that this different behavior cannot be linked to parameters commonly used to describe shear effects on cells such as the mean energy dissipation rate or the Kolmogorov length scale, even if this concept is extended by locally resolved hydrodynamic parameters. Alternatively, the hydrodynamic heterogeneity was statistically quantified by means of variance coefficients of the hydrodynamic parameters fluid velocity, shear rate, and energy dissipation rate. The calculated variance coefficients of all hydrodynamic parameters were higher in the setup with three impellers than in the single impeller setup, which might explain the rather stable process behavior in multiple impeller systems due to the reduced hydrodynamic heterogeneity. Such comprehensive insights lead to a deeper understanding of the bioprocess. Full article
Show Figures

Figure 1

Figure 1
<p>Fluid velocity u in the SIS and TIS. (<b>a</b>) Fluid velocity in the SIS; (<b>b</b>) Fluid velocity in the TIS. The shown data correspond to an agitation rate of 400 rpm. (<b>c</b>) Mean fluid velocity along the reactor height (slice plot) in the SIS for different agitation rates; (<b>d</b>) mean fluid velocity along the reactor height (slice plot) in the TIS for different agitation rates. In (<b>c</b>,<b>d</b>), the flow velocity was averaged across the diameter in steps of one millimeter along the reactor height.</p>
Full article ">Figure 2
<p>Energy dissipation rate ε in the SIS and TIS: (<b>a</b>) energy dissipation rate in the SIS; (<b>b</b>) energy dissipation rate in the TIS. The shown data correspond to an agitation rate of 400 rpm. (<b>d</b>) Mean energy dissipation rate along the reactor height (slice plot) in the SIS for different agitation rates; (<b>d</b>) mean energy dissipation rate along the reactor height (slice plot) in the TIS for different agitation rates. In (<b>c</b>,<b>d</b>), the energy dissipation rate was averaged across the diameter in steps of one millimeter along the reactor height.</p>
Full article ">Figure 3
<p>Mean growth curves (viable cell density X and viability) for the cultivations in the SIS at different agitation rates and respective averaged power inputs: 400 rpm (16 W m<sup>−3</sup>); 800 rpm (110 W m<sup>−3</sup>); 1200 rpm (356 W m<sup>−3</sup>). Data were averaged from three cultivations each.</p>
Full article ">Figure 4
<p>Growth curves (viable cell density X and viability) for the cultivations in the TIS at different agitation rates and respective averaged power inputs. The agitation rate settings range from 770 rpm or 454 W m<sup>−3</sup> to 1400 rpm or 4742 W m<sup>−3</sup>. One experiment was performed at each agitation rate.</p>
Full article ">Figure 5
<p>Maximum viable cell densities (VCD) compared for experiments in both reactor setups: cultivation results from the SIS (grey bars); cultivation results from the TIS (blue bars). Data for 400, 800, and 1200 rpm in the SIS were averaged from three cultivations each. Dashed arrows indicate qualitative trends for each reactor setup.</p>
Full article ">Figure 6
<p>Antibody space time yields (STY) compared for experiments in both reactor setups: cultivation results from the SIS (grey bars); cultivation results from the TIS (blue bars). Data for 400, 800, and 1200 rpm in the SIS were averaged from three cultivations each. Dashed arrows indicate expected trends for each reactor setup.</p>
Full article ">Figure 7
<p>Critical energy dissipation rate ε<sub>krit</sub> distributions (black circles) calculated from the cell size distributions in the SIS compared to the energy dissipation rate distribution in the reactor system (grey squares) obtained from CFD data. The averaged energy dissipation rates from CFD simulations are indicated in blue, their maxima in red, maximum energy dissipation rates calculated from slice plots are indicated in purple.</p>
Full article ">Figure 8
<p>Critical energy dissipation rate distributions (black circles) in the TIS compared to the energy dissipation rate distribution in the reactor system (grey squares). The averaged energy dissipation rates from CFD simulations are indicated in blue, their maxima in red, maximum energy dissipation rates calculated from slice plots are indicated in purple.</p>
Full article ">Figure 9
<p>Critical energy dissipation rate distributions (black circles) compared to the energy dissipation rate distribution in the SIS (600 rpm and 800 rpm) and in the TIS (450 rpm) (grey squares). The averaged energy dissipation rates from CFD simulations are indicated in blue, their maxima in red. Maximum energy dissipation rates calculated from slice plots are indicated in purple.</p>
Full article ">Figure 10
<p>Variance coefficients for hydrodynamic parameters: (<b>a</b>) fluid velocity u, (<b>b</b>) shear rate γ, and (<b>c</b>) energy dissipation rate ε in both cultivation systems. The coefficients were calculated from standard deviations from slice plot data divided by their mean value. They represent a measure of hydrodynamic homogeneity of the respective reactor system.</p>
Full article ">Figure A1
<p>Pitched blade impeller (<b>above</b>) and six blade impeller (<b>below</b>) with corresponding CAD models. For the pitched blade impeller, a Newton number Ne of 0.35 was determined with an empiric correlation [<a href="#B7-processes-10-00107" class="html-bibr">7</a>], while for the six blade impeller, 3.95 was considered, following another empiric correlation [<a href="#B31-processes-10-00107" class="html-bibr">31</a>].</p>
Full article ">Figure A2
<p>Shear rate γ in the SIS and TIS: (<b>a</b>) shear rate in the SIS; (<b>b</b>) shear rate in the TIS. The shown data correspond to an agitation rate of 400 rpm. (<b>d</b>) Mean shear rate along the reactor height (slice plot) in the SIS for different agitation rates; (<b>d</b>) mean shear rate along the reactor height (slice plot) in the TIS for different agitation rates. In (<b>c</b>,<b>d</b>), the shear rate was averaged across the diameter in steps of one millimeter along the reactor height.</p>
Full article ">Figure A3
<p>Exemplary cell size distributions from cultivation data in the SIS (800 rpm and 1200 rpm) and in the TIS (770 rpm and 1400 rpm). Distributions for all cultivations were taken from samples at 72 h cultivation time.</p>
Full article ">Figure A4
<p>Areas in the TIS with energy dissipation rates above 1000 W m<sup>−3</sup>.</p>
Full article ">Figure A5
<p>Critical reactor volume fractions with energy dissipation rates above 20 kW m<sup>−3</sup> plotted over the corresponding averaged power input. Black circles represent values for the SIS, grey squares those of the TIS.</p>
Full article ">
17 pages, 1109 KiB  
Article
Monte Carlo-Based Error Propagation for a More Reliable Regression Analysis across Specific Rates in Bioprocesses
by Julian Kager and Christoph Herwig
Bioengineering 2021, 8(11), 160; https://doi.org/10.3390/bioengineering8110160 - 26 Oct 2021
Cited by 6 | Viewed by 3220
Abstract
During process development, bioprocess data need to be converted into applicable knowledge. Therefore, it is crucial to evaluate the obtained data under the usage of transparent and reliable data reduction and correlation techniques. Within this contribution, we show a generic Monte Carlo error [...] Read more.
During process development, bioprocess data need to be converted into applicable knowledge. Therefore, it is crucial to evaluate the obtained data under the usage of transparent and reliable data reduction and correlation techniques. Within this contribution, we show a generic Monte Carlo error propagation and regression approach applied to two different, industrially relevant cultivation processes. Based on measurement uncertainties, errors for cell-specific growth, uptake, and production rates were determined across an evaluation chain, with interlinked inputs and outputs. These uncertainties were subsequently included in regression analysis to derive the covariance of the regression coefficients and the confidence bounds for prediction. The usefulness of the approach is shown within two case studies, based on the relations across biomass-specific rate control limits to guarantee high productivities in E. coli, and low lactate formation in a CHO cell fed-batch could be established. Besides the possibility to determine realistic errors on the evaluated process data, the presented approach helps to differentiate between reliable and unreliable correlations and prevents the wrong interpretations of relations based on uncertain data. Full article
(This article belongs to the Topic Bioreactors: Control, Optimization and Applications)
Show Figures

Figure 1

Figure 1
<p>Time-resolved input and rate calculation output data for an exemplary cultivation (<span class="html-italic">E. coli</span>). The rate calculation results are displayed as the black line, and the associated uncertainty, obtained by Monte Carlo resampling, is displayed as grey shading.</p>
Full article ">Figure 2
<p>Rate calculation procedure for <span class="html-italic">E. coli</span> and CHO cell cultivation datasets, presenting the inputs (in) and outputs (out) of the single calculation steps and the relative average error for the outputs obtained by a Monte Carlo error propagation. The target variables for the subsequent analysis were the specific growth rate (<math display="inline"><semantics> <mi>μ</mi> </semantics></math>), the cell-specific glycerol (<math display="inline"><semantics> <msub> <mi>q</mi> <mi>s</mi> </msub> </semantics></math>) and glutamine (<math display="inline"><semantics> <msub> <mi>q</mi> <mrow> <mi>G</mi> <mi>l</mi> <mi>n</mi> </mrow> </msub> </semantics></math>) uptake rates, the recombinant protein (<math display="inline"><semantics> <msub> <mi>q</mi> <mi>p</mi> </msub> </semantics></math>) and lactate (<math display="inline"><semantics> <mrow> <msub> <mi>q</mi> <mo>(</mo> </msub> <mrow> <mi>L</mi> <mi>a</mi> <mi>c</mi> <mo>)</mo> </mrow> </mrow> </semantics></math>) formation rates, and the cell viability. * The viable cell count was used as an additional input for the CHO cell process.</p>
Full article ">Figure 3
<p>Regression analysis between biomass-specific substrate uptake <math display="inline"><semantics> <msub> <mi>q</mi> <mi>s</mi> </msub> </semantics></math> and specific production rate <math display="inline"><semantics> <msub> <mi>q</mi> <mi>p</mi> </msub> </semantics></math> in an <span class="html-italic">E. coli</span> fed-batch process. (<b>a</b>) Regression line and (<b>b</b>) regression parameters from normal least squares (LS) and York regression (York). (<b>c</b>,<b>d</b>) One-thousand Monte Carlo regressions based on the uncertainties of the specific rates. (<b>e</b>,<b>f</b>) The obtained 68.3% parameter confidence intervals and resulting prediction confidence.</p>
Full article ">Figure 4
<p>Regression analysis between cell-specific glutamine uptake <math display="inline"><semantics> <msub> <mi>q</mi> <mrow> <mi>g</mi> <mi>l</mi> <mi>n</mi> </mrow> </msub> </semantics></math> and specific lactate formation rate <math display="inline"><semantics> <msub> <mi>q</mi> <mrow> <mi>l</mi> <mi>a</mi> <mi>c</mi> </mrow> </msub> </semantics></math> in a CHO cell fed-batch process. (<b>a</b>) Regression line and (<b>b</b>) regression parameters from normal least squares (LS) and York regression. (<b>c</b>,<b>d</b>) One-thousand Monte Carlo regressions based on the uncertainty of the specific rates. (<b>e</b>,<b>f</b>) The obtained 68.3% confidence intervals on the regression line and the regression parameters.</p>
Full article ">Figure 5
<p>Control limits for high (−0.4 g/(g h)) and low (−0.1 g/(g h)) biomass-specific substrate uptake (<math display="inline"><semantics> <msub> <mi>q</mi> <mi>s</mi> </msub> </semantics></math>) during induction with the offline-determined uptake rates and their standard deviation <math display="inline"><semantics> <msub> <mi>σ</mi> <mi>offline</mi> </msub> </semantics></math>. Resulting and predicted productivities <math display="inline"><semantics> <msub> <mi>q</mi> <mi>p</mi> </msub> </semantics></math> of the two processes with the 99% prediction confidence and standard deviation for the measured productivities.</p>
Full article ">Figure 6
<p>Control limits for cell-specific glutamine uptake (<math display="inline"><semantics> <msub> <mi>q</mi> <mrow> <mi>g</mi> <mi>l</mi> <mi>n</mi> </mrow> </msub> </semantics></math>) to avoid lactate (<math display="inline"><semantics> <msub> <mi>q</mi> <mrow> <mi>l</mi> <mi>a</mi> <mi>c</mi> </mrow> </msub> </semantics></math>) production with offline-determined specific rates with their standard deviation (<math display="inline"><semantics> <msub> <mi>σ</mi> <mi>offline</mi> </msub> </semantics></math>) of the examined CHO cell process. Deduction of control limits based on York and LS regression with their expected lactate production with 99% confidence.</p>
Full article ">Figure 7
<p>Determined CHO cell viability throughout a fed-batch experiment with a harvest threshold of 90% viability. Based on the propagated uncertainty, the probability of crossing the threshold can be determined.</p>
Full article ">Figure 8
<p>Data evaluation and predictive analysis including uncertainty and propagating it along these procedures.</p>
Full article ">
14 pages, 3502 KiB  
Article
3D Printed Microfluidic Spiral Separation Device for Continuous, Pulsation-Free and Controllable CHO Cell Retention
by Anton Enders, John-Alexander Preuss and Janina Bahnemann
Micromachines 2021, 12(9), 1060; https://doi.org/10.3390/mi12091060 - 31 Aug 2021
Cited by 14 | Viewed by 3959
Abstract
The development of continuous bioprocesses—which require cell retention systems in order to enable longer cultivation durations—is a primary focus in the field of modern process development. The flow environment of microfluidic systems enables the granular manipulation of particles (to allow for greater focusing [...] Read more.
The development of continuous bioprocesses—which require cell retention systems in order to enable longer cultivation durations—is a primary focus in the field of modern process development. The flow environment of microfluidic systems enables the granular manipulation of particles (to allow for greater focusing in specific channel regions), which in turn facilitates the development of small continuous cell separation systems. However, previously published systems did not allow for separation control. Additionally, the focusing effect of these systems requires constant, pulsation-free flow for optimal operation, which cannot be achieved using ordinary peristaltic pumps. As described in this paper, a 3D printed cell separation spiral for CHO-K1 (Chinese hamster ovary) cells was developed and evaluated optically and with cell experiments. It demonstrated a high separation efficiency of over 95% at up to 20 × 106 cells mL?1. Control over inlet and outlet flow rates allowed the operator to adjust the separation efficiency of the device while in use—thereby enabling fine control over cell concentration in the attached bioreactors. In addition, miniaturized 3D printed buffer devices were developed that can be easily attached directly to the separation unit for usage with peristaltic pumps while simultaneously almost eradicating pump pulsations. These custom pulsation dampeners were closely integrated with the separator spiral lowering the overall dead volume of the system. The entire device can be flexibly connected directly to bioreactors, allowing continuous, pulsation-free cell retention and process operation. Full article
(This article belongs to the Special Issue Bioprocess Microfluidics)
Show Figures

Figure 1

Figure 1
<p>CAD drawing of the spiral separator. (<b>A1</b>) Separator design used for cell retention testing (<b>A2</b>) side view of the design. Dolomite Microfluidic Connector (Dolomite Microfluidics, UK) on the left, slip-on luer connector for syringes on the right, sample collection port on the bottom. (<b>B</b>) Cut view of the spiral channel with a height of 200 µm and width of 600 µm. (<b>C</b>) Top view of the split at the end of the spiral channel into two 300 µm wide outlet channels.</p>
Full article ">Figure 2
<p>(<b>A</b>) Top view of the CAD drawing of the separator spiral. (<b>B</b>) Image of the fluorescent particle flow in the red region of A. Channel walls were highlighted with white dotted lines. Numbers on the right represent the number of the spiral winding. (<b>C</b>) Photo of the 3D printed spiral separator with microfluidic connector attachment points on the inlet and outlet. (<b>D</b>) Relative width of the fluorescent particle flow in relation to the channel width after each winding of the spiral at particle concentrations of 0.5, 1, 2 and 5.56 × 10<sup>6</sup> particles per mL.</p>
Full article ">Figure 3
<p>Schematic of the experimental setup to evaluate the separator performance. The inlet syringe pump is placed upright above a magnetic stirrer and a magnetic stir bar is placed inside the syringe to mix the cell solution during the experiment.</p>
Full article ">Figure 4
<p>Cell loss rate and efficiency of the cell retention device at four different cell concentrations. Blue markers: inlet flow rate of 500 µL min<sup>−1</sup>; orange markers: 700 µL min<sup>−1</sup>; grey markers: 1000 µL min<sup>−1</sup>; yellow markers: 1300 µL min<sup>−1</sup>. Error bars represent the standard deviation of three biological replications. (<b>A1</b>) Cell loss rate (× 10<sup>6</sup> cells min<sup>−1</sup>) as a function of the outer outlet flow rate (OOFR; mL min<sup>−1</sup>) at different inlet flow rates with an inlet cell concentration of 5× 10<sup>6</sup> cells mL<sup>−1</sup>. (<b>A2</b>) Separation efficiency as a function of the OOFR at different inlet flow rates with an inlet cell concentration of 5 × 10<sup>6</sup> cells mL<sup>−1</sup>. (<b>B1</b>) Cell loss rate (× 10<sup>6</sup> cells min<sup>−1</sup>) as a function of the OOFR at different inlet flow rates with an inlet cell concentration of 10 × 10<sup>6</sup> cells mL<sup>−1</sup>. (<b>B2</b>) Separation efficiency as a function of the OOFR at different inlet flow rates with an inlet cell concentration of 10 × 10<sup>6</sup> cells mL<sup>−1</sup>. (<b>C1</b>) Cell loss rate (× 10<sup>6</sup> cells min<sup>−1</sup>) as a function of the OOFR at different inlet flow rates with an inlet cell concentration of 15 × 10<sup>6</sup> cells mL<sup>−1</sup>. (<b>C2</b>) Separation efficiency as a function of the OOFR at different inlet flow rates with an inlet cell concentration of 15 × 10<sup>6</sup> cells mL<sup>−1</sup>. (<b>D1</b>) Cell loss rate (× 10<sup>6</sup> cells min<sup>−1</sup>) as a function of the OOFR at different inlet flow rates with an inlet cell concentration of 20 × 10<sup>6</sup> cells mL<sup>−1</sup>. (<b>D2</b>) Separation efficiency as a function of the OOFR at different inlet flow rates with an inlet cell concentration of 20 × 10<sup>6</sup> cells mL<sup>−1</sup>.</p>
Full article ">Figure 5
<p>(<b>A</b>) Example of an experimental setup with peristaltic pumps, buffer devices, and the cell separation device for use with a bioreactor. (<b>B</b>) Schematic side view of a buffer device, where the bottom part fills up with liquid while the trapped air works as a damper. (<b>C</b>) Side view of the smaller 1 mL buffer used behind the inlet pump with 1 mL air volume; flangeless fitting connector on the left, glue connection on the right side of the device. (<b>D</b>) Larger buffer device with a 5 mL air buffer used in front of the outlet pump; flangeless fitting connector at the right and glue connector at the left side of the device. Additional CAD views and dimensions are available in the <a href="#app1-micromachines-12-01060" class="html-app">Supplementary Information</a> (<a href="#app1-micromachines-12-01060" class="html-app">Figure S2</a>). (<b>E</b>) Top view of the 3D printed buffer devices and spiral separator fitted with PTFE tubes and flangeless fittings. The outer outlet channel (see “Old Medium” in (<b>A</b>)) is parallel to the outlet buffer device and not connected to the buffer itself. The spiral device was coated with nail polish to improve the translucency.</p>
Full article ">Figure 6
<p>(<b>A</b>) Ink width (%) in different video frames without pulsation devices. (<b>B</b>) Ink width (%) in different video frames with the outlet pulsation device attached. (<b>C</b>) Ink width (%) in different video frames with a pulsation device at the inlet and the outlet, respectively. (<b>D</b>) Ink width (%) in different video frames with syringe pumps for reference.</p>
Full article ">
18 pages, 3760 KiB  
Article
Single-Use Printed Biosensor for L-Lactate and Its Application in Bioprocess Monitoring
by Lorenz Theuer, Judit Randek, Stefan Junne, Peter Neubauer, Carl-Fredrik Mandenius and Valerio Beni
Processes 2020, 8(3), 321; https://doi.org/10.3390/pr8030321 - 9 Mar 2020
Cited by 8 | Viewed by 4805
Abstract
There is a profound need in bioprocess manufacturing for low-cost single-use sensors that allow timely monitoring of critical product and production attributes. One such opportunity is screen-printed enzyme-based electrochemical sensors, which have the potential to enable low-cost online and/or off-line monitoring of specific [...] Read more.
There is a profound need in bioprocess manufacturing for low-cost single-use sensors that allow timely monitoring of critical product and production attributes. One such opportunity is screen-printed enzyme-based electrochemical sensors, which have the potential to enable low-cost online and/or off-line monitoring of specific parameters in bioprocesses. In this study, such a single-use electrochemical biosensor for lactate monitoring is designed and evaluated. Several aspects of its fabrication and use are addressed, including enzyme immobilization, stability, shelf-life and reproducibility. Applicability of the biosensor to off-line monitoring of bioprocesses was shown by testing in two common industrial bioprocesses in which lactate is a critical quality attribute (Corynebacterium fermentation and mammalian Chinese hamster ovary (CHO) cell cultivation). The specific response to lactate of the screen-printed biosensor was characterized by amperometric measurements. The usability of the sensor at typical industrial culture conditions was favorably evaluated and benchmarked with commonly used standard methods (HPLC and enzymatic kits). The single-use biosensor allowed fast and accurate detection of lactate in prediluted culture media used in industrial practice. The design and fabrication of the biosensor could most likely be adapted to several other critical bioprocess analytes using other specific enzymes. This makes this single-use screen-printed biosensor concept a potentially interesting and versatile tool for further applications in bioprocess monitoring. Full article
(This article belongs to the Special Issue Measurement Technologies for up- and Downstream Bioprocessing)
Show Figures

Figure 1

Figure 1
<p>(<b>a</b>) An A4-sheet containing 108 screen-printed electrochemical sensors for biosensor development; (<b>b</b>) the fabricated electrochemical biosensor. The insert shows the architecture of the enzyme layer with the Pt-nanoparticle-modified carbon electrode surface covered by the enzyme/chitosan membrane.</p>
Full article ">Figure 2
<p>(<b>a</b>) Cyclic voltammetry in 50 μM H<sub>2</sub>O<sub>2</sub>, dissolved in PBS at a 25 mV/s scanning rate with a bare carbon electrode and a carbon electrode with electrodeposited Pt-NPs; (<b>b</b>) calibration curve of a L-lactate biosensor in PBS. The standard deviation was calculated using the response of ten biosensors.</p>
Full article ">Figure 3
<p>(<b>a</b>) Continuous reading of the biosensor for 24 h (the first hour inserted) in 500 µM L-lactate PBS solution; (<b>b</b>) stability of the biosensors upon storage (14 days) in dry conditions at 4 °C and at room temperature (21 °C). All readings were performed in PBS solutions containing 500 µM of L-lactate.</p>
Full article ">Figure 4
<p>Cyclic voltammetry measurements in three cell culture media using bare and Pt-NP-functionalized carbon electrodes in the absence or presence of 500 µM H<sub>2</sub>O<sub>2</sub>. (<b>a</b>) Fujifilm media diluted 1 to 60 in PBS; (<b>b</b>) RPMI medium; (<b>c</b>) CGXII medium. Voltage is cycled between −0.4 and 0.6 V at a rate of 50 mV/s with a voltage step of 10 mV.</p>
Full article ">Figure 5
<p>Monitoring of lactate off-line for seven hours of <span class="html-italic">C. glutamicum</span> cultivation in CGXII medium using the single-use biosensor, and its comparison with off-line HPLC lactate measurements. The 1× graph shows the first recordings with individual biosensors. The 3× graph shows the mean value of recordings of three individual thrice-used biosensors. The cultivation chart also shows the growth of the bacteria as dry cell weight concentrations.</p>
Full article ">Figure 6
<p>(<b>a</b>) The correlation of the different calibration methods with the HPLC measurements. The lactate measurements of the HPLC have been compared with the measurements of single and multiple use of the biosensors, two-point calibration and cross-calibration. (<b>b</b>) Five-point calibration in the industrial medium.</p>
Full article ">Figure 7
<p>Mammalian cell culture measurements at line using the single-use biosensor. (<b>a</b>) Seven-day batch cultivation showing lactate measurements with the single-use biosensor in comparison with off-line HPLC data; (<b>b</b>) Thirteen-day fed-batch cultivation showing lactate measurements with the single-use biosensor in comparison with off-line spectroscopic lactate assay data. The growth profiles of CHO cell cultures are shown in (<b>a</b>) and (<b>b</b>) from viability assays.</p>
Full article ">Figure A1
<p>Amperometric L-lactate biosensor calibration curves (at potential 0.4 V and 180 s) in PBS with different enzyme loads on the sensor.</p>
Full article ">Figure A2
<p>Preliminary comparison of different storing conditions for the biosensor. Response obtained daily in a 500 μM solution of L-lactate in PBS and normalized against response obtained on day 0.</p>
Full article ">Figure A3
<p>Example of biosensor calibration curve in CGXII medium diluted in PBS (blue line, <span class="html-italic">n</span> = 5), RPMI media (grey line, <span class="html-italic">n</span> = 5) and Fujifilm media (black line, <span class="html-italic">n</span> = 1).</p>
Full article ">
Back to TopTop