Cell Culture Process Research Articles

Enhancing real-time cell culture process monitoring through the integration of advanced machine learning techniques: A comparative analysis of Raman and capacitance spectroscopies.

Machine learning (ML) techniques have emerged as an important tool improving the capabilities of online process monitoring and control in cell culture process for biopharmaceutical manufacturing. A variety of advanced ML algorithms have been evaluated in this study for cell growth monitoring using spectroscopic tools, including Raman and capacitance spectroscopies. While viable cell density can be monitored real-time in the cell culture process, online monitoring of cell viability has not been well established. A thorough comparison between the advanced ML techniques and traditional linear regression method (e.g., Partial Least Square regression) reveals a significant improvement in accuracy with the leading ML algorithms (e.g., 31.7% with Random Forest regressor), addressing the unmet need of continuous monitoring viability in a real time fashion. Both Raman and capacitance spectroscopies have demonstrated success in viability monitoring, with Raman exhibiting superior accuracy compared to capacitance. In addition, the developed methods have shown better accuracy in a relatively higher viability range (>90%), suggesting a great potential for early fault detection during cell culture manufacturing. Further study using ML techniques for VCD monitoring also showed an increased accuracy (27.3% with Raman spectroscopy) compared to traditional linear modeling. The successful integration of ML techniques not only amplifies the potential of process monitoring but also makes possible the development of advanced process control strategies for optimized operations and maximized efficiency.

A Narrative Review on Manufacturing Methods Employed in the Production of Mesenchymal Stromal Cells for Knee Osteoarthritis Therapy.

Knee osteoarthritis (OA) is a chronic, progressive, inflammatory, and degenerative whole-joint disease. Early-stage OA treatments typically include physiotherapy, weight-loss, pain relief medications, and intra-articular knee injections, such as corticosteroids, hyaluronic acid, or platelet-rich plasma. These treatments primarily provide symptomatic relief rather than reversing or halting disease progression. Recently, mesenchymal stromal cell (MSC) injections have garnered attention due to their immunomodulatory and regenerative capacities. MSCs, which can be derived from sources such as bone marrow, umbilical cord, or adipose tissue, and can be allogeneic or autologous, have demonstrated promising results in both animal models and several human studies. However, different protocols have been employed, presenting challenges for comparing outcomes. In this review, we address these variable settings, evaluate current practices, and identify key factors critical in optimizing MSC-based therapies by critically reviewing clinical trials of ex vivo expanded MSC therapies for OA undertaken between 2008 and 2023. Specific attention was given to two key aspects: (1) the cell culture process employed in manufacturing of autologous or allogeneic MSC products, and (2) the post-culture methods employed in storage, reconstitution and administration of the MSCs. Our findings suggest that standardizing MSC production for clinical applications remains a significant challenge, primarily due to variations in tissue sources, harvesting techniques, and manufacturing protocols, and due to broad discrepancies in reporting. Thus, we propose a set of minimal reporting criteria to guide future clinical trials. A common reporting guideline is a critical step towards a more standardized MSC production across different laboratories and clinical settings, thereby enhancing reproducibility and advancing the field of regenerative medicine for knee OA, as well as other disease settings.

From Cell Clones to Recombinant Protein Product Heterogeneity in Chinese Hamster Ovary Cell Systems.

Chinese hamster ovary (CHO) cells are commonly used to produce recombinant therapeutic proteins (RTPs). The yield of RTPs in CHO cells has been greatly improved through cell editing and optimization of culture media, cell culture processes, and expression vectors. However, the heterogeneity of cell clones and product aggregation considerably affect the yield and quality of RTPs. Recently, novel technologies such as semi-targeted and site-specific transgene integration, endoplasmic reticulum-residents, and cell culture process optimization have been used to address these issues. In this review, novel developments in the field of CHO cell expression system heterogeneity are summarized. Moreover, the advantages and limitations of the new strategies are discussed, and important methods for the control of RTP quality are outlined.

Microbial lysates as low-cost serum replacements in cellular agriculture media formulation.

Microbial lysates as low-cost serum replacements in cellular agriculture media formulation.

Machine Learning-Powered Optimization of a CHO Cell Cultivation Process.

Chinese Hamster Ovary (CHO) cells are the most widely used cell lines to produce recombinant therapeutic proteins such as monoclonal antibodies (mAbs). However, the optimization of the CHO cell culture process is very complex and influenced by various factors. This study investigates the use of machine learning (ML) algorithms to optimize an established industrial CHO cell cultivation process. A ML algorithm in the form of an artificial neural network (ANN) was used and trained on datasets from historical and newly generated CHO cell cultivation runs. The algorithm was then used to find better cultivation conditions and improve cell productivity. The selected artificial intelligence (AI) tool was able to suggest optimized cultivation settings and new condition combinations, which promised both increased cell growth and increased mAb titers. After performing the validation experiments, it was shown that the ML algorithm was able to successfully optimize the cultivation process and significantly improve the antibody production. The best results showed an increase in final mAb titer up to 48%, demonstrating that the use of ML algorithms is a promising approach to optimize the productivity of bioprocesses like CHO cell cultivation processes clearly.

In-line prediction of viability and viable cell density through machine learning-based soft sensor modeling and an integrated systems approach: An industrially relevant PAT case study.

The biopharmaceutical industry is shifting toward employing digital analytical tools for improved understanding of systems biology data and production of quality products. The implementation of these technologies can streamline the manufacturing process by enabling faster responses, reducing manual measurements, and building continuous and automated capabilities. This study discusses the use of soft sensor models for prediction of viability and viable cell density (VCD) in CHO cell culture processes by using in-line optical density and permittivity sensors. A significant innovation of this study is the development of a simplified empirical model and adoption of an integrated systems approach for in-line viability prediction. The initial evaluation of this viability model demonstratedpromising accuracy with 96% of the residuals within a ±5% error limit and a Final Day mean absolute percentage error of ≤5% across various scales and process conditions. This model was integrated with a VCD prediction model utilizing Gaussian Process Regressor with Matern Kernel (nu = 0.5), selected from over a hundred advanced machine learning techniques. This VCD prediction model had an R2 of 0.92 with 89% predictions within ±10% error and significantly outperformed the commonly used partial least squares regression models. The results validated the use of these models for real-time in-line prediction of viability and VCD and highlighted the potential to substantially reduce reliance on labor-intensive discrete offline measurements. The integration of these innovative technologies aligns with regulatory guidelines and establishes a foundation for further advancements in the biomanufacturing industry, promising improved process control, efficiency, and compliance with quality standards.

Production of Recombinant Adeno-Associated Virus Through High-Cell-Density Transfection of HEK293 Cells Based on Fed-Perfusion Culture.

Adeno-associated virus (AAV)-associated gene therapy has been increasingly promising, in light of the drugs progressed to clinical trials or approved for medications internationally. Therefore, scalable and efficient production of recombinant AAV is pivotal for advancing gene therapy. Traditional methods, such as the triple-plasmid transfection of human embryonic kidney 293 cells in suspension culture, have been widely employed but often hampered by low unit yield. In this study, we optimized the cell culture process with high cell density up to 2 × 107 cells/mL by employing a perfusion culture system with centrifugation and medium exchange in shake flasks and perfusion device in bioreactor. Furthermore, we utilized a design of experiments strategy to systematically modulate a series of transfection-related variables including the quantity of plasmid DNA, the DNA-to-polyethylenimine ratio, incubation duration, and the impact of post-transfection feeding strategies on the yield of recombinant AAV (rAAV). Our comprehensive analysis and subsequent optimizations actualized a remarkable unit yield reaching nearly 2 × 1012 vector genomes (vg)/mL. Importantly, the resulting single-cell yield and biological activity were found to be comparable with those obtained from fed-batch cultures, underscoring the efficacy of our approach. Based on these findings, we investigated rAAV yield via high-density suspend culture in bioreactor, particularly focusing on cell aggregation and the use of perfusion technology. Intriguingly, we attempted to elevate the yield of an oversized recombinant coagulation factor VIII AAV843 vector by 3.5-fold, reaching a yield of 1 × 1012 vg/mL. Concurrently, the medium usage rate was only double that of batch feeding, thereby significantly shrinking the upstream cost of rAAV manufacture. In summary, this strategy significantly benefits large-scale AAV production for both commercial and clinical applications.

New deep AI‐based algorithm to assess quality of primary corneal endothelial cell cultures

Aims/Purpose: The development of cell therapies to treat corneal endothelial pathologies requires characterizing the cells during the cell culture process. The endothelial cell density (ECD) and morphometry are essential parameters for identifying cultures with sufficient phenotype and yield. However, they are difficult to measure on cells in culture using conventional image analysis tools. Aims: (1) To develop an AI‐based automatic cell segmentation method for reliable ECD and morphology measurement. (2) To select new morphometry parameters. (3) To classify endothelial cell cultures by quality.Methods: Experiments were conducted on primary cultures of human corneal endothelial cells (hCECs) after immunofluorescent labeling against NCAM revealing lateral membranes. Cultures from donors < 30 years old (high ECD and regular mosaic) were compared with cultures form older donors comprising senescent cells and cells in endothelial‐mesenchymal transition. Images were acquired with an epifluorescence microscope. We compared two AI segmentation models, a custom model and the published Cellpose model1,2. Models were trained with manually labeled images and tested on separate images. Standard image analysis methods in Python measured cell number, morphology, comparing reference and AI‐generated images3. Additional metrics were developed to qualify endothelial cell culture and correlated with known metrics. K‐Means and other classifications were tested to classify cell culture quality.Results: Cellpose outperformed the custom model in segmentation accuracy (mean relative error ± SD respectively: 0.054 ± 0.020 vs 0.0499 ± 0.0136). New reliable morphological measurements were defined, enabling classification of endothelial cell culture clusters based on quality. Coronal CV (coefficient of variation) was developed to complement classical CV used to characterize clinical endothelial cell.Conclusions: AI segmentation of hCEC in culture significantly improves quality measurement during bioengineering, facilitating rapid analysis and classification of cell batches, thereby enhancing quality assessment and success rates for cell‐based endothelial therapy.References Stringer et al (2020). Pachitariu et al (2022). Gavet et al (2012).

Proteomics Reveals Distinctive Host Cell Protein Expression Patterns in Fed-Batch and Perfusion Cell Culture Processes.

Chinese hamster ovary (CHO) cells are widely used to produce recombinant proteins, including monoclonal antibodies (mAbs), through various process modes. While fed-batch (FB) processes have been the standard, a shift toward high-density perfusion processes is being driven by increased productivity, flexible facility footprints, and lower costs. Ensuring the clearance of process-related impurities, such as host cell proteins (HCPs), is crucial in biologics manufacturing. Although purification processes remove most impurities, integrated strategies are being developed to enhance clearance of somehigh-risk HCPs. Current understanding of HCP expression dynamics in cell culture is limited. This study utilized data-independent acquisition (DIA) proteomics to compare the proteomic profiles of cell culture supernatants from 14 FB clones and three perfusion clones, all expressing the same mAb from the same host cell line. Results showed that perfusion processes enhance cell growth and productivity, exhibiting distinct proteomic profiles compared to FB processes. Perfusion processes also maintain a more comparable HCP abundance profile across clones, especially for 46 problematic HCPs monitored. Cluster analysis of FB proteomics revealed distinct abundance patterns and correlations with process parameters. Differential abundance analysis identified significant protein differences between the two processes. This is the first extensive study characterizing HCPs expressed by clones under different process modes. Further research could lead to strategies for preventing or managing problematic HCPs in biologics manufacturing.

A Novel, Site-Specific N-Linked Glycosylation Model Provides Mechanistic Insights Into the Process-Condition Dependent Distinct Fab and Fc Glycosylation of an IgG1 Monoclonal Antibody Produced by CHO VRC01 Cells.

The CHO VRC01 cell line produces an anti-HIV IgG1 monoclonal antibody containing N-linked glycans on both the Fab (variable) and Fc (constant) regions. Site-specific glycan analysis was used to measure the complex effects of cell culture process conditions on Fab and Fc glycosylation. Experimental data revealed major differences in glycan fractions across the two sites. Bioreactor pH was found to influence fucosylation, galactosylation, and sialylation in the Fab region and galactosylation in the Fc region. To understand the complex effects of process conditions on site-specific N-linked glycosylation, a kinetic model of site-specific N-linked glycosylation was developed. The model parameters provided mechanistic insights into the differences in glycan fractions observed in the Fc and Fab regions. Enzyme activities calculated from the model provided insights into the effect of bioreactor pH on site-specific N-linked glycosylation. Model predictions were experimentally tested by measuring glycosyltransferase-enzyme mRNA-levels and intracellular nucleotide sugar concentrations. The model was used to demonstrate the effect of increasing galactosyltransferase activity on site-specific N-linked glycan fractions. Experiments involving galactose and MnCl2 supplementation were used to test model predictions. The model is capable of providing insights into experimentally measured data and also of making predictions that can be used to design media supplementation strategies.

Analysis of cell lysis for improved understanding between the shake tube and stirred tank reactor perfusion CHO cell cultures.

Shake tubes (ST) are widely employed to assist the development of the stirred tank reactor (STR) perfusion cell culture. However, cell lysis may be frequently underrestimated and lead to culture performance discrepency between these systems, rendering the ST model ineffective in designing the STR perfusion cultures. In this study, perfusion culture performance bewteen the STR and ST was investigated under various conditions with the analysis of cell lysis. Comparable performance was observed bewteen the two systems at low perfusion rates ( ≤1.0 VVD), except that the specific productivity ( ) of the STR was decreased at =0.5 VVD, which was related to product degradation by cell lysis. In contrast, significant differences in cell maintenance, metabolism, and were found at =2.0 VVD. By the analysis of the authentic cell growth and death kinetics, it was found that cell growth arrest, potentially due to the limited availability of oxygen, led to the stable cell maintenance at VCD≈90 × 106 cells/ml and altered cellular metabolism for the ST, while the continuous decline of VCD and in the STR were related to excessive cell death, subsequently ascribed to the harmful hydrodynamic stress conditions. We further demonstrated that cell lysis accounted for 57.62-76.29% of the total generated biomass in both the reactors and significantly impacted the estimation of process descriptors crucial for understanding the true cellular states. With cell lysis in sight, cell performance can therefore be accurately described and this knowledge can be further leveraged to expedite process development for the perfusion cell culture processes.

Beyond viability: Advancing CHO cell culture process strategies to modulate host cell protein levels

Beyond viability: Advancing CHO cell culture process strategies to modulate host cell protein levels

Procollagen-lysine 2-oxoglutarate 5-dioxygenases are responsible for 5R-hydroxylysine modification of therapeutic T-cell bispecific monoclonal antibodies produced by Chinese hamster ovary cells.

We present a detailed mass spectrometric analysis of three 2 + 1 T-cell bispecific monoclonal antibodies (TCB mAbs), where an unexpected +15.9950Da mass shift in tryptic peptides was observed. This modification was attributed to the occurrence of 5R-hydroxylysine (Hyl) using a hybrid LC-MS/MS molecular characterization and CRISPR/Cas9 gene deletion approach. The modification was found at various sites within TCB mAbs, with a conspicuous hot spot motif mirroring a prior observation where Hyl was mapped to the CH1-VH Fab domain interface of IgGs. In contrast to the preceding report, our structural modeling analysis on TCB mAbs unveiled substantial differences in the orientation and flexibility of motifs in immediate proximity and across the artificial CH1-VL cross Fab interface and upstream elbow segment. Utilizing a hybrid database search, RNAseq, and a CRISPR/Cas9 knockout methodology in Chinese hamster ovary (CHO) production cell lines, procollagen-lysine, 2-oxoglutarate 5-dioxygenases (PLODs) were conclusively identified as the catalyzing enzymes accountable for the 5R-Hyl modification in TCB mAbs. To quantitatively inhibit Hyl formation in TCB mAbs, the activity of all three Chinese hamster PLOD isoenzymes needs to be depleted via CRISPR/Cas9 gene knockout. Moreover, our investigation identified cell culture iron availability, process duration, and clonal variability in CHO cells as elements influencing the levels of Hyl formation in TCB mAbs. This research offers a solution for circumventing Hyl formation in therapeutic complex mAb formats, such as TCB mAbs, produced in CHO cell culture processes, thereby addressing potential technical and biological challenges associated with unintended Hyl modification.

Magnetically Driven Hydrogel Surfaces for Modulating Macrophage Behavior.

During the host response toward implanted biomaterials, macrophages can shift phenotypes rapidly upon changes in their microenvironment within the host tissue. Exploration of this phenomenon can benefit significantly from the development of adequate tools. Creating cell microenvironment alterations on classical hydrogel substrates presents challenges, particularly when integrating them with cell cultivation and monitoring processes. However, having the capability to dynamically manipulate the cell microenvironment on biomaterial surfaces holds significant potential. We introduce magnetically actuated hydrogels (MadSurface) tailored to induce reversible stiffness changes on polyacrylamide hydrogel substrates with embedded magnetic microparticles in a time-controllable manner. Our investigation focused on exploring the potential of magnetic fields and MadSurfaces in dynamically modulating macrophage behavior in a programmable manner. We achieved a consistent modulation by subjecting the MadSurface to a pulsed magnetic field with a frequency of 0.1 Hz and a magnetic field flux density of 50 mT and analyzed exposed cells using flow cytometry and ELISA. At the single-cell level, we identified a subpopulation for which the dynamic stiffness conditions in conjunction with the pulsed magnetic field increased the expression of CD206 in M1-activated THP-1 cells, indicating a consistent shift toward the M2 anti-inflammatory phenotype on MadSurface. At the population level, this effect was mostly hindered in the culture period utilized in this work. The MadSurface approach advances our understanding of the interplay between magnetic field, cell microenvironment alterations, and macrophage behavior.

Bio-printing method as a novel approach to obtain a fibrin scaffold settled by limbal epithelial cells for corneal regeneration

Treatment of Limbal Stem Cell Deficiency (LSCD), based on autologous transplantation of the patient’s stem cells, is one of the few medical stem cell therapies approved by the European Medicines Agency (EMA). It relies on isolating and culturing in vivo Limbal Epithelial Stem Cells (LESC) and then populating them on the fibrin substrate, creating a scaffold for corneal epithelial regeneration. Such a solution is then implanted into the patient’s eye. The epithelial cell culture process is specific, and its results strongly depend on the initial cell seeding density. Achieving control of the density and repeatability of the process is a desirable aim and can contribute to the success of the therapy. The study aimed to test bioprinting as a potential technique to increase the control over LESCs seeding on a scaffold and improve process reproducibility. Cells were applied to 0.5 mm thick, flat, transparent fibrin substrates using extrusion bioprinting; the control was the traditional manual application of cells using a pipette. The use of 3D printer enabled uniform coverage of the scaffold surface, and LESCs density in printed lines was close to the targeted value. Moreover, printed cells had higher cell viability than those seeded traditionally (91.1 ± 8.2% vs 82.6 ± 12.8%). The growth rate of the epithelium was higher in bioprinted samples. In both methods, the epithelium had favorable phenotypic features (p63 + and CK14 +). 3D printing constitutes a promising approach in LSCD therapy. It provides favorable conditions for LESCs growth and process reproducibility. Its application may lead to reduced cell requirements, thereby to using fewer cells on lower passages, which will contribute to preserving LESCs proliferative potential.

Comprehensive stable-isotope tracing of glucose and amino acids identifies metabolic by-products and their sources in CHO cell culture

Mammalian cell culture processes are widely utilized for biotherapeutics production, disease diagnostics, and biosensors, and hence, should be optimized to support robust cell growth and viability. However, toxic by-products accumulate in cultures due to inefficiencies in metabolic activities and nutrient utilization. In this study, we applied comprehensive 13C stable-isotope tracing of amino acids and glucose to two Immunoglobulin G (IgG) producing Chinese Hamster Ovary (CHO) cell lines to identify secreted by-products and trace their origins. CHO cells were cultured in media formulations missing a single amino acid or glucose supplemented with a 13C-tracer of the missing substrate, followed by gas chromatography-mass spectrometry (GC-MS) analysis to track labeled carbon flows and identify by-products. We tracked the sources of all secreted by-products and verified the identity of 45 by-products, majority of which were derived from glucose, leucine, isoleucine, valine, tyrosine, tryptophan, methionine, and phenylalanine. In addition to by-products identified previously, we identified several metabolites including 2-hydroxyisovaleric acid, 2-aminobutyric acid, L-alloisoleucine, ketoisoleucine, 2-hydroxy-3-methylvaleric acid, desmeninol, and 2-aminobutyric acid. When added to CHO cell cultures at different concentrations, certain metabolites inhibited cell growth while others including 2-hydroxy acids, surprisingly, reduced lactate accumulation. In vitro enzymatic analysis indicated that 2-hydroxy acids were metabolized by lactate dehydrogenase suggesting a possible mechanism for lowered lactate accumulation, e.g., competitive substrate inhibition. The 13C-labeling assisted metabolomics pipeline developed and the metabolites identified will serve as a springboard to reduce undesirable by-products accumulation and alleviate inefficient substrate utilization in mammalian cultures used for biomanufacturing and other applications through altered media formulations and pathway engineering strategies.

Improving Raman-Based Models for Real-Time Monitoring the CHO Cell Culture Process with Effective Variable Selection Strategies

Research has shown that Raman spectroscopy could be applied to monitor various components in mammalian cell culture in real time. In the process of application, it is necessary to ensure the performance of the Raman-based model. The variable selection strategy is an effective method that significantly influences the model performance and simplification. In this study, different variable selection strategies were evaluated, and the optimal variable selection strategy was determined for monitoring the CHO cell culture process. Firstly, a wide variety of spectral regions involving the Raman fingerprinting region and the C-H stretching region were investigated. Secondly, six different variable selection algorithms were meticulously assessed. Thirdly, the combination of different variable selection algorithms was used to improve model performance and simplify the model. Finally, the monitoring of cell culture processes was implemented. The findings underscored that commonly used spectral regions could improve the model performance but could not simplify the model well. Moving-window partial least square (MWPLS), genetic algorithm (GA), and random frog (RF) are more suitable for Raman modeling of the cell culture process, but they must be used after the spectral region selection. The combination of three variable selection algorithms (MWPLS-GA-RF) improved the model’s performance by 16–70% by selecting 30–60 variables, effectively simplifying the model. For glucose, lactate, viable cell density, and ammonium ion, real-time monitoring was performed well. This study will be helpful for researchers to select suitable variable selection strategies for building models for the real-time monitoring of cell culture.

Hybrid modeling for in silico optimization of a dynamic perfusion cell culture process.

The bio-pharmaceutical industry heavily relies on mammalian cells for the production of bio-therapeutic proteins. The complexity of implementing and high cost-of-goods of these processes are currently limiting more widespread patient access. This is driving efforts to enhance cell culture productivity and cost reduction. Upstream process intensification (PI), using perfusion approaches in the seed train and/or the main bioreactor, has shown substantial promise to enhance productivity. However, developing optimal process conditions for perfusion-based processes remain challenging due to resource and time constraints. Model-based optimization offers a solution by systematically screening process parameters like temperature, pH, and culture media to find the optimum conditions in silico. To our knowledge, this is the first experimentally validated model to explain the perfusion dynamics under different operating conditions and scales for process optimization. The hybrid model accurately describes Chinese hamster ovary (CHO) cell culture growth dynamics and a neural network model explains the production of mAb, allowing for optimization of media exchange rates. Results from six perfusion runs in Ambr® 250 demonstrated high accuracy, confirming the model's utility. Further, the implementation of dynamic media exchange rate schedule determined through model-based optimization resulted in 50% increase in volumetric productivity. Additionally, two 5 L-scale experiments validated the model's reliable extrapolation capabilities to large bioreactors. This approach could reduce the number of wet lab experiments needed for culture process optimization, offering a promising avenue for improving productivity, cost-of-goods in bio-pharmaceutical manufacturing, in turn improving patient access to pivotal medicine.

Enhancing cross‐scale Raman in‐line monitoring capability of cell culture process in large‐scale manufacturing

AbstractThis study introduces an approach to enhance the Raman calibration model cross‐scale prediction capabilities in cell cultures. Our investigation centers on the improvement of Raman calibration models along with the scaling‐up of mammalian cell culture processes. Initially, we observed that integrating data from a 50 L run into the original dataset at lab‐scale was an effective strategy for improving prediction accuracy on 1000 L run. Further investigation indicated that this improvement could be attributed to the similarity between the spectra of 50 L with 1000 L. Subsequent investigation unveiled that the spectral variations primarily stemmed from differences between the instruments. This was verified through a direct comparison of two Raman analyzers. Then two pairwise spectral transformation methods, direct standardization (DS) and piecewise DS, both exhibited superior performance in predicting. This research underscores the importance of Raman spectral transfer method to alleviate instrument‐to‐instrument variations in cross‐scale monitoring of cell culture processes.

A novel approach for perfusion process design based on a "Grey-Box" kinetic model.

Perfusion cell-culture mode has caught industrial interest in the field of biomanufacturing in recent years. Thanks to new technology, perfusion-culture processes can support higher cell densities, higher productivities and longer process times. However, due to the inherent operational complexity and high running costs, the development and design of perfusion-culture processes remain challenging. Here, we present a model-based approach to design optimized perfusion cultures of Chinese Hamster Ovary cells. Initially, four batches of bench-top reactor continuous-perfusion-culture data were used to fit the model parameters. Then, we proposed the model-based process design approach, aiming to quickly find out the "theoretically optimal" operational parameters combinations (perfusion rate and the proportion of feed medium in perfusion medium) which could achieve the target steady-state VCD while minimizing both medium cost and perfusion rate during steady state. Meanwhile, we proposed a model-based dynamic operational parameters-adjustment strategy to address the issue of cell-growth inhibition due to the high osmolality of concentrated perfusion medium. In addition, we employed a dynamic feedback control method to aid this strategy in preventing potential nutrient depletion scenarios. Finally, we test the feasibility of the model-based process design approach in both shake flask semi-perfusion culture (targeted at 5 × 107 cells/ml) and bench-top reactor continuous perfusion culture (targeted at 1.1 × 108 cells/ml). This approach significantly reduces the number of experiments needed for process design and development, thereby accelerating the advancement of perfusion-mode cell-culture processes.