CHO cell lines are the most important mammalian expression system for producing protein-based biopharmaceuticals. Over the last decades, continuous advancements in cell line development strategies—focusing on expression system optimization, clone selection, and media and process innovations—have significantly improved product yield and quality. To keep pace with the rapid development of new and more complex therapeutics, genetic engineering of cell lines to enhance growth, productivity and optimize product quality attributes is a promising approach. In this study, we developed a genetically engineered CHO cell line as a new innovative host, aiming to create stable cell lines with further enhanced titer and productivity for biopharmaceutical manufacturing. The newly engineered host was employed alongside the wild-type (WT) cell line in various cell line development campaigns to generate production clones expressing DTE, IgG, and Fc-fusion antibodies. Sartorius' CHO CLD technology, optimized for accelerated generation and effective identification of high-producing clones, was utilized. Additional evaluations under fed-batch conditions in different Ambr® system scales using 4Cell® SmartCHO media and feeds demonstrated significantly enhanced productivity across all tested molecules, marking this new engineered cell line as a promising new host for Sartorius’ cell line development platform.

The growing number of candidate therapeutic biologics currently under development has increased the demand for innovative solutions from Process Development operations to deliver these medicines. A substantial number of biologics are manufactured using live cell systems, where mammalian Chinese Hamster Ovary (CHO) cells are predominantly used for production of recombinant protein therapeutics. To meet regulatory requirements and ensure a safe and robust process, the manufacturing cell lines must be derived from a single cell origin. Typically, Cell Line Development single cell cloning and screening campaigns are considerably challenging and resource intensive as hundreds to thousands of clones need to be isolated and evaluated to identify high quality candidates suitable for clinical and commercial manufacturing. At Amgen, we implemented the Beacon Optofluidic system for single cell cloning, which is a fully integrated nanofluidic cell culture platform that allows isolation of up to 1758 clonal cell lines on a single nanofluidic chip. Commercially available Spotlight® assay reagents, implemented in Beacon CLD workflows, rely on fluorescently conjugated anti-fragment crystallizable (Fc) or anti-Light chain (LC) probes to quantify the steady-state secreted protein level for each individual clone. However, there is a substantial amount of recombinant protein therapeutics that may not be compatible with commercially available Spotlight® reagents due to an incompatible molecule type or protein sequence engineering. We introduced differentiating clone selection strategies, including the development of novel assays to quantify protein secretion for engineered therapeutic modalities. Furthermore, we optimized the cell export workflow through media additives and export process parameters to eliminate cell adhesion of difficult to manipulate clones and achieve 99% clonality assurance with sticky cell lines. These advancements enhance confidence in single cell derivation and clone selection process at the point of Single Cell Cloning to ensure high quality of manufacturing cell lines.

The production of life-saving biotherapeutics depends heavily on Chinese hamster ovary (CHO) cell culture systems, which require precise control of essential amino acids (EAAs) supplementation. Improper EAA levels—whether in excess or deficiency—can lead to cellular stress and death, complicating manufacturing processes. Our research addresses this fundamental limitation by engineering CHO cells capable of synthesizing their own essential amino acids through the integration of heterologous synthetic biosynthetic pathways. We have successfully achieved random genomic integration of two essential amino acid biosynthetic pathways, followed by rigorous selection in amino acid dropout media. Expression of the transfected genes has been confirmed through qPCR, with pathways integrated at ~1 copy per cell. Gene expression varied due to different promoter/polyA combinations, ranging between 0.2 – 10 % of Gapdh expression. These engineered cells demonstrate the ability to grow in media lacking the matching essential amino acid, validating functional restoration of biosynthetic capacity. We will also present preliminary assessment of the transcriptional response to restoring these evolutionarily ancient, foreign pathways, providing insights into the regulatory networks activated during this metabolic reprogramming. Our approach represents a significant advancement toward creating more sustainable and robust bioproduction systems. By reducing dependency on precise external amino acid supplementation, these engineered CHO cell lines have the potential to enhance manufacturing flexibility and process control. Ongoing work focuses on pathway optimization, characterization of cellular adaptations through multi-omics approaches, and evaluation of these pathways as novel selection systems. Additionally, we will assess how these prototrophic hosts impact product titer using standard manufacturing workflows and evaluate product quality attributes such as glycosylation.

Monoclonal antibodies (mAbs) dominate the growing biopharmaceutical market, but low α-2,6-sialylation in CHO-produced mAbs limits their therapeutic potential. SialMAX is our glycoengineering platform designed to boost α-2,6-sialylation by knocking out the ST3GAL4 gene using CRISPR/Cas9. To streamline development, we created a lectin-based confocal microscopy method that visualizes cell surface glycosylation in real time. This non-invasive tool supports early clone screening and prioritization, reducing time and resources before purification. Validated in multiple engineered lines, this integrated approach accelerates cell line development and enables more consistent, effective mAbs. SialMAX shows how combining precision editing with real-time phenotyping can shape the future of biomanufacturing.

Inducible transcriptional circuits are useful tools to fine-tune gene expression and enable real-time control over desirable cellular phenotypes. In contrast to conventional inducible systems that function in an on/off manner, Linearizer circuits enable gradual gene expression in response to increasing inducer molecule concentrations, somewhat akin to dimmer light switches. Linearizer circuits have extensive potential applications across biopharmaceutical cell culture processes, for example: (i) controlled and coordinated expression of peptide chains that form multimeric therapeutic proteins to enhance the yield of bi- and multi-specifics, (ii) inducible and tuneable expression of adeno-associated viral vector (AAV) components to maximise gene therapy vector yields, and (iii) real-time glycosylation control to enable robust quality assurance of therapeutic glycoproteins. A key challenge in deploying Linearizer circuits is that the stoichiometry of their components (repressor operon sites, repressor protein, and gene of interest) must be ensured for optimal performance (minimal basal GOI expression, fold induction, and linear dose response to inducer concentrations). This study evaluates the performance of two negatively autoregulated Linearizer circuits (TetR_Lin and PhlF_Lin) in Chinese hamster ovary (CHO) cells using two different genomic integration strategies, targeted integration (TI) and random integration (RI). Each circuit consists of a fluorescent protein reporter (eGFP or mCherry) and a repressor gene (TetR or PhlF). For RI, Linearizers were deployed using transfection and antibiotic resistance selection. For TI, orthogonal landing pads with the Cre/lox and Bxb1/att recombinase mediated cassette exchange systems were used. Our results show that RI results in high basal expression, low fold induction, and a nonlinear response across inducer concentrations. In contrast, TI achieves lower basal expression, a strong linear response to inducer molecule concentration, and a broader induction range. Our results suggest that the suboptimal performance observed with RI is likely due to stoichiometric imbalances of Linearizer circuit components, and that the issues are resolved with targeted integration. By optimising Linearizer circuit performance in CHO cells, this project paves the way towards real-time control and optimisation – at the cellular level – of biopharmaceutical cell culture processes.

Chinese hamster ovary (CHO) cell lines are the predominant host for biopharmaceutical production and are continuously engineered to enhance productivity and product quality. A novel approach for eukaryotic systems, though previously applied in prokaryotes, is the minimal genome strategy, which seeks to reduce genome size by removing non-essential genomic regions. Our group previously demonstrated genomic deletions of up to ~1 megabase (Mb), but given the CHO genome size of 2.45 gigabases (Gb), larger deletions are needed to achieve substantial genome reduction. This study therefore aimed to explore the upper size limits of genomic deletions in CHO cells. CRISPR/Cas9-mediated deletions were performed by transfecting Cas9 protein and single guide RNAs (sgRNAs) flanking genomic target regions selected based on gene expression and essentiality. Genomic regions of up to 13 Mb with low gene expression were identified and deletions thereof performed with the aim of complete genomic excision. Even larger regions of up to 130 Mb including highly expressed and essential genes were addressed via monoallelic deletions to test the maximum feasible deletion size. Following single cell sorting, individual clones exhibiting deletions of the target sequences were identified and validated through PCR and sequencing. All deletions were verified, with confirmation at clonal level for up to 50 Mb and confirmation at cell pool level for up to 130 Mb deletions. Flow cytometry-based ploidy analysis revealed an increased DNA content in some clones, indicating polyploidy and emphasizing the need for comprehensive clone characterization post editing. Bioprocess compatibility was finally evaluated through batch cultivation and assessment of IgG productivity, cell growth, and viability. Interestingly, partially improved productivity and harvest viability was observed with no significant negative impacts attributed to genome reduction. This study demonstrates targeted CRISPR/Cas9-mediated genomic deletions of unprecedented size in CHO cells, thereby establishing a basis for creating a significantly genome-reduced CHO cell line with potentially enhanced recombinant protein production capabilities and reduced cellular complexity.

Monoclonal antibodies (mAbs) represent the most commercially successful category of biopharmaceuticals. However, the high production costs associated with mAb synthesis in CHO cell cultures limit accessibility for many patients. Optimizing protein productivity in biopharmaceutical manufacturing is heavily dependent on advancements in cell culture media, which play a critical role in the production process. This study aims to model how variations in culture media influence protein production to facilitate medium optimization for improved biopharmaceutical yield. Since most commercially available media consist of proprietary formulations with undisclosed compositions, industrial optimization practices typically involve blending different commercial media in varying proportions to enhance cell culture performance. While statistical methods can assist in identifying optimal conditions, they often necessitate extensive experimental trials. To reduce the cost and experimental burden of media optimization, this study proposes a novel hybrid modeling approach integrating dynamic flux balance analysis (dFBA) with partial least squares (PLS)-based empirical modeling. Using this hybrid framework, mAb production is optimized by determining the optimal proportions of commercial media in the final mixture. The resulting regression models are then incorporated as kinetic constraints within the dFBA framework with specific tolerances. The optimization problem is subsequently formulated to maximize mAb production, with decision variables representing media components and constraints defined by the PLS-dFBA hybrid model. Optimization results are experimentally validated, generating new data that iteratively refine the model, facilitating a continuous search for improved conditions. This systematic approach harnesses the complementary strengths of data-driven PLS modeling and dFBA metabolic pathway optimization, ultimately contributing to the development of enhanced cell culture media for optimized mAb production.

Protein subunit vaccines have a strong track record of efficacy and safety and have been widely applied for prevention of a variety of infectious diseases. The impacts of post-translational modifications of vaccine antigens are often overlooked, despite the fact that they can vary significantly depending on the expression hosts (e.g., bacteria, yeast, plant, insect or mammalian cells) and the culture conditions used for their manufacturing. Using SARS-CoV-2 spike trimers as model antigens, we sought to evaluate the immunological impact of modulating their state of glycosylation. Spike proteins rich in complex-type (CT), high-mannose (HM) or paucimannose (PM) N-linked glycans were produced using Chinese Hamster Ovary (CHO) cells (cultured with or without the mannosidase inhibitor kifunensine) or insect cells. We found that when these antigens are adjuvanted with liposomes composed of sulfated lactosyl archaeol (SLA), all glycoforms are highly immunogenic and induce abundant spike-specific serum IgG and IFN-γ producing T-cells in a mouse model. The spike antigen with CT glycans induces a significantly more potent neutralizing immune response, which directly correlates to more abundant receptor binding domain (RBD)-specific IgG when comparing to the antigen with HM glycans. This observation remains true whether the spike is resistin- or T4 foldon-trimerized, indicating that the glycosylation effect is not trimerization domain-specific. Spike with PM glycans induces remarkably low titers of neutralizing antibodies and RBD-specific IgG. Our results highlight the significant impacts of a vaccine antigen's glycosylation profile in directing the immune response, which should be an important consideration for design of protein-based vaccines.

H5N1, a highly pathogenic avian influenza subtype, has caused severe poultry outbreaks and sporadic human transmission with mortality rates over 50%. As the virus evolves, the need for effective preparedness becomes urgent. Immunological tools play a central role, offering insights into disease mechanisms, advancing vaccines, enhancing diagnostics, and guiding therapeutics. This poster highlights key innovations to build a comprehensive toolbox for H5N1 pandemic preparedness. Highlight the key immunological tools that form a comprehensive toolbox for H5N1 pandemic preparedness. This poster emphasizes the role of these tools in advancing vaccine development, monitoring viral evolution, analyzing immune responses, and guiding therapeutic strategies to enhance global health security and pandemic resilience.

When generating CHO based production clones, a significant amount of time is spent selecting clones and creating an optimal bioprocess. This is due to high clone to clone variation. Our previous research has indicated that the clonal variation can be nearly entirely removed simply by employing a recombinase-based targeted gene integration strategy rather than using methods relying on double stranded breaks. In this latest work, we have compared the results of a fed-batch bioprocess optimization conducted on several clones either created using RMCE or random integration. The results show that the most important factors for RMCE clones are process parameters such as e.g. pH or choice of feed. The most important factor for clones generated with random insertion is which clone you selected. In fact, the most optimal bioprocess for the RMCE clones was optimal for all clones, despite the clones producing different products, indicating that one bioprocess optimization is all you need.

The Thomson 6-well plate system provides a mid-scale culture format with individual well working volumes of 20-50mL, serving as an alternative to conical bioreactor tubes. The plate's integrated lid includes splash guards and helps reduce evaporation during culture periods. Each well's 50mL maximum working volume allows for larger culture volumes, which can be particularly useful when working with low-expressing proteins where additional culture volume is needed for protein accumulation. The multi-well design enables processing multiple samples within a single unit, increasing throughput compared to individual tubes.

Cell lines expressing specific membrane proteins are essential tools in research and drug development. Phage display, a key technology in this field, enables the identification of highly specific monoclonal antibodies. Whole-cell panning in phage display further advances this process by using membrane protein-expressing cell lines to present target antigens in their native context. These cell lines are also instrumental in screening hit compounds to identify antigen binders and performing bioassays to evaluate drug potency and efficacy. Generating stable cell lines with diverse type I, II, and III membrane proteins across various cell types, from different organisms, and tissues often necessitates customized strategies. The key challenge lies in developing efficient and streamlined strategies for stable cell line generation. Here, we evaluated the DirectedLuck® transposase from ProBioGen AG for research cell line production. Our findings highlight its ability to accelerate the selection of expressing pools with high expression levels, and maintain stability of target genes, offering a promising platform for optimizing cell line development.

Expi Protein Expression Systems have long powered your protein expression applications. For therapeutic antibody discovery, which necessitates synthesizing thousands of antibodies, high-throughput antibody production facilitates your antibody library screening. Introducing GeneArt Thermo Fisher Scientific Small-Scale HTP Antibody Production Service– a robust and efficient platform for the rapid generation mAbs. It combines the precision of GeneArt services with the robust capabilities of ExpiCHO and Expi293 to deliver consistent results and a reliable product, helping you accelerate your drug discovery efforts

Escherichia coli is a widely used host for recombinant protein production due to its fast growth, well-characterized genetics, and cost-effectiveness. However, heterologous expression often leads to inclusion body formation, requiring solubilization and protein-specific refolding protocols. Here, we describe a streamlined strategy for recovering diagnostically relevant arboviral antigens from inclusion bodies. Three arboviral proteins were expressed in E. coli Rosetta(DE3): non-structural protein 1 from dengue virus (DENVNS1), envelope domain III from dengue virus (DENVE3), and non-structural protein 1 from Zika virus (ZIKVNS1). The recombinant proteins predominantly accumulated in inclusion bodies, prompting a systematic evaluation of solubilization (8 M urea) and renaturation conditions to obtain correctly folded and immunoreactive proteins suitable for diagnostic applications. After expression and purification under denaturing conditions, proteins were refolded using a redox-controlled buffer system (reduced/oxidized glutathione) and gradual urea removal via G-25 size-exclusion chromatography. This dilution-based approach offered a simplified and reproducible strategy for oxidative protein folding. Refolded protein yields reached 8 mg/L (DENVNS1), 20 mg/L (DENVE3), and 5 mg/L (ZIKVNS1). Dynamic light scattering revealed that preparations contained high-molecular-weight aggregates and a smaller population of lower-order species, suggesting partial recovery of monomeric forms. These findings support the use of this refolding protocol as a practical approach for producing immunoreactive arboviral antigens suitable for diagnostic assay development.

The successful development of biotherapeutic molecules begins with the generation of a robust cell line. Choosing the right Cell Line Development (CLD) path is essential and should be focused on both comprehensive analysis of protein-to-be-expressed attributes and specific programs needs assessment. We have developed a robust transposase-based expression platform that overcomes these challenges and deliver consistent high-expressing cell lines up to 12g/L with sustained stability over 60 generations. To reach this goal, we have leveraged a proprietary CHO-M mammalian cell to design comprehensive CLD workflows without compromising the successful inherited assets from our former random expression platform. In this “trust our experts” approach, we offer flexible solutions and risk mitigation strategies to ensure the best cell line development path for any therapeutic protein format.

Lectins are carbohydrate-binding proteins involved in crucial biological processes like cell-cell recognition, immune responses, and host-pathogen interactions [1,2]. Their high specificity for sugar moieties enables them to recognize glycosylated proteins on host cells or pathogens, making them valuable tools for studying glycan-mediated interactions and developing targeted therapies [3]. Among the various lectin families, C-type lectins, are of particular interest due to their calcium-dependent binding mechanism and prominent roles in immune surveillance and pathogen recognition [4]. To enhance their utility, lectibodies, chimeric molecules combining lectin domains with antibody Fc regions, have been developed [5]. These fusions improve stability and avidity and convey effector functions, enabling targeted applications in glycan-rich structures such as tumors and pathogen surfaces [5,6]. Here, we describe the construction and characterization of an off-the-shelf human lectibody collection for subsequent assessment of glycan recognition and applicability against infectious agents. Over 100 annotated lectin domain sequences from the human genome were cloned in the lectibody format and are expressed via transient gene expression in mammalian cells. Proteins are purified using bio-affinity followed by size exclusion, yielding high-quality samples. Protein quality is judged based on purity and biophysical properties, while glycan-binding specificity is evaluated using a wide variety of carbohydrate structures. Finally, we outline future directions for functional assays, protein engineering, and systematic screening of lectibodies as potential anti-infective candidates. 1.Sharon N, Lis H., Glycobiology. 2004 Nov;14(11):53R–62R. 2.Varki A., Glycobiology. 2017 Jan;27(1):3–49. 3.Van Kooyk Y, Rabinovich GA., Nat Immunol. 2008 Jun;9(6):593–601. 4.Zelensky AN, Gready JE., FEBS Journal. 2005 Dec;272(24):6179–6217. 5.Hoffmann D, Mereiter S, et al., EMBO J. 2021 Aug;40: e108375 6.Oh YJ, et al., Mol Ther. 2022 Apr 6;30(4):1523-1535.

Effective T-cell activation is driven by three signals: Signal 1 delivered through antigen recognition via the T-cell receptor, signal 2 through co-stimulatory receptors, and signal 3 mediated by cytokines. CD3-targeting T-cell engagers provide signal 1 to T cells and have shown significant clinical benefit in hematological malignancies but have faced challenges in solid tumors due to on target off tumor toxicities. Emerging clinical data supports the hypothesis that for efficient and sustained activity in the presence of signal 1, engagement of co-stimulatory molecules like CD28 could be important for effective anti-tumor activity in solid tumors. We present the development of a panel of bispecific antibodies (bsAbs), targeting the costimulatory molecule CD28 and the tumor associated antigen (TAA) Nectin-4, a cell adhesion molecule overexpressed in bladder cancer and other malignancies. To select our development candidate RNDO-564, we screened a panel of bsAbs with varying CD28-potencies for both functional activity and biophysical stability to assess and reduce manufacturability risks. Functional characterization of our bsAbs showed robust tumor cytotoxicity and IL-2 secretion while posing a lower safety risk providing that the activity is dependent on presence of signal 1 and expression of the TAA. Additionally, we demonstrated that the panel of CD28 binders can be paired with various TAA-binding arms, enabling the design of new bispecific antibodies for different indications. In addition to the desired biological activity, the CD28 x Nectin-4 bsAbs also exhibited favorable developability profiles as demonstrated by subjecting the panel to accelerated stress conditions like thermal stress at elevated temperatures for extended periods of time or low pH hold. This extensive functional and biophysical characterization enabled the selection and successful manufacture of our clinical candidate RNDO-564.

The rise of H5N1, a highly pathogenic avian flu virus capable of infecting mammals, and the continuous dissemination of SARS-CoV-2 variants highlight the need for effective viral therapies. Current antiviral treatments and vaccines often fail due to viruses high mutation rate, leading to immune evasion. Camelidae produce functional immunoglobulin G molecules lacking a light chain and CH1 domain, with a variable region consisting of a single domain called VHH. Although they are much smaller than full-length antibodies (15 vs. 150 kDa), VHHs can be expressed as single domain antibodies while keeping similar binding specificity and affinity. Their small size allows them to access and recognize cryptic antigenic sites that are inaccessible to human antibodies, have high solubility and stability, and can be administered in aerosolized form. VHHs are also modular and can be fused together to generate bispecific or multispecific antibodies that typically exhibit increased avidity compared to individual monomers. The fusion of two different VHHs targeting non-overlapping epitopes on the same antigen results in biparatopic antibodies that may render them less sensitive to viral mutations. We produced libraries of VHHs against SARS-CoV-2 spike and avian flu H5 and N1 antigens. Various VHH designs against spike have been generated and produced in CHO cells including Fc fusions to enhance their in vivo half-life. These modalities will be tested in surrogate neuralization assays to identify the best candidates for further development as potential therapeutic or diagnostic tools.

The rise in cancer, autoimmune, inflammatory, and infectious diseases in recent decades has led to a surge in the development of monoclonal antibodies (mAbs) therapies, now the most widely used family of biologics. To meet the growing global demand, biopharmaceutical industries are intensifying their production processes. One approach to achieve more efficient production of effective mAbs is to develop tools for real-time quality monitoring. Specifically, the glycosylation profile of mAbs must be closely monitored, since it greatly impacts their therapeutic efficacy and innocuity, making it a critical quality attribute. In this study, we developed a surface plasmon resonance-based integrated assay allowing for the simultaneous quantification and glycosylation characterization of mAbs in crude samples, hence permitting the at-line analysis of bioreactor cell cultures. Thanks to the high specificity of the interaction between biosensor surface-bound protein A and the Fc region of mAbs, we quantified crude IgG samples under mass transport limitations. Next, by flowing running buffer on the surface, impurities contained in the mAbs samples were washed away from the biosensor surface, allowing subsequent recording of the kinetics between the captured mAbs and injected FcγRII receptors. Of interest, with this strategy, we were able to quantify terminal galactosylation and core fucosylation of IgG lots, two important glycan modifications for mAb efficacy.

Generative artificial intelligence (AI) is transforming the design of large molecule therapeutics, significantly accelerating the pace and efficiency of drug development. Traditional discovery approaches rely on screening semi-random sequences, with optimization achieved through similarly stochastic diversification. This random exploration of sequence space yields unpredictable success rates and necessitates the evaluation of vast numbers of variants, making the process both time-consuming and costly. In contrast, AI-guided design offers a more informed, targeted strategy—marking a shift from a largely empirical discovery exercise to an engineering-driven discipline grounded in iterative design-build-measure-learn cycles. At Generate Biomedicines, this technology has been deployed to generate high-quality in silico variant sequences with exceptional speed and precision. However, while sequence design has advanced rapidly, the technologies supporting production and screening of these variants have not progressed at the same pace. As a result, wet lab processes—namely protein production, purification, and analytical chemistry—have become the new rate-limiting steps in the discovery pipeline. To overcome this bottleneck, the Protein Sciences Department at Generate has expanded its capabilities in protein expression, purification, and analytical chemistry through focused investment. The resulting infrastructure is notable for a company of our size and underscores our commitment to building a platform with lasting translational potential. This robust capability underpins our biotherapeutic discovery workflow, supporting all processes from high-throughput screening of thousands of variants to the delivery of comprehensive Lead Candidate Selection (LCS) and Development Candidate Nomination (DCN) packages to bolster our clinical pipeline.

Early manufacturability prediction of antibody-producing cell lines can be a lengthy and costly process necessary in the transition of lead biotherapeutics to process development and manufacturing. Significant time is required to identify and expand single CHO cell clones to practical culture volumes for in-depth cell line and recombinant protein characterization. We present a novel approach which enables -omics level (transcriptomic and proteomic) analysis of single and up to 10 cell clusters directly from stable cell pools using microlitres of culture. The method combines two novel developments: (1) a recombinant antibody secretion-dependent quantitative immunostaining approach to identify single cells expressing at relative levels, and (2) a microfluidic and laser-cell-lysis-based approach to pool cell lysates for -omics analyses. Using a panel of stable antibody-producing CHO cell pools, we demonstrate that the immunostaining approach quickly ranks cultures with good correlation to titer measurements made from shake flasks. Indeed, the immunostaining approach required only one hour expression to detect expression from pools compared to seven days needed for standard shake culture assays. Our novel quantitative immunostaining approach gives a relative ranking of CHO cell antibody expression at single cell resolution within a pool thereby informing which individual cell should be lysed for transcriptomics and proteomics. Based on the immunostaining signal intensity with high, medium, and low expression, we lysed and collected 1-to-10-cell lysates in microdroplets for downstream -omics analysis from several stable antibody expressing CHO cell pools. Heterogeneity in transcriptomics and proteomics profiles from single CHO cells within and between stable antibody producing pools was investigated. The approach is mostly automated using digital microfluidics, with potential for a fully autonomous platform in future iterations. We envision that this method will enable rapid and powerful investigations into the manufacturability of lead biotherapeutics by elucidating proteomic and transcriptomic signatures of maximally productive cells in a shortened timeline.

Human embryonic kidney cells (HEK293) serve as cell factories in viral vector manufacturing, particularly for recombinant adeno-associated virus (rAAV) production. They find widespread utilization in industrial applications, but a comprehensive characterization of the HEK293 genome and epigenome stability is still missing. To address this knowledge gap, the study employs a systematic approach to examine the genetic landscapes of various HEK293 cell lines. The objective is to evaluate their responses to changing environmental conditions and improve the current understanding of how these molecular mechanisms might influence rAAV production processes. Therefore, adherent HEK293 cells were adapted to suspension growth using various commercially available serum-free media formulations. Following successful adaptation, whole-genome deep sequencing was performed on both adapted and parental cell lines. The sequenced reads were then aligned to the human reference genome, enabling the assessment of genome stability, by evaluation of identified structural variants. Comparative analysis of these cell lines along with publicly available genome sequences of different HEK293 derivatives revealed a characteristic genetic signature common to all HEK293 cells, independent of cultivation conditions, phenotypic divergence or phylogenetic distance. Alterations in the distribution of structural variants including insertions and deletions, and of single nucleotide polymorphisms, indicate a continuous accumulation of genetic changes over time in culture, rather than abrupt genomic shifts in response to altered cultivation conditions. In contrast, adenoviral genes integrated into HEK293 cells appear to be highly conserved, as indicated by their stable copy number and consistent integration site. Overall, this work offers novel insights into the cellular response of various HEK293 cells to different cultivation conditions. Furthermore, it lays the groundwork for more comprehensive omics characterization, to support the development of cell lines with higher production efficiency.

The structural characterization of glycoproteins both as biotherapeutic products and as endogenous biomarkers in plasma and tissues remains one of the most analytically demanding tasks in modern bioanalysis. Glycosylation is a highly complex, heterogeneous, and context-dependent post-translational modification that influences protein folding, function, pharmacokinetics, and immunogenicity. High throughput (HTP) glycomics and glycoproteomics methods have recently been developed as essential tools to dissect this complexity, and their implementation into routine workflows for industry has been tested. Accurate glycoprotein analysis requires resolving site-specific glycoform variations across batches and navigating complex plasma and tissue samples with high dynamic range and structural isomers. Tissue analysis adds challenges like limited material and extracellular matrix interference. Accurate interpretation depends on advanced bioinformatics tools and machine learning algorithms, which must navigate an enormous search space of potential glycan structures and account for variable ionization efficiencies, incomplete fragmentation, and co-eluting species. Despite these hurdles, precise glycoprotein characterization is essential for therapeutic safety and for leveraging glycans as biomarkers in diseases such as cancer and neurodegeneration. Ongoing innovation is key to advancing high-throughput glycoanalysis in clinical and biopharmaceutical settings. We have developed HTP glycomics and glycoproteomics mass spectrometry-based methods to address these challenges. We have now optimized technologies ensuring the retention of labile residues such as sialic acids and fucose, and non-carbohydrate substituents like acetylation and sulfation. We have worked out advanced workflows that combine enrichment techniques, specialized digestion and derivatization protocols, and orthogonal mass spectrometry platforms (such as LC-MS/MS with EThcD, stepped HCD and UVPD) to structurally elucidate and preserve native glycosylation patterns. These developments are a continuation of our ongoing efforts of using state-of-the-art MS instrumentation to address newly arising difficulties in glycoprotein characterization and applying these tools to assist the industry in characterizing new biologics products.

Ensuring consistent product quality in cell culture-based vaccine manufacturing requires a thorough understanding of the process parameters that affect titers, yields, and impurity levels. This study applies Quality by Design (QbD) principles to an influenza A virus (IAV) production process operated in batch mode using two monoclonal suspension MDCK cell lines, C59 and C113 (Sartorius, Germany), with distinct charcteristics, focusing on process robustness and optimization. Based on knowledge from previous process development [1], a quantitative risk assessment including biological and technical parameters was performed to identify Critical Process Parameters (CPPs). Using a Design of Experiments (DoE) approach in an Ambr® 15 scale-down system, four key CPPs (pH value, dissolved oxygen concentration, viable cell concentration at time of infection, and multiplicity of infection) were investigated at three levels. After data analysis and modeling, we obtained dedicated design spaces for each cell clone characterized by high process robustness with a less than 1% risk of failure and even some indications for virus titer and yield improvement, while keeping process-related impurities such as DNA and total protein concentration low. Scale-up experiments in a 2 L single-use stirred tank bioreactor confirmed the validity of these conditions. Total virus titers of 2.95±0.06 log10(HAU/100 µL) and 3.13±0.12 log10(HAU/100 µL) were obtained for C59 and C113 cells, respectively [2]. By applying QbD principles, this study not only improves IAV production but also demonstrates a framework applicable to manufacturing of other cell culture-based vaccines. The results provide valuable insights for optimizing manufacturing processes, reducing batch failure risks, and supporting regulatory approval through data-driven process characterization. [1] Zinnecker et al., 2024, Eng. Life Sci., https://doi.org/10.1002/elsc.202300245 [2] Zinnecker et al., 2025, Eng. Life Sci., https://doi.org/10.1002/elsc.70027

Over the past decade, cell culture media (CCM) optimization has been a key strategy for obtaining high yields and improving productivity, while ensuring product quality in biopharmaceutical production. Conventional fed-batch media formulations for CHO cell cultivation typically consist of a basal medium, a pH-neutral main feed (feed A) as well as an alkaline feed (feed B). The separate, alkaline feed B is mainly needed to dissolve the key amino acids L-cystine and L-tyrosine, which are otherwise hardly soluble around pH 7 in aqueous solutions. However, these conventional media formulations come with inherent challenges such as high process complexity based on separate preparation as well as certain quality risks resulting from potential pH spikes and operator safety concerns. The transition from dual-feed to single-feed systems in CHO cell culture is a promising strategy to streamline biopharmaceutical production. It can reduce bioprocessing complexity by decreasing the number of vessels needed for the operation and even enhance monoclonal antibody titers. This study explores how the use of the chemically defined (di)peptides N,N’-di-L-lysyl-L-cystine [(Lys-Cys)2] and glycyl-L-tyrosine (Gly-Tyr) enables a simplified system comprising only the basal medium and one feed at neutral pH.

A novel monoclonal antibody (mAb) (anti-sST2) (U. Chile) with therapeutic potential for autoimmune disease Ulcerative colitis and other similar diseases with high levels of sST2 soluble protein was designed. For the production of this mAb, a possible secretory bottle neck is hyphotesized, observed at a previous analysis. Unfolded/misfolded proteins could cause reticulum stress, activating the unfolded protein response and possible apoptosis; also an increase on mAb demand could raise reactive oxygen species (ROS) leading to an accumulation of protein in the endoplasmic reticulum. To improve this scenario, process engineering approaches were studied, supplementing the culture medium with three different molecules: a novel and promissory antioxidant, a recognized antioxidant and a chemical additive, to decrease folding and aggregation problems, and with it is expected to enhance production of the novel mAb. Independent supplementation improved specific production (qP), with a value of 47%, 43.5% and 46.8% more than control for novel antioxidant, recognized antioxidant and chemical additive, respectively. Also, CHO cells were able to grow at optimal conditions on a studied range of the novel antioxidant, maintaining cell viability over 85% in culture, and improving the specific cell growth rate (µ) (0.55 1/d), a 30% higher than control, which also improved qP. As expected, supplementation of both a known antioxidant and a novel antioxidant were able to reduce intracellular ROS.

Influenza A virus (IAV) defective interfering particles (DIPs) hold great promise for the prevention and treatment of IAV infections. DIPs inhibit IAV replication and spread and counteract unrelated viral infections by activating the innate immune system. To investigate their potential use as antivirals, we used a systems biology approach that integrates mathematical modelling with experimental data to uncover the mechanisms underlying the effects of IAV DIPs.

The overlapping circulation of SARS-CoV-2, Influenza A virus (IAV), and Respiratory Syncytial Virus (RSV) continues to strain global healthcare systems, particularly among vulnerable populations. A single vaccine targeting all three pathogens could streamline immunization efforts and enhance protection, while reducing manufacturing costs. We previously showed that CHO cell–derived enveloped virus-like particles (eVLPs) formed through expression of full-length SARS-CoV-2 spike (S) protein not only exhibit high S density and strong immunogenicity, but also serve as a platform to co-display heterologous antigens such as IAV hemagglutinin (H1) and neuraminidase (N1). Here, we extend this approach by producing a trivalent eVLP candidate that simultaneously displays SARS-CoV-2 S, IAV H1, and the RSV pre-fusogenic fusion (F) protein. These S/H1/F eVLPs were successfully produced using both transient and stable gene expression in CHO cells and were purified via affinity chromatography. The presence of all three antigens on the same particles was confirmed by Western blot and immuno-electron microscopy. Their immunogenicity is currently being evaluated in vivo to assess their potential as a single vaccine against SARS-CoV-2, IAV, and RSV.

The emergence of cell and gene therapies in recent years, combined with an increasing number of approved therapies, has meant that more and more previously incurable diseases can now be treated. However, the availability and cost of these therapies remains a major challenge that needs to be addressed. One way to make life-saving therapies more accessible and affordable for patients worldwide is through process intensification (PI). PI can be achieved by optimizing various aspects of the cell culture process, such as cell density, media formulation, and process type, resulting in more robust and scalable processes. While PI is well established for CHO-based protein production, this field is still largely uncharted territory in HEK293-based gene and cell therapies. Sartorius Xell leverages extensive expertise in developing advanced cell culture media, alongside a comprehensive portfolio of analytical techniques and equipment scaling options, to enhance cell culture processes. This toolbox has facilitated the creation of a specialized cell culture medium for HEK293 suspension cultures, capable of sustaining exceptionally high cell densities of 100 million cells per milliliter over extended durations. Building on established techniques from CHO-based therapeutic protein production, the HEK293 cell line, renowned for its applications in recombinant protein production and gene therapy, was integrated into a continuous perfusion process. This approach ensured optimal growth conditions and nutrient supply, allowing the cells to maintain their ability to be transiently transfected even at elevated densities. The study demonstrates the feasibility of gene transfer under intensified conditions, highlighting the potential of perfusion systems to revolutionize cell-based production processes. These advancements promise increased bioproduct yields while preserving the functional integrity of cells for transient transfection applications, paving the way for more efficient and scalable biomanufacturing strategies.

Optimization of recombinant adeno-associated virus (rAAV) production is essential for effective gene therapy applications. However, multiple factors affect the rAAV productivity in mammalian cells, and often they interact with each other, making the optimization process highly challenging. In our work, we show how coupling mixture design (MD) with face-centered central composite design (FCCD) is the most suitable design of experiments (DOE) approach, among other common DOE methods, for optimizing rAAV2 productivity and cell viability. moreover, we build on this method and demonstrate that combining MD with FCCD can be used to optimize the percentage of full capsids in rAAV2 upstream preparation. Additionally, we investigate the influence of the gene of interest (GOI) on the optimal conditions for viral particle production and packaging efficiency. By integrating MD and FCCD methodologies, we achieved an improvement of almost 100-fold in Log(Vp) in the case of egfp-expressing rAAV, and a 12-fold increase in bdnf-expressing full rAAV capsids, suggesting that this combined approach is a versatile and effective strategy for optimizing rAAV production processes. These findings emphasize the need for a comprehensive understanding of the factors influencing rAAV production to enhance the efficiency and efficacy of viral vector applications in gene therapy.

High expression levels of virus-like particles (VLPs) are essential for vaccination and diagnostic applications. Although the Baculovirus Expression Vector System (BEVS) in insect cells is widely employed for the production of VLPs due to its high yields, it faces significant bottlenecks. These challenges include optimization problems of the required ratios of structural proteins and the co-production of baculoviral proteins and particles, resulting in a cumbersome purification. Our plasmid-based High Five insect cell expression system overcomes these limitations. In a direct comparison of BEVS and the plasmid-based system regarding the yield and quality of Noro, Entero, and Rota VLPs, we were able to highlight the advantages of the plasmid-based system. Furthermore, we produced SARS-CoV-2 VLPs for antibody development. The quality of these VLPs was validated by Nanotracking Analysis, ELISA, cytometry, and microscopy, confirming a diameter of ~145 nm, ACE2 binding and the typical "corona" aura. These fluorescent SARS-CoV-2 VLPs have been used in cell-based assays and enable high-throughput screening of anti-SARS-CoV-2 antibody candidates for their capability to inhibit binding to ACE2 positive cells through cytometry. We now transfer this system to other viruses like Hantavirus and West Nile virus. In summary, we demonstrate the successful production of VLPs in our plasmid-based insect cell system, achieving both high quality and high yield, and their subsequent application in antibody development.

Adeno-associated virus (AAV) is a leading viral vector for in vivo gene therapies. However, current AAV production titers and quality remain insufficient for large-scale applications. To improve AAV production, understanding the biomolecular mechanism of the AAV biosynthesis pathway is crucial. In the context of an AAV manufacturing platform that relies on infection of a stable producer cell line (PCL) with a wild-type adenovirus, a particular interest is how the host cell machinery is exploited by both the adenoviral helper virus and AAV. As AAV is replication-deficient on its own, it requires co-infection with a helper virus such as adenovirus, which reprograms the host environment to enable AAV genome replication and capsid assembly. However, the molecular steps underlying this coordinated process remain poorly understood, making it challenging to pinpoint where deficiencies arise in AAV-producing PCL clones. Here we present a genome-scale reconstruction of the AAV biosynthesis pathway, comprising 2,863 manually curated genes annotated into 125 process terms across 22 subsystems. This reconstruction also incorporates the induction of AAV production via wild-type adenovirus infection, providing a comprehensive resource for systems-level analysis of AAV biosynthesis. We overlaid transcriptomic data from AAV-producing clones onto the biosynthetic pathways to identify functional modules associated with variations in AAV productivity and quality. Comparative gene set enrichment analyses using the AAV-specific reconstruction versus conventional Gene Ontology (GO) annotations demonstrate improved contextualization of regulated pathways under different conditions. Network-based visualization and clustering approaches enable the mapping of gene expression dynamics onto the reconstructed network, revealing condition- and time-dependent changes in pathway activity. Identified pathways of interest may be further correlated with AAV titer and vector quality to investigate potential mechanisms driving productivity variance. Together, the reconstruction serves as a systems-level framework for interpreting gene regulation during AAV production. This resource supports transcriptomic profiling, facilitates bioprocess optimization, and lays the groundwork for future identification of key functional interplay modules between producer cell line, AAV and adenovirus during AAV biosynthesis.