Search

Purpose: With the increasing demand for intensified processes in biopharmaceutical manufacturing, there is a growing requirement for a rapid and precise analytical method to measure product yield. Aligning with FDA guidelines, this requirement aims to enhance process monitoring and control. Currently, the monitoring of monoclonal antibody (mAb) titer during manufacturing heavily relies on offline testing conducted in quality control laboratories. However, the lengthy turnaround time for analytical results, often taking hours or even days, undermines their usefulness for process control and decision-making. Methods: This manuscript presents a case study on HaLCon® protein analyzer, a compact Liquid Chromatography (LC) system equipped with a Protein A affinity column. HaLCon® is specifically designed for at-line/on-line titer measurement directly at the biopharmaceutical manufacturing site, enabling the rapid acquisition of product yield information. The study investigates the capabilities, performance, and potential applications of HaLCon® at the laboratory scale. Results and Conclusion: The results obtained demonstrate the effectiveness and reliability of HaLCon® as an analytical instrument for at-line titer measurement, offering valuable insights into its potential application within the biopharmaceutical manufacturing industry. [ABSTRACT FROM AUTHOR]

Advanced process control in the biopharmaceutical industry often lacks real‐time measurements due to resource constraints. Raman spectroscopy and Partial Least Squares (PLS) models are often used to monitor bioprocess cultures in real‐time. In spite of the ease of training, the accuracy of the PLS model is impacted if it is not used to predict quality attributes for the cell lines it is trained on. To address this issue, a deep convolutional neural network (CNN) is proposed for offline modeling of metabolites using Raman spectroscopy. By utilizing asymmetric least squares smoothing to adjust Raman spectra baselines, a generic training data set is created by amalgamating spectra from various cell lines and operating conditions. This data set, combined with their derivatives, forms a two‐dimensional model input. The CNN model is developed and validated for predicting different quality variables against measurements from various continuous and fed‐batch experimental runs. Validation results confirm that the deep CNN model is an accurate generic model of the process to predict real‐time quality attributes, even in experimental runs not included in the training data. This model is robust and versatile, requiring no recalibration when deployed at different sites to monitor various cell lines and experimental runs. [ABSTRACT FROM AUTHOR]

BACKGROUND: This study aimed to establish chemometric models using Raman spectroscopy data for biochemical monitoring of rabies Virus‐Like Particles (VLP) production based on baculovirus/insect cell system. The models were developed using fresh and stored samples from the initial development stages (Schott culture flasks). The following modeling techniques were assessed: partial least squares (PLS) and artificial neural networks (ANN). The effects of spectral filtering approaches, spectral ranges (400–1850 cm−1; 100–3425 cm−1), and sample cryopreservation were also considered. The applicability of the models was evaluated using experimental data from assays carried out in a benchtop bioreactor. RESULTS: The results showed that the prediction capacity of the chemometric models was negatively impacted when samples from rabies VLP production were cryopreserved. Further studies are needed to confirm the maximum storage time for samples (< 4 months) without a significant difference in model predictions compared to those from an in line database. The dilution of the sample should be kept constant throughout the rabies VLP development stages. A nonlinear correlation was observed between dilution and the predicted values of biochemical parameters from Raman spectral data. The choice of spectral filtering has a major impact on the prediction accuracy of chemometric models. CONCLUSION: The optimal filtering approach should be individually optimized for each biochemical parameter. The ANN models were significantly more suitable for biochemical monitoring than the PLS approach. The 400–1850 cm−1 Raman shift range is recommended for biochemical monitoring of rabies VLP using a baculovirus/insect cell platform when samples are cell‐free. © 2023 Society of Chemical Industry (SCI). [ABSTRACT FROM AUTHOR]

Fermentation monitoring is a powerful tool for bioprocess development and optimization. On‐line metabolomics is a technology that is starting to gain attention as a bioprocess monitoring tool, allowing the direct measurement of many compounds in the fermentation broth at a very high time resolution. In this work, targeted on‐line metabolomics was used to monitor 40 metabolites of interest during three Escherichia coli succinate production fermentation experiments every 5 min with a triple quadrupole mass spectrometer. This allowed capturing high‐time resolution biological data that can provide critical information for process optimization. For nine of these metabolites, simple univariate regression models were used to model compound concentration from their on‐line mass spectrometry peak area. These on‐line metabolomics univariate models performed comparably to vibrational spectroscopy multivariate partial least squares regressions models reported in the literature, which typically are much more complex and time consuming to build. In conclusion, this work shows how on‐line metabolomics can be used to directly monitor many bioprocess compounds of interest and obtain rich biological and bioprocess data. [ABSTRACT FROM AUTHOR]

Pharmaceutical manufacturing is reliant upon bioprocessing approaches to generate the range of therapeutic products that are available today. The high cost of production, susceptibility to process failure, and requirement to achieve consistent, high-quality product means that process monitoring is paramount during manufacturing. Process analytic technologies (PAT) are key to ensuring high quality product is produced at all stages of development. Spectroscopy-based technologies are well suited as PAT approaches as they are non-destructive and require minimum sample preparation. This study explored the use of a novel attenuated total reflection Fourier transform infrared (ATR-FTIR) spectroscopy platform, which utilises disposable internal reflection elements (IREs), as a method of upstream bioprocess monitoring. The platform was used to characterise organism health and to quantify cellular metabolites in growth media using quantification models to predict glucose and lactic acid levels both singularly and combined. Separation of the healthy and nutrient deficient cells within PC space was clearly apparent, indicating this technique could be used to characterise these classes. For the metabolite quantification, the binary models yielded R2 values of 0.969 for glucose, 0.976 for lactic acid. When quantifying the metabolites in tandem using a multi-output partial least squares model, the corresponding R2 value was 0.980. This initial study highlights the suitability of the platform for bioprocess monitoring and paves the way for future in-line developments.

The unstructured mechanistic model (UMM) allows for modeling the macro-scale of a phenomenon without known mechanisms. This is extremely useful in biomanufacturing because using the UMM for the joint estimation of states and parameters with an extended Kalman filter (JEKF) can enable the real-time monitoring of bioprocesses with unknown mechanisms. However, the UMM commonly used in biomanufacturing contains ordinary differential equations (ODEs) with unshared parameters, weak variables, and weak terms. When such a UMM is coupled with an initial state error covariance matrix P (t = 0) and a process error covariance matrix Q with uncorrelated elements, along with just one measured state variable, the joint extended Kalman filter (JEKF) fails to estimate the unshared parameters and state simultaneously. This is because the Kalman gain corresponding to the unshared parameter remains constant and equal to zero. In this work, we formally describe this failure case, present the proof of JEKF failure, and propose an approach called SANTO to side-step this failure case. The SANTO approach consists of adding a quantity to the state error covariance between the measured state variable and unshared parameter in the initial P(t = 0) of the matrix Ricatti differential equation to compute the predicted error covariance matrix of the state and prevent the Kalman gain from being zero. Our empirical evaluations using synthetic and real datasets reveal significant improvements: SANTO achieved a reduction in root-mean-square percentage error (RMSPE) of up to approximately 17% compared to the classical JEKF, indicating a substantial enhancement in estimation accuracy. [ABSTRACT FROM AUTHOR]

Fluorescence spectroscopy is a non-invasive and highly sensitive method for bioprocess monitoring. The use of fluorescence spectroscopy is not very well established in the industry for in-line monitoring. In the present work, a 2-D fluorometer with two excitation lights (365 and 405 nm) and emission spectra in the range of 350–850 nm were used for in-line monitoring of two strains of Bordetella pertussis cultivation operated in batch and fed batch. A Partial Least Squares (PLS) based regression model was used for the estimation of cell biomass, amino acids (glutamate and proline) and antigen (Pertactin) produced. It was observed that accurate predictions were achieved when models were calibrated separately for each cell strain and nutrient media formulation. Also, prediction accuracy was improved when dissolved oxygen, agitation and culture volume are added as additional features in the regression model. The proposed approach of combining in-line fluorescence and other online measurements is shown to have good potential for in-line monitoring of bioprocesses. [ABSTRACT FROM AUTHOR]

In this study, we developed a method to build Raman calibration models without culture data for cell culture monitoring. First, Raman spectra were collected and then analyzed for the signals of all the mentioned analytes: glucose, lactate, glutamine, glutamate, ammonia, antibody, viable cells, media, and feed agent. Using these spectral data, the specific peak positions and intensities for each factor were detected. Next, according to the design of the experiment method, samples were prepared by mixing the above-mentioned factors. Raman spectra of these samples were collected and were used to build calibration models. Several combinations of spectral pretreatments and wavenumber regions were compared to optimize the calibration model for cell culture monitoring without culture data. The accuracy of the developed calibration model was evaluated by performing actual cell culture and fitting the in-line measured spectra to the developed calibration model. As a result, the calibration model achieved sufficiently good accuracy for the three components, glucose, lactate, and antibody (root mean square errors of prediction, or RMSEP = 0.23, 0.29, and 0.20 g/L, respectively). This study has presented innovative results in developing a culture monitoring method without using culture data, while using a basic conventional method of investigating the Raman spectra of each component in the culture media and then utilizing a design of experiment approach. Graphical Abstract This is a visual representation of the abstract. [ABSTRACT FROM AUTHOR]

The present contribution deals with the simultaneous estimation of the reaction rates and input disturbances in a bioprocess. Online estimation of just the reaction rates is possible by sliding mode observer techniques or high-gain observers, in finite-time or with exponential convergence, respectively. However, these techniques are not robust against unpredictable input bioreactor disturbances and are not readily adaptable to treat external disturbances. Motivated by these facts, we propose a novel extended super-twisting algorithm for a class of nonlinear systems, such that the estimation of uncertain parameters, unknown internal dynamics, and external disturbances converge in a finite time or a neighborhood near its nominal values. The effectiveness of the developed algorithm is tested through simulations in an anaerobic digestion process using a batch and a continuous bioreactor. Simulation results show good robustness performance of the extended super-twisting algorithm. • An extended super-twisting observer-based estimator is proposed. • The stability of the proposed algorithm is proved within Lyapunov's formalism. • Bioprocess reaction rates and input disturbance can be estimated simultaneously. [ABSTRACT FROM AUTHOR]

This work aimed to quantify growth and biochemical parameters (viable cell density, Xv; cell viability, CV; glucose, lactate, glutamine, glutamate, ammonium, and potassium concentrations) in upstream stages to obtain rabies virus-like particles (rabies VLP) from insect cell-baculovirus system using on-line and off-line Raman spectra to calibrate global models with minimal experimental data. Five cultivations in bioreactor were performed. The first one comprised the growth of uninfected Spodoptera frugiperda (Sf9) cells, the second and third runs to obtain recombinant baculovirus (rBV) bearing Rabies G glycoprotein and matrix protein, respectively. The fourth one involved the generation of rabies VLP from rBVs and the last one was a repetition of the third one with cell inoculum infected by rBV. The spectra were acquired through a Raman spectrometer with a 785-nm laser source. The fitted Partial Least Square models for nutrients and metabolites were comparable with those previously reported for mammalian cell lines (Relative error < 15 %). However, the use of this chemometrics approach for Xv and CV was not as accurate as it was for other parameters. The findings from this work established the basis for bioprocess Raman spectroscopical monitoring using insect cells for VLP manufacturing, which are gaining ground in the pharmaceutical industry. [Display omitted] • Modeling from Raman spectra using PLS for rabies VLP production monitoring. • Raman spectral range pick for monitoring biochemically rabies VLP process. • Effective online biochemical monitoring over rabies VLP production with Raman data. • Suitable biochemical monitoring over rabies VLP production with offline Raman data. • Assessed parameters relative error was less than 15 % using online or offline data. [ABSTRACT FROM AUTHOR]

The unstructured mechanistic model (UMM) allows for modeling the macro-scale of a phenomenon without known mechanisms. This is extremely useful in biomanufacturing because using the UMM for the joint estimation of states and parameters with an extended Kalman filter (JEKF) can enable the real-time monitoring of bioprocesses with unknown mechanisms. However, the UMM commonly used in biomanufacturing contains ordinary differential equations (ODEs) with unshared parameters, weak variables, and weak terms. When such a UMM is coupled with an initial state error covariance matrix P(t=0) and a process error covariance matrix Q with uncorrelated elements, along with just one measured state variable, the joint extended Kalman filter (JEKF) fails to estimate the unshared parameters and state simultaneously. This is because the Kalman gain corresponding to the unshared parameter remains constant and equal to zero. In this work, we formally describe this failure case, present the proof of JEKF failure, and propose an approach called SANTO to side-step this failure case. The SANTO approach consists of adding a quantity to the state error covariance between the measured state variable and unshared parameter in the initial P(t = 0) of the matrix Ricatti differential equation to compute the predicted error covariance matrix of the state and prevent the Kalman gain from being zero. Our empirical evaluations using synthetic and real datasets reveal significant improvements: SANTO achieved a reduction in root-mean-square percentage error (RMSPE) of up to approximately 17% compared to the classical JEKF, indicating a substantial enhancement in estimation accuracy.

Over the last decades, the success of advanced cell therapies and the increasing production volumes of vaccines, proteins, or viral vectors have raised the need of robust cell‐based manufacturing processes for ensuring product quality and satisfying good manufacturing practice requirements. The cultivation process of cells needs to be highly controlled for improved productivity, reduced variability, and optimized bioprocesses. Cell cultures can be easily monitored using different technologies, which could deliver direct or indirect assessment of the cells' viability. Among these techniques, nuclear magnetic resonance (NMR) spectroscopy is a powerful technology that permits the evaluation and the identification of key endogenous metabolites. NMR can provide information on the cell metabolic pathways, on the bioprocesses, and is also capable to quickly test for impurities. In this study, NMR was successfully used as a technology for monitoring cell viability and expansion in different supports for cell growth (including bioreactors), to predict the bioprocess output and for the early identification of key metabolites linked to cell starvation. This investigation will allow the timely control of culture conditions and favor the optimization of the bioprocesses. [ABSTRACT FROM AUTHOR]

Recombinant Escherichia coli grown in large-scale fermenters are used extensively to produce plasmids and biopharmaceuticals. One method commonly used to control culture growth is predefined glucose feeding, often an exponential feeding profile. Predefined feeding profiles cannot adjust automatically to metabolic state changes, such as the metabolic burden associated with recombinant protein expression or high-cell density associated stresses. As the culture oxygen consumption rates indicates a culture's metabolic state, there exist several methods to estimate the oxygen uptake rate (OUR). These common OUR methods have limited application since these approaches either disrupt the oxygen supply, rely on empirical relationships, or are unable to account for latency and filtering effects. In this study, an oxygen transfer rate (OTR) estimator was developed to aid OUR prediction. This non-disruptive OTR estimator uses the dissolved oxygen and the off-gas oxygen concentration, in parallel. This new OTR estimator captures small variations in OTR due to physical and chemical manipulations of the fermenter, such as in stir speed variation, glucose feeding rate change, and recombinant protein expression. Due its sensitivity, this non-disruptive real-time OTR estimator could be integrated with feed control algorithms to maintain the metabolic state of a culture to a desired setpoint. [Display omitted] • Non-disruptive k L a method allowed real-time OTR estimate and ensuing OUR assessment. • Recursive least squares method captured more high-frequency information. • First-order mixing model used to addressed latency and filtering effects of off-gas sensing. • Non-linear k L a and stir speed correlation continuously updated throughout fermentation. [ABSTRACT FROM AUTHOR]

Single-chain fragment variables (scFvs), composed of variable heavy and light chains joined together by a peptide linker, can be produced using a cost-effective bacterial expression system, making them promising candidates for pharmaceutical applications. However, a versatile method for monitoring recombinant-protein production has not yet been developed. Herein, we report a novel anti-scFv aptamer-based biosensing system with high specificity and versatility. First, anti-scFv aptamers were screened using the competitive systematic evolution of ligands by exponential enrichment, focusing on a unique scFv-specific peptide linker. We selected two aptamers, P1-12 and P2-63, with K D = 2.1 μM or K D = 1.6 μM toward anti-human epidermal growth factor receptor (EGFR) scFv, respectively. These two aptamers can selectively bind to scFv but not to anti-EGFR Fv. Furthermore, the selected aptamers recognized various scFvs with different CDRs, such as anti-4-1BB and anti-hemoglobin scFv, indicating that they recognized a unique peptide linker region. An electrochemical sensor for anti-EGFR scFv was developed using anti-scFv aptamers based on square wave voltammetry. Thus, the constructed sensor could monitor anti-EGFR scFv concentrations in the range of 10–500 nM in a diluted medium for bacterial cultivation, which covered the expected concentration range for the recombinant production of scFvs. These achievements promise the realization of continuous monitoring sensors for pharmaceutical scFv, which will enable the real-time and versatile monitoring of large-scale scFv production. [ABSTRACT FROM AUTHOR]

In protein refolding processes the scarce availability of online measurements hampers effective process monitoring. In this work we developed a mechanistic soft-sensor for protein refolding based on online intrinsic fluorescence measurements of tryptophan and tyrosine. In validation experiments using two model proteins, lactate dehydrogenase (LDH) and galactose oxidase, the soft-sensor showed accurate estimates for the prediction of the total sum of folding products (NRMSE < 6. 1 %) by calculating the changing rate of the average emission wavelength. For refolding of the enzyme LDH it was possible to obtain separate predictions of native protein and insoluble aggregates. The soft-sensor design was further extended by a model-based observer approach using particle filtering to incorporate kinetic formulations as well as physical constraints. The novel approach enabled the analysis of kinetic mechanisms during rapid reaction dynamics and can therefore be seen as an enabler to achieve a better understanding of kinetic refolding mechanisms. [Display omitted] • Soft-sensor development for protein refolding based on online intrinsic fluorescence. • Using model-based observer over direct calculation allowed for physical constraints. • Soft-sensor enabled analysis of highly dynamic kinetic folding mechanisms. [ABSTRACT FROM AUTHOR]

The real‐time monitoring of metabolites (RTMet) is instrumental for the industrial production of biobased fermentation products. This study shows the first application of untargeted on‐line metabolomics for the monitoring of undiluted fermentation broth samples taken automatically from a 5 L bioreactor every 5 min via flow injection mass spectrometry. The travel time from the bioreactor to the mass spectrometer was 30 s. Using mass spectrometry allows, on the one hand, the direct monitoring of targeted key process compounds of interest and, on the other hand, provides information on hundreds of additional untargeted compounds without requiring previous calibration data. In this study, this technology was applied in an Escherichia coli succinate fermentation process and 886 different m/z signals were monitored, including key process compounds (glucose, succinate, and pyruvate), potential biomarkers of biomass formation such as (R)‐2,3‐dihydroxy‐isovalerate and (R)‐2,3‐dihydroxy‐3‐methylpentanoate and compounds from the pentose phosphate pathway and nucleotide metabolism, among others. The main advantage of the RTMet technology is that it allows the monitoring of hundreds of signals without the requirement of developing partial least squares regression models, making it a perfect tool for bioprocess monitoring and for testing many different strains and process conditions for bioprocess development. [ABSTRACT FROM AUTHOR]

The properties of the convergence region of the estimation error of a robust observer for second-order systems are determined, and a new algorithm is proposed for setting the observer parameters, considering persistent but bounded disturbances in the two observation error dynamics. The main contributions over closely related studies of the stability of state observers are: (i) the width of the convergence region of the observer error for the unknown state is expressed in terms of the interaction between the observer parameters and the disturbance terms of the observer error dynamics; (ii) it was found that this width has a minimum point and a vertical asymptote with respect to one of the observer parameters, and their coordinates were determined. In addition, the main advantages of the proposed algorithm over closely related algorithms are: (i) the definition of observer parameters is significantly simpler, as the fulfillment of Riccati equation conditions, solution of LMI constraints, and fulfillment of eigenvalue conditions are not required; (ii) unknown bounded terms are considered in the dynamics of the observer error for the known state. Finally, the algorithm is applied to a model of microalgae culture in a photobioreactor for the estimation of biomass growth rate and substrate uptake rate based on known concentrations of biomass and substrate. [ABSTRACT FROM AUTHOR]

Biopharmaceutical process development often involves the use of small-scale bioreactors (SSBR) for optimizing media formulations and process conditions during scale up to commercial scale production. Two key process parameters (CPP) used in SSBR studies are protein titre and viable cell density (VCD). Here, we explore the efficacy of parallel polarized total synchronous fluorescence spectroscopy (TSFS || ) and Synchronous Light Scattering (SyLS || ) to qualitatively monitor these CPPs and quantitatively predict titre and VCD for a large-scale cell culture media optimization SSBR study. The study involved 71 different media formulations (50+ components each), and the bioprocess was run for 13 days or more. Samples were extracted at set times (Day 0, 3, 9, and 13) and clarified by centrifugation. TSFS || spectra showed significant emission changes along with increased light scatter over course of the bioprocess. SyLS || measurements strongly correlated with particle size data obtained from Dynamic Light Scattering but did not correlate well with VCD probably because of the centrifugation-based sample preparation. Statistical and principal component analysis (PCA) of the pTSFS data showed that spectral variation was greater between media formulations than due to the evolving bioprocess. This prevented the development of accurate global prediction models for media performance (e.g., predicting product titre at day 9 from media spectra measured at day 0). However, classification methods were successfully used to select media subsets with better quantitative prediction accuracy based on spectral similarities. A practical binary (high/low performance) classification model based on Support Vector Machines was generated for media formulation screening. Combining emission and scatter measurements with multivariate data analysis provides a more holistic, multi-attribute bioprocess monitoring method that minimizes the need to use different offline analytical methods. This methodology can be used to monitor process trajectories and deviations, and ultimately be used to predict bioprocess CPPs when implemented on production scale processes where there is much less compositional variation in the media. We believe this SSBR-pTSFS/SyLS approach will provide a valuable resource to develop the design/parameter space for in process monitoring at production scale from early-stage process/media development studies., Competing Interests: Declaration of Competing Interest The authors declare the following financial interests/personal relationships which may be considered as potential competing interests. Alan G. Ryder reports financial support was provided by Science Foundation Ireland. Alan G. Ryder reports equipment, drugs, or supplies was provided by Agilent Technologies (Mulgrave Victoria, Australia). The samples used in this study were provided by a third party Biopharmaceutical manufacturer, who for confidentiality issues do not wish to be identified. This has no bearing on the scientific content of the manuscript. If there are other authors, they declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. CONFLICT OF INTEREST The authors declare no financial or commercial conflict of interest., (Copyright © 2024. Published by Elsevier B.V.)

The Zika disease caused by the Zika virus was declared a Public Health Emergency by the World Health Union (WHO), with microcephaly as the most critical consequence. Aiming to reduce the spread of the virus, biopharmaceutical organizations invest in vaccine research and production, based on multiple platforms. A crescent vaccine production approach is based on virus-like particles (VLP), for not having genetic material in its composition, hypoallergenic and non-mutant character. For bioprocess, it is essential to have means of real-time monitoring, which can be assessed using process analysis techniques such as Near-infrared (NIR) spectroscopy, that can be combined with chemometric methods, like Partial-Least Squares (PLS) and Artificial Neural Networks (ANN) for prediction of biochemical variables. This work proposes a biochemical Zika VLP upstream production at-line monitoring model using NIR spectroscopy comparing sampling conditions (with or without cells), analytical blank (air, ultrapure water), and spectra pre-processing approaches. Seven experiments in a benchtop bioreactor using recombinant baculovirus/Sf9 insect cell platform in serum-free medium were performed to obtain biochemical and spectral data for chemometrics modeling (PLS and ANN), composed by a random data split (80 % calibration, 20 % validation) for cross-validation of the PLS models and 70 % training, 15 % testing, 15 % validation for ANN. The best models generated in the present work presented an average absolute error of 1.59 × 10 5  cell/mL for density of viable cells, 2.37 % for cell viability, 0.25 g/L for glucose, 0.007 g/L for lactate, 0.138 g/L for glutamine, 0.18 g/L for glutamate, 0,003 g/L for ammonium, and 0.014 g/L for potassium., Competing Interests: Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper., (Copyright © 2024 Elsevier B.V. All rights reserved.)

The monitoring and control of bioprocesses rely on the measurement of the main metabolite concentrations. To this end, infrared spectroscopy (IR) is a good candidate with which to perform rapid and non-destructive measurements. However, IR-based measurements rely on a calibration step linking the measured spectra to the concentrations of the compounds of interest. This calibration may suffer with problems of robustness when the measuring conditions change, such as when some chemicals not present in the calibration spectra are added when using the IR sensor. In this study, a method based on orthogonal projection, dynamic orthogonal projection (DOP), was tested for its ability to cope with the robustness problem caused by the addition of ammonia in a pilot-scale anaerobic digester, whose volatile fatty acid concentrations were monitored by mid-IR spectrometry. The results demonstrate that DOP has significant potential as a form of process analytical technology. [ABSTRACT FROM AUTHOR]

[Display omitted] • Monitoring of Rabies virus-like particles production using online Raman spectra. • Nonlinear models showed a better fit than linear ones in biochemical monitoring. • The models can be used in rabies virus-like particle commercial applications. • Variables absolute error was like those for equivalent offline work. • Model accuracies were alike chemometrics models for other cells hosts. This work aimed to set inline Raman spectroscopy models to monitor biochemically (viable cell density, cell viability, glucose, lactate, glutamine, glutamate, and ammonium) all upstream stages of a virus-like particle-making process. Linear (Partial least squares, PLS; Principal components regression, PCR) and nonlinear (Artificial neural networks, ANN; supported vector machine, SVM) modeling approaches were assessed. The nonlinear models, ANN and SVM, were the more suitable models with the lowest absolute errors. The mean absolute error of the best models within the assessed parameter ranges for viable cell density (0.01–8.83 × 106 cells/mL), cell viability (1.3–100.0 %), glucose (5.22–10.93 g/L), lactate (18.6–152.7 mg/L), glutamine (158–1761 mg/L), glutamate (807.6–2159.7 mg/L), and ammonium (62.8–117.8 mg/L) were 1.55 ± 1.37 × 106 cells/mL (ANN), 5.01 ± 4.93 % (ANN), 0.27 ± 0.22 g/L (SVM), 4.7 ± 2.6 mg/L (SVM), 51 ± 49 mg/L (ANN), 57 ± 39 mg/L (SVM) and 2.0 ± 1.8 mg/L (ANN), respectively. The errors achieved, and best-fitted models were like those for the same bioprocess using offline data and others, which utilized inline spectra for mammalian cell lines as a host. [ABSTRACT FROM AUTHOR]

This work assessed the impact of laser wavelength and sample conditioning on offline monitoring (viable cell density, cell viability, virus titer, glucose, lactate, glutamine, glutamate, and ammonium) of SARS-CoV-2 virus-like particles production upstream stage by Raman spectroscopy. The evaluated chemometrics techniques were Partial Least Squares (PLS) and Artificial Neural Networks (ANN), and different spectral filtering approaches were also considered. ANN showed better prediction capacity for most of the parameters, but ammonium and lactate, better predicted by PLS, and glutamine, no difference between modeling techniques was detected. For cell growth parameters and virus titer, samples without cells measured at 785 nm Raman laser wavelength originated better-adjusted models. This laser wavelength was also more appropriate for the remaining monitored experimental parameters except for glucose, in which the best model came from the spectral database acquired at 1064 nm wavelength. Cell remotion significantly increased the accuracy of viable cell density, cell viability, glutamate, and virus titer models. However, glucose, lactate, and ammonium models showed better prediction capacity for samples containing cells. Thus, it was demonstrated that laser wavelength, sample conditioning, spectral preprocessing, and chemometric modeling techniques need to be tailored for each experimental parameter to improve accuracy. • Cell growth and virus titer were better modeled using 785 nm Raman laser. • Biochemical parameters were also better modeled at 785 nm except for glucose. • Log 10 transformation of virus titer significantly improved the prediction quality. • Cell remotion improved model's accuracy for cell growth, glutamate, and virus titer. • Laser wavelength, sample conditioning, spectra filtering must be studied together. [ABSTRACT FROM AUTHOR]

Glucose is one of the most vital nutrients in all living organisms, so its monitoring is critical in healthcare and bioprocessing. Enzymatic sensors are more popular as a technology solution to meet the requirement. However, periplasmic binding proteins have been investigated extensively for their high sensitivity, enabling microdialysis sampling to replace existing complex and expensive glucose monitoring solutions based on enzymatic sensors. The binding proteins are used as optical biosensors by introducing an environment-sensitive fluorophore to the protein. The biosensor's construction, characterization, and potential application are well studied, but a complete glucose monitoring system based on it is yet to be reported. This work documents the development of the first glucose sensor prototype based on glucose binding protein (GBP) for automatic and continuous glucose measurements. The development includes immobilizing the protein into reusable chips and a low-cost solution for non-invasive glucose sampling in bioprocesses using microdialysis sampling technique. A program was written in LabVIEW to accompany the prototype for the complete automation of measurement. The sampling technique allowed glucose measurements of a few micromolar to 260 mM glucose levels. A thorough analysis of the sampling mode and the device's performance was conducted. The reported measurement accuracy was 81.78%, with an RSD of 1.83%. The prototype was also used in online glucose monitoring of E. coli cell culture. The mode of glucose sensing can be expanded to the measurement of other analytes by switching the binding proteins. [Display omitted] [ABSTRACT FROM AUTHOR]

This paper addresses the design and evaluation of a robust observer for second order bioprocesses considering unknown bounded disturbance terms and uncertainty in the dynamics of the unknown and known states. The observer design and the stability analysis are based on dead-zone Lyapunov functions, and a detailed procedure is provided. The transient response bounds and the convergence region of the unknown observer error are determined in terms of the disturbance bounds, considering persistent but bounded disturbances in the dynamics of both the known and unknown observer errors. This is a significant contribution to closely related observer design studies, in which the transient response bounds are determined, but persistent and bounded disturbances are not considered in the dynamics of the known observer error. Other important contributions are: (i) the procedure for defining the observer parameters is significantly simpler than common observer designs, since a solution to the Ricatti equation, solution to LMI constraints, or the accomplishment of eigenvalue inequality conditions are not required; (ii) discontinuous signals are not used in the observer; and (iii) the effect of the gain sign associated with the unknown state in the dynamics of the known state is explicitly and clearly considered in the observer design and in the convergence study. In addition, the guidelines for selecting the observer parameters are provided. Numerical simulation confirms the stability analysis results: the observer errors converge within a short time, with a low estimation error, if observer-parameters are properly defined. [ABSTRACT FROM AUTHOR]

Flow cytometry and its technological possibilities have greatly advanced in the past decade as analysis tool for single cell properties and population distributions of different cell types in bioreactors. Along the way, some solutions for automated real‐time flow cytometry (ART‐FCM) were developed for monitoring of bioreactor processes without operator interference over extended periods with variable sampling frequency. However, there is still great potential for ART‐FCM to evolve and possibly become a standard application in bioprocess monitoring and process control. This review first addresses different components of an ART‐FCM, including the sampling device, the sample‐processing unit, the unit for sample delivery to the flow cytometer and the settings for measurement of pre‐processed samples. Also, available algorithms are presented for automated data analysis of multi‐parameter fluorescence datasets derived from ART‐FCM experiments. Furthermore, challenges are discussed for integration of fluorescence‐activated cell sorting into an ART‐FCM setup for isolation and separation of interesting subpopulations that can be further characterized by for instance omics‐methods. As the application of ART‐FCM is especially of interest for bioreactor process monitoring, including investigation of population heterogeneity and automated process control, a summary of already existing setups for these purposes is given. Additionally, the general future potential of ART‐FCM is addressed. [ABSTRACT FROM AUTHOR]

Microalgae offer a promising source of biofuel and a wide array of high-value biomolecules. Large-scale cultivation of microalgae at low density poses a significant challenge in terms of water management. High-density microalgae cultivation, however, can be challenging due to biochemical changes associated with growth dynamics. Therefore, there is a need for a biomarker that can predict the optimum density for high biomass cultivation. A locally isolated microalga Cyanobacterium aponinum CCC734 was grown with optimized nitrogen and phosphorus in the ratio of 12:1 for sustained high biomass productivity. To understand density-associated bottlenecks secretome dynamics were monitored at biomass densities from 0.6 ± 0.1 to 7 ± 0.1 g/L (2 to 22 OD) in batch mode. Liquid chromatography coupled with mass spectrometry identified 880 exometabolites in the supernatant of C. aponinum CCC734. The PCA analysis showed similarity between exometabolite profiles at low (4 and 8 OD) and mid (12 and 16 OD), whereas distinctly separate at high biomass concentrations (20 and 22 OD). Ten exometabolites were selected based on their role in influencing growth and are specifically present at low, mid, and high biomass concentrations. Taking cues from secretome dynamics, 5.0 ± 0.5 g/L biomass concentration (16 OD) was optimal for C. aponinum CCC734 cultivation. Further validation was performed with a semi-turbidostat mode of cultivation for 29 days with a volumetric productivity of 1.0 ± 0.2 g/L/day. The secretomes-based footprinting tool is the first comprehensive growth study of exometabolite at the molecular level at variable biomass densities. This tool may be utilized in analyzing and directing microalgal cultivation strategies and reduction in overall operating costs. [ABSTRACT FROM AUTHOR]

Pharmaceutical manufacturing is reliant upon bioprocessing approaches to generate the range of therapeutic products that are available today. The high cost of production, susceptibility to process failure, and requirement to achieve consistent, high-quality product means that process monitoring is paramount during manufacturing. Process analytic technologies (PAT) are key to ensuring high quality product is produced at all stages of development. Spectroscopy-based technologies are well suited as PAT approaches as they are non-destructive and require minimum sample preparation. This study explored the use of a novel attenuated total reflection Fourier transform infrared (ATR-FTIR) spectroscopy platform, which utilises disposable internal reflection elements (IREs), as a method of upstream bioprocess monitoring. The platform was used to characterise organism health and to quantify cellular metabolites in growth media using quantification models to predict glucose and lactic acid levels both singularly and combined. Separation of the healthy and nutrient deficient cells within PC space was clearly apparent, indicating this technique could be used to characterise these classes. For the metabolite quantification, the binary models yielded R 2 value was 0.980. This initial study highlights the suitability of the platform for bioprocess monitoring and paves the way for future in-line developments.R 2 value was 0.980. This initial study highlights the suitability of the platform for bioprocess monitoring and paves the way for future in-line developments., Competing Interests: Authors LC, DP, MB, and HB were employed by the Dxcover Ltd. The remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest., (Copyright © 2024 Christie, Rutherford, Palmer, Baker and Butler.)

Microalgal predation is one of the imminent threats for mass algae cultivation in open ponds. Invasion of predators results in total clearance of algal biomass within 24–48 h. Detection of contamination in microalgal cultures using online spectroscopy has attracted considerable interest among researchers. Currently reported spectral markers such as hyperspectral and multispectral tools, using visible wavelengths of light, mainly detect changes in the pigment composition or degradation associated with contaminants, especially predators. Unlike monitoring of pigment composition, our work leverages the species-discriminatory potential of Fourier transform infrared (FTIR) spectroscopy in combination with chemometric tools for detection of predators. Here we report FTIR-based signature features, at 1346, 1363, and 1382 cm−1, for detection of Oxyrrhis marina–mediated grazing in cultures of Dunaliella tertiolecta. Based on a partial least square regression (PLSR) model (R2 = 0.894), the signature spectra can indicate the presence of O. marina at a concentration of 5 × 102 cells mL−1 and at least 72 h prior to the culture crash. As opposed to offline grazer monitoring tools, the potential for FTIR-based flow-through design in combination with multivariate methods could enable real-time and non invasive means of early detection of algal grazers. [ABSTRACT FROM AUTHOR]

The aim of this study was to validate an analytical method for the simultaneous quantification of 2,3-butanediol, glycerol, acetoin, ethanol and phosphate (PO43-) by high-performance liquid chromatography (HPLC) coupled with a refractive index detector. The validation was performed according to the International Council for Harmonization of Technical Requirements for Pharmaceuticals for Human Use (ICH) - Guideline Q2 (R1). The parameters of linearity, limits of detection and quantification, precision and accuracy were evaluated. The accuracy was evaluated using standards solutions and a microbial cultivation sample. The linearity (r ≥ 0.99, n = 3) of the method was defined in the range from 0.18 to 2.52 g L−1 for phosphate (PO43-); 0.5 to 10.0 g L−1 for glycerol, acetoin and ethanol; 0.375 to 7.5 g L−1 for meso-2,3-butanediol; and 0.125 to 2.5 g L−1 for (S,S)- or (R,R)-2,3-butanediol. The limits of detection and quantification were below the concentration range used in the method. The average recovery rate, intra-day and inter-day precisions for all compounds were 98.71, 0.09 and 0.50%, respectively. The results for the accuracy ranged between 97.97 and 101.18% for all compounds in the standard solution. The method was demonstrated to be precise, specific, and reproducible for the quantification of all compounds evaluated, ensuring the reliability. Further, this method was successfully applied to samples from bioprocesses with low matrix effects and an average recovery of 99.55%. [ABSTRACT FROM AUTHOR]

The properties of the convergence region of the estimation error of a robust observer for second-order systems are determined, and a new algorithm is proposed for setting the observer parameters, considering persistent but bounded disturbances in the two observation error dynamics. The main contributions over closely related studies of the stability of state observers are: (i) the width of the convergence region of the observer error for the unknown state is expressed in terms of the interaction between the observer parameters and the disturbance terms of the observer error dynamics; (ii) it was found that this width has a minimum point and a vertical asymptote with respect to one of the observer parameters, and their coordinates were determined. In addition, the main advantages of the proposed algorithm over closely related algorithms are: (i) the definition of observer parameters is significantly simpler, as the fulfillment of Riccati equation conditions, solution of LMI constraints, and fulfillment of eigenvalue conditions are not required; (ii) unknown bounded terms are considered in the dynamics of the observer error for the known state. Finally, the algorithm is applied to a model of microalgae culture in a photobioreactor for the estimation of biomass growth rate and substrate uptake rate based on known concentrations of biomass and substrate.

Protein quantification during bioprocess monitoring is essential for biopharmaceutical manufacturing and is complicated by the complex chemical composition of the bioreactor broth. Here we present the early‐stage development and optimization of a polarized total synchronous fluorescence spectroscopy (pTSFS) method for protein quantification in a hydrolysate‐protein model (mimics clarified bioreactor broth samples) using a standard benchtop laboratory fluorometer. We used UV transmitting polarizers to provide wider range pTSFS spectra for screening of the four different TSFS spectra generated by the measurement: parallel (||), perpendicular (⊥), unpolarized (T) intensity spectra and anisotropy maps. TSFS|| (parallel polarized) measurements were the best for protein quantification compared to standard unpolarized measurements and the Bradford assay. This was because TSFS|| spectra had a better analyte signal to noise ratio (SNR), due to the anisotropy of protein emission. This meant that protein signals were better resolved from the background emission of small molecule fluorophores in the cell culture media. SNR of >5000 was achieved for concentrations of bovine serum albumin/yeastolate 1.2/10 g L–1 with TSFS||. Optimization using genetic algorithm and interval partial least squares based variable selection enabled reduction of spectral resolution and number of excitation wavelengths required without degrading performance. This enables fast (<3.5 min) online/at‐line measurements, and the method had an LOD of 0.18 g L–1 and high accuracy with a predictive error of <9%. [ABSTRACT FROM AUTHOR]

The biopharmaceutical market is dominated by monoclonal antibodies, the majority of which are produced in Chinese hamster ovary (CHO) cell lines. Intense cell engineering, in combination with optimization of various process parameters results in increasing product titers. To enable further improvements in manufacturing processes, detailed information about how certain parameters affect cellular mechanisms in the production cells, and thereby also the expressed drug substance, is required. Therefore, in this study the effects of commonly applied changes in bioprocessing parameters on an anti-IL8 IgG1 producing CHO DP-12 cell line were investigated on the level of host cell proteome expression combined with product quality assessment of the expressed IgG1 monoclonal antibody. Applying shifts in temperature, pH and dissolved oxygen concentration, respectively, resulted in altered productivity and product quality. Furthermore, analysis of the cells using two-dimensional liquid chromatography-mass spectrometry employing tandem mass tag based isotopic quantitation and synchronous precursor selection-MS3 detection revealed substantial changes in the protein expression profiles of CHO cells. Pathway analysis indicated that applied bioprocessing conditions resulted in differential activation of oxidative phosphorylation. Additionally, activation of ERK5 and TNFR1 signaling suggested an affected cell cycle. Moreover, in-depth product characterization by means of charge variant analysis, peptide mapping, as well as structural and functional analysis, revealed posttranslational and structural changes in the expressed drug substance. Taken together, the present study allows the conclusion that, in anti-IL8 IgG1 producing CHO DP-12 cells, an improved energy metabolism achieved by lowering the cell culture pH is favorable when aiming towards high antibody production rates while maintaining product quality.

• Soluble product localization is not trivial in Gram-negative E. coli. • Inner and outer membrane integrity strongly affect up- and downstream process. • We review monitoring tools for differentiating inner and outer membrane integrity. • We review factors affecting E. coli membrane integrity. • We discuss how E. coli membrane integrity can be controlled. Soluble expression of recombinant proteins in E. coli is often done by translocation of the product across the inner membrane (IM) into the periplasm, where it is retained by the outer membrane (OM). While the integrity of the IM is strongly coupled to viability and impurity release, a decrease in OM integrity (corresponding to increased "leakiness") leads to accumulation of product in the extracellular space, strongly impacting the downstream process. Whether leakiness is desired or not, differential monitoring and control of IM and OM integrity are necessary for an efficient E. coli bioprocess in compliance with the guidelines of Quality by Design and Process Analytical Technology. In this review, we give an overview of relevant monitoring tools, summarize the research on factors affecting E. coli membrane integrity and provide a brief discussion on how the available monitoring technology can be implemented in real-time control of E. coli cultivations. [ABSTRACT FROM AUTHOR]

This paper addresses the design and evaluation of a robust observer for second order bioprocesses considering unknown bounded disturbance terms and uncertainty in the dynamics of the unknown and known states. The observer design and the stability analysis are based on dead-zone Lyapunov functions, and a detailed procedure is provided. The transient response bounds and the convergence region of the unknown observer error are determined in terms of the disturbance bounds, considering persistent but bounded disturbances in the dynamics of both the known and unknown observer errors. This is a significant contribution to closely related observer design studies, in which the transient response bounds are determined, but persistent and bounded disturbances are not considered in the dynamics of the known observer error. Other important contributions are: (i) the procedure for defining the observer parameters is significantly simpler than common observer designs, since a solution to the Ricatti equation, solution to LMI constraints, or the accomplishment of eigenvalue inequality conditions are not required; (ii) discontinuous signals are not used in the observer; and (iii) the effect of the gain sign associated with the unknown state in the dynamics of the known state is explicitly and clearly considered in the observer design and in the convergence study. In addition, the guidelines for selecting the observer parameters are provided. Numerical simulation confirms the stability analysis results: the observer errors converge within a short time, with a low estimation error, if observer-parameters are properly defined.

Abstract Background A glucose concentration is an important factor for a fed-batch process of Saccharomyces cerevisiae. Therefore, it is necessary to be controlled under a critical concentration to avoid overflow metabolism and to gain high productivity of biomass. In the study, 2D fluorescence spectroscopy was applied for an online monitoring and controlling of the yeast cultivations to attain the pure oxidative metabolism. Results The characteristic of the NADH intensity can effectively identify the metabolic switch between oxidative and oxidoreductive states. Consequently, the feed rate was regulated using the single signal based on the fluorescence intensity of NADH. With this closed-loop control of the glucose concentration, a biomass yield was obtained at 0.5 g biomass/g glucose. In addition, ethanol production could be avoided during the controlled feeding phase. Conclusions The fluorescence sensor with a single signal of the NADH fluorescence intensity has potential to control a glucose concentration under the critical value in real time. Therefore, this achievement of the feedback control is promising to build up a compact and economical fluorescence sensor with the specific wavelength using light-emitting diodes and photodiodes. The sensor could be advantageous to the bioprocess monitoring because of a cost-effective and miniaturized device for routine analysis.

Real-time measurements and adjustments of critical process parameters are essential for the precise control of fermentation processes and thus for increasing both quality and yield of the desired product. However, the measurement of some crucial process parameters such as biomass, product, and product precursor concentrations usually requires time-consuming offline laboratory analysis. In this work, we demonstrate the in-line monitoring of biomass, penicillin (PEN), and phenoxyacetic acid (POX) in a Penicilliumchrysogenum fed-batch fermentation process using low-cost microspectrometer technology operating in the near-infrared (NIR). In particular, NIR reflection spectra were taken directly through the glass wall of the bioreactor, which eliminates the need for an expensive NIR immersion probe. Furthermore, the risk of contaminations in the reactor is significantly reduced, as no direct contact with the investigated medium is required. NIR spectra were acquired using two sensor modules covering the spectral ranges 1350–1650 nm and 1550–1950 nm. Based on offline reference analytics, partial least squares (PLS) regression models were established for biomass, PEN, and POX either using data from both sensors separately or jointly. The established PLS models were tested on an independent validation fed-batch experiment. Root mean squared errors of prediction (RMSEP) were 1.61 g/L, 1.66 g/L, and 0.67 g/L for biomass, PEN, and POX, respectively, which can be considered an acceptable accuracy comparable with previously published results using standard process spectrometers with immersion probes. Altogether, the presented results underpin the potential of low-cost microspectrometer technology in real-time bioprocess monitoring applications. [ABSTRACT FROM AUTHOR]

In a biotechnological process, standard monitored process variables are pH, partial oxygen pressure (pO2), and temperature. These process variables are important, but they do not give any information about the metabolic activity of the cell. The ISICOM is an in situ combi-sensor that is measuring the cell-specific oxygen uptake rate (qOUR) online. This variable allows a qualitative judgement of metabolic cell activity. The measuring principle of the ISICOM is based on a volume element enclosed into a small measuring chamber. Inside the measuring chamber, the pO2 and the scattered light is measured. Within a defined measuring interval, the chamber closes, and the oxygen supply for the cells is interrupted. The decreasing oxygen concentration is recorded by the pO2 optode. This measuring principle, known as the dynamic method, determines the oxygen uptake rate (OUR). Together with the scattered light signal, the cell concentration is estimated and the qOUR is available online. The design of the ISICOM is focused on functionality, sterility, long-term stability, and response time behavior so the sensor can be used in bioprocesses. With the ISICOM, measurement of online and in situ measurement of the OUR is possible. The OUR and qOUR online measurement of an animal cell batch cultivation is demonstrated, with maximum values of OUR = 2.5 mmol L−1 h−1 and a qOUR = 9.5 pmol cell−1 day−1. Information about limitation of the primary and secondary substrate is derived by the monitoring of the metabolic cell activity of bacteria and yeast cultivation processes. This sensor contributes to a higher process understanding by offering an online view on to the cell behavior. In the sense of process analytical technology (PAT), this important information is needed for bioprocesses to realize a knowledge base process control. [ABSTRACT FROM AUTHOR]

Compact 1H NMR and Raman spectrometers were used for real‐time process monitoring of alcoholic fermentation in a continuous flow reactor. Yeast cells catalyzing the sucrose conversion were immobilized in alginate beads floating in the reactor. The spectrometers proved to be robust and could be easily attached to the reaction apparatus. As environmentally friendly analysis methods, 1H NMR and Raman spectroscopy were selected to match the resource‐ and energy‐saving process. Analyses took only a few seconds to minutes compared to chromatographic procedures and were, therefore, suitable for real‐time control realized as a feedback loop. Both compact spectrometers were successfully implemented online. Raman spectroscopy allowed for faster spectral acquisition and higher quantitative precision, NMR yielded more resolved signals thus higher specificity. By using the software Matlab for automated data loading and processing, relevant parameters such as the ethanol, glycerol, and sugar content could be easily obtained. The subsequent multivariate data analysis using partial linear least‐squares regression type 2 enabled the quantitative monitoring of all reactants within a single model in real time. [ABSTRACT FROM AUTHOR]

• 2D Fluorescence (2DF) can provide a status fingerprint for microalgae cultures. • ML is implemented to derive cell count and pigment concentration from 2DF scans. • Spectral variable selection is shown crucial for obtaining satisfactory models. • Dataset division overfitting was successfully identified and eliminated. We propose a systematic approach for monitoring important productivity parameters in a Dunaliella salina culture using 2D fluorescence data. For this purpose, a methodology based on Machine Learning algorithm Projection to Latent Structures Regression (PLSR) coupled with variable selection strategies was used. Additionally, a robustness analysis is proposed to support the validation of the yielded models and provide a measure of their reliability. Robust (i.e., Q2 ≥ 0.5) and parsimonious (i.e., selecting down to 3 % of the fluorescence variables present in a 250–700 nm wavelength excitation-emission matrix) models were obtained for monitoring cell count, chlorophyll b , total carotenoids and β-carotene culture concentration, and the ratio between total carotenoids and total chlorophylls, all of which were validated with a left-out batch performing with R2 higher than 0.7 except for β-carotene (R2 = 0.54). [ABSTRACT FROM AUTHOR]

Raman spectroscopy is widely used in monitoring and controlling cell cultivations for biopharmaceutical drug manufacturing. However, its implementation for culture monitoring in the cell line development stage has received little attention. Therefore, the impact of clonal differences, such as productivity and growth, on the prediction accuracy and transferability of Raman calibration models is not yet well described. Raman OPLS models were developed for predicting titer, glucose and lactate using eleven CHO clones from a single cell line. These clones exhibited diverse productivity and growth rates. The calibration models were evaluated for clone-related biases using clone-wise linear regression analysis on cross validated predictions. The results revealed that clonal differences did not affect the prediction of glucose and lactate, but titer models showed a significant clone-related bias, which remained even after applying variable selection methods. The bias was associated with clonal productivity and lead to increased prediction errors when titer models were transferred to cultivations with productivity levels outside the range of their training data. The findings demonstrate the feasibility of Raman-based monitoring of glucose and lactate in cell line development with high accuracy. However, accurate titer prediction requires careful consideration of clonal characteristics during model development., (© 2023 The Authors. Biotechnology Journal published by Wiley-VCH GmbH.)

• Spectroscopic methods or unstructured biological models lead to poor state estimates. • Using PLS-based estimates as inputs to a model-based estimator leads to good results. • Experiments show the successful application of a combination of NIR and a SPKF. • Low model quality can therefore be compensated for by this hybrid approach. Typically available online measurements in cultivations do not yield enough information to build up an optimization-based control for which model-based estimation techniques are necessary. More specifically, real-time information about the concentrations of substrates and biomass is the key to controlling and optimizing cultivations in a bioreactor. In addition to established techniques such as the measurement of dissolved oxygen, off-gas composition, and the amount of spent correction fluids for pH control, inline near-infrared spectroscopy (NIR) offers the possibility of gathering information online to estimate those concentrations, without the need for manual sampling. The NIR data can be transformed via the widely applied partial least squares (PLS) regression to estimate substrate and biomass concentrations. However, the spectra are corrupted by disturbances, such as gas bubbles from sparging, that can result in noisy and defective spectra and thus may lead to highly imprecise measurements. For feedback control, such imprecisions can lead to wrong decisions with respect to feeding rates, which may be hard to compensate for at a later stage. Moreover, information about operating conditions and physical constraints, such as feeding profiles and kinetic relations between substrates that could be used to rectify erroneous predictions, are disregarded in the PLS model. In this paper, we therefore present a hybrid approach of NIR spectroscopy and nonlinear, model-based state estimation to enable improved quality in the online estimation of substrates and biomass in aerated yeast cultivations. The feeding rates, off-gas concentrations, NIR spectra, and knowledge about the dynamics of the process are integrated seamlessly in a sigma-point Kalman filter (SPKF). The latter is formulated such that it is able to handle physical boundaries, which is especially important for fed-batch cultivations. In order to evaluate the hybrid approach, it is compared to different scenarios typically applied in fermentations. One uses only NIR data with a PLS regression model. The other works with an SPKF, but without NIR information. It is shown that the presented hybrid method outperforms the methods applied separately for the estimation of biomass, glucose, ethanol, ammonium, and phosphate in cultivations of a new, genetically modified strain of Saccharomyces cerevisiae. As only a few cultivations are needed for calibration, a reliable online state estimation is available in the early stage of development of this process. [ABSTRACT FROM AUTHOR]

This study was performed in order to evaluate a new LED‐based 2D‐fluorescence spectrometer for in‐line bioprocess monitoring of Chinese hamster ovary (CHO) cell culture processes. The new spectrometer used selected excitation wavelengths of 280, 365, and 455 nm to collect spectral data from six 10‐L fed‐batch processes. The technique provides data on various fluorescent compounds from the cultivation medium as well as from cell metabolism. In addition, scattered light offers information about the cultivation status. Multivariate data analysis tools were applied to analyze the large data sets of the collected fluorescence spectra. First, principal component analysis was used to accomplish an overview of all spectral data from all six CHO cultivations. Partial least square regression models were developed to correlate 2D‐fluorescence spectral data with selected critical process variables as offline reference values. A separate independent fed‐batch process was used for model validation and prediction. An almost continuous in‐line bioprocess monitoring was realized because 2D‐fluorescence spectra were collected every 10 min during the whole cultivation. The new 2D‐fluorescence device demonstrates the significant potential for accurate prediction of the total cell count, viable cell count, and the cell viability. The results strongly indicated that the technique is particularly capable to distinguish between different cell statuses inside the bioreactor. In addition, spectral data provided information about the lactate metabolism shift and cellular respiration during the cultivation process. Overall, the 2D‐fluorescence device is a highly sensitive tool for process analytical technology applications in mammalian cell cultures. [ABSTRACT FROM AUTHOR]

[Display omitted] • A rapid and efficient technology for on-line monitoring kombucha is developed. • Near-infrared spectroscopy is used as a monitoring tool for total sugars in kombucha. • Linear and nonlinear calibration methods are combined and compared. • The combined model can be used to monitor the total sugar. Kombucha is widely recognized for its health benefits, and it facilitates high-quality transformation and utilization of tea during the fermentation process. Implementing on-line monitoring for the kombucha production process is crucial to promote the valuable utilization of low-quality tea residue. Near-infrared (NIR) spectroscopy, together with partial least squares (PLS), backpropagation neural network (BPANN), and their combination (PLS-BPANN), were utilized in this study to monitor the total sugar of kombucha. In all, 16 mathematical models were constructed and assessed. The results demonstrate that the PLS-BPANN model is superior to all others, with a determination coefficient (R2p) of 0.9437 and a root mean square error of prediction (RMSEP) of 0.8600 g/L and a good verification effect. The results suggest that NIR coupled with PLS-BPANN can be used as a non-destructive and on-line technique to monitor total sugar changes. [ABSTRACT FROM AUTHOR]

On-line monitoring of key growth indicators is essential to regulate the kombucha fermentation process and improve utilization of summer-autumn tea. This study developed an on-line visible/near-infrared spectroscopy (V-NIR) detection system and method to monitor the concentration of soluble sugars, total acids, and bacteria during kombucha fermentation. The partial least squares method (PLS) was coupled with the sparrow search algorithm (SSA), whale optimization algorithm (WOA), and African vulture optimization algorithm (AVOA) to select characteristic variables, extract high-frequency variables, and establish calibration models. In general, the model performance of the high-frequency variable subset was better than that of the selected variable subset. The correlation coefficients of the optimal model of the six indexes in the two stages were 0.8692, 0.8513, 0.9392, 0.9746, 0.9570, and 0.9830, respectively. The developed monitoring system and modelling strategy can be applied to the real-time online determination of kombucha fermentation and further applied to the food industry. [Display omitted] • An on-line visible/near-infrared spectroscopy monitoring system is developed. • The key growth indicators are monitored during kombucha fermentation. • The swarm intelligence algorithms are used to select feature variables. • The high-frequency variables are extracted in selected feature variable subsets. [ABSTRACT FROM AUTHOR]