Cell culture product quality attribute prediction using convolutional neural networks and Raman spectroscopy

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.

On-line targeted metabolomics for real-time monitoring of relevant compounds in fermentation processes

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.

How Not to Make the Joint Extended Kalman Filter Fail with Unstructured Mechanistic Models

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.

Feasibility and performance of cross-clone Raman calibration models in CHO cultivation

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.

A Simplified Algorithm for Setting the Observer Parameters for Second-Order Systems with Persistent Disturbances Using a Robust Observer

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.

Capillary and microchip electrophoresis method development for amino acid monitoring during biopharmaceutical cultivation

The increased use of biopharmaceuticals calls for improved means of bioprocess monitoring. In this work, capillary electrophoresis (CE) and microchip electrophoresis (MCE) methods were developed and applied for the analysis of amino acids (AAs) in cell culture supernatant. In samples from different days of a Chinese hamster ovary cell cultivation process, all 19 proteinogenic AAs containing primary amine groups could be detected using CE, and 17 out of 19 AAs using MCE. The relative concentration changes in different samples agreed well with those measured by high-performance liquid chromatography (HPLC). Compared to the more commonly employed HPLC analysis, the CE and MCE methods resulted in faster analysis, while significantly lowering both the sample and reagent consumption, and the cost per analysis. This article is protected by copyright. All rights reserved.

Advances in automated real-time flow cytometry for monitoring of bioreactor processes

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.

Continuous optical in-line glucose monitoring and control in CHO cultures contributes to enhanced metabolic efficiency while maintaining darbepoetin alfa product quality

Great efforts are directed towards improving productivity, consistency and quality of biopharmaceutical processes and products. One particular area is the development of new sensors for continuous monitoring of critical bioprocess parameters by using online or in-line monitoring systems. Recently, we developed a glucose biosensor applicable in single-use, in-line and long-term glucose monitoring in mammalian cell bioreactors. Now, we integrated this sensor in an automated glucose monitoring and feeding system capable of maintaining stable glucose levels, even at very low concentrations. We compared this fed-batch feedback system at both low (< 1 mM) and high (40 mM) glucose levels with traditional batch culture methods, focusing on glycosylation and glycation of the recombinant protein darbepoetin alfa (DPO) produced by a CHO cell line. We evaluated cell growth, metabolite and product concentration under different glucose feeding strategies and show that continuous feeding, even at low glucose levels, has no harmful effects on DPO quantity and quality. We conclude that our system is capable of tight glucose level control throughout extended bioprocesses and has the potential to improve performance where constant maintenance of glucose levels is critical.

Proteomic Profiling of IgG1 Producing CHO Cells Using LC/LC-SPS-MS3: The Effects of Bioprocessing Conditions on Productivity and Product Quality

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.

Development of a rapid polarized total synchronous fluorescence spectroscopy (pTSFS) method for protein quantification in a model bioreactor broth

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%.

A Nuclear Magnetic Resonance (NMR) Platform for Real-Time Metabolic Monitoring of Bioprocesses

We present a Nuclear Magnetic Resonance (NMR) compatible platform for the automated real-time monitoring of biochemical reactions using a flow shuttling configuration. This platform requires a working sample volume of ∼11 mL and it can circulate samples with a flow rate of 28 mL/min, which makes it suitable to be used for real-time monitoring of biochemical reactions. Another advantage of the proposed low-cost platform is the high spectral resolution. As a proof of concept, we acquire 1H NMR spectra of waste orange peel, bioprocessed using Trichoderma reesei fungus, and demonstrate the real-time measurement capability of the platform. The measurement is performed over more than 60 h, with a spectrum acquired every 7 min, such that over 510 data points are collected without user intervention. The designed system offers high resolution, automation, low user intervention, and, therefore, time-efficient measurement per sample.

Recent Progress in Lab-On-a-Chip Systems for the Monitoring of Metabolites for Mammalian and Microbial Cell Research

Lab-on-a-chip sensing technologies have changed how cell biology research is conducted. This review summarises the progress in the lab-on-a-chip devices implemented for the detection of cellular metabolites. The review is divided into two subsections according to the methods used for the metabolite detection. Each section includes a table which summarises the relevant literature and also elaborates the advantages of, and the challenges faced with that particular method. The review continues with a section discussing the achievements attained due to using lab-on-a-chip devices within the specific context. Finally, a concluding section summarises what is to be resolved and discusses the future perspectives.

Application of green analytical chemistry to a green chemistry process: Magnetic resonance and Raman spectroscopic process monitoring of continuous ethanolic fermentation

Compact ¹ H 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, ¹ H 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.

A novel LED-based 2D-fluorescence spectroscopy system for in-line bioprocess monitoring of Chinese hamster ovary cell cultivations-Part II

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.

Extraction, Enrichment, and in situ Electrochemical Detection on Lab-on-a-Disc: Monitoring the Production of a Bacterial Secondary Metabolite

Development of microsystems, which enable "sample-to-answer" detection from real samples, is often challenging. We present the first integration of supported liquid membrane extraction combined with electrochemical detection on a centrifugal fluidic platform. The developed lab-on-a-disc (LoD) system enabled the separation, enrichment, and subsequent electrochemical detection of the target analyte from a complex sample mixture. As a case study, we quantified the amount of a dietary supplement and pharmaceutical precursor, p-coumaric acid, from bacterial growth media at different time points during production. The assay, extraction, and detection, performed on the LoD device, proved to be a low cost and environmentally friendly approach, requiring only a few tens of microliters of organic solvent and enabled detection in a 3 μL volume. In addition, the data obtained from the centrifugal platform showed a good correlation with data obtained from the high performance liquid chromatography analysis.

Media photo-degradation in pharmaceutical biotechnology - impact of ambient light on media quality, cell physiology, and IgG production in CHO cultures

BACKGROUND

Many vital components in bioprocess media are prone to photo-conversion or photo-degradation upon exposure to ambient light, with severe negative consequences for biomass yield and overall productivity. However, there is only limited awareness of light irradiation as a potential risk factor when working in transparent glass bioreactors, storage vessels or disposable bag systems. The chemical complexity of most media renders a root-cause analysis difficult. This study investigated in a novel, holistic approach how light-induced changes in media composition relate to alterations in radical burden, cell physiology, morphology, and product formation in industrial Chinese hamster ovary (CHO) bioprocesses.

RESULTS

Two media formulations from proprietary and commercial sources were tested in a pre-hoc light exposure scenario prior to cultivation. Using fluorescence excitation/emission (EEM) matrix spectroscopy, a photo-sensitization of riboflavin was identified as a likely cause for drastically decreased IgG titers (up to -80%) and specific growth rates (-50% to -90%). Up to three-fold higher radical levels were observed in photo-degraded medium. On the biological side, this resulted in significant changes in cell morphology and aberrations in the normal IgG biosynthesis/secretion pathway.

CONCLUSION

These findings clearly illustrate the underrated impact of room light after only short periods of exposure, occurring accidentally or knowingly during bioprocess development and scale- up. The detrimental effects, which may share a common mechanistic cause at the molecular level, correlate well with changes in spectroscopic properties. This offers new perspectives for online monitoring concepts, and improved detectability of such effects in future. © 2018 The Authors. Journal of Chemical Technology & Biotechnology published by JohnWiley & Sons Ltd on behalf of Society of Chemical Industry.

Determining Biogenic Content of Biogas by Measuring Stable Isotopologues 12CH₄, 13CH₄, and CH₃D with a Mid-Infrared Direct Absorption Laser Spectrometer

A tunable laser absorption spectrometer (TLAS) was developed for the simultaneous measurement of δ¹³C and δD values of methane (CH₄). A mid-infrared interband cascade laser (ICL) emitting around 3.27 µm was used to measure the absorption of the three most abundant isotopologues in CH₄ with a single, mode-hop free current sweep. The instrument was validated against methane samples of fossil and biogenic origin with known isotopic composition. Three blended mixtures with varied biogenic content were prepared volumetrically, and their δ¹³C and δD values were determined. Analysis demonstrated that, provided the isotopic composition of the source materials was known, the δ¹³C and δD values alone were sufficient to determine the biogenic content of the blended samples to within 1.5%.

Developing global regression models for metabolite concentration prediction regardless of cell line

Following the Process Analytical Technology (PAT) of the Food and Drug Administration (FDA), drug manufacturers are encouraged to develop innovative techniques in order to monitor and understand their processes in a better way. Within this framework, it has been demonstrated that Raman spectroscopy coupled with chemometric tools allow to predict critical parameters of mammalian cell cultures in-line and in real time. However, the development of robust and predictive regression models clearly requires many batches in order to take into account inter-batch variability and enhance models accuracy. Nevertheless, this heavy procedure has to be repeated for every new line of cell culture involving many resources. This is why we propose in this paper to develop global regression models taking into account different cell lines. Such models are finally transferred to any culture of the cells involved. This article first demonstrates the feasibility of developing regression models, not only for mammalian cell lines (CHO and HeLa cell cultures), but also for insect cell lines (Sf9 cell cultures). Then global regression models are generated, based on CHO cells, HeLa cells, and Sf9 cells. Finally, these models are evaluated considering a fourth cell line(HEK cells). In addition to suitable predictions of glucose and lactate concentration of HEK cell cultures, we expose that by adding a single HEK-cell culture to the calibration set, the predictive ability of the regression models are substantially increased. In this way, we demonstrate that using global models, it is not necessary to consider many cultures of a new cell line in order to obtain accurate models. Biotechnol. Bioeng. 2017;114: 2550-2559. © 2017 Wiley Periodicals, Inc.

Estimating Extrinsic Dyes for Fluorometric Online Monitoring of Antibody Aggregation in CHO Fed-Batch Cultivations

Multi-wavelength fluorescence spectroscopy was evaluated in this work as tool for real-time monitoring of antibody aggregation in CHO fed-batch cultivations via partial least square (PLS) modeling. Therefore, we used the extrinsic fluorescence dyes 1-anilinonaphthalene-8-sulfonate (ANS), 4,4′-bis-1-anilinonaphthalene-8-sulfonate (Bis-ANS), or Thioflavin T (ThT) as medium additives. This is a new application area, since these dyes are commonly used for aggregate detection during formulation development. We determined the half maximum inhibitory concentrations of ANS (203 ± 11 µmol·L⁻¹), Bis-ANS (5 ± 0.5 µmol·L⁻¹), and ThT (3 ± 0.2 µmol·L⁻¹), and selected suitable concentrations for this application. The results showed that the emission signals of non-covalent dye antibody aggregate interaction superimposed the fluorescence signals originating from feed medium and cell culture. The fluorescence datasets were subsequently used to build PLS models, and the dye-related elevated fluorescence signals dominated the model calibration. The soft sensors based on ANS and Bis-ANS signals showed high predictability with a low error of prediction (1.7 and 2.3 mg·mL⁻¹ aggregates). In general, the combination of extrinsic dye and used concentration influenced the predictability. Furthermore, the ThT soft sensor indicated that the intrinsic fluorescence of the culture might be sufficient to predict antibody aggregation online.

Does Nonlinear Modeling Play a Role in Plasmid Bioprocess Monitoring Using Fourier Transform Infrared Spectra?

The monitoring of biopharmaceutical products using Fourier transform infrared (FT-IR) spectroscopy relies on calibration techniques involving the acquisition of spectra of bioprocess samples along the process. The most commonly used method for that purpose is partial least squares (PLS) regression, under the assumption that a linear model is valid. Despite being successful in the presence of small nonlinearities, linear methods may fail in the presence of strong nonlinearities. This paper studies the potential usefulness of nonlinear regression methods for predicting, from in situ near-infrared (NIR) and mid-infrared (MIR) spectra acquired in high-throughput mode, biomass and plasmid concentrations in Escherichia coli DH5-α cultures producing the plasmid model pVAX-LacZ. The linear methods PLS and ridge regression (RR) are compared with their kernel (nonlinear) versions, kPLS and kRR, as well as with the (also nonlinear) relevance vector machine (RVM) and Gaussian process regression (GPR). For the systems studied, RR provided better predictive performances compared to the remaining methods. Moreover, the results point to further investigation based on larger data sets whenever differences in predictive accuracy between a linear method and its kernelized version could not be found. The use of nonlinear methods, however, shall be judged regarding the additional computational cost required to tune their additional parameters, especially when the less computationally demanding linear methods herein studied are able to successfully monitor the variables under study.

Mammalian cell culture monitoring using in situ spectroscopy: Is your method really optimised?

In recent years, as a result of the process analytical technology initiative of the US Food and Drug Administration, many different works have been carried out on direct and in situ monitoring of critical parameters for mammalian cell cultures by Raman spectroscopy and multivariate regression techniques. However, despite interesting results, it cannot be said that the proposed monitoring strategies, which will reduce errors of the regression models and thus confidence limits of the predictions, are really optimized. Hence, the aim of this article is to optimize some critical steps of spectroscopic acquisition and data treatment in order to reach a higher level of accuracy and robustness of bioprocess monitoring. In this way, we propose first an original strategy to assess the most suited Raman acquisition time for the processes involved. In a second part, we demonstrate the importance of the interbatch variability on the accuracy of the predictive models with a particular focus on the optical probes adjustment. Finally, we propose a methodology for the optimization of the spectral variables selection in order to decrease prediction errors of multivariate regressions. © 2017 American Institute of Chemical Engineers Biotechnol. Prog., 33:308-316, 2017.

Observability analysis of biochemical process models as a valuable tool for the development of mechanistic soft sensors

By enabling the estimation of difficult-to-measure target variables using available indirect measurements, mechanistic soft sensors have become important tools for various bioprocess monitoring and control scenarios. Despite promising higher process efficiencies and increased process understanding, widespread application of soft sensors has been stalled by uncertainty about the feasibility and reliability of their estimations given present process analytical constraints. Observability analysis can provide an indication of the possibility and reliability of soft sensor estimations by analyzing the structural properties of first-principle (mechanistic) models. In addition, it can provide a criteria for selection of suitable measurement methods with respect to their information content; thereby leading to successful implementation of soft sensors in bioprocess development and manufacturing environments. We demonstrate the utility of observability analysis for two classes of upstream bioprocesses: the processes involving growth and ethanol formation by Saccharomyces cerevisiae and the process of penicillin production by Penicillium chrysogenum. Results obtained from laboratory-scale cultivations in addition to in-silico experiments enable a comparison of theoretical aspects of observability analysis and the real-life performance of soft sensors. By taking the expected error of measurements provided to the soft sensor into account, an innovative scaling approach facilitates a higher degree of comparability of observability results among various measurement configurations and process conditions.

A tool for selective inline quantification of co-eluting proteins in chromatography using spectral analysis and partial least squares regression

Selective quantification of co-eluting proteins in chromatography is usually performed by offline analytics. This is time-consuming and can lead to late detection of irregularities in chromatography processes. To overcome this analytical bottleneck, a methodology for selective protein quantification in multicomponent mixtures by means of spectral data and partial least squares regression was presented in two previous studies. In this paper, a powerful integration of software and chromatography hardware will be introduced that enables the applicability of this methodology for a selective inline quantification of co-eluting proteins in chromatography. A specific setup consisting of a conventional liquid chromatography system, a diode array detector, and a software interface to Matlab® was developed. The established tool for selective inline quantification was successfully applied for a peak deconvolution of a co-eluting ternary protein mixture consisting of lysozyme, ribonuclease A, and cytochrome c on SP Sepharose FF. Compared to common offline analytics based on collected fractions, no loss of information regarding the retention volumes and peak flanks was observed. A comparison between the mass balances of both analytical methods showed, that the inline quantification tool can be applied for a rapid determination of pool yields. Finally, the achieved inline peak deconvolution was successfully applied to make product purity-based real-time pooling decisions. This makes the established tool for selective inline quantification a valuable approach for inline monitoring and control of chromatographic purification steps and just in time reaction on process irregularities.