Bioprocess systems engineering: transferring traditional process engineering principles to industrial biotechnology

Mask region-based CNNs for cervical cancer progression diagnosis on pap smear examinations

This research presents a novel approach for cervical cancer detection and segmentation using tissue images with multiple cells. The study employs a novel deep learning architecture based on Mask Region-Based Convolutional Neural Network (RCNN) and statistical analysis. This new architecture enables us to achieve a high percentage of detection and pix-to-pix area segmentation. A mean Average Precision (mAP) higher than 60% for 3-class and 5-class was achieved. In addition, higher F1-scores of 70% for 3-class and 5-class were obtained. This investigation is a collaborative work, where a medical consultant collected the samples from the Papanicolaou (Pap) Smear examination and labeled the cells presented to the liquid-based cytology (LBC). Furthermore, the online available benchmark data set, SIPaKMeD, was also utilized. Additionally, sample images from the Mendeley data set were also labeled by the trained medical consultant for comparison. The proposed scheme automatically generates a full report for a medical consultant to identify the location of the malicious cells in the given images and expedite the diagnosis and treatment process.

Population balance modelling captures host cell protein dynamics in CHO cell cultures

Monoclonal antibodies (mAbs) have been extensively studied for their wide therapeutic and research applications. Increases in mAb titre has been achieved mainly by cell culture media/feed improvement and cell line engineering to increase cell density and specific mAb productivity. However, this improvement has shifted the bottleneck to downstream purification steps. The higher accumulation of the main cell-derived impurities, host cell proteins (HCPs), in the supernatant can negatively affect product integrity and immunogenicity in addition to increasing the cost of capture and polishing steps. Mathematical modelling of bioprocess dynamics is a valuable tool to improve industrial production at fast rate and low cost. Herein, a single stage volume-based population balance model (PBM) has been built to capture Chinese hamster ovary (CHO) cell behaviour in fed-batch bioreactors. Using cell volume as the internal variable, the model captures the dynamics of mAb and HCP accumulation extracellularly under physiological and mild hypothermic culture conditions. Model-based analysis and orthogonal measurements of lactate dehydrogenase activity and double-stranded DNA concentration in the supernatant show that a significant proportion of HCPs found in the extracellular matrix is secreted by viable cells. The PBM then served as a platform for generating operating strategies that optimise antibody titre and increase cost-efficiency while minimising impurity levels.

Sustainable Design Approach for Modeling Bioprocesses from Laboratory toward Commercialization: Optimizing Chitosan Production

Enhancing the biochemical supply chain towards sustainable development requires more efforts to boost technology innovation at early design phases and avoid delays in industrial biotechnology growth. Such a transformation requires a comprehensive step-wise procedure to guide bioprocess development from laboratory protocols to commercialization. This study introduces a process design framework to guide research and development (R&D) through this journey, bearing in mind the particular challenges of bioprocess modeling. The method combines sustainability assessment and process optimization based on process efficiency indicators, technical indicators, Life Cycle Assessment (LCA), and process optimization via Water Regeneration Networks (WRN). Since many bioprocesses remain at low Technology Readiness Levels (TRLs), the process simulation module was examined in detail to account for uncertainties, providing strategies for successful guidance. The sustainability assessment was performed using the geometric mean-based sustainability footprint metric. A case study based on Chitosan production from shrimp exoskeletons was evaluated to demonstrate the method’s applicability and its advantages in product optimization. An optimized scenario was generated through a WRN to improve water management, then compared with the case study. The results confirm the existence of a possible configuration with better sustainability performance for the optimized case with a sustainability footprint of 0.33, compared with the performance of the base case (1.00).

Modeling the time course of ComX: towards molecular process control for Bacillus wild-type cultivations

Wild-type cultivations are of invaluable relevance for industrial biotechnology when it comes to the agricultural or food sector. Here, genetic engineering is hardly applicable due to legal barriers and consumer’s demand for GMO-free products. An important pillar for wild-type cultivations displays the genus Bacillus. One of the challenges for Bacillus cultivations is the global ComX-dependent quorum sensing system. Here, molecular process control can serve as a tool to optimize the production process without genetic engineering. To realize this approach, quantitative knowledge of the mechanism is essential, which, however, is often available only to a limited extent. The presented work provides a case study based on the production of cyclic lipopeptide surfactin, whose expression is in dependence of ComX, using natural producer B. subtilis DSM 10 ^T. First, a surfactin reference process with 40 g/L of glucose was performed as batch fermentation in a pilot scale bioreactor system to gain novel insights into kinetic behavior of ComX in relation to surfactin production. Interestingly, the specific surfactin productivity did not increase linearly with ComX activity. The data were then used to derive a mathematic model for the time course of ComX in dependence of existing biomass, biomass growth as well as a putative ComX-specific protease. The newly adapted model was validated and transferred to other batch fermentations, employing 20 and 60 g/L glucose. The applied approach can serve as a model system for molecular process control strategies, which can thus be extended to other quorum sensing dependent wild-type cultivations.

Modeling of Continuous PHA Production by a Hybrid Approach Based on First Principles and Machine Learning

Polyhydroxyalkanoates (PHA) are renewable alternatives to traditional oil-derived polymers. PHA can be produced by different microorganisms in continuous culture under specific media composition, which makes the production process both promising and challenging. In order to achieve large productivities while maintaining high yield and efficiency, the continuous culture needs to be operated in the so-called dual nutrient limitation condition, where both the nitrogen and carbon sources are kept at very low concentrations. Mathematical models can greatly assist both design and operation of the bioprocess, but are challenged by the complexity of the system, in particular by the dual nutrient-limited growth phenomenon, where the cells undergo a metabolic shift that abruptly changes their behavior. Traditional, non-structured mechanistic models based on Monod uptake kinetics can be used to describe the bioreactor operation under specific process conditions. However, in the absence of a model description of the metabolic phenomena inside the cell, the extrapolation to a broader operation domain (e.g., different feeding concentrations and dilution rates) may present mismatches between the predictions and the actual process outcomes. Such detailed models may require almost perfect knowledge of the cell metabolism and omic-level measurements, hampering their development. On the other hand, purely data-driven models that learn correlations from experimental data do not require any prior knowledge of the process and are therefore unbiased and flexible. However, many more data are required for their development and their extrapolation ability is limited to conditions that are similar to the ones used for training. An attractive alternative is the combination of the extrapolation power of first principles knowledge with the flexibility of machine learning methods. This approach results in a hybrid model for the growth and uptake rates that can be used to predict the dynamic operation of the bioreactor. Here we develop a hybrid model to describe the continuous production of PHA by Pseudomonas putida GPo1 culture. After training, the model with experimental data gained under different dilution rates and medium compositions, we demonstrate how the model can describe the process in a wide range of operating conditions, including both single and dual nutrient-limited growth.

Bioreactor control systems in the biopharmaceutical industry: a critical perspective

Industrial-scale bioprocessing underpins much of the production of pharmaceuticals, nutraceuticals, food, and beverage processing industries of the modern world. The profitability of these processes increasingly leverages the economies of scale and scope that are critically dependent on the product yields, titers, and productivity. Most of the processes are controlled using classical control approaches and represent over 90% of the industrial controls used in bioprocessing industries. However, with the advances in the production processes, especially in the biopharmaceutical and nutraceutical industries, monitoring and control of bioprocesses such as fermentations with GMO organisms, and downstream processing has become increasingly complex and the inadequacies of the classical and some of the modern control systems techniques is becoming apparent. Therefore, with increasing research complexity, nonlinearity, and digitization in process, there has been a critical need for advanced process control that is more effective, and easier process intensification and product yield (both by quality and quantity) can be achieved. In this review, industrial aspects of a process and automation along with various commercial control strategies have been extensively discussed to give an insight into the future prospects of industrial development and possible new strategies for process control and automation with a special focus on the biopharmaceutical industry.

Supplementary Information

The online version contains supplementary material available at 10.1007/s43393-021-00048-6.

Bio-Electrochemical System Depollution Capabilities and Monitoring Applications: Models, Applicability, Advanced Bio-Based Concept for Predicting Pollutant Degradation and Microbial Growth Kinetics via Gene Regulation Modelling

Microbial fuel cells (MFC) are an emerging technology for waste, wastewater and polluted soil treatment. In this manuscript, pollutants that can be treated using MFC systems producing energy are presented. Furthermore, the applicability of MFC in environmental monitoring is described. Common microbial species used, release of genome sequences, and gene regulation mechanisms, are discussed. However, although scaling-up is the key to improving MFC systems, it is still a difficult challenge. Mathematical models for MFCs are used for their design, control and optimization. Such models representing the system are presented here. In such comprehensive models, microbial growth kinetic approaches are essential to designing and predicting a biosystem. The empirical and unstructured Monod and Monod-type models, which are traditionally used, are also described here. Understanding and modelling of the gene regulatory network could be a solution for enhancing knowledge and designing more efficient MFC processes, useful for scaling it up. An advanced bio-based modelling concept connecting gene regulation modelling of specific metabolic pathways to microbial growth kinetic models is presented here; it enables a more accurate prediction and estimation of substrate biodegradation, microbial growth kinetics, and necessary gene and enzyme expression. The gene and enzyme expression prediction can also be used in synthetic and systems biology for process optimization. Moreover, various MFC applications as a bioreactor and bioremediator, and in soil pollutant removal and monitoring, are explored.

About Model Validation in Bioprocessing

In bioprocess engineering the Qualtiy by Design (QbD) initiative encourages the use of models to define design spaces. However, clear guidelines on how models for QbD are validated are still missing. In this review we provide a comprehensive overview of the validation methods, mathematical approaches, and metrics currently applied in bioprocess modeling. The methods cover analytics for data used for modeling, model training and selection, measures for predictiveness, and model uncertainties. We point out the general issues in model validation and calibration for different types of models and put this into the context of existing health authority recommendations. This review provides a starting point for developing a guide for model validation approaches. There is no one-fits-all approach, but this review should help to identify the best fitting validation method, or combination of methods, for the specific task and the type of bioprocess model that is being developed.

Potential of Integrating Model-Based Design of Experiments Approaches and Process Analytical Technologies for Bioprocess Scale-Down

Typically, bioprocesses on an industrial scale are dynamic systems with a certain degree of variability, system inhomogeneities, and even population heterogeneities. Therefore, the scaling of such processes from laboratory to industrial scale and vice versa is not a trivial task. Traditional scale-down methodologies consider several technical parameters, so that systems on the laboratory scale tend to qualitatively reflect large-scale effects, but not the dynamic situation in an industrial bioreactor over the entire process, from the perspective of a cell. Supported by the enormous increase in computing power, the latest scientific focus is on the application of dynamic models, in combination with computational fluid dynamics to quantitatively describe cell behavior. These models allow the description of possible cellular lifelines which in turn can be used to derive a regime analysis for scale-down experiments. However, the approaches described so far, which were for a very few process examples, are very labor- and time-intensive and cannot be validated easily. In parallel, alternatives have been developed based on the description of the industrial process with hybrid process models, which describe a process mechanistically as far as possible in order to determine the essential process parameters with their respective variances. On-line analytical methods allow the characterization of population heterogeneity directly in the process. This detailed information from the industrial process can be used in laboratory screening systems to select relevant conditions in which the cell and process related parameters reflect the situation in the industrial scale. In our opinion, these technologies, which are available in research for modeling biological systems, in combination with process analytical techniques are so far developed that they can be implemented in industrial routines for faster development of new processes and optimization of existing ones.

Model-Based Optimization of a Fed-Batch Bioreactor for mAb Production Using a Hybridoma Cell Culture

Production of monoclonal antibodies (mAbs) is a well-known method used to synthesize a large number of identical antibodies, which are molecules of huge importance in medicine. Due to such reasons, intense efforts have been invested to maximize the mAbs production in bioreactors with hybridoma cell cultures. However, the optimal control of such sensitive bioreactors is an engineering problem difficult to solve due to the large number of state-variables with highly nonlinear dynamics, which often translates into a non-convex optimization problem that involves a significant number of decision (control) variables. Based on an adequate kinetic model adopted from the literature, this paper focuses on developing an in-silico (model-based, offline) numerical analysis of a fed-batch bioreactor (FBR) with an immobilized hybridoma culture to determine its optimal feeding policy by considering a small number of control variables, thus ensuring maximization of mAbs production. The obtained time stepwise optimal feeding policies of FBR were proven to obtain better performances than those of simple batch operation (BR) for all the verified alternatives in terms of raw material consumption and mAbs productivity. Several elements of novelty (i–iv) are pointed out in the “conclusions” section (e.g., considering the continuously added biomass as a control variable during FBR).

Machine learning for metabolic engineering: A review

Machine learning provides researchers a unique opportunity to make metabolic engineering more predictable. In this review, we offer an introduction to this discipline in terms that are relatable to metabolic engineers, as well as providing in-depth illustrative examples leveraging omics data and improving production. We also include practical advice for the practitioner in terms of data management, algorithm libraries, computational resources, and important non-technical issues. A variety of applications ranging from pathway construction and optimization, to genetic editing optimization, cell factory testing, and production scale-up are discussed. Moreover, the promising relationship between machine learning and mechanistic models is thoroughly reviewed. Finally, the future perspectives and most promising directions for this combination of disciplines are examined.

Aqueous Two-Phase Systems at Large Scale: Challenges and Opportunities

Aqueous two-phase systems (ATPS) have proved to be an efficient and integrative operation to enhance recovery of industrially relevant bioproducts. After ATPS discovery, a variety of works have been published regarding their scaling from 10 to 1000 L. Although ATPS have achieved high recovery and purity yields, there is still a gap between their bench-scale use and potential industrial applications. In this context, this review paper critically analyzes ATPS scale-up strategies to enhance the potential industrial adoption. In particular, large-scale operation considerations, different phase separation procedures, the available optimization techniques (univariate, response surface methodology, and genetic algorithms) to maximize recovery and purity and economic modeling to predict large-scale costs, are discussed. ATPS intensification to increase the amount of sample to process at each system, developing recycling strategies and creating highly efficient predictive models, are still areas of great significance that can be further exploited with the use of high-throughput techniques. Moreover, the development of novel ATPS can maximize their specificity increasing the possibilities for the future industry adoption of ATPS. This review work attempts to present the areas of opportunity to increase ATPS attractiveness at industrial levels.

Controlled gas exchange in whole lung bioreactors

In cellular, tissue-level or whole organ bioreactors, the level of dissolved oxygen is one of the most important factors requiring control. Hypoxic environments may lead to cellular apoptosis, while hyperoxic environments may lead to cellular damage or dedifferentiation, both resulting in loss of overall tissue function. This manuscript describes the creation, characterization and validation of a bioreactor system that can control oxygen delivery based on real-time metabolic demand of cultured whole lung tissue. A mathematical model describing and predicting gas exchange within the tunable bioreactor system is developed. In addition, the inherent gas exchange properties of the bioreactor and the inherent oxygen consumption rates of native rat lungs are determined, thereby providing a quantitative relationship between system parameters and levels of dissolved oxygen. Finally, the mathematical model is validated during whole lung culture under a range of system parameters. The system presented here provides a quantitative relationship between the concentration of dissolved oxygen, tissue oxygen consumption rates, and controllable system parameters that introduce gasses into the bioreactor. This relationship not only enables the maintenance of constant levels of dissolved oxygen throughout a culture period during which cells are replicating, but also provides noninvasive and real-time estimation of the metabolic and proliferative states of native or engineered lung tissue simply through dissolved oxygen measurements. Copyright © 2017 John Wiley & Sons, Ltd.

Towards stable kinetics of large metabolic networks: Nonequilibrium potential function approach

While the biochemistry of metabolism in many organisms is well studied, details of the metabolic dynamics are not fully explored yet. Acquiring adequate in vivo kinetic parameters experimentally has always been an obstacle. Unless the parameters of a vast number of enzyme-catalyzed reactions happened to fall into very special ranges, a kinetic model for a large metabolic network would fail to reach a steady state. In this work we show that a stable metabolic network can be systematically established via a biologically motivated regulatory process. The regulation is constructed in terms of a potential landscape description of stochastic and nongradient systems. The constructed process draws enzymatic parameters towards stable metabolism by reducing the change in the Lyapunov function tied to the stochastic fluctuations. Biologically it can be viewed as interplay between the flux balance and the spread of workloads on the network. Our approach allows further constraints such as thermodynamics and optimal efficiency. We choose the central metabolism of Methylobacterium extorquens AM1 as a case study to demonstrate the effectiveness of the approach. Growth efficiency on carbon conversion rate versus cell viability and futile cycles is investigated in depth.

Economic analysis of uricase production under uncertainty: Contrast of chromatographic purification and aqueous two-phase extraction (with and without PEG recycle)

Uricase is the enzyme responsible for the breakdown of uric acid, the key molecule leading to gout in humans, into allantoin, but it is absent in humans. It has been produced as a PEGylated pharmaceutical where the purification is performed through three sequential chromatographic columns. More recently an aqueous two‐phase system (ATPS) was reported that could recover Uricase with high yield and purity. Although the use of ATPS can decrease cost and time, it also generates a large amount of waste. The ability, therefore, to recycle key components of ATPS is of interest. Economic modelling is a powerful tool that allows the bioprocess engineer to compare possible outcomes and find areas where further research or optimization might be required without recourse to extensive experiments and time. This research provides an economic analysis using the commercial software BioSolve of the strategies for Uricase production: chromatographic and ATPS, and includes a third bioprocess that uses material recycling. The key parameters that affect the process the most were located via a sensitivity analysis and evaluated with a Monte Carlo analysis. Results show that ATPS is far less expensive than chromatography, but that there is an area where the cost of production of both bioprocesses overlap. Furthermore, recycling does not impact the cost of production. This study serves to provide a framework for the economic analysis of Uricase production using alternative techniques. © 2015 American Institute of Chemical Engineers Biotechnol. Prog., 32:126–133, 2016

Economic analysis of Royalactin production under uncertainty: Evaluating the effect of parameter optimization

Royalactin is a protein with several different potential uses in humans. Research, in insects and in mammalian cells, has shown that it can accelerate cell division and prevent apoptosis. The method of action is through the use of the epidermal growth factor receptor, which is present in humans. Potential use in humans could be to lower cholesterolemic levels in blood, and to elicit similar effects to those seen in bees, e.g., increased lifespan. Mass production of Royalactin has not been accomplished, though a recent article presented a Pichia pastoris fermentation and recovery by aqueous two-phase systems at laboratory scale as a possible basis for production. Economic modelling is a useful tool with which compare possible outcomes for the production of such a molecule and in particular, to locate areas where additional research is needed and optimization may be required. This study uses the BioSolve software to perform an economic analysis on the scale-up of the putative process for Royalactin. The key parameters affecting the cost of production were located via a sensitivity analysis and then evaluated by Monte Carlo analysis. Results show that if titer is not optimized the strategy to maintain a low cost of goods is process oriented. After optimization of this parameter the strategy changes to a product-oriented and the target output becomes the critical parameter determining the cost of goods. This study serves to provide a framework for the evaluation of strategies for future production of Royalactin, by analyzing the factors that influence its cost of manufacture.

Model-based cell number quantification using online single-oxygen sensor data for tissue engineering perfusion bioreactors

Online and non-invasive quantification of critical tissue engineering (TE) construct quality attributes in TE bioreactors is indispensable for the cost-effective up-scaling and automation of cellular construct manufacturing. However, appropriate monitoring techniques for cellular constructs in bioreactors are still lacking. This study presents a generic and robust approach to determine cell number and metabolic activity of cell-based TE constructs in perfusion bioreactors based on single oxygen sensor data in dynamic perfusion conditions. A data-based mechanistic modeling technique was used that is able to correlate the number of cells within the scaffold (R(2)  = 0.80) and the metabolic activity of the cells (R(2)  = 0.82) to the dynamics of the oxygen response to step changes in the perfusion rate. This generic non-destructive measurement technique is effective for a large range of cells, from as low as 1.0 × 10(5) cells to potentially multiple millions of cells, and can open-up new possibilities for effective bioprocess monitoring.