Euler-Lagrange computational fluid dynamics for (bio)reactor scale down: An analysis of organism lifelines

An industrial perspective on metabolic responses of Penicillium chrysogenum to periodic dissolved oxygen feast-famine cycles in a scale-down system

While traveling through different zones in large-scale bioreactors, microbes are most likely subjected to fluctuating dissolved oxygen (DO) conditions at the timescales of global circulation time. In this study, to mimic industrial-scale spatial DO gradients, we present a scale-down setup based on dynamic feast/famine regime (150 s) that leads to repetitive cycles with rapid changes in DO availability in glucose-limited chemostat cultures of Penicillium chrysogenum. Such DO feast/famine regime induced a stable and repetitive pattern with a reproducible metabolic response in time, and the dynamic response of intracellular metabolites featured specific differences in terms of both coverage and magnitude in comparison to other dynamic conditions, for example, substrate feast/famine cycles. Remarkably, intracellular sugar polyols were considerably increased as the hallmark metabolites along with a dynamic and higher redox state (NADH/NAD+) of the cytosol. Despite the increased availability of NADPH for penicillin production under the oscillatory DO conditions, this positive effect may be counteracted by the decreased ATP supply. Moreover, it is interesting to note that not only the penicillin productivity was reduced under such oscillating DO conditions, but also that of the unrecyclable byproduct ortho-hydroxyphenyl acetic acid and degeneration of penicillin productivity. Furthermore, dynamic flux profiles showed the most pronounced variations in central carbon metabolism, amino acid (AA) metabolism, energy metabolism and fatty acid metabolism upon the DO oscillation. Taken together, the metabolic responses of P. chrysogenum to DO gradients reported here are important for elucidating metabolic regulation mechanisms, improving bioreactor design and scale-up procedures as well as for constructing robust cell strains to cope with heterogenous industrial culture conditions.

Repeated glucose oscillations in high cell-density cultures influence stress-related functions of Escherichia coli

Engineering microbial cells for the commercial production of biomolecules and biochemicals requires understanding how cells respond to dynamically changing substrate (feast-famine) conditions in industrial-scale bioreactors. Scale-down methods that oscillate substrate are commonly applied to predict the industrial-scale behavior of microbes. We followed a compartment modeling approach to design a scale-down method based on the simulation of an industrial-scale bioreactor. This study uses high cell-density scale-down experiments to investigate Escherichia coli knockout strains of five major glucose-sensitive transcription factors (Cra, Crp, FliA, PrpR, and RpoS) to study their regulatory role during glucose oscillations. RNA-sequencing analysis revealed that the glucose oscillations caused the down-regulation of several stress-related functions in E. coli. An in-depth analysis of strain physiology and transcriptome revealed a distinct phenotype of the strains tested under glucose oscillations. Specifically, the knockout strains of Cra, Crp, and RpoS resulted in a more sensitive transcriptional response than the control strain, while the knockouts of FliA and PrpR responded less severely. These findings imply that the regulation orchestrated by Cra, Crp, and RpoS may be essential for robust E. coli production strains. In contrast, the regulation by FliA and PrpR may be undesirable for temporal oscillations in glucose availability.

Integration Approaches to Model Bioreactor Hydrodynamics and Cellular Kinetics for Advancing Bioprocess Optimisation

Large-scale bioprocesses are increasing globally to cater to the larger market demands for biological products. As fermenter volumes increase, the efficiency of mixing decreases, and environmental gradients become more pronounced compared to smaller scales. Consequently, the cells experience gradients in process parameters, which in turn affects the efficiency and profitability of the process. Computational fluid dynamics (CFD) simulations are being widely embraced for their ability to simulate bioprocess performance, facilitate bioprocess upscaling, downsizing, and process optimisation. Recently, CFD approaches have been integrated with dynamic Cell reaction kinetic (CRK) modelling to generate valuable information about the cellular response to fluctuating hydrodynamic parameters inside large production processes. Such coupled approaches have the potential to facilitate informed decision-making in intelligent biomanufacturing, aligning with the principles of "Industry 4.0" concerning digitalisation and automation. In this review, we discuss the benefits of utilising integrated CFD-CRK models and the different approaches to integrating CFD-based bioreactor hydrodynamic models with cellular kinetic models. We also highlight the suitability of different coupling approaches for bioprocess modelling in the purview of associated computational loads.

Developing rational scale-down simulators for mimicking substrate heterogeneities based on cell lifelines in industrial-scale bioreactors

The influence of extracellular variations on the cellular metabolism and thereby the process performance at large-scale can be evaluated using the so-called scale-down simulators. Nevertheless, the major challenge is to design an appropriate scale-down simulator, which can accurately mimic the cell lifelines that record the flow paths and experiences of cells circulating in large-scale bioreactors. To address this, a dedicated SDSA (scale-down simulator application) was purposedly developed on the basis of black box model and process reaction model established for Penicillium chrysogenum strain as well as cell lifelines or trajectories information in an industrial-scale fermentor. Guided by the SDSA, the industrial-relevant metabolic regimes for substrate availability, i.e., excess, limitation and starvation, were successfully reproduced at laboratory-scale three-compartment scale-down (SD) system. In addition, such SDSA can also display individual process dynamics in each compartment, and demonstrate how individual factors influence the entire bioprocess performance, thus serving both educational and research purposes.

Milligrams to kilograms: making microbes work at scale

If biomanufacturing can become a sustainable route for producing chemicals, it will provide a critical step in reducing greenhouse gas emissions to fight climate change. However, efforts to industrialize microbial synthesis of chemicals have met with varied success, due, in part, to challenges in translating laboratory successes to industrial scale. With a particular focus on Escherichia coli, this review examines the lessons learned when studying microbial physiology and metabolism under conditions that simulate large-scale bioreactors and methods to minimize cellular waste through reduction of maintenance energy, optimizing the stress response and minimizing culture heterogeneity. With general strategies to overcome these challenges, biomanufacturing process scale-up could be de-risked and the time and cost of bringing promising syntheses to market could be reduced.

Mimicked Mixing-Induced Heterogeneities of Industrial Bioreactors Stimulate Long-Lasting Adaption Programs in Ethanol-Producing Yeasts

Commercial-scale bioreactors create an unnatural environment for microbes from an evolutionary point of view. Mixing insufficiencies expose individual cells to fluctuating nutrient concentrations on a second-to-minute scale while transcriptional and translational capacities limit the microbial adaptation time from minutes to hours. This mismatch carries the risk of inadequate adaptation effects, especially considering that nutrients are available at optimal concentrations on average. Consequently, industrial bioprocesses that strive to maintain microbes in a phenotypic sweet spot, during lab-scale development, might suffer performance losses when said adaptive misconfigurations arise during scale-up. Here, we investigated the influence of fluctuating glucose availability on the gene-expression profile in the industrial yeast Ethanol Red™. The stimulus-response experiment introduced 2 min glucose depletion phases to cells growing under glucose limitation in a chemostat. Even though Ethanol Red™ displayed robust growth and productivity, a single 2 min depletion of glucose transiently triggered the environmental stress response. Furthermore, a new growth phenotype with an increased ribosome portfolio emerged after complete adaptation to recurring glucose shortages. The results of this study serve a twofold purpose. First, it highlights the necessity to consider the large-scale environment already at the experimental development stage, even when process-related stressors are moderate. Second, it allowed the deduction of strain engineering guidelines to optimize the genetic background of large-scale production hosts.

Downscaling Industrial-Scale Syngas Fermentation to Simulate Frequent and Irregular Dissolved Gas Concentration Shocks

In large-scale syngas fermentation, strong gradients in dissolved gas (CO, H₂) concentrations are very likely to occur due to locally varying mass transfer and convection rates. Using Euler-Lagrangian CFD simulations, we analyzed these gradients in an industrial-scale external-loop gas-lift reactor (EL-GLR) for a wide range of biomass concentrations, considering CO inhibition for both CO and H₂ uptake. Lifeline analyses showed that micro-organisms are likely to experience frequent (5 to 30 s) oscillations in dissolved gas concentrations with one order of magnitude. From the lifeline analyses, we developed a conceptual scale-down simulator (stirred-tank reactor with varying stirrer speed) to replicate industrial-scale environmental fluctuations at bench scale. The configuration of the scale-down simulator can be adjusted to match a broad range of environmental fluctuations. Our results suggest a preference for industrial operation at high biomass concentrations, as this would strongly reduce inhibitory effects, provide operational flexibility and enhance the product yield. The peaks in dissolved gas concentration were hypothesized to increase the syngas-to-ethanol yield due to the fast uptake mechanisms in C. autoethanogenum. The proposed scale-down simulator can be used to validate such results and to obtain data for parametrizing lumped kinetic metabolic models that describe such short-term responses.

Elucidating yeast glycolytic dynamics at steady state growth and glucose pulses through kinetic metabolic modeling

Microbial cell factories face changing environments during industrial fermentations. Kinetic metabolic models enable the simulation of the dynamic metabolic response to these perturbations, but their development is challenging due to model complexity and experimental data requirements. An example of this is the well-established microbial cell factory Saccharomyces cerevisiae, for which no consensus kinetic model of central metabolism has been developed and implemented in industry. Here, we aim to bring the academic and industrial communities closer to this consensus model. We developed a physiology informed kinetic model of yeast glycolysis connected to central carbon metabolism by including the effect of anabolic reactions precursors, mitochondria and the trehalose cycle. To parametrize such a large model, a parameter estimation pipeline was developed, consisting of a divide and conquer approach, supplemented with regularization and global optimization. Additionally, we show how this first mechanistic description of a growing yeast cell captures experimental dynamics at different growth rates and under a strong glucose perturbation, is robust to parametric uncertainty and explains the contribution of the different pathways in the network. Such a comprehensive model could not have been developed without using steady state and glucose perturbation data sets. The resulting metabolic reconstruction and parameter estimation pipeline can be applied in the future to study other industrially-relevant scenarios. We show this by generating a hybrid CFD-metabolic model to explore intracellular glycolytic dynamics for the first time. The model suggests that all intracellular metabolites oscillate within a physiological range, except carbon storage metabolism, which is sensitive to the extracellular environment.

Elevated energy costs of biomass production in mitochondrial respiration-deficient Saccharomyces cerevisia

Microbial growth requires energy for maintaining the existing cells and producing components for the new ones. Microbes therefore invest a considerable amount of their resources into proteins needed for energy harvesting. Growth in different environments is associated with different energy demands for growth of yeast Saccharomyces cerevisiae, although the cross-condition differences remain poorly characterized. Furthermore, a direct comparison of the energy costs for the biosynthesis of the new biomass across conditions is not feasible experimentally; computational models, on the contrary, allow comparing the optimal metabolic strategies and quantify the respective costs of energy and nutrients. Thus in this study, we used a resource allocation model of S. cerevisiae to compare the optimal metabolic strategies between different conditions. We found that S. cerevisiae with respiratory-impaired mitochondria required additional energetic investments for growth, while growth on amino acid-rich media was not affected. Amino acid supplementation in anaerobic conditions also was predicted to rescue the growth reduction in mitochondrial respiratory shuttle-deficient mutants of S. cerevisiae. Collectively, these results point to elevated costs of resolving the redox imbalance caused by de novo biosynthesis of amino acids in mitochondria. To sum up, our study provides an example of how resource allocation modeling can be used to address and suggest explanations to open questions in microbial physiology.

CFD supported scale up of perfusion bioreactors in biopharma

The robust scale up of perfusion systems requires comparable conditions over all scales to ensure equivalent cell culture performance. As cells in continuous processes circulate outside the bioreactor, performance losses may arise if jet flow and stirring cause a direct connection between perfusion feed and return. Computational fluid dynamics can be used to identify such short circuit flows, assess mixing efficiencies, and eventually adapt the perfusion setup. This study investigates the scale up from a 2 L glass bioreactor to 100 L and 500 L disposable pilot scale systems. Highly resolved Lattice Boltzmann Large Eddy simulations were performed in single phase and mixing efficiencies (Emix) furthermore experimentally validated in the 2 L system. This evaluation gives insight into the flow pattern, the mixing behavior and information on cell residence time inside the bioreactors. No geometric adaptations in the pilot scale systems were necessary as Emix was greater than 90% for all conditions tested. Two different setups were evaluated in 2 L scale where the direction of flow was changed, yielding a difference in mixing efficiency of 10%. Nevertheless, since Emix was confirmed to be >90% also for both 2 L setups and the determined mixing times were in a similar range for all scales, the 2 L system was deemed to be a suitable scale down model. The results demonstrate how computational fluid dynamic models can be used for rational process design of intensified production processes in the biopharmaceutical industry.

Microbial lifelines in bioprocesses: From concept to application

Bioprocesses are scaled up for the production of large product quantities. With larger fermenter volumes, mixing becomes increasingly inefficient and environmental gradients get more prominent than in smaller scales. Environmental gradients have an impact on the microorganism's metabolism, which makes the prediction of large-scale performance difficult and can lead to scale-up failure. A promising approach for improved understanding and estimation of dynamics of microbial populations in large-scale bioprocesses is the analysis of microbial lifelines. The lifeline of a microbe in a bioprocess is the experience of environmental gradients from a cell's perspective, which can be described as a time series of position, environment and intracellular condition. Currently, lifelines are predominantly determined using models with computational fluid dynamics, but new technical developments in flow-following sensor particles and microfluidic single-cell cultivation open the door to a more interdisciplinary concept. We critically review the current concepts and challenges in lifeline determination and application of lifeline analysis, as well as strategies for the integration of these techniques into bioprocess development. Lifelines can contribute to a successful scale-up by guiding scale-down experiments and identifying strain engineering targets or bioreactor optimisations.

Influence of oxygen concentration on the metabolism of Penicillium chrysogenum

In large-scale bioreactors, there is often insufficient mixing and as a consequence, cells experience uneven substrate and oxygen levels that influence product formation. In this study, the influence of dissolved oxygen (DO) gradients on the primary and secondary metabolism of a high producing industrial strain of Penicillium chrysogenum was investigated. Within a wide range of DO concentrations, obtained under chemostat conditions, we observed different responses from P. chrysogenum: (i) no influence on growth or penicillin production (>0.025 mmol L-1); (ii) reduced penicillin production, but no growth limitation (0.013-0.025 mmol L-1); and (iii) growth and penicillin production limitations (<0.013 mmol L-1). In addition, scale down experiments were performed by oscillating the DO concentration in the bioreactor. We found that during DO oscillation, the penicillin production rate decreased below the value observed when a constant DO equal to the average oscillating DO value was used. To understand and predict the influence of oxygen levels on primary metabolism and penicillin production, we developed a black box model that was linked to a detailed kinetic model of the penicillin pathway. The model simulations represented the experimental data during the step experiments; however, during the oscillation experiments the predictions deviated, indicating the involvement of the central metabolism in penicillin production.

Computational fluid dynamics simulation improves the design and characterization of a plug-flow-type scale-down reactor for microbial cultivation processes

The scale-up of bioprocesses remains one of the major obstacles in the biotechnology industry. Scale-down bioreactors have been identified as valuable tools to investigate the heterogeneities observed in large-scale tanks at the laboratory scale. Additionally, computational fluid dynamics (CFD) simulations can be used to gain information about fluid flow in tanks used for production. Here, we present the rational design and comprehensive characterization of a scale-down setup, in which a flexible and modular plug-flow reactor was connected to a stirred-tank bioreactor. With the help of CFD using the realizable k-ε model, the mixing time difference between a 20 and 4000 L bioreactor was evaluated and used as scale-down criterion. CFD simulations using a shear stress transport (SST) k-ω turbulence model were used to characterize the plug-flow reactor in more detail, and the model was verified using experiments. Additionally, the model was used to simulate conditions where experiments technically could not be performed due to sensor limitations. Nevertheless, verification is difficult in this case as well. This was the first time a scale-down setup was tested on high-cell-density Escherichia coli cultivations to produce industrially relevant antigen-binding fragments (Fab). Biomass yield was reduced by 11% and specific product yield was reduced by 20% during the scale-down cultivations. Additionally, the intracellular Fab fraction was increased by using the setup. The flexibility of the introduced scale-down setup in combination with CFD simulations makes it a valuable tool for investigating scale effects at the laboratory scale. More information about the large scale is still necessary to further refine the setup and to speed up bioprocess scale-up in the future.

An analysis of organism lifelines in an industrial bioreactor using Lattice-Boltzmann CFD

Euler-Lagrange CFD simulations, where the biotic phase is represented by computational particles (parcels), provide information on environmental gradients inside bioreactors from the microbial perspective. Such information is highly relevant for reactor scale-down and process optimization. One of the major challenges is the computational intensity of CFD simulations, especially when resolution of dynamics in the flowfield is required. Lattice-Boltzmann large-eddy simulations (LB-LES) form a very promising approach for simulating accurate, dynamic flowfields in stirred reactors, at strongly reduced computation times compared to finite volume approaches. In this work, the performance of LB-LES in resolving substrate gradients in large-scale bioreactors is explored, combined with the inclusion of a Lagrangian biotic phase to provide the microbial perspective. In addition, the hydrodynamic performance of the simulations is confirmed by verification of hydrodynamic characteristics (radial velocity, turbulent kinetic energy, energy dissipation) in the impeller discharge stream of a 29 cm diameter stirred tank. The results are compared with prior finite volume simulation results, both in terms of hydrodynamic and biokinetic observations, and time requirements.

Performing in spite of starvation: How Saccharomyces cerevisiae maintains robust growth when facing famine zones in industrial bioreactors

In fed-batch operated industrial bioreactors, glucose-limited feeding is commonly applied for optimal control of cell growth and product formation. Still, microbial cells such as yeasts and bacteria are frequently exposed to glucose starvation conditions in poorly mixed zones or far away from the feedstock inlet point. Despite its commonness, studies mimicking related stimuli are still underrepresented in scale-up/scale-down considerations. This may surprise as the transition from glucose limitation to starvation has the potential to provoke regulatory responses with negative consequences for production performance. In order to shed more light, we performed gene-expression analysis of Saccharomyces cerevisiae grown in intermittently fed chemostat cultures to study the effect of limitation-starvation transitions. The resulting glucose concentration gradient was representative for the commercial scale and compelled cells to tolerate about 76 s with sub-optimal substrate supply. Special attention was paid to the adaptation status of the population by discriminating between first time and repeated entry into the starvation regime. Unprepared cells reacted with a transiently reduced growth rate governed by the general stress response. Yeasts adapted to the dynamic environment by increasing internal growth capacities at the cost of rising maintenance demands by 2.7%. Evidence was found that multiple protein kinase A (PKA) and Snf1-mediated regulatory circuits were initiated and ramped down still keeping the cells in an adapted trade-off between growth optimization and down-regulation of stress response. From this finding, primary engineering guidelines are deduced to optimize both the production host's genetic background and the design of scale-down experiments.

Digital models in biotechnology: Towards multi-scale integration and implementation

Industrial biotechnology encompasses a large area of multi-scale and multi-disciplinary research activities. With the recent megatrend of digitalization sweeping across all industries, there is an increased focus in the biotechnology industry on developing, integrating and applying digital models to improve all aspects of industrial biotechnology. Given the rapid development of this field, we systematically classify the state-of-art modelling concepts applied at different scales in industrial biotechnology and critically discuss their current usage, advantages and limitations. Further, we critically analyzed current strategies to couple cell models with computational fluid dynamics to study the performance of industrial microorganisms in large-scale bioprocesses, which is of crucial importance for the bio-based production industries. One of the most challenging aspects in this context is gathering intracellular data under industrially relevant conditions. Towards comprehensive models, we discuss how different scale-down concepts combined with appropriate analytical tools can capture intracellular states of single cells. We finally illustrated how the efforts could be used to develop digitals models suitable for both cell factory design and process optimization at industrial scales in the future.

Aerobic bioreactors: condensers, evaporation rates, scale-up and scale-down

OBJECTIVES

Hydrodynamics, mixing and shear are terms often used when explaining or modelling scale differences, but other scale differences, such as evaporation, can arise from non-hydrodynamic factors that can be managed with some awareness and effort.

RESULTS

We present an engineering approach to the prediction of evaporation rates in bioreactors based on gH₂O/Nm³ of air entering and leaving the bioreactor and confirm its usefulness in a 28-run design of experiments investigating the effects of aeration rate (0.02 to 2.0 VVM), condenser temperature (10 to 20 °C), fill (2.5 to 5 kg), broth temperature (25 to 40 °C) and agitator speed (25 to 800 rpm). Aeration rate and condenser temperature used in the engineering prediction provided a practically useful estimate of evaporation; the other factors, while statistically identified as having some influence, were of negligible practical usefulness. Evaporation rates were never found to be zero, and could be at least 10% different to those expected at scale.

CONCLUSIONS

An assessment of evaporation rates for any project is encouraged, and it is recommended that the effects are accounted for by measurements, modelling or by tuning the exhaust cooling device temperature to minimize scale differences.

Bio-Inspired Propulsion: Towards Understanding the Role of Pectoral Fin Kinematics in Manta-like Swimming

Through computational fluid dynamics (CFD) simulations of a model manta ray body, the hydrodynamic role of manta-like bioinspired flapping is investigated. The manta ray model motion is reconstructed from synchronized high-resolution videos of manta ray swimming. Rotation angles of the model skeletal joints are altered to scale the pitching and bending, resulting in eight models with different pectoral fin pitching and bending ratios. Simulations are performed using an in-house developed immersed boundary method-based numerical solver. Pectoral fin pitching ratio (PR) is found to have significant implications in the thrust and efficiency of the manta model. This occurs due to more optimal vortex formation and shedding caused by the lower pitching ratio. Leading edge vortexes (LEVs) formed on the bottom of the fin, a characteristic of the higher PR cases, produced parasitic low pressure that hinders thrust force. Lowering the PR reduces the influence of this vortex while another LEV that forms on the top surface of the fin strengthens it. A moderately high bending ratio (BR) can slightly reduce power consumption. Finally, by combining a moderately high BR = 0.83 with PR = 0.67, further performance improvements can be made. This enhanced understanding of manta-inspired propulsive mechanics fills a gap in our understanding of the manta-like mobuliform locomotion. This motivates a new generation of manta-inspired robots that can mimic the high speed and efficiency of their biological counterpart.

Microfluidic Reproduction of Dynamic Bioreactor Environment Based on Computational Lifelines

The biotechnological production of fine chemicals, proteins and pharmaceuticals is usually hampered by loss of microbial performance during scale-up. This challenge is mainly caused by discrepancies between homogeneous environmental conditions at laboratory scale, where bioprocesses are optimized, and inhomogeneous conditions in large-scale bioreactors, where production takes place. Therefore, to improve strain selection and process development, it is of great interest to characterize these fluctuating conditions at large-scale and to study their effects on microbial cells. In this paper, we demonstrate the potential of computational fluid dynamics (CFD) simulation of large-scale bioreactors combined with dynamic microfluidic single-cell cultivation (dMSCC). Environmental conditions in a 200 L bioreactor were characterized with CFD simulations. Computational lifelines were determined by combining simulated turbulent multiphase flow, mass transport and particle tracing. Glucose availability for Corynebacterium glutamicum cells was determined. The reactor was simulated with average glucose concentrations of 6 g m−3, 10 g m−3 and 16 g m−3. The resulting computational lifelines, discretized into starvation and abundance regimes, were used as feed profiles for the dMSCC to investigate how varying glucose concentration affects cell physiology and growth rate. In this study, each colony in the dMSCC device represents a single cell as it travels through the reactor. Under oscillating conditions reproduced in the dMSCC device, a decrease in growth rate of about 40% was observed compared to continuous supply with the same average glucose availability. The presented approach provides insights into environmental conditions observed by microorganisms in large-scale bioreactors. It also paves the way for an improved understanding of how inhomogeneous environmental conditions influence cellular physiology, growth and production.

Stochastic parcel tracking in an Euler-Lagrange compartment model for fast simulation of fermentation processes

The compartment model (CM) is a well‐known approach for computationally affordable, spatially resolved hydrodynamic modeling of unit operations. Recent implementations use flow profiles based on Computational Fluid Dynamics (CFD) simulations, and several authors included microbial kinetics to simulate gradients in bioreactors. However, these studies relied on black‐box kinetics that do not account for intracellular changes and cell population dynamics in response to heterogeneous environments. In this paper, we report the implementation of a Lagrangian reaction model, where the microbial phase is tracked as a set of biomass‐parcels, each linked with an intracellular composition vector and a structured reaction model describing their intracellular response to extracellular variations. A stochastic parcel tracking approach is adopted, in contrast to the resolved trajectories used in CFD implementations. A penicillin production process is used as a case study. We show good performance of the model compared with full CFD simulations, both regarding the extracellular gradients and intracellular pool response, using the mixing time as a matching criterion and taking into account that the mixing time is sensitive to the number of compartments. The sensitivity of the model output towards some of the inputs is explored. The coarsest representative CM requires a few minutes to solve 80 h of flow time, compared with approximately 2 weeks for a full Euler–Lagrange CFD simulation of the same case. This alleviates one of the major bottlenecks for the application of such CFD simulations towards the analysis and optimization of industrial fermentation processes.

Monitoring Intracellular Metabolite Dynamics in Saccharomyces cerevisiae during Industrially Relevant Famine Stimuli

Carbon limitation is a common feeding strategy in bioprocesses to enable an efficient microbiological conversion of a substrate to a product. However, industrial settings inherently promote mixing insufficiencies, creating zones of famine conditions. Cells frequently traveling through such regions repeatedly experience substrate shortages and respond individually but often with a deteriorated production performance. A priori knowledge of the expected strain performance would enable targeted strain, process, and bioreactor engineering for minimizing performance loss. Today, computational fluid dynamics (CFD) coupled to data-driven kinetic models are a promising route for the in silico investigation of the impact of the dynamic environment in the large-scale bioreactor on microbial performance. However, profound wet-lab datasets are needed to cover relevant perturbations on realistic time scales. As a pioneering study, we quantified intracellular metabolome dynamics of Saccharomyces cerevisiae following an industrially relevant famine perturbation. Stimulus-response experiments were operated as chemostats with an intermittent feed and high-frequency sampling. Our results reveal that even mild glucose gradients in the range of 100 µmol·L−1 impose significant perturbations in adapted and non-adapted yeast cells, altering energy and redox homeostasis. Apparently, yeast sacrifices catabolic reduction charges for the sake of anabolic persistence under acute carbon starvation conditions. After repeated exposure to famine conditions, adapted cells show 2.7% increased maintenance demands.

Kinetic Modeling of Saccharomyces cerevisiae Central Carbon Metabolism: Achievements, Limitations, and Opportunities

Central carbon metabolism comprises the metabolic pathways in the cell that process nutrients into energy, building blocks and byproducts. To unravel the regulation of this network upon glucose perturbation, several metabolic models have been developed for the microorganism Saccharomyces cerevisiae. These dynamic representations have focused on glycolysis and answered multiple research questions, but no commonly applicable model has been presented. This review systematically evaluates the literature to describe the current advances, limitations, and opportunities. Different kinetic models have unraveled key kinetic glycolytic mechanisms. Nevertheless, some uncertainties regarding model topology and parameter values still limit the application to specific cases. Progressive improvements in experimental measurement technologies as well as advances in computational tools create new opportunities to further extend the model scale. Notably, models need to be made more complex to consider the multiple layers of glycolytic regulation and external physiological variables regulating the bioprocess, opening new possibilities for extrapolation and validation. Finally, the onset of new data representative of individual cells will cause these models to evolve from depicting an average cell in an industrial fermenter, to characterizing the heterogeneity of the population, opening new and unseen possibilities for industrial fermentation improvement.

Changes in Oxygen Availability during Glucose-Limited Chemostat Cultivations of Penicillium chrysogenum Lead to Rapid Metabolite, Flux and Productivity Responses

Bioreactor scale-up from the laboratory scale to the industrial scale has always been a pivotal step in bioprocess development. However, the transition of a bioeconomy from innovation to commercialization is often hampered by performance loss in titer, rate and yield. These are often ascribed to temporal variations of substrate and dissolved oxygen (for instance) in the environment, experienced by microorganisms at the industrial scale. Oscillations in dissolved oxygen (DO) concentration are not uncommon. Furthermore, these fluctuations can be exacerbated with poor mixing and mass transfer limitations, especially in fermentations with filamentous fungus as the microbial cell factory. In this work, the response of glucose-limited chemostat cultures of an industrial Penicillium chrysogenum strain to different dissolved oxygen levels was assessed under both DO shift-down (60% → 20%, 10% and 5%) and DO ramp-down (60% → 0% in 24 h) conditions. Collectively, the results revealed that the penicillin productivity decreased as the DO level dropped down below 20%, while the byproducts, e.g., 6-oxopiperidine-2-carboxylic acid (OPC) and 6-aminopenicillanic acid (6APA), accumulated. Following DO ramp-down, penicillin productivity under DO shift-up experiments returned to its maximum value in 60 h when the DO was reset to 60%. The result showed that a higher cytosolic redox status, indicated by NADH/NAD+, was observed in the presence of insufficient oxygen supply. Consistent with this, flux balance analysis indicated that the flux through the glyoxylate shunt was increased by a factor of 50 at a DO value of 5% compared to the reference control, favoring the maintenance of redox status. Interestingly, it was observed that, in comparison with the reference control, the penicillin productivity was reduced by 25% at a DO value of 5% under steady state conditions. Only a 14% reduction in penicillin productivity was observed as the DO level was ramped down to 0. Furthermore, intracellular levels of amino acids were less sensitive to DO levels at DO shift-down relative to DO ramp-down conditions; this difference could be caused by different timescales between turnover rates of amino acid pools (tens of seconds to minutes) and DO switches (hours to days at steady state and minutes to hours at ramp-down). In summary, this study showed that changes in oxygen availability can lead to rapid metabolite, flux and productivity responses, and dynamic DO perturbations could provide insight into understanding of metabolic responses in large-scale bioreactors.

Automated Compartment Model Development Based on Data from Flow-Following Sensor Devices

Due to the heterogeneous nature of large-scale fermentation processes they cannot be modelled as ideally mixed reactors, and therefore flow models are necessary to accurately represent the processes. Computational fluid dynamics (CFD) is used more and more to derive flow fields for the modelling of bioprocesses, but the computational demands associated with simulation of multiphase systems with biokinetics still limits their wide applicability. Hence, a demand for simpler flow models persists. In this study, an approach to develop data-based flow models in the form of compartment models is presented, which utilizes axial-flow rates obtained from flow-following sensor devices in combination with a proposed procedure for automatic zoning of volume. The approach requires little experimental effort and eliminates the necessity for computational determination of inter-compartmental flow rates and manual zoning. The concept has been demonstrated in a 580 L stirred vessel, of which models have been developed for two types of impellers with varying agitation intensities. The sensor device measurements were corroborated by CFD simulations, and the performance of the developed compartment models was evaluated by comparing predicted mixing times with experimentally determined mixing times. The data-based compartment models predicted the mixing times for all examined conditions with relative errors in the range of 3–27%. The deviations were ascribed to limitations in the flow-following behavior of the sensor devices, whose sizes were relatively large compared to the examined system. The approach provides a versatile and automated flow modelling platform which can be applied to large-scale bioreactors.

Pseudomonas putida KT2440 endures temporary oxygen limitations

The obligate aerobic nature of Pseudomonas putida, one of the most prominent whole-cell biocatalysts emerging for industrial bioprocesses, questions its ability to be cultivated in large-scale bioreactors, which exhibit zones of low dissolved oxygen tension. P. putida KT2440 was repeatedly subjected to temporary oxygen limitations in scale-down approaches to assess the effect on growth and an exemplary production of rhamnolipids. At those conditions, the growth and production of P. putida KT2440 were decelerated compared to well-aerated reference cultivations, but remarkably, final biomass and rhamnolipid titers were similar. The robust growth behavior was confirmed across different cultivation systems, media compositions, and laboratories, even when P. putida KT2440 was repeatedly exposed to dual carbon and oxygen starvation. Quantification of the nucleotides ATP, ADP, and AMP revealed a decrease of intracellular ATP concentrations with increasing duration of oxygen starvation, which can, however, be restored when re-supplied with oxygen. Only small changes in the proteome were detected when cells encountered oscillations in dissolved oxygen tensions. Concluding, P. putida KT2440 appears to be able to cope with repeated oxygen limitations as they occur in large-scale bioreactors, affirming its outstanding suitability as a whole-cell biocatalyst for industrial-scale bioprocesses.

Parametric numerical study and optimization of mass transfer and bubble size distribution in a gas-liquid stirred tank bioreactor equipped with Rushton turbine using computational fluid dynamics

Designing and optimizing a bioreactor can be an especially challenging process. Computational modelling is an effective tool to investigate the effects of various operating parameters on bioreactor performance and identify the optimum ones. In this work, a computational fluid dynamics-population balance model (CFD-PBM) was developed to elucidate the effect of different geometrical and operating parameters on the hydrodynamics and mass transfer coefficient of a batch stirred tank bioreactor. The validated model was projected to predict the effect of different parameters including the gas flow rate, the impeller off-bottom clearance, the number of agitator blades, and rotational speed of the impeller on the velocity profiles, air volume fraction, bubble size distribution, and the local gas mass transfer coefficient (K l a) in the bioreactor. Air bubble breakup and coalescence phenomena were considered in all simulations. Factorial experimental design approach was employed to statistically investigate the impacts of the aforementioned operating and geometrical parameters on K l a and bubble size distribution in the bioreactor in order to determine the most significant parameters. This can give an essential insight into the most impactful factors when it comes to designing and scaling up a bioreactor.

Euler-Lagrangian Simulations: A Proper Tool for Predicting Cellular Performance in Industrial Scale Bioreactors

Eulerian-Lagrangian approach to investigate cellular responses in a bioreactor has become the center of attention in recent years. It was introduced to biotechnological processes about two decades ago, but within the last few years, it proved itself as a powerful tool to address scale-up and -down topics of bioprocesses. It can capture the history of a cell and reveal invaluable information for, not only, bioprocess control and design but also strain engineering. This way it will be possible to shed light on the actual environment that cell experiences throughout its lifespan. Lifelines of a microorganism in a bioreactor can serve as the missing link that encompasses the biological timescales and the physical timescales. For this purpose digitalization of bioreactors provides us with new insights that are not achievable in industrial reactors easily if at all, namely, substrate and product gradients; high-shear regions are among the most interesting factors that can be reproduced adequately with help of a digital twin. In this chapter basic principles of this method will be introduced, and later on some practical aspects of particle tracking technique will be illustrated. In the final section, some of the advantages and challenges associated with this method will be discussed.

On the verge of the market - Plant factories for the automated and standardized production of biopharmaceuticals

The market for biopharmaceuticals is dominated by recombinant proteins and is driven mainly by the development of vaccines and antibodies. Manufacturing predominantly relies on fermentation-based production platforms, which have limited scalability and suffer from high upstream process costs. As an alternative, the production of recombinant proteins in whole plants (plant molecular farming) provides a scalable and cost efficient upstream process because each plant functions as a self-contained bioreactor, avoiding costs associated with single-use devices and cleaning-in-place. Despite many proof-of-concept studies and the approval of a few products as medical devices, the only approved pharmaceutical proteins manufactured in whole plants have been authorized under emergency protocols. The absence of approvals under standard clinical development pathways in part reflects the lack of standardized process equipment and unit operations, leading to industry inertia based on familiarity with fermenter systems. Here we discuss the upstream production steps of plant molecular farming by transient expression in intact plants, including seeding, plant cultivation, infiltration with Agrobacterium tumefaciens, post-infiltration incubation, and harvesting. We focus on cultivation techniques because they strongly affect the subsequent steps and overall process design. We compare the benefits and drawbacks of open field, greenhouse and vertical farm strategies in terms of upfront investment costs, batch reproducibility, and decoupling from environmental impacts. We consider process automation, monitoring and adaptive process design in the context of Industry 4.0, which can boost process efficiency and batch-to-batch uniformity to improve regulatory compliance. Finally, we discuss the costs-benefit aspects of the different cultivation systems.

The impact of clearance on mixing time for interface-added substrate

This study was carried out to find the optimum clearance (impeller to bottom distance) for Rushton and pitch-blade turbine impellers in a stirred tank bioreactor for improved substrate mixing time added at interface, taking advantage of computational fluid dynamics. In this regard, the time needed for a thin layer of liquid, resembling substrate-rich or poor part, getting homogenously dispersed within the tank was calculated. The mixing time calculated in this way is called the surface aeration related mixing time (SARMT). SARMT was calculated using two approaches and was compared with each other. For the pitch-blade turbine impeller, a criterion which guarantees accurate mixing time by simulation was not satisfied, so the SARMT profile against clearance was not achieved. For the Rushton impeller, a general descending order of SARMT against impeller-bottom clearance was observed.

Understanding gradients in industrial bioreactors

Gradients in industrial bioreactors have attracted substantial research attention since exposure to fluctuating environmental conditions has been shown to lead to changes in the metabolome, transcriptome as well as population heterogeneity in industrially relevant microorganisms. Such changes have also been found to impact key process parameters like the yield on substrate and the productivity. Hence, understanding gradients is important from both the academic and industrial perspectives. In this review the causes of gradients are outlined, along with their impact on microbial physiology. Quantifying the impact of gradients requires a detailed understanding of both fluid flow inside industrial equipment and microbial physiology. This review critically examines approaches used to investigate gradients including large-scale experimental work, computational methods and scale-down approaches. Avenues for future work have been highlighted, particularly the need for further coordinated development of both in silico and experimental tools which can be used to further the current understanding of gradients in industrial equipment.

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.

Flow-following sensor devices: A tool for bridging data and model predictions in large-scale fermentations

Production-scale fermentation processes in industrial biotechnology experience gradients in process variables, such as dissolved gases, pH and substrate concentrations, which can potentially affect the production organism and therefore the yield and profitability of the processes. However, the extent of the heterogeneity is unclear, as it is currently a challenge at large scale to obtain representative measurements from different zones of the reactor volume. Computational fluid dynamics (CFD) models have proven to be a valuable tool for better understanding the environment inside bioreactors. Without detailed measurements to support the CFD predictions, the validity of CFD models is debatable. A promising technology to obtain such measurements from different zones in the bioreactors are flow-following sensor devices, whose development has recently benefitted from advancements in microelectronics and sensor technology. This paper presents the state of the art within flow-following sensor device technology and addresses how the technology can be used in large-scale bioreactors to improve the understanding of the process itself and to test the validity of detailed computational models of the bioreactors in the future.

Data-driven in silico prediction of regulation heterogeneity and ATP demands of Escherichia coli in large-scale bioreactors

Escherichia coli exposed to industrial-scale heterogeneous mixing conditions respond to external stress by initiating short-term metabolic and long-term strategic transcriptional programs. In native habitats, long-term strategies allow survival in severe stress but are of limited use in large bioreactors, where microenvironmental conditions may change right after said programs are started. Related on/off switching of genes causes additional ATP burden that may reduce the cellular capacity for producing the desired product. Here, we present an agent-based data-driven model linked to computational fluid dynamics, finally allowing to predict additional ATP needs of Escherichia coli K12 W3110 exposed to realistic large-scale bioreactor conditions. The complex model describes transcriptional up- and downregulation dynamics of about 600 genes starting from subminute range covering 28 h. The data-based approach was extracted from comprehensive scale-down experiments. Simulating mixing and mass transfer conditions in a 54 m³ stirred bioreactor, 120,000 E. coli cells were tracked while fluctuating between different zones of glucose availability. It was found that cellular ATP demands rise between 30% and 45% of growth decoupled maintenance needs, which may limit the production of ATP-intensive product formation accordingly. Furthermore, spatial analysis of individual cell transcriptional patterns reveal very heterogeneous gene amplifications with hot spots of 50%-80% messenger RNA upregulation in the upper region of the bioreactor. The phenomenon reflects the time-delayed regulatory response of the cells that propagate through the stirred tank. After 4.2 h, cells adapt to environmental changes but still have to bear an additional 6% ATP demand.

Applying multiple approaches to deepen understanding of mixing and mass transfer in large-scale aerobic fermentations

Different methods are used at Corteva^® Agriscience to improve our understanding of mixing in large-scale mechanically agitated fermentors. These include (a) use of classical empirical correlations, (b) use of small-scale models, and (c) computational fluid dynamics (CFD). Each of these approaches has its own inherent strengths and limitations. Classic empirical or semi-empirical correlations can provide insights into mass transfer, blending, shear, and other important factors but are dependent on the geometry and condition used to develop the correlations. Laboratory-scale modelling can be very useful to study mixing and model the effect of heterogeneity on the culture, but success is highly dependent on the methodology applied. CFD provides an effective means to accelerate the exploration of alternative design strategies through physics-based computer simulations that may not be adequately described by existing knowledge or correlations. However, considerable time and effort is needed to build and validate these models. In this paper, we review the various approaches used at Corteva Agriscience to deepen our understanding of mixing in large-scale fermentation processes.

Investigation of cell line specific responses to pH inhomogeneity and consequences for process design

With increasing bioreactor volumes, the mixing time of the reactor increases as well, which creates an inhomogeneous environment for the cells. This can result in impaired process performance in large-scale production reactors. Particularly the addition of base through the reactor headspace can be problematic, since it creates an area, where cells are repeatedly exposed to an increased pH. The aim of this study is to simulate this large-scale phenomenon at lab-scale and investigate its impact. Two different cell lines were exposed to pH amplitudes of a maximal magnitude of 0.05 units (pH of 6.95). Both cell lines showed similar responses, like decreased viable cell counts, but unaffected lactate levels. However, cell line B showed an initially increased specific productivity in response to the introduced amplitudes, whereas cell line A showed a consistently lower specific productivity. Furthermore, the time point at which base addition is started influences the impact, which pH amplitudes have on process performance. When pH control was started earlier in the process, maximal viable cell counts decreased and the lactate metabolic shift was less pronounced. These results show that the potential negative impact of pH amplitudes can be minimized by strategic process design.

Synergistically applying 1-D modeling and CFD for designing industrial scale bubble column syngas bioreactors

The reduction of greenhouse gas emissions and future perspectives of circular economy ask for new solutions to produce commodities and fine chemicals. Large-scale bubble columns operated by gaseous substrates such as CO, CO2, and H2 to feed acetogens for product formations could be promising approaches. Valid in silico predictions of large-scale performance are needed to dimension bioreactors properly taking into account biological constraints, too. This contribution deals with the trade-off between sophisticated spatiotemporally resolved large-scale simulations using computationally intensive Euler-Euler and Euler-Lagrange approaches and coarse-grained 1-D models enabling fast performance evaluations. It is shown that proper consideration of gas hold-up is key to predict biological performance. Intrinsic bias of 1-D models can be compensated by reconsideration of Sauter diameters derived from uniquely performed Euler-Lagrange computational fluid dynamics.

Escherichia coli metabolism under short-term repetitive substrate dynamics: adaptation and trade-offs

Background

Microbial metabolism is highly dependent on the environmental conditions. Especially, the substrate concentration, as well as oxygen availability, determine the metabolic rates. In large-scale bioreactors, microorganisms encounter dynamic conditions in substrate and oxygen availability (mixing limitations), which influence their metabolism and subsequently their physiology. Earlier, single substrate pulse experiments were not able to explain the observed physiological changes generated under large-scale industrial fermentation conditions.

Results

In this study we applied a repetitive feast–famine regime in an aerobic Escherichia coli culture in a time-scale of seconds. The regime was applied for several generations, allowing cells to adapt to the (repetitive) dynamic environment. The observed response was highly reproducible over the cycles, indicating that cells were indeed fully adapted to the regime. We observed an increase of the specific substrate and oxygen consumption (average) rates during the feast–famine regime, compared to a steady-state (chemostat) reference environment. The increased rates at same (average) growth rate led to a reduced biomass yield (30% lower). Interestingly, this drop was not followed by increased by-product formation, pointing to the existence of energy-spilling reactions. During the feast–famine cycle, the cells rapidly increased their uptake rate. Within 10 s after the beginning of the feeding, the substrate uptake rate was higher (4.68 μmol/g_CDW/s) than reported during batch growth (3.3 μmol/g_CDW/s). The high uptake led to an accumulation of several intracellular metabolites, during the feast phase, accounting for up to 34% of the carbon supplied. Although the metabolite concentrations changed rapidly, the cellular energy charge remained unaffected, suggesting well-controlled balance between ATP producing and ATP consuming reactions.

Conclusions

The adaptation of the physiology and metabolism of E. coli under substrate dynamics, representative for large-scale fermenters, revealed the existence of several cellular mechanisms coping with stress. Changes in the substrate uptake system, storage potential and energy-spilling processes resulted to be of great importance. These metabolic strategies consist a meaningful step to further tackle reduced microbial performance, observed under large-scale cultivations.

Scale-down simulators for mammalian cell culture as tools to access the impact of inhomogeneities occurring in large-scale bioreactors

During the scale-up of a bioprocess, not all characteristics of the process can be kept constant throughout the different scales. This typically results in increased mixing times with increasing reactor volumes. The poor mixing leads in turn to the formation of concentration gradients throughout the reactor and exposes cells to varying external conditions based on their location in the bioreactor. This can affect process performance and complicate process scale-up. Scale-down simulators, which aim at replicating the large-scale environment, expose the cells to changing environmental conditions. This has the potential to reveal adaptation mechanisms, which cells are using to adjust to rapidly fluctuating environmental conditions and can identify possible root causes for difficulties maintaining similar process performance at different scales. This understanding is of utmost importance in process validation. Additionally, these simulators also have the potential to be used for selecting cells, which are most robust when encountering changing extracellular conditions. The aim of this review is to summarize recent work in this interesting and promising area with the focus on mammalian bioprocesses, since microbial processes have been extensively reviewed.

Dynamic Environmental Control in Microfluidic Single-Cell Cultivations: From Concepts to Applications

Microfluidic single-cell cultivation (MSCC) is an emerging field within fundamental as well as applied biology. During the last years, most MSCCs were performed at constant environmental conditions. Recently, MSCC at oscillating and dynamic environmental conditions has started to gain significant interest in the research community for the investigation of cellular behavior. Herein, an overview of this topic is given and microfluidic concepts that enable oscillating and dynamic control of environmental conditions with a focus on medium conditions are discussed, and their application in single-cell research for the cultivation of both mammalian and microbial cell systems is demonstrated. Furthermore, perspectives for performing MSCC at complex dynamic environmental profiles of single parameters and multiparameters (e.g., pH and O₂ ) in amplitude and time are discussed. The technical progress in this field provides completely new experimental approaches and lays the foundation for systematic analysis of cellular metabolism at fluctuating environments.

Development and characterization of Escherichia coli triple reporter strains for investigation of population heterogeneity in bioprocesses

Background

Today there is an increasing demand for high yielding robust and cost efficient biotechnological production processes. Although cells in these processes originate from isogenic cultures, heterogeneity induced by intrinsic and extrinsic influences is omnipresent. To increase understanding of this mechanistically poorly understood phenomenon, advanced tools that provide insights into single cell physiology are needed.

Results

Two Escherichia coli triple reporter strains have been designed based on the industrially relevant production host E. coli BL21(DE3) and a modified version thereof, E. coli T7E2. The strains carry three different fluorescence proteins chromosomally integrated. Single cell growth is followed with EmeraldGFP (EmGFP)-expression together with the ribosomal promoter rrnB. General stress response of single cells is monitored by expression of sigma factor rpoS with mStrawberry, whereas expression of the nar-operon together with TagRFP657 gives information about oxygen limitation of single cells. First, the strains were characterized in batch operated stirred-tank bioreactors in comparison to wildtype E. coli BL21(DE3). Afterwards, applicability of the triple reporter strains for investigation of population heterogeneity in bioprocesses was demonstrated in continuous processes in stirred-tank bioreactors at different growth rates and in response to glucose and oxygen perturbation simulating gradients on industrial scale. Population and single cell level physiology was monitored evaluating general physiology and flow cytometry analysis of fluorescence distributions of the triple reporter strains. Although both triple reporter strains reflected physiological changes that were expected based on the expression characteristics of the marker proteins, the triple reporter strain based on E. coli T7E2 showed higher sensitivity in response to environmental changes. For both strains, noise in gene expression was observed during transition from phases of non-growth to growth. Apparently, under some process conditions, e.g. the stationary phase in batch cultures, the fluorescence response of EmGFP and mStrawberry is preserved, whereas TagRFP657 showed a distinct response.

Conclusions

Single cell growth, general stress response and oxygen limitation of single cells could be followed using the two triple reporter strains developed in this study. They represent valuable tools to study population heterogeneity in bioprocesses significantly increasing the level of information compared to the use of single reporter strains.