Lagrangian Trajectories to Predict the Formation of Population Heterogeneity in Large-Scale Bioreactors

Abstract

Successful scale-up of bioprocesses requires that laboratory-scale performance is equally achieved during large-scale production to meet economic constraints. In industry, heuristic approaches are often applied, making use of physical scale-up criteria that do not consider cellular needs or properties. As a consequence, large-scale productivities, conversion yields, or product purities are often deteriorated, which may prevent economic success. The occurrence of population heterogeneity in large-scale production may be the reason for underperformance. In this study, an in silico method to predict the formation of population heterogeneity by combining computational fluid dynamics (CFD) with a cell cycle model of Pseudomonas putida KT2440 was developed. The glucose gradient and flow field of a 54,000 L stirred tank reactor were generated with the Euler approach, and bacterial movement was simulated as Lagrange particles. The latter were statistically evaluated using a cell cycle model. Accordingly, 72% of all cells were found to switch between standard and multifork replication, and 10% were likely to undergo massive, transcriptional adaptations to respond to extracellular starving conditions. At the same time, 56% of all cells replicated very fast, with µ ≥ 0.3 h−1 performing multifork replication. The population showed very strong heterogeneity, as indicated by the observation that 52.9% showed higher than average adenosine triphosphate (ATP) maintenance demands (12.2%, up to 1.5 fold). These results underline the potential of CFD linked to structured cell cycle models for predicting large-scale heterogeneity in silico and ab initio.

Bioprocess scale‐up/down as integrative enabling technology: from fluid mechanics to systems biology and beyond

“…The consequence is that every time when a new position of a cell is computed (based on the computed fluid flow), a random side step is added that mimics the effect of the turbulence. By storing the positions of the cells as function of time, the 'lifeline' of each simulated cell is recorded (Enfors et al, 2001;Haringa et al, 2016;Kuschel et al, 2017).…”

Coupled metabolic‐hydrodynamic modeling enabling rational scale‐up of industrial bioprocesses

“…The computational demand is considerable since O (10 5 +) parcels may be required for a proper representation of the population (Haringa, Noorman, & Mudde, ), but accounting for additional pools and pool interaction is straightforward, thus averting the shortcomings of the Eulerian approach. An additional benefit is that the extra‐ and intracellular status of each parcel can be tracked, providing so‐called “lifelines” (Haringa et al, ; Kuschel, Siebler, & Takors, ; Lapin et al, , ; Y. Liu et al, ) and providing information on the environment through the organism's eyes. With this methodology in place, simulations can be utilized as an integral part of the bioreactor design cycle (Figure ), starting with assessment of the large‐scale design (Figure a), downscaling and analysis of the organism's response (Figure b), including said response in the simulation (Figure c) and in‐silico improvement of the design (Figure d).…”

“…A comparison of the cases also provided valuable insights into how substrate gradients manifest, depending on the substrate uptake characteristics and mixing behavior. McClure, Kavanagh, Fletcher, and Barton () studied exposure to excess sugar concentrations in a bubble column, whereas Kuschel et al () studied the presence of different regimes in replication behavior of P. putida in an industrial reactor. Most recently, Siebler, Lapin, Hermann, and Takors () used the approach to study CO limitations in syngas fermentation with C. Ljungdahlii , indicating 97% of cells experience substrate limitations, and 84% are likely to undergo transcription changes, after stress exposures of more than 70 s. Y. Liu et al () focused on strain‐rate variations observed by plant cells and their influence on viability loss, in what constitutes an evolution of the “Energy dissipation circulation function (EDCF)” method (Jüsten, Paul, Nienow, & Thomas, ).…”

“…Three different methods for lifeline analysis were developed by Haringa et al (). In regime analysis, the lifeline gets discretized into a limited number of regimes, each with a certain characteristic, such as substrate starvation and excess, overflow/byproduct formation, and so forth (Haringa et al, ; Kuschel et al, ; Siebler et al, ). The frequency at which cells switches between regimes and the duration of exposure to each regime are quantified.…”

“…The computational demand is considerable since O (10 5 +) parcels may be required for a proper representation of the population , but accounting for additional pools and pool interaction is straightforward, thus averting the shortcomings of the Eulerian approach. An additional benefit is that the extra-and intracellular status of each parcel can be tracked, providing so-called "lifelines" Kuschel, Siebler, & Takors, 2017;Lapin et al, 2004Lapin et al, , 2006Y. Liu et al, 2016) and providing information on the environment through the organism's eyes.…”