Generic biomass estimation methods targeting physiologic process control in induced bacterial cultures

Abstract

Within this contribution we have outlined and compared four generic applicable methods for biomass estimation in early bioprocess development. The accuracy and robustness of a hard sensor, softsensor and hybrid sensor were discussed based on the coefficient of variation of the root mean squared error (cvRMSE) for two strains and three different levels of metabolic activity. This comparison facilitates a comprehensive overview of appropriate methods for biomass estimation in bioprocess development. Depending on the scope of the planned experiments as well as on the available infrastructure and historic data, the outlined data ease method selection. Hereby, we aim to alleviate physiologic bioprocess development requiring physiological process control which is necessarily based on real time biomass estimation.www.els-journal.com Page 2 Engineering in Life SciencesThis article is protected by copyright. All rights reserved. 2 AbstractAdvanced bioprocess development strategies focus on the control of physiological entities which rely on accurate real time determination of biomass concentration. Various methods have been proposed in literature but up to this date a comprehensive and differentiated comparison of biomass estimation approaches for early stage bioprocess development is missing. In this contribution, we compared hard sensor, soft-sensor and data-driven approaches for real-time biomass estimation in respect to accuracy, transferability and costs. The outlined methods were tested with two different microbial strains and recombinant products using E. coli. To investigate the applicability of the outlined methods, method performance was assessed in correspondence to metabolic activity. Based on statistical descriptors the methods were compared and discussed. The results indicate no significant impact of strain or biomass estimation approach on the measurement quality. The average relative error of 11-13% can be greatly reduced by over 85% combining the outlined methods by the means of weighted average. This approach proved to be highly robust even during highly dynamic process conditions of oscillating specific substrate uptake rates. Concluding, the combination of low cost first principle soft-sensor approaches in combination with a hybrid soft-sensor yields the best information to effort ratio.

Inclusion Body Bead Size in E. coli Controlled by Physiological Feeding

“…At high q s,C early C-source accumulation could be observed with present product degradation. For easy comparison between different physiological feeding strategies and different used C-sources, the cumulative sugar uptake value, dSn [g/g], (Equation (2)) could be used making a normalization on the fed mass C-source in respect to the total biomass at the induction time [34]:normaldnormalSnormaln=∫0tmCnormaldnormaltnormalX(t0) with m C [g] being the fed mass of C-source, dt [h] the respective time interval and X(t 0 ) [g] the total biomass at the start of induction. The induction time scale can now be exchanged for the dSn value during the induction phase.…”

“…Exchanging a static feeding strategy with the feeding control shown in Figure 3a, does implement a size growth even at late induction times and is triggering maximum size towards the time-point of harvest. Optimized feeding strategies, varying µ over induction time, seem to boost the overall product titer [34]. The probability density plot shown in Figure 3b presents the size as a function of induction time for the given run.…”

“…Within this study, it was also shown that a stepwise increase, such as a q s -ramp, is superior to a constant feeding profile in Pichia pastoris protein production. The influence of q s ramps onto inclusion body formation was performed by Reichelt et al [34]. Within this study also the term “q s,crit ” was determined, describing the maximum physiological substrate uptake rate.…”

Perspectives of inclusion bodies for bio-based products: curse or blessing?

“…Physiological feeding strategies are based on the specific substrate uptake rate ( q s,C ) of the respective C-source and can be exclusively performed in the controlled environment of a bioreactor. Cell stress and reduction of cell viability during the induction lead to a reduced uptake of C-source and therefore to a possible overfeeding during induction (Reichelt et al 2016, Slouka Christoph 2018). Adapting of q s,C based on the physiological state of the cell is essential for preventing cell death and IB degradation during cultivation.…”

Physiological capacities decline during induced bioprocesses leading to substrate accumulation

“…Nevertheless, in the setting of bioprocess development, historic process data are commonly scarce, which impairs the use of data‐based algorithms. Consequently, hard type sensor or first principle mass balance based approaches are regarded as more feasible, especially in the early stages of bioprocess development .…”