Experimentally implemented dynamic optogenetic optimization of ATPase expression using knowledge-based and Gaussian-process-supported models

TL;DR

The paper addresses optimizing dynamic ATP turnover by optogenetically regulating ATPase in E. coli lactate fermentation. It introduces a simplified, unsegregated macro-kinetic model augmented with Gaussian processes to handle model uncertainty and enables an open-loop optimal-control framework constrained by experimental data. Experimental validation shows that light-driven ATPase induction can modestly increase lactate yield relative to zero-light conditions, with hybrid models offering enhanced predictive capability. The work demonstrates a practical, scalable approach toward metabolic cybergenetics, paving the way for more robust, feedback-based control strategies in bioprocess optimization.

Abstract

Optogenetic modulation of adenosine triphosphatase (ATPase) expression represents a novel approach to maximize bioprocess efficiency by leveraging enforced adenosine triphosphate (ATP) turnover. In this study, we experimentally implement a model-based open-loop optimization scheme for optogenetic modulation of the expression of ATPase. Increasing the intracellular concentration of ATPase, and thus the level of ATP turnover, in bioprocesses with product synthesis coupled with ATP generation, can lead to increased product formation and substrate uptake. Previous simulation studies formulated optimal control problems using dynamic constraint-based models to find optimal light inputs in fermentations with optogenetically mediated ATPase expression. However, using these models poses challenges due to resulting bilevel optimizations and complex parameterization. Here, we outline a simplified unsegregated and quasi-unstructured kinetic modeling approach that reduces the number of dynamic states and leads to single-level optimizations. The models can be augmented with Gaussian processes to compensate for model uncertainties. We implement optimal control constrained by knowledge-based and hybrid models for optogenetic ATPase expression in $\textit{Escherichia coli}$ with lactate as the main product. To do so, we genetically engineer $\textit{E. coli}$ to obtain optogenetic expression of ATPase using the CcaS/CcaR system. This represents the first experimental implementation of model-based optimization of ATPase expression in bioprocesses.

Figures (5)

• Figure 1: Open-loop control of ATP turnover via optogenetic modulation of the ATPase in batch processes. In gray we show other potential configurations such as closed-loop control and fed-batch fermentations, although these fall out of the scope of this study.
• Figure 2: A) Simplified core metabolism of the microorganism used in this study, i.e., E. coli sGB015 with enforced ATP wasting. It shows the conversion of glucose into lactate. Relevant redox and energy co-factors are presented. B) The light-inducible ATP wasting is managed by three heterologous genetic elements in E. coli sGB015: pPLPCB(S), pGB-MPI-23 and the chromosomal insertion of PcpcG2$\Delta$59-atpAGD-rrnBT1. Plasmids are shown circular, while the chromosome is shown linear. Note that genes (italicized) and their related proteins (non-italicized) are shown with the same color. pPLPCB(S) expresses ho1 (dark blue) and pcyA (purple), thereby enabling the conversion of heme (red pentagon) into phycocyanobilin (green pentagon), the chromophore necessary for CcaS to detect light. The expression of ccaS (orange) and ccaR (pale blue) is enabled through pGB-MPI-23. CcaS autophosphorylates after a conformational change induced by green light with the photon-protein interaction enabled through phycocyanobilin. Afterward, CcaS phosphorylates CcaR (phosphate group P is represented by circles), leading to CcaR dimerization and functioning as a transcription factor for PcpcG2$\Delta$59 on the chromosome, thereby initiating the expression of atpAGD (red). Promoters are presented as arrows on the plasmids and the chromosome. Open reading frames are shown as arrows next to their related genes. Terminators are marked as black perpendicular lines. Origins of replication are shown in yellow and antibiotic resistances in pale green. C) Scheme of the fermentation setup with the green and red light delivery system based on LEDs. D) Photograph of the actual setup shown in C). The temperature regulation chamber, magnetic stirrer plate, LED arrays, and actuation system are labeled. The actuation system is shown with the two tunable regulators for determining the current that is delivered to the green and red LED arrays.
• Figure 3: A) Average lactate on glucose yield $Y_\mathrm{LG,batch}$, B) biomass on glucose yield $Y_\mathrm{BG,batch}$, and C) lactate volumetric productivity $r_\mathrm{L,batch}$. Blue bars: experiments used to model the system. Orange bars: predicted open-loop optimization results using the nominal (OLO_nom_pred) and hybrid (OLO_hyb_pred) models. Green bars: actual experimental results for the open-loop optimizations using the nominal (OLO_nom_exp) and hybrid (OLO_hyb_exp) models. These metrics correspond to average batch results, i.e., considering initial and final points. The units of $u_l$ are $\mu$mol$\ $m$^{-2}\ $s$^{-1}$.
• Figure 4: Fitting of the nominal and hybrid models to the experimental data under different constant light inputs.
• Figure 5: Results of the open-loop optimization prediction and experimental implementation. The optimizations based on the nominal and hybrid models are shown on the left and right sides, respectively.