Energetic constraints of metal-reducing bacteria as biocatalysts for microbial electrosynthesis

Abstract

Background: As outlined by the Intergovernmental Panel on Climate Change, we need to approach global net zero CO₂ emissions by approximately 2050 to prevent warming beyond 1.5 °C and the associated environmental tipping points. Future microbial electrosynthesis (MES) systems could decrease net CO₂ emissions by capturing it from industrial sources. MES is a process where electroactive microorganisms convert the carbon from CO₂ and reduction power from a cathode into reduced organic compounds. However, no MES system has attained an efficiency compatible with a financially feasible scale-up. To improve MES efficiency, we need to consider the energetic constraints of extracellular electron uptake (EEU) from an electrode to cytoplasmic electron carriers like NAD⁺. In many microbes, EEU to the cytoplasm must pass through the respiratory quinone pool (Q-pool). However, electron transfer from the Q-pool to cytoplasmic NAD⁺ is thermodynamically unfavorable. Here, we model the thermodynamic barrier for Q-pool dependent EEU using the well-characterized bidirectional electron transfer pathway of Shewanella oneidensis, which has NADH dehydrogenases that are energetically coupled to proton-motive force (PMF), sodium-motive force (SMF), or uncoupled. We also tested our hypothesis that Q-pool dependent EEU to NAD⁺ is ion-motive force (IMF)-limited in S. oneidensis expressing butanediol dehydrogenase (Bdh), a heterologous NADH-dependent enzyme. We assessed membrane potential changes in S. oneidensis + Bdh on a cathode at the single-cell level pre to post injection with acetoin, the substrate of Bdh.

Results: We modeled the Gibbs free energy change for electron transfer from respiratory quinones to NADH under conditions reflecting changes in membrane potential, pH, reactant to product ratio, and energetically coupled IMF. Of the 40 conditions modeled for each method of energetic coupling (PMF, SMF, and uncoupled), none were thermodynamically favorable without PMF or SMF. We also found that membrane potential decreased upon initiation of EEU to NAD⁺ for S. oneidensis on a cathode.

Conclusions: Our results suggest that Q-pool-dependent EEU is both IMF-dependent and is IMF-limited in a proof-of-concept system. Because microbes that rely on Q-pool-dependent EEU are among the most genetically tractable and metabolically flexible options for MES systems, it is important that we account for this thermodynamic bottleneck in future MES platform designs.

Fig. 1

A Extracellular electron uptake (EEU) through the respiratory quinone pool (Q-pool). B EEU that bypasses the Q-pool via H₂ diffusion to soluble cytoplasmic hydrogenases. Note that these examples assume either a wildtype or engineered organism capable of EEU to the cytoplasm. Many species in these categories will not be capable of EEU in their wildtype form

Fig. 2

A EEU in S. oneidensis from an electrode to NAD⁺, where PMF is supplied by a proton-pumping terminal oxidase. The thermodynamically distinct paths of electron transfer to three differently coupled NADH dehydrogenases (Nuo, Nqr, and Ndh). B Free energy (ΔG) associated with each step of the pathway in (A). For Step 7, a negative free energy means that EEU across the quinone pool is thermodynamically favorable and a positive free energy means that EEU across the quinone pool to cytoplasmic NAD⁺ is thermodynamically unfavorable. MQ, menaquinone-7; MQH₂, menaquinol-7; FMN, flavin mononucleotide; FMNH₂, reduced flavin mononucleotide; STC, small tetraheme cytochrome; MtrABC, proteins of the Mtr pathway; Bdh, butanediol dehydrogenase; bc1, bc1 complex; e-, electron [68]

Fig. 3

An example of our experimental setup and analysis for the microscopy-compatible bioelectrochemical system (BES). A Diagram of our BES, the design of which was based on Pirbadian et al. (2020). To minimize O₂ changes in the BES during injection, the BES was connected to an anaerobic bottle containing a media reservoir and head space. On the right-hand side is a picture of our BES setup. The aluminum foil on the bottle was present to protect ThT in the media from photodegradation while N₂ bubbled into the media bottle overnight. B For each timepoint, the background fluorescence was subtracted in ImageJ. Regions of interest (ROI) were defined for individual cells, or small groups of cells, that fit into the defined area of 0.25 to 20 µm². For each timepoint of a biological replicate, the mean ThT intensity was the background-subtracted mean fluorescence intensity of the ROIs. Some cell drift out of the field of view (FOV) was uncontrollable, however, we strove to maintain the FOV for each biological replicate to keep the laser exposure time, and any resulting changes in membrane permiability to ThT, consistent for each timepoint

Fig. 4

Gibbs free energy predicted for Reaction 1 under various biologically relevant conditions (non-extremophile). In (A–C) a diagram of the associated EEU pathway is on the left and the results of our thermodynamic model are on the right. Each heatmap shows the ΔG calculated for Reaction 1, where the Q-pool is either primarily oxidized or reduced, 0.1% reduced and 90% reduced, respectively. For both an oxidized and reduced Q-pool, we modeled various combinations of membrane potential and periplasmic pH. As Reaction 1 occurs across a membrane, we calculated ΔG using Eqs. 1 and 2. ‘Mtr’ represents the MtrA, MtrB, and MtrC proteins from the Mtr pathway. Small tetraheme cytochromes (STC) are electron carriers in the periplasmic space. ‘FMN’ and ‘FMNH₂’ are the oxidized and reduced forms of flavin mononucleotide, respectively. Arrows indicate the direction of electron flow (excluding the ion-translocating arrows). The quinone pool (Q-pool) contains menaquinone (MQ) and menaquinol (MQH₂)

Fig. 5

ThT fluorescence and electrical current simultaneously measured in microscopy-compatible BESs. A, B Applied electrode potential (yellow), ThT fluorescence (green), and current (purple) plotted by time. In (A) the two injections were acetoin and CCCP. B The same experimental procedure performed with the solvent controls injected, water and ethanol respectively. C Mean ± SEM (standard error of the mean) for the pre- to post-injection difference in ThT fluorescence (left, green) and electric current (purple, right). In (C) ‘Intra-experiment’ refers to the pre- to post-injection change, for either acetoin or water within a single biological replicate; ΔThT or Δcurrent are the mean ± SEM of three biological replicates for either the acetoin injection (Acetoin, n = 3) or the solvent control (Water, n = 3). The * symbol indicates statistical significance