Electrical-biological hybrid system for CO₂ reduction

Highlights

• Carbon fixation by combining biological catalysis and electrocatalysis.

• A RuBisCO-independent CO2 fixation pathway was constructed.

• The pathway converts two formate and one CO2 to one pyruvate in E. coli.

Abstract

Here we have developed an electrochemical-biological hybrid system to fix CO₂. Natural biological CO₂ fixation processes are relatively slow. To increase the speed of fixation we applied electrocatalysts to reduce CO₂ to formate. We chose a user-friendly organism, Escherichia coli, as host. Overall, the newly constructed CO₂ and formate fixation pathway converts two formate and one CO₂ to one pyruvate via glycine and L-serine in E. coli. First, one formate and one CO₂ are converted to one glycine. Second, L-serine is produced from one glycine and one formate. Lastly, L-serine is converted to pyruvate. E. coli's genetic tractability allowed us to balance various parameters of the pathway. The carbon flux of the pathway was sufficient to compensate L-serine auxotrophy in the strain. In total, we integrated both electrocatalysis and biological systems into a single pot to support E. coli growth with CO₂ and electricity. Results show promise for using this hybrid system for chemical production from CO₂ and electricity.

Introduction

Global energy and environmental problems have stimulated increased efforts towards synthesizing chemicals from renewable resources, reducing greenhouse gas emissions, and utilizing atmospheric CO₂. Employing CO₂ as a substrate creates either a carbon sink or closed loop production system, both of which have numerous environmental and industrial benefits. Biological production systems have been developed to sequester CO₂ indirectly using microbial fermentation with plant biomass based sugars (e.g. corn ethanol) (Lynd et al., 1991) or directly by photosynthetic microorganisms (e.g. algae and cyanobacteria) (Wijffels and Barbosa, 2010, Case and Atsumi, 2016). However, such biological systems need to overcome many issues including low areal productivity, massive water usage, and competition with food production. Biological systems using autotrophic organisms have resulted in moderate improvements to achieve chemical production from CO₂ (Ref. Zhang et al. (2017), Fast and Papoutsakis (2012)). Inorganic methods such as electrocatalysts have been developed to reduce CO₂ to useful one carbon molecules by utilizing a wide range of novel chemistries. However, these electrocatalysts are unable to form carbon-carbon bonds (Goeppert et al., 2014, Riplinger et al., 2014, Taheri et al., 2015). Combining biological and electrochemical systems into one approach would help overcome their individual challenges. Here, in the common industrial organismEscherichia coli, we designed and constructed a biosynthetic pathway integrated with electrocatalysis to synergistically improve CO₂ fixation. This integration provides a fundamentally new approach for carbon fixation to produce valuable chemical products.

Recently a few electro-biochemical hybrid production systems have been developed with the potential to be more effective compared to their traditional biological counterparts (Li et al., 2012, Liu et al., 2016, Claassens et al., 2016). A pivotal study was reported by Li et al. (2012) wherein the chemolithotrophic bacterium, Ralstonia eutropha, was used to produce isobutanol from electrochemically reduced CO₂ via formate. While the work provided clear proof-of-concept, advancement of hybrid systems has been limited due to inadequate tools and scarce knowledge of readily engineered host organisms. This is also compounded by R. eutropha being an obligate aerobe. Oxygen is a strong electron acceptor so it inhibits the reduction of CO₂ and during electrochemical reactions can be converted to various bacterial toxins such as hydrogen peroxide (H₂O₂) (Li et al., 2012). To avoid these issues E. coli was chosen as the host for the hybrid system due to its anaerobic metabolic capabilities and native possession of many required enzymes.

Biological CO₂ fixation systems typically use autotrophs, mainly photoautotrophs, to convert CO₂ and light energy into useful chemicals such as fuel, bulk feedstocks, and pharmaceuticals (Case and Atsumi, 2016, Nybo et al., 2015), but none have achieved commercial-level titers. This is in part due to ribulose bisphosphate carboxylase-oxygenase (RuBisCO), the major autotrophic carboxylase by which > 99.5% of global inorganic carbons are fixed (Raven, 2009). Although attempts have been made for decades to increase the rate and efficiency of this key carbon fixation enzyme, improvements have been marginal and do not significantly increase the overall rate of carbon fixation (Whitney et al., 2011, Raines, 2011). To overcome this limitation we explored a RuBisCO-independent CO₂ fixation pathway in E. coli.

The constructed pathway from CO₂ and formate to pyruvate was named the reductive glycine pathway (RGP) (Fig. 1). It has been suggested that RGP is the most theoretically efficient route for formate assimilation, although RGP has not been constructed or discovered in any organisms (Bar-Even et al., 2013). RGP has the following advantages as a carbon fixation pathway: 1) it is a linear pathway, which makes improvements easier compared to circular pathways such as the Calvin-Benson (CB) cycle, 2) its reactions are thermodynamically favorable (Bar-Even et al., 2013), one of the most important factors in efficiency, and 3) it doesn’t utilize oxygen-sensitive enzymes, which represents a robust pathway functional in both aerobic and anaerobic conditions. The main challenge of RGP is its requirement of two different carbon sources, CO₂ and formate. This is why the electrocatalysis aspect is critical, as it converts CO₂ to formate, eliminating the need for two carbon sources. Thus, in our hybrid system only CO₂ is fed into the system.

To assimilate formate that is electro-chemically generated from CO₂, exogenous formate-tetrahydrofolate ligase (Fhs) (Paukert and Rabinowitz, 1980) was utilized since E. coli lacks a native Fhs. To fix CO₂ the native glycine cleavage system (GCS) (Kikuchi et al., 2008) was installed. This protein complex contains glycine dehydrogenase (GcvP), aminomethyltransferase (GcvT), lipoic acid-containing protein (GcvH), and lipoamide dehydrogenase (Lpd). GCS catalyzes the cleavage of glycine to 5,10-methylene-THF (CH₂-THF) and CO₂ (Fig. 1) (Kikuchi et al., 2008). We hypothesize that this reaction is reversible, allowing the hybrid to utilize the electrochemically derived CO₂ to produce glycine, a reversal of GCS (rGCS).

The overall hybrid RGP consists of three modules (Fig. 1). The first module is glycine synthesis from formate and CO₂ via a CH₂-THF intermediate (Fig. 1, green line). This module utilizes rGCS and a segment of the Wood-Ljungdahal pathway (WLP) (Fast and Papoutsakis, 2012, Schuchmann and Muller, 2014). The second module is L-serine synthesis from CH₂-THF and glycine (Fig. 1, blue line). In this module, formate is again converted to CH₂-THF, which is then coupled with glycine to synthesize L-serine by an L-serine hydroxymethyltransferase (GlyA) in the Serine-Glycine cycle (SGC) (Schirch et al., 1985). Recently, Yishai et al. (2017) have shown that the SGC, including GlyA, is functional in E. coli. The third module is the irreversible L-serine deamination to produce pyruvate (Fig. 1, red line). Overall, one molecule of pyruvate is synthesized from two formate, one CO₂, three NAD(P)H, and two ATP molecules. Thus, RGP requires less ATP and NAD(P)H to produce one pyruvate from CO₂ than the CB cycle, which requires five NADPH molecules and seven ATP molecules to produce one pyruvate (Bar-Even, 2016, Bar-Even et al., 2012). This is in part due to formate, a reduced form of CO₂, being more readily fixed than CO₂ thereby decreasing the required energy for carbon fixation.

Here, we present an electrochemical-biological hybrid approach to convert CO₂ to the central metabolite pyruvate. In particular, we employed rGCS for biological CO₂ fixation to avoid the utilization of RuBisCO. We integrated both electrocatalyst and biological systems into a single pot to investigate whether they are compatible with each other. This integration would improve energy efficiency for CO₂ fixation, which is a bottleneck in current biochemical production. In this study we used the engineered system to support E. coli growth, but because the product of this approach is a central metabolite, this system could be combined with almost any biochemical production pathway to produce a wide range of chemicals from CO₂ and electricity. As such, this system would be more renewable than those based on fast-growing heterotrophic hosts that use sugar precursors, and more efficient than those based on (photo)autotrophic hosts that use RuBisCO.