process (Siegert and Banks, 2005). Hence, stimulating the protein and SCFAs biodegradation is of great interest to assist lipid extractions. Previous studies have reported that MECs have the potential to accelerate the fermentation of protein along with other complex organics, such as SCFAs (Lu et al., 2012; Velasquez-Orta et al., 2009). Furthermore, with the increased removal of protein, hydrogen bonds between membrane proteins and lipids can be disrupted, rupturing the cell membrane and exposing intracellular lipids for much simpler extraction (Cooney et al., 2009; Sheng et al., 2011). Hence, this was the main motivation behind coupling MEC with SF, also known as the electro-selective fermentation (ESF), to significantly enhance lipid wet-extractions (Liu et al., 2019; Liu et al., 2020b; Liu et al., 2020c).
Recently, several studies have successfully demonstrated superior lipid extraction from microalgae using the ESF (Liu et al., 2019; Liu et al., 2020b; Liu et al., 2020c). Liu et al. (2019) first investigated the relative performance of ESF compared to SF, in terms of protein and COD removal. When the ESF and SF (Control) reactors were fed with microalgae biomass (Scenedesmus acutus), significant differences were found in TCOD removal (12% vs. 2%), TCOD-to-SCOD conversion (17% vs. 11%), and protein biodegradation (42% vs. 10%) for ESF and SF, respectively. Interestingly, although the anodic respiration only contributed to < 1% of the total electron input, the ESF system yielded much higher protein and COD degradations. More importantly, even with the greater loss of total lipids due to β-oxidation-linked-to-biohydrogenation (e.g., shifting LCFAs from C18:1 to C16:0 or C14:0) in the ESF system, it still demonstrated significantly higher lipid yields (37%) compared with the SF (12%) and the lipid extractions with hexane-isopropanol solvents (~4%). In summary, despite currently being a small component of overall COD balance, ESF exhibited superior hydrolysis of protein of microalgae biomass (Scenedesmus acutus), a higher saturation ratio of LCFA, and higher lipid yields (~3 folds) over SF (Liu et al., 2019). On the other hand, Liu et al. (2020c) have introduced a flat-plate MEC based ESF to observe the performance in lipid extraction by having a large anode surface area for electroactive bacteria biofilm. The benefit was in the enhanced electron consumptions, where the electrons in the microalgae biomass were greatly scavenged by electroactive bacteria in the anode biofilm. Furthermore, the SCFAs were also efficiently converted to current (e.g., preventing the accumulation of SCFAs), maintaining the optimal working environment (e.g., preventing pH drop) for the fermentation process. Compared with Liu et al. (2019), the ESF significantly improved in microalgae lipid extraction, achieving 56% (vs. 37%) extraction from the total lipids.
Figure 1.5 A conceptual schematic showing lipid extraction from microalgae using electro-selective fermentation. The figure is drawn with modification after Liu et al. (2019).
Finally, Liu et al. (2020b) have compared various SRTs on the ESF performance. In their study, ESF systems with different SRTs were compared (System A: start with a 6 d SRT, then switch to 2 d SRT vs. System B: start with a 2 d SRT, then switch to 6 d SRT). The ESF system A exhibited the highest lipid extractability (25%) when operated with a 6 d SRT, and gave the highest lipid productivity (450 mg/L/d) when switched to a 2 d SRT. This was attributed to establishing the microbial community containing protein fermenters at a 6 d SRT and washing out of lipid fermenters when the system A was switched to the 2 d SRT. Opposingly, the system B failed to enhance the lipid extractability. It initially (e.g., at a 2 d SRT) washed out the lipid fermenters, but it was unable to recover them when the system switched to a 6 d SRT. Ultimately, demonstrated by multiple studies, the further studies are warranted to bring this technology forward towards further development, scaling-up and potential commercialization.
1.3.5 Acetoin
Acetoin is known to be a volatile compound, which is widely utilized in chemical synthesis, cigarettes, cleaning products, food, personal care products, plant growth, pest controls, wine, etc. (Cheynier et al., 2010; Förster et al., 2017; Xiao and Lu, 2014). Especially, acetoin has been largely applied as food flavoring (e.g., has an intense butter flavor) and fragrance manufacturing (Xiao and Lu, 2014). Moreover, acetoin is also highly promising, ranking in the top 30 for its potential applications as the bio-based building block precursors (Werpy and Petersen, 2004), which can embrace a green environment (e.g., petro-free economy). Due to these several industrial applications, in recent years, the production of acetoin became highly desirable due to a significantly large annual consumption across the globe (e.g., 2-4 mg per day per person, equal to several thousand tons per year worldwide) (Xiao and Lu, 2014).
In general, acetoin is a substance that cannot be generated as the sole end product during the anaerobic fermentation due to being more oxidized form than other typical substrates, such as glucose (Förster et al., 2017). Hence, the acetoin is generated under a specific condition during fermentation by Bacillus subtilis and serval Enterobacteria species, for instance: (1) occurs as an intermediate in the formation of 2,3-butanediol in butanediol fermentation; (2) produced in a two-step process from pyruvate (Cheynier et al., 2010; Förster et al., 2017; Härtig and Jahn, 2012; Krieg and Padgett, 2011). Briefly describing the two-step process, the acetolactate synthase (AlsS) first catalyzes the condensation of two molecules of pyruvate to acetolactate, where one molecule of CO2 is released during this reaction. Then, acetolactate is decarboxylated by the acetolactate decarboxylase to produce acetoin (Nicholson, 2008; Ramos et al., 2000).
To date, several limitations in acetoin synthesis have been reported. For instance, only low acetoin yields can be achieved using wild-type strains because the majority of other products are transformed to 2,3-butanediol (Förster et al., 2017). Hence, significant efforts have been dedicated to engineering the strains (e.g., Bacillus subtilis, Serratia marcescens, Clostridium acetobutylicum, Candida glabrata) to maximize acetoin output (Bai et al., 2015; Li et al., 2015; Liu et al., 2015; Zhang et al., 2016). However, despite the high acetoin yields (e.g., ~70-96% of maximum theoretical), acetoin-producing strains can be pathogenic, which may limit their use in the markets without appropriate post-treatment (Bursac et al., 2017; Förster et al., 2017; Xiao & Lu, 2014). Moreover, the related studies were performed using oxic (either anoxic or aerobic) processes, which are highly unfavorable for the biosynthesis of value-added products due to the requirement of aeration and a higher ratio of anabolism over catabolism (Weusthuis et al., 2011). Therefore, a more suitable approach that is capable of producing a high amount of environmental-friendly acetoin is greatly needed.
In response, studies have been implemented to produce acetoin using EF systems (Bursac et al., 2017; Förster et al., 2017) (see Table 1.1). Bursac et al. (2017) studied Shewanella oneidensis to produce acetoin from lactate in an EF system. In their study, first, the Shewanella strains were engineered for the acetoin synthesis, where the Bacillus subtilis genes, capable of acetolactate synthase and acetolactate decarboxylase, were codon-optimized and integrated into pBAD202 plasmid, and then the plasmid was transformed into different Shewanella stains. Using the engineered strains, acetoin production of 40% of the theoretical maximum was achieved by complete degradation of lactate (50 mM) after 48 hours of fermentation. To further exhibit higher acetoin production, their study developed the strains again (e.g., deleting the -prophage to increase cell adhesion on the anode) and introduced an EF system. Hence, the EF system with engineered strains demonstrated higher acetoin yield (78% of the maximum theoretical vs. 40% in conventional fermentation) (Bursac et al., 2017). On the other hand, Förster et al. (2017) investigated acetoin production in an EF system operated with glucose. Their study chose Escherichia coli because it was found to be the most suitable host organism, due to its versatile metabolism and genetic tractability (Nielsen et al., 2010; Xu et al., 2015; Xu et al., 2014). Similar to Bursac et al. (2017), Förster et al. (2017) engineered the strain, such as the quadruple
