Current Status of Biological Production Using C2 Feedstocks

The conversion of C2 substances to acetyl-CoA involves relatively fewer steps and is often devoid of carbon loss. As a key substance in central carbon metabolism, acetyl-CoA can enter the tricarboxylic acid (TCA) cycle downstream and the gluconeogenesis pathway upstream. In biological production, most products can be further metabolized from the TCA cycle products and glycolysis products. Therefore, converting C2 substances to acetyl-CoA is the first and most important step for the utilization of C2 substances in biological systems. In this part, we will introduce the pathways of converting C2 substances to acetyl-CoA (as shown in Figure 2) and then list some examples of production using C2 substances. 3.1. Ethylene Glycol to Acetyl-CoA As an unconventional C2 carbon source, ethylene glycol can’t be utilized for growth by most conventional strains such as Escherichia coli. However, E. coli possesses endogenous pathways that can channel glycolaldehyde into the tricarboxylic acid (TCA) cycle. Therefore, the key to converting ethylene glycol into acetyl-CoA lies in transforming ethylene glycol into glycolaldehyde. In some studies, it has been found that overexpression of the fucO gene, which encodes 1,2-propanediol reductase in E. coli, can enhance the utilization efficiency of ethylene glycol by E. coli [41]. This enzyme is Fe²⁺-dependent and loses activity during aerobic growth. When the point mutations I7L and L8V are introduced, the enzyme can function under aerobic conditions [42]. In some other organisms, natural pathways for ethylene glycol utilization do exist. For example, it was found that Pseudomonas putida JM37 (isolated from soil rich in m-xylene) can naturally use ethylene glycol as a carbon source to grow, while the wild-type strain KT2440 can’t. Proteomic analysis discovered a series of genes associated with ethylene glycol utilization, among which the pedE gene encodes a PQQ-dependent alcohol dehydrogenase (pyrroloquinoline quinone-dependent alcohol dehydrogenase, PQQ-ADH) that can convert ethylene glycol to glycolaldehyde [43]. Rhodococcus jostii RHA1 can also grow using ethylene glycol. Genomic analysis identified the egaA gene, which encodes a mycofactocin (MFT)-associated dehydrogenase, as the key to ethylene glycol utilization, and it can also catalyze the oxidation of ethylene glycol to glycolaldehyde [44]. Additionally, in a study on Gluconobacter oxydans, the enzyme expressed by the Gox0313 gene was found to convert ethylene glycol to glycolaldehyde [45]. In this work, the pedE, fucO^I7L,L8V, and Gox0313 genes were introduced into E. coli to compare their abilities to utilize ethylene glycol, respectively. The results showed that the Gox0313 gene was significantly more effective than the other two. When using ethylene glycol as the sole carbon source, E. coli overexpressing the Gox0313 gene reached a cell concentration of OD₆₀₀~6 in 48 h and remained relatively stable, even with the production of glycolaldehyde [45]. After converting ethylene glycol to glycolaldehyde, for conventional engineered strains of E. coli, the endogenous pathway mainly involves the conversion of glycolaldehyde to glyoxylate via the genes aldA and glcDEF. Subsequently, two molecules of glyoxylate are condensed into tartronate semialdehyde by tartronate semialdehyde synthase encoded by the gcl gene, with the release of one molecule of CO₂. Tartronate semialdehyde is then converted to 2-phosphoglycerate via GlxRK and GarK, entering the glycolytic pathway and ultimately being converted to acetyl-CoA, with another molecule of CO₂ being released. Therefore, in the endogenous pathway, the carbon utilization rate for converting glycolaldehyde to acetyl-CoA is only 50%. For a bioproduction process, such a high carbon loss is clearly uneconomical, so finding a pathway with higher carbon utilization or even no carbon loss is crucial. Many studies have focused on this direction. In one work, researchers overexpressed the endogenous fsaA, kdsD, and rpe genes from E. coli and the pkt gene from Clostridium acetobutylicum in an E. coli with the endogenous glycolaldehyde utilization gene aldA knocked out [46]. This artificial cycle can convert glycolaldehyde to acetyl-CoA without carbon loss. It firstly fixes glycolaldehyde with glyceraldehyde-3-phosphate to form arabitol-5-phosphate, which is then converted to xylulose-5-phosphate and back to glyceraldehyde-3-phosphate with the release of acetyl phosphate. The acetyl phosphate is then converted to acetyl-CoA by the endogenous pta gene in E. coli. However, the metabolic efficiency was very low for this cyclic pathway. Isotope tracing revealed that only 16% of the target metabolites, mevalonate and acetyl-CoA, originated from glycolaldehyde. This is far from sufficient for using glycolaldehyde, or even ethylene glycol, as a carbon source for microbial growth or production. In another study, researchers proposed a novel approach for glycolaldehyde utilization [47]. They aimed to identify an enzyme that could convert glycolaldehyde directly to acetyl phosphate in a single step, thereby significantly shortening the pathway from glycolaldehyde to acetyl-CoA. Although the natural enzyme that catalyzes this reaction was not reported, some enzymes, known as phosphoketolases (PKs), can produce acetyl phosphate from fructose-6-phosphate or xylulose-5-phosphate [48]. Therefore, researchers focused on these enzymes and selected eight PKs for experimental verification following a phylogenetic tree construction using 111 bacterial species. An enzyme, BaXfspk, was identified as having high efficiency in converting glycolaldehyde to acetyl phosphate. By introducing this system into E. coli, the ability of E. coli to grow using glycolaldehyde was significantly enhanced, reaching a biomass yield of 0.681 ± 0.028 gCDW/g (g cell dry weight per g glycolaldehyde) [47]. However, research on this topic is very limited, and others are still using the endogenous glycolaldehyde conversion pathways in E. coli [41,45,49,50]. In summary, using ethylene glycol as a carbon source for microbial growth or production mainly involves integrating the oxidation of ethylene glycol and the utilization of glycolaldehyde. In previous studies, engineered E. coli strains could grow or produce target products using ethylene glycol, with yields higher than those achieved using glucose as a carbon source [50]. It is particularly notable that the endogenous glycolaldehyde utilization pathway in E. coli has a carbon utilization efficiency of only 50%. This suggests that there is still a long way to go in optimizing the use of ethylene glycol as a carbon source for bioproduction, and it also holds great potential and significant development value. It is also worth noting that the pathways for ethylene glycol utilization share many similarities with those for ethanol utilization. This implies that future work could leverage bioinformatics tools to explore ethanol metabolic pathways, particularly by identifying more enzymes that could be used in the utilization of ethylene glycol. 3.2. Ethanol to Acetyl-CoA Ethanol, as a common C2 metabolite, exists with a natural utilization pathway as a carbon source. Currently, many researchers are attempting to use ethanol as a carbon source for bioproduction, including fatty acid ethyl esters(biodiesel), squalene, and acetyl-CoA derived drugs [51,52,53,54]. The pathways they use can be roughly divided into two types: directly converting ethanol to acetyl-CoA, or first converting ethanol to acetic acid and then converting it to acetyl-CoA via the acetic acid utilization pathway. In this section, we mainly introduce the pathways for the direct conversion of ethanol to acetyl-CoA or the conversion to acetic acid, while the conversion of acetic acid to acetyl-CoA will be separately introduced in the next section. In E. coli, there is a natural pathway for converting ethanol to acetyl-CoA. This pathway is mediated by a bifunctional enzyme, AdhE, which can convert acetyl-CoA to ethanol via acetaldehyde in two steps [55]. Although the reaction catalyzed by this enzyme is reversible, in the actual growth of E. coli, due to various reasons, ethanol cannot often be converted to acetyl-CoA via this pathway, but it has been found that the AdhE mutant AdhE^A267T/E568K can enhance the ability of E. coli to utilize ethanol under aerobic conditions [56]. In a recent study, another ethanol utilization pathway in E. coli has been discovered. In this pathway, ethanol is first converted to acetaldehyde by alcohol dehydrogenase AdhP, and then converted to acetyl-CoA by the mutant of acetaldehyde dehydrogenase MhpF [57]. In another study comparing these two metabolic pathways, it was found that the pathway using Adh^EA267T/E568K not only significantly increased the growth rate of E. coli in a medium with ethanol as the sole carbon source, but also resulted in a high yield of 3-hydroxypropionic acid (3-HP) [58]. However, the production method in this study is whole-cell catalysis. Although this approach of separating bacterial growth from the production process can effectively improve the production efficiency of the product, the bacterial growth process still relies on traditional carbon sources, so there is room for further improvement in this study, such as increasing the effect of the ethanol utilization pathway and comprehensively regulating central metabolic network. In a subsequent work of this group, they used ethanol as a carbon source in the fermentation process. The yield of the target product isopropanol, increased to 67.45% of the maximum theoretical yield, which is higher than that using glucose as a carbon source [59], possibly because ethanol induced the expression of endogenous genes related to product synthesis. In addition to modifying the existing pathways in E. coli to utilize ethanol, attempts have also been made to construct artificial ethanol utilization pathways. Researchers combined the aldehyde dehydrogenase from Dickeya zeae (encoded by the ada gene) and the alcohol dehydrogenase from Saccharomyces cerevisiae (encoded by the adh2 gene) to construct a metabolic pathway in E. coli that converts ethanol to acetyl-CoA via acetaldehyde in two steps. Using this pathway, the researchers achieved the growth of E. coli with ethanol as the sole carbon source and obtained the corresponding acetyl-CoA derived products [60]. In the commonly used engineered strain S. cerevisiae, the pathway for growth using ethanol as a carbon source is naturally present. Therefore, by artificially regulating the endogenous pathways and the expression of regulatory genes in yeast to increase the production of acetyl-CoA or reduce its consumption in non-target pathways, the ability of yeast to utilize ethanol for growth and production can be effectively enhanced at the same time [53,61,62,63]. In other microorganisms, ethanol is first converted to acetic acid and then enters the endogenous acetic acid utilization pathway to further transform to acetyl-CoA. In this process, alcohol dehydrogenase (ADH) and aldehyde dehydrogenase (ALDH) play key roles. These two enzymes are present in many species and often rely on NAD⁺ or NADP⁺ as proton acceptors [64,65,66]. A novel oxidative pathway (ExaABC) derived from Pseudomonas aeruginosa has been identified, which relies on a cytochrome to oxidize ethanol [67]. However, converting ethanol to acetic acid before utilization significantly lengthens the metabolic pathway compared to direct conversion to acetyl-CoA. Therefore, such pathways are often not cost-effective and are rarely used in production studies. Of course, even for pathways that directly convert ethanol to acetyl-CoA, the most critical step is the oxidation of ethanol. The enzymes in these pathways can serve as alternative options, providing researchers with new ideas. It is important to note that these ethanol oxidation pathways may generate reducing power in the form of NADH or NADPH, or may not generate reducing power at all. While this difference may not significantly impact normal microbial growth, it becomes crucial when ethanol is used as the primary or sole carbon source for growth and production. Therefore, these different ethanol oxidation pathways and their enzymes can be selected by researchers to meet their specific needs for generating NADH, NADPH, or no reducing power, thereby better matching the production of various target compounds. 3.3. Acetic Acid to Acetyl-CoA Acetic acid can be produced naturally by many microorganisms. There are also natural metabolic pathways for utilizing acetic acid as a carbon source, which is why research on acetic acid utilization pathways is more extensive and in-depth compared to the utilization of other C2 compounds. In E. coli, there are two main pathways for acetyl-CoA utilization: the ACS pathway and the ACKA-PTA pathway. The ACS pathway is catalyzed by a single enzyme, acetyl-CoA synthetase (Acs). Under the catalysis of Acs, acetic acid first binds with ATP to form acetyl-AMP, releasing one molecule of pyrophosphate. Subsequently, the AMP group is exchanged with the CoA group to produce acetyl-CoA [68,69]. In the ACKA-PTA pathway, two enzymes are involved: acetate kinase (AckA) and phosphotransacetylase (Pta). In this pathway, acetic acid is first catalyzed by AckA to form acetyl phosphate and ADP, and then acetyl phosphate is converted to acetyl-CoA by Pta. Both pathways require an intermediate product to convert acetic acid to acetyl-CoA, but there is a significant difference between them. In the ACS pathway, the formation of pyrophosphate during the conversion of acetic acid to acetyl-AMP makes the pathway irreversible, as pyrophosphate is rapidly removed in the cell. In contrast, the ACKA-PTA pathway lacks such a constraint and is reversible. Additionally, the affinity of Acs for acetic acid is higher than that of AckA (Km values of 0.2 mM for Acs and 7 mM for AckA [70]), making Acs more suitable for acetic acid utilization. Overexpression of the ACS pathway has been shown to improve the growth of E. coli in acetate-containing media, while overexpression of AckA and Pta can inhibit cell growth [71]. This may be caused by the difference in whether the reaction is reversible. Numerous studies have been conducted on the use of acetate as a carbon source for bioproduction. Since acetate is directly converted to acetyl-CoA upon assimilation, it has a natural advantage as a carbon source for the production of acetyl-CoA-derived products. For example, a strain of E. coli was engineered to produce isopropanol from acetate, achieving a yield of 13.3 g/L in a bioreactor with a stable acetate concentration of 30 mM [72]. In another study, 2.54 g/L of γ-aminobutyric acid was produced from 5.91 g/L of acetate [73]. Both studies used the E. coli W strain, which is known for its better acetate utilization compared to other common strains such as BL21 (DE3), W3110, and MG1655[71]. The W strain was originally isolated from soil near Rutgers University in 1943 [74] and has been found to have better tolerance under various stress conditions [75,76]. These studies suggest that the W strain may be a naturally optimized strain for acetate utilization, potentially eliminating the need for overexpression of acetate utilization pathways. In addition to E. coli, other microorganisms with high acetate tolerance and utilization capabilities have been identified, such as Corynebacterium glutamicum [77]. For instance, by introducing relevant pathway enzymes and regulating carbon flux-related genes, C. glutamicum was able to produce itaconic acid from 20 g/L of acetate, achieving a yield of 116 mM, which is 35% of the maximum theoretical yield [78]. This indicates that selecting naturally acetate-tolerant and -utilizing strains as hosts may achieve or even surpass the effects of artificially engineered common strains for acetate-based production. 3.4. Acetyl-CoA back to Gluconeogenesis Pathway and Other Metabolic Regulation C2 carbon sources can be more directly converted into acetyl-CoA, which is advantageous for bioproduction but also brings some challenges. Not all high-value compounds in biological production are derivatives of acetyl-CoA. For example, the production of various aromatic compounds in E. coli requires the shikimate pathway, which begins with the condensation of phosphoenolpyruvate (PEP) and erythrose-4-phosphate (E4P). Clearly, PEP is an upstream metabolite of acetyl-CoA in glycolysis, while E4P is a product of the pentose phosphate pathway. Therefore, when producing such compounds, it is necessary to replenish acetyl-CoA into the gluconeogenesis pathway (as shown in Figure 3). This allows microorganisms to use C2 carbon sources to produce a wider variety of high-value products, rather than being limited to acetyl-CoA derivatives. It was discovered that overexpression of the ydbK gene enhances the ability of E. coli to utilize acetate for the production of non-acetyl-CoA derivatives [79]. The ydbK gene encodes pyruvate-ferredoxin oxidoreductase, which catalyzes the reversible reaction between pyruvate and acetyl-CoA with CO₂. Although overexpression of ydbK effectively improved bacterial growth and product yield, subsequent research found that co-overexpression of pckA and maeB yielded even better results [80]. The pckA gene encodes phosphoenolpyruvate carboxykinase, which catalyzes the conversion of oxaloacetate to phosphoenolpyruvate. The maeB gene encodes NADP⁺-dependent malate dehydrogenase, which catalyzes the conversion of malate to pyruvate. However, in this study, the pckA and maeB genes were overexpressed simultaneously, but neither was more effective than ydbK. It is also worth noting that, unlike the other two genes, the pathway catalyzed by the enzyme encoded by ydbK is involved in carbon fixation rather than carbon loss. Therefore, this pathway may potentially offer better performance. When ethylene glycol is used as a carbon source, if it is converted to glycolaldehyde and then assimilated via the endogenous pathway of E. coli, it can directly enter the glycolysis/gluconeogenesis pathway in the form of 2-phosphoglycerate, thus eliminating the need for additional acetyl-CoA supplementation. Among the other methods for acetyl-CoA replenishment, except for the overexpression of the ydbK gene, the substrates used are derived from the TCA cycle. Withdrawing these substances from the TCA cycle may severely disrupt the normal function of the TCA cycle, thereby affecting bacterial growth. Therefore, the glyoxylate cycle becomes an attractive alternative. In the key reaction of the glyoxylate cycle, one molecule of isocitrate and one molecule of acetyl-CoA are converted into one molecule of malate and one molecule of succinate. Thus, without considering the conversion of substances outside the TCA cycle, the glyoxylate cycle can increase the number of substances within the TCA cycle. Therefore, appropriately regulating the expression levels of the glyoxylate cycle and the replenishment pathways can achieve the replenishment of acetyl-CoA into the gluconeogenesis pathway without compromising the stability of the TCA cycle. Some studies have also shown that using promoters with different strengths to regulate the intensity of these two pathways can effectively increase the yield of target products [81,82]. Additionally, the regulatory gene of the glyoxylate cycle, such as iclR [83], can be knocked out to enhance the carbon flux through the glyoxylate cycle. However, another study suggests that the knockout of the iclR gene may reduce the carbon flux through the pentose phosphate pathway [84], which could be disadvantageous for the shikimate pathway that relies on products from the pentose phosphate pathway. Therefore, directly regulating the expression of key enzyme genes in the glyoxylate cycle, such as aceA [85] or icd, may be a better approach. When ethylene glycol and ethanol are used as carbon sources, their conversion to acetyl-CoA requires oxidation reactions, which may lead to an increased demand for oxygen by the bacteria. To satisfy the O₂ demand, overexpression of oxygen-binding proteins such as vitreous hemoglobin (VHb) may be helpful [86]. Additionally, it should be noted that these three C2 carbon sources can inhibit bacterial growth at high concentrations. In E. coli, acetate at concentrations above 5 g/L can inhibit growth [87], while ethanol shows an inhibitory effect just at a low concentration of 2% (v/v). There is less research on ethylene glycol, and the specific toxic concentration has not been clearly identified, but it may likely be similar to ethanol. These concentrations are too low for bioproduction. Therefore, to utilize these carbon sources for bioproduction, it is necessary to enhance the tolerance of the host strain to the carbon sources. To improve the tolerance of strains, many methods, such as adaptive laboratory evolution (ALE) or targeted tolerance-related modifications based on omics studies, have been developed [88]. Given the limited understanding of global metabolism in biology, irrational adaptive evolution is a more commonly used and convenient method. The important aspect of adaptive evolution is the introduction of mutations at a desired degree. The atmospheric and room temperature plasma (ARTP) mutagenesis has been proved to be highly efficient for kinds of strains [89]. Additionally, new methods such as targeted artificial DNA replisome (TADR) [90], which introduce mutations in bacteria during growth, make adaptive evolution more efficient and sustainable. To efficiently produce target products using C2 carbon sources, many other regulations are possibly involved, such as reducing the loss of key substances through side pathways [91], regulating the ATP/ADP and NAD(P)⁺/NAD(P)H ratios, and controlling specific metabolic pathways. These are more related to the specific target products and will not be elaborated on here.