Review Open Access
Thermoanaerobacter Species: The Promising Candidates for Lignocellulosic Biofuel Production
Received: 31 January 2023 Accepted: 09 March 2023 Published: 15 March 2023
© 2023 The authors. This is an open access article under the Creative Commons Attribution 4.0 International License (https://creativecommons.org/licenses/by/4.0/).
1. Introduction
Fossil fuel, consumed worldwidely, is the non-renewable energy, the exhaustion of which has inspired the interest in the environmentally friendly energy—biofuels [1,2]. The usage of biofuels such as bioethanol can reduce fossil fuel consumption, cut down the emission of SO2 and NOX, and consequently do good to the environment [3]. Lignocellulose is seen as an abundant, sustainable and low-cost resource on the earth for biofuels production, which can not only avoid the conflict between food and fuels, but also help to deal with the globally abundant agricultural residues to protect the local environment [4,5]. Lignocelluloses such as wheat straw, rice straw and sugarcane bagasse were adopted for biofuels production [6,7,8], suggesting bio-processing of biomass is widely accepted due to the good perspective [9]. Unfortunately, biofuel engineers must face the fact that cellulosic ethanol still cannot match the low cost of petro-chemicals and first-generation (1G) ethanol (usually produced from cereal grains, sugar and amylaceous plants, popularly in USA and Brazil) [5,10]. The second and third-generation ethanol is obtained from lignocellulosic materials and algae, respectively, while the latter is not well developed due to some limited factors like the technology for extracting bioethanol from seaweed [11]. The conversion of lignocellulosic feedstock into biofuels at competitive prices depends on the efficient conversion of lignocellulosic biomass. In order to achieve this purpose, changes need to be done urgently, such as reducing the enzyme costs, application of enzyme-free saccharification processes, enforcing the co-fermentation of mixed sugars for biofuels production [5,12,13,14].
Generally, lignocellulose mainly contained lignin (15–20%), cellulose (40–50%) and hemicellulose (25–35%) [15], which could be decomposed to monosaccharides and disaccharides then followed by biofuels conversion [16]. Traditional method for lignocellulose pretreatment aims at obtaining glucose by cellulases, including endoglucanase, exoglucanase and β-glucosidase. However, complete hydrolysis of lignocellulose faces many challenges in biomass application [17,18]: (1) β-glucosidase is the rate limiting enzyme for cellulose hydrolysis and large amount usage of β-glucosidase will increase the cost; (2) Glucose in the hydrolysate is an inhibitor of cellulase, which leads to low efficiency of complete hydrolysis of cellulose; (3) The presence of glucose in the complete hydrolysate can not only easily cause the microbial pollution, but also induce the carbon catabolite repression (CCR), prolonging the fermentation period. Therefore, pretreatment with incomplete hydrolysate of lignocellulose began to emerge [19,20].
Among microorganisms that can ferment biomass to biofuels, thermophilic anaerobic bacteria, with an optimal growth temperature range of 45–70 °C, have some advantages over mesophilic bacteria [21], such as the ability to co-utilize glucose and xylose by Thermoanaerobacter species [22,23] (they are further subdivided into the genus Thermoanaerobacter and the genus Thermoanaerobacterium due to some differences such as the absence or presence of the nitrogenase genes [24]), lower risk of the contamination for high temperature fermentation [25], less heat exchange following pretreatment and higher economy to recover ethanol at temperatures over 50 °C by continuous distillation [26]. Another common and well-studied thermophile is Clostridium thermocellum, which could degrade cellulose but the titer of biofuels produced was low [27,28,29]. Here, we mainly focus on Thermoanaerobacter species, which might be the potential chassis microorganisms for lignocellulosic biorefinery with the ability to directly utilize hemicellulose or even cellulose [30,31]. This review will comprehensively discuss the utilization of substrates, genetic operation platforms constructed and the strategies employed for enhancing the biofuel production of Thermoanaerobacter species (Figure 1). Additionally, the potential of Thermoanaerobacter species co-cultured with C. thermocellum for consolidated bioprocessing (CBP) will be also addressed.
2. Excellent Performance on Substrates Utilization
2.1. Xylan and Xylose Metabolism in Thermoanaerobacter Species
Xylan is the second most abundant plant polysaccharide on the earth, and the D-xylose is the primary ingredient of the hemicellulose from lignocellulosic biomass, both of which are not the common or preferred substrates for most microorganisms [32]. Xylose metabolism appears to occur exclusively via the traditional xylose isomerase-xylulokinase-pentose phosphate pathway in Thermoanaerobacter species [33].
As shown in Figure 2, the genus Thermoanaerobacter has been reported to be an excellent utilizer of xylan and xylose [30,34], the feature of which is of particular industrial interest. In Thermoanaerobacterium aotearoense SCUT27 (SCUT27), results of the genome annotation revealed 10 genes related to xylose utilization, including one xylose isomerase, one xylulose kinase, and 8 ABC-type xylose transporters [35]. Compared to T. thermosaccharolyticum DSM571, strain SCUT27 possesses two more D-xylose ABC transporter-related genes for xylose utilization. Strain SCUT27 also can degrade xylan directly and four genes related to xylan degradation, such as one gene for β-1,4-xylanase, one for β-xylosidase, and the other 2 for xylanase/chitin deacetylase were found in the genome [35]. In strain SCUT27, xylanase and β-xylosidase that belong to the glycoside hydrolase family 10 (GH10) have been cloned and characterized [36,37]. The function of Tsac_1445 and Tsac_1464 for xylan/xylose metabolism was also characterized in T. saccharolyticum [38]. Additionally, xylose delays cell lysis by sustaining coenzyme and ion metabolism, thus enhancing the viability of the thermophiles [23]. Most of mesophilic strains use glucose preferentially to xylose owing to CCR. However, Thermoanaerobacter species can co-utilize xylose and glucose without CCR [22,39] as discussed below.
2.2. Glucose and Xylose Co-Fermentation in Thermoanaerobacter Species
When cultured with mixed carbon source, microorganisms would selectively use preferred carbon sources by preventing the expression of enzymes participating in secondary substrates [19]. This regulation is called CCR, which benefits microorganisms for competitions in natural environments but prolongs the fermentation periods and increases the costs for industrial applications. Co-fermentation of multiple carbon sources for the production of high value-added biofuels is a promising production strategy. Among the abundant microorganisms in nature, thermophilic anaerobic bacteria such as Thermoanaerobacter sp. X514, T. saccharolyticum and strain SCUT27 were reported to have the native ability to utilize glucose and xylose simultaneously [22,23,39]. To reveal molecular mechanism of the glucose and xylose co-utilization in Thermoanaerobacter sp. X514, the glycobiome was performed and analyzed via two-dimensional transcriptome sequencing and results revealed that the functional modules of glucose and xylose metabolism were independent and regulated by different transcriptional anti-terminators BglGs, which was quite different from xylose transporters regulated by XylR in Escherichia coli, Bacillus subtilis and Clostridium tyrobutyricum [23]. However, the functional modules of glucose and xylose metabolism regulated by BglGs independently in Thermoanaerobacter sp. X514 could not perfectly explain the release of CCR between glucose and xylose. It’s reported that BglG family transcription antiterminators are key players in the control of carbohydrate catabolism in bacteria [40,41]. These antiterminators are generally composed of three modules, an N-terminal RNA-binding domain (CAT) followed by two homologous regulation modules PRD1 and PRD2 that control the RNA binding activity of the effector via phosphorylation on the conserved histidines [41]. LicT is one member of BglG family transcription antiterminator. The phosphorylation of LicT in PRD1 is catalyzed by P~EIIBBgl in the absence of β-glucosides and inhibits the activity of LicT. In contrast, when β-glucosides are present, P~EIIBBgl is preferably transferring the phosphate group to β-glucosides with the PRD1 of LicT unphosphorylated. However, this format of LicT is also non-functional and requires P~His-HPr to phosphorylate PRD2, in the absence of preferred carbon sources such as glucose. When His-100 and His-169 in PRD1 are unphosphorylated, and one or both conserved His in PRD2 are phosphorylated, LicT is activated and binds bound to RNA on the ribonucleic antiterminator to stabilize an RNA secondary structure for the successful transcription as shown in Figure 3. The transcription of operons mediated by BglG is also regulated by the preferred carbon sources. In addition to BglG, both MtlR and LevR, as the transcriptional activators, are also activated by P~His-HPr in the absence of preferred carbon sources [39,42,43], which could explain the catabolite control protein A (CcpA)-independent CCR between glucose and cellobiose in strain SCUT27 [20].
In addition, some other sugars can also be co-utilizaed during fermentation (Figure 2). With the lignocellulosic hydrolysate as substrate, once glucose and xylose concentrations approached zero, mannose and galactose were consumed simultaneously, the concentration of which were decreased by 98% and 92%, respectively. Finally, 37 g/L ethanol was produced from mixed sugars with an average ethanol productivity of 1.5 g/L/h in T. saccharolyticum engineered strain (ALK2) [34]. Although strain SCUT27 could co-utilize glucose and xylose, the presence of glucose still repressed xylose catabolism in some extent [44]. Therefore, new microorganism should be obtained by further screening and metabolic engineering. Thermophilic strain like TG57 was selected under the pressure of 2-deoxyglucose with ccr and xylR deletion/silence and thus was able to simultaneously ferment glucose, xylose and arabinose to produce butanol (7.33 g/L) as the sole solvent [45]. Other strategies for the enhancing utilization of substrates were also discussed below.
2.3. Starch Metabolism in Thermoanaerobacter
Starch is the abundant carbon source in the nature and high-temperature tolerant amylase is urgently needed to degrade the starch for industrial applications. Many types of thermostable enzymes have been studied for decades and one of them is amylopullulanase (APU), which possesses hydrolytic activity towards α-1,4 and α-1,6-glucosidic linkages and has been well-studied for industrial applications [46,47].
Thermoanaerobacter species such as T. saccharolyticum and strain SCUT27 can use starch, suggesting that thermostable amylases could be obtained from them [22,38,48]. It’s reported that the uitlization of starch was strongly inhibited by 2-deoxyglucose (an analogue of glucose), indicating that the CCR existed between starch and glucose (Figure 2) [22]. The CCR related genes, ccpA, ptsH and hprK were confirmed in strain SCUT27 and the high affinity between CcpA and phosphorylated HPr was proved via affinity chromatography [22]. Currie et al. also studied key hydrolases and transporters (Tsac_1344-1349) employed by T. saccharolyticum for starch metabolism, and genes expressions among starch, glucose or xylan were compared, showing a host of differences in terms of transporters, sensory genes, and hydrolases [38]. The highest ethanol titer achieved was 70 g/L in batch culture with a mixture of 90 g/L maltodextrin and 60 g/L cellobiose by T. saccharolyticum [49].
2.4. Intracellular Cellobiose Metabolism in Thermoanaerobacter
Cellulose is the major component in most of the lignocelluloses and can be hydrolyzed to glucose by cellulases, including endoglucanase, exoglucanase and β-glucosidase. Cellobiose is abundant in partial hydrolysis of lignocellulosic hydrolysates due to the weak activity of β-glucosidase, and the rate-limiting enzyme in the enzymatic hydrolysis [50]. There are three types of industrial hosts with the capability to use cellulose: (1) uptaking glucose from cellulose with cellulase pretreatment; (2) utilizing cellobiose extracellularly from partial hydrolysis of cellulose with the ability to produce β-glucosidases; (3) assimilating cellobiose intracellularly from partial cellulose hydrolysis with transporters and intracellular β-glucosidases [51]. Several critical barriers existed with the complete extracellular hydrolysis of cellulose, such as high costs for the supplementation of β-glucosidase, CCR caused by the released glucose during hydrolysis and energy waste for glucose metabolism instead of cellobiose [17,52].
The genus Thermoanaerobacter was engineered to overcome these obstacles for intracellular assimilation of cellobiose. Like strain SCUT27, it can assimilate cellobiose intracellularly with CCR existing between cellobiose and glucose [20,22]. Cel2 was the only intact operon for cellobiose utilization including integral PTS transport elements and glycoside hydrolase, the transcription of which was regulated by σ54-dependent transcription activator CelRcel2. By replacing the σ54-dependent promoter of Cel2 with the strong promoter from gene adhE, the mutant obtained the ability to co-utilize glucose and cellobiose [20]. In hemicellulose-degrading T. saccharolyticum, the biochemical and structure of a glucose-tolerant β-glucosidase was analyzed by Kim et al and results showed that TsaBgl had the unique conformation compared with other structurally known Bgls, and it also could convert cellobiose as well as other soluble cello-oligomers into glucose, providing energy sources for living organisms [53]. In T. saccharolyticum, 41 g/L ethanol was achieved from enzymatic hydrolysate of hardwood containing 96 g/L mixture of glucose and cellobiose [54] and 70 g/L ethanol was obtained in batch fermentation with a mixture of cellobiose and maltodextrin [49].
2.5. Traits about Metabolism of Other Substrates
In addition to the characteristics of sugar metabolism mentioned above, there are some other traits about sugar metabolism in the genus Thermoanaerobacter. In T. saccharolyticum, it was found that the utilization of mannose and arabinose was started before glucose and xylose were exhausted, owing to some thermostable enzymes efficient in Thermoanaerobater, such as β-mannanase [34,55]. However, in strain SCUT27, the metabolism of galactose and arabinose was inhibited by glucose, while mannose could be co-fermented with glucose [22]. In T. saccharolyticum, arabinose was completely utilized even at high concentrations, the metabolism of which was suppressed significantly at the presence of glucose or mannose. Directed mutations of hpr resulted in a reversal of carbon source preferences, and the mutant (His15Asp) utilized arabinose first, then followed by glucose, slowly and incompletely [54]. Notably, in T. saccharolyticum, CCR of galactose, arabinose and cellobiose in the presence of 2-deoxyglucose could be relieved by the HPr (His15Asp) mutation [39].
The central carbon metabolic network of T. saccharolyticum was established based on isotopic tracing and results showed that the metabolism of glucose was exclusively depended on EMP pathway, and both the Entner-Doudoroff (ED) pathway and the oxidative pentose phosphate pathway (oxPPP) were inactive [56]. Additionally, some thermophiles such as T. pseudethanolicus 39E and T. thermosaccharolyticum are capable of metabolizing sucrose [33]. Interestingly, T. butyriciformans sp. nov. USBA-019T and T. thermosaccharolyticum MJ1, presented different products composition under various substrates [57,58]. The main fermentation products of T. butyriciformans sp. nov. USBA-019T under glucose, lactose, cellobiose, galactose, arabinose, xylose, starch and xylan were acetate and butyrate. Other products, such as ethanol and lactic acid, also could be produced under starch and cellobiose [57]. Mannitol and glycerol were the more reduced substrates than glucose, and T. mathranii could use mannitol and glycerol to acquire higher ethanol yield [59]. Recently, it was reported that T. kivui as an acetogenic model organism could reduce CO2 with electrons derived from H2 or CO, or from organic substrates in the Wood–Ljugdahl pathway (WLP) [60,61], further broadening the substrate spectrum of Thermoanaerobater.
3. Genetic Operation Platforms
3.1. Vectors for Gene Overexpression
This section would provide a comprehensive description about the current genetic tools and techniques for modification of the genus Thermoanaerobacter, which belongs to low-G+C thermophilic and obligate anaerobes in the class Clostridia, and is considered to be resistant to genetic engineering due to the difficulty of introducing foreign DNA to cells. Shaw et al. reported the evidence of natural genetic competence in 13 kinds of Thermoanaerobacter and Thermoanaerobacterium strains, which were previously believed to be difficult to be transformed or genetically recalcitrant [62]. And in T. saccharolyticum JW/SL-YS485, natural competence-mediated DNA incorporation occurs during the exponential growth phase with both replicating plasmid and circular or linear DNA [62]. The transformation protocol and methods for transformants selection of T. saccharolyticum are described by Hon et al [48].
The earliest and most common vector for transformation and gene-overexpression in the genus Thermoanaerobacter such as Thermoanaerobacterium sp. strain JW/SL-YS485 was the plasmid pIKM1 containing kanamycin resistance gene (resistance gene of high-temperature kanamycin, htk) [54,63]. This replicating plasmid was originated from a 2.4 kb thermophilic replicating plasmid in T. saccharolyticum strain B6A-RI with KanR, AmpR, and a pUC origin, and then developed to create a convenient cloning vector [54]. Additionally, pTE16 was constructed based on the E. coli-Clostridium perfringens shuttle vector pJIR715 with a thermostable chloramphenicol (Cm) resistance cassette. The electro-transformation of T. ethanolicus JW200 was performed with pTE16 and the transformation efficiency was 50 ± 7.4 transformants per μg plasmid DNA [64]. Based on the hyperthermophilic Caldicellulosiruptor bescii replicon pBAS2, Joseph et al. constructed vectors which could be transformed into some thermophiles at 60 °C and maintained in multiple copy with the replication initiation protein RepL [65]. Le et al. successfully constructed an E. coli-T. ethanolicus shuttle expression vector pBlu10-htk for effective heterologous expression [66]. Some heterologous genes expression gave the strains expected performances successfully. For example, a heterologous gene gldA encoding an NAD+-dependent glycerol dehydrogenase was expressed to facilitate NADH regeneration for enhancing ethanol formation in T. mathranii [59]. Genes encoding the urease were integrated in T. saccharolyticum to hydrolyze urea [67].
3.2. Gene Knockout by Homologous Recombination
Kanamycin, erythromycin, chloramphenicol and thiamphenicol are four common antibiotics that can be used for the positive selection of the transformants from the genus Thermoanaerobacter (Table 1) [34,64]. Kanamycin and thiamphenicol are generally preferred due to the lower background and high efficiency of transformation. Notably, thiamphenicol or chloramphenicol is not used for T. thermosaccharolyticum, which carries acyltransferase gene like thiolase (thl). And these antibiotics can work in conditions that are suitable for the optimal growth of the genus Thermoanaerobacter (i.e., temperature and pH) under the responding working concentration as shown in Table 1 [34,64,68]. Haloacetates, 5-fluoroorotic acid (5-FOA) and 5-fluoro-2’-deoxyuridine (5-FUDR) are the negative selective drugs and can be adopted for the selection of transformants from genus Thermoanaerobacter with the corresponding endogenous negative selection markers genes, pta-ack, pryF and tdk deleted, respectively [48,69]. Several kinds of editing ways such as deletions, insertions, or a combination of a deletion and an insertion by homologous recombination in the genome could be achieved by the transformation of the suicide plasmid (circular DNA) like pBluescript II SK(+) or PCR products (linear DNA). Markerless gene deletion strategies were developed based on the counter-selectable markers, such as pyrF, tdk and pta/ack genes, which could eliminate the constraints imposed by the small number of positive selection markers for T. saccharolyticum [70]. The design of the suicide vector for markerless editing was also described in several studies [48,69,70].
3.3. Application of CRISPR/Cas Genome Editing System in Thermophiles
Recently, there are some application of Clustered Regularly-Interspaced Short Palindromic Repeat (CRISPR)/cas (CRISPR associated) system in thermophiles. Compared with traditional makerless editing, the advantage of the CRISPR system is to make the markerless editing easier for just one step. Mougiakos et al. identified and characterized ThermoCas9 from the thermophilic bacterium Geobacillus thermodenitrificans T12, which was active between 20 and 70 °C in vitro [74]. And thermoCas9-based engineering tools for gene editing such as gene deletion and transcriptional silencing at 55 °C in Bacillus smithii were performed successfully [75]. Le et al. developed a thermostable Cas9-based system for genome editing of T. ethanolicus, in which three genes (tdk, adhE and rsp) were edited successfully [66].
However, the heterologous Cas9 generally was poisonous to the cells [76,77]. Therefore, the native CRISPR/Cas system of some thermophiles was developed. For Thermoanaerobacter species, Dai et al. developed the endogenous type I-B CRISPR/Cas system with thymidine kinase gene (tdk) as a negative selection marker for multiplex genome editing in strain SCUT27, and the engineered strain SCUT27/Δtdk/Δldh/ΔargR exhibited the significant improvement over the wild type on ethanol production and xylose utilization [76]. In C. thermocellum, Walker et al. also repurposed the native Type I–B CRISPR/cas systems for genome editing [78]. Additionally, three thermophilic Cas9 variants (heterologous Type II) were tested and results showed that only GeoCas9, isolated from Geobacillus stearothermophilus, was active in C. thermocellum [78]. To overcome the limitation of homologous recombination between the repair template and the genome for both the Type I–B and the Type II GeoCas9 system, three novel thermophilic recombinases were tested, and only the exo/beta homologs, isolated from Acidithiobacillus caldus, exhibited the activity in C. thermocellum and led to higher genome editing efficiency [78].
The CRISPR-based transcriptional regulation of repression and activation (CRISRPi & CRISRPa) is a simple and high-throughput genome engineering tool and has been successfully performed in the genus Clostridium but rarely in thermophiles [79,80,81,82]. Nowadays, Ganguly et al. developed a ThermoCas9-based CRISPR interference (CRISPRi) system, which was applied to repress the transcription of lactate dehydrogenase (LDH) and phosphotransacetylase (PTA) genes in T. thermocellum with 67% and 62% enzymatic activity loss, respectively, without gene damage [83]. In addition to the heterologous Type II CRISPRi system, native Type I system was developed in Haloferax volcanii, E. coli, Listeria monocytogenes and S. thermophilus as well as in plants and even human cells [84,85,86], implying the potential application of endogenous type I CRISRPi or CRISRPa system in the genus Thermoanaerobacter.
4. Strategies for Enhancing the Biofuel Production
4.1. Enhancement of Substrates Utilization
The performance of the utilization of various substrates in Thermoanaerobacter has been reviewed above (section 2). However, there are not so many researches about the localization and class of each sugar transporters in Thermoanaerobacter, as well as the responding modification. To improve the tolerance to substrates, the strain Δldh/Δpta-ack of T. saccharolyticum was continuously cultured for approximately 3000 h with progressively increased feed xylose concentrations and finally the mutant exhibited a greater capacity for xylose consumption. More than 99% of the feed xylose was metabolized at concentrations up to 70 g/L with a mean ethanol yield of 0.46 g/g in continuous culture at pH of 5.2–5.4 without base addition [34]. In addition to the adaption evolution under high concentration substrates, genetic modification is also a key strategy to improve substrate utilization. Four enzymes from Herbinix spp. strain LL1355 which could hydrolyze almost all of glucuronoarabinoxylan linkages were successfully expressed in T. thermosaccharolyticum, allowing it to consume more substrates than that of the parent strain (78% vs. 53%) with ethanol yield improved by 24% from corn fiber [87]. Arginine repressors (ArgRs), the transcriptional repressors of the arginine biosynthesis, were reported to be involved in several key metabolic pathways and might be the global regulator [88,89]. In strain SCUT27, the deficiency of argR (V518_1864) could greatly improve the co-utilization of glucose and xylose, due to the enhanced activity of xylose isomerase, xylulokinase and the higher energy level [44]. Deletion of the redox-sensing transcriptional repressor (Rex) also improved the ability of sugar consumption of SCUT27 about 74–131% with the improvement of total ATP concentration and the cofactors NADH and NAD+ concentrations [90].
Consequently, further work can be carried out to improve the substrates utilization from the following three aspects: (i) Figure out and strengthen the transporters of sugars in Thermoanaerobacter; (ii) Understand the components and effectors of CCR in Thermoanaerobacter, and realize the co-utilization among the sugars and even expand the substrate spectra; (iii) High substrate tolerant strain should be screened or engineered by adaptive evolution and random mutation.
4.2. Enforcement of the Metabolic Flux to Biofuels Production
Biofuels are emerged as a promising alternative to fossil fuels for meeting the rapid increases of energy consumption and accumulation of greenhouse gases (climate change) [14]. For the genus Thermoanaerobacter, there are mainly three biofuels products: ethanol, hydrogen and butanol. Ethanol and hydrogen are principal products for majority of the Thermoanaerobacter species while butanol is only produced in a part of the genus Thermoanaerobacter like T. thermosaccharolyticum [91], which has the metabolic pathways of butyrate and butanol (Figure 4). In addition to biofuels, other bioproducts like lactic acid and acetic acid were rarely reported to be produced by Thermoanaerobacter species [60,61,92]. So, here we mainly discuss the research of Thermoanaerobacter species on biofuels production.
4.2.1. Ethanol Production
Bioethanol is one of the most attractive biofuels as it can serve as a blendstock for gasoline [93]. As a result, several countries, including the USA, Brazil, China, Canada, India and many EU members, have already proclaimed commitments to reduce their dependence on fossil fuels towards developing bioethanol production [94]. Currently, bioethanol is commonly produced by Saccharomyces cerevisiae [95,96], Zymomonas mobilis [97,98] and E. coli [96,99]. Thermoanaerobacter species are good ethanol producers at 55 °C (the temperature more economical to recover ethanol by continuous distillation) [54,100,101]. Strategies for enhancing the ethanol production mainly include strengthening the pathway of ethanol production, lowering the pathway of by-products, regulating the intracellular cofactor regeneration system as well as ATP level, or adaptive evolution (Table 2).
Both pyruvate formate lyase (PFL) and pyruvate ferredoxin oxidoreductase (PFOR) were found to participate in acetyl-CoA formation from pyruvate in genus Thermoanaerobacter [102]. PFOR is the key enzyme for high ethanol production in genus Thermoanaerobacter, because it is not inhibited by accumulation of ethanol or NADH while PFOR of C. thermocellum is [103]. Acetyl-CoA would be catalyzed by three major adh genes (adhA, adhB, and adhE) for ethanol production [33]. In addition, a few thermophilic bacterial species within the genus Thermoanaerobacter harbored a copy of aldehyde:ferredoxin oxidoreductase (aor), which reduced a variety of organic acids to their corresponding aldehydes, showing the AOR-ADH pathway in strains like Thermoanaerobacter sp. strain X514, T. brockii ssp. Finnii and T. pseudethanolicus [104,105,106]. AdhE is the bifunctional enzyme including alcohol dehydrogenase (ADH) and aldehyde dehydrogenase (ALDH) activities, the deletion of which would reduce the ethanol yield by ~95% [107]. NADH was the preferred cofactor for both ALDH and ADH activities [108]. It was reported that deletion of adhE reduced NADH-dependent ADH activity by 98%, which did not affect the NADPH-dependent ADH activity [72]. Notably, some mutants with high ethanol production, showed the increased NADPH-dependent ALDH and ADH activities, in which deletion of adhA dramatically reduced NADPH-dependent ADH activity and ethanol production [72,108]. Moreover, multiple gene deletions of all annotated alcohol dehydrogenase genes showed that AdhA and AdhE were the two main alcohol dehydrogenases involved in ethanol production in Thermoanaerobacter specie [72]. Thus, AdhA is responsible for the NADPH-ADH activity in Thermoanaerobacterium specie, especially in high-ethanol-producing strains, and may also serve as an additional acetaldehyde scavenger [72]. However, the overexpression of bifunctional or monofunctional ADH genes in wild type or engineered strains seems no improvement in ethanol titer [54].
For higher ethanol production by redistribution of carbon flow, metabolic engineering of the genus Thermoanaerobacter began with the disruption of the lactate production pathway [100,109]. Compared with the wild type, the biomass of mutant Δldh without lactic acid production, was increased by 31.0–31.4% under glucose or xylose as well as 2.1 to 2.4-fold increases in the yield of ethanol (mole/mole substrate) [100]. Shaw et al. knocked out the genes involved in organic acid formation (acetate kinase, phosphate acetyltransferase, and L-lactate dehydrogenase) and the resultant strain was able to produce ethanol as the sole detectable product (90% of the theoretical maximum yield) and 37 g/L ethanol was produced from mixed sugars with a maximum ethanol productivity of 2.7 g/L/h [34]. Shaw et al. investigated three putative hydrogenase enzyme systems in T. saccharolyticum and found that the strain with specified hfs and ldh deletions exhibited an increased ethanol yield from consumed carbohydrates [110,111]. The deletion of hfsB was also performed in strain SCUT27 and results indicated the enhanced ethanol production was obtained, due to the increased NADH supply and alcohol dehydrogenase activity [112]. Pyruvate formate lyase-activating protein (PflA) could assist the formation of pyruvate formate lyase (Pfl), which catalyzed pyruvate into acetyl-CoA and formate. The formation of formate also affected the ethanol production by changing the reducing power supply [101]. In strain SCUT27, the mutant Δldh/ΔpflA was constructed and it could produce 45 g/L ethanol during immobilized cell fermentation, owing to the increased electrons available for ethanol production [101]. The enhanced ethanol/acetate ratio was also important for ethanol production, which contributed to the resistant to 5% (v/v) exogenous ethanol [101]. To decrease amino acid secretion and reduce ammonium assimilation in C. thermocellum, type I glutamine synthetase gene (glnA) was knocked out and the levels of secreted valine and total amino acids were reduced by 53% and 44%, respectively, along with ethanol yields increased by 53% [27].
Additionally, reducing equivalents are key factors of alcohol production in Thermoanaerobacter species. There are three types of ferredoxin:NAD(P)H oxidoreductase (FNOR) reactions: uncoupled FNOR reaction, proton- or Na+-translocating FNOR reaction (RNF) and NADH-dependent FNOR reaction (NFN) [113]. NADH-FNOR (Tsac_1705) which transfers electrons from the reduced ferredoxin to NAD+, is the main enzyme responsible for ferredoxin oxidization in the NADH-based ethanol pathway in T. saccharolyticum [113]. In the high-ethanol-producing mutant (Δack/pta), NADPH-FNOR activity was increased compared with that of the wild type, which presumably led to the increased NADPH production [72,114]. The electron-bifurcating enzyme complex NfnAB (one of NADH-dependent reduced ferredoxin:NADP+ oxidoreductases, NFN), which transfers electrons from the reduced ferredoxin and NADH to 2 NADP+, is thought to play a key role in linking NAD(P)(H) metabolism with ferredoxin metabolism and has been hypothesized to be the main enzyme for ferredoxin oxidization in mutants of T. saccharolyticum for increasing ethanol production [114]. In strain SCUT27, the NADH/NAD+ ratio in the ∆nfnAB mutant was increased [68]. Interestingly, adhA is directly located in the upstream nfnAB in T. saccharolyticum, showing the potential link between reducing power supply [114]. Two different mechanisms for stoichiometric yield of ethanol were proposed in T. saccharolyticum. One was based on NADPH and NfnAB, and the other was dependent on NADH and a yet undescribed NADH-FNOR (like unknown rnf complexes) [114]. The redox-sensing transcriptional repressor Rex, which was widely reported as the targeting adhE by sensing the redox state, was knocked out in strain SCUT27. Results showed that the ethanol yield of strain SCUT27/Δrex was increased by 75–91% within different carbon sources, along with the improvement of total ATP, cofactors concentration as well as the activities of alcohol dehydrogenase [90]. To regenerate cofactor, T. mathranii was fed with mannitol to increase ethanol yield [59]. To further facilitate NADH regeneration for ethanol formation, a heterologous gene gldA encoding a NAD+-dependent glycerol dehydrogenase was overexpressed in T. mathranii, and the resultant mutant showed increased the ethanol yield in the presence of glycerol undr xylose [59]. However, it was also reported that a wide range of genetic modifications for the mixed acid fermentation pathways did not significantly alter the NAD(P)H/NAD(P)+ ratios as well as their concentrations in T. saccharolyticum [115]. It appeared that for each species, the ratios were maintained in a somewhat narrow range: 0.4–0.7 (NADH/NAD+) and 0.5–0.9 (NADPH/NADP+) [115].
Moreover, adaptive evolution under exogenous ethanol did not always result in mutants of T. saccharolyticum that could produce more ethanol and in fact, the mutant obtained usually produced lower titer of ethanol than the wild type [54]. A strain of T. pseudethanolicus with enhanced ethanol tolerance showed high yields of ethanol but lower titer than the parent strain, and NADH-dependent ADH, ALDH, and FNOR activities of the mutant were lost, but NADPH-dependent activities remained [116]. Interestingly, addition of acetate showed an unexpected stimulatory effect on ethanolic fermentation by T. ethanolicus strain 39E and strain X514 while lactate was in general inhibitory to ethanolic fermentation. In T. saccharolyticum M2886, the highest ethanol titer achieved was 70 g/L (also the maximun ethanol tolerance) in batch culture under the mixture of cellobiose and maltodextrin by removing exopolysaccharide synthesis gene (perR) and replacing the gene of methyglyoxal synthase with pta/ack genes, after multiple rounds of adaptation and selection [49]. Furthermore, it suggested that vitamin B12 de novo synthesis in strain X514 had a significant positve effect on ethanol yields in both mono- and coculture [33]. The vitamin B12 pathway may appear to promote ethanologenesis through ethanolamine and 1,2-propanediol [23]. Additionally, conversion of acetic acid to acetone in engineered T. saccharolyticum resulted in the improved ethanol productivity and titer, and was an attractive low-cost solution to acetic acid inhibition [117].
To sum up, ethanol production in Thermoanaerobacter species could be further improved by conducting more carbon flux to ethanol and weakening the reducing power consumed by the branch-pathways for enhancing ethanol production. In addition, high-ethanol-tolerant strain should be screened by adaptive evolution or random mutation. Finally, future studies on ethanol production would focus on the cheaper and less toxic feedstocks to improve their application value.
4.2.2. H2 Production
Hydrogen produced from biomass feedstocks is considered as an effective solution toward a decarbonized economy [118]. Through the circular economy framework, it is possible to produce the sustainable biohydrogen and other bioproducts while address the issues such as waste management [119]. In Thermoanaerobacter species, hydrogen is a necessary product for the disposal of the reducing equivalents. Strains like T. tengcongensis are able to convert the reducing equivalents generated during the fermentation from glucose to the acetate and CO2 by reducing H+ to H2 (Table 3). A unique combination of hydrogenases, a ferredoxin-dependent [NiFe] hydrogenase and an NADH-dependent Fe-only hydrogenase, was found to be responsible for H2 formation in T. tengcongensis [120].
Three putative hydrogenase systems were identified in T. saccharolyticum: A four-gene operon containing two [FeFe]-hydrogenase genes (hfsBD), provisionally termed hfs (hydrogenase-Fe-S), of which the deletion resulted in a growth defect [111], was found to be the main enzymatic catalyst of hydrogen production; A second hydrogenase gene cluster, a bifurcating NAD(P)H-linked [FeFe]-hydrogenase (hyd), exhibited methyl viologen-linked hydrogenase enzymatic activity and was specific to the transfer of electrons from NAD(P)H to H+; A third gene cluster, a putative [NiFe]-hydrogenase with homology to the ech genes, did not exhibit the hydrogenase activity [110,112]. Strains with a deletion of the hfs genes exhibited a 95% decrease in both hydrogen and acetic acid production, and hfsB seemed as the most important gene of the operon hfsABCD [110]. The genes hyd, ech and hfsABCD, highly homologic to that of T. saccharolyticum and T. tengcongensis, were also found in strain SCUT27 and proved to be related to H2 production by gene knockout [112].
To redirect the NADH flow to hydrogen production, the gene encoding L-lactate dehydrogenase was knocked out, resulting in 2 and 2.5-folds increase of H2 yield and production rate, respectively [69]. In the wild type of T. saccharolyticum, deletion of nfnAB showed a 46% increase in H2 formation owing to the enhanced NADH generatation [114]. Also, in strain SCUT27, the concentration of hydrogen production was increased by 41.1% due to the enhanced NADH/NAD+ ratio in the ∆nfnAB mutant [68]. Additionally, deletion of PFL, potential H2-uptake (NAD(P)H-dependent) H2ases and ethanol producing pathways and overexpression of Fd-dependent H2ases could also increase H2 yields and production [122]. However, when H2 was accumulated in the closed system, the fermentation was partially shifted to ethanol production, due to the high hydrogen partial pressure [p(H2)] leading to the low NADH-dependent hydrogenase activity [121].
Furthermore, a number of H2-producing Thermoanaerobacter species were screened and co-cultured with other strains for higher H2 production. Hu et al. isolated a new T. thermosaccharolyticum named MJ1 from paper sludge, whose hydrogen yields could reach 11.18 mol‑H2/mol cellobiose at an initial sugar concentration of 5 g/L, and the hydrogen production was 111.8 mM under sugarcane bagasse hydrolysate pretreated with 60% dilute‑acid (1%, g/v) [58]. Li & Zhu took a biphasic fermentation approach for the production of ethanol and hydrogen from cassava pulp. The glucose generated by co-culturing of C. thermocellum and strain SCUT27 under cassava pulp was utilized rapidly by subsequently inoculated S. cerevisiae, and the final productions of 8.8 g/L ethanol and 4.1 mmol/L hydrogen were achieved [123]. C. thermocellum and strain SCUT27 were also co-cultured with 40 g/L alkali-pretreated sugarcane bagasse (SCB) for bio-hydrogen production, with a final hydrogen production of ~50 mmol/L [124].
For enhanced thermophilic biohydrogen production with cellulosic wastes, future work can be carried out from the following aspects: First, digging out new thermophilic strains which can naturally produce high level of hydrogen efficiently. Second, providing new strategies for genetic modifications on Thermoanaerobacter species for higher hydrogen production, such as developing favorable thermophilic biohydrogen pathways and strengthening enzymes stability. Third, co-cultureing with other microorganisms for further improving the substrates utilization or product yields. Finally, enhancing the biohydrogen production technology under the dark fermentation mode with the implementation of agroindustrial wastes as potential substrates, and exploring the suitable temperature and pH which are the imperative parameters during the dark fermentation.
4.2.3. Butanol Production
Except for the general products of hydrogen and ethanol, some Thermoanaerobacter species like T. thermosaccharolyticum also have great potential to produce butanol (Table 4), which is a renewable alternative to gasoline in vehicles without any modifications in engine system [2] due to its higher boiling point [125].
A novel thermophilic and butanogenic T. thermosaccharolyticum M5 was isolated and could directly produce butanol from hemicellulose via a unique ethanol–butanol (EB) pathway, with the efficient expression of xylanase, β-xylosidase and the bifunctional alcohol/aldehyde dehydrogenase (AdhE) [91]. In addition, process optimization based on the characteristic of AdhE was carried out, which further improved the final butanol titer to 1.17 g/L from xylan [126]. Furthermore, the co-culture of Thermoanaerobacterium sp. M5 and C. acetobutylicum NJ4 was established to improve the butanol titer to 8.34 g/L from xylan through CBP [126]. Moreover, Li et al. reported a wild-type T. thermosaccharolyticum strain TG57, which was capable of converting microcrystalline cellulose directly to butanol (1.93 g/liter) as the only final product. And the strain TG57 was also able to simultaneously ferment glucose, xylose, and arabinose to produce 7.33 g/L butanol [45]. Integration of heterologous components of n-butanol pathway (thl-hbd-crt-bcd-etfAB) from T. thermosaccharolyticum DSM 571 resulted in 0.85 g/L butanol from 10 g/L xylose in T. saccharolyticum JW/SL-YS485 (21% of the theoretical maximum yield), and resulted in 1.05 g/L butanol from 10 g/L xylose in lactate deficient strain (26% of the theoretical maximum yield) [127].
In general, the titer of butanol produced by metabolic engineering strategies needs to be improved due to the high toxicity of butanol and the low catalytic activity of enzymes. The following works can be adopted to optimize the butanol biosynthesis: (i) eliminate the byproduct like ethanol to improve the yield of butanol; (ii) obtain the efficient and thermotolerant enzymes by rational engineering or direct evolution for butanol producing; (iii) improve cell tolerance to butanol via transporter engineering or evolution engineering; (iv) screen novel thermophilic butanol producers that could produce much higher butanol titer than that of C. acetobutylicum.
4.3. Improvement of the Tolerance to Lignocellulose-Derived Inhibitors
T. saccharolyticum was able to utilize most of the carbohydrates in the hydrolysates of pretreated hardwood, indicating that the strain possessed all the proteins required for efficiently breaking down hemicellulose and could produce twice more ethanol from detoxified hydrolysate than that of untreated hydrolysate (31 g/L vs. 16 g/L) in fermenters [54]. It’s reported that lignocellulose-derived inhibitors could trigger the accumulation of reactive oxygen species (ROS), resulting in cell damage. Some researches also suggested that inhibitors presented a redox challenge to cells [54,130]. Although the inhibitors such as HMF and furfural could be metabolized by strains like T. saccharolyticum, some inhibitors such as furfural could delay growth, and others such as vanillin could affect cell yield [54,76].To make full use of the lignocellulose, detoxification of lignocellulosic hydrolysates was usually necessary.
Activated carbon was used to reduce the toxicity of both the solid and the liquid feeds [49]. The treatment of the hydrolysate with diafiltration and/or 1% activated carbon resulted in much higher ethanol titers [54]. Tolerance of the strains to inhibitors was greatly improved by the presence of cysteine in the medium [54]. Additionally, acetic acid was also a kind of inhibitor. The ethanol production was stopped when acetic acid levels reached 10–14g/L [54]. Overliming was demonstrated as an effective method to detoxify polymeric hemicellulose [54]. An excess of calcium carbonate provided excellent buffering at a pH of 5.5, which is close to the optimum pH for T. saccharolyticum [49]. Since many inhibitors had greater negative effect in minimal media than in yeast extract containing media, addition of yeast extract seemed a simple way for alleviating the suppression [24,49].
Moreover, genetic modification also played a vital role in strengthening Thermoanaerobacter species against lignocellulose-derived inhibitors, as shown in Table 5. Elimination of the gene perR in T. saccharolyticum led to dramatical improvement in inhibitor tolerance and higher oxygen tolerance as well as higher ethanol titer [49,130]. The pta/ack genes seemed the key genes for the inhibitor tolerance phenotype and the dramatic inhibitor tolerance was conferred by restoring the pta/ack genes into a pta/ack(−) strain [49]. In strain SCUT27/ΔargR1864, the improved concentration of intracellular ATP and NAD(H) provided more energy to respond the stresses, which gave the mutant the better cell viability to utilize lignocellulosic hydrolysates for enhancing ethanol production [44]. In addition, the transcriptome data between SCUT27/ΔargR1864 and wild type showed that ArgR1864 was the negative regulator for chaperonin synthesis. With argR1864 knockout, the mutant showed the upregulated expression of DnaK-DnaJ-GrpE system and GroEL-GroES chaperonin, which not only endowed the strain with the enhanced ability to scavenge the reactive oxygen species (ROS), but also gave the mutant the capability to maintain better cell growth, xylose consumption, acetic acid and ethanol production under the stress of various lignocellulose-derived inhibitors [131].
It was reported that T. saccharolyticum was inhibited by monosaccharides at concentrations higher than 40 g/L, and with methylglyoxal synthase deletion, the mutant could grow at 100 g/L glucose and produce more ethanol from hardwood hydrolysate [49]. Furthermore, during continuous cultures of T. thermosaccharolyticum, the cessation of growth and fermentation at feed exceeding 70 g/L xylose seemed to be attributed to salt formed by neutralizing organic acid rather than the toxicity of ethanol [132]. Notably, salt inhibition could be avoided by introduction of the urease operon [67].
5. Applications of Thermoanaerobacter Species for Various Bioprocessing
When enzymatic hydrolysis is adopted for cellulosic ethanol production, different levels of process integration can be visualized: (i) separate (or sequential) hydrolysis and fermentation (SHF), where the enzymes (cellulases) are used separately from fermentation tanks; (ii) simultaneous saccharification and fermentation (SSF), which consolidates enzymatic hydrolysis with the fermentation of hexose and pentose sequentially; (iii) simultaneous saccharification and cofermentation (SSCF), which further combines the fermentation of hexose and pentose together; and (iv) consolidated bioprocessing (CBP), where cellulases and ethanol are produced in a single reactor [33]. For Thermoanaerobacter species, cellulose hydrolysis is the main limiting factor in the progress of fermentation [34].
A fed-batch fermentation with T. saccharolyticum M2886 was performed and 50 g/L ethanol after 60 h cultivation under pretreated hardwood was obtained in SHF mode [49]. 37 g/L ethanol was produced by strain ALK2 in SHF at 50°C which had a 2.5-fold decrease in cellulase loading compared with that of S. cerevisiae at 37 °C [34]. For SSF, mesophiles ferment well at ≤37 °C and thus require substantially higher cellulase loadings compared with the thermophiles cultured under 50~60 °C (the optimal temperature for both enzymes activity and the growth of Thermoanaerobacter) [34,54]. Pre-hydrolysis for a long time facilitates the cellulose conversion, but the high initial sugar concentrations >50 g/L (usually due to high initial enzyme dose) would inhibit the cells growth [54]. SSF of avicel with T. saccharolyticum in batch mode at an initial concentration of 50 g/L after pretreatment with a commercial cellulase showed the better performance than that that of yeast (2.5-fold increase of cellulase loading). And further reductions in cellulase loading could be anticipated as more thermostable cellulases are developed [34,133]. Under the optimal substrate “mock hydrolysate” and conditions such as temperature, pH-control strategy, inoculum size, cellulase mixture and feeding regimen, T. saccharolyticum could produce over 61 g/L ethanol in thermophilic SSF [54]. Moreover, 31 g/L ethanol was produced in SSF of pre-treated hardwood and 26 g/L ethanol was obtained with a hardwood hemicellulose extract [49]. Fermentation conditions were also developed to get the highest possible ethanol titer in a SSCF configuration under the conditions expected from pre-treated hardwood (100 g/L purified cellulose, 10 g/L acetic acid, 35 g/L xylose and 20 g/L glucose), and more than 60 g/L ethanol (greater than 90% yield) was achieved at 93 h cultivation by strain M1442 [49]. Additionally, cell lysis behavior was observed when the strain grew too fast, but it could be prevented under mixed sugars with a slower growth rate, or in the medium with high concentration of magnesium or low pH [54].
A preferred CBP system can directly convert both cellulose and hemicellulose into target products without the costly hydrolytic enzyme cocktail. Cellulolytic strains like C. thermocellum are necessary to be used to hydrolyze and ferment all of the sugars in the broth with Thermoanaerobacter species. It was reported that the coculture of C. thermocellum JN4 and T. thermosaccharolyticum GD17 could significantly improve the efficiency of cellulosic biofuel [134]. Biodegradation of sugarcane bagasse (SCB) was remarkably improved with the addition of non-ionic surfactant (etc. Triton X-100) to the CBP system with C. thermocellum [135]. The applications of CBP for hydrogen production have been shown above (section 4.2.2). Lynd et al. also listed some detailed advances in CBP using C. thermocellum and T. saccharolyticum [136]. While these production schemes represent increasing levels of simplification through process consolidation, consolidation of multiple steps often results in a loss of process efficiency. Improving the efficiency of individual step, such as cellulose hydrolysis and ethanol fermentation, is still the important task for the development of economically feasible cellulosic biofuel.
6. Conclusions
In this paper, the traits of Thermoanaerobacter species on substrate utilization, their current platforms of genetic operation, and the strategies for enhanced biofuels (ethanol, hydrogen and butanol) production were reviewed. Numerous ways have been developed to enhance the sugar consumption and biofuel production in Thermoanaerobacter species. However, the production of ethanol is far less than that of S. cerevisiae and Z. mobilis. Achievement of biobutanol with high titer is still difficult due to the overwhelming metabolic stress. Further research should be performed as follows to establish efficient microbial cell factories for the genus Thermoanaerobacter to produce biofuels or other bioproducts:
(i) Better understand the genes and pathways involved in carbohydrate transport and metabolism, and address the mechanism of the co-utilization between glucose and xylose, which can shorten the fermentation period and make full use of the lignocellulosic materials.
(ii) Further develop effective elements of genetic operation such as useable fluorescent proteins for Thermoanaerobacter species, and technologies in large-scale engineering like CRISPRi and CRISPRa, for conveniently analyzing metabolic flux, engineering the high-producing strains and exploring the physiological mechanism of the bacteria.
(iii) The whole-cell models of Thermoanaerobacter species should be built to analyze the cellular networks and find key bottlenecks in biofuels production. Then the corresponding metabolic engineering strategies could be used for enhancing titer and yield of the products.
(iv) Tolerance to the inhibitors, high concentration of substrates and products, always poses a great challenge for Thermoanaerobacter species. It is necessary to reveal the mechanism and lay the fundation to improve the resistance of cells to several inhibitors for higher biofuels production.
(v) Optimize the fermentation process and construct microbial consortia to reduce the metabolic burden. More strategies and techniques should be practiced on perspective applications of Thermoanaerobacter species for consolidated bioprocessing.
(vi) Other valuable biochemicals and bioproducts such as poly-3-hydroxybutyrate (PHB), polylactic acid (PLA) and fatty acid esters could be further developed in Thermoanaerobacter species once getting the responding thermostable enzymes.
Ethics Statement
Not applicable.
Informed Consent Statement
Not applicable.
Funding
This work was supported by the National Natural Science Foundation of China (21878103) and the Natural Science Foundation of Guangdong Province of China (2018A030310368).
Declaration of Competing Interest
The authors declare that they have no competing interests.
