View PDF VersionPrevious ArticleNext Article
 
DOI: 10.1039/D1GC04288B (Critical Review) Green Chem., 2022, 24, 1390-1403

Recent advances in biological production of 1,3-propanediol: new routes and engineering strategies

Fanghuan Zhu abc, Dehua Liu *abcd and Zhen Chen *abcd
aDepartment of Chemical Engineering, Tsinghua University, Beijing 100084, China. E-mail: zhenchen2013@mail.tsinghua.edu.cn; Fax: +86-10-62792128; Tel: +86-10-62772130
bKey Lab of Industrial Biocatalysis, Ministry of Education, Beijing 100084, China. E-mail: dhliu@mail.tsinghua.edu.cn; Fax: +86-10-62792128; Tel: +86-10-62792128
cTsinghua Innovation Center in Dongguan, Dongguan 523808, China
dCenter of Synthetic and Systems Biology, Tsinghua University, Beijing 100084, China

Received 17th November 2021 , Accepted 12th January 2022

First published on 12th January 2022

Abstract

1,3-Propanediol (1,3-PDO) is an important chemical which has been widely used in the cosmetics, pharmaceutical, and especially polymer industries. The production of 1,3-PDO from renewable feedstocks by green processes is attracting wide attention. Although biological production of 1,3-PDO has been commercialized, the development of more efficient microbial cell factories and new bioprocesses to use diversified cheap feedstocks, eliminate by-product formation, and avoid the requirement of adding expensive cofactors is highly desired to further reduce production costs. This review details recent advancements of metabolic engineering and synthetic biology strategies in the biological production of 1,3-PDO. In particular, we elaborate on the comprehensive development of biological routes for 1,3-PDO production from glycerol, sugars, and other carbon sources using engineered producers, new chassis, and new pathways. The recently developed bioprocesses using microbial consortia and electro-fermentation are also included. The challenges and prospects of current bioproduction systems and their potential industrial applications are discussed.

image file: d1gc04288b-p1.tif

Fanghuan Zhu

Miss Fanghuan Zhu obtained her B.S. degree from Zhejiang University in 2018 and master's degree in Food Science from Zhejiang University in 2021. She is now a PhD student in the Department of Chemical Engineering, Tsinghua University, China. Her research focuses on metabolic engineering of microorganisms for the production of chemicals.

image file: d1gc04288b-p2.tif

Dehua Liu

Dr Dehua Liu is a professor and the director of the Institute of Applied Chemistry, Department of Chemical Engineering, Tsinghua University, China. He received his PhD in Department of Chemical Engineering, Tsinghua University in 1991. His research field of interest includes the production of biofuels and biochemicals from natural renewable biomass, especially focusing on biodiesel production from oil feedstock and microbial production of 1,3-propanediol from the biodiesel by-product glycerol. He has published more than 250 academic papers. He has been awarded the First Award of the S&T Progress of China Petroleum and Chemical Industry Federation (CPCIF)(2006), “BlueSky Awards” for World's Leading Technology of Renewable Energy with Best Value of Investment (2016, 2018) and so on.

image file: d1gc04288b-p3.tif

Zhen Chen

Dr Zhen Chen is an associate professor in the Department of Chemical Engineering, Tsinghua University, China. He received his PhD at the Institute of Bioprocess and Biosystems Engineering in Hamburg University of Technology in 2012, and performed his post-doctoral research work at Hamburg University of Technology (2012–2013). His current research focuses on the integration of metabolic engineering and synthetic biology for the development of bioprocesses to produce biofuels and chemicals. He was awarded “BlueSky Awards” for World's Leading Technology of Renewable Energy with Best Value of Investment (2018).

1. Introduction

1,3-Propanediol (1,3-PDO), a hygroscopic and viscous liquid, is widely used as an important organic solvent in the cosmetics industry and raw material or intermediate in polymer and pharmaceutical industries. In particular, 1,3-PDO can be used as a key monomer for the production of high-performance polyester polytrimethylene terephthalate (PTT).1 With the superior advantages of softness, stretch resilience, and high resistance to stains over other polyesters, PTT has been widely applied in the manufacture of decoration materials, textile fibers, clothing, engineering plastics, and films.2 The production of 1,3-PDO was originally developed based on chemical synthetic routes, including ethylene oxide hydroformylation and hydrogenation process and acrolein hydration–hydrogenation process, which were developed and commercialized by two chemical industry giants, Shell and Degussa.3 However, considering the high investment, technical difficulty, substrate toxicity and price, and environmental issues, the industrial application of the mentioned chemical processes has been gradually abandoned after the early 2010s. Some new chemical routes such as hydrogenolysis of glycerol to 1,3-PDO are still under development and the obtained productivity and selectivity are not high enough for commercialization.4,5

In contrast, the biological production of 1,3-PDO from renewable resources via fermentation has been attracting wide attention in the past twenty years, and large efforts have been made to develop new microorganisms and processes to increase the economic competitiveness of biological routes. Compared to chemical routes, microorganisms can utilize diversified carbon sources to produce targeted products and biological processes are generally safer and more environmentally friendly. In particular, DuPont and Genencor developed an efficient biological route to produce 1,3-PDO from glucose based on recombinant Escherichia coli in the early 2000s6 and commercialized the process in 2006.3 Our group has also developed and commercialized an alternative biological route with industrial partners for 1,3-PDO production based on glycerol, a cheap and abundant raw material from the biodiesel industry.7

The economic competitiveness of biological routes is strongly dependent on the titer, yield, and productivity of the targeted products. In particular, the cost of substrates accounts for 50%–60% of the total production cost of 1,3-PDO.8 Thus, the development of 1,3-PDO bioproduction systems from different cheap substrates with high yields is highly desired. Natural microorganisms can only utilize glycerol to produce 1,3-PDO with the accumulation of many byproducts. Thus, large efforts have been made to reduce byproduct formation in order to increase product yield and simplify the downstream process in the past few years.9–11 Moreover, metabolic engineering strategies have also been implemented to develop new chassis and processes in order to produce 1,3-PDO from other substrates.

In this review, we summarize and discuss the recent advancements in applying metabolic engineering and synthetic biology strategies for microbial biosynthesis of 1,3-PDO. Particularly, strain engineering of natural producers and new chassis to produce 1,3-PDO from glycerol, glucose, and other alternative substrates is elaborated. We also discuss the remaining challenges of current 1,3-PDO bioproduction systems and propose future perspectives to accelerate their industrial application.

2. Production of 1,3-PDO from glycerol

The surplus of glycerol from biodiesel and bioethanol industries is leading to a dramatic decline in the price of crude glycerol, making it an attractive feedstock for biorefineries. In nature, several microorganisms can naturally synthesize 1,3-PDO from glycerol, including Klebsiella pneumoniae, Clostridium butyricum, Clostridium pasteurianum, Enterobacter agglomerans, Citrobacter freundii, Lactobacillus brevis, etc.12 Under anaerobic or microaerobic conditions, glycerol is initially dehydrated to 3-hydroxypropionaldehyde (3-HPA) via B12-dependent or S-adenosylmethionine-dependent glycerol dehydratase, and the latter is reduced to 1,3-PDO by the NADH-dependent 1,3-PDO oxidoreductases (Fig. 1). The NADH should be generated via the glycerol oxidative pathway, which also inevitably generates different kinds of fermentative products under anaerobic or microaerobic conditions (Fig. 1). Under optimized conditions, different wild-type strains can produce 40–90 g L−1 1,3-PDO in fed-batch fermentation with the accumulation of different byproducts13–18 (Table 1). The accumulation of byproducts not only inhibits cell growth and reduces product yield but also complicates the purification process.19 Accordingly, many strategies have been proposed to address the issues of byproduct accumulation and reduce equivalent insufficiency. In addition, metabolic engineering strategies have also been implemented to engineer several non-natural hosts to produce 1,3-PDO from glycerol (Table 2).
image file: d1gc04288b-f1.tif
Fig. 1 Metabolic pathways and engineering strategies for the production of 1,3-propanediol (1,3-PDO). The red arrows indicate heterologous pathways. The metabolites in blue indicate main byproducts during 1,3-PDO production. Abbreviation: PEP, phosphoenolpyruvate; galP, galactose permease; glk, glucose kinase; ptsG, EIIAGlu; crr, EIICBGlu; tpiA, triosephosphate isomerase; dhaD/gldA, glycerol dehydrogenase; dhaK, dihydroxyacetone kinase; glpD/glpABC, glycerol-3-phosphate dehydrogenase; glpK, glycerol kinase; gdp1, glycerol 3-phosphate dehydrogenase; gpp2, glycerol-3-phosphate phosphatase; glpF, glycerol transporter; dhaBCE, glycerol dehydratase; gdrAB, glycerol dehydratase activator; dhaT, 1,3-PDO oxidoreductase; yqhD, alcohol dehydrogenase; gapA, NAD+-dependent glyceraldehyde-3-phosphate dehydrogenase; ldhA, lactate dehydrogenase; budABC, 2,3-butanediol dehydrogenase; frdABCD, fumarate reductase; pflB, pyruvate formate lyase; poxB, pyruvate oxidase; adhE, alcohol dehydrogenase; pta, phosphoacetyl transferase; ackA, acetate kinase; lpdA, dihydrolipoyl dehydrogenase; gltA, citrate synthase; icd, isocitrate dehydrogenase; sucAB, α-ketoglutarate dehydrogenase complex; sucCD, succinyl coA synthase; sdhABCD, succinate dehydrogenase; fumABCD, fumarase; mdh, malate dehydrogenase; nuo, NADH dehydrogenase I; ndh, NADH dehydrogenase II.
Table 1 Production of 1,3-PDO by typical wild-type strains
Strain Titer Yield Productivity Substrate Culture Ref.
K. pneumoniae 80.1 g L−1 0.53 molmol–1 2.22 g L−1 h−1 Glycerol Fed-batch 13
K. oxytoca 50.1 g L−1 0.48 mol mol−1 0.55 g L−1 h−1 Crude glycerol Fed-batch 14
C. butyricum 93.7 g L−1 0.63 mol mol−1 3.3 g L−1 h−1 Glycerol Fed-batch 15
76.2 g L−1 0.62 mol mol−1 2.3 g L−1 h−1 Crude glycerol Fed-batch
C. freundii 68.1 g L−1 0.48 mol mol−1 0.79 g L−1 h−1 Crude glycerol Fed-batch 16
L. diolivorans 85.4 g L−1 0.57 mol mol−1 substrate 0.85 g L−1 h−1 Glycerol + glucose Fed-batch 17
L. reuteri 49.9 g L−1 1.03 mol mol−1 0.92 g L−1 h−1 Glycerol + glucose Fed-batch 18

Table 2 Production of 1,3-PDO from glycerol by typical metabolically engineered strains
Strain Strategies Titer Yield Productivity Substrate Culture Ref.
Natural producer
K. pneumoniae Δldh 102.1 g L−1 0.52 mol mol−1 2.13 g L−1 h−1 Glycerol Fed-batch 21
C. beijerinckii Cell immobilized; overexpression of dhaD and dhaKLM 26.1 g L−1 0.55 mol mol−1 0.99 g L−1 h−1 Glycerol Fed-batch 41
L. reuteri Δadh2; overexpression of pduQ 52.3 g L−1 0.86 mol mol−1 1.09 g L−1 h−1 Glycerol and glucose Fed-batch 37
C. freundii Overexpression dhaT 35.6 g L−1 0.49 mol mol−1 0.32 g L−1 h−1 Glycerol Fed-batch 42
New chassis
E. coli JA 03 ΔptsG; overexpression of dhaB, yqhD, gdrAB, galP, glk, gapN 13.5 g L−1 0.64 mol mol−1 0.22 g L−1 h−1 Glycerol and glucose Batch 45
V. natriegens ΔadhE, Δldh, Δpfl, Δpta-ackA, ΔfrdABCD, ΔaldAB, ΔarcA, ΔglpR, ΔsthA; overexpression of phaP and pntAB 56.2 g L−1 0.61 mol mol−1 2.36 g L−1 h−1 Glycerol Fed-batch 50
44.6 g L−1 0.52 mol mol−1 1.76 g L−1 h−1 Crude glycerol
P. denitrificans ΔaldH13, Δpta-ack, ΔnuoA 33.5 g L−1 0.89 mol mol−1 0.66 g L−1 h−1 Glycerol Fed-batch 51
C. acetobutylicum Overexpression of dhaB and dhaT 84.0 g L−1 0.65 mol mol−1 1.70 g L−1 h−1 Glycerol Fed-batch 53

2.1 Engineering of natural producers

Among the natural 1,3-PDO producers, K. pneumoniae is one of the most popular cell factories due to its high glycerol consumption rate and 1,3-PDO production rate. During 1,3-PDO production, K. pneumoniae also accumulates many byproducts, including succinate, formate, lactate, 2,3-butanediol (2,3-BDO), ethanol, and acetate (Fig. 1). The theoretical yield of 1,3-PDO is 0.76 mol mol–1 glycerol if pyruvate dehydrogenase in the glycerol oxidative pathway is activated and only acetate is generated as a byproduct. The theoretical yield of 1,3-PDO can be increased to 0.84 mol mol−1 glycerol if the tricarboxylic acid cycle (TCA cycle) is activated for acetyl-CoA metabolism and NADH generation and no other fermentative byproducts are accumulated.20 The reduction or elimination of the formation of a specific byproduct can be easily achieved by deleting the corresponding genes.21–23 However, only blocking a single byproduct synthesis pathway was shown to be ineffective in increasing the yield of 1,3-PDO.24,25 For example, deletion of the ldhA gene encoding lactate dehydrogenase almost completely abolished lactate accumulation (from 37 g L−1 to <0.5 g L−1) but simultaneously increased the accumulation of 2,3-BDO to 26.6 g L−1 (from <5 g L−1)26 due to flux flexibility at the pyruvate node.27 Xin et al. tried to eliminate all byproduct formation pathways except lactate by deleting the adhE, budAB, frdA, poxB, and pta-ackA genes, which are responsible for the synthesis of ethanol, 2,3-BDO, succinate, and acetate. The resulting strain can coproduce 76 g L−1 1,3-PDO and 112 g L−1D-lactate with a total conversion yield of 0.95 mol mol−1 glycerol in fed-batch fermentation and the total accumulation of other byproducts (succinate and acetate) was lower than 5 g L−1.28 Although the atom economy of the process is high, the yield of 1,3-PDO (∼0.43 mol mol−1) is not satisfactory considering that 1,3-PDO is more expensive than lactate. Simultaneously blocking lactate and 2,3-BDO synthesis pathways was shown to be deleterious for cell growth and glycerol consumption and a high accumulation of pyruvate and acetate was observed.27,29 It was probably caused by the surplus of NADH after blocking the two major NADH consumption routes29 in the glycerol oxidative pathways. Although NADH is required for 1,3-PDO production, a high NADH/NAD+ ratio also results in the inhibition of several key enzymes in the glycerol oxidative pathway, including glycerol-3-phosphate dehydrogenase, pyruvate dehydrogenase, and citrate synthase.30 Thus, how to maintain the balance between NADH generation and consumption is a dilemma that needs to be addressed. To enhance NADH consumption for the biosynthesis of 1,3-PDO, the strict genetic regulation of the dha operon should be rationally adjusted. By overexpressing the dhaT gene encoding 1,3-propanediol dehydrogenase in a K. pneumoniae mutant lacking both lactate and 2,3-BDO synthesis pathways, the intracellular redox balance was partially restored, but the yield of 1,3-PDO was not increased.29 Lama et al. tried to overexpress the whole 1,3-PDO synthesis module in a plasmid (dhaBCE and gdrAB genes encoding glycerol dehydratase and its activator under the tac promoter and dhaT gene under the T5 promoter) and showed that it was efficient in reducing pyruvate accumulation and increasing the production of 1,3-PDO.30 The reduction of pyruvate and acetate accumulation may also be achieved by activating pyruvate dehydrogenase and the TCA cycle. Citrate synthase with a mutation of K167A or R164K and dihydrolipoyl dehydrogenase (a subunit of pyruvate dehydrogenase) with a mutation of E354G are known to be less sensitive to NADH inhibition. Deleting the arcA gene encoding a key transcriptional regulator of the TCA cycle and introducing K167A into citrate synthase enhanced the metabolic flux toward the TCA cycle and was shown to be able to increase the titer of 1,3-PDO from 9.58 g L−1 to 16.71 g L−1 under microaerobic conditions.31 Introducing R164K into citrate synthase and E354G into dihydrolipoyl dehydrogenase also reduced the accumulation of acetate.30 Deletion of the pck gene encoding phosphoenolpyruvate carboxykinase to reduce the formation of phosphoenolpyruvate can also be helpful in reducing pyruvate and acetate accumulation.32 Alternatively, Wang et al. tried to introduce a polyhydroxybutyrate (PHB) synthesis pathway in K. pneumoniae which provided another route to consume acetyl-CoA and was shown to be useful for reducing acetate accumulation.33

Although K. pneumoniae is an efficient chassis for 1,3-PDO production, it is a conditionally pathogenic microorganism and its industrial application is limited in some countries. Lactobacillus species which are considered GRAS microorganisms have also been explored for 1,3-PDO production. For example, L. diolivorans and L. reuteri strains have been widely screened and the culture conditions have been optimized to increase the production of 1,3-PDO.18,34–36L. reuteri cannot utilize glycerol as a sole carbon source for 1,3-PDO production since it does not contain a glycerol oxidation pathway. Thus, glucose should be used as a co-substrate and some byproducts, including lactate, acetate, and ethanol, are produced during the fermentation. With the development of a genetic toolbox for Lactobacillus sp., L. reuteri was metabolically engineered by deleting the afh2 gene encoding alcohol dehydrogenase and overexpressing pduQ encoding 1,3-PDO dehydrogenase, and the engineered strain can accumulate 687 mM 1,3-PDO.37 However, in the later period of fermentation, the 1,3-PDO production rate was seriously reduced due to the accumulation of lactate. Thus, more systematic genetic modifications should be carried out to engineer Lactobacillus sp. as a potential 1,3-PDO producer.

Clostridium species are another category of natural 1,3-PDO producers. Wild-type C. butyricum and C. pasteurianum have been widely investigated for 1,3-PDO production38,39 which can produce 70–100 g L−1 1,3-PDO with the accumulation of acetate and butyrate as the main byproducts. However, due to the lack of efficient genetic toolboxes, metabolic engineering strategies have not been widely explored for engineering these strains. Recently, the CRISPR/Cas9 system was successfully introduced into C. pasteurianum to increase the production of butanol.40 However, genetic engineering of C. pasteurianum in order to increase 1,3-PDO production has not been reported. A recombinant C. beijerinckii was previously constructed by overexpressing the gdh gene encoding glycerol dehydrogenase and dhaKLM genes encoding dihydroxyacetone kinase.41 However, the achieved titer (26.1 g L−1) in fed-batch fermentation was significantly lower than that of other 1,3-PDO producers.

Citrobacter freundii has also been screened and engineered for 1,3-PDO production. A recent study showed that introducing a heterologous 1,3-PDO oxidoreductase gene in C. freundii could increase the titer of 1,3-PDO from 25.5 g L−1 to 35.6 g L−1 in fed-batch fermentation.42 However, the titer and productivity are still significantly lower than those of other 1,3-PDO producers and the prominent accumulation of lactate and acetate also needs to be addressed.42,43

2.2 Engineering of new chassis

In the past ten years, several non-natural 1,3-PDO producers have also been engineered to produce 1,3-PDO from glycerol. Compared to natural producers, classical microbial chassis such as E. coli have a clear genetic background and can be easily modified by well-developed genetic toolboxes. By introducing a 1,3-PDO synthesis module (dhaBCE, gdrAB, and dhaT genes) from K. pneumoniae, a recombinant E. coli can produce 8.6 g L−1 1,3-PDO in fed-batch fermentation using glycerol and glucose as co-substrates. The titer of 1,3-PDO increased to 13.2 g L−1 when the dhaT gene was substituted by the yqhD gene encoding NADPH-dependent alcohol from E. coli.44 To regenerate NADPH, Yang et al. tried to introduce an NADP+-dependent glyceraldehyde-3-phosphate dehydrogenase encoded by the gapN gene into the recombinant E. coli.45 By tuning the expression level of the gapN gene by adjusting the 5′-untranslated regions (5′-UTR) and substituting the phosphoenolpyruvate (PEP)-dependent glucose transport system with the ATP-dependent transport system, the engineered E. coli can produce 13.47 g L−1 1,3-PDO with a yield of 0.64 mol mol−1 glycerol in batch fermentation using glycerol and glucose as co-substrates. Similarly, Przystałowska et al. constructed a recombinant E. coli by introducing dhaBCE and gdrAB genes from C. freundii and the dhaT gene from K. pneumoniae, which produced 10.6 g L−1 1,3-PDO with a yield of 0.4 mol mol−1 using glycerol as the sole carbon source.46

The marine bacterium Vibrio natriegens is a potential new industrial chassis which has a remarkably short doubling time of less than 10 min and a high substrate uptake rate.47–49 Recently, V. natriegens was systematically engineered by our group to efficiently produce 1,3-PDO from both refined and crude glycerol.50 To construct a superior strain, five different strategies were employed, including (1) optimizing the 1,3-PDO synthesis modules and reducing the accumulation of the toxic intermediate 3-HPA by combinatorial optimization of the expression levels of glycerol dehydratase and alcohol dehydrogenase genes; (2) blocking all byproduct formation pathways (ΔadhEΔldhAΔpflΔpta-ackAΔfrdABCDΔaldAΔaldB); (3) enhancing NADPH generation via the overexpression of pntAB genes encoding membrane-bound pyridine nucleotide transhydrogenase and deletion of the sthA gene encoding soluble pyridine nucleotide transhydrogenase; (4) accelerating glycerol consumption and TCA cycle via the deactivation of two transcriptional regulators ArcA and GlpR; and (5) enhancing the production of 1,3-PDO by process optimization. The final engineered strain can produce 56.2 g L−1 1,3-PDO with a yield of 0.61 mol mol−1 glycerol and a productivity of 2.36 g L−1 h−1. More importantly, no other byproducts accumulated during fermentation, providing a competitive advantage for a simplified downstream process.

Although E. coli and V. natriegens can be easily modified via genetic tools, a key disadvantage is that both strains cannot naturally synthesize coenzyme B12, a key cofactor for B12-dependent glycerol dehydratase. Pseudomonas denitrificans, which can synthesize vitamin B12 aerobically,51,52 is considered another promising host for 1,3-PDO production. By introducing dhaBCE and gdrAB genes from K. pneumoniae and the yqhD gene from E. coli, blocking the acetate and 3-hydroxypropionic acid (3-HP) synthesis pathways, and simultaneously disturbing NADH dehydrogenase, the engineered strain can produce 440 mM 1,3-PDO with a yield of 0.89 mol mol−1 glycerol via a two-stage fed-batch fermentation using gluconate and glycerol as co-substrates.51 However, the process accumulated a few byproducts and the productivity (0.66 g L−1 h−1) and titer were still too low for industrial application. Another strategy to develop a B12-free bioprocess is to introduce B12-independent glycerol dehydratase into the desired chassis. By introducing the dhaB1B2 genes encoding S-adenosylmethionine-dependent glycerol dehydratase and the dhaT gene from C. butyricum, C. acetobutylicum was engineered to produce 1104 mM 1,3-PDO with a yield of 0.65 mol mol−1 and a productivity of 1.70 g L−1 h−1. However, the strain also accumulated butyrate and acetate as byproducts.53

2.3 Other strategies

2.3.1 Design of co-production system to increase atom economy. The production of 1,3-PDO as the sole product is generally considered the primary goal of most metabolic engineering work as this can simplify the downstream process. However, since 1,3-PDO is a reduced product and the obtained yield of 1,3-PDO is generally lower than 0.70 mol mol−1 glycerol, the atom economy of these processes is not satisfactory. To increase the atom economy and economic viability of the bioprocess, it is possible to design a co-production system by coupling the production of 1,3-PDO with another value-added oxidized product (Table 3). Several factors should be considered for such systems. First, NADH should be well recycled by the designed system. Second, the physical properties of the two products should be significantly different, enabling easy separation of the two products. Finally, the designed products should not be very toxic to cells. As mentioned before, co-production of 1,3-PDO and lactate was achieved by blocking other byproduct synthesis pathways in K. pneumoniae and a total conversion yield of 0.95 mol mol−1 glycerol was obtained.28 The co-production of 1,3-PDO with 3-HP has also been widely investigated in K. pneumoniae.54–56 By introducing an aldehyde dehydrogenase and blocking byproduct formation pathways, the highest yield of 1,3-PDO and 3-HP reached 0.82 mol mol−1 glycerol in K. pneumoniae.57 We also designed a 1,3-PDO-glutamate co-production system by introducing the 1,3-PDO synthesis pathway into Corynebacterium glutamicum.58 During the conversion of glucose into glutamate by the engineered C. glutamicum, excessive NADH can be utilized for 1,3-PDO production, boosting the co-production of the two desired products. Since C. glutamicum does not bear other glycerol degradation pathways, the designed process can reach an extremely high yield of 1,3-PDO (approximately 1.0 mol mol−1 glycerol).58
Table 3 Co-production of 1,3-PDO with another product by typical metabolically engineered strains
Co-products Strain Titer of 1,3-PDO Titer of Co-product Yield Culture Ref.
a PHB: polyhydroxybutyrate.
1,3-PDO and 3-HP K. pneumoniae 49.3 g L−1 24.4 g L−1 0.61 mol of (1,3-PDO + 3-HP) per mol of glycerol Fed-batch 54
1,3-PDO and isoprene E. coli 2.5 g L−1 0.67 g L−1 1.0 mol of 1,3-PDO per mol of glycerol, 0.3 mol isoprene per mol of glucose Flask 59
1,3-PDO and D-lactate K. oxytoca 76.2 g L−1 111.9 g L−1 0.95 mol of (1,3-PDO + D-lactate) per mol of glycerol Fed-batch 28
1,3-PDO and glutamate C. glutamicum 14.4 g L−1 32.5 g L−1 0.89 mol of 1,3-PDO per mol of glycerol, 0.50 mol of glutamate per mol of glucose Flask 58
1,3-PDO and PHBa K. pneumoniae 91.2 g L−1 2.56 g L−1 0.59 mol of 1,3-PDO per mol of glycerol Fed-batch 33

The co-production of 1,3-PDO with another volatile product such as isoprene has also been reported. By introducing genes of 1,3-PDO and isoprene synthesis pathways, and deleting the glycerol kinase gene (ΔglpK), and overexpressing pntAB genes to satisfy the cofactor balance, the engineered E. coli produced 2532.1 mg L−1 1,3-PDO and 665.1 mg L−1 isoprene in flask fermentation with yields of 1.0 mol of 1,3-PDO per mol of glycerol and 0.3 mol of isoprene per mol of glucose.59 This process does not increase the complexity or additional cost of the separation process, providing an interesting route for co-producing value-added compounds. However, the economic competitiveness of these processes should be further evaluated by considering the price of 1,3-PDO and another targeted product and the additional cost for product separation.

2.3.2 Production of 1,3-PDO by microbial consortium. While pure culture is widely used in current fermentation processes, mixed culture using natural or artificial microbial consortium has also been explored for the production of 1,3-PDO from glycerol. The advantages of using microbial consortium for bioproduction include wider substrate utilization spectrum, higher tolerance to inhibition of substrates or fermentation products, reduced metabolic burden, no need for sterilization, etc. Natural microbial consortia from different environments (e.g., waste sludge) have been adapted for 1,3-PDO production, which can tolerate up to 200 g L−1 glycerol and produce 70–90 g L−1 1,3-PDO under non-sterilized conditions.60 The obtained yields and productivities by mixed culture are comparable to pure culture but with fewer byproducts due to the synergy of different microorganisms (e.g. consumption of acetate by some microorganisms). However, the stability of the system for large-scale fermentation should be further tested.

Besides natural consortia, artificial co-culture systems have also been tested. By cultivating L. reuteri, which can secrete 3-HPA, with E. coli overexpressing the aldehyde dehydrogenase gene gabD4 and the 1,3-PDO dehydrogenase gene pduQ, the engineered system can accumulate 125.93 g L−1 3-HP and 88.46 g L−1 1,3-PDO in fed-batch fermentation.61 Another co-cultivation system consisted of two recombinant E. coli strains, one overexpressing dhaB1B2 from C. butyricum and another overexpressing dhaT from K. pneumoniae. This co-fermentation system took glycerol and glucose as substrates and produced 41.65 g L−1 1,3-PDO with a yield of 0.67 mol mol−1 glycerol.62 For these systems, different cells should be cultured separately, and the collected cells should be mixed at a specific ratio during the fermentation stage, which may be difficult to operate in an industrial environment.

2.3.3 Production of 1,3-PDO via electrofermentation. Production of 1,3-PDO from glycerol requires NADH. Apart from increasing the intrinsic NADH supply which will inevitably lead to carbon loss of the substrate, another possible approach is to introduce electrodes to provide electrons. Electrofermentation is a recently developed platform which is particularly interesting for the production of oxidized or reduced products. In the cathodic compartment with electrons provided, the metabolism of suspended cells would shift to NADH-consuming pathways such as 1,3-PDO formation. Bioelectrochemical systems (BESs) were first applied to mixed culture fermentation of glycerol and it was found that reductive current can be integrated into glycerol metabolism to increase the yield of 1,3-PDO and that the metabolic flux was redirected from propionate fermentation to 1,3-PDO production.63 The production of 1,3-PDO was positively correlated with the current intensity;64 however, the electroactive microorganisms in the mixed culture would be outcompeted from the biofilm during long-term operation,65 leading to decreased abundance of 1,3-PDO producers and reduced yield of 1,3-PDO.64 The limitation of cell succession in the mixed culture could be avoided by using pure electroactive microorganisms. For instance, Clostridium pasteurianum DSM 525 and Klebsiella pneumoniae L17 were identified to be electroactive heterotrophs and have been applied in BESs to enhance the production of 1,3-PDO from glycerol.66–68 Only a small amount of electrons from the cathode could considerably compensate for the shortcomings of NADH and enhance the production of 1,3-PDO.66 Different redox mediators have also been tested to increase the interaction between the cathode and the producers. Among various electron shuttles, the application of neutral red was shown to be beneficial for 1,3-PDO production by Klebsiella pneumoniae L17 while utilization of Brilliant Blue was more favorable for 1,3-PDO production by Clostridium pasteurianum DSM 525.67,68

3. Production of 1,3-PDO from sugars

3.1 Production of 1,3-PDO from sugars via recombination of natural pathways

Production of 1,3-PDO from abundant sugars is an appealing alternative for the industry since the supply and price of sugar feedstocks are more stable than that of glycerol. No natural microorganisms can directly produce 1,3-propanediol from sugars. Thus, combining natural glycerol and 1,3-PDO synthesis modules in one producer is the most direct way to develop a sugar-based 1,3-PDO production process (Fig. 1 and Table 4). The most successful case is the development of a recombinant E. coli by DuPont and Genencor which can produce 135 g L−1 1,3-PDO from glucose with a high yield (1.20 mol mol−1) and productivity (3.5 g L−1 h−1).6 Although the details for the development of the industrial strain are not fully known, the key strategies include (1) blocking native glycerol oxidative pathways by deleting the glpK gene encoding glycerol kinase and the gldA gene encoding glycerol dehydrogenase; (2) introducing a glycerol synthesis module by overexpressing the glycerol-3-phosphate dehydrogenase gene gpd1 and the glycerol-3-phosphate phosphatase gene gpp2 from Saccharomyces cerevisiae; (3) introducing a 1,3-PDO synthesis module by overexpressing the dha operon from K. pneumoniae; (4) destroying the PEP-dependent glucose phosphotransferase system (PTS) by deleting ptsHI and crr genes and upregulating the non-PTS-dependent glucose transport system by overexpressing the galP gene encoding galactose permease and the glk gene encoding glucose kinase; and (5) downregulation of the gapA gene encoding glyceraldehyde-3-phosphate dehydrogenase to reduce the flux toward glycolysis.6 In particular, utilization of the endogenous NADPH-dependent alcohol dehydrogenase gene yqhD to substitute the dhaT gene from K. pneumoniae was shown to be important to achieve a high titer of 1,3-PDO under aerobic conditions. This is probably because the NADPH/NADP+ ratio is higher than NADH/NAD+ under aerobic conditions. On the other hand, the substitution of the PTS-dependent glucose transport system by the non-PTS-dependent system is important for increasing the yield of 1,3-PDO in E. coli since the phosphorylation of 1 mol of glucose via the PTS system consumes 1 mol of PEP and generates 1 mol of pyruvate, limiting the maximum yield of 1,3-PDO to 1 mol mol−1 if pyruvate cannot be reconverted to PEP. This limitation can be overcome by bacteria harboring the pyruvate–PEP recycling route. For example, C. glutamicum can regenerate PEP from pyruvate via the pyruvate–oxoacetate–PEP route by two active enzymes, namely pyruvate carboxylase and PEP carboxykinase. Thus, the maximum yield of 1,3-PDO in C. glutamicum can reach 1.33 mol mol−1 glucose even with the PTS system. Our group has constructed a recombinant C. glutamicum which can produce 110 g L−1 1,3-PDO with a yield of 1.0 mol mol−1 (0.42 g g−1) glucose without disrupting the native PTS system69 (Table 4).
Table 4 Production of 1,3-PDO from glucose by typical metabolically engineered strains
Strain Strategies Titer Yield Productivity Substrate Culture Ref.
K. pneumoniae J2B ΔptsG, ΔglpF, ΔldhA, ΔadhE, ΔfrdA, ΔpflB, Δedd-eda, ΔmgsA, Δpta-ackA, ΔpoxB, Δmdh, Δndh, ΔbudB; gltAR164K, lpdAE324G; overexpression of gdp1, gpp2, glf, dhaB, gdrAB, dhaT 62 g L−1 1.27 mol mol−1 0.94 g L−1 h−1 Glucose Fed-batch 30
S. cerevisiae Overexpression of dhaB and dhaT 1.2 g L−1 Glucose Flask 91
E. coli Overexpression of gpd1, gpp2, dha operon except dhaT, yqhD, galP, and glk; ΔglpK, ΔgldA, ΔptsHI, Δcrr, Δedd-eda, ΔmgsA, Δpta-ackA, Δndh, ΔarcA; downregulation of gapA 135 g L−1 1.20 mol mol−1 3.5 g L−1 h−1 Glucose Fed-batch 6
C. glutamicum Overexpression of pduCEDGH and yqhD; overexpression of gpd1, gpp2, and hdpA-gldA; ΔldhA, ΔpoxB, Δpyk, Δald, Δadh, Δppc, Δzwf; downregulation of gapA 110 g L−1 1.00 mol mol−1 2.3 g L−1 h−1 Glucose Fed-batch 69

Recently, K. pneumoniae was also engineered to produce 1,3-PDO from glucose by introducing the glycerol synthesis module from S. cerevisiae.70,71 By overexpressing gpd1 and gpp2 in a K. pneumoniae mutant lacking glycerol oxidative pathways (ΔglpKΔglpDΔgldAΔdhaD), the engineered strain can accumulate a high amount of glycerol.70,71 However, the yield of 1,3-PDO was low due to the low expression of the dha operon and the limited availability of coenzyme B12. By overexpressing the dhaBCE, gdrAB, and dhaT genes, the yield of 1,3-PDO was increased to 0.27 mol mol−1.70,71 To further increase the yield of 1,3-PDO, several additional strategies were also employed, including (1) destroying the PTS system (ΔptsG) and overexpressing the glucose facilitator gene glf from Zymomonas mobilis; (2) blocking byproduct formation pathways (ΔldhAΔadhEΔfrdAΔpflBΔedd-edaΔmgsAΔpta-ackAΔpoxBΔmdh) and glycerol secretion pathways (ΔglpF1ΔglpF2); and (3) modulating the TCA cycle and electron transport chain by introducing the E354G mutation to dihydrolipoyl dehydrogenase, and the R164K mutation to citrate synthase, and deleting ndh gene encoding NADH dehydrogenase II. The final engineered strain could produce 62 g L−1 1,3-PDO with a high yield of 1.27 mol mol−1 glucose in two-stage fed-batch fermentation.30 However, vitamin B12 was still needed to be added during fermentation due to the limited supply of coenzyme B12 under the employed fermentation conditions.

3.2 Production of 1,3-PDO from sugars via non-natural pathways

Although 1,3-PDO can be produced from glucose via the recombination of natural glycerol and 1,3-PDO synthesis pathways, all of the developed bioprocesses based on this route still need to add vitamin B12 during fermentation. Considering the high price of vitamin B12, three non-natural metabolic pathways have been proposed and experimentally verified, enabling the de novo production of 1,3-PDO from different sugars without using vitamin B12 (Table 5).
Table 5 Production of 1,3-PDO via non-natural pathways
Pathways Strain Titer Culture Substrate Ref.
Homoserine-derived pathway E. coli W3110 0.63 g L−1 Flask Glucose 73
E. coli MG1655 3.03 g L−1 Fed batch Glucose 74
Malate-derived pathway E. coli MG1655 7.61 mg L−1 Flask Glucose 76
Malonyl-CoA derived pathway E. coli W3110 7.98 g L−1 Fed batch Glucose 78
2.36 g L−1 Flask Glycerol
0.82 g L−1 Xylose
0.44 g L−1 Acetate

The first glycerol-independent route was established by converting homoserine into 1,3-PDO via three enzymatic steps (Fig. 2, the red line). In this artificial pathway, homoserine is first converted into 2-keto-4-hydroxybutyrate by an amino acid transferase or amino acid dehydrogenase, and 2-keto-4-hydroxybutyrate is further converted into 3-HPA by an α-keto acid decarboxylase. 3-HPA can be easily reduced to 1,3-PDO by alcohol dehydrogenase. The key challenge for establishing this non-natural pathway is the lack of efficient and specific enzymes for catalyzing the first two steps. Glutamate dehydrogenase was initially modified to increase its activity toward homoserine by introducing two mutations K92V/T195S. By overexpressing the mutated glutamate dehydrogenase gene together with the pdc gene encoding pyruvate decarboxylase (PDC) from Z. mobilis, the engineered E. coli can accumulate 51.5 mg L−1 1,3-PDO using homoserine and glucose as co-substrates.72 Further enzyme screening showed that aspartate transaminase encoded by the aspC gene was more efficient for homoserine deamination and PDC with a mutation of I472A had higher specificity toward 2-keto-4-hydroxybutyrate. By overexpressing aspC and pdcI472A and increasing the biosynthesis of homoserine, the engineered E. coli could produce 0.63 g L−1 1,3-PDO in shake flasks using glucose as the sole carbon source.73 Similarly, by overexpressing a mutated phosphoserine aminotransferase gene serCR42W/R77W and pdc, another E. coli strain was engineered to be able to accumulate 3.03 g L−1 1,3-PDO in fed-batch fermentation.74 Further increasing the activities and specificities of the two key enzymes by protein engineering and the precursor and NADPH availability by metabolic engineering is desired for improving the titer of 1,3-PDO for practical application.


image file: d1gc04288b-f2.tif
Fig. 2 Non-natural metabolic pathways for the production of 1,3-PDO. Abbreviations: thrA, threonine-insensitive bifunctional aspartokinase/homoserine dehydrogenase; gdhA, glutamate dehydrogenase; aspC, aspartate transaminase; serC, phosphoserine aminotransferase; pdc, pyruvate decarboxylase; yqhD, alcohol dehydrogenase; lysC*, malate kinase (mutation at V115A/E119S/E250K/E434V); asd, malate semialdehyde dehydrogenase; ssr, malate semialdehyde reductase; lldD, lactate dehydrogenase; ppc, phosphoenolpyruvate carboxylase; pyc, pyruvate carboxylase; mdh, malate dehydrogenase; mcrC*, malonyl-CoA reductase (mutation at N940V/K1106W/S1114R); prpE, 3-hydroxypropionyl-CoA synthetase; car, carboxylic acid reductase; pduP, aldehyde dehydrogenase; mdh2, NAD-dependent methanol dehydrogenase; khb, 2-keto-4-hydroxybutyrate aldolase; kdc, α-keto acid decarboxylase; adh1, alcohol dehydrogenase; DERA, deoxyribose-5-phosphate aldolase.

Frazao et al. proposed another non-natural 1,3-PDO biosynthetic pathway which is derived from malate75,76 (Fig. 2, the orange line). The whole pathway consists of six steps and can be divided into two modules: a 2,4-dihydroxybutyrate synthetic module and a 1,3-PDO synthesis module. In the first module, malate is transferred to malyl-4-phosphate by malate kinase; malyl-4-phosphate is further converted to malate-4-semialdehyde by malate semialdehyde dehydrogenase, and malate-4-semialdehyde can be oxidized to 2,4-dihydroxybutyrate by aldehyde dehydrogenase. In the downstream module, 2,4-dihydroxybutyrate is first converted into 2-keto-4-hydroxybutyrate by dehydrogenase and 2-keto-4-hydroxybutyrate can be converted into 1,3-PDO by an α-keto acid decarboxylase and an alcohol dehydrogenase as mentioned before. Although this pathway is thermodynamically feasible, it is highly challenging to implement due to the lack of corresponding enzymes for the first five steps. Although detectable activities can be obtained by protein engineering, simultaneous expression of the enzymes in an engineered E. coli only resulted in a low amount of 1,3-PDO (0.1 mM). Large efforts should be made to further improve the activities of the five enzymes and to balance the expression of pathway genes.

Our group has proposed a third B12-independent pathway to produce 1,3-PDO by the reduction of 3-HP (Fig. 2, the yellow line). 3-HP can be converted into 1,3-PDO via two potential routes: a 3-hydroxypropionyl-CoA dependent pathway and a direct reduction pathway. Direct reduction of 3-HP into 1,3-PDO can be achieved by employing carboxylic acid reductase (CAR) and alcohol dehydrogenase.77,78 This pathway is not effective enough due to the low activity of CARs toward short-chain carboxylic acids. The other pathway employs 3-hydroxypropionyl-CoA synthetase from Metallosphaera sedula to convert 3-HP into 3-hydroxypropionyl-CoA and the latter can be further converted into 1,3-PDO via a CoA-acylating aldehyde dehydrogenase from Salmonella typhimurium and the alcohol dehydrogenase YqhD. This pathway is very promising for 1,3-PDO production due to the high activities of pathway enzymes. By combining the malonyl-CoA-derived 3-HP synthesis pathway and the 3-hydroxypropionyl-CoA-dependent 1,3-PDO synthesis pathway, an engineered E. coli strain was constructed which could produce 7.98 g L−1 1,3-PDO from glucose in fed-batch fermentation without adding vitamin B12. Moreover, the strain can also be applied to convert different substrates to 1,3-PDO, including glycerol, xylose, and acetate (Table 5). Further engineering should be conducted to increase the specificity of CoA-acylating dehydrogenase since this enzyme is also highly active to acetyl-CoA which results in the accumulation of ethanol as a byproduct.

The design and successful implementation of non-natural pathways open the way to develop new and economical bioprocesses for 1,3-PDO production from diversified feedstocks. Since the new pathways employ common metabolic intermediates (homoserine, malate, or acetyl-CoA) as precursors and the involved enzymes do not need special cofactors, it is possible to transfer the pathways into different microorganisms to develop industrially interesting bioprocesses. For example, different microorganisms have been engineered to produce 3-HP from CO2 or fatty acids.79–81 With the introduction of the 3-HP to the 1,3-PDO module in the engineered microorganisms, it is possible to directly produce 1,3-PDO from CO2 or fatty acids. Direct production of 1,3-PDO from CO2 or industrial waste gas represents a green process for 1,3-PDO production. Waste oil, which can be hydrolyzed to fatty acids and glycerol, can also be a very promising feedstock for 1,3-PDO production due to its high reduction degree. However, the efficiency of the new pathways is still low due to the low activities and specificities of the pathway enzymes. Thus, protein engineering and metabolic engineering strategies should be combined to further improve the pathway efficiency toward practical application.

4. Production of 1,3-PDO from other carbon sources

Besides glycerol and sugars, other feedstocks including CO2 and alcohols have also been explored for the biological production of 1,3-PDO.

4.1 Production of 1,3-PDO from CO2

Cyanobacteria which can use light energy to convert CO2 into value-added products are becoming an attractive platform for bioproduction. Synechococcus elongatus was employed as a chassis to produce 1,3-PDO by introducing the glycerol synthesis module (gpd1 and gpp2) from S. cerevisiae, glycerol dehydratase and its activator (dhaBCE and gdrAB) from K. pneumoniae, and yqhD from E. coli. By optimizing the selective promoters and culture medium, 1.22 g L−1 1,3-PDO and 0.87 g L−1 glycerol could be accumulated in 20 days. Notably, it was not necessary to add vitamin B12 during fermentation due to the presence of intrinsic pseudo-vitamin B12 in S. elongatus.82–84 Co-cultivation of S. elongatus expressing only a heterologous glycerol synthesis module with K. pneumoniae also allowed the accumulation of 40 mg L−1 1,3-PDO from CO2 within 5 days.85 Furthermore, considering the incompatibility of the highly oxidative environment in cyanobacteria with the oxygen-sensitive glycerol dehydratase, the heterologous 1,3-PDO synthesis pathway was encapsulated into the heterocyst of cyanobacterium Anabaena PCC7120 to protect the oxygen-sensitive enzyme.86 Compared with the situation without heterocyst, the engineered Anabaena strain accumulated a 1.7-fold higher amount of 1,3-PDO (46.0 mg L−1). However, the titer and productivity are still too low for practical application.

4.2 Production of 1,3-PDO from alcohols

Methanol can be a potential feedstock for biorefinery due to its high abundance, low price, and high reduction degree. Direct production of 1,3-PDO from methanol has not been achieved thus far. Instead, two new pathways to use methanol as a co-substrate for 1,3-PDO production were proposed and conceptually verified. The first route employs methanol dehydrogenase (encoded by the mdh2 gene) to catalyze the reduction of methanol to formaldehyde and 2-keto-4-hydroxybutyrate aldolase (encoded by the khb gene) for the condensation of pyruvate with formaldehyde, resulting in the production of 2-keto-4-hydroxybutyrate (Fig. 2, the green line). 2-Keto-4-hydroxybutyrate can be converted into 1,3-PDO by an α-keto acid decarboxylase and an alcohol dehydrogenase as mentioned before. However, since methanol dehydrogenase and 2-keto-4-hydroxybutyrate aldolase have very high Km values for methanol and formaldehyde (>500 mM), the engineered E. coli overexpressing khb, kdc encoding α-keto acid decarboxylase, and dhaT genes only accumulated 32.7 mg L−1 1,3-PDO using glucose and methanol as co-substrates.87 The second biological route tried to use methanol (or formaldehyde) and ethanol as co-substrates for 1,3-PDO production88 (Fig. 2, the blue line). The two alcohols can be converted into formaldehyde and acetaldehyde by methanol dehydrogenase and alcohol dehydrogenase, which can be further condensed to 3-HPA by deoxyribose-5-phosphate aldolase (DERA) and reduced to 1,3-PDO by alcohol dehydrogenase. Principally, this is a short pathway and it has a very high atom economy. However, due to the low activities of methanol dehydrogenase and deoxyribose-5-phosphate aldolase, only 1.05 mM 1,3-PDO was accumulated using methanol and ethanol as co-substrates. When a high concentration of formaldehyde and ethanol were used as co-substrates, 1.32 g L−1 1,3-PDO could be accumulated by fed-batch fermentation.

5. Summary and prospects

The development of green processes with low production costs is highly important for the broad application of 1,3-PDO. Considering the price of ethylene glycerol and 1,4-butanediol, the two most important monomers for the synthesis of polyethylene terephthalate (PET) and polybutylene terephthalate (PBT), it is estimated that the price of 1,3-PDO should be lower than US$3000 per ton for its large-scale application in the polymer industry. Considering the significant fluctuation of glucose or glycerol price (US$300–500 per ton), the titer, yield, and productivity of 1,3-PDO should be higher than 100 g L−1, 0.4 g g−1, and 2.5 g L−1 h−1 for commercial production.8 Since a large proportion of 1,3-PDO production cost comes from substrates and downstream processes, it is highly desirable to develop efficient production routes to convert cheap and abundant feedstocks into 1,3-PDO with high yield and low accumulation of byproducts. Thus, there are four major challenges for developing industrially feasible biological routes, including (1) expanding the scope of substrate utilization; (2) increasing the available electron supply (in the form of NAD(P)H) to increase the yield of 1,3-PDO; (3) reducing or eliminating the accumulation of byproducts to simplify the downstream process; and (4) avoiding the supplementation of expensive cofactors (such as vitamin B12) and nutrients (such as yeast extract). The recent development of metabolic engineering, synthetic biology, and new bioproduction systems could provide some interesting strategies to address the abovementioned challenges.

Currently, only glycerol and glucose have been used as feedstocks for the biological production of 1,3-PDO in industry. The glycerol-based production systems using natural or engineered 1,3-PDO producers (K. pneumoniae or C. butyricum) still accumulate different kinds of byproducts.32,53 Although it is principally possible to develop a homo-1,3-PDO production system by eliminating all byproduct formation pathways and activating the TCA cycle, it is very challenging to completely eliminate the complex native metabolic regulation. Blocking all pathways leading to byproduct formation often significantly disturbs cellular metabolism, resulting in retarded cell growth and reduced productivity. It is important to apply different omics tools in the future to identify potential bottlenecks and use synthetic biology tools to fine-tune the expression of targeted genes. For example, although the inhibition of citrate synthase by NADH is already known,30,31 the expression of other genes in the TCA cycle may also be bottlenecks limiting the flux to the TCA cycle in microaerobic conditions. Combinatorial optimization of different genes may be necessary to activate the TCA cycle in order to reduce pyruvate and acetate accumulation.30,31 Alternatively, adaptive laboratory evolution may be used as a potential tool to recover cell growth and productivity.

Production of 1,3-PDO from glucose or glycerol using non-natural chassis is generally based on aerobic conditions. The advantage of aerobic fermentation is the low accumulation of byproducts since the TCA cycle can be fully activated. However, the increased energy input during fermentation should be considered when comparing different processes. Moreover, most new producers with heterologous 1,3-PDO synthesis pathways still need to use expensive vitamin B12 during fermentation. The establishment of non-natural pathways74,76,78 circumvents the necessity of B12 addition and also enables the conversion of different substrates into 1,3-PDO. For example, the introduction of 3-HP-based 1,3-PDO synthesis has enabled the conversion of glucose, glycerol, xylose, and acetate to 1,3-PDO by the same strain.78 Principally, microorganisms which can naturally utilize methanol, starch, lignocellulose, CO2, or syngas can be engineered to produce 1,3-PDO with the introduction of this pathway although the efficiency of the pathway should be further increased by protein engineering and metabolic engineering.78,84,88,89 Development of new processes to efficiently utilize these cheap feedstocks may provide potentially sustainable routes for large-scale production of 1,3-PDO. In particular, the utilization of CO2 for 1,3-PDO is promising in the future considering the extremely low price and high abundance of CO2. Future work should focus on increasing the titer and productivity of 1,3-PDO by cyanobacteria or other microorganisms.

Although at an early stage, the recently developed new bioproduction systems based on microbial consortium or electro-fermentation can be interesting for 1,3-PDO production in the future.67,68 Microbial consortia show higher tolerance to the inhibition of substrates and fermentation products compared to pure culture, and no sterilization is required during fermentation which could save some energy.60 Evaluation of the stability of the systems in large-scale fermentation is an important step toward the real application of microbial consortia. Additionally, electro-fermentation provides a potential new way to produce reductive products such as 1,3-PDO by obtaining electrons from a reductive current, avoiding the necessity of generating NADH from a carbon source, which is highly important for increasing the product yield.66–68 New electroactive heterotrophs and more efficient redox mediators can be further developed and applied in 1,3-PDO production. Moreover, the combination of metabolically engineered strains with electro-fermentation may be used to reduce byproduct accumulation and further promote 1,3-PDO production. The development of efficient fermenters for electro-fermentation and integrated processes should also be considered for its practical application.

The application of cell-free systems in 1,3-PDO production can also be a potential direction. Rieckenberg et al. have previously demonstrated an in vitro system for 1,3-PDO production by integrating glycerol dehydratase, NADPH-dependent alcohol dehydrogenase, and hydrogenase. A very high yield of 1,3-PDO (∼0.95 mol mol−1) was reached due to the use of hydrogen as the electron donor.90 Other non-natural pathways may also be tested in cell-free systems with the consideration of pathway length and the requirement of cofactors. For example, the alcohol-based (methanol and ethanol) pathway may be technologically feasible for large-scale production of 1,3-PDO using cell-free systems since it is very short and has a high theoretical yield. Screening or engineering of efficient and stable enzymes to reduce the cost of enzymes is important for its practical application.

Author contributions

ZC: conceptualization, funding acquisition, writing – review and editing, visualization. FHZ: investigation, resources, writing – original draft, visualization. DHL: funding acquisition, supervision.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

This work was supported by the National Key R&D Program of China (No. 2018YFA0901500, 2019YFE0196900, and 2021YFC2100900), the National Natural Science Foundation of China (Grant No. 21938004, 22078172, and 21878172) and the DongGuan Innovative Research Team Program (No. 201536000100033).

References

  1. S. Jun, C. Moon, C. Kang, S. Kong, B. Sang and Y. Um, Appl. Biochem. Biotechnol., 2010, 161, 491–501 CrossRef CAS PubMed .
  2. H. Liu, Y. Xu, Z. Zheng and D. Liu, Biotechnol. J., 2010, 5, 1137–1148 CrossRef CAS PubMed .
  3. J. Kurian, J. Polym. Environ., 2005, 13, 159–167 CrossRef CAS .
  4. F. Wu, H. Jiang, X. Zhu, R. Lu, L. Shi and F. Lu, ChemSusChem, 2021, 14, 569–581 CrossRef CAS PubMed .
  5. J. Wang, M. Yang and A. Wang, Chin. J. Catal., 2020, 41, 1311–1319 CrossRef CAS .
  6. C. Nakamura and G. Whited, Curr. Opin. Biotechnol., 2003, 14, 454–459 CrossRef CAS PubMed .
  7. D. Liu, Abstr. Pap. Am. Chem. Soc., 2014, 248, 487 Search PubMed .
  8. B. Hermann and M. Patel, Appl. Biochem. Biotechnol., 2007, 136, 361–388 CrossRef CAS PubMed .
  9. X. Wang, J. Zhou, J. Shen, Y. Zheng, Y. Sun and Z. Xiu, Appl. Microbiol. Biotechnol., 2020, 104, 9179–9191 CrossRef CAS PubMed .
  10. Y. Lan, J. Feng, X. Guo, H. Fu and J. Wang, Bioprocess Biosyst. Eng., 2021, 44, 2375–2385 CrossRef CAS PubMed .
  11. J. Yun, M. Yang, T. Magocha, H. Zhang, Y. Xue, G. Zhang, X. Qi and W. Sun, Bioresour. Technol., 2018, 247, 838–843 CrossRef CAS PubMed .
  12. Y. Zhang, D. Liu and Z. Chen, Biotechnol. Biofuels, 2017, 10, 299 CrossRef PubMed .
  13. S. Zhou, L. Li, M. Perseke, Y. Huang, J. Wei and Q. Qin, Sci. Bull., 2015, 60, 511–521 CrossRef CAS .
  14. M. Metsoviti, K. Paraskevaidi, A. Koutinas, A. Zeng and S. Papanikolaou, Process Biochem., 2012, 47, 1872–1882 CrossRef CAS .
  15. E. Wilkens, A. Ringel, D. Hortig, T. Willke and K. Vorlop, Appl. Microbiol. Biotechnol., 2012, 93, 1057–1063 CrossRef CAS PubMed .
  16. M. Metsoviti, A. Zeng, A. Koutinas and S. Papanikolaou, J. Biotechnol., 2013, 163, 408–418 CrossRef CAS PubMed .
  17. S. Pflugl, H. Marx, D. Mattanovich and M. Sauer, Bioresour. Technol., 2012, 119, 133–140 CrossRef PubMed .
  18. J. Jolly, B. Hitzmann, S. Ramalingam and K. Ramachandran, J. Biosci. Bioeng., 2014, 118, 188–194 CrossRef CAS PubMed .
  19. M. Yang, J. Yun, H. Zhang, T. Magocha, H. Zabed, Y. Xue, E. Fokum, W. Sun and X. Qi, Food Technol. Biotechnol., 2018, 56, 3–15 CAS .
  20. Q. Zhang and Z. Xiu, Biotechnol. Prog., 2009, 25, 103–115 CrossRef CAS PubMed .
  21. Y. Xu, N. Guo, Z. Zheng, X. Ou, H. Liu and D. Liu, Biotechnol. Bioeng., 2009, 104, 965–972 CrossRef CAS PubMed .
  22. Y. Zhang, Y. Li, C. Du, M. Liu and Z. Cao, Metab. Eng., 2006, 8, 578–586 CrossRef CAS PubMed .
  23. M. Seo, J. Seo, S. Heo, J. Baek, D. Rairakhwada, B. Oh, P. Seo, M. Choi and C. Kim, Appl. Microbiol. Biotechnol., 2009, 84, 527–534 CrossRef CAS PubMed .
  24. B. Oh, S. Lee, S. Heo, J. Seo and C. Kim, Microb. Cell Fact., 2018, 17, 92 CrossRef PubMed .
  25. S. Zhou, Y. Huang, X. Mao, L. Li, C. Guo, Y. Gao and Q. Qin, PLoS One, 2019, 14, e0200978 CrossRef CAS PubMed .
  26. M. Durgapal, V. Kumar, T. Yang, H. Lee, D. Seung and S. Park, Bioresour. Technol., 2014, 159, 223–231 CrossRef CAS PubMed .
  27. V. Kumar, M. Durgapal, M. Sankaranarayanan, A. Somasundar, C. Rathnasingh, H. Song, D. Seung and S. Park, Bioresour. Technol., 2016, 214, 432–440 CrossRef CAS PubMed .
  28. B. Xin, F. Tao, Y. Wang, H. Liu, C. Ma and P. Xu, Metab. Eng., 2017, 41, 102–114 CrossRef CAS PubMed .
  29. Y. Cui, J. Zhou, L. Gao, C. Zhu, X. Jiang, S. Fu and H. Gong, J. Appl. Microbiol., 2014, 117, 690–698 CrossRef CAS PubMed .
  30. S. Lama, E. Seol and S. Park, Metab. Eng., 2020, 62, 116–125 CrossRef CAS PubMed .
  31. J. Lee, H. Jung, M. Jung and M. Oh, Biotechnol. Bioprocess Eng., 2019, 24, 95–102 CrossRef CAS .
  32. Y. Zhang, Z. Jia, J. Lin, D. Xu, S. Fu and H. Gong, J. Appl. Microbiol., 2017, 123, 678–687 CrossRef CAS PubMed .
  33. W. Wang, X. Yu, Y. Wei, R. Ledesma-Amaro and X. Ji, Chem. Eng. Sci., 2021, 236, 116539 CrossRef CAS .
  34. K. Lindlbauer, H. Marx and M. Sauer, New Biotechnol., 2017, 34, 32–39 CrossRef CAS PubMed .
  35. M. Ricci, A. Russo, I. Pisano, L. Palmieri, M. de Angelis and G. Agrimi, J. Microbiol. Biotechnol., 2015, 25, 893–902 CrossRef CAS PubMed .
  36. J. Ju, D. Wang, S. Heo, M. Kim, J. Seo, Y. Kim, D. Kim, S. Kang, C. Kim and B. Oh, Microb. Cell Fact., 2020, 19, 6 CrossRef CAS PubMed .
  37. K. Singh, S. Ainala and S. Park, Appl. Geochem., 2021, 131, 125590 Search PubMed .
  38. T. Colin, A. Bories and G. Moulin, Appl. Microbiol. Biotechnol., 2000, 54, 201–205 CrossRef CAS PubMed .
  39. A. Chatzifragkou, G. Aggelis, M. Komaitis, A. Zeng and S. Papanikolaou, Bioresour. Technol., 2011, 102, 10625–10632 CrossRef CAS PubMed .
  40. R. Schmitz, W. Sabra, P. Arbter, Y. Hong, T. Utesch and A. Zeng, Eng. Life Sci., 2019, 19, 412–422 CrossRef CAS PubMed .
  41. D. Wischral, J. Zhang, C. Cheng, M. Lin, L. De Souza, F. Pessoa, N. Pereira and S. Yang, Bioresour. Technol., 2016, 212, 100–110 CrossRef CAS PubMed .
  42. E. Celinska, A. Drozdzynska, M. Jankowska, W. Bialas, K. Czaczyk and W. Grajek, Process Biochem., 2015, 50, 48–60 CrossRef CAS .
  43. F. Barbirato, C. Camarasaclaret, J. Grivet and A. Bories, Appl. Microbiol. Biotechnol., 1995, 43, 786–793 CrossRef CAS .
  44. F. Wang, H. Qu, D. Zhang, P. Tian and T. Tan, Mol. Biotechnol., 2007, 37, 112–119 CrossRef CAS PubMed .
  45. B. Yang, S. Liang, H. Liu, J. Liu, Z. Cui and J. Wen, Bioresour. Technol., 2018, 267, 599–607 CrossRef CAS PubMed .
  46. H. Przystałowska, J. Zeyland, D. Szymanowska-Powałowska, M. Szalata, R. Słomski and D. Lipiński, Microbiol. Res., 2015, 171, 1–7 CrossRef PubMed .
  47. R. Eagon, J. Bacteriol., 1962, 83, 736–737 CrossRef CAS PubMed .
  48. M. Weinstock, E. Hesek, C. Wilson and D. Gibson, Nat. Methods, 2016, 13, 849–851 CrossRef CAS PubMed .
  49. H. Lee, N. Ostrov, B. Wong, M. Gold, A. Khalil and G. Church, Nat. Microbiol., 2019, 4, 1105–1113 CrossRef CAS PubMed .
  50. Y. Zhang, Z. Li, Y. Liu, X. Cen, D. Liu and Z. Chen, Metab. Eng., 2021, 65, 52–65 CrossRef CAS PubMed .
  51. S. Zhou, S. Lama, M. Sankaranarayanan and S. Park, Bioresour. Technol., 2019, 292, 121933 CrossRef CAS PubMed .
  52. M. Arasu, R. Sarkar, B. Sekar, V. Kumar, C. Rathnasingh, J. Choi, H. Song, D. Seung and S. Park, Biotechnol. Bioprocess Eng., 2013, 18, 43–51 CrossRef CAS .
  53. M. Gonzalez-Pajuelo, I. Meynial-Salles, F. Mendes, J. Andrade, I. Vasconcelos and P. Soucaille, Metab. Eng., 2005, 7, 329–336 CrossRef CAS PubMed .
  54. Y. Huang, Z. Li, K. Shimizu and Q. Ye, Bioresour. Technol., 2012, 103, 351–359 CrossRef CAS PubMed .
  55. Y. Huang, Z. Li, K. Shimizu and Q. Ye, Bioresour. Technol., 2013, 128, 505–512 CrossRef CAS PubMed .
  56. S. Ashok, S. Raj, C. Rathnasingh and S. Park, Appl. Microbiol. Biotechnol., 2011, 90, 1253–1265 CrossRef CAS PubMed .
  57. Y. Ko, E. Seol, B. Sekar, S. Kwon, J. Lee and S. Park, Bioresour. Technol., 2017, 244, 1096–1103 CrossRef CAS PubMed .
  58. J. Huang, Y. Wu, W. Wu, Y. Zhang, D. Liu and Z. Chen, Sci. Rep., 2017, 7, 42246 CrossRef CAS PubMed .
  59. J. Guo, Y. Cao, H. Liu, R. Zhang, M. Xian and H. Liu, Appl. Microbiol. Biotechnol., 2019, 103, 2597–2608 CrossRef CAS PubMed .
  60. L. Jiang, H. Liu, Y. Mu, Y. Sun and Z. Xiu, Eng. Life Sci., 2017, 17, 635–644 CrossRef CAS PubMed .
  61. Y. Zhang, H. Zabed, J. Yun, G. Zhang, Y. Wang and X. Qi, ACS Sustainable Chem. Eng., 2021, 9, 4625–4637 CrossRef CAS .
  62. J. Yun, H. Zabed, Y. Zhang, A. Parvez, G. Zhang and X. Qi, Bioresour. Technol., 2021, 319, 124218 CrossRef CAS PubMed .
  63. M. Zhou, J. Chen, S. Freguia, K. Rabaey and J. Keller, Environ. Sci. Technol., 2013, 47, 11199–11205 CrossRef CAS PubMed .
  64. M. Zhou, S. Freguia, P. Dennis, J. Keller and K. Rabaey, Microb. Biotechnol., 2015, 8, 483–489 CrossRef CAS PubMed .
  65. N. Xafenias, M. Anunobi and V. Mapelli, Process Biochem., 2015, 50, 1499–1508 CrossRef CAS .
  66. O. Choi, T. Kim, H. Woo and Y. Um, Sci. Rep., 2014, 4, 6961 CrossRef CAS PubMed .
  67. C. Kim, J. Lee, J. Baek, D. Kong, J. Na, J. Lee, E. Sundstrom, S. Park and J. Kim, ChemSusChem, 2020, 13, 564–573 CrossRef CAS PubMed .
  68. T. Utesch, W. Sabra, C. Prescher, J. Baur, P. Arbter and A. Zeng, Biotechnol. Bioeng., 2019, 116, 1627–1643 CrossRef CAS PubMed .
  69. Z. Li, Y. Dong, Y. Liu, X. Cen, D. Liu and Z. Chen, Metab. Eng., 2022, 70, 79–88 CrossRef CAS PubMed .
  70. H. Zong, X. Liu, W. Chen, Z. Bin and J. Sun, Biotechnol. Bioprocess Eng., 2017, 22, 549–555 CrossRef CAS .
  71. S. Lama, E. Seol and S. Park, Bioresour. Technol., 2017, 245, 1542–1550 CrossRef CAS PubMed .
  72. Z. Chen, F. Geng and A. Zeng, Biotechnol. J., 2015, 10, 284–289 CrossRef CAS PubMed .
  73. W. Zhong, Y. Zhang, W. Wu, D. Liu and Z. Chen, ACS Synth. Biol., 2019, 8, 587–595 CrossRef CAS PubMed .
  74. Y. Zhang, C. Ma, W. Dischert, P. Soucaille and A. Zeng, Biotechnol. J., 2019, 14, 1900003 CrossRef PubMed .
  75. T. Walther, C. Topham, R. Irague, C. Auriol, A. Baylac, H. Cordier, C. Dressaire, L. Lozano-Huguet, N. Tarrat, N. Martineau, M. Stodel, Y. Malbert, M. Maestracci, R. Huet, I. Andre, M. Remaud-Simeon and J. Francois, Nat. Commun., 2017, 8, 15828 CrossRef CAS PubMed .
  76. C. Frazao, D. Trichez, H. Serrano-Bataille, A. Dagkesamanskaia, C. Topham, T. Walther and J. Francois, Sci. Rep., 2019, 9, 11576 CrossRef PubMed .
  77. J. Wang, C. Li, Y. Zou and Y. Yan, Proc. Natl. Acad. Sci. U. S. A., 2020, 117, 19159–19167 CrossRef CAS PubMed .
  78. Z. Li, Z. Wu, X. Cen, Y. Liu, Y. Zhang, D. Liu and Z. Chen, ACS Synth. Biol., 2021, 10, 478–486 CrossRef CAS PubMed .
  79. B. Liu, S. Xiang, G. Zhao, B. Wang, Y. Ma, W. Liu and Y. Tao, Metab. Eng., 2019, 51, 121–130 CrossRef CAS PubMed .
  80. Y. Wang, T. Sun, X. Gao, M. Shi, L. Wu, L. Chen and W. Zhang, Metab. Eng., 2016, 34, 60–70 CrossRef PubMed .
  81. E. Lan, D. Chuang, C. Shen, A. Lee, S. Ro and J. Liao, Metab. Eng., 2015, 31, 163–170 CrossRef CAS PubMed .
  82. F. Watanabe, E. Miyamoto and Y. Nakano, J. Agric. Food Chem., 2001, 49, 5685–5688 CrossRef CAS PubMed .
  83. M. Taga and G. Walker, J. Bacteriol., 2008, 190, 1157–1159 CrossRef CAS PubMed .
  84. Y. Hirokawa, Y. Maki and T. Hanai, Metab. Eng., 2017, 39, 192–199 CrossRef CAS PubMed .
  85. Y. Wang, F. Tao, J. Ni, C. Li and P. Xu, Green Chem., 2015, 17, 3100–3110 RSC .
  86. H. Liu, J. Ni, P. Xu and F. Tao, ACS Synth. Biol., 2018, 7, 2436–2446 CrossRef CAS PubMed .
  87. S. Grady, J. Wang and E. Dekker, Biochemistry, 1981, 20, 2497–2502 CrossRef CAS PubMed .
  88. H. Meng, C. Wang, Q. Yuan, J. Ren and A. Zeng, ACS Synth. Biol., 2021, 10, 799–809 CrossRef CAS PubMed .
  89. R. Sato, T. Tanaka, H. Ohara and Y. Aso, Curr. Microbiol., 2020, 77, 3704–3710 CrossRef CAS PubMed .
  90. F. Rieckenberg, I. Ardao, R. Rujananon and A. Zeng, Eng. Life Sci., 2014, 14, 380–386 CrossRef CAS .
  91. Z. Ma, Z. Rao, L. Xu, X. Liao, H. Fang, B. Zhuge and J. Zhuge, Curr. Microbiol., 2010, 60, 191–198 CrossRef CAS PubMed .
This journal is © The Royal Society of Chemistry 2022
Click here to see how this site uses Cookies. View our privacy policy here.