Production of succinic acid by metabolically engineered microorganisms

Available online at www.sciencedirect.com

ScienceDirect Production of succinic acid by metabolically engineered microorganisms Jung Ho Ahn, Yu-Sin Jang1 and Sang Yup Lee Succinic acid (SA) has been recognized as one of the most important bio-based building block chemicals due to its numerous potential applications. For the economical biobased production of SA, extensive research works have been performed on developing microbial strains by metabolic engineering as well as fermentation and downstream processes. Here we review metabolic engineering strategies applied for bio-based production of SA using representative microorganisms, including Saccharomyces cerevisiae, Pichia kudriavzevii, Escherichia coli, Mannheimia succiniciproducens, Basfia succiniciproducens, Actinobacillus succinogenes, and Corynebacterium glutamicum. In particular, strategies employed for developing engineered strains of these microorganisms leading to the best performance indices (titer, yield, and productivity) are showcased based on the published papers as well as patents. Those processes currently under commercialization are also analyzed and future perspectives are provided. Address Metabolic and Biomolecular Engineering National Research Laboratory, Department of Chemical and Biomolecular Engineering (BK21 Plus Program), BioProcess Engineering Research Center, and Center for Systems and Synthetic Biotechnology, Institute for the BioCentury, KAIST, 291 Daehak-ro, Yuseong-gu, Daejeon 34141, Republic of Korea Corresponding author: Lee, Sang Yup ([email protected]) Present address: Institute of Agriculture & Life Science (IALS), Department of Agricultural Chemistry and Food Science, Gyeongsang National University, Jinju, Gyeongsangnam-do 52828, Republic of Korea.

1

Current Opinion in Biotechnology 2016, 42:54–66 This review comes from a themed issue on Pharmaceutical biotechnology Edited by Blaine Pfeifer and Yi Tang

http://dx.doi.org/10.1016/j.copbio.2016.02.034 0958-1669/# 2016 Elsevier Ltd. All rights reserved.

Introduction Succinic acid (SA), a four carbon dicarboxylic acid currently produced by chemical conversion of maleic anhydride [1] is an almost ubiquitous metabolite in many organisms, and thus can be produced by microbial fermentation. In recent years, our increasing concerns on climate change and other environmental problems have Current Opinion in Biotechnology 2016, 42:54–66

been urging us to move away from fossil resource-dependent chemical processes and move towards more sustainable processes for the bio-based production of chemicals and materials from renewable resources [2]. In 2004 and 2010, the U.S. Department of Energy (DOE) reported SA as one of the five most promising bio-based platform chemicals. Fermentative production of SA would provide even more environmental benefits as one mole of CO2 is fixed per one mole of SA during the fermentation. Recognizing its potential importance, extensive research has been carried out globally, which led to the development of several cost-effective processes for fermentative SA production from renewable resources. The cost for fermentative SA production is estimated to be $0.55–1.10 per kg, which is competitive to that of petrochemical process. Several plants producing SA have been established by companies such as Bioamber, Myriant, Succinity, and Reverdia [3]. SA is currently used as surfactant, ion chelator, additive in agricultural and food, and in pharmaceutical industries. The demand of SA as a platform chemical is expected to rapidly increase to an anticipated-market size of >700 000 tons per year by 2020 [3]. A much bigger market of SA is expected as a precursor for numerous industrially valuable chemicals including adipic acid (a precursor for Nylon x,6), 1,4-butanediol (1,4-BDO; a precursor for polyesters and Spandex), tetrahydrofuran (THF; an important solvent and a precursor for poly[tetramethylene ether] glycol), N-methylpyrrolidone (NMP; an important solvent in chemical and lithium-ion battery industries), 2-pyrrolidone (a precursor for pharmaceuticals and vinylpyrrolidone), gamma-butyrolactone (GBL; a precursor for pesticides, herbicides, and pharmaceuticals), and other green solvents and chemicals. Furthermore, the use of SA can be extended to the synthesis of bio-based and/or biodegradable polymers such as polyesters: for example, polybutylene succinate (PBS) and polyamides (Nylon x,4) [4]. In this paper, we review the metabolic characteristics, metabolic engineering strategies, and fermentation performance indices of the most prominent SA producing microorganisms. In particular, metabolic engineering strategies employed for developing these SA producers to reduce the formation of byproducts as well as to maximize the yield and productivity of SA are revisited (Figure 1 and Table 1). Furthermore, the advantages and disadvantages of the prominent SA producing microorganisms are summarized in Table 2. Finally, perspectives www.sciencedirect.com

Succinic acid production by engineered microorganisms Ahn, Jang and Lee 55

Figure 1

aspC, citF, poxB, tdcDE, mgsA, focA

CO2 ppc

PhosphoenolCO2 pyruvate

ppc

pykF IdhA

Pyruvate

Oxaloacetate

pdh

mdh

Malate fumABC

frdABCD

Lactate Formate

pfl

Oxaloacetate

Ethanol

Pyruvate

Succinate E. coli (anaerobic)

pfI

Lactate Formate

pdh

Malate fumABC

AcetylCoA adhE pta

Fumarate

ackA

Acetate

frdABCD

Phosphoenolpyruvate

ppc

IdhA

ackA

Acetate

CO2

pykF

pycR.etli

mdh

AcetylCoA adhE pta

Fumarate

fruA

ptsG

Phosphoenolpyruvate

EtOH

CO2

pykF IdhA

pckA

Oxaloacetate

Pyruvate

mdh

maeB

Malate fumC

MQH2

Lactate Formate

AcetylCoA adhE pta

Fumarate frdABCD

pfIB

EtOH

ackA

Acetate

MQ Succinate

Succinate E. coli (dual-phase)

M. succiniciproducens (anaerobic) Current Opinion in Biotechnology

Metabolic pathways of E. coli [21,52] and M. succiniciproducens [32] and their best metabolic engineering strategies for the enhanced production of SA. Genes knocked out for the enhanced production of SA are marked with ‘x’. For dual phase fermentation of E. coli, cells were first grown aerobically to a high concentration before SA production under anaerobic condition in the second phase. The pyc gene overexpressed in E. coli for dual phase fermentation is shown in bold arrow. The engineered E. coli strain KJ122 (DldhA, DadhE, DackA, DfocA-pflB, DmgsA, DpoxB, DtdcDE, DcitF, DaspC, DsfcA) shown in the left produced 88 g/L of SA with the yield and productivity of 1.29 mol/mol glucose and 0.73 g/L/h, respectively, by anaerobic fed-batch fermentation. The engineered E. coli strain AFP111 (Dpfl, DldhA, DptsG) overexpressing the Rhizobium etli pyc gene shown in the center produced 99.2 g/L of SA with the yield and productivity of 1.74 mol/mol glucose and 1.3 g/L/h, respectively; aerobic cell propagation stage was not taken into account for the calculation of SA yield and productivity. The engineered M. succiniciproducens PALFK (DldhA, Dpta-ackA, DfruA) shown in the right produced 78.41 g/L of SA with the yield and productivity of 1.64 mol/mol and 6.03 g/L/h, respectively, from sucrose and glycerol by anaerobic fed-batch fermentation. Genes shown are: focA, formate transporter; mgsA, methylglyoxal synthase; poxB, pyruvate dehydrogenase; tdcDE, propionate kinase/acetate kinase; citF, citrate lyase; aspC, aspartate aminotransferase; pfl, pyruvate formate lyase; ldhA, lactate dehydrogenase; ptsG, glucose-specific PTS enzyme; pyc; pyruvate carboxylase; pta, phosphate acetyltransferase; ackA, acetate kinase; fruA, fructose specific PTS system.

on further performance improvement and industrial-scale bio-based production of SA from renewable resources are discussed.

Production of succinic acid SA, an intermediate of the tricarboxylic acid (TCA) cycle and one of the end products of anaerobic metabolism, is synthesized in almost all microbe, plants, and animal cells. Among these organisms, bacteria and fungi have been recognized as suitable hosts for the efficient production of SA. Much effort has been exerted to develop processes for the bio-based production of SA using several fungal/yeast strains such as Aspergillus niger, Aspergillus fumigatus, Byssochlamys nivea, Candida tropicalis, Lentinus degener, Paecilomyces varioti, Penicillium viniferum, Saccharomyces cerevisiae, and Pichia kudriavzevii (Issatchenkia orientalis). In the case of bacteria, the following strains have been employed for SA production: Actinobacillus succinogenes, Anaerobiospirillum succiniciproducens, Corynebacterium glutamicum, Escherichia coli, Mannheimia succiniciproducens, and Basfia succiniciproducens (very similar strain to M. succiniciproducens). Among these, this paper reviews the metabolic engineering strategies employed www.sciencedirect.com

for developing succinic acid producers based on S. cerevisiae, P. kudriavzevii, E. coli, M. succiniciproducens, B. succiniciproducens, A. succinogenes, and C. glutamicum, as representative examples of those currently under commercialization. In comparing the results, those that employed dual phase fermentation (e.g., cell growth phase followed by SA production phase in the cases of E. coli and C. glutamicum) are indicated so that the true performance indices can be better understood. Saccharomyces cerevisiae

S. cerevisiae is a well characterized eukaryotic microorganism that has been most widely used for industrial bioethanol production. Also, it has been employed as a platform strain for the production of various chemicals thanks to the availability of numerous genetic, metabolic engineering, and omics tools. S. cerevisiae does not normally produce SA as a fermentation end product. However, unlike bacterial succinic acid producers that prefer to grow at neutral pH, S. cerevisiae can grow within a wide pH range of 3–6, which offers a great advantage for SA production. The ability to grow at low pH reduces the need for neutralization to produce SA, and thus, generation of salts (e.g., gypsum) can Current Opinion in Biotechnology 2016, 42:54–66

56 Pharmaceutical biotechnology

Table 1 Summary of bio-based SA production reported in selected papers and patents Strain

Description

Fermentation type

Saccharomyces cerevisiae (Reverdia) CEN.PK113-7D Wild type strain  N/A a SUC-200  Batch, 30 8C, pH 3, aerobic RWB064; +pckA from A. succinogenes, +fumR from R. (OUR = 5 mmol/L/h, 10% CO2), oryzae, +mdh3 from S. glucose, KOH cerevisiae, +mae1 from S. pombe SUC-194 CEN.PK113-7D; +gsh1, +cys3,  Fed-batch, 30 8C, dual phase +glr1, +mdh2 from S. cerevisiae, (1. aerobic, pH 5, KOH, +pckA from M. OUR = 30 mmol/h, 10% CO2; succinciproducens, +frdm1 from 2. anaerobic, pH release, 10% T. brucei, +fumR from R. oryzae, CO2, 90% N2) glucose +mae1 from S. pombe SUC-297 CEN.PK113-7D; Dadh1, Dadh2,  Fed-batch, 30, aerobic, dual Dgpd1, +pckA from M. phase (1. pH 5; NH3, pO2 = 20%; succinciproducens, +gsh1, 2. pH 3; KOH, 50% CO2), glucose +cys3, +glr1, +mdh3, and +pyc2p from S. cerevisiae, +fumR from R. oryzae, +frdm1 from T. brucei, +mae1 from S. pombe Pichia kudriavzevii (Issatchenkia orientalis, Bioamber) CD1822  N/A Evolved strain from the P. kudriavzevii ATCC PTA-6658, can grow in presence of 150 g/L SA at pH 2.5  Batch, 30 8C, pH 3, aerobic 13171 CD1822; Dcyb2a, +pyc1 and +fum1 from C. krusei, +frd1 from (DO < 5%, OUR = 10 mmol/L/h, S. cerevisiae, +mdh from Z. 10% CO2), glucose, KOH rouxii, +frd1 from T. brucei Evolved strain from the P. 13723  Batch, 30 8C, pH 3, aerobic kudriavzevii ATCC PTA-6658 (DO < 10%, OUR = 18 mmol/L/ (high glucose consumption and h, 10% CO2), glucose, KOH growth rates); Dura, Dpdc, +pyc1 and +fum from C. krusei, +mae from S. pombe, +frd from L. mexicana, +mdh from R. delemar Escherichia coli (Myriant) KJ122 DaspC, DcitF, DtdcDE, DsfcA,  Batch, 37 8C, pH 7, anaerobic, D( focA-pflB), DldhA, DadhE, glycerol, K2CO3 + KOH DmgsA, DackA, DpoxB TG400 KJ122; 4galP (G297D)  Batch, 37 8C, pH 7, anaerobic, glucose + xylose, K2CO3 + KOH KJ122; 4glpK (A55 T, R157H),  Batch, 39 8C, pH 7, aerobic, MH28 4glpR glycerol, K2CO3 Mannheimia succiniciproducens (MBEL, KAIST) MBEL55E Wild type  Batch, 39 8C, pH 6.5, anaerobic, glucose, NH4OH LPK7 DldhA, Dpfl,  Fed-batch, 39 8C, pH 6.5, 4pta-ackA anaerobic, glucose, NH4OH PALK DldhA, Dpta-ackA  Fed-batch, 39 8C, pH 6.5, anaerobic, glucose, NH4OH PALFK PALK; DfruA  Fed-batch, 39 8C, pH 6.5, anaerobic, sucrose + glycerol, NH4OH  Fed-batch/initialOD600 of 9.03, 39 8C, pH 6.5, anaerobic, sucrose + glycerol, NH4OH  MCRB, 39 8C, pH 6.5, anaerobic, sucrose + glycerol, NH4OH PALK; +glpK from E. coli  Fed-batch, 39 8C, pH 6.5, PALKG anaerobic, sucrose + glycerol, NH4OH

Current Opinion in Biotechnology 2016, 42:54–66

Titer (g/L)

Yield (mol/mol)

Yield (g/g)

Productivity (g/L/h)

References

N/A 16.00

N/A N/A

N/A N/A

N/A 0.18

[53] [54]

19.50

N/A

N/A

0.22

[55]

43.00

N/A

N/A

0.45

[50]

N/A

N/A

N/A

N/A

[16]

23.00

N/A

N/A

0.26

[16]

48.20

0.69

0.45

0.97

[17]

88.00

1.29

0.85

0.73

(Figure 1; [56])

96.00

1.46

0.96

0.80

[56]

84.30

1.53

1.00

1.76

[49]

10.50

0.69

0.45

1.17

[29]

52.43

1.16

0.76

1.80

[29]

45.79

1.32

0.86

2.36

[31]

68.41

1.57

1.03

2.46

[32]

78.41

1.64

1.07

6.03

(Figure 1; [32])

29.00

1.54

1.01

29.73

[32]

64.67

1.39

0.91

N/A

[32]

www.sciencedirect.com

Succinic acid production by engineered microorganisms Ahn, Jang and Lee 57

Table 1 (Continued ) Description

Strain

Basfia succiniciproducens (Succinity) DD1 Wild type LU15224

DD1; Dpfl, Dldh

Fermentation type

 Batch, 37 8C, pH 6.5, anaerobic, glucose  Batch, 37 8C, pH 6.5, anaerobic, glycerol, NH4OH  Batch, 37 8C, pH 6.5, anaerobic, glycerol + maltose, NH4OH  Batch, 37 8C, pH 6.5, anaerobic, glucose, NH4OH

LU15224/ LU15224; +glyoxylate shunt operon, +malate synthase from pJFF224 (icl ms Y.m.) Y. molaretii Actinobacillus succinogenes (Michigan Biotechnology Institute; MBI) 130Z Wild type strain  Batch, 39 8C, pH 6.5, anaerobic, glucose, MgCO3 Corynebacterium glutamicum (Ajinomoto) C. glutamicum Dldh, +pyruvate carboxylase  Batch, 31.5 8C, pH 7.6, from Brevibacterium flavum, anaerobic (cell proliferation under +sdh from C. glutamicum aerobic condition), (NH4)2CO3 a

Titer (g/L)

Yield (mol/mol)

Yield (g/g)

Productivity (g/L/h)

References

20.00

0.75

0.49

N/A

[35,57]

36.20

1.90

1.26

1.51

[58]

69.80

1.70

1.11

2.91

[51]

46.30

1.33

0.87

N/A

[59]

67.20

N/A

N/A

0.80

[40]

N/A

N/A

N/A

N/A

[60]

N/A, not available.

be avoided in downstream process. Unlike most of the SAproducing microorganisms (see below), which produce SA through the reductive TCA cycle under anaerobic or micro-aerobic condition, S. cerevisiae possesses rather weak reductive TCA cycle. Production of SA through the reductive TCA cycle is thermodynamically unfavorable and the enzymes corresponding to the reductive pathway are subjected to glucose repression in S. cerevisiae. Thus, SA production under aerobic condition has been suggested to be more advantageous for S. cerevisiae. To develop a S. cerevisiae mutant strain for the enhanced production of SA under aerobic condition, several genes in the TCA cycle were disrupted. A series of studies were performed on examining the effect of disrupting S. cerevisiae succinate dehydrogenase, which is composed of four subunits encoded by the sdh1, sdh2, sdh3, and sdh4 genes, on SA production. Also, the effect of disrupting the sdh1b gene, which is a homologue of the sdh1 gene that suppresses the sdh1 deletion phenotype in S. cerevisiae, was examined. Double disruption of the sdh1 and sdh1b genes causing complete loss of succinate dehydrogenase activity allowed 1.9-fold increase in SA titer accompanied with reduced malate production [5]. On the other hand, the sdh1 and sdh2-deleted S149 strain was found to produce rather high level of ethanol. S. cerevisiae possess six NADH-dependent alcohol dehydrogenase isozymes encoded by the adh1, adh2, adh3, adh4, adh5, and sfa1 genes. In order to reduce ethanol production, five alcohol dehydrogenase genes (adh1 to adh5) were deleted in the S149 strain to make the S149sdh12 strain. Compared to the control strain, the S149sdh12 strain showed 75% decrease in ethanol yield and 20-fold increase in SA yield (0.217 mol/mol glucose) [6]. Deletion of all alcohol dehydrogenase isozyme genes stably prevented ethanol production [7]. www.sciencedirect.com

The above results demonstrated that disruption of ethanol production pathway in succinate dehydrogenase mutant is effective for improving SA production in S. cerevisiae. However, cell growth was reduced upon disruption of alcohol dehydrogenases because cellular growth is strongly dependent on ethanol production. In addition, reduction in ethanol production was accompanied by increased glycerol production because of cell’s need to oxidize NADH; the ethanol production pathway in S. cerevisiae is responsible for oxidizing NADH generated through glycolysis, for the maintenance of intracellular balance of NADH and NAD+ [8]. Furthermore, cellular metabolite profiling based on metabolomics showed intracellular accumulation of SA in the S149sdh12 strain. S. cerevisiae does not naturally carry a SA transporter. Hence, malic acid transporter (mae1) from Schizosaccharomyces pombe, known to increase SA transport, was introduced to the S149sdh12 strain to improve SA production. As a result, the SA yield could be increased to 0.236 mol/mol glucose [6]. Instead of using the TCA cycle for SA production under aerobic condition, the use of glyoxylate shunt that does not involve oxidative decarboxylation can potentially offer higher yield. S. cerevisiae possesses three isocitrate dehydrogenase isozymes that vary in subunit structure, subcellular location, and cofactor specificity. Among them, two mitochondrial isozymes, IDH1 and IDP1, that are NADspecific and NADP-specific, respectively, can contribute to a-ketoglutarate formation. a-Ketoglutarate is used to synthesize glutamate, essential for cell growth. Thus, double knockout of the idh1 and idp1 genes results in glutamate auxotrophy, which can be monitored as a sign of blockage from isocitrate to a-ketoglutarate. A mutant strain, AH22ura3, was developed by the deletion of the sdh1, sdh2, idh1, and idp1 genes. Shake flask culture of this strain Current Opinion in Biotechnology 2016, 42:54–66

58 Pharmaceutical biotechnology

Table 2 Advantages and disadvantages of different SA producers and future research direction SA producer S. cerevisiae and P. kudriavzevii

C. glutamicum and E. coli

M. succiniciproducens and B. succiniciproducens

Other rumen bacteria (A. succinogenes)

Advantages

Disadvantages

 Well-established industrial microorganism (physiology, genetic tools)  Low pH fermentation (pH 3–5) allows simple separation/ purification process  Good tolerance to SA  Well-established industrial microorganism (physiology, genetic tools)  Engineered strain can produce high amount of SA (titer, yield, productivity)  SA is naturally formed as an endproduct  Simple anaerobic fermentation  Genome sequenced and genetic engineering tools available  Excellent CO2-fixing anaploretic pathway  Engineered strain can produce high amount of SA (titer, yield, productivity) with almost no byproduct formation  High amount of SA is naturally formed as an end-product  Simple anaerobic fermentation

 Does not naturally produce SA as an end product  Low SA producing ability (titer, yield, productivity)

 Prevent/reduce byproducts formation  Cheap substrates utilization  Optimization of redox balance  Enhance SA titer, yield, productivity

 Does not naturally produce SA as an end product  Byproducts formation  pH sensitive  Aerobic/anaerobic shift often required for optimum SA production  Harbor several auxotrophies  pH sensitive

 Broad/cheap substrates utilization  Fermentation at low pH to decrease downstream costs  Improve fermentation process

 Harbor several auxotrophies  Lack of useful genetic tools  pH sensitive

on glucose as a carbon source resulted in production of 3.62 g/L of SA with a yield of 0.11 mol/mol glucose [9]. However, as mentioned earlier, external glutamate supplementation was required to overcome auxotrophy. Genome-scale metabolic simulation coupled with evolutionary programming method predicted three knockout target genes, sdh3, ser3p, and ser33, to link cell growth with SA production [10]. This strategy includes knockout of sdh3 that encodes cytochrome b subunit of succinate dehydrogenase to remove the primary SA consuming reaction and deletion of ser3p and ser33p encoding 3-phosphoglycerate dehydrogenase isoenzyme to disable serine biosynthesis. Serine required for cell growth, can be synthesized from glycine, which is coupled to SA production through the glyoxylate pathway. Using this strategy, a SA yield of 0.03 mol/mol glucose was achieved under aerobic condition with glycine supplementation. Directed evolution was employed to obtain a prototrophic strain that does not require glycine supplementation. Finally, the icl1 gene encoding native isocitrate lyase 1 was overexpressed in the evolved strain to achieve increased SA yield of 0.07 mol/ mol glucose without glycine supplementation. SA production through the reductive TCA pathway offers a theoretical maximum yield of 1.7 mol/mol glucose, which Current Opinion in Biotechnology 2016, 42:54–66

Future research direction

 Overcome auxotrophies  Fermentation at low pH to decrease downstream costs  Increase SA tolerance

 Develop genetic tools for strain development  Overcome auxotrophies  Fermentation at low pH to decrease downstream costs  Increase SA tolerance

is superior to those using oxidative TCA cycle or glyoxylate shunt [11]. However, there are difficulties of employing the reductive TCA pathway in S. cerevisiae. Fumarate hydratase exhibits an irreversible conversion of fumarate to malate, and the frds1 and osm1genes encoding fumarate reductase are only expressed under anaerobic condition [12]. In addition, the reductive pathway consumes 2 mol of NADH to produce 1 mol of SA, while less is required for malate and fumarate. Hence, it is rather difficult to produce SA compared with malate or fumarate due to the limited supply of NADH in the cytosol. Metabolic engineering was performed on a pyruvate decarboxylase-deficient S. cerevisiae strain to overcome these difficulties and produce SA using the reductive TCA pathway. To reduce the flux towards malate from fumarate, fumarate hydratase ( fum1) was deleted and the E. coli fumarate hydratase ( fumC) was overexpressed. In addition, pyruvate carboxylation was enhanced by overexpressing the pyc2 gene encoding pyruvate carboxykinase 2. The mdh3 and frd1 genes encoding malate dehydrogenase and fumarate reductase, respectively, were overexpressed as well. Also, glycerol-3-phosphate dehydrogenase that controls glycerol synthesis under aerobic condition was deleted. This strain cultured in a medium supplemented with CaCO3, urea, and biotin produced 13 g/L of SA with a yield of 0.21 mol/mol glucose at pH 3.8 [13]. Genome-scale flux balance analysis was www.sciencedirect.com

Succinic acid production by engineered microorganisms Ahn, Jang and Lee 59

performed and identified an inner dicarboxylate mitochondrial transporter as a deletion target. Indeed, deletion of decarboxylate mitochondrial transporter led to the improvement of SA yield to 0.03 mol/mol glucose [14]. Again, the ability to grow and produce SA at low pH is a great economical advantage of yeast strains over other SA producing microorganisms. Furthermore, numerous studies on producing SA from cheap lignocellulosic feedstock are in progress [15]. One weak point of yeast as a SA producer is that the SA titer, yield, and productivity of engineered S. cerevisiae are relatively lower than those obtainable with other SA producers. Reverdia is employing engineered S. cerevisiae strain for the production of SA and has been operating a 10 000 ton capacity plant since 2011 (Figure 3). Performance indices reported in the patents by Reverdia are summarized in Table 1.

Pichia kudriavzevii

Some Pichia and Candida species exhibit good tolerance to succinic acid. Five succinic acid tolerant Pichia and Candida strains, P. kudriavzevii strains ATCC PTA-6658, ATCC 60585, and ATCC 24210, C. lambica ATCC 38617, and C. sorboxylosa ATCC 24120, have been identified by culturing cells in a medium containing 150 g/L of succinic acid at pH 2.5–2.8 [16]. Further evolution of P. kudriavzevii ATCC PTA-6658 strain was performed under the glucose limited chemostat condition, to isolate the CD1822 strain [16]. The CD1822 strain was metabolically engineered by the deletion of cyb2a gene that encodes lactate cytochrome C oxidoreductase, while the following genes were heterologously introduced: the Candida krusei pyc1 and fum1, S. cerevisiae frd1, Zygosaccharomyces rouxii mdh, and Trypanosoma brucei frd1. The resulting engineered P. kudriavzevii 13171 strain produced 23 g/L of succinic acid from glucose in a batch fermentation at pH 3.0 [16]. In another evolutionary engineering study, the P. kudriavzevii 11-1 strain having higher glucose consumption and high growth rates was isolated from the ATCC PTA-6658 strain. The P. kudriavzevii 11-1 strain was engineered by knocking out the ura and pdc genes encoding orotidine 50 -phosphate decarboxylase and pyruvate decarboxylase, respectively, while the C. krusei pyc1 and fum, S. pombe mae, Leishmania mexicana frd, and Rhizopus delemar mdh genes were heterologously introduced. The resulting engineered P. kudriavzevii 13723 strain produced 48.2 g/L of SA from glucose in a batch fermentation at pH 3.0 [17]. Bioamber is employing engineered P. kudriavzevii for the production of SA and has been operating a 30 000 ton capacity plant since 2014 (Figure 3). Performance indices reported in the patents by Bioamber are summarized in Table 1. www.sciencedirect.com

Escherichia coli

The wild-type E. coli strains produce SA only as a minor product under anaerobic condition. Many researchers have worked on developing engineered E. coli strains capable of efficiently producing SA by taking advantages of its fast growth and availability of well-established methods for metabolic engineering. For this, several strategies including rational metabolic engineering, random mutagenesis, and evolutionary engineering have been employed. The wild-type E. coli, which produces SA at a theoretical yield of 1 mol/mol glucose under anaerobic condition was engineered by deleting enzymes that compete with SA formation. The most notable engineered strain NZN111 is constructed by knocking out pyruvate formate lyase ( pfl) and lactate dehydrogenase (ldhA) genes to prevent production of formic and lactic acids. Unfortunately, the mutant strain could not grow anaerobically on glucose and pyruvate accumulation was observed [18]; inactivation of NADH-dependent lactate dehydrogenase limits the regeneration of NAD+ and consequently disables proper growth under anaerobic condition. This problem could be overcome by overexpressing E. coli malate dehydrogenase (mdh) in NZN111. Using this strain, 31.9 g/L of SA was produced with a yield of 1.19 mol/mol glucose by fedbatch fermentation [19]. Another approach was taken to restore cell growth on glucose and enhance anaerobic SA production by NZN111 strain through spontaneous chromosomal mutation of glucose phosphotransferase ( ptsG). The ptsG mutant strain, AFP111, with suppressed normal glucose repression successfully grew on glucose and showed increase in SA production with the yield and productivity of 1 mol/mol glucose and 0.87 g/L/h, respectively [20]. Then, the AFP111 strain was further engineered by introducing the Rhizobium etli pyc gene encoding pyruvate carboxylase. Dual-phase fermentation of this engineered strain was performed by achieving high cell density in the first aerobic growth phase and then SA production in the second anaerobic production phase. When the optimal transition point from aerobic to anaerobic phase was applied, the AFP111 strain expressing the pyc gene produced 99.2 g/L of SA with the yield and productivity of 1.74 mol/mol glucose and 1.3 g/L/h, respectively [21]; here, the amount of glucose consumed and time required for aerobic growth were not taken into account in calculating SA yield and productivity. Similarly, the approach of evolution and selection was taken to overcome poor growth and glucose fermentation of the ldhA, adhE, and ackA deleted mutant E. coli strain. An isolate with recovered cell growth in anaerobic condition was obtained and pyruvate formate-lyase ( pflB) and formate transporter ( focA) were eliminated to reduce the loss of flux in the forms of formate and acetyl-CoA (Figure 1). As a result, this strain could not grow without acetate. Another round of natural selection was performed to Current Opinion in Biotechnology 2016, 42:54–66

60 Pharmaceutical biotechnology

overcome acetate auxotrophy and an evolved KJ060 strain was obtained. Anaerobic batch fermentation of KJ060 produced 87 g/L of SA with the yield and productivity of 1.41 mol/mol glucose and 0.72 g/L/h, respectively [22]. As an alternative strategy, SA can be produced using the glyoxylate shunt pathway, which requires lower reducing equivalent compared to the anaerobic route. The SBS550MG strain that utilizes dual phase SA production scheme was developed by the deletion of adhE, ldhA, ptaackA, and iclR genes. The glyoxylate shunt was activated by the inactivation of iclR, which encodes transcriptional repressor for glyoxylate shunt. Using this strain, 40 g/L of SA was produced in an anaerobic fermentation with pulse feeding of glucose. Further engineering was conducted on the SBS550MG by co-expressing the Lactococcus lactis pyc and Candida boidinii fdh1 genes to make the SBS550MG-Cms243 strain. The NAD+-dependent formate dehydrogenase converts 1 mol of formate into 1 mol of NADH and CO2, resulting in increase in NADH availability and 6% increase in SA yield compared to the parent SBS550MG strain [23]. Aerobic production of SA offers advantages in terms of rapid cell growth and higher carbon source uptake rate, allowing higher overall SA productivity. To construct glyoxylate cycle for aerobic SA production, the glucose transferase system ( ptsG), succinate dehydrogenase (sdhAB), iclR, and byproduct formation pathways ( poxB, pta-ackA) were deactivated, while the Sorghum vulgare ppc gene was overexpressed. As a result, this mutant E. coli strain produced 58.3 g/L SA in 59 h under aerobic condition with the average yield and productivity of 0.94 mol/ mol glucose and 1.08 g/L/h, respectively [24]. In another study, comparative analysis of the genomes of E. coli and a high-level succinic acid producer M. succiniciproducens identified that succinate dehydrogenase (sdhABCD), malate dehydrogenase (mqo), isocitrate lyase (aceA), malate synthase (aceB), pyruvate kinase ( pykF), and glucosespecific PTS enzyme ( ptsG) were only present in E. coli, while pykA was present in both strains [25]. Sequential disruption of these genes only present in E. coli, however, did not enhance SA production in E. coli. Comparative in silico genome-scale metabolic simulations suggested ptsG, pykF, and pykA triple knockout mutant strain (later experimentally constructed as W3110GFA strain) for enhanced SA production. The W3110GFA strain showed 3.4-fold increase in SA production compared with the wild-type strain by anaerobic fermentation with significant reduction in byproducts [25]. As shown in Figure 3, Myriant is employing engineered E. coli strain for the production of SA and has been operating a 15 000 ton capacity plant since 2013 [26]. Performance indices reported in the patents by Myriant are summarized in Table 1. Current Opinion in Biotechnology 2016, 42:54–66

Mannheimia succiniciproducens and Basfia succiniciproducens

M. succiniciproducens MBEL55E, a capnophilic bacterium, was first isolated from the bovine rumen in the late 1990s. This rumen bacterium is a facultatively anaerobic, nonspore forming, mesophilic, Gram negative bacterium that efficiently fixes CO2 and produce SA as major fermentation product [27]. Menaquinone is employed for anaerobic respiration by M. succiniciproducens and fumarate functions as a final electron acceptor. In addition, M. succiniciproduens carries anaerobic fumarate hydratase, fumarate reductase, and NADH dehydrogenase, demonstrating its strong oxygen-independent metabolic characteristics. M. succiniciproducens can utilize a wide variety of carbon sources such as glucose (starch, cellulose hydrolysates), fructose (sugarcane), sucrose (sugar cane, beet), maltose (starch), lactose (whey), mannitol (starch), and xylose (hemicellulose hydrolysate) [28]. Also, it was later found that glycerol (byproduct of biodiesel industry) can also be used. As represented by its small genome size, this bacterium is auxotrophic for several amino acids and vitamins. Genome-scale metabolic simulations followed by actual medium formulation studies were performed to develop a chemically defined medium that contains two amino acids and four vitamins. Availability of the whole genome sequence of M. succiniciproducens allowed much better understanding of the metabolic characteristics and consequently genome-wide metabolic engineering towards enhanced SA production [28]. The entire cellular, membrane, and secreted proteins were analyzed and mapped to identify target proteins that have potential to improve the performance of microorganism by metabolic engineering. The in silico flux analysis identified that PEP carboxylation by PEP carboxykinase, rather than PEP carboxylase, is the key step for CO2-fixing, and have direct relationship with SA flux in the reductive TCA branch. To verify that PEP carboxykinase plays the most important role in C3 to C4 conversion, genes corresponding to malic enzyme (maeB), PEP carboxylase ( ppc), and PEP carboxykinase ( pckA) were individually disrupted and cell growth profiles were examined. Agreeing with the results from in silico flux analysis, the strain with pckA inactivation showed severe growth retardation, while strains inactivated with ppc or maeB did not show much effect on cell growth [29]. In addition, major enzymes that take away metabolic flux from SA formation were identified: lactate dehydrogenase, pyruvate formate lyase, phosphotransacetylase, and acetate kinase. Since there had been no genetic engineering tools for M. succinciproducens, an effective and robust gene deletion and overexpression systems had to be developed first. The wild-type M. succiniciproducens produces 10.5 g/L of SA, 4.96 g/L of acetic acid, 4.1 g/L of formic acid, and 3.47 g/L of lactic acid under anaerobic batch fermentation www.sciencedirect.com

Succinic acid production by engineered microorganisms Ahn, Jang and Lee 61

using glucose as a carbon source [29]. For improved SA production with reduced byproducts formation, genes encoding lactate dehydrogenase (ldhA), pyruvate formate lyase ( pflB), phosphate acetyltransferase ( pta), and acetate kinase (ackA) were sequentially deleted to make the mutant LPK7 strain. Fed-batch fermentation of LPK7 resulted in production of 52.4 g/L of SA with a yield and productivity of 1.16 mol/mol of glucose and 1.8 g/L/h, respectively. The formation of byproducts was greatly reduced to 0.81 g/L of acetic acid, 0.25 g/L of lactic acid, and no formic acid. However, cell growth was somewhat retarded and excretion of pyruvate was observed. Possibility of carrying out continuous fermentation of LPK7 was also examined [30]. To develop a redox-balanced, growth maximized, and enhanced homo-SA-producing strain, a systematic approach combining metabolic engineering and in silico genome-scale metabolic simulation was taken. The best in silico simulation result suggested a new engineered strain, PALK, in which the ldhA and pta-ackA genes were deleted [31]. Fed-batch fermentation of the PALK strain resulted in the production of 45.8 g/L of SA with the yield and productivity of 1.32 mol/mol of glucose and 2.36 g/L/h, respectively [31]. The byproducts formed during the course of fermentation was reduced to 0.45 g/L of acetic acid, 0.24 g/L of pyruvic acid, and no formic and lactic acids [31]. A notable aspect of the PALK strain is that even without the inactivation of the pfl gene, formic acid is not excreted. Formic acid seems to be utilized for the generation of redox equivalents by formate dehydrogenase complex. In addition, the maximum specific growth rate of the PALK was significantly increased to 0.69 h1, which is much higher than that (0.3 h1) of the LPK7 strain [31]. Glycerol is an attractive carbon source for the production of reduced chemicals such as SA (Figure 1). Glycerol is an inexpensive and abundant byproduct of biodiesel industry, and a relatively more reduced carbon source. In particular, if equal number of carbon is provided, twice the amount of reducing equivalents (NADH) are produced during the production of PEP using glycerol as a carbon source compared to other carbohydrates such as glucose. Studies on the mechanism of glycerol uptake in the presence of sucrose were conducted to construct an efficient sucrose utilizing M. succinciproducens strain for SA production, based on the identified sucrose phosphotransferase system of M. succiniciproducens. Furthermore, the PALFK strain, which is fructose phosphotransferase ( fruA) deleted PALK strain, and the PALKG strain that is the PALK strain expressing the E. coli glycerol kinase encoding gene (glpK) were developed (Figure 1). The PALFK strain produced 68.4 g/L of SA with a yield of 1.57 mol/mol and the PALKG produced 64.7 g/L of SA with a yield of 1.39 mol/mol in fed-batch fermentations using sucrose and glycerol [32]. It is notable that the ratio of total amount of byproducts to SA was only 0.02, using www.sciencedirect.com

the PALFK, approaching true homo-SA production. When the initial seed culture OD600 was slightly increased to 9.03, 78.41 g/L of SA was produced with the yield and productivity of 1.64 mol/mol and 6.03 g/L/h, respectively, from sucrose and glycerol by anaerobic fedbatch fermentation (Figure 1). Different fermentation approaches, such as high cell density fed-batch fermentation and membrane cell recycling bioreactor (MCRB) system, were taken to further enhance SA production using the PALFK strain [32]. Using such fermentation system, SA could be produced at a high productivity of 29.7 g/L/h with a yield of 1.54 mol/ mol from sucrose and glycerol [32]. To date, the engineered M. succinciproducens strains are capable of producing homo-SA with highest productivity and yield, and thus at the lowest SA production cost. Although the unpublished results cannot be reported here, the performance indices we reported in the patents by KAIST are summarized in Table 1. B. succiniciproducens, a member of Pasteurellaceae, is a Gram negative, facultative, anaerobic bacterium used by Succinity, a joint venture between BASF and Corbion Purac, for the industrial-scale production of SA [3]. The wild-type B. succiniciproducens DD1 isolated from the rumen of German cow in 2008 shows high similarity to M. succinciproducens MBEL55E, which was isolated from the rumen of Korean cow and first reported in 2002 [27,33]. The genome sizes of the MBEL55E and the DD1 strains are similar at 2 314 078 bp and 2 340 000 bp, respectively, and have almost the same GC content of 42.5 mol%. In the 2380 ORFs of MBEL55E and 2363 ORFs of DD1, 2006 ORFs were homologous, and in overall an average of 95% homology was observed [34]. A 16S rRNA analysis based phylogenetic tree shows that MBEL55E and DD1 strains are almost identical strains (Figure 2). It should be noted that MBEL55E was classified as Mannheimia when published in 2002 based on the 16S rRNA sequences available at that time. On the other hand, DD1 strain was named as Basfia in 2008 when more complete genome sequences became available. Systems-wide 13C metabolic flux analysis was performed to elucidate intracellular fluxes in B. succiniciproducens [35]. Based on the analysis, metabolic engineering of B. succiniciproducens was performed by deleting pflD and ldhA, which eliminated formic acid production, drastically reduced lactic acid production, and made pyruvic acid a major byproduct. This engineered strain showed improved SA yield of 1.08 mol/mol glucose compared to the wild type DD1 strain (the SA yield of 0.75 mol/mol glucose). These results are similar to those reported for M. succinciproducens LPK strain, which is also a lactate dehydrogenase and pyruvate formate lyase inactivated strain, showing a SA yield of 0.97 mol/mol glucose with pyruvic acid as a dominating byproduct [29,36]. It will be interesting to see if Current Opinion in Biotechnology 2016, 42:54–66

62 Pharmaceutical biotechnology

Figure 2

(a)

0.00

Mannheimia succiniciproducens MBEL55E (NR074906; 22-JAN-2003)

0.02 0.00

0.05

Basfia succiniciproducens DD1 (FJ463880; 13-NOV-2008)

0.03

Actinobacillus succinogenes 130Z (NR024860, 12-SEP-1997) Anaerobiospirillum succiniciproducens (U96412; 03-APR-1997)

0.08 0.00 0.01

(b)

0.01

0.00

0.01

0.02

0.02 0.01

0.01

0.01 0.02

0.01

Mannheimia glucosida P925 (AF053889) Mannheimia haemolytica NCTC 9380 (NR 11448) Mannheimia ruminalis HPA92 (AF053900) Mannheimia granulomatis ATCC 49244 (AF053902) Mannheimia varigena 177 (AF053893) Actinobacillus capsulatus CCUG 12396 (M75069)

0.00 0.01

0.01

0.02

0.01 0.02

0.00

0.02

0.03 0.02

0.00 0.01 0.02

0.00

0.04

0.11

Pasteurella dagmatis ATCC 43325 (M75051) Pasteurella canis CCUG 12400 (NR 042882) Avibacterium volantium NCTC 3438 (NR 025937) Avibacterium avium NCTC 11297 (NR 044750) Actinobacillus succinogenes 130Z (NR 024860) Mannheimia succiniciproducens MBEL55E (NR 074906) Basfia succiniciproducens DD1 (FJ 463880) Pasteurella testudinis CCUG 19802 (AY362926) Anaerobiospirillum succiniciproducens (U96412) Current Opinion in Biotechnology

Phylogenetic analysis of SA producers. (a) 16S rRNA sequence (NCBI database)-based phylogenetic tree of four natural SA high producers. (b) 16S rRNA sequence-based phylogenetic tree of microorganisms including four natural SA high producers (bold).

restoration of pyruvate formate lyase while knocking out the pta-ackA genes in B. succiniciproducens will also lead to enhanced SA production as reported for M. succiniciproducens [31]. Succinity is employing the engineered B. succiniciproducens strain for the production of SA and has been operating a 10 000 ton capacity plant since 2013 (Figure 3). Performance indices reported in the patents by Succinity are summarized in Table 1. Actinobacillus succinogenes

A. succinogenes 130Z belongs to the family of Pasteurellaceae based on the 16S rRNA sequence analysis, and is a capnophilic, Gram negative, and facultative anaerobic rumen bacterium. A. succinogenes can utilize a wide range of carbon sources, including arabinose, fructose, glucose, glycerol, sucrose, and lactose, and produces high concentration of SA. A. succinogenes has moderate tolerance to high concentration of glucose compared to E. coli or A. succiniciproducens. The supplementation of hydrogen, an electron donor, and electrically reduced neutral red significantly enhanced the production of SA; consistent results were observed when reduced sugars such as mannitol were used compared to glucose [37]. Current Opinion in Biotechnology 2016, 42:54–66

A. succinogenes 130Z strain can produce 66.4 g/L of SA with the yield and productivity of 1.02 mol/mol glucose and 0.79 g/L/h, respectively, in aerobic batch fermentation [38]. Continuous and repeat-batch biofilm fermentation of the 130Z strain resulted in significant increase of SA productivity to 8.8 g/L/h [39]. Since A. succinogenes 130Z produces relatively large amounts of byproducts, fluoroacetate-based screening was performed to reduce production of formic and acetic acids. The mutant strain, FZ53, produced 105.8 g/L of SA with the yield and productivity of 1.22 mol/mol glucose and 1.34 g/L/ h, respectively, in anaerobic batch fermentation [40]. Although less byproducts were produced by the mutant strain, production of byproducts could not be completely prevented. In addition to pyruvic acid, propionic acid that is normally not observed with other SA producers were accumulated. Relatively less studies have been reported on metabolic engineering of A. succinogenes. Michigan Biotechnology Institute (MBI) has been employing mutant and engineered A. succinogenes strains for the production of SA since 1981 (Figure 3). Performance indices reported in the patent by MBI is summarized in Table 1. www.sciencedirect.com

Succinic acid production by engineered microorganisms Ahn, Jang and Lee 63

Figure 3

10,000 tonnes/yr SA production plant in Montmeló, Spain (2013) Joint venture between BASF & Corbion Purac Basfia succiniciproducens

30,000 tonnes/yr SA production plant in Sarnia, Canada (2014) Low-pH yeast technology Pichia kudriavzevii

Tokyo, Japan (2006) Joint project partner with Mitsubishi Chemical Corynebacterium glutamicum

Lansing, U.S.A (1981) Affiliated with Michigan State University Actinobacillus succinogenes

15,000 tonnes/yr SA production plant in Lake providence, U.S.A (2013) Escherichia coli

10,000 tonnes/yr SA production plant in Cassano Italy (2011) Joint venture between Royal DSM & Roquette Frères Saccharomyces cerevisiae

Daejeon, Korea (2000) Affiliated with KAIST Mannheimia succiniciproducens

Current Opinion in Biotechnology

Overview on industrial production of SA by companies including Bioamber, Myriant, Reverdia, and Succinity. Activities on highly efficient succinic acid production by Michigan Biotechnology Institute and KAIST are also shown.

Corynebacterium glutamicum

C. glutamicum is a Gram positive, facultative anaerobic, and heterotrophic bacterium, and is well established for industrial-scale production of various amino acids and industrially relevant products. Full genome sequence of C. glutamicum has been determined and numerous genetic tools are available for strain development. When aerobically grown C. glutamicum is incubated in anaerobic condition, cell growth is arrested and a mixture of lactic acid, acetic acid, and SA is produced from glucose. C. glutamicum Dldh-pCRA717 strain was developed by deleting lactate dehydrogenase and overexpressing the native pyruvate carboxylase [41]. This mutant strain produced 146 g/L of SA with a yield of 1.4 mol/mol glucose by high cell density fermentation using the aerobically pre-cultured cells at 50 g dry cells/L under anaerobic condition; thus, the amount of glucose consumed and time needed for aerobic growth phase were not taken into account for SA yield and productivity calculations. Pyruvate carboxylase utilizes bicarbonate for the conversion of pyruvate to oxaloacetate. Hence, supplementing calcium bicarbonate during fermentation was found to strongly increase SA production showing high proportionality between the rates of bicarbonate consumption and SA production. To reduce the production of lactic and acetic acids, the pta, ackA, pqo, cat, and ldhA genes that encode phosphotransacetylase, acetate kinase, pyruvate:menaquinone oxidoreductase, acetyl CoA:CoA transferase, and lactate www.sciencedirect.com

dehydrogenase, respectively, were deleted [42]. Pyruvate carboxylase was integrated into the chromosome to increase carboxylation flux, and the NAD+-coupled Mycobacterium vaccae formate dehydrogenase was chromosomally integrated to provide additional reducing equivalents [42]. Finally, glyceraldehyde 3-phosphate dehydrogenase was overexpressed to develop BOL3/ pAN6-gap strain [42]. The resulting strain produced 134 g/L of SA with and yield and productivity of 1.67 mol/mol glucose and 2.53 g/L/h, respectively, using anaerobic high cell density (using the aerobically precultured cells at 50 g dry cells/L) fed-batch fermentation with formic acid supplementation [42]. Again, the aerobic cell propagation stage was not included in the calculation of SA yield and productivity. Glyceraldehyde 3-phosphate dehydrogenase was found to stimulate glucose consumption under anaerobic conditions and leads to high ratio of NADH to NAD+ [43]. Another strain that uses dual synthesis route to anaerobically produce SA without the need for formic acid supplementation was developed. Similar to other SA producers, the ldhA, pqo, ackA, and cat genes were deleted and the pyc and ppc genes were overexpressed [44]. Then, the glyoxylate shunt was reconstructed by the overexpression of isocitrate lyase, malate synthase, and citrate synthase [44]. Citrate synthase was overexpressed to direct carbon flux towards glyoxylate shunt. Lastly, the sucE gene encoding SA exporter under anaerobic condition was overexpressed to make the SA5 strain [44]. Using the aerobically pre-cultured cells at 27.5 g dry Current Opinion in Biotechnology 2016, 42:54–66

64 Pharmaceutical biotechnology

cells/L, fed-batch fermentation resulted in production of 109 g/L of SA with yield and productivity of 1.32 mol/mol glucose and 1.1 g/L/h and, respectively [44]; here, the aerobic cell propagation stage was not taken to account in the calculation of SA yield and productivity. There have been a few reports on aerobic SA production as well. In a study, the combined overexpression of the pycP458S and ppc with disruption of succinate dehydrogenase and acetate-forming pathway resulted in development of the BL-1(pAN6-pycP458Sppc) strain capable of producing 9.7 g/L of SA with a yield of 0.36 mol/mol glucose in aerobic batch fermentation [45]. In another study, the sdhCAB, ldhA, pqo, cat, and pta disrupted C. glutamicum mutant, ZX1 strain expressing the B. subtilis acetyl-CoA synthase (acs) gene was developed to increase SA production while reducing acetic acid production under aerobic condition [46]. Finally, the ZX1 (pEacsAgltA) strain was developed by overexpressing the inherent gltA gene encoding citrate synthase, while the native promoters for ppc and pyc genes were replaced by a stronger sod promoter to pull more carbon flux towards SA production [46]. Aerobic fed-batch fermentation of the ZX1 (pEacsAgltA) strain produced 28.5 g/L of SA with yield and productivity of 0.63 mol/mol glucose and 0.42 g/L/h, respectively [46].

The costs for SA recovery can be high like other bulk chemicals. Thus, efficient recovery methods such as selective precipitation of SA without generating gypsum and novel cost-effective extraction need to be developed. In our own experience, homo-SA production by M. succiniproducens allowed development of cost-effective recovery process without gypsum formation. Also, integrated optimization of fermentation and downstream processes will be essential to reduce the overall cost of SA production. It should be noted that strain might need to be further engineered during the process optimization and scale-up processes and also abundant inexpensive feedstock to be employed [48]. When the oil price declines, there is always concern on economic competitiveness of bio-based chemicals. As the recent COP-21 in Paris revealed, however, achieving sustainability without causing environmental problems including climate change will become essential practice of chemical industry. SA is in fact in much better position than other bio-based chemicals with respect to production cost (competitive with petrochemical route) and environmental benefits (additional fixing of carbon dioxide in addition to the use of renewable non-food biomass as a substrate). It is expected that bio-based SA will be produced in large amounts and used in many applications in the near future.

Acknowledgements Ajinomoto and Mitsubishi Chemical jointly developed SA process using engineered C. glutamicum since 2006 (Figure 3) [47]. Performance indices reported in the patent by Ajinomoto is summarized in Table 1.

Conclusions and future perspectives As briefly reviewed in this paper, great advances have been made in developing metabolically engineered microorganisms capable of efficiently producing SA in industrial scale. Several companies started producing SA in increasingly large scale. Each company is employing its own uniquely engineered microorganism. Since the performance indices of SA production shown in Table 1 are those explicitly reported in the literature (journal papers and patents), it should be noted that actual performance indices at companies are expected to be much better. What lies in the future? Just like other bulk bioproducts, the strategies for pushing the performance indices to their maximum limits will be continuously developed and employed for further improvement of strain performance; the general strategies of systems metabolic engineering can be consulted for this purpose [48]. Fermentation strategy will depend on the strain to be employed, and those processes such as semi-continuous and repeated fed-batch fermentation giving high volumetric productivity (without sacrificing yield) will be developed and employed in large-scale fermentation to reduce direct fixed capital cost as well as annual operating cost (reduced depreciation costs due to the use of smaller bioreactor). Current Opinion in Biotechnology 2016, 42:54–66

This work was supported by the Technology Development Program to Solve Climate Changes on Systems Metabolic Engineering for Biorefineries from the Ministry of Science, ICT and Future Planning (MSIP) through the National Research Foundation (NRF) of Korea (NRF-2012M1A2A2026556 and NRF-2012M1A2A2026557). The expanded version of this paper will appear as a book chapter ‘Succinic Acid’ in Wiley’s Advanced Biotechnology book series edited by SY Lee, J Nielsen, and G Stephanopoulos. Authors declare that there is conflict of interest in authors’ work on M succiniciproducens technology as it is of commercial interest, and thus strains might not be available for distribution.

References and recommended reading Papers of particular interest, published within the period of review, have been highlighted as:  of special interest  of outstanding interest 1.

Jansen ML, van Gulik WM: Towards large scale fermentative production of succinic acid. Curr Opin Biotech 2014, 30:190-197.

2.

Pinazo JM, Domine ME, Parvulescu V, Petru F: Sustainability metrics for succinic acid production: a comparison between biomass-based and petrochemical routes. Catal Today 2015, 239:17-24.

3.

Choi S, Song CW, Shin JH, Lee SY: Biorefineries for the production of top building block chemicals and their derivatives. Metabol Eng 2015, 28:223-239.

4.

Cok B, Tsiropoulos I, Roes AL, Patel MK: Succinic acid production derived from carbohydrates: an energy and greenhouse gas assessment of a platform chemical toward a bio-based economy. Biofuel Bioprod Bior 2014, 8:16-29.

5.

Kubo Y, Takagi H, Nakamori S: Effect of gene disruption of succinate dehydrogenase on succinate production in a sake yeast strain. J Biosci Bioeng 2000, 90:619-624.

6.

Ito Y, Hirasawa T, Shimizu H: Metabolic engineering of Saccharomyces cerevisiae to improve succinic acid www.sciencedirect.com

Succinic acid production by engineered microorganisms Ahn, Jang and Lee 65

production based on metabolic profiling. Biosci Biotechnol Biochem 2014, 78:151-159. 7.

Ida Y, Furusawa C, Hirasawa T, Shimizu H: Stable disruption of ethanol production by deletion of the genes encoding alcohol dehydrogenase isozymes in Saccharomyces cerevisiae. J Biosci Bioeng 2012, 113:192-195.

8.

Jouhten P, Rintala E, Huuskonen A, Tamminen A, Toivari M, Wiebe M, Ruohonen L, Penttila¨ M, Maaheimo H: Oxygen dependence of metabolic fluxes and energy generation of Saccharomyces cerevisiae CEN. PK113-1A. Bmc Syst Biol 2008, 2:60.

9.

Raab AM, Gebhardt G, Bolotina N, Weuster-Botz D, Lang C: Metabolic engineering of Saccharomyces cerevisiae for the biotechnological production of succinic acid. Metabol Eng 2010, 12:518-525.

10. Otero JM, Cimini D, Patil KR, Poulsen SG, Olsson L, Nielsen J: Industrial systems biology of Saccharomyces cerevisiae enables novel succinic acid cell factory. PloS One 2013, 8:54144. 11. Li Y, Li M, Zhang X, Yang P, Liang Q, Qi Q: A novel whole-phase succinate fermentation strategy with high volumetric productivity in engineered Escherichia coli. Bioresour Technol 2013, 149:333-340. 12. Enomoto K, Arikawa Y, Muratsubaki H: Physiological role of soluble fumarate reductase in redox balancing during anaerobiosis in Saccharomyces cerevisiae. FEMS Microbiol Lett 2002, 215:103-108. 13. Yan D, Wang C, Zhou J, Liu Y, Yang M, Xing J: Construction of reductive pathway in Saccharomyces cerevisiae for effective succinic acid fermentation at low pH value. Bioresour Technol 2014, 156:232-239. 14. Agren R, Otero JM, Nielsen J: Genome-scale modeling enables metabolic engineering of Saccharomyces cerevisiae for  succinic acid production. J Ind Microbiol Biotechnol 2013, 40:735-747. Genome-scale metabolic modeling and flux balance analysis were performed for the prediction of succinic acid overproduction strategies in S. cerevisiae. Deletion of decarboxylate mitochondrial transporter followed by led to the improvement of SA yield to 0.03 mol/mol glucose. Subsequent transcriptional profiling studies identified that succinate formation is coupled to mitochondrial redox balancing. 15. Sandstro¨m AG, Almqvist H, Portugal-Nunes D, Neves D, Lide´n G, Gorwa-Grauslund MF: Saccharomyces cerevisiae: a potential host for carboxylic acid production from lignocellulosic feedstock? Appl Microbiol Biot 2014, 98:7299-7318. 16. Finley KR, Huryta JM, Mastel BM, McMullin TW, Poynter GM, Rush BJ, Watts KT, Fosmer AM, McIntosh VL, Brady KM: Compositions and methods for succinate production. US patent 2013, US20130302866A1. 17. Rush BJ, Fosmer AM: Methods for succinate production. US  patent 2014, US20140363862A1. C. krusei mutant strain was developed by Bioamber for the enhanced production of SA under low pH condition. The engineered strain produced 48.2 g/L of SA with the yield and productivity of 0.69 mol/mol glucose and 0.97 g/L/h, respectively, by aerobic batch fermentation at pH 3. 18. Singh A, Lynch MD, Gill RT: Genes restoring redox balance in fermentation-deficient E. coli NZN111. Metabol Eng 2009, 11:347-354. 19. Wang W, Li Z, Xie J, Ye Q: Production of succinate by a pflB ldhA double mutant of Escherichia coli overexpressing malate dehydrogenase. Bioprocess Biosyst Eng 2009, 32:737-745. 20. Chatterjee R, Millard CS, Champion K, Clark DP, Donnelly MI: Mutation of the ptsG gene results in increased production of succinate in fermentation of glucose by Escherichia coli. Appl Environ Microbiol 2001, 67:148-154. 21. Vemuri G, Eiteman M, Altman E: Succinate production in dualphase Escherichia coli fermentations depends on the time of transition from aerobic to anaerobic conditions. J Ind Microbiol Biotechnol 2002, 28:325-332. www.sciencedirect.com

22. Jantama K, Haupt M, Svoronos SA, Zhang X, Moore J, Shanmugam K, Ingram L: Combining metabolic engineering and metabolic evolution to develop nonrecombinant strains of Escherichia coli C that produce succinate and malate. Biotechnol Bioeng 2008, 99:1140-1153. 23. Balzer GJ, Thakker C, Bennett GN, San K-Y: Metabolic engineering of Escherichia coli to minimize byproduct formate and improving succinate productivity through increasing NADH availability by heterologous expression of NAD+dependent formate dehydrogenase. Metabol Eng 2013, 20:1-8. 24. Lin H, Bennett GN, San KY: Fed-batch culture of a metabolically engineered Escherichia coli strain designed for high-level succinate production and yield under aerobic conditions. Biotechnol Bioeng 2005, 90:775-779. 25. Lee SJ, Lee D-Y, Kim TY, Kim BH, Lee J, Lee SY: Metabolic engineering of Escherichia coli for enhanced production of succinic acid, based on genome comparison and in silico gene knockout simulation. Appl Environ Microbiol 2005, 71:7880-7887. 26. Myriant: Succinic Acid & Derivatives SBU. http://www.myriant. com/pdf/myriant-succinic-acid-customer-presentation-english. pdf. 27. Lee P, Lee S, Hong S, Chang H: Isolation and characterization of a new succinic acid-producing bacterium, Mannheimia succiniciproducens MBEL55E, from bovine rumen. Appl Microbiol Biotechnol 2002, 58:663-668. 28. Hong SH, Kim JS, Lee SY, In YH, Choi SS, Rih J-K, Kim CH, Jeong H, Hur CG, Kim JJ: The genome sequence of the capnophilic rumen bacterium Mannheimia succiniciproducens. Nat Biotechnol 2004, 22:1275-1281. 29. Lee SJ, Song H, Lee SY: Genome-based metabolic engineering of Mannheimia succiniciproducens for succinic acid production. Appl Environ Microbiol 2006, 72:1939-1948. 30. Oh IJ, Lee HW, Park CH, Lee SY, Lee J: Succinic acid production by continuous fermentation process using Mannheimia succiniciproducens LPK7. J Microbiol Biotechnol 2008, 18:908-912. 31. Lee SY, Lim SW, Song H: Novel engineered microorganism  producing homo-succinic acid and method for preparing succinic acid using the same. EP patent 2014, EP2054502B1. M. succiniciproducens mutant strain was developed by in silico prediction of gene knock-out targets. Fed-batch fermentation of the PALK strain (lactate dehydrogenase and phosphotransacetylase-acetate kinase mutant) resulted in the production of 45.8 g/L of SA with the yield and productivity of 1.32 mol/mol of glucose and 2.36 g/L/h, respectively. Almost homo-succinic acid production could be accomplished. 32. Lee SY, Lee JW, Choi S, Yi J: Mutant microorganism producing  succinic acid simultaneously using sucrose and glycerol, and method for preparing succinic acid using same. US patent 2014, US8691516B2. M. succiniciproducens mutant strain was developed to simultaneously utilize sucrose and glycerol for the enhanced production of SA. In addition to fed-batch fermentation, other fermentation processes, such as high cell density fermentation and membrane cell recycle bioreactor (MCRB) fermentation, were developed. Using MCRB fermentation, SA could be produced at a high productivity of 29.7 g/L/h with a yield of 1.54 mol/mol from sucrose and glycerol. 33. Scholten E, Da¨gele D: Succinic acid production by a newly isolated bacterium. Biotechnol Lett 2008, 30:2143-2146. 34. Kuhnert P, Scholten E, Haefner S, Mayor D, Frey J: Basfia succiniciproducens gen. nov., sp. nov., a new member of the family Pasteurellaceae isolated from bovine rumen. Int J Syst Evol Microbiol 2010, 60:44-50. 35. Becker J, Reinefeld J, Stellmacher R, Scha¨fer R, Lange A, Meyer H, Lalk M, Zelder O, von Abendroth G, Schro¨der H: Systems-wide analysis and engineering of metabolic pathway fluxes in bio-succinate producing Basfia succiniciproducens. Biotechnol Bioeng 2013, 110:3013-3023. 36. Lee SY, Lee SJ: Rumen bacteria variants and process for preparing succinic acid employing the same. US patent 2008, US7470530B2. Current Opinion in Biotechnology 2016, 42:54–66

66 Pharmaceutical biotechnology

37. Van der Werf MJ, Guettler MV, Jain MK, Zeikus JG: Environmental and physiological factors affecting the succinate product ratio during carbohydrate fermentation by Actinobacillus sp. 130Z. Arch Microbiol 1997, 167:332-342. 38. Song H, Lee SY: Production of succinic acid by bacterial fermentation. Enzyme Microb Technol 2006, 39:352-361. 39. Urbance SE, Pometto AL III, DiSpirito AA, Denli Y: Evaluation of succinic acid continuous and repeat-batch biofilm fermentation by Actinobacillus succinogenes using plastic composite support bioreactors. Appl Microbiol Biotechnol 2004, 65:664-670. 40. Guettler MV, Jain MK, Rumler D: Method for making succinic acid, bacterial variants for use in the process, and methods for obtaining variants. US patent 1996, US5573931A. 41. Inui M, Murakami S, Okino S, Kawaguchi H, Verte`s AA, Yukawa H: Metabolic analysis of Corynebacterium glutamicum during lactate and succinate productions under oxygen deprivation conditions. J Mol Microb Biotech 2004, 7:182-196. 42. Litsanov B, Brocker M, Bott M: Toward homosuccinate fermentation: metabolic engineering of Corynebacterium glutamicum for anaerobic production of succinate from glucose and formate. Appl Environ Microbiol 2012, 78:3325-3337. 43. Jojima T, Fujii M, Mori E, Inui M, Yukawa H: Engineering of sugar metabolism of Corynebacterium glutamicum for production of amino acid L-alanine under oxygen deprivation. Appl Microbiol Biotechnol 2010, 87:159-165. 44. Zhu N, Xia H, Yang J, Zhao X, Chen T: Improved succinate  production in Corynebacterium glutamicum by engineering glyoxylate pathway and succinate export system. Biotechnol Lett 2014, 36:553-560. A dual route for anaerobic SA production was introduced into C. glutamicum. Combined engineering of glyoxylate shunt and SA export system effectively enhanced SA production. The engineered strain produced 109 g/L with the yield and productivity of 1.32 mol/mol glucose and 1.1 g/L/h, respectively, by aerobic cell cultivation followed by anaerobic fed-batch fermentation; the aerobic cell propagation stage was not taken to account in the calculation of SA yield and productivity. 45. Litsanov B, Kabus A, Brocker M, Bott M: Efficient aerobic succinate production from glucose in minimal medium with Corynebacterium glutamicum. Microb Biotechnol 2012, 5:116-128. 46. Zhu N, Xia H, Wang Z, Zhao X, Chen T: Engineering of acetate  recycling and citrate synthase to improve aerobic succinate production in Corynebacterium glutamicum. PloS One 2013, 8:e60659. An acetate assimilation pathway from B. subtilis was introduced into C. glutamicum to reduce acetate accumulation and enhance SA production. As a result, this engineered strain produced 28.5 g/L of SA with an average yield and productivity of 0.63 mol/mol glucose and 0.4 g/L/h, respectively under aerobic condition. 47. Nishi K, Kaneko H, Tasaki H, Koga M, Tanegawa T, Furuya S, Fujiwara K, Sato T: Method of manufacturing succinic acid and ammonium succinate solutions. US patent 2011, US8034975B2.

Current Opinion in Biotechnology 2016, 42:54–66

48. Lee SY, Kim HU: Systems strategies for developing industrial microbial strains. Nat Biotechnol 2015, 33:1061-1072. 49. Yocum RR, Hermann T, Hu X: Fermentation of glycerol to  organic acids. US patent 2014, US20140234923A1. A glycerol utilizing E. coli mutant strain was developed by Myriant for the enhanced production of SA. This engineered strain produced 84.3 g/L of SA with the yield and productivity of 1.53 mol/mol glycerol and 1.76 g/L/h, respectively, by batch fermentation under aerobic condition. 50. Van De Graaf MJ, Valianpoer F, Fiey G, Delattre L, Schulten EAM:  Process for the crystallization of succinic acid. US patent 2015, US20150057425A1. A S. cerevisiae mutant strain was developed by Reverdia for the enhanced production of SA under low pH condition. The engineered strain produced 43 g/L of SA in aerobic condition, using dual phase fedbatch fermentation at pH 3. 51. Schroder H, Haefner S, Von Abendroth G, Hollmann R, Raddatz A,  Ernst H, Gurski H: Microbial succinic acid producers and purification of succinic acid. US patent 2014, US8673598B2. B. succiniciproducens mutant strain was developed by Succinity for the enhanced production of SA. The engineered strain used glycerol and maltose to produce 70 g/L of SA with the yield and productivity of 1.7 mol/mol and 2.91 g/L/h, respectively, by anaerobic fed-batch fermentation 52. Lee SJ, Lee D-Y, Kim TY, Kim BH, Lee J, Lee SY: Metabolic engineering of Escherichia coli for enhanced production of succinic acid, based on genome comparison and in silico gene knockout simulation. Appl Environ Microbiol 2005, 71:7880-7887. 53. Ask M, Mapelli V, Hock H, Olsson L, Bettiga M: Engineering glutathione biosynthesis of Saccharomyces cerevisiae increases robustness to inhibitors in pretreated lignocellulosic materials. Microb Cell Fact 2013, 12:87. 54. Jansen MLA, Verwaal R: Dicarboxylic acid production by fermentation at low pH. US patent 2011, US20110229945A1. 55. Jansen MLA, Van De Graaf MJ, Verwaal R: Dicarboxylic acid production process. US patent 2012, US20120040422A1. 56. Grabar T, Gong W, Yocum RR: Metabolic evolution of Escherichia coli strains that produce organic acids. US patent 2012, US20120202259A1. 57. Scholten E, Da¨gele D, Haefner S, Schro¨der H: Bacterial strain and process for the fermentative production of organic acids. US patent 2013, US8574875B2. 58. Scholten E, Daegele D, Haefner S, Schroeder H: Novel microbial succinic acid producer. US patent 2011, US20110300589. 59. Scholten E, Haefner S, Schro¨der H: Bacterial cells having a glyoxylate shunt for the manufacture of succinic acid. US patent 2014, US8877466B2. 60. Murase M, Aoyama R, Ikuta M, Yamagishi K, Moriya M, Nakamura J, Kojima H: Process for producing succinic acid. EP patent 2015, EP1672077B1.

www.sciencedirect.com