Increasing succinic acid production using the PTS-independent glucose transport system in a Corynebacterium glutamicum PTS-defective mutant

Abstract

Succinic acid synthesized from glucose shows potential as a bio-based platform chemical. However, the need for a high glucose concentration, and the accompanying low yields, limit its industrial applications. Despite efficient glucose uptake by the phosphotransferase system (PTS), 1 mol of phosphoenolpyruvate is required for each mole of internalized glucose. Therefore, a PTS-defective Corynebacterium glutamicum mutant was constructed to increase phosphoenolpyruvate availability for succinic acid synthesis, resulting in a lower glucose utilization rate and slower growth. The transcriptional regulator iolR was also deleted to enable the PTS-defective mutant to utilize glucose via iolT-mediated glucose transport. Deletion of iolR and overexpression of iolT1 and ppgk (polyphosphate glucokinase) in the PTS-deficient C. glutamicum strain completely restored glucose utilization, increasing production by 11.6 % and yield by 32.4 % compared with the control. This study revealed for the first time that iolR represses the expression of the two glucokinase genes (glk and ppgk).

Zhihui Zhou and Chen Wang contributed equally to this work.

Introduction

As a member of the C4-dicarboxylic acid family, succinic acid is a platform chemical with potential for bio-based synthesis. It is a precursor for the synthesis of various important chemicals, including 1,4-butanediol, tetrahydrofuran, and adipic acid [30], and demand for succinic acid is increasing [22]. Succinic acid is currently produced from fossil fuels by a synthetic process. However, the high environmental cost of this process motivates the development of a cleaner and more economical biological routes to succinic acid production.

The bio-based succinic acid industry is mainly focused on succinic acid-producing organisms, including Corynebacterium glutamicum [18, 25], Anaerobiospirillum succiniciproducens [27], Actinobacillus succinogenes [5], Escherichia coli [16], and Mannheimia succiniciproducens [17]. In particular, C. glutamicum, a rapidly growing Gram-positive bacterium, is commonly used to produce various amino and organic acids [23, 26, 34]. Under anaerobic conditions, C. glutamicum can ferment glucose, to give lactic acid, succinic acid, and acetic acid [8, 23].

Theoretically, 1.71 mol of succinic acid can be produced per mole of glucose (plus CO₂) depending on the electron availability. This theoretical yield can be increased to 2.00 mol of succinic acid per mol of glucose if CO₂ and additional reducing power are supplied [22]. Wild-type C. glutamicum contains a metabolic pathway for succinic acid production that yields only 0.29 mol per mole of glucose. However, metabolic engineering can increase succinic acid yields in this species. For example, a C. glutamicum ∆ldhA (lactate dehydrogenase) mutant with enhanced pyruvate carboxylase activity showed an increased succinic acid yield of 1.4 mol/mol [24], while a BOL-1 strain overexpressing pyc, along with an NAD⁺-coupled formate dehydrogenase, accumulated succinic acid with a yield of 1.7 mol/mol [18].

In C. glutamicum, metabolite interconversion at the phosphoenolpyruvate (PEP)–pyruvate–oxaloacetate node (PPON) involves a set of reactions that connects the major pathways of carbon metabolism. These reactions are responsible for the distribution of the carbon flux among catabolism, anabolism, and the cellular energy supply (Fig. 1). Sauer et al. [28] showed that the PPON is a good target for metabolic engineering to achieve optimal metabolite production. Under anaerobic conditions, succinic acid is primarily synthesized from PEP, which is converted to oxaloacetate by phosphoenolpyruvate carboxylase (encoded by ppc). Oxaloacetate is then converted to succinic acid via the reductive tricarboxylic acid cycle (TAC) branch [8].

Fig. 1

Schematic representation of the metabolic pathways relevant for succinic acid production by C. glutamicum under anaerobic condition. Bold line predominant pathways under oxygen deprivation. Dashed lines the oxidative branch of TCA pathway where the enzymes of are repressed by glucose and hypoxia [2]. IMGlc glucose import system, IolT1, IolT2 inositol transporters, Glk, Ppgk glucokinases, IolR transcriptional regulator, PtsI, PtsH, PtsG phosphoenolpyruvate:sugar phosphotransferase system (PTS), Pck: PEP carboxykinase, Ppc:PEP carboxylase, Pyc pyruvate carboxylase, Pyk pyruvate kinase, Mdh malate dehydrogenase, Fum fumarase, Sdh succinodehydrogenase, MalE malic enzyme

Under normal conditions, half of the available PEP in the cell is used for glucose uptake and phosphorylation, which reduces the amount of PEP available for succinic acid production [3]. Wittmann et al. [37] investigated the metabolic flux distribution in a pts⁻  C. glutamicum strain versus the control strain and showed that the PTS-defective mutant had increased PEP available for the synthesis of aromatic compounds. However, the slow growth and low glucose utilization rates of the PTS-defective mutant meant that it was not suitable for industrial applications [3, 20, 21, 33, 36]. C. glutamicum contains two glucokinases: glk (cg2399), and ppgk (cg2091). Improved glucose utilization and l-lysine production were achieved by overexpressing iolT1 and ppgk in the PTS⁻ strain [19]. In an E. coli PTS⁻ mutant, overexpression of galP (encoding galactose permease) and glk (encoding glucokinase) restored glucose transport and increased the glycolytic flux to fermentation products [6]. Two myo-inositol transporters capable of mediating glucose uptake, encoded by iolT1 (Cgl0181) and iolT2 (Cgl3058), have been identified in C. glutamicum [9]. These transporters are responsible for myo-inositol utilization and can be induced by a specific concentration of inositol [15]. Glucose phosphorylation is not involved in PTS-independent glucose uptake; thus, glucokinase activity may be required.

In this study, two strategies were attempted to restore the glucose utilization rate of a PTS-defective C. glutamicum strain. Ikeda et al. [9] previously identified a suppressor mutant derived from a PTS-defective C. glutamicum strain on glucose agar plates, which was later confirmed to have an iolR mutation. The promoter operator region of iolT1 could be repressed in this suppressor mutant, and real-time polymerase chain reaction (RT-PCR) analysis was used to investigate the effects of iolR on the expression of PTS-independent glucose utilization genes. However, little is known about these genes in the iolR-deficient mutant, particularly the effects of this mutation on the expression of ppgk and glk [9, 11]. Therefore, our first strategy was to experimentally inactivate iolR and observe the effects on gene expression. The second strategy was to increase the transcriptional levels of iolT1 and compensate for the deletion of ppgk in the PTS⁻ strain through overexpression.

Materials and methods

Microorganisms and medium

Strain genotypes and plasmids are shown in Table 1. For plasmid construction, E. coli JM109 was cultured at 37 °C in lysogeny broth complex medium (LB) [29]. C. glutamicum strains were routinely cultivated at 30 °C in LB medium. For selection of pK18mobsacB and its derivatives, 50 and 25 μg/ml kanamycin was added to E. coli and C. glutamicum cultures, respectively. Nutrient-rich medium (A medium) was used for aerobic growth and the mineral salts medium (BT medium) was used for anaerobic fermentation [38].

Bacterial strains and plasmids used in this study

Strain, plasmids	Genotypes, properties	Sources or references
Strains
C. glutamicum NC-3	C. glutamicum ATCC 13032 derivative with an in-frame deletion of the ldhA, with integration of xylose metabolic gene (xylA ,xylB) and gapA	[35]
C. glutamicum NC-3b	C. glutamicum NC-3 derivative with inactivation of ptsG	This study
C. glutamicum NC-3b-1	C. glutamicum NC-3b derivative with in-frame deletion of iolR	This study
C. glutamicum NC-3b-iolT1	C. glutamicum NC-3b derivative with chromosomal integration of gapA promoter and iolT1 gene into the ptsG locus	This study
C. glutamicum NC-3b-iolT1(pXMJ19-ppgk)	C. glutamicum NC-3b-iolT1 derivative with pXMJ19-ppgk	This study
C. glutamicum NC-3b-2	C. glutamicum NC-3b-1 derivative with chromosomal integration of gapA promoter and iolT1 gene into the ptsG locus	This study
C. glutamicum NC-3b-3	C. glutamicum NC-3b-2 derivative with chromosomal integration of ppgk gene under the control of the gapA promoter into the iolR locus	This study
E. coli JM109	recA1 endA1 gyrA96 thi hsdR17 supE44 relA1 Δ(lac-proAB)/F′ [traD36 proAB + lacIq  lacZΔM15]	Takara
Plasmids
pXMJ19	Cmr; E. coli-C. glutamicum shuttle vector, tac promoter	[12]
pXMJ19-ppgk	Cmr, pXMJ19 derivative containing the ppgk gene under the control of an IPTG-inducible tac promoter	This study
pk18mobsacB	Kanr; vector for allelic exchange in C. glutamicum	[32]
pk18mobsacB∆ptsG	Derived from pk18mobsacB, containing up- and downstream regions of ptsG	This study
pk18mobsacB∆iolR	Derived from pk18mobsacB, containing up- and downstream regions of iolR	This study
pk18mobsacB∆ptsG::PgapA-iolT1	Derived from pk18mobsacB∆ptsG with 2.8-kb EcoRV-SphI DNA fragment containing PgapA-iolT1 gene	This study
pk18mobsacB∆iolR::PgapA-ppgk	Derived from pk18mobsacB∆iolR with 1.7-kb XbaI-SphI DNA fragment containing PgapA-ppgk gene	This study

Anaerobic succinic acid production in a lidded bottle

A single colony of the appropriate C. glutamicum strain was inoculated into 5 ml of LB medium from a fresh LB agar plate, and the culture was incubated on a rotary shaker for 12 h at 30 °C. A 2-ml aliquot of this starter culture was then inoculated into a 500-ml baffled shake flask containing 50 ml of A medium supplemented with 110 mM glucose for 18 h, 200 rpm at 30 °C. The cells were then harvested by centrifugation (8000×g, 4 °C, 10 min), resuspended in 25 ml of BT medium, and transferred into a 100-ml lidded bottle. The cell suspension was incubated for 18 h, 150 rpm at 30 °C. Oxygen deprivation (dissolved oxygen concentration <0.01 ppm) was achieved using a gassing manifold with oxygen-free CO₂ for 30 s according to the manufacturer’s instructions.

Anaerobic succinic acid production in fed-batch mode

A single colony was picked from plates and inoculated into 5 ml LB medium and cultivated at 30 °C for 12 h. Aliquots (2 ml) were inoculated into three 500-ml baffled shake flasks containing 100 ml of A medium supplemented with 110 mM glucose. After approximately 16 h of incubation at 200 rpm and 30 °C, or when cultures reached an optical density at 600 nm (OD₆₀₀) of 10, a 5-l fermentation bioreactor (BIOTECH-5JG, China) containing 3 l of A medium supplemented with 185 mM glucose was inoculated with the starter culture and run for 16 h, to a final OD₆₀₀ of 35. The cells were then harvested by centrifugation (8000×g, 4 °C, 10 min), resuspended in 1 l of BT medium containing 367 mM glucose and 300 mM NaHCO₃, and transferred into a 1-l bioreactor (Biotech-5JG-9000A, China). Oxygen deprivation of oxygen (dissolved oxygen concentration <0.01 ppm) was achieved using a gassing manifold with oxygen-free CO₂ for 5 min. Under anaerobic conditions, the cell suspension was incubated at 30 °C and 150 rpm for 48 h. In two independent fermentations, the pH was maintained at 7.2 by automated addition of 20 % (v/v) NH₃·H₂O.

Recombinant DNA techniques

Standard protocols were used for construction, purification, and analysis of plasmid DNA, and for E. coli transformation. Extraction of C. glutamicum chromosomal DNA and transformation of C. glutamicum by electroporation were performed as described previously [9]. PCR was performed using a DNA thermal cycler (2720 Thermal cycler, Applied Biosystems, USA) using Prime STARHS DNA Polymerase (Takara Bio, Japan). The restriction endonuclease, ligase, and other materials for the polymerase chain reaction (PCR) were purchased from Takara. PCR was performed in a 50-µl reaction system containing 5 µl MgCl₂ (20 mmol/l), 5 µl dNTP_S (20 mmol/l), 3 µl chromosomal DNA (400 µg/ml), 1 µl DNA polymerase, and 5 µl oligonucleotide primers, made up to the final volume with ddH₂O. The thermal cycler conditions were 30 cycles of 95 °C for 30 s, 62 °C for 15 s, and 72 °C for 2 min 30 s. All primers used in this study are shown in Table 2 and were synthesized by Genscript Biotechnology (Nanjing, China).

Sequences of primers

Name	Sequence (5′–3′)	Function and restriction site
ppgk-F	GATAAGCTT  AAAGGAGGACAACCATGACTGAGACTGGATTT	Expression of ppgk; HindIII
ppgk-R	ACAGAATTCTTATGGGGTGAGGTGTTG	Expression of ppgk; EcoRI
∆ptsG-F1	TCCCCCGGGATGGCGTCCAAACTG	ptsG deletion; SmaI
∆ptsG-R1	ACAT GCATGC TTAG GATATC CGGTGGCAGGAAGTAGAA	ptsG deletion; SphI, EcoRV
∆ptsG-F2	CCG GATATC CTAA GCATGC ATGTACCAGGCATTGCAAT	ptsG deletion; EcoRV, SphI
∆ptsG-R2	GATAAGCTTTACTCGTTCTTGCCG	ptsG deletion; HindIII
∆iolR-F1	TCCCCCGGGACTTCGTGAGTGCTC	iolR deletion; SmaI
∆iolR-R1	ACAT GCATGC TTAG TCTAGA CTAGGGGAGCTTCGGTGGTCAT	iolR deletion; SphI, XbaI
∆iolR-F2	CTAG TCTAGA CTAA GCATGC ATGTCGGTGCGCGCCGAGCAGT	iolR deletion; XbaI, SphI
∆iolR-R2	GATAAGCTTGAAACCAGCCCATGT	iolR deletion; HindIII
PgapA-iolT1-F1	CCGGATATCCGGCGAAAACGAAA	Overexpression of iolT1 under the control of the gapA promoter; EcoRV
PgapA-iolT1-R1	CGGCCTGAATGAAGGTACTAGCCATGTTGTGTCTCCTCTAAAGATTGTAGG
PgapA-iolT1-F2	CCTACAATCTTTAGAGGAGACACAACATGGCTAGTACCTTCATTCAGGCCG
PgapA-iolT1-R2	ACATGCATGCGCTGTGATCACACCATG	Overexpression of iolT1 under the control of the gapA promoter; SphI
PgapA-ppgk-F1	CTAGTCTAGACGGCGAAAACGAAA	Overexpression of ppgk under the control of the gapA promoter; XbaI
PgapA-ppgk-R1	CAATTCCAAATCCAGTCTCAGTCATGTTGTGTCTCCTCTAAAGATTGTAGG
PgapA-ppgk-F2	CCTACAATCTTTAGAGGAGACACAACATGACTGAGACTGGATTTGGAATTG
PgapA-ppgk-R2	ACATGCATGCTTATGGGGTGAGGTGTTG	Overexpression of ppgk under the control of the gapA promoter; SphI

Restriction sites are underlined and linker sequences for crossover PCR are shown in italics. The bold letters mean the Shine–Dalgarno sequence

Construction of plasmids and C. glutamicum mutant strains

A C. glutamicum mutant with an in-frame deletion of ptsG, which encodes Glc, a membrane-bound, glucose-specific permease of the phosphotransferase system, was constructed in a two-step homologous recombination procedure as described previously [10]. Briefly, regions flanking ptsG were amplified by PCR using the primer pairs ∆ptsG-F1/∆ptsG-R1 and ∆ptsG-F2/∆ptsG-R2. The resulting amplicons were then joined using a crossover PCR protocol using the primer pair ∆ptsG-F1/∆ptsG-R2. The resulting 906-bp product was digested with SmaI/HindIII and cloned into pK18mobsacB digested with the same restriction enzymes, generating pK18mobsacB∆ptsG. pK18mobsacB∆ptsG was electroporated into C. glutamicum NC-3, generating C. glutamicum NC-3b. The deletion of ptsG was verified by PCR using the primers ∆ptsG-F1 and ∆ptsG-R2. Primers used are listed in Table 2.

iolR was inactivated by a crossover PCR. Flanking regions of iolR were amplified using the primer pairs ∆iolR-F1/∆iolR-R1 and ∆iolR-F2/∆iolR-R2. The two amplicons were then joined by crossover PCR using the primers ∆iolR-F1 and ∆iolR-R2. The resulting 1-kb product was digested with SmaI/HindIII and cloned into pK18mobsacB digested with the same restriction enzymes, generating pK18mobsacB∆iolR. The resulting plasmid was isolated and used for gene disruption. Integration into the genome in the resulting strain C. glutamicum NC-3b-1 was verified by PCR using primers ∆iolR-F1 and ∆iolR-R2.

Plasmid pK18mobsacB∆ptsG::P_gapA-iolT1 was used to express iolT1 by inserting iolT1 into the ptsG locus. To construct pK18mobsacB∆ptsG::P_gapA-iolT1, the coding region of iolT1 was amplified using primers P_gapA-iolT1-F2 and P_gapA-iolT1-R2, with C. glutamicum genomic DNA as a template. The upstream region of gapA (from −1 to −927 bp), containing the promoter, was amplified using primers P_gapA-iolT1-F1 and P_gapA-iolT1-R2. The two fragments were fused by PCR using primers P_gapA-iolT1-F1 and P_gapA-iolT1-R2. The resulting 2.8-kb fragment was digested with EcoRV/SphI, and then ligated into the corresponding sites of pK18mobsacB∆ptsG to yield pK18mobsacB∆ptsG::P_gapA-iolT1. Plasmid pK18mobsacB∆ptsG::P_gapA-iolT1 was electroporated into C. glutamicum NC-3b and C. glutamicum NC-3b-1, resulting in C. glutamicum NC-3b-iolT1 and C. glutamicum NC-3b-2, respectively.

Vector pXMJ19 was used for IPTG-inducible overexpression. ppgk was amplified from C. glutamicum ATCC13032 genomic DNA using the primers listed in Table 2. To construct pXMJ19-ppgk, the PCR product was digested with HindIII/EcoRI and cloned into pXMJ19 digested with the same restriction enzymes. pXMJ19-ppgk was electroporated into C. glutamicum NC-3b-iolT1, resulting in C. glutamicum NC-3b-iolT1(pXMJ19-ppgk).

To construct pK18mobsacB∆iolR::P_gapA-ppgk, the coding region of ppgk was amplified using primers P_gapA-ppgk-F2 and P_gapA-ppgk-R2, while the gapA promoter region was amplified using primers P_gapA-ppgk-F1 and P_gapA-ppgk-R1. The two fragments were fused by PCR using the primers P_gapA-ppgk-F1 and P_gapA-ppgk-R2. The resulting 1.7-kb fragment was digested with XbaI/SphI and then ligated to the corresponding sites of pK18mobsacB∆iolR, yielding pK18mobsacB∆iolR::P_gapA-ppgk. This plasmid was then electroporated into C. glutamicum NC-3b-2, resulting in C. glutamicum NC-3b-3. All primers were designed based on the genome sequences of C. glutamicum ATCC13032.

RNA extraction, cDNA synthesis, and RT-PCR

Exponential growth-phase cells (1.5 ml) cultured in A medium were harvested by centrifugation at 8000×g for 1 min at 4 °C. Total RNA was extracted from C. glutamicum strains and purified as described previously [7]. cDNA was synthesized from 500 ng of RNA and was analyzed by RT-PCR, as previously described [13, 14], with gene expression levels standardized to 16S rRNA expression and calculated by the comparative cycle threshold method [31]. Oligonucleotides used in RT-PCR are listed in Table 3.

Oligonucleotides used in RT-PCR

Gene	Sequence (5′–3′)
iolT1	F: GTTGCACTAGTTGCGACGTT R: TTAGTCCGAGCTCACGTGTC
iolT2	F: CTCCATGCAGACTTTCCTCA R: CGATACCCTTCATTCGGACT
glk	F: CTGGAAGATTTCAGCGAGTG R: GGTCAAGGACATCAGCAATG
ppgk	F: CCAACACAGAACTCGGTCAC R: TTCTCGTATTCGCTCAGCAC

Analytical methods

Cell growth was monitored by measuring the cell density of cultures at OD₆₀₀ using a UV–Visible spectroscopy system (Mapada Co., China). Glucose concentration was determined using a glucose analyzer (SBA-40E, Biology Institute of Shandong Academy of Sciences, China) containing glucose oxidase according to the manufacturer’s instructions. The concentrations of succinic acid and its by-products (acetic acid and pyruvic acid) were determined using a high-performance liquid chromatography system (Agilent, USA) equipped with a UV detector and a conductivity meter, and a Grace Prevail column (length, 250 nm; internal diameter, 4.6 nm). The mobile phase consisted of 25 mM KH₂PO₄ (pH 2.5) at a flow rate of 1.0 ml/min, and the column was operated at 25 °C, with detection at OD₂₁₅. The mass yields of the products were defined as the amount of product generated per mole of glucose consumed, and expressed in mol/mol. Dry cell weight (DCW) was estimated by correlating OD₆₀₀ values with 0.3 g DCW/l as described previously [1].

Results

PtsG-deficient C. glutamicum exhibits higher succinic acid yield and lower glucose utilization

To increase the metabolic flux of PEP to oxaloacetate, a C. glutamicum ptsG mutant (NC-3b) was constructed by two-step homologous recombination. The growth of C. glutamicum NC-3b in minimal medium with 110 mM glucose as the sole carbon source was severely impaired compared with that of the wild type (growth rates of 0.17 ± 0.03 and 0.5 ± 0.02 h⁻¹, respectively) (Fig. 2a). The glucose consumption was also lower than that of C. glutamicum NC-3, with glucose consumption rates of 1.49 ± 0.3 and 3.94 ± 0.2 mM/h, respectively (Fig. 2b). However, the succinic acid yield of C. glutamicum NC-3b was lower than that of C. glutamicum NC-3, with respective succinic acid yields of 1.30 ± 0.1 and 0.95 ± 0.08 mol/mol. The C. glutamicum NC-3b by-product yields were significantly lower than those of the wild type, with pyruvic acid yields of 0.11 ± 0.02 and 0.18 ± 0.01 mol/mol, and acetic acid yields of 0.24 ± 0.04 and 0.40 ± 0.08 mol/mol, respectively (Table 4).

Fig. 2

Growth (a) and glucose consumption (b) of C. glutamicum NC-3, C. glutamicum NC-3b, C. glutamicum NC-3b-1, C. glutamicum NC-3b-2 and C. glutamicum NC-3b-3 in minimal medium with 110 mM glucose as the sole carbon source under aerobic condition. Filled squares, C. glutamicum NC-3; filled circles, C. glutamicum NC-3b; filled triangles, C. glutamicum NC-3b-1; filled stars, C. glutamicum NC-3b-2; filled diamonds, C. glutamicum NC-3b-3. Arithmetic means and absolute errors from at least three independent cultivation are given

 

Succinic acid production of C. glutamicum NC-3 and NC-3b and derivatives from 183 mM glucose

Strain	Anaerobic biomass (g L−1)	Succinic acid (mM g−1 DCW)	Succinic acid yield (mol/mol)	Pyruvate caid yield (mol/mol)	Acetate acid (mol/mol)
C.glutamicum NC-3	8.19 ± 0.2	15.73 ± 0.4	0.95 ± 0.03	0.18 ± 0.01	0.40 ± 0.08
C.glutamicum NC-3b	3.78 ± 0.1	11.68 ± 0.3	1.30 ± 0.15	0.11 ± 0.02	0.24 ± 0.04

Results of sealed bottles experiments using 183 mM glucose in anaerobic medium after 18 h for incubation, and the values shown are averages of triplicate experiments ± standard deviations

Derepression of the transcriptional regulator iolR has positive effects on PTS-independent glucose transport

To restore the glucose uptake rate in the ptsG mutant C. glutamicum NC-3b, an iolR-deficient mutant, C. glutamicum NC-3b-1, was constructed by homologous recombination. The growth rate of C. glutamicum NC-3b-1 (∆iolR) in minimal medium with 110 mM glucose as the sole carbon source was higher than that of C. glutamicum NC-3b (∆ptsG), with observed growth rates of 0.44 ± 0.07 and 0.17 ± 0.03 h⁻¹, respectively (Fig. 2a). The glucose consumption rate was also higher than that of C. glutamicum NC-3b (∆ptsG), with rates of 2.68 and 1.49 ± 0.3 mM/h, respectively (Fig. 2b). Accordingly, we used RT-PCR analysis to investigate the transcript levels of the PTS-independent glucose transport genes iolT1, iolT2, ppgk, and glk in C. glutamicum NC-3b (∆ptsG). As shown in Fig. 3, inactivation of iolR resulted in a 5.11 ± 0.22-fold increase in iolT1 transcript, and a 3.25 ± 0.29-fold increase in the transcription of iolT2, while ppgk and glk transcript levels were increased by 2.23 ± 0.29- and 2.14 ± 0.23-fold, respectively.

Fig. 3

Relative mRNA levels of the PTS-independent glucose PTS-independent glucose transport genes in C. glutamicum NC-3b-1 carrying the mutation of iolR. Total RNAs were prepared from cells grown to the exponential phase in A medium. Aliquots of RNAs were reverse transcribed and subjected to qPCR. The transcript levels of iolT1 (white bars), iolT2 (black bars), ppgk (hatched bars), and glk (dotted bars) were standardized to the constitutive expression level of 16S rRNA. The transcript levels in C. glutamicum NC-3b were set to 1.0. Data represent mean values from three independent cultures, and the standard deviation from the mean is indicated as error bars

Expression of either iolT1 or ppgk partly restores glucose uptake in C. glutamicum NC-3b

To compare the effects of iolT1 and ppgk overexpression with the effects of iolR deletion in the PTS inactivation mutant (NC-3b), C. glutamicum NC-3b-iolT1 and C. glutamicum NC-3b-iolT1(pXMJ19-ppgk) were constructed and analyzed for their glucose uptake capabilities. The growth of C. glutamicum NC-3b was slightly increased following expression of iolT1, with growth rates of 0.28 ± 0.07 and 0.17 ± 0.03 h⁻¹ for strains NC-3b-iolT1 and NC-3b, respectively. The glucose consumption rate of strain NC-3b-iolT1 was also higher than that of C. glutamicum NC-3b (2.17 ± 0.5 and 1.49 ± 0.3 mM/h, respectively). Overexpression of ppgk in C. glutamicum NC-3b-iolT1 further increased the growth rate and glucose uptake of the mutant to levels closer to those of wild-type strain C. glutamicum NC-3 (growth rate, 0.5 ± 0.02 h⁻¹; glucose consumption rate, 3.94 ± 0.2 mM/h), with strain NC-3b-iolT1(pXMJ19-ppgk) demonstrating a growth rate of 0.36 ± 0.03 h⁻¹ and a glucose consumption rate of 2.76 ± 0.3 mM/h. However, these results showed that expression of iolT1 and ppgk only partly restored glucose uptake in C. glutamicum NC-3b; thus, a combination of the two strategies was examined in an attempt to completely restore glucose utilization in the PTS-deficient strain.

Expression of iolT1 and ppgk completely restores glucose utilization in C. glutamicum NC-3b-1

To test if the combined deletion of iolR and overexpression of the PTS-independent glucose transport genes could increase succinic acid production in the absence of PTS, C. glutamicum NC-3b-3 (∆iolR-iolT1-ppgk) was constructed and its succinic acid production was analyzed. Notably, the growth of C. glutamicum strains NC-3b-2 (∆iolR-iolT1) and NC-3b-3 (∆iolR-iolT1-ppgk) was much faster than that of C. glutamicum NC-3b, but still slightly slower than that of C. glutamicum NC-3 (Fig. 2a). The glucose consumption rate of C. glutamicum strains NC-3b-2 and C. glutamicum NC-3b-3 was also much faster than that of C. glutamicum NC-3b, and was equal to that of wild-type strain C. glutamicum NC-3 (Fig. 2b). Succinic acid accumulation in media containing 183 mM glucose was enhanced by 25 % for these strains, with yields of 191 ± 3.65 mM for C. glutamicum NC-3b-3 and 153 ± 3.43 mM for C. glutamicum NC-3 (Fig. 4a). The succinic acid yield of C. glutamicum NC-3b-3 (1.43 mol/mol) was 50.5 % higher than that of C. glutamicum NC-3 (0.95 mol/mol) (Fig. 4b), while the formation of by-products by strain NC-3b-3 was dramatically lower than that of C. glutamicum, the wild-type strain (pyruvic acid production, 5.82 and 20.3 mM; acetic acid production, 30.2 and 45.5 mM, respectively).

Fig. 4

Succinic acid production, glucose consumption (a) and succinic acid yield (b) of C. glutamicum NC-3, C. glutamicum NC-3b, C. glutamicum NC-3b-1, C. glutamicum NC-3b-2 and C. glutamicum NC-3b-3 in sealed bottles using anaerobic medium supplemented with 183 mM glucose. Black bars glucose consumption, light gray bars succinic acid production, sparse line bars succinic acid yield. Arithmetic means and absolute errors from at least three independent cultivation are given

Fed-batch succinic acid production using C. glutamicum strains NC-3 and NC-3b-3

To further investigate whether the production of succinic acid could be improved by increasing the available PEP, two independent fed-batch fermentations were carried out under identical conditions. The two experiments had comparable final succinic acid yields, succinic acid production rates, and formation of by-products. The fermentation with C. glutamicum NC-3 is shown in Fig. 5a, and described in detail below. Initially, the anaerobic fermentation medium contained 367 mM glucose and 300 mM NaHCO₃. At 8 and 44 h post-inoculation, NaHCO₃ was added to the medium (stock concentrations of 250 and 100 mM, respectively). At 14 and 44 h, glucose was added to a concentration of 90 mM. Fermentation using C. glutamicum NC-3b-3 is shown in Fig. 5b. Initially, the anaerobic fermentation medium contained 367 mM glucose and 300 mM NaHCO₃. At 8, 22, and 38 h post-inoculation, NaHCO₃ was added to 100 mM, and at 8 and 22 h post-inoculation, glucose was added to 90 mM.

Fig. 5

Representative anaerobic fed-batch fermentation with C. glutamicum NC-3 (a) and C. glutamicum NC-3b-3 (b) showing succinic acid production during the utilization of glucose and formation of the by-products. a Succinic acid (open circles), glucose consumption (open triangles), acetate acid (open squares), pyruvate acid (open diamonds), succinic acid yield (open stars). b Succinic acid (filled circles), glucose consumption (filled triangles), acetate acid (filled squares), pyruvate acid (filled diamonds), succinic acid yield (filled stars). The experiment was performed in a 3-L fermentation bioreactor system, and the cells, pregrown aerobically, were resuspended in 1-L anaerobic fermentation medium. Two independent fermentations were performed, with both showing comparable results with respect to product yield and production

As shown in Fig. 5, the final succinic acid concentration of the C. glutamicum NC-3b-3 culture at 48 h was 11.6 % greater than that of C. glutamicum NC-3 (769 and 689 mM, respectively). Over the course of the 48-h fermentation, 522.2 and 616.7 mM glucose had been consumed (Fig. 5). Additionally, a higher succinic acid yield was obtained from C. glutamicum NC-3b-3 (1.47 mol/mol glucose) than from wild-type C. glutamicum NC-3 (1.11 mol/mol glucose). By-product formation in the C. glutamicum NC-3b-3 fermentation was dramatically lower than that of C. glutamicum NC-3 (pyruvic acid production, 44.3 and 56.1 mM; acetic acid production, 108.3 and 250.9 mM, respectively). Thus, the deletion of iolR and combined overexpression of iolT1 with ppgk in the PTS-deficient C. glutamicum NC-3b strain enabled 11.6 % greater succinic acid production and a 32.4 % higher succinic acid yield than that of C. glutamicum NC-3 in a 3-l fermentation bioreactor system.

Discussion

Previous studies have shown that C. glutamicum strains with a PTS⁻ glucose⁺ phenotype have characteristics that are useful for biotechnological applications [4]. Mutation of ptsG increased production of succinic acid from the fermentation of glucose by C. glutamicum, and decreased conversion of PEP to pyruvate reduced the formation of by-products derived from pyruvate [3]. In C. glutamicum, each mole of glucose taken up by the PTS requires the consumption of 1 mol of PEP, and PTS inactivation has a strong effect on the metabolic flux distribution throughout the central metabolic network [37]. In a PTS mutant, the pentose phosphate pathway and tricarboxylic acid cycle fluxes were enhanced, and there was an increase in the exchange rate between PEP and oxaloacetic acid, involving Ppc and Pck, compared with the wild-type strain [37]. This suggests that available PEP relative to pyruvate in the PTS mutant strain contributes to a better balance in the supply of carbon from central metabolism into the succinic acid-biosynthetic pathway through the two anaplerotic reactions involving pyruvate carboxylase and PEP carboxylase.

Although the PTS is the major route of glucose uptake, a lower glucose utilization rate and slower growth were observed in the ptsG-deficient C. glutamicum strain (Fig. 2). This study has developed a strategy for engineering C. glutamicum to use a PTS-independent glucose uptake route instead of the original PTS pathway. This strategy contains just two steps: (1) disruption of iolR, and (2) overexpression of the PTS-independent glucose utilization genes. RT-PCR was applied to detect the effects of iolR on the expression of iolT1, iolT2, ppgk, and glk. Deletion of iolR partly restored the glucose utilization and growth rates of C. glutamicum NC-3b (Fig. 2), and increased the transcript levels of iolT1, iolT2, ppgk, and glk. The inactivation of iolR also relieved repression of iolT, most likely acting on the promoter of this gene. Introduction of a single-base deletion (320delA) into the transcriptional regulator gene iolR also allowed a ptsG-deficient C. glutamicum strain to utilize glucose through the PTS-independent glucose transport system [11]. The current study is the first to demonstrate that transcription of ppgk and glk is repressed by the transcriptional regulator iolR. It is probable that iolR has negative effects on the promoter operator region of glucokinase genes.

To more precisely analyze the effects of deleting iolR and independently overexpressing iolT1 and ppgk, iolT1 was overexpressed in C. glutamicum NC-3b to rule out the effect of iolR. iolT1 and ppgk were chosen because their overexpression in a ∆hpr strain had a greater positive effect on biomass formation from a medium containing 10 mM inositol and 200 mM glucose than the overexpression of iolT2 and glk, as well as affecting the growth of the GSM strains [19]. The results showed that overexpression of iolT1 and ppgk was necessary to restore glucose transport; however, growth and glucose uptake were partly restored by overexpression of iolT1 and ppgk on their own. Although Lindner et al. improved the yield of l-lysine by overexpressing ppgK with either iolT1 or iolT2 in a PTS-deficient strain, growth needed to be prolonged, which would decrease the productivity of the reaction [19]. In the current study, the engineered strain C. glutamicum NC-3b-3 showed glucose uptake and growth rates comparable to those of the wild-type strain C. glutamicum NC-3. Batch fermentation of C. glutamicum NC-3 and C. glutamicum NC-3b-3 revealed that strains engineered for PTS-independent glucose uptake showed improved production of succinic acid because of reduced pyruvate and increased PEP availability in the cells.

By decoupling glucose transport from PEP consumption, the metabolic availability of this intermediate molecule was significantly higher than with a PTS⁺ strain. In C. glutamicum, PEP also participates as a substrate in the reaction catalyzed by pyruvate kinase (pyk). Future work will include the deletion of pyk to further increase PEP availability. It can also be expected that the production of other chemicals, such as phenylalanine, shikimate, and dehydroshikimate, which are synthesized using PEP as a precursor, will be improved in a PTS⁻ glucose⁺ strain.

Acknowledgments

This work was supported by the 973 Program of China (Grant no. 2011CB707405).

Conflict of interest

The authors declare that they have no conflict of interest.