Production of γ-Aminobutyrate (GABA) in Recombinant Corynebacterium glutamicum by Expression of Glutamate Decarboxylase Active at Neutral pH.

γ-Aminobutyrate (GABA) is an important chemical by itself and can be further used for the production of monomer used for the synthesis of biodegradable polyamides. Until now, GABA production usingCorynebacterium glutamicum harboring glutamate decarboxylases (GADs) has been limited due to the discrepancy between optimal pH for GAD activity (pH 4.0) and cell growth (pH 7.0). In this study, we developed recombinant C. glutamicum strains expressing mutated GAD from Escherichia coli (EcGADmut) and GADs from Lactococcus lactis CICC20209 (LlGAD) and Lactobacillus senmaizukei (LsGAD), all of which showed enhanced pH stability and adaptability at a pH of approximately 7.0. In shake flask cultivations, the GABA productions of C. glutamicum H36EcGADmut, C. glutamicum H36LsGAD, and C. glutamicum H36LlGAD were examined at pH 5.0, 6.0, and 7.0, respectively. Finally, C. glutamicum H36EcGADmut (40.3 and 39.3 g L-1), H36LlGAD (42.5 and 41.1 g L-1), and H36LsGAD (41.6 and 40.2 g L-1) produced improved GABA titers and yields in batch fermentation at pH 6.0 and pH 7.0, respectively, from 100 g L-1 glucose. The recombinant strains developed in this study could be used for the establishment of sustainable direct fermentative GABA production from renewable resources under mild culture conditions, thus increasing the availability of various GADs.

Introduction

γ-Aminobutyrate (GABA) is a non-proteinogenic amino acid that functions as an inhibitory neurotransmitter in animals and mediates the acid stress response in plants.[1] Additionally, GABA has widely been used in the pharmaceutical industry as the active compound in tranquilizers, analgesics, diuretics, and antianxiety medicines.[2−4] Particularly, the production of GABA using bacteria has been assessed as an ideal strategy for functional food industrial applications.[5−7] Furthermore, current environmental crises such as global warming and plastic waste accumulation have led to the exploration of new applications of GABA, a carbon-neutral chemical that can be used for further synthesis of 2-pyrrolidone, an important C4 platform chemical and monomer of bio-based polyamides such as polyamide 4.[8,9] GABA biosynthesis is initiated from l-glutamate by glutamate decarboxylases (GADs; EC. 4.1.1.15), a pyridoxal 5′-phosphate (PLP)-dependent enzyme, via α-decarboxylation. Numerous studies have assessed the synthesis of GABA with GADs through enzymatic reactions or whole-cell reactions using Escherichia coli(9−13) and lactic acid bacteria (LAB)[14−16] as biocatalysts (Table ). In enzymatic GABA production, immobilization of E. coli GAD could produce 223.8 g L–1 of GABA in a 1 L reactor within 100 min, retaining more than half of its activity after 10 sequential uses.[12] Although enzymatic reactions are among the most powerful strategies for GABA production, whole-cell reactions are considered a more cost-efficient alternative because they can lower the total production cost related to the removal of GAD purification and immobilization steps. Whole-cell reaction, in which an endogenous gadA is overexpressed in E. coli BL21(DE3), resulted in the production of 280–300 g L–1 of GABA with an 89.5% yield via direct l-glutamate biotransformation.[10] It was also reported that whole-cell conversion using the engineered E. coli XL1-blue strain expressing gadB from Lactococcus lactis subsp. lactis Il1403 could produce 94.8 g L–1 of GABA from 200 g L–1 monosodium glutamate (MSG).[9] Although high titers of GABA could be obtained from the conversion of l-glutamate or MSG through enzymatic or whole-cell biocatalysis reactions, there is a growing need for the establishment of more sustainable and cost-effective alternative microbial processes, the direct fermentative production of GABA from renewable carbon sources such as glucose.[17,18]

Table 1

Summary of Recently Reported GABA Production in Recombinant Microbial Hostsa

strain	features	scale	carbon source	titer (g L–1)	productivity (g L–1 h–1)	yield (g g–1)	ref
E. coli BL21(DE3)	PT7: GAD from S. cerevisiae	whole-cell	l-glutamate	245	5.83*	0.99 (mol mol–1)	(52)
E. coli BL21(DE3)	PT7: GAD from L. lactis FJNUGA01	whole-cell	l-glutamate	204.1	34	0.99 (mol mol–1)	(53)
E. coli BW25113	ΔgadCΔgadAB, ParaBAD: gadB(M4)- groES-groEL, gadB mutant from L. lactis IL1403	whole-cell	l-glutamate	308.26	44.04	0.996 (mol mol–1)	(54)
E. coli EDK11	PLetO: gadB, gadC, gabT, and gltA	shake flask	glucose	1.2	0.05*		(55)
E. coli Nissle 1917 pMT1-G/pMT2- R/EcNP	Ptrc: gadB	whole-cell	MSG	17.9			(56)
E. coliBL21 (DE3)/GADZ11	PT7: GAD from Bacillus spp.	whole-cell	l-glutamate	1640	96.4*	0.985 (mol mol–1)	(44)
Corynebacterium glutamicum ATCC 13032	ΔgabT, ΔgabP; Ptac: gadB2 from L. brevis ATCC 367	shake flask	glucose	28.7	0.3		(57)
C. glutamicum SH	Ptac: gadB1-gadB2-ppc	fed-batch	glucose	26.3	0.365		(58)
C. glutamicum H36GD1852	PH36: gadBmutxylAB from E. coli	fed-batch	empty fruit bunch (EFB) solution	35.47	0.68		(22)
C. glutamicum SH	PtacM: R4a-gabB2B1mut from L. brevis	shake flask	glucose	26.5	0.442*	0.269*	(59)
C. glutamicum H36EcGADmut	PH36: GAD from E. coli	batch	glucose	40.3	1.12	0.40	this study
C. glutamicum H36LlGAD	PH36: GAD from L. lactis CICC20209	batch	glucose	42.5	1.18	0.43	this study
C. glutamicum H36LsGAD	PH36: GAD from L. senmaizukei	batch	glucose	41.6	1.16	0.42	this study

* indicates calculated data based on reported studies.

To establish an efficient system for one-step fermentative GABA production from glucose, a sufficient amount of intracellular l-glutamate should be ensured because it is a key precursor for GABA synthesis. Among various host strains, C. glutamicum is considered the ideal host since it is already known as a robust industrial commercial l-glutamate-producing strain. Moreover, C. glutamicum is a Gram-positive and nonpathogenic bacterium that is generally recognized as safe (GRAS) serving benefits on target chemical production.[19,20] In a previous study, an engineered C. glutamicum 13 032 strain expressing the EcGADmut produced 38.6 g L–1 of GABA from glucose during fed-batch fermentation at pH 6.0.[21] Recently, the heterologous expression of EcGADmut along with E. coli xylose isomerase (xylA) and xylulose kinase (xylB) in an l-glutamate overproducing C. glutamicum KCTC 1852 strain resulted in the production of 35.5 g L–1 of GABA from an empty fruit bunch biosugar solution (a mixture of glucose and xylose) during batch fermentation at pH 6.0.[22]

One of the main challenges of direct GABA synthesis from glucose is that there is a trade-off between GAD activity and cell growth. During fermentation, GABA production is limited due to the significant difference between the optimal pH for GAD activity (pH 4.0) and the optimal pH for cell growth (pH 7.0). Mostly, GADs used for enzymatic conversion or whole-cell conversion of MSG into GABA have shown specific activity under acidic conditions. Particularly, GADs from Lactobacillus zymae,(23)E. coli,(24)Lactobacillus sakei A156,[25] and L. brevis HYE1[26] were reported to be only active at pH 4.0–5.0. Given that the optimal pH for cell growth is higher than that for optimal GAD activity, these GADs cannot be used in direct fermentative GABA production from renewable resources, as they are unable to support the efficient production of GABA during the exponential growth phase of the host strain. Moreover, GAD is active during the stationary phase when both pH and cell growth have begun to significantly decrease, which ultimately results in retarded productivity of GABA requiring long cultivation times of over 120 h.[24,27]

Therefore, the employment of various GADs exhibiting enhanced activity within a wider pH range (pH 5.0–7.0) has been attempted to produce high GABA yields in a growth-associated manner. To this end, many studies have contributed to the screening of GADs active in a broad pH range using two primary strategies: (i) site-directed mutagenesis on previously reported GADs and (ii) discovery of novel GADs. For example, (i) a site-specific mutant of E. coli GAD (EcGADmut; E89Q/Δ452–466) was developed based on the structural changes observed during pH shifts, which led to enhanced activity within a wider pH range[21,22,28] and (ii) novel GADs active at pH 7.0 were elucidated in LAB via genome mining.[29]

In this study, recombinant C. glutamicum strains capable of producing high amounts of GABA under near-neutral pH conditions were developed by employing several GADs from E. coli (EcGADmut), L. lactis (LlGAD), and L. senmaizukei (LsGAD), which had been reported to be active at a broad pH range, as demonstrated by in vitro assays.[21,29] Therefore, our study sought to extend the applicability of GADs active at a pH close to 7.0 and to secure a diverse gene pool of GADs with a broad pH range during in vivo experiments. First, the GABA biosynthesis pathway was established in recombinant C. glutamicum KCTC 1852 strains via the construction of plasmids for heterologous expression of three GADs under the control of two different strong synthetic promoters, H30 and H36 (Table ).[29,30] Then, further control of pH in flask cultivation by the intermittent addition of an acid or a base was implemented to investigate the optimal pH for GABA production. Finally, batch fermentation of the engineered strains with heterologous expression of EcGADmut, LsGAD, and LlGAD under the H36 promoter was conducted by maintaining the culture pH at 6.0 and 7.0 (Figure ). Taken together, we have, here, demonstrated the extended applicability of GADs in neutral pH conditions.

Table 2

Strains and Plasmids Used in This Study

strains and plasmids	relevant characteristics	reference or source
Strains
E. coli XL1-Blue	recA1 endA1 gyrA96 thi-1 hsdR17 supE44 relA1 lac [FA1proAB lacIqZΔM15 Tn10 (TetR)]	stratagene
C. glutamicum KCTC 1852	l-glutamate producers	KCTC
C. glutamicum 1852 H36EcGADmut	C. glutamicum KCTC 1852 harboring pCES208H36EcGADmut	(22)
C. glutamicum 1852 H30LsGAD	C. glutamicum KCTC 1852 harboring pCES208H30LsGAD	this study
C. glutamicum 1852 H36LsGAD	C. glutamicum KCTC 1852 harboring pCES208H36LsGAD	this study
C. glutamicum 1852 H30LlGAD	C. glutamicum KCTC 1852 harboring pCES208H30LlGAD	this study
C. glutamicum 1852 H36LlGAD	C. glutamicum KCTC 1852 harboring pCES208H36LlGAD	this study
Plasmids
pCES208H30GFP	pCES208 derivative; H30 promoter, eGFP, Kmr	(30)
pCES208H36GFP	pCES208 derivative; H36 promoter, eGFP, Kmr	(30)
pCES208H36EcGADmut	pCES208H36GFP derivative; H36 promoter, GAD mutant from E. coli (Glu89Gln/Δ452–466), Kmr	(21)
pCES208H30LsGAD	pCES208H30GFP derivative; H30 promoter, GAD from L. senmaizukei, Kmr	this study
pCES208H36LsGAD	pCES208H36GFP derivative; H36 promoter, GAD from L. senmaizukei, Kmr	this study
pCES208H30LlGAD	pCES208H30GFP derivative; H30 promoter, GAD from L. lactis CICC20209, Kmr	this study
pCES208H36LlGAD	pCES208H36GFP derivative; H36 promoter, GAD from L. lactis CICC20209, Kmr	this study

Figure 1

Diagram of the GABA biosynthetic pathway in C. glutamicum and proposed production strategies for GABA using various GADs through flask and batch fermentation. The abbreviation shown is AKG, α-ketoglutarate.

Materials and Methods

Bacterial Strains and Plasmids

Table summarizes all of the bacterial strains and plasmids used in this study. E. coli XL1-Blue (Stratagene, La Jolla, CA) was used as the main host strain for gene cloning and plasmid maintenance. The l-glutamate overproducing C. glutamicum KCTC 1852 strain was purchased from the Korean Collection for Type Cultures (KCTC, Korea). The pCES208H30GFP, pCES208H36GFP, and pCES208H36EcGADmut plasmids containing mutated GAD from E. coli (EcGADmut) were constructed, as previously described.[21,30]

Plasmid Construction

All DNA manipulations were conducted following standard procedures.[31] A C1000 Thermal Cycler (Bio-Rad, Hercules, CA) was used for polymerase chain reactions (PCRs). The primers used in this study were synthesized by Bioneer (Daejeon, Korea) (Table ). For the expression of GAD from L. lactis CICC20209 and L. senmaizukei, each gene was codon-optimized and synthesized by Cure bio (Seoul, Korea) based on sequence data from a previous report.[30] The DNA fragments were digested with BamHI/NotI and ligated into the BamHI/NotI sites of pCES208H30GFP or pCES208H36GFP to make pCES208H30LsGAD and pCES208H30LlGAD or pCES208H36LsGAD and pCES208H30LlGAD, respectively (Table ), in which the gfp gene was replaced with the corresponding GAD gene.

Table 3

List of Primers Used in PCR Experiments

glutamate decarboxylase	primer
EcGADmut-F	GGATCCATGGATAAGAAGCAAGTAACGG
EcGADmut-R	GCGGCCGCTTAGGTATGTTTAAAGCTGTTCT
LsGAD-F	GGATCCATGAGTAAAAACGATCAGGAAACGC
LsGAD-R	GCGGCCGCCTAACGAACAGTCGTCTTGTTCTTG
LlGAD-F	GGATCCATGTTATACGGAAAAGAAAATCGCG
LlGAD-R	GCGGCCGCTTAGTGAGTAAATCCATATGTTTTATTTTC

Culture Conditions

E. coli was grown in the Luria–Bertani (LB) medium containing 10 g L–1 tryptone, 5 g L–1 yeast extract, and 5 g L–1 NaCl at 37 °C, 200 rpm. Shake flask culture for GABA production using the recombinant C. glutamicum 1852 strains was conducted using 250 mL baffled flasks with 50 mL of the GP1 medium. These flasks were incubated at 30 °C and 250 rpm. The GP1 medium contained 50 g L–1 glucose, 50 g L–1 (NH4)2SO4, 1 g L–1 K2HPO4, 3 g L–1 urea, 0.4 g L–1 MgSO4·7H2O, 50 g L–1 peptone, 0.01 g L–1 FeSO4, and 0.01 g L–1 MnSO4·5H2O. Glucose was separately autoclaved and added to the culture medium. Each flask culture was supplemented with 25 μg mL–1 kanamycin (Km), 0.1 mM pyridoxal 5-phosphate hydrate (PLP), 50 μg L–1 biotin, and 200 μg L–1 thiamine. These supplements were prepared as stock solutions and sterilized using 0.22 μm filter membranes prior to addition to the culture medium. To minimize pH fluctuations during cultivation, 10 g L–1 of calcium carbonate (CaCO3) was added to the culture medium. For pH-controlled flask culture, the pH of flask cultures was checked every 24 h prior to dropwise addition of 1 M HCl or 1 M NaOH to adjust pH 5.0, 6.0, or 7.0.

For batch fermentation, 50 mL of seed culture was prepared by making 4 mL overnight cultures of recombinant C. glutamicum 1852 strains for inoculation of 50 mL of the GP1 medium in a 500 mL baffled flask. These were incubated at 30 °C and 250 rpm. Batch fermentations were carried out at 30 °C and 600 rpm using a 2.5 L jar fermenter (BioCNS, Korea), which contains a total volume of 500 mL of the CG100 medium. The CG100 medium is composed of 100 g L–1 of glucose, 30 g L–1 of yeast extract, 30 g L–1 of (NH4)2SO4·7H2O, 0.5 g L–1 of KH2PO4, 0.5 g L–1 of MgSO4·7H2O, 0.01 g L–1 of MnSO4·H2O, 0.01 g L–1 of FeSO4·7H2O, 0.5 mg of biotin, and 0.3 mg of thiamine-HCl. Kanamycin and PLP were added to the culture medium at 25 μg mL–1 and 0.1 mM, respectively. Two different pH values (6.0 and 7.0) were examined for GABA production and maintained via automatic addition of 28% (v/v) NH4OH. Foam formation was suppressed by adding Antifoam 204 (Sigma-Aldrich, St. Louis, MO).

Analytical Procedures

Cell growth was monitored by measuring absorbance at 600 nm (OD600) using a UV-1700 spectrophotometer (Shimadzu, Japan). Glucose concentration was analyzed by high-performance liquid chromatography [HPLC, Infinity 1260 (Agilent Technologies, Santa Clara, CA)] equipped with a refractive index detector (RID) and a Bio-Rad Aminex HPX-87H column. Elution was performed with 5 mM of H2SO4 at a 0.8 mL min–1 flow rate at 50 °C. The concentrations of l-glutamate and GABA were analyzed by HPLC equipped with a variable wavelength detector (VWD) and a C18 column (RStech, Daejeon, Korea) according to the DEEMM method.[32] Elution was performed with mobile phase A (acetonitrile) and mobile phase B (25 mM sodium acetate buffer, pH 4.8).

Results and Discussion

Construction of the GABA Biosynthesis Pathway in C. glutamicum KCTC 1852

Most bacterial GADs are only active under acidic conditions, especially at 4.0, and GAD activity is strongly diminished as the pH of the culture medium increases.[23,25,26] In contrast, C. glutamicum is widely regarded as a moderately alkali-tolerant bacterium with an optimal pH range between 7.0 and 8.5.[33,34] Therefore, the efficient conversion of l-glutamate into GABA largely relies on maintaining an optimal balance between maximum bacterial growth and GAD activity.

To identify GADs that efficiently produce GABA under a wider pH range, several studies have reported the production of GABA with engineered or novel GADs that possess a broad active pH range from 5.0 to 7.0.[21,28,29,35] Recently, there has been an attempt to search for novel GADs from GRAS probiotic bacteria including LAB based on genome mining and sequence alignment analysis, and consequently, 16 putative GADs were screened.[29] Among them, LsGAD and LlGAD were found to possess a wider optimal pH range and a more durable enzymatic stability than GAD from L. brevis, which was used as a probe.[29] Therefore, in this study, we hereby examined three different GADs (EcGADmut, LsGAD, and LlGAD) that were previously reported to exhibit superior activity under a wide pH range (pH 5.0–7.0) for GABA production from glucose at near-neutral pH in l-glutamate-overproducing C. glutamicum KCTC 1852 strains. To this end, we evaluated five different recombinant C. glutamicum KCTC 1852 strains expressing EcGADmut, LsGAD, or LlGAD under the control of a synthetic H30 or H36 promoter: (i) H36EcGADmut strain (C. glutamicum KCTC 1852 harboring pCES208H36EcGADmut), (ii) H30LsGAD strain (C. glutamicum KCTC 1852 harboring pCES208H30LsGAD), (iii) H36LsGAD strain (C. glutamicum KCTC 1852 harboring pCES208H36LsGAD), (iv) H30LlGAD strain (C. glutamicum KCTC 1852 harboring pCES208H30LlGAD), and (v) H36LlGAD strain (C. glutamicum KCTC 1852 harboring pCES208H36LlGAD) (Figure ). Strong synthetic H30 and H36 promoters have been successfully implemented for the construction of recombinant C. glutamicum strains that can produce various chemicals such as cadaverine, glutaric acid, GABA, and 5-HV.[22,36−38] Only the H36 promoter was examined for the expression of EcGADmut in C. glutamicum KCTC 1852, as we had previously reported that EcGADmut expression under the H36 promoter in C. glutamicum KCTC 1852 supported higher GABA production from an empty fruit bunch biosugar solution.[22]

Comparison of GADs in Recombinant C. glutamicum during pH-Controlled Flask Cultivation

Flask cultivation using each recombinant C. glutamicum strain was conducted under three different pH conditions (pH 5.0, 6.0, and 7.0) to determine which GAD and promoter combinations produced the highest titers of GABA at near-neutral pH. Even though the initial pH values of the flask cultures were established according to previous studies,[21,22] cell growth and production of l-glutamate could be adversely affected as the pH in the culture medium decreased over time. Therefore, the pH in the flask was manually controlled during cultivation to maintain a target pH range. To this end, flask cultivation was performed to produce GABA from 50 g L–1 of glucose, and proper amounts of 1 M HCl or 1 M NaOH were added every 24 h to adjust the pH to 5.0, 6.0, and 7.0 (Figure ).

Figure 2

Comparison of GABA production (A) and l-glutamate accumulation (B) during shake flask culture by different recombinant C. glutamicum 1852 strains with different combinations of GADs under the control of the strong synthetic promoters H30 or H36: C. glutamicum H30LsGAD, C. glutamicum H30LlGAD, C. glutamicum H36EcGADmut, C. glutamicum H36LsGAD, and C. glutamicum H36LlGAD. All flask cultures were conducted in triplicate. All measurements are presented as mean ± standard deviation.

At pH 5.0, the final titers of GABA achieved by the five different engineered strains were 7.42 ± 0.08 g L–1 (H30LsGAD), 7.75 ± 0.11 g L–1 (H30LlGAD), 6.95 ± 0.16 g L–1 (H36EcGADmut), 7.69 ± 0.15 g L–1 (H36LsGAD), and 7.99 ± 0.10 g L–1 (H36LlGAD) (Figure A). Consistent with previous studies,[21,22] the recombinant C. glutamicum strains with the H36 promoter exhibited higher GABA production than those with the H30 promoter. However, compared with the recombinant C. glutamicum H36EcGADmut strain, C. glutamicum expressing novel GADs (LsGAD and LlGAD) achieved higher GABA production with lower l-glutamate accumulation.

At pH 6.0, C. glutamicum H30LsGAD and H30LlGAD yielded 8.79 ± 0.02 and 8.50 ± 0.16 g L–1 of GABA, respectively. Similarly, C. glutamicum H36EcGADmut, H36LsGAD, and H36LlGAD strains yielded 8.10 ± 0.15, 9.25 ± 0.16, and 9.22 ± 0.34 g L–1 of GABA, respectively (Figure A). Therefore, in addition to the higher GABA concentrations obtained by the novel GADs compared to EcGADmut, the recombinant strains with the H36 promoter also yielded higher GABA titers than those with the H30 promoter.

In the case of cultivation at pH 7.0, C. glutamicum H30LsGAD and H30LlGAD produced 9.56 ± 0.03 and 9.89 ± 0.14 g L–1 of GABA, respectively, whereas C. glutamicum H36EcGADmut, H36LsGAD, and H36LlGAD could produce 8.05 ± 0.13, 10.20 ± 0.15, and 10.36 ± 0.18 g L–1 of GABA, respectively (Figure A). Similar to the other flask culture conditions, the H36 promoter induced a higher GABA production. Additionally, higher pH conditions resulted in higher GABA concentrations, and therefore, the largest differences in GABA concentration between EcGADmut and novel GADs were observed at pH 7.0.

In accordance with previous studies,[21,22,29] expressing GADs from LAB and engineered GADs from E. coli in host strains that are reported to maintain their activities even at near-neutral pH could efficiently produce GABA at pH 7.0. The produced GABA concentrations at various pH conditions obtained by C. glutamicum H36LsGAD (pH 5.0, 7.69 ± 0.15 g L–1; pH 6.0, 9.25 ± 0.16 g L–1; pH 7.0, 10.20 ± 0.15 g L–1) and C. glutamicum H36LlGAD (pH 5.0, 7.99 ± 0.10 g L–1; pH 6.0, 9.22 ± 0.34 g L–1; pH 7.0, 10.36 ± 0.18 g L–1) were marginally improved compared to those obtained from C. glutamicum H36EcGADmut (pH 5.0, 6.95 ± 0.16 g L–1; pH 6.0, 8.10 ± 0.15 g L–1; pH 7.0, 8.05 ± 0.13 g L–1). This demonstrated that LsGAD and LlGAD have a slightly stronger in vivo catalytic activity at pH 5.0–7.0. In addition, recombinant strains with the H36 promoter produced consistently higher concentrations of GABA than those with the H30 promoter. The strains with the H36 promoter harboring LsGAD produced 3% (pH 5.0), 5% (pH 6.0), and 6% (pH 7.0) higher GABA yields. Moreover, the strains harboring LlGAD under the control of the H36 promoter resulted in 3% (pH 5.0), 8% (pH 6.0), and 5% (pH 7.0) enhanced production of GABA than those of the H30 strains (Figure A). This may be because a stronger promoter could support the high-level expression of related enzymes along with the overproduction of target compounds.[21,30,39,40] Furthermore, the C. glutamicum H36EcGADmut, H36LsGAD, and H36LlGAD strains exhibited lower accumulations of unconverted l-glutamate at pH 7.0, with final concentrations of 1.21 ± 0.09, 1.16 ± 0.10, and 1.00 ± 0.19 g L–1 of l-glutamate, respectively, while C. glutamicum H30LsGAD and H30LlGAD strains have also accumulated less l-glutamate at pH 7.0 (1.10 ± 0.16 and 1.06 ± 0.15 g L–1, respectively) than at pH 5.0 (Figure B).

Batch Fermentations of Recombinant C. glutamicum for the Enhanced Production of GABA under Near-Neutral pH

Even though pH adjustment was conducted during flask cultivation, strict regulation of fermentation environments by tightly adjusting pH is essential to achieve high-level production of the desired products. Previously, GABA was produced using GAD that was only active at low pH, thus, the following fermentation strategies were carried out. For example, two-stage cultivation is employed for the production of target compounds in low pH conditions, which is divided into two parts: (i) the early cell growth phase, where cells grow well using a carbon source but without much GABA production and (ii) the GABA production phase, where the pH decreases to approximately 4.0, after which GAD becomes active and substantial amounts of GABA production are initiated.[41,42] Nevertheless, it is more beneficial to proceed with the fermentation with GADs, which show activity in almost neutral pH, as they can render high GABA yields in short periods. Therefore, we conducted batch fermentation of recombinant C. glutamicum strains to obtain high GABA yields in a 2.5 L lab-scale bioreactor maintained at pH 6.0 and 7.0 without compromising production efficiency. This was achieved by utilizing three GADs exhibiting comparable activities at near-neutral pH conditions. Since recombinant strains expressing EcGADmut, LsGAD, and LlGAD under the control of the H36 promoter have supported higher productions of GABA in flask cultures, batch fermentation of C. glutamicum H36EcGADmut, H36LsGAD, and H36LlGAD was conducted herein. Figure illustrates the time profiles of each batch fermentation of the final C. glutamicum strains at different pH values.

Figure 3

Time profiles of GABA, l-glutamate, glucose concentration, and cell density during batch fermentation in recombinant C. glutamicum H36EcGADmut at pH 6.0 (A), recombinant C. glutamicum H36EcGADmut at pH 7.0 (B), recombinant C. glutamicum H36LsGAD at pH 6.0 (C), recombinant C. glutamicum H36LsGAD at pH 7.0 (D), recombinant C. glutamicum H36LlGAD at pH 6.0 (E), and recombinant C. glutamicum H36LlGAD at pH 7.0 (F). All batch fermentations were done in duplicates and the figure shows a representative fermentation profile.

The final GABA titers obtained from batch fermentation at pH 6.0 were 40.3 g L–1 (H36EcGADmut), 41.6 g L–1 (H36LsGAD), and 42.5 g L–1 (H36LlGAD), with a final OD600 of 96.8 (H36EcGADmut), 90.8 (H36LlGAD), and 99.8 (H36LsGAD) (Figure A,C,E). After 36 h of cultivation, a GABA yield of 0.40 g g–1 glucose, 0.42 g g–1 glucose, and 0.43 g g–1 glucose and GABA productivities of 1.12, 1.16, and 1.18 g L–1 h–1 were achieved using the C. glutamicum H36EcGADmut, H36LsGAD, and H36LlGAD strains, respectively. l-Glutamate began to accumulate in the early stage of fermentation, and the final concentrations of l-glutamate were 1.34 g L–1 (H36EcGADmut), 1.26 g L–1 (H36LsGAD), and 1.35 g L–1 (H36LlGAD) after 36 h of cultivation.

During batch fermentation at pH 7.0, the GABA titers and final cell concentrations achieved by the engineered C. glutamicum H36EcGADmut, H36LsGAD, and H36LlGAD strains were 39.3 g L–1 with an OD600 of 99.6, 40.2 g L–1 with an OD600 of 102.1 and 41.1 g L–1 with an OD600 of 101.6, respectively (Figure B,D,F). Additionally, the C. glutamicum H36EcGADmut, H36LsGAD, and H36LlGAD strains could produce GABA with a yield of 0.39 g g–1 glucose, 0.39 g g–1 glucose, and 0.41 g g–1 glucose and a productivity of 1.09, 1.12, and 1.14 g L–1 h–1, respectively. The accumulation of l-glutamate was observed at the end of the cultivation with final concentrations of 1.11 g L–1 (H36EcGADmut), 1.19 g L–1 (H36LsGAD), and 1.12 g L–1 (H36LlGAD).

The highest GABA production was achieved in the batch fermentation of the C. glutamicum H36LlGAD strain at pH 6.0. It was first expected to achieve the enhanced production of GABA at pH 7.0 as with the flask cultivation results; however, a slightly higher concentration of GABA with improved productivity was observed at pH 6.0. It was reasoned that slightly acidic environments are still preferred in fermentative GABA production even with the employment of GADs active in a wide pH range, because the conversion of l-glutamate to GABA is a proton-consuming reaction.[43] Furthermore, LAB GADs were reported to favor lower pH conditions (approximately pH 5.0–6.0) in vitro analyses. For example, both GADs retained more than 80% of their relative activity at pH 5.0–6.0 but substantially lost their stability (up to 50%) if they deviate from their optimum pH.[29] Nevertheless, comparable amounts of GABA could be obtained by the recombinant C. glutamicum strains at pH 7.0. Furthermore, approximately 1 g L–1 of unconverted l-glutamate accumulation was observed during the batch fermentations (Figure ). Given that the unconverted l-glutamate concentrations during flask cultivation were over 10% of the final GABA concentrations (Figure ), we concluded that batch fermentation where pH was tightly regulated was effective for GABA production with substantially reduced l-glutamate accumulation. Compared to our previous study,[22] the employment of three different GADs demonstrated a higher production of GABA with a much lower l-glutamate accumulation. In our previous study,[22] batch fermentation of C. glutamicum H36EcGADmut at pH 6.0 and 7.0 rendered 30.8 and 33.8 g L–1 of GABA production, respectively, from 90 g L–1 of glucose and 10 g L–1 of xylose, with over 8 g L–1 of l-glutamate accumulation. These high amounts of l-glutamate accumulated in the culture medium led to lower GABA productivity. In contrast, the present study achieved l-glutamate accumulation as low as 1.0 g L–1, with GABA concentrations over 40 g L–1 during the batch fermentation of the C. glutamicum H36EcGADmut, H36LsGAD, and H36LlGAD strains under both pH 6.0 and pH 7.0. These findings indicated that all three GADs used in this study could efficiently produce GABA from glucose at a near-neutral pH.

More recently, novel GADs from Bacillus spp. were identified, and engineered E. coli expressing GADZ11 could convert l-glutamate into GABA via whole-cell reaction at pH 5.0 with a 98.6% average conversion yield.[44] However, these GADs were not considered feasible for the industrial production of GABA from glucose via direct fermentation, as they were only active under low pH conditions. In another study, the authors found that the PucR-type transcriptional regulator (gadR), GABA permease, GABA transaminase, and succinate semialdehyde dehydrogenase (gadTDP) are involved in GABA degradation pathways,[45] suggesting that the deletion of related enzymes could improve GABA production. Therefore, further enzyme engineering (crystal structure analysis and enzymatic characterization, de novo design of enzymes, and alteration of enzyme–substrate specificity) and evolutionary engineering (adaptive laboratory evolution (ALE)) of LsGAD and LlGAD could provide important insights into the exact mechanisms that drive the structural and kinetic changes of key enzymes in response to pH variations. Moreover, a combination of metabolic engineering and synthetic biology tools has enabled the development of engineered GABA-producing C. glutamicum strains with broader metabolic capacities to enhance the biosynthesis of target products.[46−49]

Conclusions

Herein, we examined the E. coli GAD mutant and novel LAB-derived GADs, which were reported to be active in expanded pH range in vivo analysis for growth-coupled fermentative GABA production in l-glutamate-overproducing C. glutamicum, thus widening the usefulness of GADs active at near-neutral culture conditions. First, we compared the activities of two novel GADs from L. senmaizukei and L. lactis(29) and engineered GAD from E. coli(21,22) under the control of two strong synthetic promoters (H30 and H36) having different strengths in C. glutamicum. During the flask cultivation, the concentration of GABA increased proportionally to the pH in the culture medium up to 7.0, indicating that the recombinant strains with engineered GADs could result in enhanced GABA production from glucose even at an almost neutral pH value. Further batch fermentation of C. glutamicum H36EcGADmut, H36LsGAD, and H36LlGAD rendered slightly higher final concentrations of GABA from 100 g L–1 of glucose at pH 6.0, reaching yields of up to 40.3, 41.6, and 42.5 g L–1, respectively. Since comparable amounts of GABA can be produced by the engineered strains even at a neutral pH value, the recombinant C. glutamicum strains developed in this study could be used for direct fermentative production of GABA in a growth-coupled manner, thus increasing the applicability of GADs that can produce GABA from glucose at an almost neutral pH. In addition, developed competitive bioprocesses with high amounts of GABA titer, productivity, and yield using glucose as a sole carbon source could contribute to the establishment of the economically feasible and efficient one-step fermentative GABA production system. However, the remaining issues must be addressed to ensure the widespread adoption of the approaches proposed herein, and therefore, the following strategies will be explored in our future studies: (i) characterizing the crystal structure of the enzymes and their functions, (ii) additional expression of pyridoxal kinase without the addition of external PLP for enhanced GAD activity,[50,51] and (iii) blocking competitive pathways related to GABA degradation.[45]