Metabolic engineering of high-salinity-induced biosynthesis of γ-aminobutyric acid improves salt-stress tolerance in a glutamic acid-overproducing mutant of an ectoine-deficient Halomonas elongata

Abstract

A moderately halophilic eubacterium, Halomonas elongata, has been used as cell factory to produce fine chemical 1,4,5,6-tetrahydro-2-methyl-4-pyrimidinecarboxylic acid (ectoine), which functions as a major osmolyte protecting the cells from high-salinity stress. To explore the possibility of using H. elongata to biosynthesize other valuable osmolytes, an ectoine-deficient salt-sensitive H. elongata deletion mutant strain KA1 (ΔectABC), which only grows well in minimal medium containing up to 3% NaCl, was subjected to an adaptive mutagenesis screening in search of mutants with restored salt tolerance. Consequently, we obtained a mutant, which tolerates 6% NaCl in minimal medium by overproducing L-glutamic acid (Glu). However, this Glu-overproducing (GOP) strain has a lower tolerance level than the wild-type H. elongata, possibly because the acidity of Glu interferes with the pH homeostasis of the cell and hinders its own cellular accumulation. Enzymatic decarboxylation of Glu to γ-aminobutyric acid (GABA) by a Glu decarboxylase (GAD) could restore cellular pH homeostasis; therefore, we introduced an engineered salt-inducible HopgadBmut gene, which encodes a wide pH-range GAD mutant, into the genome of the H. elongata GOP strain. We found that the resulting H. elongata GOP-Gad strain exhibits higher salt tolerance than the GOP strain by accumulating high concentration of GABA as an osmolyte in the cell (176.94 µmol/g cell dry weight in minimal medium containing 7% NaCl). With H. elongata OUT30018 genetic background, H. elongata GOP-Gad strain can utilize biomass-derived carbon and nitrogen compounds as its sole carbon and nitrogen sources, making it a good candidate for the development of GABA-producing cell factories.IMPORTANCEWhile the wild-type moderately halophilic H. elongata can synthesize ectoine as a high-value osmolyte via the aspartic acid metabolic pathway, a mutant H. elongata GOP strain identified in this work opens doors for the biosynthesis of alternative valuable osmolytes via glutamic acid metabolic pathway. Further metabolic engineering to install a GAD system into the H. elongata GOP strain successfully created a H. elongata GOP-Gad strain, which acquired higher tolerance to salt stress by accumulating GABA as a major osmolyte. With the ability to assimilate biomass-derived carbon and nitrogen sources and thrive in high-salinity environment, the H. elongata GOP-Gad strain can be used in the development of sustainable GABA-producing cell factories.

Keywords: Halomonas elongata; L-glutamic acid; compatible osmolyte; metabolic engineering; γ-aminobutyric acid.

Fig 1

Growth of wild-type H. elongata OUT30018 (WT), ectoine-deficient salt-sensitive KA1, and KA1-derived spontaneous mutant strains under different salt-stress conditions. H. elongata strains were precultured in M63 medium supplemented with 4% glycerol and 3% NaCl until optical density at 600 nm (OD₆₀₀) reached around 0.8 and used as 5% inoculum for main cultures in M63 medium containing 4% glycerol with different NaCl concentrations (3%, 6%, 7%, or 8% NaCl). Photos of the cultures were taken after 48 h of incubation to show differences in cell density.

Fig 2

Profiles of L-glutamic acid and L-alanine in the cells of wild-type H. elongata OUT30018, ectoine-deficient salt-sensitive KA1, and KA1-derived spontaneous mutant glutamic acid-overproducing strains culturing in M63 medium containing 4% glycerol with 3% (open column), 6% (hatched column), or 7% (filled column) NaCl. H. elongata strains were precultured in M63 medium containing 4% glycerol with 3%, 6%, or 7% NaCl until OD₆₀₀ reached more than 0.80 and used as a 5% inoculum for the main cultures in same salinity (3%, 6%, or 7% NaCl) medium. When OD₆₀₀ of the main cultures reached more than 0.80, free amino acids including Glu and Ala were extracted from the cells by dissolving cell pellets in pure water (20 µL pure water per 1 mg cell fresh weight), and the extracts were analyzed by HPLC. Data were normalized with internal standard norvaline. Values are mean ± standard deviation (n = 3). *P ≤ 0.05. ND, no data; because H. elongata KA1 strain is unable to grow in M63 medium containing 6% or 7% NaCl. (A) Profile of Glu and Ala in WT H. elongata OUT30018 cells. (B) Profile of Glu and Ala in H. elongata KA1 cells. (C) Profile of Glu and Ala in H. elongata GOP cells.

Fig 3

Schematic of major osmolyte biosynthetic pathways operating in H. elongata OUT30018, KA1, GOP, KA1-Gad, and GOP-Gad strains. (A) H. elongata OUT30018 accumulates ectoine as a result of an expression of the salt-inducible ectABC operon, which contains genes that encode the three enzymes of the ectoine biosynthesis pathway; L-2,4-diaminobutyric acid (DABA) transaminase (DAT) encoded by ectB gene, DABA acetyltransferase (DAA) encoded by ectA gene, and ectoine synthase (ES) encoded by ectC gene. (B) H. elongata KA1 strain does not accumulate ectoine due to the lack of the ectABC gene cluster. This strain can only grow well in the medium containing 3% NaCl. (C) H. elongata GOP strain does not accumulate ectoine due to the lack of the ectABC gene cluster; however, spontaneous mutation in its genome made this strain produces and accumulates higher Glu than the KA1 strain, possibly due to enhanced activity of either glutamate synthetase (GOGAT) or glutamate dehydrogenase (GDH). As a result, this strain has higher salt tolerance than the KA1 strain and can grow in the medium containing 6% and 7% NaCl. (D) H. elongata KA1-Gad strain is engineered to contain a salt-inducible artificial bicistronic mCherry-HopGadBmut operon encoding a red fluorescent reporter protein (mCherry) and a wide pH-range mutant of an L-glutamic acid decarboxylase (GAD), which converts Glu to GABA. This strain does not accumulate ectoine due to the lack of the ectABC gene and does not accumulate Glu to the concentration that is high enough to support GABA accumulation under high-salinity conditions. Therefore, this strain cannot grow in the medium containing more than 4% NaCl. (E) H. elongata GOP-Gad strain is engineered to contain a salt-inducible artificial bicistronic mCherry-HopGadBmut operon encoding an mCherry and a wide pH-range GAD mutant, which converts Glu accumulated in this spontaneous mutant into GABA. This strain grows better than the GOP strain in medium containing 6% and 7% NaCl due to its ability to accumulate GABA as a major osmolyte. Gly, glycerol; DHAP, dihydroxyacetone phosphate; PYR, pyruvate; AcCoA, acetyl coenzyme A; Ac-P, acetyl phosphate; A-AMP, acetyl-AMP; CIT, citrate; ICT, isocitrate; α-KG, α-ketoglutarate; SUC-CoA, succinyl-coenzyme A; SUC, succinate; FUM, fumarate; MAL, malate; OAA, oxaloacetate; Asp, aspartate; ASA, aspartic β-semialdehyde; ADABA, N-γ-acetyl-L-2,4-diaminobutyric acid; GS, glutamine synthetase.

Fig 4

Schematic of the genomic structure at ectABC locus in H. elongata OUT30018, KA1, GOP, KA1-Gad, and GOP-Gad strains. U_ectA, 1-kb upstream region of the ectA gene, which contains an ectA promoter with putative binding sites for the osmotically induced sigma factor σ³⁸ and the vegetative sigma factor σ⁷⁰. This region was used as a target for homologous recombination at the ectABC locus. D_ectC, 1-kb downstream region of the ectC gene, which contains an ectC terminator. This region was used as a target for homologous recombination at the ectABC locus. ectA, gene, which encodes an L-2,4-diaminobutyric acid (DABA) acetyltransferase; ectB, gene, which encodes a DABA transaminase; ectC, gene, which encodes an ectoine synthase; mCherry, gene, which encodes a red fluorescent reporter protein mCherry. HopGadBmut, synthetic H. elongata’s codon-usage-optimized (Hop) GadB mutant gene (HopGadBmut), which encodes a mutant glutamate decarboxylase (GAD) with activity across broader pH range than the wild-type GAD.

Fig 5

Salt-inducible production of mCherry reporter protein in recombinant H. elongata KA1-Gad and GOP-Gad strains. H. elongata OUT30018 (WT), GOP-Gad, and KA1-Gad strains cultured in M63 medium containing 4% glycerol with 3% or 6% NaCl until OD₆₀₀ reached more than 1.00 were used as a 5% inoculum for the main cultures in same salinity (3% or 6% NaCl) M63 medium containing 4% glycerol. When OD₆₀₀ of the main cultures reached more than 1.00, the cells were pelleted and subjected to tests. (A) Visualization of the salt-inducible production of the mCherry fluorescent reporter protein in the H. elongata GOP-Gad and KA1-Gad strains as shown by mCherry fluorescence of the cell pellets under fluorescent light (Fl) in comparison with bright-field (BF) images. Cell pellet of H. elongata OUT30018 (WT) was used as a negative control. (B) Detection of mCherry protein produced in the H. elongata GOP-Gad and KA1-Gad strains by western blot (WB) analysis. Proteins extracted from the cell pellets shown in (A) were electrophoresed in two identical 5%–20% gradient SDS-polyacrylamide gels. One gel was stained with Coomassie Brilliant Blue (CBB; right panels) for visualization of total protein separated on each lane, while proteins on the other gel were transferred to PVDF membrane and probed with antibody to red fluorescent protein (RFP) in WB analysis (left panels). Protein extracted from H. elongata OUT30018 (WT) was used as a negative control. Rat anti-RFP tag was used as a primary antibody, and goat anti-Rat IgG/IgM (H + L) HRP was used as a secondary antibody. mCherry protein bands were detected at 26 kDa (*).

Fig 6

Profiles of major osmolytes in H. elongata KA1, KA1-Gad, GOP, and GOP-Gad cells grown in M63 medium containing 3% NaCl. Intracellular concentration of major osmolytes, Glu (open columns), Ala (hatched columns), and GABA (filled columns) in H. elongata KA1, KA1-Gad, GOP, and GOP-Gad cells cultured in M63 medium containing 4% glycerol with 3% NaCl was profiled. Precultures were grown in M63 medium containing 4% glycerol with 3% NaCl to the OD₆₀₀ of more than 1.00 and used as a 5% inoculum for a main culture in fresh M63 medium containing 4% glycerol with 3% NaCl. When OD₆₀₀ of the main cultures was 0.5–0.8 during exponential growth phase, osmolytes were extracted from the cells and analyzed by HPLC. Data were normalized with internal standard norvaline. Values are mean ± standard deviation (n = 5). BDL, below detection limit.

Fig 7

Effect of medium salinity on growths of H. elongata GOP and GOP-Gad strains. H. elongata GOP and GOP-Gad strains were precultured in M63 medium containing 4% glycerol with 3% NaCl until OD₆₀₀ reached 1.00 before used as an inoculum for the main cultures in M63 medium containing 4% glycerol with 3% (▲), 6% (◆), 7% (●), or 8% (■) NaCl. The starting OD₆₀₀ of all main cultures was adjusted to 0.01, and OD₆₀₀ of each cell culture was measured at different time points. Values are mean ± standard deviation (n = 5). (A) Growth curve of H. elongata GOP strain. (B) Growth curve of H. elongata GOP-Gad strain. (C) Growth comparison between GOP (open columns) and GOP-Gad (hatched columns) strains cultured in M63 medium containing 4% glycerol with 3%, 6%, 7%, or 8% NaCl for 7 days. *P ≤ 0.05.

Fig 8

Differences in major osmolytes composition among H. elongata GOP-Gad cultures, which are growing at different rates in high-salinity medium. H. elongata GOP-Gad cells, cultured in M63 medium containing 4% glycerol and 3% NaCl to OD₆₀₀ of more than 0.80, were used as a 5% inoculum for 30 main cultures in M63 medium containing 4% glycerol with 7% NaCl. When OD₆₀₀ of the main cultures reached 0.80–1.11, intracellular osmolytes were extracted from the cells and analyzed by HPLC. (A) Profiles of major osmolytes of H. elongata GOP-Gad strain cultured in M63 medium containing 4% glycerol and 7% NaCl. Concentration of Glu (open columns), Ala (hatched columns), and GABA (filled columns) was normalized with internal standard norvaline. Values are mean ± standard deviation (n = 30). (B) Growth profiles of H. elongata GOP-Gad cultures growing at different rates. The cultures were sorted by their OD₆₀₀ at day 7 of cultivation into two groups. Out of 30 cultures, 21 cultures with OD₆₀₀ that was equal to or more than 0.80 (OD₆₀₀ = 0.80–1.11) were categorized as fast-growing cultures (filled columns), and 9 cultures with OD₆₀₀ that was less than 0.80 (OD₆₀₀ = 0.51–0.75) were categorized as slow-growing cultures (open columns). Dashed horizontal line indicates the point, where OD₆₀₀ = 0.80, which separates fast-growing cultures from the slow-growing cultures. (C) Profiles of major osmolytes in the cells of the fast- and the slow-growing GOP-Gad cultures. Because intracellular osmolytes were extracted from the cultured with OD₆₀₀ between 0.80 and 1.11, intracellular osmolytes of the fast-growing cultures (n = 21 of 30) were extracted after 7 days of cultivation, while those of the slow-growing cultures (n = 9 of 30) were extracted after 8 or 9 days of cultivation.

Fig 9

Effect of medium salinity on GABA accumulation in the H. elongata GOP-Gad strain. (A) Profiles of major osmolytes in the cells of H. elongata GOP-Gad strain cultured in M63 medium containing 4% glycerol with 3%, 6%, or 7% NaCl. The strain was precultured in M63 medium containing 3% NaCl to OD₆₀₀ of more than 0.8 and used as a 5% inoculum for the main culture in M63 medium containing 4% glycerol with 3%, 6%, or 7% NaCl. After two subcultures, osmolytes of the cells from the third culture were extracted when OD₆₀₀ reached 0.9–1.2 and were analyzed by HPLC. Concentration of Glu, Ala, and GABA was normalized with internal standard norvaline. Values are mean ± standard deviation (n = 5). Glu, open columns; Ala, hatched columns; GABA, filled columns. (B) Molar ratio of GABA to Glu (GABA/Glu) in the cellular extracts of H. elongata GOP-Gad strain cultured in M63 medium containing 4% glycerol with 3%, 6%, or 7% NaCl. Ratio was calculated from data shown in Fig. 9A. Values are mean ± standard deviation (n = 5). **P ≤ 0.01, ***P ≤ 0.001.

Fig 10

Comparison of major osmolytes extraction methods: wet cells, hypo-osmotic extraction (bacterial milking) method vs freeze-dried cells, conventional phase-separation method. H. elongata GOP-Gad strain was precultured in M63 medium containing 4% glycerol and 3% NaCl until OD₆₀₀ was more than 0.8 and used as a 5% inoculum for the main culture in 120 mL M63 medium containing 4% glycerol and 7% NaCl (n = 3). When the culture reached late-log phase (OD₆₀₀ = 1.0–1.2), cells were harvested from 50 mL of the culture in duplicate, and the weights of the wet cell pellets were recorded as cell fresh weight (CFW). Pure water was added to one of the pellet samples to extract the major osmolytes by hypo-osmotic bacterial milking method, while the other pellet sample was freeze dried, and the weight of the dried cell pellet was recorded as cell dry weight (CDW). Then, the major osmolytes were extracted from the dried pellet by adding methanol/chloroform/water (10:5:3.4, by volume) in the conventional phase-separation method. The amount of Glu, Ala, and GABA in the extracts was determined by HPLC. Concentration of Glu (open columns), Ala (hatched columns), and GABA (filled columns) was normalized with internal standard norvaline. (A) Major osmolyte profile of H. elongata GOP-Gad strain growing under high-salinity stress condition derived from sample extracted by wet cells, hypo-osmotic extraction (bacterial milking) method. (B) Major osmolyte profile of H. elongata GOP-Gad strain growing under high-salinity stress condition derived from sample extracted by freeze-dried cells, conventional phase-separation method. (C) A unit conversion of the data shown in B from μmol/g CDW to μmol/g CFW by calculating with original CFW of each sample measured before the freeze-drying process. (D) Ratios of the major osmolytes in the extract derived by wet cells, hypo-osmotic extraction (bacterial milking) method (Fig. 10A) to those derived by freeze-dried cells, conventional phase-separation method (Fig. 10C).