Altered Metabolic Flux due to Deletion of odhA causes l-Glutamate Overproduction in Corynebacterium glutamicum

Bacterial strains and growth conditions. C. glutamicum ATCC 13869 (formerly B. lactofermentum) was used as a wild-type strain. The Escherichia coli K-12 strain JM109 [recA1 endA1 gyrA96 thi hsdR17 supE44 relA1 Δ(lac-proAB)/F′ (traD36 proAB⁺lacI^qlacZΔM15)] (40) was used in the recombinant DNA procedures. Antibiotics at the following concentrations were used for the selection of cells carrying plasmids: 25 μg kanamycin ml⁻¹ or 4 μg chloramphenicol ml⁻¹ for C. glutamicum and 25 μg kanamycin ml⁻¹ or 40 μg chloramphenicol ml⁻¹ for E. coli. C. glutamicum and E. coli were grown on CM2B medium (per liter, 10 g yeast extract, 10 g tryptone, 5 g NaCl, and 10 μg biotin) at 30°C and LB medium (per liter, 5 g yeast extract, 10 g tryptone, and 5 g NaCl) at 37°C, respectively.

Batch fermentation.Batch fermentations were conducted at 30°C in 1,000-ml stirred vessels (1,200 rpm) containing 300 ml of a solution comprising (per liter) 65 g glucose, 1.5 g KH₂PO₄, 1 g MgSO₄, 0.01 g FeSO₄, 0.01 g MnSO₄, 0.45 mg vitamin B₁, 0.45 mg biotin, and 40 ml soybean protein hydrolysate. They were maintained at pH 7.0 by the automatic addition of ammonia gas. The fermentations were initiated by inoculating cells grown overnight on CM2B agar at an optical density at 660 nm (OD₆₆₀) of 0.15. During fermentation, samples were removed to measure OD₆₆₀, l-glutamate, and glucose. When all the glucose had been consumed, the fermentation was stopped, and fatty acid proportions and dry weight of cells were determined.

Flask culture.To analyze the clones isolated from the originally obtained odhA deletion strain AJ13133, to investigate the effects of l-glutamate-producing inductions on l-glutamate production in AJ110214 cells, and to assay enzyme activities, cells were cultured in a 500-ml flask containing 20 ml of a medium comprising (per liter) 30 g glucose, 1 g KH₂PO₄, 0.4 g MgSO₄·7H₂O, 0.01 g FeSO4·7H₂O, 0.01 g MnSO₄·4H₂O, 15 g (NH₄)₂SO₄, 200 μg vitamin B₁ hydrochloride, 300 μg biotin, and 40 ml soybean protein hydrolysate (pH 8.0). To inoculate the flask cultures, cells grown overnight on CM2B agar were added to an OD₆₆₀ of 0.15. Tween 40 (Wako, Tokyo, Japan) or penicillin (Wako) was added to the cultures at the desired concentrations. For biotin-limiting conditions, 1-ml aliquots of precultures grown in the appropriate concentrations of biotin were inoculated into 20 ml of the main culture medium without biotin, so that the biotin was depleted during subsequent growth. Samples were removed every 2 to 3 h to measure OD₆₆₀, l-glutamate, and glucose.

Analysis of fatty acids.Cells were harvested by centrifugation, washed twice in H₂O, and freeze dried. Fatty acids were analyzed by gas chromatography (Japan Food Research Laboratories' Custom Service, Tokyo, Japan).

Analysis of growth, l-glutamate production, and glucose consumption.Growth was monitored by measuring OD₆₆₀. l-Glutamate and glucose concentrations were measured with a Biotech-Analyzer AS-210 (Sakura Seiki, Tokyo, Japan) using a glutamate oxidase and a glucose oxidase sensor, respectively. Specific rates of l-glutamate production and glucose consumption (expressed as grams per gram [dry weight] of cells per hour) were calculated. The dry weight of cells was calculated from the OD₆₆₀ by the formula obtained experimentally by measuring the dry weight of cells and the OD₆₆₀s of six samples with different cell densities (OD₆₆₀, 36 to 60); dry weight of cells (expressed in grams per liter⁻¹) = 26.052 × (OD₆₆₀/101)^1.1224.

Analysis of organic acids and amino acids.Organic acid concentrations were analyzed by high-performance liquid chromatography using an ULTRON DW-80 AD0415 column with phosphate buffer (pH 1.9) as the mobile phase (1 ml⁻¹; 45°C) and were detected at 210 nm. Amino acid concentrations were measured by ninhydrin reaction followed by high-performance liquid chromatography using an L-8500 automatic amino acid analyzer (Hitachi, Tokyo, Japan).

Analysis of intracellular metabolites.Cells of a 500-μl sample were separated from the medium by centrifugation on 500 μl of KF-54 silicon oil (specific gravity, 1.07) (Shinetsu Kagaku Kogyo, Tokyo, Japan) as the separation layer and 300 μl 21% HClO₄ as an acid fixation layer (64,000 × g, 5 min, and 4°C) (13). The sedimented cells in the acid layer were further disrupted using a freeze-thaw procedure, and the resulting extracts were neutralized in the cold by adding 225 μl of 2 M Na₂CO₃. The supernatant obtained by centrifugation (64,000 × g, 5 min, and 4°C) was used for determination of intracellular amino acids and organic acids. The cytoplasmic volume was assumed to be 1.6 μl (mg [dry weight] of cells)⁻¹ (10).

Analysis of N-terminal amino acids.Approximately 200 μg of protein in the crude cell extract was separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and blotted onto a polyvinylidene fluoride membrane. The membrane was stained with Coomassie brilliant blue R-250, the region corresponding to the 136-kDa band was cut out, and its N-terminal sequence was analyzed by automated Edman degradation employing a gas phase protein sequencer (TaKaRa Custom Service Center, Kyoto, Japan).

Construction of the odhA deletion strain.A 4.4-kbp SalI-XhoI fragment (Fig. 2) from plasmid pPKSX4.4 (37) containing the odhA gene was cloned into the SalI site of E. coli vector plasmid pHSG299 (GenBank accession no. M19415 ; TaKaRa, Kyoto, Japan), yielding plasmid pHSGodhASX. A 1.9-kbp KpnI-KpnI region in the middle of odhA was deleted by digesting pHSGodhASX with KpnI, followed by self-ligation. Then a temperature-sensitive C. glutamicum plasmid replicon, a 2.9-kbp BamHI-KpnI fragment from pHSC4 (34), which is derived from pHM1519 (23) and cannot replicate at 34°C in C. glutamicum, was ligated into its BamHI site to generate plasmid pdodhA. It harbors the deleted odhA, and its replication is blocked at the nonpermissive temperature of 34°C. A C. glutamicum derivative lacking odhA (designated AJ13133) was obtained by homologous recombination replacing the chromosomal odhA gene with the deleted odhA on pdodhA by the following procedures. The C. glutamicum wild-type strain ATCC 13869 was transformed with pdodhA at 30°C, cultured at 34°C with kanamycin, and the clones resistant to kanamycin at 34°C were selected. They had their chromosomal odhA gene integrated with pdodhA. The cells were then cultured at 34°C without kanamycin, and the clones sensitive to kanamycin at 34°C were selected. The deletion in chromosomal odhA was confirmed by dideoxy sequencing and Southern hybridization, in which chromosomal DNA was digested with BamHI and probed with the 1.5-kbp 5′-terminal region of odhA (Fig. 2). A 9.5-kbp BamHI restriction fragment hybridized to an odhA probe was observed in the mutant lacking odhA, whereas this restriction fragment was 11.4 kbp in the wild type. Spontaneous mutants from the odhA deletion strain AJ13133 that grew faster on CM2B agar were isolated. One of these, named AJ110214, had no ODHC activity.

• Open in new tab
• Download powerpoint

FIG. 2.

Map of the odhA gene. Double-headed arrows indicate the fragment used. Arrow A, 4.4-kbp SalI-XhoI fragment containing the odhA gene cloned into pPKSX4.4. Arrow B, 1.5-kbp 5′-terminal fragment used for Southern hybridization. Double-headed dashed arrow C indicates the 1.9-kbp KpnI-KpnI region deleted in the odhA deletion strain.

Construction of AJ110214 derivatives with various GDH activities.A 2.2-kbp DNA fragment containing the GDH coding region and its putative promoter (2) was amplified by PCR with oligonucleotides 5′-GCTAGCCTCGGGAGCTCT-3′ and 5′-GATCTTTCCCAGACTCTG-3′ as the primers and with C. glutamicum ATCC 13869 chromosomal DNA as the template. The resulting fragment was cloned into the SmaI site of the E. coli vector plasmid pSTV28 (TaKaRa), yielding plasmid pSTVgdhwt. The C. glutamicum plasmid pHM1519 (23) was then ligated into the SalI site of pSTVgdhwt to produce plasmid pgdhwt. The gdh promoter variants of pgdhwt (Table 2) were obtained by site-directed mutagenesis of the putative promoter region, as follows. The first PCR was performed with oligonucleotides 5′-GCTAGCCTCGGGAGCTCTAGGAGAT-3′ and 5′-ACGTTCAATTATAGCAGTGTCGCAC-3′ (primer 2) as the primers and with C. glutamicum ATCC 13869 chromosomal DNA as the template. The second PCR was performed using the amplified DNA fragment as one of the primers, oligonucleotide 5′-GATCTTTCCCAGACTCTG-3′ as another primer, and C. glutamicum ATCC 13869 chromosomal DNA as the template. The resulting fragment was cloned into the SmaI site of the E. coli vector plasmid pSTV28 (TaKaRa), yielding plasmid pSTVgdh2. The C. glutamicum plasmid pHM1519 (23) was then ligated into the SalI site of pSTVgdhwt to produce plasmid pgdh2. Oligonucleotides 5′-ACGTTCAATTATAGCAGTGTCGCACAGATATGTCAACAAAGAATTAAAATTGTTTGAAA-3′, 5′-ACGTTCAATTATAGCAGTGTCGCACAGATATGACAACAAAGAATTAAAATTGTTTGAAA-3′, and 5′-ACGTTCAATTATAGCAGTGTCGCACAGATATGGCAACAAAGAATTAAAATTGTTTGAAA-3′ were used instead of primer 2 to produce pghd3, pdhh4, and pghd7, respectively. The direction of gdh transcription on the plasmid was opposite that of the lacZ transcription originally located on pSTV28. Sequences were confirmed by dideoxy sequencing whenever PCR was performed. The resulting plasmids were introduced into C. glutamicum AJ110214 by electroporation to generate strains with various GDH activities (Table 2).

Chromosomal gdh promoter variants were also obtained from AJ110214 by homologous recombination, replacing the putative promoter region with the various nucleotide sequences shown in Table 2, by the following procedures. A 1.9-kbp DNA fragment containing the putative gdh promoter (2) was amplified by PCR with oligonucleotides 5′-CAGTTGTGGCTGATCCGCCAAGGCC-3′ and 5′-TCGCCCTTAGCCTTGATCATTTCAC-3′ as the primers and with C. glutamicum ATCC 13869 chromosomal DNA as the template. The resulting fragment, which had the promoter region in its middle so that two homologous crossovers would occur equally on both sides of the promoter region, was cloned into the SmaI site of the E. coli vector plasmid pHSG299 (GenBank accession no. M19415 ; TaKaRa), yielding plasmid pHSGgdhprwt. Into the XbaI site of pHSGgdhprwt, a temperature-sensitive C. glutamicum plasmid, comprising a 2.96-kbp BamHI-KpnI fragment from pHSC4 (34), which is derived from pHM1519 (23), was ligated to give plasmid pgdhprwt. Then, the NheI-BglII region of pgdhprwt containing the promoter within the insert of pHSGgdhprwt was replaced by the corresponding region of pgdh2, pgdh3, or pgdh7 (Table 2) to give plasmid pgdhpr2, pgdhpr3, or pgdhpr7, respectively. AJ110214 was transformed with pgdhpr2, pgdhpr3, or pgdhpr7 by electroporation, and the clones with the chromosomal gdh promoter replaced by the corresponding region on the temperature-sensitive plasmid pghdpr2, pgdhpr3, or pgdhpr7 were selected as in the construction of the odhA deletion strain. AJ110214gdh2 was obtained using pghdpr2, AJ110214gdh3 was obtained using pghdpr3, and AJ110214gdh7 and AJ110214gdh7′ were obtained using pghdpr7. The chromosomal gdh promoter sequences were confirmed by dideoxy sequencing.

Enzyme assays.OHDC and GDH were assayed as described previously (2, 14) using the cells from flask cultures.

The N-terminal amino acid sequence of OdhA.We previously cloned the C. glutamicum odhA gene, which encodes the E1o subunit of ODHC. Its deduced amino acid sequence revealed that OdhA has a unique N-terminal extension similar to the catalytic domain of the E2o subunit of ODHC (37). In order to identify the precise protein-coding region of the odhA gene, we determined the N-terminal amino acid sequence of OdhA. As we reported previously (37), a 136-kDa band was enhanced by sodium dodecyl sulfate-polyacrylamide gel electrophoretic analyses of cell extracts of C. glutamicum harboring plasmid pPKSX4.4 carrying the odhA gene. The sequence of 11 N-terminal amino acids was NH₂-Xaa-Xaa-Ala-Ser-Thr-Phe-Gly-Asn-Ala-Xaa-Leu (the first, second and tenth residues could not be identified). This sequence was identical to the sequence from ³⁸Ser to ⁴⁹Leu of the odhA open reading frame in D84102. The translation start was found to be ⁵⁵¹GTG. odhA encodes a 1,221-amino-acid polypeptide with a molecular mass of 134,623 Da. A ribosome binding-like sequence, GAGG, is located 12 bp upstream of the initiation codon.

This N-terminal amino acid analysis confirmed that OdhA was translated from its N-terminal extension. Recent genome analysis has revealed that the open reading frame of the odhA gene of Corynebacterium efficiens starts with GTG, which corresponds to the initiation codon ⁵⁵¹GTG of C. glutamicum odhA (26). An E2o-like N-terminal extension is commonly found in the OdhA proteins of gram-positive bacteria with high GC contents, such as Mycobacterium tuberculosis (3), Mycobacterium leprae (4), and Streptomyces coelicolor (1). C. glutamicum has another gene annotated as an E2o subunit within its genome (gene ID, 1020158), which has the characteristic feature of E2 subunits: three copies of a lipoyl-binding domain followed by the catalytic domain (12, 30). It would be interesting to know the function of the N-terminal extension of OdhA.

Construction of the odhA deletion strain.To construct a strain that completely lacked ODHC activity, we deleted the region between two KpnI sites in odhA, which corresponds to the segment from ²⁶⁵Thr to ⁹⁰⁵Gly in OdhA (see Materials and Methods). As expected, no ODHC activity was detected in the resulting odhA deletion strain, AJ13133, whereas the ODHC-specific activity of the wild-type strain was 2.7 nmol min⁻¹ (mg of protein)⁻¹.

Growth and l-glutamate production of the odhA deletion strain.AJ13133 produced a remarkably high level of l-glutamate without any inductions (Fig. 3). This was as high as the level of the wild type under l-glutamate-producing conditions: 27.8 mmol (g [dry weight] of cells)⁻¹ compared with 1.0 mmol (g [dry weight] of cells)⁻¹ for the wild type without any induction (Fig. 3). Moreover, AJ13133 accumulated l-glutamate during growth, unlike the wild type under l-glutamate-producing conditions, where l-glutamate production takes place after growth arrest due to the induction treatments (14, 33). l-Glutamate production by AJ13133 continued until all the glucose was consumed.

• Open in new tab
• Download powerpoint

FIG. 3.

Growth (A) and l-glutamate production (B) of the C. glutamicum wild-type strain ATCC 13869 and its derivative lacking odhA. Batch fermentations were conducted in stirred vessels at 30°C and pH 7.0 until all the glucose was consumed. AJ13133 produced a remarkably high level of l-glutamate. It accumulated l-glutamate during growth, which continued to the end of the fermentation. Solid circles, wild type; open circles, odhA deletion strain.

Fatty acid composition.To determine whether alteration of the fatty acid composition of the membrane was necessary for l-glutamate production, we analyzed the fatty acid composition of whole cells of AJ13133 (Table 1). We assumed that the fatty acid composition of whole cells paralleled that of the membrane, since almost all fatty acid exists in the membrane.

The data on the wild-type cells were in good agreement with published results; we noted a relative decrease of about 50% in oleic acid levels, and a relative increase of a similar amount in stearic acid levels, with Tween 40 addition (11, 31). The fatty acid composition of the cells lacking odhA, which were producing l-glutamate as efficiently as the wild-type cells with Tween 40 added, was almost the same as that of wild-type cells under the non-l-glutamate-producing condition. Similar results were obtained in two independent experiments. Evidently, l-glutamate production takes place without alterations in the fatty acid composition of the membrane. We therefore conclude that l-glutamate production in C. glutamicum can be caused by metabolic flux change due to odhA deletion alone, without alteration of the cell membrane.

Clones isolated from the odhA deletion strain.The original odhA deletion strain, AJ13133, appeared to be genetically unstable, because the colonies formed on CM2B agar differed in size after storage at 4°C for about 2 weeks. Although the reason was unclear, a cell lacking odhA might accumulate considerable l-glutamate, which could be harmful to the cell. It might also accumulate intermediates of glycolysis or the TCA cycle (such as 2-oxoglutarate, acetate, or pyruvate), which could inhibit some important enzymes of glycolysis or the TCA cycle or could be toxic due to acidity. Mutants in which such problems are somehow overcome might arise spontaneously, for example, mutants in which the organic acids might be metabolized more rapidly, or mutants in which l-glutamate might be excreted more actively. We isolated 6 clones that grew rapidly and 12 that grew slowly on CM2B agar and cultured them in flasks in order to know their growth and l-glutamate production characteristics (Fig. 4). OD₆₆₀ and l-glutamate levels were measured after glucose was completely consumed, when the cultivation time was 16 to 20 h for the first six clones and 40 to 46 h for the last 12 clones. The first six clones were of two types: three clones produced l-glutamate as efficiently as the original strain, while the other three clones did not produce l-glutamate and accumulated metabolites such as 145.8 mM acetate, 6.3 mM alanine, 4.3 mM valine, 1.7 mM lysine, and 0.7 mM 2-oxoglutarate (Fig. 4 A, B, and C). The 12 clones that grew slowly on CM2B agar produced 60 to 70% of the l-glutamate of the original strain. We picked one clone from the first group, which produced l-glutamate as efficiently as the original strain, named it AJ110214, and used it in further experiments. The growth rates of the wild type, AJ13133, and AJ110214 on CM2B liquid medium were 0.47 ± 0.03 h⁻¹, 0.22 ± 0.04 h⁻¹, and 0.40 ± 0.01 h⁻¹, respectively (averages ± standard deviations of three independent experiments). No ODHC activity was detected in AJ110214, and it was not unstable like the original mutant, AJ13133. AJ110214 transformed by plasmid pPKSX4.4, containing the odhA gene, did not overproduce l-glutamate.

• Open in new tab
• Download powerpoint

FIG. 4.

Clones isolated from the original strain AJ13133 lacking odhA. (A) Growth and l-glutamate production of clones isolated from the original odhA deletion strain AJ13133. The isolated clones could be divided into three groups: the first, rapidly growing clones that produced l-glutamate as efficiently as the parent strain (squares); the second, rapidly growing clones that did not overproduce l-glutamate (circles); and the third, clones whose growth was slower than that of the parent and whose l-glutamate production was 60 to 70% of that by the parent (triangles). OD₆₆₀ and l-glutamate levels were measured after glucose was completely consumed, when the cultivation time was 16 to 20 h for the clones in the first and second groups and 40 to 46 h for the clones in the third group. (B and C) By-products of the isolated clones. (B) Amino acids; (C) organic acids. Results for two clones from each group are shown. Solid bars, clones in the first group; open bars, clones in the second group; shaded bars, clones in the third group.

Effect of l-glutamate-producing inductions on l-glutamate production by strain AJ110214 lacking odhA.To clarify the relationship between l-glutamate-producing inductions, ODHC activity, and l-glutamate production, we analyzed the effects of l-glutamate-producing inductions, biotin limitation, and the addition of Tween 40 or penicillin on l-glutamate production in AJ110214.

Under biotin-limited conditions, where the total amount of biotin supplied to the culture was 0.014 μg/ml or less, l-glutamate production was induced in the wild type, as reported previously. In AJ110214, significant l-glutamate production was observed, as in AJ13133, even in the presence of biotin levels above 0.3 μg/ml, at which the wild type does not overproduce l-glutamate. When biotin was limited to 0.0014 μg/ml or less, the l-glutamate yield (millimoles of glutamate produced per millimole of glucose at the end of the experiment) of AJ110214 increased; when biotin was limited to 0.0005 μg/ml, the l-glutamate yield of AJ110214 increased by as much as 8.3% over that under biotin-rich conditions (average of three independent experiments) (Fig. 5C). In addition, the l-glutamate yield of AJ110214 was 11.8% higher than that of the wild type (average of three independent experiments) when the biotin concentration was 0.0014 μg/ml. These findings show that ohdA deletion plus biotin limitation gave a higher l-glutamate yield than either ohdA deletion alone or biotin limitation alone.

• Open in new tab
• Download powerpoint

FIG. 5.

Effects of the l-glutamate-producing condition of biotin limitation on l-glutamate production in AJ11024. Wild-type (solid circles) and AJ110214 (open circles) cells were cultured under biotin-limited conditions. Biotin concentrations are shown along the x axis. For biotin-limited conditions, seed cultures were prepared with 10, 30, 60, 100, and 300 μg of biotin liter⁻¹, and 1 ml was inoculated into 20 ml of the main medium lacking biotin so that biotin was depleted during the main cultures. For biotin-rich conditions, a seed culture was prepared in 300 μg of biotin liter⁻¹, and 1 ml was inoculated into 20 ml of the main medium containing 300 μg of biotin liter⁻¹. The biotin concentration was calculated by dividing the total amounts of biotin supplied to the seed and main cultures by the volume of the main culture. The specific l-glutamate production rate (A), the specific glucose consumption rate (B), and the l-glutamate yield (C) are shown. Specific rates were calculated during the periods when they were approximately constant. l-Glutamate yield is defined as millimoles of l-glutamate produced per millimole of glucose at the end of the experiment. Error bars, standard deviations from the means, calculated from three independent experiments.

Furthermore, under biotin-rich conditions, the specific glucose consumption rate of AJ110214 was much higher than that under biotin-limited conditions (Fig. 5B). The specific l-glutamate production rate under biotin-rich conditions was also high, probably as a consequence of the high glucose consumption rate (Fig. 5A), although the l-glutamate yield was lower than that under biotin-limited conditions. The specific glucose consumption rate was more sensitive to the biotin concentration in AJ110214 than in the wild type (Fig. 5B).

When Tween 40 was added, l-glutamate production was induced in the wild type, as reported previously (7, 35) (Fig. 6). In AJ110214, significant l-glutamate production was observed even without Tween 40 addition, as in AJ13133, and when 0.1 g of Tween 40 liter⁻¹ was added, both the specific l-glutamate production rate and the specific glucose consumption rate increased (Fig. 6A and B), and the l-glutamate yield also increased, by as much as 7.1% (average of three independent experiments). With 0.3 g of Tween 40 liter⁻¹, the l-glutamate yield of AJ110214 was much higher than that of the wild-type strain (70.9% and 45.5%, respectively [averages of three independent experiments]). These findings show that ohdA deletion plus Tween 40 addition yielded higher l-glutamate production than either ohdA deletion alone or Tween 40 addition alone.

• Open in new tab
• Download powerpoint

FIG. 6.

Effects of the l-glutamate-producing condition of Tween 40 addition on l-glutamate production in AJ110214. Wild-type (solid circles) and AJ110214 (open circles) cells were cultured with Tween 40 addition; Tween 40 was added to 0, 0.1, 0.3, or 3 g liter⁻¹. The specific l-glutamate production rate (A), the specific glucose consumption rate (B), and the l-glutamate yield (C) are shown. Specific rates were calculated during the periods when they were approximately constant. l-Glutamate yield is defined as millimoles of l-glutamate produced per millimole of glucose at the end of the experiment. Error bars, standard deviations from the means, calculated from three independent experiments.

When penicillin was added, l-glutamate production was induced in the wild type, as reported previously (7, 35). In AJ110214, significant l-glutamate production was observed even without penicillin addition, as in AJ13133, and when 0.2 U of penicillin ml⁻¹ was added, though the growth rate and the OD remained approximately unchanged (Fig. 7A), both the glucose consumption rate and the l-glutamate production rate increased (Fig. 7B and C), and the l-glutamate yield also increased (Fig. 7D). The l-glutamate yield of AJ110214 was much higher than that of the wild-type strain (75.2% versus 15.2%) when 0.2 U of penicillin ml⁻¹ was added (Fig. 7D). These findings show that ohdA deletion plus penicillin addition yielded higher l-glutamate production than either ohdA deletion alone or penicillin addition alone.

• Open in new tab
• Download powerpoint

FIG. 7.

Effects of the l-glutamate-producing condition of penicillin addition on l-glutamate production in AJ11024. Wild-type (solid symbols) and AJ110214 (open symbols) cells were cultured without penicillin (circles) or with 0.2 U of penicillin ml⁻¹ (triangles). (A) Time courses of growth; (B) glucose consumption; (C) l-glutamate production; (D) l-glutamate yield.

In summary, AJ110214 produced l-glutamate without l-glutamate-producing inductions, and its productivity was increased by biotin limitation, addition of Tween 40, or addition of penicillin. These results show that each of the l-glutamate-producing inductions and the loss of ODHC activity have additive effects with respect to l-glutamate production.

Analysis of the C. glutamicum gdh promoter.We found that the odhA deletion strains, AJ13113 and AJ110214, accumulated 2-oxoglutarate in the medium; AJ13113 accumulated about 12 mM, and AJ110214 accumulated about 5 to 6 mM 2-oxoglutarate when cultured in flasks (Fig. 4C). This suggests that the GDH activities of these strains are not high enough to deal with the increased intracellular pool of 2-oxoglutarate resulting from the loss of ODHC activity. To investigate the GDH activity that optimizes the balance between the metabolic flux to 2-oxoglutarate via the TCA cycle and the flux from 2-oxoglutarate to l-glutamate, we constructed AJ110214 derivatives with GDH activities ranging from 1-fold to 95.1-fold higher than that of the parent strain. For this purpose, we analyzed the promoter of gdh (Table 2). The sequence upstream of the GDH-coding region in C. glutamicum ATCC 13869 was slightly different from the sequence reported for C. glutamicum ATCC 13032 (2); however, we were able to locate the −35 TGGTCA and −10 CATAAT motifs 33 and 10 bp upstream of the transcription start, respectively, as reported previously (2). The regions between −10 and the transcription start and between −35 and −10 were shorter by 1 nucleotide in ATCC 13869 than in ATCC 13032. The extended −10 motif “TGn” found in E. coli (21) was also present just upstream of the putative −10 hexamer. We mutated the putative −35 and −10 regions to make them more similar to the E. coli or C. glutamicum consensus promoter motifs (Table 2): TTGACA for the −35 sequence in E. coli, TTGCCA for the −35 sequence in C. glutamicum (28), and TATAAT for the −10 sequence in both species (28). The −35 region is reported to be much less conserved in C. glutamicum than in E. coli or Bacillus spp. (28, 29). Replacement of the putative −10 region by TATAAT enhanced GDH activity 4.5-fold. In addition, replacement of the putative −35 region by TTGCCA enhanced GDH activity sixfold, and replacement by TTGACA enhanced it sevenfold. T was more favorable than G for the second nucleotide in the −35 hexamer.

Optimizing metabolic flux for l-glutamate production by tuning GDH activity.To identify the GDH activity that optimizes the metabolic flux from 2-oxoglutarate to l-glutamate, we analyzed the effect of the GDH activity level on l-glutamate production (Fig. 8). When GDH activity was 7.1- to 13.5-fold higher than that in AJ110214, 2-oxoglutarate accumulation fell to 1.1 mM and l-glutamate production was maximally increased, from 108 to 130 mM (average of two independent experiments). When the GDH activity was increased more than 13.5-fold, l-glutamate production decreased. Thus, further metabolic changes optimizing the flux from 2-oxoglutarate to l-glutamate improved l-glutamate production resulting from loss of ODHC activity.

• Open in new tab
• Download powerpoint

FIG. 8.

Optimization of GDH activity for l-glutamate production in AJ110214. Derivatives of AJ110214 with various GDH activities were cultured in flasks, and production of l-glutamate (A) and 2-oxoglutarate (B) was analyzed at 15 to 16 h, when glucose was completely consumed. GDH activities are shown relative to that of AJ110214 (taken as 1). l-Glutamate production was maximally increased from 108 to 130 mM, at which concentration GDH activity was 7.1- to 13.5-fold higher than that in AJ110214. Two independent experiments were performed. Open diamonds, AJ110214gdh2; open squares, AJ110214gdh3; open triangles, AJ110214gdh7; asterisks, AJ110214gdh7′; solid diamonds, AJ110214/pgdh2; solid squares, AJ110214/pgdh3; small bars, AJ110214/pgdh4; solid triangles, AJ110214/pgdh7; crosses, AJ110214/pgdhwt; open circles, AJ110214; solid circles, AJ110214/pSAC4 (vector).

In addition, we noted that the accumulation of 2-oxoglutarate also fell when the GDH activity was increased more than 13.5-fold (Fig. 8B). We investigated whether the intracellular concentrations of 2-oxoglutarate and l-glutamate would reflect the extracellular production of the two metabolites under conditions of wild-type-level or high-level (6.0-, 13.5-, and 95.1-fold higher than the wild-type level) GDH activity (Fig. 9A). Intracellular 2-oxoglutarate was detected in several samples, but we could not determine the precise concentrations, because they were too low. The intracellular l-glutamate concentration was elevated in parallel with the increase in the GDH activity level from 22.5 mM to 245.5 mM (average of two independent experiments) (Fig. 9A), unlike extracellular l-glutamate production (Fig. 8 and 9D), indicating that the l-glutamate efflux is limiting when the GDH activity level is high.

• Open in new tab
• Download powerpoint

FIG. 9.

Effects of Tween 40 addition on l-glutamate production in the odhA deletion strains (AJ110214) with wild-type levels of GDH activity. Intracellular and extracellular metabolites were analyzed under conditions of no Tween 40 addition (A and D) or addition of 0.1 (B and E) or 0.3 (C and F) g of Tween 40 liter⁻¹, using the odhA deletion strains (AJ110214) with 1-, 6.0-, 13.5-, and 95.1-fold-higher levels of GDH activity. (A, B, and C) Intercellular concentrations of amino acids (left panels) and organic acids (right panels); (C, D, and E) extracellular production of l-glutamate (left panels), other amino acids (center panels), and organic acids (right panels) per gram of cells (dry weight). Results of two independent experiments are shown. Magenta, AJ110214; pink, AJ11024/pSAC4 (vector); yellow, AJ110214gdh7 (with 6.0-fold-higher GDH levels); light blue, AJ11024/pghdwt (13.5-fold-higher GDH levels); violet, AJ11024/pgdh7 (95.1-fold-higher GDH levels).

Analysis of the effects of Tween 40 addition on l-glutamate production in the odhA deletion strain.We show that each of the l-glutamate-producing inductions had an additive effect on l-glutamate production brought about by loss of ODHC activity. To understand the effects of those inductions, we investigated the intracellular and extracellular production of l-glutamate, other amino acids, and organic acids including 2-oxoglutarate under the condition of Tween 40 addition, using odhA deletion strains with wild-type-level or high-level (6.0-, 13.5-, and 95.1-fold higher than that of the parent) GDH activity.

For AJ110214, intracellular concentrations of lactate, aspartate, alanine, succinate, and 2-oxoglutarate were increased by the addition of 0.1 g of Tween 40 liter⁻¹, and this increase was enhanced when the Tween 40 concentration was elevated to 0.3 g liter⁻¹, except for succinate. Accordingly, the intracellular l-glutamate concentration of AJ110214 rose from 22.5 mM to 553.6 mM and 726.1 mM (averages of two independent experiments) (Fig. 9A) by addition of 0.1 g (Fig. 9B) and 0.3 g (Fig. 9C) of Tween 40 liter⁻¹, respectively. This indicates that addition of Tween 40 might dramatically change the metabolic flux, leading to more-efficient l-glutamate synthesis. A similar result was obtained in the same experiment using the wild-type: increases in the intracellular alanine (from 0.5 ± 0.2 mM to 10.9 ± 0.9 mM [average ± standard deviation of three independent experiments]), succinate (from 26.1 ± 7.3 mM to 87.1 ± 9.4 mM), and 2-oxoglutarate (from 0.75 ± 0.14 mM to 4.96 ± 0.54 mM) concentrations were also observed by the addition of 0.3 g of Tween 40 liter⁻¹, besides the increase in arginine concentrations (from 0.4 ± 0.1 mM to 10.0 ± 0.9 mM), when the intracellular l-glutamate concentration was elevated from 35.6 ± 10.6 mM to 202.9 ± 28.0 mM.

As for the effects of GDH activity levels on intracellular metabolites, 2-oxoglutarate concentrations decreased in parallel with the increase in the GDH activity level in the presence of 0.1 g and 0.3 g of Tween 40 liter⁻¹. The stronger the GDH activity, the more strongly the metabolic flux would be pulled to l-glutamate synthesis, as expected. The intracellular l-glutamate concentration rose with the increase in the GDH activity level in the absence of Tween 40. In the presence of 0.1 and 0.3 g of Tween 40 liter⁻¹, however, when the metabolic flux was drastically altered toward more-efficient l-glutamate synthesis, intracellular l-glutamate concentrations increased only slightly, from 553.6 mM to 601.9 mM under 0.1 g of Tween 40 liter⁻¹ (Fig. 9B) and from 726.1 mM to 772.3 mM under 0.3 g of Tween 40 liter⁻¹ (Fig. 9C) (average of two independent experiments), when the GDH level was elevated 6.0-fold. The intracellular l-glutamate concentration decreased when the GDH level was higher than 13.5-fold. These findings might be attributed to the result of metabolic balance; for example, the intracellular aspartate and alanine concentrations were increased with the increase in the GDH activity level at either of the Tween 40 concentrations, indicating altered equilibrium in aspartate aminotransferase and alanine aminotransferase reactions. In accordance with this, extracellular aspartate and alanine production was elevated, and extracellular l-glutamate production was decreased, when the GDH activity level was higher than 13.5-fold at either of the Tween 40 concentrations.

In summary, addition of Tween 40 was indicated to induce a drastic alteration in metabolic flux that might lead to more-efficient l-glutamate synthesis.