Previous Article | Next Article ![]()
Applied and Environmental Microbiology, October 2005, p. 5920-5928, Vol. 71, No. 10
0099-2240/05/$08.00+0 doi:10.1128/AEM.71.10.5920-5928.2005
Copyright © 2005, American Society for Microbiology. All Rights Reserved.
Institute of Biotechnology 1, Research Center Juelich, D-52428 Juelich, Germany,1 Laboratoire des Sciences du Genie Chimique, Institut National Polytechnique de Lorraine, 54505 Vandoeuvre-les-Nancy, France2
Received 18 March 2005/ Accepted 6 May 2005
|
|
|---|
|
|
|---|
Glutamate production by C. glutamicum is characterized by a low pentose phosphate pathway flux (22); low conversion of 2-oxoglutarate to oxaloacetate in the tricarboxylic acid cycle (22), likely due to low activity of the oxoglutarate dehydrogenase complex (3, 12, 37, 38); pyruvate carboxylase fulfilling the anaplerotic demand, since phosphoenolpyruvate carboxylase is inhibited by glutamate (2, 28); and glutamate excretion by a specific export system (5, 9, 12).
A peculiarity of glutamate production is that it has to be elicited by external triggers, e.g., by addition of penicillin or detergents, by biotin limitation, or by ethambutol addition (5,17, 20, 33, 37, 38). In some strains glutamate production can be induced by a temperature shift (3). This is the case with C. glutamicum 2262. However, the production properties of this strain are unstable, and upon prolonged continuous cultivation, a non-glutamate-producing mutant, referred to below as C. glutamicum 2262NP, was isolated (41). The temperature shift was found to induce a metabolic transition in C. glutamicum 2262 that was not observed in the mutant C. glutamicum 2262NP. This transition typically involves a decrease in the activities of the 2-oxoglutarate dehydrogenase complex and of pyruvate dehydrogenase, a redirection of the 2-oxoglutarate flux toward glutamate production, and the formation of L-lactate as a by-product (41).
Since glutamate production by C. glutamicum has not yet been analyzed with respect to global gene expression changes, we employed DNA microarray analysis for transcriptome comparisons between glutamate-producing C. glutamicum 2262 and its non-glutamate-producing mutant 2262NP. Moreover, the existence of this pair of strains, which grow very similarly in continuous cultivation and differ in their glutamate-producing capacity, provided an opportunity to determine gene expression changes specific for glutamate production under the defined conditions of a chemostat. Here we characterize an operon, induced during temperature-triggered glutamate production, that is necessary for the (re)utilization of L-lactate.
(Part of the work described here belongs to the planned dissertation of C. Stansen at the Faculty of Mathematics and Natural Sciences of the Heinrich-Heine-Universität, Düsseldorf, Germany.)
|
|
|---|
|
View this table: [in a new window] |
TABLE 1. Bacterial strains and plasmids used in this study
|
Quinone-dependent L-lactate dehydrogenase assay.
For determination of enzyme activities, exponentially growing cells were harvested by centrifugation (4,500 x g, 5 min, 4°C) and washed twice with 50 mM ice-cold KH2PO4, pH 7.0. Cell extracts were prepared directly or after storage at 70°C by resuspension in 1 ml of 50 mM KH2PO4, pH 7.0, and disruption by ultrasonic treatment at 4°C (UP 200S; Dr. Hielscher GmbH, Teltow, Germany) at an amplitude of 55% and a duty cycle of 0.5 for 6 min. After centrifugation at 4°C for 60 min at 13,000 x g, enzyme activity was determined immediately in the cell-free supernatant. L-Lactate dehydrogenase activity was determined by a modified assay according to reference 23. Reaction mixtures of 1 ml contained 100 mM KH2PO4 (pH 7.5)-50 µM 2,6-dichloroindophenol (DCPIP) and 20 µl crude extract. Quinone-dependent L-lactate dehydrogenase was assayed spectrophotometrically at 30°C by determining the decrease in absorbance of DCPIP (
600 = 20 cm2 µmol1). The reaction was started by addition of 20 mM L-lactate, 10 mM D-lactate, or 20 mM racemic malic acid.
Comparative genome and transcriptome analysis using DNA microarrays.
Generation of C. glutamicum DNA microarrays, total RNA preparation, cDNA synthesis, preparation and labeling of genomic DNA, DNA microarray hybridization, and statistical data analysis were performed as described previously (14, 21, 29, 42).
Construction of plasmids and strains.
The genome sequence of C. glutamicum deposited in GenBank as accession number NC_003450 was used. Plasmids were constructed in Escherichia coli DH5
from PCR-generated fragments (Expand High Fidelity PCR kit; Roche Diagnostics) by using C. glutamicum ATCC 13032 genomic DNA prepared according to the method of Eikmanns et al. (6) as a template. E. coli was transformed by the RbCl2 method (10), while C. glutamicum was transformed via electroporation (39). All transformants were analyzed by plasmid analysis and/or PCR with appropriate primers, respectively. To construct an NCgl2816 inactivation mutant of C. glutamicum ATCC 13032, an internal 468-bp fragment was PCR amplified using primers NCgl2816-int_fw (5'-GCAACGAGTTCCTCTCCACAATC-3' [the nucleotide corresponding to nucleotide {nt} 3118442 of NC_003450 is underlined]) and NCgl2816-int_rev (5'-ACAGGCCGATGAACACTAGGAG-3' [the nucleotide corresponding to nt 3118909 of NC_003450 is underlined]). Subsequently, this fragment was blunted and cloned into the SmaI site of pK18mob (35). Gene inactivation with pK18mobNCgl2816int was carried out as described previously (27a). The correct genotype of the inactivation mutant was verified by PCR analysis using primer NCgl2816-int_proof_fw (5'-TGACCATGAGGTTGGGCCAATC-3' [the nucleotide corresponding to nt 3118215 of NC_003450 is underlined]) and the universal primer M13-24F (on pK18mob). In order to construct pEKEx3-lldD, lldD was amplifed by PCR using primers lldD_fw (5'-AAGGAGATATAGATATGGTGAAACGTCAACTGCC-3' [the nucleotide corresponding to nt 3119621 of NC_003450 is underlined, and italicized nucleotides correspond to the ribosome binding site of T7 gene 10]) and lldD_rev (5'-TTAAATCTCCGCCGCTGC AGAACGAGTT-3' [the nucleotide corresponding to nt 3120883 of NC_003450 is underlined]). The PCR fragment was blunted and cloned into the SmaI site of pEKEx3.
Reverse transcription-PCR (RT-PCR) analysis.
RNA was prepared as described previously (14, 21, 29, 42) and incubated in 0.1 M sodium acetate, pH 3.0, 5 mM MgSO4, and 0.3 U µl1 DNase I (Roche Diagnostics, Mannheim, Germany) for 30 min at 37°C. After heat inactivation for 5 min at 75°C and phenol-chloroform extraction, the RNA was precipitated with LiCl as described elsewhere (34). After denaturation for 5 min at 65°C, reverse transcription of 350 ng RNA was performed with Omniscript Reverse Transcriptase (QIAGEN, Hilden, Germany) according to the manufacturer's instructions by using primer lldD_rev (described in the preceding section) for reaction cDNA-A and primer NCgl2816-int_proof_rev (5'-TGCTGCTCCGACTTGGTAACC-3' [the nucleotide corresponding to nt 3119433 of NC_003450 is underlined]) for reaction cDNA-B. Subsequently, the cDNAs were amplified using combinations of the above-mentioned primers NCgl2816-int_proof_fw, NCgl2816-int_proof_rev, lldD_fw, and lldD_rev. As a control, dnaE RNA was reverse transcribed using RT-dnaE-rev (5'-CTGGAACCATGTCGTCCTAGAG-3' [the nucleotide corresponding to nt 2251091 of NC_003450 is underlined]) and amplified using primers RT-dnaE-fw (5'-TGCCCTTCCGGCGATGTCCAA-3' [the nucleotide corresponding to nt 2251460 of NC_003450 is underlined]) and RT-dnaE-rev.
|
|
|---|
![]() View larger version (22K): [in a new window] |
FIG. 1. Continuous culture of glutamate-producing C. glutamicum 2262 (A and B) and its non-glutamate-producing mutant 2262NP (C and D) at 39°C and a dilution rate of 0.05 h1. (A and C) Optical densities at 570 nm ( ); (B and D) intracellular (x) and extracellular ( ) glutamate concentrations. Arrows indicate the time when both the continuous supplementation of medium and the temperature increase from 33 to 39°C were imposed. The pH was maintained at 7.
|
![]() View larger version (14K): [in a new window] |
FIG. 2. L-Lactate concentrations in supernatants ( ) and specific activities of quinone-dependent L-lactate dehydrogenase ( ) in crude extracts of glutamate-producing C. glutamicum 2262 (A) and of its non-glutamate-producing mutant 2262NP (B) during continuous culture. Samples analyzed were taken from the continuous cultures described in the legend to Fig. 1.
|
The transcriptomes of glutamate-producing C. glutamicum 2262 and the non-glutamate-producing strain C. glutamicum 2262NP were compared using the samples withdrawn during the chemostat cultivations at 70 h of cultivation (see Fig. 1). Three independent DNA microarray experiments based on different RNA isolations were performed. These revealed that only two genes showed an mRNA level significantly (P < 0.05) changed by at least fourfold. NCgl2816 and NCgl2817 showed 11- and 9-fold-higher mRNA levels, respectively, in C. glutamicum 2262 than in C. glutamicum 2262NP. The two genes are adjacent to each other on the C. glutamicum chromosome, and their organization suggests that they are cotranscribed. The NCgl2816 gene encodes a putative transport protein and is situated upstream of NCgl2817, which putatively encodes a quinone-dependent L-lactate dehydrogenase. This led to the hypothesis that NCgl2817 and NCgl2816 constitute an operon for uptake and oxidation of L-lactate.
To follow up this hypothesis, samples were withdrawn from the chemostat cultures to assay for quinone-dependent L-lactate dehydrogenase in crude extracts. Whereas the glutamate-producing strain showed a transient increase in the specific activity of quinone-dependent L-lactate dehydrogenase, peaking at 45 nmol mg1 min1 (Fig. 2), this was not the case with the non-glutamate-producing strain 2262NP, where the activity did not exceed 20 nmol mg1 min1 (Fig. 2). Thus, the increased expression of NCgl2817 was reflected in an increased quinone-dependent L-lactate dehydrogenase activity. Importantly, the increase in the L-lactate concentration preceded the increase in the specific activity of quinone-dependent L-lactate dehydrogenase (Fig. 2). The L-lactate concentration in the culture medium decreased more rapidly than that calculated on the basis of the dilution kinetics of the continuous cultivation, indicating reutilization of L-lactate by C. glutamicum. These observations and the fact that, for thermodynamic reasons, quinone-dependent L-lactate dehydrogenase (EC 1.1.2.3; encoded by NCgl2817, lldD) operates in the L-lactate oxidation direction (1) are commensurate with the view that this enzyme is involved in reutilization of excreted L-lactate.
Inactivation of NCgl2816 in C. glutamicum ATCC 13032.
To study the function of the putative lactate utilization genes, the genetically tractable wild-type strain C. glutamicum ATCC 13032 was employed (15). The inactivation mutant 13032::pK18mobNCgl2816int was constructed and analyzed in order to determine whether the NCgl2816 gene is involved in L-lactate utilization. This mutant grew like the parent strain with pyruvate as a carbon source but was not able to utilize L-lactate (data not shown). Both the mutant and the parent strain grew similarly, albeit poorly, on D-lactate (data not shown).
In order to test whether the altered growth of the mutant is due to inactivation of the NCgl2816 gene or to polar effects on the downstream gene NCgl2817, the latter gene was cloned into pEKEx3 and expressed in the inactivation mutant to yield C. glutamicum 13032::pK18mobNCgl2816int(pEKEx3-NCgl2817). This recombinant was able to utilize L-lactate again (data not shown), demonstrating that only the function of NCgl2817 is essential for growth of C. glutamicum on L-lactate as the sole carbon source.
Properties and carbon source-dependent regulation of quinone-dependent L-lactate dehydrogenase from C. glutamicum ATCC 13032.
To test whether NCgl2817 encodes quinone- dependent L-lactate dehydrogenase, the specific activity of this enzyme was determined under various conditions. Only specific activities of quinone-dependent L-lactate dehydrogenase lower than 0.03 µmol mg1 · min1 were observed for strains 13032::pK18mobNCgl2816int and ATCC 13032 during growth on D-lactate or glucose (Table 2), and both strains showed specific activities of quinone-dependent L-lactate dehydrogenase of 0.02 µmol mg1 · min1 during growth on pyruvate. However, during growth on a mixture of L-lactate and glucose, C. glutamicum ATCC 13032 exhibited a high specific activity of 0.23 µmol mg1· min1, while that of 13032::pK18mobNCgl2816int was lower than 0.03 µmol mg1 · min1 (Table 2 and data not shown). Overexpression of NCgl2817 in ATCC 13032 and 13032::pK18mobNCgl2816int increased the specific activity of quinone-dependent L-lactate dehydrogenase during growth on glucose 47- and 29-fold, respectively (Table 2). Taking these data together, NCgl2817 is the gene encoding quinone-dependent L-lactate dehydrogenase, and we therefore renamed NCgl2817 as lldD.
|
View this table: [in a new window] |
TABLE 2. Specific activities of quinone-dependent L-lactate dehydrogenase and quinone-dependent D-lactate dehydrogenase in various C. glutamicum strains during growth on glucose or L-lactate
|
![]() View larger version (18K): [in a new window] |
FIG. 3. Substrate dependence of C. glutamicum WT(pEKEx3-lldD) quinone-dependent L-lactate dehydrogenase. Shown are specific activities of quinone-dependent L-lactate dehydrogenase in crude extracts from C. glutamicum WT(pEKEx3-lldD) with varying concentrations of the substrate L-lactate. (Inset) Double-reciprocal plot of the data.
|
Comparative transcriptome analysis of L-lactate- and pyruvate-grown C. glutamicum ATCC 13032.
We compared the transcriptomes of pyruvate- and L-lactate-grown cultures to test whether L-lactate-grown C. glutamicum ATCC 13032 cells showed high lldD mRNA levels and to determine whether NCgl2816 and lldD were the only genes specifically induced by L-lactate. The DNA microarray experiment revealed higher mRNA levels of NCgl2816 and lldD during growth on L-lactate than during growth on pyruvate (Table 3). Although NCgl2816 and lldD showed the highest increases in their mRNA levels on L-lactate compared to pyruvate, L-lactate utilization was characterized by additional gene expression changes. With L-lactate as the carbon source, expression of 20 further genes was more than fourfold higher, and expression of 10 genes more than fourfold lower, than for pyruvate-grown cultures (Table 3). The genes showing higher expression during growth on L-lactate mostly code for hypothetical proteinsputative regulatory, transport, and secretory proteins. Pyruvate-grown cells, on the other hand, showed higher expression of the pta-ack operon, coding for phosphotransacetylase (sixfold [Table 3]) and acetate kinase (twofold), and about fivefold-higher expression of aceA, coding for the glyoxylate cycle enzyme isocitrate lyase, and of the sdhCAB operon, coding for the tricarboxylic acid cycle enzyme succinate dehydrogenase (Table 3). These genes were proposed to belong to the regulon of the regulator of acetate metabolism RamB (7, 8). During growth on L-lactate, expression of genes of the iron starvation stimulon (19), like that of genes encoding proteins for the uptake and utilization of iron siderophores or heme, increased as follows: NCgl0377, 5.2-fold; NCgl0378, 1.8-fold; NCgl0380, 2.2-fold; NCgl0381, 2.1-fold; NCgl0382, 1.6-fold; NCgl0482, 4.2-fold; NCgl0483, 2.3-fold; NCgl0636, 2.1-fold; NCgl0637, 2.5-fold; NCgl0639, 3.6-fold; NCgl0773, 3.0-fold; NCgl00774, 3.3-fold; NCgl0776, 2.7-fold; NCgl0777, 1.9-fold; NCgl0778, 2.1-fold; NCgl00779, 2.9-fold; NCgl2970, 2.6-fold (see also Table 3). The genes for an AraC- and an ArsR-like regulator (NCgl0430 and NCgl0943) showed 4.2- and 10.1-fold-higher RNA levels, respectively, than during growth on pyruvate (Table 3).
|
View this table: [in a new window] |
TABLE 3. Comparative gene expression analysis of C. glutamicum ATCC 13032 grown on pyruvate or L-lactate as the sole carbon source
|
![]() View larger version (34K): [in a new window] |
FIG. 4. Transcriptional organization of the NCgl2816-lldD locus in C. glutamicum analyzed by RT-PCR. (A) Scheme showing the NCgl2816-lldD locus in C. glutamicum and the RT-PCRs used to determine cotranscription of NCgl2816 and lldD. RNA from wild-type C. glutamicum was transcribed into cDNA with two different primers in the two separate reverse transcriptase reactions cDNA-A and cDNA-B. Subsequently, these cDNAs were used as templates for the PCRs labeled 1 to 6. (B) Results from the RT-PCR analyses described above. The lower DNA fragment visible in lanes 1 to 6 represents dnaE, and RT-PCR of dnaE served as a positive control in all reactions. The upper bands correspond to the products of PCRs 1 to 6, diagramed in panel A. Lanes 7 to 12 represent control reactions confirming the absence of DNA in the RNA preparation. The reactions were identical to PCRs 1 to 6 (for which results are shown in lanes 1 to 6, respectively) except that reverse transcriptase was omitted in reactions cDNA-A and cDNA-B.
|
|
|
|---|
Quinone-dependent L-lactate dehydrogenase from C. glutamicum belongs to the cluster of orthologous groups COG1304 and shows 88% sequence identity to Corynebacterium efficiens CE2762. The related mycobacteria possess two genes for quinone-dependent L-lactate dehydrogenases (lldD1 and lldD2). Sequence similarities of C. glutamicum LldD to the LldD2 enzymes from Mycobacterium tuberculosis (57%) (Rv1872c), Mycobacterium avium subsp. paratuberculosis (58%) (MAP1585c), Mycobacterium bovis (58%) (Mb1903c), and Mycobacterium leprae (58%) (ML2046) are higher than those to LldD1 enzymes from these mycobacteria (31% each for M. tuberculosis Rv0694, M. avium subsp. paratuberculosis MAP4154, and M. bovis Mb0713) and higher than those to LldD from E. coli (34%) and L-lactate ferricytochrome c reductase (29%) (Cyb2p) from the yeast Saccharomyces cerevisiae.
Besides quinone-dependent L-lactate dehydrogenase (EC 1.1.2.3), the following enzymes are known to occur in Corynebacterianeae: L-lactate monooxygenase (EC 1.13.12.4), NAD-dependent L-lactate dehydrogenase (EC 1.1.1.27), and quinone-dependent D-lactate dehydrogenase (EC 1.1.1.28). C. glutamicum possesses quinone-dependent D-lactate dehydrogenase (EC 1.1.1.28) activity likely to be encoded by NCgl0865 (dld) (1). An L-lactate monooxygenase (EC 1.13.12.4) has been characterized in Mycobacterium smegmatis as a flavoprotein utilizing molecular oxygen to oxidize L-lactate to acetate, CO2, and water (24). However, there are no indications that the genomes of C. glutamicum, C. efficiens, M. tuberculosis, M. leprae, and E. coli contain genes for L-lactate monooxygenases. On the other hand, genes for NAD-dependent L-lactate dehydrogenases (EC 1.1.1.27) are present in Corynebacterineae. These enzymes catalyze the exergonic NAD-dependent reduction of pyruvate to L-lactate and thus operate in the reverse direction from quinone-dependent lactate dehydrogenases, which catalyze the highly exergonic menaquinone-dependent oxidation of lactate (1). In C. glutamicum, ldhA (NCgl2810) encodes a NAD-dependent L-lactate dehydrogenase and is crucial for L-lactate production (13).
Recently, a membrane proteomics study identified the quinone-dependent L-lactate dehydrogenase LldD as a membrane-associated protein in C. glutamicum (36). Membrane association of LldD either could facilitate oxidation of L-lactate immediately after its uptake or, as proposed recently (25), could function in NADH oxidation. In a type II NADH dehydrogenase gene (ndh) disruption mutant of C. glutamicum KY9714, membrane fractions showed increased L-lactate oxidase activity, while ndh overexpression led to decreased L-lactate as well as malate oxidase activities (25). The authors suggested that coupling between cytoplasmic NAD-dependent L-lactate dehydrogenase and membrane-associated quinone-dependent L-lactate dehydrogenase and/or coupling between cytoplasmic NAD-dependent malate dehydrogenase and membrane-associated malate:quinone oxidoreductase results in effective NADH oxidation when type II NADH dehydrogenase function is defective (25). It remains to be studied whether double or triple mutants lacking ndh and either the L-lactate or the malate oxidation cycle or both cycles exhibit impaired growth relative to that of the parent strain.
NCgl2816 is transcribed together with lldD in an operon and codes for a putative permease of the major facilitator superfamily (TC 2.A.1) (43). Disruption of the NCgl2816-lldD operon resulted in the inability to grow on L-lactate as a sole carbon source unless the mutant was complemented with a plasmid carrying lldD, indicating that NCgl2816 is not essential for growth on L-lactate. Thus, if NCgl2816 codes for an L-lactate transporter, as suggested from its primary sequence and the operon organization, additional carriers for uptake of L-lactate are likely to exist in C. glutamicum. In fact, L-lactate is imported into E. coli cells either by the L-lactate permease LldP (4) or by the glycolate uptake system GlcA (27). Both systems are redundant with respect to uptake of 2-hydroxymonocarboxylates (26, 27). Since the transcriptome analysis of L-lactate-grown C. glutamicum revealed higher mRNA levels of NCgl2713 and NCgl2965, and both encode putative permeases, these are possible candidates for 2-hydroxymonocarboxylate carriers.
Expression of the NCgl2816-lldD operon was regulated by the carbon source. When L-lactate was present in the medium, increased mRNA levels of NCgl2816 and lldD as well as an increased specific activity of quinone-dependent L-lactate dehydrogenase were observed. Transcription of the lld operon of E. coli is repressed by the two-component signal transduction system ArcB/ArcA during anaerobiosis, is activated by the pyruvate dehydrogenase repressor PdhR, and might be repressed by LldR (4, 32). However, in C. glutamicum, a lactate-responsive regulatory system has not yet been identified.
L-Lactate is a by-product excreted into the medium during glutamate production. In the continuous cultivation of C. glutamicum 2262, the excreted L-lactate was reutilized after the specific activity of quinone-dependent L-lactate dehydrogenase increased; this reutilization might have contributed to glutamate formation. To improve glutamate production, it is proposed to eliminate the concurrent L-lactate formation by deletion of the ldhA gene, since the encoded NAD-dependent L-lactate dehydrogenase was shown to be the only L-lactate-generating enzyme in C. glutamicum (13). Alternatively, increasing the activity of pyruvate dehydrogenase by overexpressing the respective genes or by introducing alleles for deregulated pyruvate dehydrogenase might improve glutamate production, as it was proposed that the reduced activity of pyruvate dehydrogenase observed during glutamate production is responsible for by-product formation (41).
Supplemental material for this article may be found at http://aem.asm.org/. ![]()
|
|
|---|
This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Copyright © 2009 by the American Society for Microbiology. For an alternate route to Journals.ASM.org, visit: http://intl-journals.asm.org | More Info»