OUP user menu

Isolation of the Corynebacterium glutamicum glnA gene encoding glutamine synthetase I

Marc Jakoby , Martin Tesch , Hermann Sahm , Reinhard Krämer , Andreas Burkovski
DOI: http://dx.doi.org/10.1111/j.1574-6968.1997.tb12627.x 81-88 First published online: 1 September 1997


The Corynebacterium glutamicum glutamine synthetase I (GSI) structural gene glnA was cloned by a PCR approach using oligonucleotide primers derived from conserved amino acid sequences of the GSI proteins from various bacteria. Disruption or deletion of this gene in C. glutamicum led to a glutamine auxotrophic phenotype and complete loss of glutamine synthetase activity, indicating the key role of this enzyme in nitrogen metabolism. Additionally, indications for a second glutamine synthetase, GSII, were found.

  • Corynebacterium
  • Nitrogen metabolism
  • Glutamine synthetase
  • glnA

1 Introduction

Corynebacterium glutamicum is a Gram-positive soil bacterium, which excretes considerable amounts of l-glutamic acid under conditions of biotin limitation [10, 14]. Due to this remarkable ability this C. glutamicum is applied in fermentation processes on an industrial scale. With the help of different mutant strains large amounts of not only glutamate (800 000 tons year−1) but also lysine (250 000 tons year−1) are produced beside smaller amounts of alanine, isoleucine, proline, threonine, and tryptophan [19].

Regulation of nitrogen metabolism plays a pivotal role in the cell in general (for review, see [21]) and in the production of amino acids in particular. Glutamate and glutamine serve as key nitrogen donors for biosynthetic reactions. There are two major pathways which facilitate the incorporation of nitrogen into these amino acids. Glutamate dehydrogenase (GDH), encoded by the gdhA gene, is used under high ammonium supply, while the glutamine synthetase (GS)/glutamate synthase (GOGAT) system allows the assimilation of ammonia when present in the medium at concentrations lower than 0.1 mM. In bacteria the ubiquitously found form of glutamine synthetases (for review, see [34]) is GSI which is subject to posttranslational modification in many species. Adenylation is well-known in Enterobacteriaceae and is also found in Strepyomyces coelicolor[35] and Thiobacillus ferrooxidans[2], while a second form of modification, namely ADP-ribosylation, was found for example in Rhodospirillum rubrum[33] and Synechocystis strain PCC 6803 [29].

Since we are working on the transport [15] and the production of amino acids [16] we were interested in a detailed understanding of the mechanism of glutamine synthetase action in C. glutamicum. Therefore, we started an approach to isolate the corresponding gene(s) via PCR. The cloning, sequencing, and characterization of glnA is reported in this communication.

2 Materials and methods

2.1 Bacterial strains, plasmids, and growth

Strains and plasmids used in this study are shown in Table 1. Cells were routinely grown aerobically in Luria-Bertani (LB) medium [24] at 30°C (C. glutamicum) or 37°C (Escherichia coli). For nitrogen starvation of C. glutamicum, a nitrogen-free minimal medium was used [28]. For complementation cells were inoculated in LB medium with and without the addition of l-glutamine (5 mM final concentration). If appropriate, media were supplemented with antibiotics (50 μg ml−1 carbenicillin, 50 μg ml−1 nalidixic acid, 15 μg ml−1 kanamycin, final concentrations).

View this table:

Strains and plasmids used in this study

Strains and plasmidsRelevant genotype/characteristicsaReference/source
Bacterial strains
C. glutamicum
ATCC 13032wild-type[1]
MJ3glnA disruption strainthis study
MJ4-26glnA deletion strainthis study
E. coli
DH5αmcrsupE44, hsdR17, recA1, endA1, gyrA96, thi1, relA, mcrA, Δ(mrr-hsdRMS-mcrBC)[9]
S17-1res, RP4-2-Tc::Mu-Kanr::Tn7[30]
pJC1E. coli/C. glutamicum shuttle vector, Ampr, Kanr[6]
pJCglnA3.1 kb NheI fragment carrying the glnA gene ligated to pJC1this study
pK18moboriV, oriT, mob, Kanr[26]
pk19ΔglnApK19mobsacB carrying the flanking regions of glnAthis study
pK19mobsacBoriV, oriT, mob, sacB, Kanr[26]
pUC1.5a1.4 kb BamHI/HindIII fragment in pUC18this study
pUC18plac, Ampr[36]
pUC19plac, Ampr[36]
pUC5400.5 kb BamHI fragment in pUC19this study
pUCΔglnApUC19 carrying the regions flanking a 0.9 kb internal BamHI deletion in glnAthis study
  • aAmpr, resistance to ampicillin; Kanr, resistance to kanamycin.

2.2 Molecular biology techniques

Cloning of DNA and transformation of E. coli was carried out using standard techniques [24]. C. glutamicum cells were made competent as described, with the exception that 10% glycerol instead of sodium phosphate buffer was used [17]. Electroporation was carried out according to Kronemeyer et al. [17]. Conjugation between E. coli S17-1 and C. glutamicum was performed as described by Schäfer et al. [25] and transconjugants were selected on Brain Heart Infusion agar (Difco, Detroit, MI, USA) containing 5 mM glutamine, nalidixic acid, and kanamycin. For PCR a mixture of oligonucleotide primers was used derived from conserved amino acid regions from different GSI proteins (see Fig. 1; primer A (sense) 5′-AC(ACGT) TT(CT) ATG CC(ACGT) AA(AG) CC-3′; primer B (antisense) 5′-TT(AG) TT(CT) TT(AGT) AT(ACGT) CC(AG) TC-3′). PCR (30 s 94°C, 60 s 44°C, 60 s 72°C, for 35 cycles) was carried out using Taq polymerase (Boehringer, Mannheim, Germany), buffer conditions as recommended by the manufacturer, 3 μM primer A, 9 μM primer B, and a Thermo-Cycler 480 (Perkin-Elmer, Norwalk, CT, USA). Chromosomal C. glutamicum DNA as template was prepared as described [8]. Plasmid DNA was isolated with the Qiagen plasmid kit (Qiagen, Hilden, Germany). For Southern blot hybridization plasmid and genomic DNA was digested with restriction endonucleases (Boehringer) and the resulting DNA fragments were separated by agarose gel electrophoresis and transferred to a nylon filter for hybridization with a digoxigenin-labelled DNA fragment. Hybridization and detection were performed with the Boehringer DIG DNA Labeling and Detection Kit as recommended by the manufacturer.

Figure 1

Alignment of GSI proteins from different sources. C.GLUT, C. glutamicum (GenBank accession number Y13221); B.SUBT, B. subtilis[31]; E.COLI, E. coli[5, 22]; and S.COEL, S. coelicolor[35]. Amino acids identical in three proteins are marked by dots, those identical in all four proteins are marked by asterisks. Amino acids shaded in gray comprise conserved regions with the following functions [12, 32]: I, subunit interaction with regions II–V; II, binding of Mn2+; III, nucleotide binding; IV, binding of glutamate and Mn2+; V, unknown; VI, adenylation region, the adenylation site (Tyr-405) is marked. Amino acid sequences from which the oligonucleotide primers were derived are indicated by arrows.

2.3 Computer work

For computer-assisted nucleotide and protein sequence analyses the HUSAR package (EMBL, Heidelberg, Germany) and the PCGene program (release 6.26; Genofit, Geneva, Switzerland) were used.

2.4 Determination of glutamine synthetase activity

C. glutamicum cells were grown overnight in LB medium supplemented with 5 mM glutamine and 2% glucose, subsequently transferred to nitrogen-free minimal medium [28] containing 1% glucose and incubated for 4 h at 30°C. After harvesting the cells were disrupted by ultrasonic treatment. Intact cells and cell debris were removed by centrifugation. After protein determination [7] the raw extract was assayed for GS activity essentially as described by Shapiro and Stadtman [27].

3 Results

3.1 Cloning of the glnA gene

In order to isolate the glnA gene, oligonucleotide primers were derived from the GSI sequences of S. coelicolor[35], S. viridochromogenes[11], and Mycobacterium tuberculosis (GenBank accession number Q10377) (Fig. 1). Using these primers in a PCR approach with chromosomal DNA from C. glutamicum as template, a 401 bp DNA fragment was amplified which was ligated to SmaI-digested pUC19 DNA for sequence determination using the SureClone Ligation Kit (Pharmacia, Freiburg, Germany). DNA sequence analysis revealed a high similarity of this fragment with glnA genes from different sources. A similar approach for the isolation of a glnII gene, using oligonucleotide primers derived from various glnII sequences, failed (data not shown).

The isolated 401 bp glnA fragment was used for a gene disruption experiment. For this purpose it was isolated from the pUC19 construct (see above) by SacI/XbaI restriction and cloned into SacI/XbaI-digested DNA of insertion vector pK18mob which does not replicate in C. glutamicum. Transformation of C. glutamicum wild-type strain ATCC 13032 via conjugation with E. coli S17-1 resulted in kanamycin-resistant C. glutamicum cells only when the plasmid was integrated in the genome via recombination. Growth tests revealed that the insertion mutants were unable to grow without glutamine added to the medium. This was the result of a completely abolished glutamine synthetase activity as shown by in vitro testing (Table 2).

View this table:

Glutamine synthetase activities of different C. glutamicum strains

StrainDescriptionSpecific activity (U mg protein−1)
ATCC 13032wild-type0.43±0.02
MJ3glnA::pK18mobnot detectable
MJ4-26ΔglnAnot detectable
MJ4-26/pJC1ΔglnAnot detectable
MJ4-26/pJCglnAplasmid-encoded glnA2.06±0.72
  • The data shown represent mean values and standard deviations of at least three independent experiments.

In order to determine the complete sequence of the glnA gene, a plasmid rescue experiment was carried out (for restriction sites, see Fig. 2). For this approach chromosomal DNA of strain insertion mutant strain MJ3 was prepared, digested with Asp718 or HindIII, and religated. After transformation of restriction-deficient E. coli strain DH5αmcr, to circumvent the wild-type restriction system, with the two different ligation mixtures several hundred kanamycin-resistant clones arose. Twelve clones of each transformation were chosen for plasmid preparation. Restriction analyses revealed that all plasmids derived from the Asp718 digestion shared a common 8.4 kb insert of C. glutamicum chromosomal DNA, while the HindIII-derived plasmids shared a 1.7 kb chromosomal DNA fragment. Subclones were prepared in plasmids pUC18 and pUC19 and the complete DNA sequences of both strands were determined. The nucleotide sequence data were submitted to GenBank (EMBL) and assigned accession number Y13221.

Figure 2

Schematic presentation of the organization of the C. glutamicum glnA gene. A box marks the position of the DNA fragment amplified by PCR using the degenerated primers. The location of the putative terminator (T) and restriction sites for Asp718 (A), BamHI (B), HindIII (H), NheI (N), SalI (S), and StuI (St) are indicated. The Asp718 site is located approximately 7.5 kb upstream of the glnA gene.

Southern blot hybridizations using the 401 bp PCR fragment as a probe confirmed that the isolated DNA insert was derived from C. glutamicum ATCC 13032 and that no chromosomal rearrangements within the cloned DNA fragment had occurred (data not shown).

3.2 Sequence analysis

The deduced start of the glnA gene is a GTG codon at position 1274 of the isolated DNA fragment. This is the first possible start site for glnA in the cloned sequence and it is strongly favored by multiple alignments. From the deduced amino acid sequence several conserved amino acid regions were identified (Fig. 1; shaded in gray) which comprise to the following functions [12, 32]: region I is responsible for subunit interaction with regions II–V for the formation of the active site of the enzyme. Region II is involved in the binding of Mn2+, together with region IV, which additionally resembles a glutamate binding motif. Region III is responsible for nucleotide binding, the function of region V is unknown and region VI resembles an adenylation motif with Tyr-405 as the adenylation site. Twenty-two nucleotides downstream of the TAA stop codon (position 2705–2707) a stem-loop structure was found showing the typical properties of a rho-independent terminator. The free energy of this terminator spanning nucleotides 2730–2769 is ΔG°′−35.3 kJ mol−1.

3.3 Deletion of glnA

For the construction of a glnA deletion strain the strategy described by Schäfer et al. [26] was applied (for restriction sites, see Fig. 2). A 0.5 kb BamHI fragment carrying the 3′ end of the glnA gene was isolated from plasmid pUC540 and ligated to BamHI-digested and dephosphorylated plasmid pUC1.5a carrying the upstream region and the start of glnA. The resulting plasmid pUCΔglnA contained a glnA gene with an internal in-frame deletion of 0.9 kb BamHI fragment. A 1.4 kb NheI fragment carrying this deletion was ligated to XbaI-restricted and dephosphorylated plasmid pK19mobsacB. The resulting plasmid pK19ΔglnA was transferred to C. glutamicum via conjugation with E. coli strain S17-1. Since this plasmid does not replicate in C. glutamicum the resulting kanamycin-resistant clones carried the plasmid integrated in the chromosome by homologous recombination. Due to the sacB gene, encoding levan sucrase, these clones were not able to grow on medium containing 10% sucrose. To select double cross-over events cells were grown under non-selective conditions and subsequently screened for sucrose resistance resulting from the loss of the integrated plasmid. This second recombination event either restored the wild-type situation or resulted in glutamine-auxotrophic cells (deletion mutant). The deletion was verified by PCR and Southern blotting (data not shown). In the deletion mutant, glutamine synthetase activity was abolished (Table 2). The growth defect and the loss of GS activity could be complemented by plasmid pJCglnA carrying the glnA gene (Table 2). In this case, MJ4-26pJCglnA showed an approximately 5-fold higher enzyme activity compared to the wild-type due to the higher gene dosage of the glnA gene.

3.4 Indication for glnII

Southern blot experiments were carried out using a 0.8 kb HincII/NaeI fragment from the glnII gene of S. viridochromogenes as a probe. Specific signals were observed when hybridization was carried out at 55°C (Fig. 3). Neither the 3.7 kb BamHI fragment nor the 5.1 kb SalI fragment showed cross-reaction with the glnA probe (data not shown). Additionally, a short DNA sequence was isolated in an unrelated plasmid rescue project which exhibited similarity to different glnII genes (data not shown).

Figure 3

Southern blot. Chromosomal DNA of C. glutamicum wild-type ATCC 13032 was digested with BamHI (2) and SalI (3), resulting DNA fragments were separated by agarose gel electrophoresis and probed with a 0.8 kb HincII/NaeI DNA fragment of the S. viridochromogenes glnII gene. (1) DIG-labelled marker VII (Boehringer); from top to bottom 8576, 7427, 6106, 4899, 3639, 2799, 1953, 1882, 1515, 1482, 1164, and 992 bp.

4 Discussion

In this study we present data on the cloning and sequencing of the C. glutamicum glnA gene encoding glutamine synthetase I. The amino acid sequence deduced from this gene comprises the five typical motifs conserved in GS proteins throughout bacteria [12, 32]. Additionally, an adenylation motif was found with Tyr-405 as the putative adenylation site, strongly favoring this regulation mechanism rather than ADP-ribosylation. Multiple alignments revealed that the C. glutamicum GSI is very similar to that of M. tuberculosis (GenBank accession number Q10377) with 70% identical amino acids over the whole protein. Mycobacterium belongs, together with Corynebacterium, Gordona, Nocardia, and Rhodococcus, to the group of the mycolic acid-containing actinomycetes [4]. The overall amino acid identity of GSI with the corresponding proteins of the actinomycetes S. coelicolor[35] and S. viridochromogenes[11] was 66% and 65%. Besides to GSI from actinomycetes, this enzyme is phylogenetically more related to the proteins from Gram-negative bacteria like Salmonella typhimurium[13], E. coli[5, 22], and Vibrio alginolyticus[20] showing 52, 51, and 51% identity, respectively. Less identity is found with low G+C Gram-positive bacteria like Bacillus cereus[23] and B. subtilis[31] with 43 and 39% identical amino acids.

Insertion and deletion experiments revealed that GSI plays a pivotal role in the nitrogen metabolism of C. glutamicum. Inactivation of the corresponding gene, glnA, by disruption or deletion led to a complete loss of glutamine synthetase activity. However, in S. hygroscopicus and S. viridochromogenes two genes, encoding GSI and GSII, respectively, were found [3, 11, 18]. A putative glnII gene was also reported for M. tuberculosis (GenBank accession number Q10378). On the basis of the close phylogenetic relationship, also confirmed by the high glnA homology found in this study, a glnII gene is also expected in C. glutamicum. In fact, Southern blot data and a short DNA sequence from a different sequencing project hint at the presence of a gene encoding putatively GSII. The role of this enzyme and its regulation is unclear at the moment and will be the subject of further investigations.


A plasmid containing the glnII gene of S. viridochromogenes was kindly provided by W. Wohlleben (Tübingen). This work was supported by the Fonds der Chemischen Industrie.


  • 1Institut für Biochemie, Universität zu Köln, Cologne, Germany.

  • 2 Institut für Biochemie, Universität zu Köln, Cologne, Germany.


View Abstract