The Alcohol Dehydrogenase Gene adhA in Corynebacterium glutamicum Is Subject to Carbon Catabolite Repression^▿

Bacteria, plasmids, oligonucleotides, and culture conditions.

Bacterial strains, plasmids and oligonucleotides, as well as their relevant characteristics and sources, are given in Table 1. TY medium (42) was used as complex medium for C. glutamicum and Escherichia coli. The minimal medium used for C. glutamicum has been described previously (14) and contained ethanol, glucose, or acetate or a mixture of those carbon sources in the concentrations indicated in Results. When appropriate, the medium contained kanamycin (50 μg ml⁻¹ for strains carrying plasmids and 15 μg ml⁻¹ for integration mutants and for transformant selection). C. glutamicum was grown aerobically at 30°C and E. coli at 37°C in 50-ml cultures in 500-ml baffled Erlenmeyer flasks on a rotary shaker at 120 rpm. The growth of the bacteria was followed by measuring the optical density at 600 nm.

DNA preparation and transformation.

The isolation of chromosomal and plasmid DNA from C. glutamicum was performed as described previously (14), and plasmid isolation from E. coli was carried out according to the method of Birnboim (5). DNA transfer into C. glutamicum was performed by electroporation, and recombinant strains were selected on LBHIS agar plates (48). Transformation of E. coli was carried out according to the method of Inoue et al. (27).

PCR techniques.

The PCR experiments were performed in a Biometra personal cycler (Biotron) with Taq polymerase (MBI-Fermentas). The oligonucleotides were obtained from MWG-Biotech or from biomers.net GmbH. The cycling times and temperatures were chosen according to fragment length and primer constitution. The PCR products were separated on agarose gels and purified using a Nucleospin extract kit (Macherey Nagel).

DNA manipulation and Southern hybridization.

Restriction enzymes, T4 DNA ligase, calf intestinal phosphatase, RNase A, and proteinase K were obtained from MBI-Fermentas and used as instructed by the manufacturer. DNA hybridization experiments were performed as described previously (14). An adhA-specific 880-bp DNA fragment was amplified from the chromosomal DNA of wild-type (WT) C. glutamicum, labeled with digoxigenin-dUTP by PCR with primers adhA-SB-hin and adhA-rev, and used as a probe. The labeling, hybridization, washing, and detection were performed by using a nonradioactive DNA labeling and detection kit from Roche Diagnostics, following the manufacturer's instructions.

Construction of a C. glutamicum adhA integration mutant.

The inactivation of the chromosomal adhA gene in C. glutamicum was performed by integration of the vector pDrive into the adhA gene. An internal fragment covering the region between base pair 321 of the 5′ end and base pair 208 of the 3′ end of the adhA gene was generated by using primers adhA-hin and adhA-rev and ligated via the A/T overhangs into pDrive, and the ligation mixture was transformed into E. coli. The identification of positive E. coli clones was performed by blue/white screening on TY agar plates containing kanamycin, isopropyl-β-d-thiogalactopyranoside (IPTG; 50 μM), and 5-bromo-4-chloro-3-indolyl-β-d-galactopyranoside (X-Gal; 80 μg/ml). Recombinant plasmids were isolated from E. coli and electroporated into WT C. glutamicum. The integration of pDrive into the chromosomal adhA locus of the resulting strain, C. glutamicum AA1, was verified by Southern blot analysis. For this purpose, a labeled adhA probe was hybridized to HindIII-restricted and size-fractioned chromosomal DNA from WT C. glutamicum and C. glutamicum AA1. The hybridization resulted in one signal of about 10 kb with DNA from the WT strain and in two signals of about 11 kb and 3.3 kb with DNA from the mutant. According to the restriction map of the adhA locus, these sizes were expected.

Cloning of adhA promoter fragments.

The adhAp₁ (401 bp), adhAp₂ (236 bp), and adhAp₃ (213 bp) promoter fragments were amplified from the chromosomal DNA of WT C. glutamicum. The PCR products were digested with SalI and BamHI, ligated into the SalI/BamHI-restricted plasmid pET2, and transformed into E. coli. The recombinant plasmids pET2-adhAp₁, pET2-adhAp₂, and pET2-adhAp₃ were isolated from E. coli and introduced into C. glutamicum.

RNA techniques.

For RNA isolation, C. glutamicum cells were grown in minimal medium containing glucose or ethanol, harvested in the exponential growth phase (optical density at 600 nm of about 4.5), and treated with 1 volume of ice-cold killing buffer (20 mM Tris-HCl, pH 8.0, 20 mM NaN₃, 5 mM MgCl₂). The isolation procedure was performed as described previously (43), and aliquots of the RNA were stored at −70°C until use.

For Northern (RNA) hybridization, an 1,149-bp DNA fragment was amplified from the chromosomal DNA of WT C. glutamicum by PCR with primers adhA-SB-hin and adhA-NB-rev, digested with SalI and XbaI, purified with a Nucleospin extract kit, and used as a probe. The [α-³²P]dATP labeling of the probe was performed with Klenow fragment (MBI-Fermentas). The separation of unincorporated nucleotides was done by using MicroSpin G-25 columns (Amersham Biosciences). For the hybridization, about 10 μg of total RNA from WT C. glutamicum was separated on an agarose gel containing 17% (vol/vol) formaldehyde and transferred onto a nylon membrane (14). Filter hybridization and detection were accomplished with conventional protocols (42).

The primer extension reactions were carried out as described previously (43) with IRD800-labeled primers CM4 and CM5 and RNA from ethanol-grown C. glutamicum(pET2-adhAp₂) cells, and the primer extension products were analyzed with an automatic sequencer (LI-COR 4000L; Licor, Inc.) using a 6% (wt/vol) polyacrylamide gel at 1,500 V and 50°C. For the exact localization of the transcriptional start site, sequencing reactions using plasmid pET2-adhAp₂ and the same oligonucleotide used for the respective primer extension reaction were coelectrophoresed in the sequencing reaction mixture.

Promoter binding assays with His-tagged RamA and RamB fusion protein.

The construction of plasmids for the preparation of His-tagged RamA and RamB and the synthesis and purification of these proteins have been described previously (10, 19). The binding of purified RamA or RamB to the adhA promoter region was tested by DNA electrophoretic mobility shift assays (EMSAs) using the fragments adhAp₁, adhAp₂, and adhAp₃. The EMSA experiments were carried out as described previously (10).

Enzyme assays.

For the determination of enzyme activities in cell extracts, C. glutamicum cells were grown in minimal medium containing the respective carbon source(s) and harvested in the exponential growth phase. Cell extracts were prepared as described previously (3).

ADH activity was analyzed in the acetaldehyde-forming direction as described previously (3). Chloramphenicol acetyltransferase (CAT) activity was determined as described by Schreiner et al. (43).

Computational analysis.

For the analysis of the ΔG°′ value (free energy under standard conditions) of the adhA terminator structures, the program Clone Manager 7 (Sci Ed Central) was used. Databank searches and alignments were carried out using BLAST, CLUSTAL W, and BioEdit software (1, 24, 47).

Identification and characterization of the adhA gene.

We previously found relatively high specific activities of ADH in ethanol-grown cells of C. glutamicum (3). Five genes (cg0273, cg0387, cg0400, cg2714, and cg3107) were annotated as potential ADH genes in the genome of C. glutamicum (GenBank accession numbers NC_003450 and NC_006958) (26, 29). Amino acid sequence alignments of the deduced polypeptides with ADH enzymes from organisms known to utilize ethanol indicated Cg3107 to be the most likely candidate as ADH involved in the ethanol oxidation of C. glutamicum. Cg3107 showed 47% identity to the functionally analyzed ADHs of Pichia stipidis and Saccharomyces cerevisiae, 49% identity to the thermostable ADH of the gram-positive Bacillus stearothermophilus, and 57% identity to the NAD⁺-dependent ADH of the gram-negative Pseudomonas aeruginosa. The other four proteins (Cg0273, Cg0387, Cg0400, and Cg2714) showed less similarity (<40%) to all these ADHs. Therefore, cg3107 was inactivated and the resulting mutant C. glutamicum AA1 was analyzed for growth on ethanol. In fact, this mutant was not able to grow in minimal medium containing ethanol as the single carbon source, whereas its growth on glucose or acetate was not affected (data not shown). For confirmation, the specific ADH activities in cell extracts of WT C. glutamicum and AA1 were determined. In contrast to the WT strain, the mutant did not show any ADH activity on any substrate tested (Table 2). This result and the inability of the mutant to grow on ethanol indicate that cg3107 in fact represents the ADH gene, and therefore, we designated it as adhA.

The adhA gene of C. glutamicum consists of 1,038 bp and is preceded by a typical ribosomal binding site (AGAAGG). A region of dyad symmetry is located 193 bp downstream of the adhA stop codon, and the mRNA hairpin loop predicted from this sequence has a ΔG°′ value of −21.5 kcal/mol at 25°C. This structure indicates transcriptional termination downstream of adhA. According to the nucleotide sequence, adhA encodes a polypeptide of 345 amino acid residues with a predicted molecular mass of 36.8 kDa. Upstream and downstream of adhA, there are genes encoding hypothetical or putative proteins of unknown functions.

Amino acid sequence alignments and BLAST analysis with the C. glutamicum ADH revealed homologous proteins in Corynebacterium efficiens YS-314 (GenBank accession no. CE0053) and Corynebacterium diphtheriae NCTC 13129 (GenBank accession no. DIP2114), showing identities of 72% and 74%, respectively, whereas no significant similarities to proteins of Corynebacterium jeikeium K411 were detected. Further comparisons of the C. glutamicum enzyme with E. coli proteins revealed no significant similarity to the bifunctional AdhE, which physiologically catalyzes ethanol formation from acetyl-CoA (22).

Transcriptional analysis of the adhA gene.

To determine the size of the adhA transcript and to test whether the different specific ADH activities observed in C. glutamicum cells grown on either glucose or ethanol are due to transcriptional control, Northern blot hybridization experiments were performed. As shown in Fig. 1A, the hybridizations revealed their main signals at about 1.5 kb. With RNA from ethanol-grown cells, the signal was much more intense than that obtained with RNA from glucose-grown cells (Fig. 1A). These results indicate that the C. glutamicum adhA gene is monocistronic and that it is subject to carbon source-dependent transcriptional regulation.

To further investigate the carbon source-dependent transcription of the adhA gene, the three promoter fragments adhAp₁, adhAp₂, and adhAp₃ (Fig. 2A) were tested for activity. For this purpose, the plasmids pET2-adhAp₁, pET2-adhAp₂, and pET2-adhAp₃ were transformed into WT C. glutamicum, and the specific CAT activities in the resulting strains were determined during growth in minimal medium containing ethanol, glucose, acetate, and the substrate mixtures ethanol plus glucose or ethanol plus acetate. As shown in Table 3, WT C. glutamicum(pET2-adhAp₁) showed specific CAT activities of 6.89 U/mg protein on ethanol and about fourfold lower activities on glucose or ethanol plus glucose. Interestingly, on acetate or acetate plus ethanol, the specific CAT activities were even about 14-fold lower than on ethanol. In WT C. glutamicum(pET2-adhAp₂) containing the shortened adhA promoter fragment, the specific CAT activities on all substrates were five- to ninefold lower than in WT C. glutamicum(pET2-adhAp₁). WT C. glutamicum(pET2-adhAp₃) showed no CAT activity at all. These results indicate (i) that the adhA promoter resides in the adhA upstream region covered by fragment adhAp₂ (position −218 to +1 with respect to the translational start codon), (ii) that the adhA promoter drives the carbon source-dependent expression of adhA, (iii) that fragment adhAp₃ does not contain a functional promoter, and (iv) that the region covered by fragment adhAp₃ contains an activating element leading in a carbon source-independent manner to an increased adhA promoter activity.

To investigate the carbon source-dependent expression of adhA in more detail, we performed successive growth experiments with C. glutamicum(pET2-adhAp₂) and C. glutamicum(pET2-adhAp₁) in minimal medium containing glucose (1%), a glucose-ethanol substrate mixture (0.125% and 1%, respectively), ethanol (1%), and glucose again (1%) and determined the CAT activities (i.e., adhA promoter activities) in the course of the cultivations. Representative results for C. glutamicum(pET-adhAp₂) are shown in Fig. 3. During growth on glucose, the specific CAT activity was relatively low (Fig. 3A). In the subsequent culture containing glucose plus ethanol, the growth was biphasic; CAT activity was low in the glucose-metabolizing growth phase and increased in the ethanol-metabolizing growth phase to a maximum of about 1 U/mg protein (Fig. 3B). When the strain was then grown on ethanol as the single carbon source, CAT activity was high in the exponential growth and slightly lower in the stationary phase (Fig. 3C). The following cultivation on glucose led to a decrease of the specific CAT activity to the level observed in the first cultivation with glucose as the sole substrate (Fig. 3D). The CAT activities determined during the cultivations with C. glutamicum(pET-adhAp₁) correlated well with those of C. glutamicum(pET-adhAp₂) except that they were five- to ninefold higher under all conditions. The CAT activities observed in all these cultivations also correlate well with the ADH activities of WT C. glutamicum previously observed in identical successive growth experiments (3). Taken together, the results corroborate that the carbon source-dependent regulation of ADH in C. glutamicum and, also, the growth phase-dependent regulation of ADH on ethanol as the sole substrate take place at the transcriptional level.

To precisely localize the transcriptional initiation site of the C. glutamicum adhA gene, primer extension experiments were performed. With the use of oligonucleotide CM4 and 50 μg total RNA from C. glutamicum(pET2-adhAp₂), a signal corresponding to a G residue 130 nucleotides upstream of the adhA translational start codon ATG was observed (Fig. 1B). This result was confirmed in an independent experiment with a different primer (CM5) (data not shown). Centered 10 bp upstream of the adhA transcriptional initiation site, the sequence TAGGGT was found and designated as the −10 region. However, there is only relatively weak similarity to the consensus −10 sequence (TAT/CAAT) of corynebacterial promoters (39, 40). No apparent −35 region could be recognized, which is a common feature of C. glutamicum promoters (39, 40).

RamA and RamB bind to the adhA promoter region.

AK, PTA, ICL, and MS are essential enzymes for acetate metabolism in C. glutamicum and are also required for the growth of C. glutamicum on ethanol (3). The expression of the genes encoding these four enzymes has been shown to be transcriptionally controlled by RamA and RamB (see the introduction). Interestingly, we observed three potential RamA binding sites in the promoter region of adhA, i.e., AGGGGGT, located 164 bp upstream of the transcriptional start, and CGGGGGT, located 112 bp and 57 bp upstream of the transcriptional start (Fig. 2A). Furthermore, two potential RamB binding sites were detected, located 132 bp upstream (TATTTTTCGAAAT; completely conserved nucleotides underlined) and 96 bp downstream (AGCAATTTGCCAA) of the adhA transcriptional start site (Fig. 2A). These observations suggested a direct interaction of RamA and RamB with the adhA promoter region and prompted us to perform EMSAs with His-tagged RamA and RamB proteins and the adhAp₁, adhAp₂, and adhAp₃ fragments (Fig. 2B and C). The incubation of adhAp₁ and adhAp₃ with RamA resulted in a three-step retardation of the probe (Fig. 2B), probably due to the binding of RamA to the three postulated RamA binding sites. The incubation of RamA with adhAp₂ resulted in a one-step retardation of the probe, probably due to the binding of RamA to the single RamA binding site 57 bp upstream of the transcriptional start. The incubation of adhAp₁ with RamB led to a two-step retardation signal (Fig. 2C), whereas the incubation of RamB with adhAp₂ and adhAp₃ led to a one-step retardation of the respective probes. We also tested for an interaction of RamA and RamB during the binding to the adhAp₁, adhAp₂, and adhAp₃ fragments. However, neither combined incubation with RamA and RamB nor the prebinding of RamA or RamB blocked or enhanced the binding of the other regulator (data not shown). These results are in agreement with the postulated RamA and RamB binding sites and indicate a direct participation of RamA and of RamB in adhA gene expression control in C. glutamicum.

RamA and RamB control the adhA gene expression.

To test for a regulatory function of RamA and RamB in adhA expression, the specific ADH activities were determined in the ramB-deficient mutant C. glutamicum RG1 and in the ramA-deficient mutant C. glutamicum RG2 (Table 2). When grown in minimal medium containing ethanol, C. glutamicum RG1 showed about the same specific ADH activity as WT C. glutamicum. When grown on glucose or ethanol plus glucose and, also, when grown on acetate or a mixture of ethanol and acetate, C. glutamicum RG1 showed significantly (two- to fivefold) higher specific ADH activities than WT C. glutamicum. In contrast to C. glutamicum RG1, the RamA-deficient strain RG2 was not able to grow on ethanol, on acetate, or on ethanol plus acetate and showed no ADH activity on the other substrates tested (Table 2).

To study the direct influence of RamA and RamB on adhA promoter activity, the specific CAT activities were determined in the respective mutant strains containing plasmids pET2-adhAp₁ and pET2-adhAp₂ (Table 3). The RamA-negative strains C. glutamicum RG2(pET2-adhAp₁) and C. glutamicum RG2(pET2-adhAp₂) showed no promoter activities at all. However, in C. glutamicum RG1(pET2-adhAp₁), we detected twofold higher specific CAT activities on glucose and on ethanol plus glucose and about threefold higher activities on acetate and ethanol plus acetate than in WT C. glutamicum(pET2-adhAp₁). The specific CAT activities on ethanol were similar in both WT C. glutamicum(pET2-adhAp₁) and C. glutamicum RG1(pET2-adhAp₁) (Table 3). Interestingly, the specific CAT activities in C. glutamicum RG1(pET2-adhAp₂) were similar to the activities determined in WT C. glutamicum(pET2-adhAp₂) on all substrates tested. Taken together, the results indicate an involvement of both RamA and RamB in the expression control of adhA. RamA is absolutely required for the expression of the C. glutamicum adhA gene and, thus, for ADH activity. RamB exerts a negative control when the cells grow in the presence of glucose or acetate as the sole or as an additional carbon source. However, the results also indicate that neither RamA nor RamB alone is responsible for the carbon source-dependent ADH regulation, i.e., for complete adhA repression in the presence of glucose or acetate in the growth medium. It is feasible that different RamA abundance within the cells (and thus, different adhA activation), together with RamB, controls adhA expression during growth on different substrates. Alternatively, C. glutamicum might possess an additional transcriptional regulator(s) controlling adhA expression and, together with RamB, being responsible for catabolite repression in medium containing ethanol and glucose or acetate.