Molecular Characterization of the Glycerol-Oxidative Pathway of Clostridium butyricum VPI 1718

Abstract

The glycerol oxidative pathway of Clostridium butyricum VPI 1718 plays an important role in glycerol dissimilation. We isolated, sequenced, and characterized the region coding for the glycerol oxidation pathway. Five open reading frames (ORFs) were identified: dhaR, encoding a putative transcriptional regulator; dhaD (1,142 bp), encoding a glycerol dehydrogenase; and dhaK (995 bp), dhaL (629 bp), and dhaM (386 bp), encoding a phosphoenolpyruvate (PEP)-dependent dihydroxyacetone (DHA) kinase enzyme complex. Northern blot analysis demonstrated that the last four genes are transcribed as a 3.2-kb polycistronic operon only in glycerol-metabolizing cultures, indicating that the expression of this operon is regulated at the transcriptional level. The transcriptional start site of the operon was determined by primer extension, and the promoter region was deduced. The glycerol dehydrogenase activity of DhaD and the PEP-dependent DHA kinase activity of DhaKLM were demonstrated by heterologous expression in different Escherichia coli mutants. Based on our complementation experiments, we proposed that the HPr phosphoryl carrier protein and His9 residue of the DhaM subunit are involved in the phosphoryl transfer to dihydroxyacetone-phosphate. DhaR, a potential regulator of this operon, was found to contain conserved transmitter and receiver domains that are characteristic of two-component systems present in the AraC family. To the best of our knowledge, this is the first molecular characterization of a glycerol oxidation pathway in a Gram-positive bacterium.

INTRODUCTION

Glycerol can be utilized as a carbon source by bacteria via several metabolic pathways that convert glycerol to dihydroxyacetone-phosphate (DHAP) before DHAP enters the glycolytic pathway. Under aerobic conditions, Escherichia coli phosphorylates glycerol using an ATP-dependent glycerol kinase (19), and glycerol-3-phosphate then is oxidized to DHAP by a membrane-bound FAD-dependent glycerol-3-phosphate dehydrogenase (35). Under anaerobic conditions, Klebsiella pneumoniae and Citrobacter freundii oxidize glycerol using a soluble NAD⁺-dependent glycerol dehydrogenase (9, 13), and dihydroxyacetone (DHA) then is phosphorylated by a DHA kinase to DHAP (9, 13, 21). DHA kinases can be grouped into two structurally related families according to the source of the high-energy phosphate: ATP or phosphoenolpyruvate (PEP). ATP-dependent DHA kinases are single-polypeptide two-domain proteins (37), while the PEP-dependent DHA kinases consist of three subunits: DhaK, DhaL, and DhaM (18). DhaK and DhaL are homologous to the amino-terminal K domain and the carboxy-terminal L domain of the ATP-dependent kinases. DhaK contains a binding site for DHA, and DhaL contains an ADP binding site. DhaM is a phosphohistidine protein that transfers phosphoryl groups from a phosphoryl carrier protein of the phosphotransferase system (PTS) (HPr or enzyme I) to the DhaL-ADP complex (3, 18).

In E. coli, genetic and biochemical studies have demonstrated that the dha operon is controlled by DhaR and the two kinase subunits, DhaK and DhaL (4, 5). DhaK and DhaL act antagonistically; DhaK functions as a corepressor and DhaL as a coactivator of DhaR (4). In the presence of DHA, when the phosphoryl group is transferred from DhaL::ATP to DHA, the now-dephosphorylated DhaL::ADP binds to the DhaR receiver domain and activates the expression of the dha operon. In the absence of DHA, DhaL::ADP is rephosphorylated by DhaM to DhaL::ATP, which does not bind to DhaR (4).

Clostridium butyricum VPI 3266 can convert glycerol reductively to 1,3-propanediol and oxidatively via DHAP to acetate and butyrate (14, 31). The physiology of cells metabolizing glycerol has been studied in chemostat cultures (30). Glycerol consumption is associated with the induction of (i) a glycerol dehydrogenase and a dihydroxyacetone kinase that feed glycerol into the central metabolism (30), and (ii) a B12-independent glycerol dehydratase and an NAD⁺-dependent 1,3-propanediol dehydrogenase involved in propanediol formation (27). Although the molecular characterization of the 1,3-propanol production pathway has provided insight into anaerobic glycerol metabolism in C. butyricum, both the identity and the regulation of the genes involved in glycerol oxidation remain to be elucidated.

In this paper, we report the cloning, sequencing, and molecular characterization of the genes encoding glycerol dehydrogenase and DHA kinase in C. butyricum VPI 1718. Furthermore, we demonstrate that the C. butyricum DHA kinase is PEP dependent and obtains its phosphoryl group from the HPr phosphoryl carrier protein. To the best of our knowledge, this is the first molecular characterization of genes involved in a glycerol oxidation pathway in a Gram-positive bacterium.

MATERIALS AND METHODS

Bacterial strains and plasmids.

All bacterial strains and plasmids used or derived from this study are listed in Table 1. E. coli BW25113 ΔgldA ΔglpK, derived from the Keio collection, was constructed by (i) combination of mutations using P1 transduction (32) and (ii) removal of the Km^r marker by FLP recombinase (10). The E. coli BW25113 ΔgldA ΔglpK ΔdhaKLM strain was constructed using the gene deletion method previously described (10): (i) replacing the dhaKLM genes with a Km^r marker in the E. coli BW25113 ΔgldA ΔglpK strain and (ii) removing the Km^r marker by using FLP recombinase. The E. coli BW25113 ΔgldA ΔglpK ΔdhaKLM ΔptsH::km strain was constructed using the P1 transduction of E. coli BW25113 ΔgldA ΔglpK ΔdhaKLM with a P1 lysate of the E. coli BW25113 ΔptsH::km^R strain from the Keio collection.

Table 1.

Bacterial strains and plasmids used in this study

Strain or plasmid	Relevant characteristic(s)a	Reference or source
Strain
Clostridium butyricum VPI 1718	Obtained from Virginia Polytechnic Institute in 1990; identical to VPI 3266, except that it is missing pCB101 and pCB102	30
Escherichia coli DH5α	supE44ΔlacU169 (φ80 lacZΔM15) hsdR17 recA1 endA1 gyrA96 thi-1 relA1	New England Biolabs (Beverly, MA)
Escherichia coli BW25113 ΔgldA::km	Deletion of the gldA gene that is replaced by a Kmr marker	2
Escherichia coli BW25113 ΔglpK::km	Deletion of the glpK gene that is replaced by a Kmr marker	2
Escherichia coli BW25113 ΔptsH::km	Deletion of the ptsH gene that is replaced by a Kmr marker	2
Escherichia coli BW25113 ΔgldA ΔglpK	Deletion of the gldA and glpK genes	This study
Escherichia coli BW25113 ΔgldA ΔglpK ΔdhaKLM	Deletion of the gldA, glpK, and dhaKLM genes	This study
Escherichia coli BW25113 ΔgldA ΔglpK ΔdhaKLM ΔptsH::km	Deletion of the gldA, glpK, dhaKLM and ptsH genes; ptsH is replaced by a Kmr marker	This study
Plasmids
pUC18	Ampr, origin ColE1; 2.7 kb; used for construction of the genomic libraries	42
pGEM-T	Ampr, origin ColE1; 3 kb; used for PCR product cloning	Promega
pPS1-2	pGEMT + 550-bp insert of the PCR-amplified C. butyricum glycerol dehydrogenase gene	This study
pPSH1	pUC18 + HindIII fragment (2 kbp) containing the C. butyricum glycerol dehydrogenase gene	This study
pPSB10	pUC18 + BglII fragment (10 kbp) of C. butyricum containing the dhaDKLM operon	This study
pSE13	pUC18 + EcoRI fragment (7.5 kbp) of C. butyricum containing the upstream sequence of the dhaDKLM operon	This study
pSOS95 pSOS95del	Apr, MLS (Em)r, acetone operon, repL gene, ColE1 origin pUC18 + replicon, and MLSr marker	12, 39
pVO K1-3	pUC18 + replicon and MLSr marker + C. butyricum synthetic dhaDKLM operon under the control of the thiolase promoter	This study
pVO K1-3 ΔD1	pUC18 + replicon and gene MLSR + C. butyricum synthetic dha DKLM operon without dhaD	This study
pVO K1-3 ΔD1 H10A	Same as pVOK1-3 ΔD1, but with a mutation of dhaM that replaces the histidine 9 residue of DhaM with an alanine residue	This study

^aMLS, macrolides-lincosamides-streptogramin B.

The cloning of the dhaDKLM genes was performed by the PCR amplification of genomic DNA using the Expand Long Template PCR system. A pair of primers that introduced the BamHI and SfoI restriction sites upstream and downstream of dhaDKLM (GDH1 and GDH2) (Table 2) was used. The amplified fragment of 3.2 kb was subcloned into the pGEM-T Easy vector (Promega, Charbonnières, France) and sequenced to ensure that no mutations were introduced. The 3.2-kb BamHI/SfoI fragment cut from the vector was ligated to the 5-kb BamHI/SfoI fragment of pSOS95, yielding the 8.2-kb plasmid pVOK1-3 (Table 1). The inactivation of dhaD in pVOK1-3 was done by the elimination of the 600-bp XbaI-NcoI fragment, yielding the 7.6-kb plasmid pVOK1-3 ΔD1 (Table 1). This construct contains the constitutive thlA (thiolase) promoter from C. acetobutylicum so that the operon can be expressed in both E. coli and C. acetobutylicum (27).

Table 2.

Sequences of primers used in this study

Primer	Sequence (5′–3′)	Description
GDH1	GAAGATCTTCGGATATAAAGGAGATATTGATATGAG	Forward primer for cloning of dhaD, dhaK, dhaL and dhaM
GDH3	TAACCCGGGATAATCCTAAATTTTATTTAGTTTCATCCC	Reverse primer for cloning of dhaD, dhaK, dhaL and dhaM
Pat1	AAAAC(A/T)(T/C)T(A/T)GATAC(A/T)GC(A/T)AAAGC	Forward primer for the PCR amplification of a dhaD probe
Pat2	CC(A/G)AA(A/T)GC(A/T)ACTTTTTC(A/T)CC(A/G)TG	Reverse primer for the PCR amplification of a dhaD probe
mutH10-D	GTAATAGTATCAGCTAGTGATTTGG	Forward primer for site directed mutagenesis of dhaM
mutH10-R	CCAAATCACTAGCTGATACTATTAC	Reverse primer for site directed mutagenesis of dhaM
dhaPE	CCTTTACACGTGTAACATCATC	Primer for primer extension analysis of the dhaDKLM operon
dhaD1-D	ATGAGAAAAGCATTTATTTGCCCAAC	Forward primer for northern hybridization of dhaD
dhaD1-R	GCTTGTACTTCTACCTATCTTATCAGC	Reverse primer for northern hybridization of dhaD
dhaK3-D	ATGGTTGGAATTGTAATAGTATCACATAG	Forward primer for northern hybridization of dhaM
dhaK3-R	tTTTATTTAGTTTCATTTCCCTCTAGCG	Reverse primer for northern hybridization of dhaM
dhaKLM-D	ATGAAAAAATTGATCAATGATGTGCAAGACGTACTGGACGAACAACTGGCAGGACTGGCGAAA GCGCATCCATCGCTGACACTGCATCAGGATCCGGCATATGAATATCCTCCTTAG	Forward primer for dhaKLM deletion
dhaKLM-R	TTAACCCTGACGGTTGAAACGTTGCGTTTTAACGTCCAGCGTTAGCGTTTCTTCTGGTTGTATCGCATACAGTTTCTCACCCTGCTGGCAAATCCAGCCAGCCAATCCCCAGTTCACGGGCGTGTAGGCTGGAGCTGCTTCG	Reverse primer for dhaKLM deletion

The histidine residue at position 9 of DhaM was mutated to an alanine residue using the QuikChange site-directed mutagenesis kits (Stratagene, La Jolla, CA). pVOK1-3 ΔD1, mutH9-D, and mutH9-R were used as template and mutation primers, respectively, yielding the 7.6-kb plasmid pVOK1-3 Η9Α ΔD1.

Growth conditions.

Batch cultures of C. butyricum VPI 1718 for DNA preparation were incubated anaerobically at 37°C in 2YT (yeast extract, tryptone) medium (32) supplemented with 2% glucose. Continuous cultures (dilution rate of 0.05 h⁻¹) used for Northern experiments (RNA extraction) were performed at 37°C in a phosphate-limited synthetic medium containing (per liter): KH₂PO₄, 0.1 g; KCl, 0.65 g; MgSO₄·7H₂O, 0.2 g; FeSO₄·7H₂O, 0.028 g; NH₄Cl, 1.5 g; CoCl₂·6H₂O, 0.01 g; para-aminobenzoic acid, 8 mg; biotin, 0.04 mg; and J633 antifoam, 0.1 g. Either glucose (30 g) or glycerol (27 g) plus glucose (3 g) was used as the carbon and energy source. The pH was automatically maintained at 6.0 by the addition of 6 M NH₄OH. E. coli strains were routinely grown aerobically at 37°C in Luria-Bertani (LB) medium (32) supplemented with ampicillin (100 μg/ml⁻¹) or erythromycin (200 μg/ml⁻¹) when needed.

For enzyme assays and heterologous gene expression, recombinant E. coli BW25113 ΔgldA ΔglpK ΔdhaKLM(pVOK1-3), E. coli BW25113 ΔgldA ΔglpK ΔdhaKLM(pVOK1-3 ΔD1), and E. coli BW25113 ΔgldA ΔglpK ΔdhaKLM(pSOS95del) were grown anaerobically in medium containing (per liter): tryptone, 10 g; yeast extract, 5 g; HEPES, 2.3 g; FeSO₄, 50 mg; nitrilotriacetic acid (NTA), 200 mg; K₂HPO₄, 0.5 g; NaCl, 2 g; DHA, 2 g; and NaNO₃, 0.85 g. The medium also contained ampicillin (100 μg/ml⁻¹). The pH of the medium was adjusted to 7.3 by the addition of 6 M NH₄OH.

For the growth complementation of different E. coli mutants, M9 DHA (10 mM) and/or M9 glycerol (10 mM) liquid medium supplemented with ampicillin (100 μg.ml⁻¹) was used (32).

Nucleic acid isolation and manipulation.

Chromosomal DNA of C. butyricum VPI 1718 was extracted as described previously (7). The isolation of plasmids from E. coli DH5α was performed using Qiaprep spin mini- and midiprep columns (Qiagen, Courtaboeuf, France). Restriction and modification enzymes were purchased from New England BioLabs (Ozyme; Saint Quentin, France) or Gibco BRL (Eragny, France) and used according to the recommendations of the manufacturer. DNA fragments were extracted from agarose gels using the Qiaquick system (Qiagen). Total RNA of C. butyricum VPI 1718 was extracted from continuous cultures using the RNeasy midi kit (Qiagen).

Hybridization.

Chromosomal DNA of C. butyricum VPI 1718 was digested to completion with restriction enzymes corresponding to suitable cloning sites in pUC18, and the resulting fragments were separated by agarose gel electrophoresis in 0.5× TAE buffer (20 mM Tris, 20 mM acetate, 0.5 mM EDTA). Southern blotting (32) was performed by capillary transfer to Hybond-N⁺ membranes (Amersham Pharmacia Biotech, Les Ulis, France) and the fixation of the separated DNA fragments. To create a DNA probe for Southern blot hybridization, a 1.8-kbp internal fragment of the C. butyricum VPI 1718 dhaD and dhaK genes was obtained after HindIII digestion of pSPB10. The probe was radiolabeled with [α-³²P]dATP (specific activity, 3,000 Ci/mmol; Amersham Pharmacia Biotech) by the random-primer method using the MegaPrime DNA labeling system (Amersham Pharmacia Biotech). Stringent conditions were used for hybridization. Prehybridization (1 h) and overnight hybridization reactions were performed at 65°C in 6× SSC (1× SSC is 0.15 M NaCl plus 0.015 M sodium citrate)-2× Denhardt solution (1× Denhardt solution is 0.02% Ficoll-0.02% polyvinylpyrrolidone-0.02% bovine serum albumin)-0.1% sodium dodecyl sulfate (SDS) buffer. Washing steps were performed at room temperature for 15 min in 1× SSC-0.1% SDS buffer and then in 0.1× SSC-0.1% SDS buffers. Hybridization signals were detected by autoradiography using Kodak X-Omat films at −80°C with intensifying screens.

For Northern blot experiments, total RNA of C. butyricum VPI 1718 was separated on a 1% (wt/vol) denaturing formaldehyde-agarose gel and transferred to a nylon membrane (Nytran; Schleicher and Schuell, Inc., Keene, NH) as previously described (32). A 993-bp fragment of dhaD and a 384-bp fragment of dhaM were amplified by PCR using C. acetobutylicum ATCC 824 genomic DNA as a template. Primers used in this analysis are described in Table 2. The PCR fragments were labeled with [α-³²P]dATP (specific activity, 3,000 Ci/mmol; Amersham Pharmacia Biotech) using the Megaprime DNA labeling system (Amersham Pharmacia Biotech) and were used as probes in hybridization. Prehybridization (for 1 h) and overnight hybridization were performed at 68°C in a 1 mM EDTA, 7% (wt/vol) SDS, and 0.5 M Na₂HPO₄ (pH 7.2) solution. Washing steps were performed at room temperature twice in 2× SSC, 0.1% (wt/vol) SDS buffer for 15 min and twice in 0.1× SSC, 0.1% (wt/vol) SDS buffer for 15 min.

Construction and screening of gene libraries.

Partial genomic libraries were constructed from chromosomal DNA of C. butyricum VPI 1718 that was completely digested with restriction enzymes. The 6- to 9-kbp fraction of EcoRI fragments and the 10- to 15-kbp fraction of BglII were agarose purified. Once ligated with EcoRI- or BamHI-digested and dephosphorylated pUC18, the preparation was used to transform competent E. coli DH5α obtained by the method previously described (20). The mixture then was spread on LB agar plates supplemented with ampicillin, 5-bromo-4-chloro-3-indolyl-β-d-galactopyranoside (X-Gal), and isopropyl-β-d-thiogalactopyranoside (IPTG). The transformants were grown overnight at 37°C and then transferred to Hybond-N⁺ membranes (Amersham Pharmacia Biotech) by replica plating, after which they were lysed by alkaline treatment as recommended by the manufacturer. DNA was fixed by being heated at 80°C for 2 h. The membranes then were screened by hybridization as described for Southern blot experiments. Positive clones were tested by restriction analysis and sequencing reactions.

DNA sequencing.

Both strands of DNA were sequenced using the dideoxy chain termination method as described previously (33) with M13 reverse or universal primers or synthetic oligonucleotide primers derived from the sequence obtained from Genome Express (Grenoble, France).

Determination of the transcription start site.

Primer extension reactions were performed as described previously (17), except that superscript reverse transcriptase (Promega) was employed. The dhaPE oligonucleotide (Table 2) complementary to the 5′ end of the dhaD transcript was end labeled with [γ-³²P]ATP (specific activity, 5,000 Ci/mmol; Amersham Pharmacia Biotech) using T4 polynucleotide kinase (Amersham Pharmacia Biotech). The cDNA was analyzed on an 8% polyacrylamide sequencing gel. To map the exact transcriptional start site, the cDNA was electrophoresed in a lane next to a standard sequencing reaction that was prepared using the same oligonucleotide.

DNA and amino acid analyses.

DNA and amino acid analyses were performed using the Vector NTI program (Invitrogen, Cergy Pontoise, Paris). Sequence comparisons and homology searches were performed using BLAST (1) and PRODOM (8).

Preparation of cell extracts.

C. butyricum VPI 1718 and E. coli extracts were prepared by sonication using the entire anaerobic procedure described previously (40).

Enzyme activity assay.

Glycerol dehydrogenase activity was measured by following the procedure described previously (29). The glycerol-dependent formation of NADH was spectrophotometrically monitored at 340 nm. The 1-ml assay mixture contained 30 mM ammonium sulfate, 100 mM potassium carbonate buffer, pH 9.0, 0.06 mM NAD⁺, and 100 mM glycerol, and the reaction was initiated by adding the extract.

DHA kinase activity was measured using the procedure described previously (21). NADH consumption was spectrophotometrically measured at 340 nm using a coupled system in which the DHAP produced by the DHA kinase was reduced by the NADH-dependent glycerol-3-phosphate dehydrogenase. The 1-ml assay mixture contained 1 mM DHA, 1 mM MgCl₂, 0.1 mM NADH, 10 U glycerol 3-phosphate dehydrogenase, 100 mM potassium carbonate buffer, pH 9.0, and 2 mM dithiothreitol. To measure ATP-dependent DHA kinase activity, 1 mM ATP was added. To measure PEP-dependent DHA kinase activity, 1 mM PEP and 100 μM ADP were added. The reaction was started by the addition of DHA.

Nucleotide sequence accession number.

The sequence data reported here have been submitted to the GenBank database and assigned accession number AY138581.

RESULTS

Cloning of the DNA region encoding enzymes of the glycerol oxidation pathway.

C. butyricum VPI 1718 is known to oxidize glycerol through a glycerol dehydrogenase and a DHA kinase (30). Based on highly conserved domains of glycerol dehydrogenases (NH2-gly-gly-gly-lys-thr-leu-asp-thr-ala-lys-ala-COOH [around amino acid residue 100] and NH2-his-gly-glu-lys-val-ala-phe-gly-COOH [around amino acid residue 275]) (9, 11, 24, 38), we designed degenerate oligonucleotides PAT1 and PAT2 (Table 1) to obtain a PCR-amplified 550-bp DNA fragment (pPS1-2). Partial BglII and EcoRI genomic libraries were constructed in pUC18 and screened with pPS1-2 to obtain pPSB10 (10-kb insert) and pSE13 (7.6-kb insert) plasmids, respectively.

Nucleotide sequence analysis.

The pPSB10 10,245-bp insert and the pSE13 7,594-bp insert were sequenced. Homology searches of protein and DNA sequence databases allowed the identification of five open reading frames (ORFs) relevant to the glycerol oxidation pathway (Fig. 1A). The ORFs were identified as dhaR (1,032 bp), encoding a putative bacterial transcriptional activator; dhaD (1,140 bp), encoding a putative glycerol dehydrogenase; and dhaK (993 bp), dhaL (627 bp), and dhaM (384 bp), encoding a putative dihydroxyacetone kinase complex. Interestingly, two inverted-repeat sequences followed by a stretch of T residues, resembling a rho-independent terminator, were found between dhaR and dhaD (centered on base 1474; TCATTtTaatAcAATGA; ΔG = −2.4 kcal·mol⁻¹) and at the end of the dhaM gene (centered on base 4888; AATGCCTGTAgATATTtaatagATATgTACAGGCATT; ΔG = −13.8 kcal·mol⁻¹). The putative genomic arrangement of dhaD, dhaK, dhaL, and dhaM strongly suggests that these genes are organized in an operon.

Fig. 1.

(A) Genomic arrangement of the genes encoding C. butyricum glycerol dehydrogenase (dhaD) and DHA kinase (dhaKLM) and surrounding regions. (B) Promoter region sequence of the dhaDKLM operon. The transcription start site is indicated with a vertical arrow. The putative −35 and −10 regions are overlined. The putative ribosome binding site is underlined, and the putative ATG start codon of the dhaD gene is shown by a bent arrow. (C) Northern blot analysis. Total RNA was isolated from C. butyricum cells grown in phosphate-limited continuous cultures containing either glucose (lane 1) or glucose-glycerol (lane 2) as the substrate.

Amino acid sequence analysis.

The amino acid sequences of the proteins encoded by dhaR, dhaD, dhaK, dhaL, and dhaM were deduced. The amino acid sequence of DhaD of C. butyricum is approximately 47% identical and 64% similar to the bacterial glycerol dehydrogenases of C. freundii (9), E. coli (38), Bacillus stearothermophilus (24), and Clostridium beijerinckii (23).

DhaK amino acid sequence analysis shows high identity to the DhaK subunits of DHA kinases from both E. coli (18) and Lactococcus lactis subsp. lactis (6, 44) (42 and 53% identity, respectively) but low identity to the N-terminal domain of the DHA kinases of C. freundii (9) and Saccharomyces cerevisiae (25) (38 and 39% identity, respectively).

The DhaL amino acid sequence shows 42% identity to the DhaL subunits of the DHA kinases of E. coli (18) and L. lactis subsp. lactis (6, 44) but only 31 and 32% identity to the C-terminal region of the DHA kinases of C. freundii (9) and S. cerevisiae (25), respectively.

The DhaM amino acid sequence is 40% identical to the DhaM subunit of the DHA kinase of E. coli (18) and 36% identical to the DHA kinase of L. lactis subsp. lactis (6, 44). The sequence analyses of DhaKLM revealed that the C. butyricum DHA kinase is a heteromeric enzyme that consists of three subunits that are similar to the PEP-dependent DHA kinases of E. coli (18) and L. lactis subsp. lactis (44).

The dhaR gene encodes a potential transcriptional regulator.

DhaR has a multidomain organization, with N- and C-terminal domains with high similarity to archetypal prokaryotic two-component signal transduction system (TCS) proteins (16, 26). The N-terminal sequence of DhaR (amino acids 1 to 120) has high identity to the N-terminal sensor (or input) domain of the histidine kinase DhaS of Clostridium perfringens (GenBank accession no. NP 561843) and C. butyricum (27, 34) (52 and 46% identity, respectively). The C-terminal domain (amino acids 250 to 344) of DhaR has high identity to the C-terminal output domains of the DhaS-associated DhaA response regulators of C. perfringens (GenBank accession no. NP 561844.1) and C. butyricum (27, 34) (43 and 49% identity, respectively). The C-terminal domain of DhaR contains two helix-turn-helix (HTH) motifs that show sequence similarity to AraC/XylS-type DNA binding proteins found in the TCS response regulator proteins in the other bacteria discussed above (15, 28) (Fig. 2).

Fig. 2.

Modular organization of DhaR. (A) Schematic structure showing the domain organization and special sequence features of the DhaR regulator of C. butyricum. The location of the highly conserved amino acid residues forming the active site of the receiver domain of response regulators are indicated with vertical lines (corresponding to Asp12, Asp13, Asp57, and Lys109 in the CheY protein), whereas predicted α-helices of the C-terminal output and effector domains are indicated in gray boxes. (B) Comparison of the output domain amino acid sequence of DhaR with homologous regions in the AraC/XylS family members AraC (P03021) and MarA (P27246). Residues matching the consensus (bottom) sequence derived from more than 100 AraC/XylS family members (15) are shown in black boxes, whereas additional residues identical in two or more proteins are shown in gray boxes. α-Helices identified in the crystal structure of the DNA-binding domain of MarA are depicted in filled boxes (28). In MarA, the two HTH motifs are separated by an additional helix (helix 4 MarA).

The third domain, or intermediate domain (amino acids 130 to 250), has no homology to proteins of known function but contains the five invariant residues of the N-terminal receiver domains of all response regulators characterized thus far: Asp-140, Asp-141, Asp-191, Thr-221, and Lys-242 of DhaR correspond to Asp-12, Asp-13, Asp-57, Thr-87, and Lys-109 of the CheY protein, respectively (22, 41). DhaR homologs are found in the gene cluster of other organisms using this glycerol oxidation system, like Clostridium botulinum (CLL_0536), Paenibacillus larvae (Plarl_010100008220), Bacillus megaterium (Bmq_1804), and Enterococcus faecalis (Vimsss_355569). However, DhaR of C. butyricum presented no significant homology to DhaR of E. coli.

Transcriptional analysis of the dhaDKLM gene cluster.

Northern blot analyses were carried out using total RNA isolated from cells grown in phosphate-limited continuous cultures containing glucose or a mixture of glucose and glycerol as growth substrates (Fig. 1C). The fermentation product profiles of these two cultures were within the experimental errors of previously published data (30). When dhaD or dhaM probe was used, a transcript around 3.2 kb was detected specifically in the glycerol-metabolizing culture. The size of the transcript was in agreement with the polycistronic dhaDKLM operon inferred by sequence analysis. Furthermore, the signal intensities of the 23S and 16S rRNA control bands in the ethidium bromide-stained agarose gels used for the Northern blot analysis were similar for all cultures (Fig. 1C). The Northern blot results were in good agreement with the glycerol dehydrogenase and DHA kinase activities of the corresponding cell extracts (30), demonstrating that the expression of dhaDKLM is regulated at the transcriptional level.

A primer extension study was performed using total RNA extracted from the glycerol-metabolizing continuous culture to identify the transcription start site of the dhaDKLM operon. A unique transcriptional start site (at a C base) corresponding to position 37 upstream of the dhaD start codon was detected (Fig. 1B). The deduced promoter sequence 5′-ATGAGT-3′ (−35) and 5′-TATAAT-3′ (−10) with a 19-nucleotide (nt) spacing is similar to the σ^A RNA polymerase recognition sequence found in Gram-positive bacteria (43).

Characterization of the glycerol oxidation operon in E. coli.

To express the genes encoding the glycerol oxidative pathway of C. butyricum in E. coli constitutively, the dhaDKLM operon first was PCR amplified and cloned into the shuttle vector pSOS95 containing the thiolase promoter from C. acetobutylicum (27), resulting in pVOK1-3. To examine DHA kinase activity without interference from the glycerol dehydrogenase activity of DhaD, the dhaD gene from pVOK1-3 was partially deleted to obtain the pVOK1-3 ΔD1 plasmid. pVOK1-3, pVOK1-3 ΔD1, and pSOS95 del (used as a control) plasmids were introduced into an E. coli ΔgldA ΔglpK ΔdhaKLM strain that does not possess glycerol dehydrogenase, glycerol kinase, or DHA kinase activity. The growth of three strains was evaluated on M9 glycerol liquid medium (Fig. 3A). Growth was observed only for the E. coli ΔgldA ΔglpK ΔdhaKLM(pVOK1-3) strain. Furthermore, crude extracts from E. coli ΔgldA ΔglpK ΔdhaKLM harboring pVOK1-3, pVOK1-3 ΔD1, or pSOS95 del were assayed for glycerol dehydrogenase activity, and two strains harboring either pVOK1-3 or pSOS95 del were assayed further for DHA kinase activity (Table 3). A glycerol dehydrogenase-specific activity of 288 mU/mg was measured for E. coli ΔgldA ΔglpK ΔdhaKLM(pVOK1-3), whereas specific activities lower than 1 mU/mg were measured in both E. coli ΔgldA ΔglpK ΔdhaKLM(pVOK1-3 ΔD1) and E. coli ΔgldA ΔglpK ΔdhaKLM(pSOS95 del). When PEP was used as a phosphate donor for DHA kinase, a specific activity of 11 mU/mg was measured for E. coli ΔgldA ΔglpK ΔdhaKLM(pVOK1-3 ΔD1), whereas specific activities lower than 1 mU/mg were measured in E. coli ΔgldA ΔglpK ΔdhaKLM and E. coli ΔgldA ΔglpK ΔdhaKLM(pSOS95 del). In contrast, when ATP was investigated as a phosphate donor for DHA kinase, specific activities lower than 1 mU/mg were measured for both strains. To investigate phosphotransfer in the C. butyricum DHA kinase, two new strains were constructed: E. coli ΔgldA ΔglpK ΔdhaKLM ΔptsH(pVOK1-3 ΔD1), with a defect in the synthesis of HPr, and E. coli ΔgldA ΔglpK ΔdhaKLM(pVOK1-3 Η9Α ΔD1), in which the histidine residue at position 9 of DhaK3 was mutated to an alanine residue. The growth of the newly constructed strains was evaluated in M9 DHA liquid medium [together with that of E. coli ΔgldA ΔglpK ΔdhaKLM(pVOK1-3 ΔD1) and E. coli ΔgldA ΔglpK ΔdhaKLM(pSOS95 del)] (Fig. 3B). Growth was observed only for E. coli ΔgldA ΔglpK ΔdhaKLM(pVOK1-3 ΔD1), indicating that HPr and the His9 residue of DhaM are involved in the transfer of a phosphate group from PEP to DHA.

Fig. 3.

Growth curves of different mutants in M9 glycerol and M9 DHA liquid culture. Cultures were performed at 37°C under aerobic conditions in M9 mineral medium supplemented with either 10 mM glycerol (A) or 10 mM DHA (B). ●, E. coli ΔgldA ΔglpK ΔdhaKLM (pVOK1-3); □, E. coli ΔgldA ΔglpK ΔdhaKLM (pVOK1-3 ΔD1); ⧫, E. coli ΔgldA ΔglpK ΔdhaKLM ΔptsH (pVOK1-3 ΔD1); ×, E. coli ΔgldA ΔglpK ΔdhaKLM (pVOK1-3 H9A ΔD1); ▴, E. coli ΔgldA ΔglpK ΔdhaKLM (pSOS95). OD₆₀₀, optical density at 600 nm.

Table 3.

Glycerol dehydrogenase and DHA kinase activities in crude cell extracts of different strains^a

Strain	Activity (mU·mg−1) of:
	Glycerol dehydrogenase	DHA kinase (+ATP)	DHA kinase (+PEP)
BW25113 ΔgldA ΔglpK ΔdhaKLM(pVOK1-3)	288.0 ± 41	ND	ND
BW25113 ΔgldA ΔglpK ΔdhaKLM(pVOK1-3 ΔD1)	0.7 ± 0.18	0.22 ± 0.08	11.0 ± 2.3
BW25113 ΔgldA ΔglpK ΔdhaKLM(pSOS95 del)	0.6 ± 0.16	0.19 ± 0.07	0.9 ± 0.2

^aMean values and standard deviations from three independent experiments are given. ND, not determined.

DISCUSSION

In C. butyricum VPI 1718, four genes are involved in the conversion of glycerol into DHAP: dhaD, dhaK, dhaL, and dhaM. The operon arrangement of the four genes, inferred by DNA sequence analysis, was confirmed by Northern blot experiments (Fig. 1C). The dhaDKLM operon was expressed only in the glycerol-metabolizing culture, indicating that this operon is regulated at the transcriptional level. It is likely that the regulation is mediated by the putative regulator DhaR, which is encoded by the dhaR gene located immediately upstream of the operon in the C. butyricum chromosome. DhaR shows significant homology with transcriptional activators from a subgroup of the AraC/XylS family that belongs to the TCS group. These two-component systems usually consist of two different proteins: a sensor protein with a sensor and a transmitter domain and a response regulator that possesses a receiver and a regulator domain. Signal transduction between the sensor and the response regulator is mediated by phosphorelay; the sensor protein is autophosphorylated at a conserved His residue in the transmitter domain and then transfers the phosphate group to a conserved Asp residue in the receiver domain in the response regulator (26). Surprisingly, DhaR in C. butyricum appears to be a hybrid protein that comprises a sensor domain and an output domain that contains two HTHs similar to the one found in the DNA binding domains of regulator proteins of the AraC/XylS family. A recent study of E. coli demonstrated that DHA kinase controls gene expression by binding to the sensor domain of the DhaR transcription factor (4). In the absence of DHA, DhaK binds to the sensing domain of DhaR and thereby keeps DhaR in a transcriptionally inactive state. In the presence of DHA, DhaL is mostly dephosphorylated and displaces DhaK and activates DhaR.

The heterologous expression of dhaD in E. coli demonstrated that it encodes an NAD⁺-dependent glycerol dehydrogenase. DhaD in C. butyricum is homologous to several other glycerol dehydrogenases from C. freundii (9), E. coli (38), and B. stearothermophilus (24). All of these glycerol dehydrogenases can be categorized into a subclass of type III alcohol dehydrogenases based on the presence of the highly conserved ADH_iron_1 sequence (http://www.expasy.ch/cgi-bin/scanprosite) (Fig. 4).

Fig. 4.

Partial alignments and homology analyses of the glycerol dehydrogenases of C. butyricum, B. stearothermophilus, C. freundii, and E. coli, which all contain the ADH_iron_1 sequence.

The amino acid sequence analysis of DhaK, DhaL, and DhaM of C. butyricum revealed high similarity to DhaK, DhaL, and DhaM of E. coli (18) and L. lactis subsp. lactis (44), respectively, which have been characterized as PEP-dependent DHA kinases. The heterologous expression of the dhaK, dhaL, and dhaM genes restored the growth of a ΔgldA ΔglpK ΔdhaKLM E. coli mutant on a mineral medium supplemented with DHA as a carbon source. A 12-fold increase in PEP-dependent DHA kinase activity was observed in this strain compared to that of the control, while only basal ATP-dependent DHA kinase activity was measured for both strains. To investigate the involvement of PEP-PTS in the DHA kinase reaction, dhaK, dhaL, and dhaM genes were heterologously expressed in a ΔgldA ΔglpK ΔdhaKLM E. coli mutant with an additional deletion of the ptsH gene. As expected, the growth of the mutant on DHA could not be restored, indicating that HPr is involved in the phosphate transfer during the DHA kinase reaction. As E. coli HPr can be phosphorylated only by enzyme I of the PEP-PTS system (36), these results indicate that the dhaK, dhaL, and dhaM genes encode a PEP-dependent DHA kinase.

PEP-dependent DHA kinases have been biochemically characterized only in E. coli (18) and L. lactis subsp. lactis (44). The DhaL subunits of both E. coli and L. lactis possess an ADP molecule associated with two molecules of Mg²⁺ (3, 44). This ADP plays a key role in a phosphoryl transfer between DhaM and DhaK. The DhaM subunit of the E. coli enzyme is very large (472 amino acids) and possesses three His residues potentially involved in the phosphoryl transfer. In contrast, all of the DhaM homologues from Gram-positive bacteria are smaller (<140 amino acids) than that of E. coli and possess only one His residue that has been shown to be involved in phosphoryl transfer in L. lactis subsp. lactis. DhaM of C. butyricum also is small (128 amino acids) and possesses only one His residue (His9), which is consistent with the findings for other Gram-positive bacteria. When this His9 residue was mutated to alanine, the E. coli ΔgldA ΔglpK ΔdhaKLM(pVOK1-3 Η9Α ΔD1) strain was unable to grow on DHA, strongly suggesting that this His9 residue is the phosphorylation site of DhaK3.

A working hypothesis of phosphoryl transfer mediated by DHA kinase of C. butyricum, using PEP as a phosphoryl donor, is presented in Fig. 5. A phosphate group is transferred sequentially from PEP via enzyme I to H15 of HPr, and hence to His9 of DhaM, the ADP cofactor of DhaL, and finally to DHA when bound to the DhaK subunit.

Fig. 5.

PEP-dependent DHA kinase of C. butyricum. Shown is a depiction of a working hypothesis of a phosphoryl group transfer from PEP to DHA mediated by the PTS system and the different DhaK subunits. The phosphate is transferred sequentially from PEP via enzyme I to H15 of HPr and then to H9 from DhaM, the ADP cofactor of DhaL, and finally to DHA bound to the DhaK subunit. Adapted from Bächler et al. (4).

To the best of our knowledge, this is the first molecular characterization of the glycerol oxidation pathway in a Gram-positive bacterium. In future studies, we hope to investigate how DhaR activates the dhaDKLM operon.