![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
Departamento de Bioquímica, Centro de Neurociências e Biologia Celular, and Departamento de Zoologia, Universidade de Coimbra, 3004-517 Coimbra,1 Centro de Biologia Celular, Departamento de Biologia, Universidade de Aveiro, 3810-193 Aveiro,3 Instituto de Tecnologia Química e Biológica, Universidade Nova de Lisboa, 2780-156 Oeiras, Portugal,4 Department of Microbiology, Cornell University, Ithaca, New York 148532
Received 27 January 2004/ Accepted 22 March 2004
 |
ABSTRACT |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
INTRODUCTION |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Identification of the genes and biosynthetic pathways involved in the synthesis of MG is essential to elucidate the physiological role of this compatible solute in the osmotic stress responses of thermophiles and hyperthermophiles and to understand its puzzling distribution among prokaryotes. Two independent biosynthetic pathways for the synthesis of MG have been elucidated; one involves the direct condensation of GDP-mannose and D-glycerate into MG by mannosylglycerate synthase (MGS). This pathway has only been encountered in R. marinus (29). The alternative pathway channels GDP-mannose and D-3-phosphoglycerate (3-PGA) through the intermediate mannosyl-3-phosphoglycerate (MPG) to MG by mannosyl-3-phosphoglycerate synthase (MPGS), encoded by mpgs, and by mannosyl-3-phosphoglycerate phosphatase (MPGP), encoded by mpgp. This pathway has been found in Pyrococcus spp., T. thermophilus, and, recently, in R. marinus as well. The genes of the two-step pathway are always contiguous in the thermophilic and hyperthermophilic archaea and bacteria that have been examined (6, 12, 13).
The bacterium tentatively named Dehalococcoides ethenogenes is an unusual microorganism whose energy metabolism relies on the reductive dechlorination of chlorinated ethene solvents, which it uses as electron acceptors (30). This strictly anaerobic organism, with an optimum growth temperature near 35°C, represents a deep line of descent in the bacterial phylum Chloroflexi (16). D. ethenogenes and a closely related isolate designated CBDB1 were recovered from sewage sludge (1, 30). Preliminary experiments have shown that D. ethenogenes is capable of growth at NaCl concentrations as high as 0.5 M, near the salinity of seawater (L. Hsu and S. Zinder, unpublished data). Moreover, environmental 16S rRNA gene sequences of organisms that are closely related to D. ethenogenes have also been detected in marine environments, indicating that they are halotolerant or slightly halophilic and may therefore require compatible solutes for osmotic adjustment (22, 23, 37) (GenBank accession nos. AJ431246 and AJ431247). However, the extremely poor growth of D. ethenogenes does not allow the determination of low-molecular-weight intracellular solutes.
The ability to completely dechlorinate polychlorinated compounds prompted the sequencing of the D. ethenogenes genome. A BLAST search of genes involved in the synthesis of MG at the TIGR unfinished-genomes database (http://tigrblast.tigr.org/ufmg/) showed that D. ethenogenes has a gene sequence potentially coding for two functional domains, one with high homology to MPGS (EC 2.4.1.217) and the other to MPGP (EC 3.1.3.70).
The identification of a putative bifunctional gene for the synthesis of MG in D. ethenogenes raises several important questions about the function of the enzyme, the synthesis of MG by the natural host, and the role of this compatible solute in D. ethenogenes. To answer some of these questions, we expressed the gene in Escherichia coli and ultimately in Saccharomyces cerevisiae. Our results show that this gene, designated mgsD, encodes a bifunctional mannosylglycerate synthase (MGSD) that synthesizes MG at low temperatures. Furthermore, expression of the gene in S. cerevisiae led to in vivo accumulation of MG, strongly suggesting that the function of the gene product is to synthesize MG in D. ethenogenes. This is the first report of a bifunctional gene and enzyme for the synthesis of MG in a mesophilic bacterium and is a novelty of evolutionary relevance.
|
MATERIALS AND METHODS |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Saccharomyces cerevisiae strain CEN.PK2 (MATa/MATb
ura3-52/ura3-52 trp1-289/trp1-289 leu2-3,112/leu2-3,112 his3
1/his31) and a yeast shuttle vector derived from pRS425 (ATCC 77106) obtained from LGC Promochem (American Type Culture Collection, Barcelona, Spain; http://www.lgcpromochem.com/atcc/) were used for gene expression. S. cerevisiae cells were grown in YPD medium (10 g of yeast extract, 20 g of peptone, and 20 g of glucose per liter), transformed by the lithium acetate protocol (17), and recombinants selected on MM-LEU medium (6.7 g of yeast nitrogen base without amino acids [Difco] per liter, 20 g of glucose per liter, and a leucine drop-out mixture of amino acids). The cultures were grown in 1-liter baffled metal-capped Erlenmeyer flasks containing 250 ml of medium at 30°C in an orbital shaker at 160 rpm.
Isolation of DNA from D. ethenogenes and PCR. DNA was extracted from D. ethenogenes with phenol-chloroform-isoamyl alcohol and purified with a Bio 101 (Vista) Geneclean spin kit as previously described (15). All PCR amplifications were carried out with 2.5 U of Pfu Turbo DNA polymerase (Stratagene) and the conditions described elsewhere (13).
Identification and analysis of mgsD and flanking genes in the D. ethenogenes genome. BLAST searches were carried out at the Institute for Genomic Research unfinished genomes database (http://tigrblast.tigr.org/ufmg/) with amino acid sequences from Pyrococcus horikoshii MPGS and MPGP (PhMPGS, GenPept BAA30023; PhMPGP, BAA30022). Preliminary sequence data were obtained from TIGR at http://www.tigr.org. All open reading frames (ORFs) surrounding mgsD were identified by BLAST with the translated sequences at the National Center for Biotechnology Information (http://www.ncbi.nlm.nih.gov/BLAST).
Cloning the bifunctional mgsD and overexpression in E. coli. D. ethenogenes genomic DNA was used as the template to amplify mgsD with primers DET1 and DET2 (Table 1), constructed based on the nucleotide sequence obtained from TIGR. The PCR product was purified after digestion with EcoRI and HindIII and ligated into the corresponding sites of pKK223-3, yielding pDET. The insert was sequenced on both strands (AGOWA, Berlin, Germany) for confirmation purposes.
|
Construction of monofunctional enzymes and expression in E. coli. The 694-amino-acid sequence of the bifunctional enzyme was aligned with the MPGS and MPGP sequences from the Pyrococcus sp. and T. thermophilus with the ClustalW Service at the European Bioinformatics Institute (http://www.ebi.ac.uk/clustalw). The sequence encoding the D. ethenogenes MPGS domain (DeMPGS) of the bifunctional MGSD was amplified by primers DET1 and DETS (Table 1), the latter designed to introduce a stop codon 3' of Leu399, excluding the amino acid linker region between MPGS and MPGP, without homology to either domain (see Fig. 2). The alternative primer DETS-L (Table 1) was designed to introduce a stop codon 3' of Lys417. DET1 and DETS-L were used to amplify the MPGS gene, including the linker region (DeMPGS-L), comprising 417 amino acids. Two additional mpgs variants were amplified with primers DETS-450 and DETS-549 (Table 1), designed to introduce stop codons 3' of Leu450 and 3' of Ser549, that were used with DET1 in separate reactions to create genes encoding 450- and 549-amino-acid polypeptides (DeMPGS-450 and DeMPGS-549, respectively). All engineered mpgs genes were cloned into pKK223-3 and transformed into E. coli BL21-Rosetta as described above.
|
Preparation of cell extracts. All cultures of recombinant E. coli were centrifuged (7,000 x g, 10 min, 4°C) and suspended in 25 mM Tris-HCl (pH 7.5) containing 5 mM MgCl2, DNase I (10 µg/ml), and protease inhibitors (13). Cells were disrupted twice by French press (120 MPa), followed by centrifugation (20,000 x g, 30 min, 4°C). The supernatants were filtered through a 0.22-µm filter and used for purification of the enzymes. S. cerevisiae cells were collected by centrifugation (5,000 x g, 10 min, 4°C) and suspended in the buffer described above. Cells were disrupted and centrifuged as described above, the cell debris was discarded, and the supernatant was used for the enzymatic assays described below.
Enzyme assays. The reaction mixture (100 µl) used to detect enzyme activity of the recombinant MGSD in E. coli cell extracts and to monitor activity during purification contained 25 µl of sample, 2.5 mM GDP-mannose, and 2.5 mM 3-phosphoglycerate (3-PGA, sodium salt) in 25 mM Tris-HCl (pH 7.5) with 20 mM CoCl2. The mixture was incubated at 30°C for 30 min and cooled on ice. The assay for detection of the activity of recombinant DeMPGS was as described above, followed by incubation at 90°C for 5 min with 2 µg of pure recombinant MPGP from T. thermophilus HB27 (13).
The assay for purification of DeMPGP was based on the dephosphorylation of MPG previously synthesized by T. thermophilus MPGS (13). The reaction mixture contained 25 µl of sample, 2 mM MPG, 25 mM Tris-HCl (pH 7.5), and 25 mM CoCl2. The mixture was incubated at 30°C for 30 min and cooled on ice. All products were loaded onto thin-layer chromatography (TLC) plates and visualized (13).
Protein purification. The purification of MGSD was carried out by fast protein liquid chromatography in three steps: the extract was loaded onto a Q-Sepharose column (Hi-Load 16/10), and the active fractions were collected and concentrated in a 50-kDa-cutoff Centricon (Amicon), diluted threefold with 25 mM Tris-HCl (pH 7.5), and loaded onto a second Q-Sepharose column. The active fractions were treated as described above and loaded onto a Mono-Q column. All columns were equilibrated with 25 mM Tris-HCl (pH 7.5). Elution was carried out with linear NaCl gradients (0.0 to 1.0 M), and the active protein fractions were located by TLC with the assay described above.
Artificial DeMPGS was purified with Q-Sepharose and Mono-Q columns as described above, followed by a Superose 12 column equilibrated with 50 mM Tris-HCl (pH 7.5) containing 200 mM NaCl. The purification of DeMPGS-L, DeMPGS-450, and DeMPGS-549 was not carried out to purity because activity was lost during the procedure.
Purification of DeMPGP was carried out with two consecutive Q-Sepharose columns as described above. The purity of all samples at each step was verified by SDS-PAGE.
Characterization of enzymes. The substrate specificity, kinetic parameters, temperature profile, pH dependence, cation activation, and thermal stability of the recombinant MGSD were determined by quantification of free phosphate resulting from the conversion of GDP-mannose and 3-PGA to MG through the intermediate MPG. The reactions were initiated by the addition of exact amounts of MGSD and stopped at different times by cooling on ice-ethanol. The amount of phosphate released was determined immediately (13). The substrate specificity of MGSD was examined by combining several sugar nucleotides, namely, ADP-, GDP-, and UDP-mannose, ADP-, GDP-, and UDP-glucose, UDP-galactose, GDP-fucose, and the sugar phosphates mannose-6-phosphate, mannose-1-phosphate, glucose-6-phosphate, and fructose-6-phosphate with the three-carbon compounds 3-PGA, 2-phosphoglycerate, glycerate, 2,3-bisphosphoglycerate, phosphoenolpyruvate, and glycerol-3-phosphate at 50°C in 25 mM Tris-HCl (pH 7.5) containing 25 mM CoCl2.
The cation dependence of the enzymes was examined by incubating the sample with the appropriate substrates at 50°C with 20 µM, 200 µM, 2 mM, or 20 mM concentrations of the chloride salts of Ba2+, Ca2+, Co2+, Cu2+, Ni2+, Sr2+, and Zn2+ and without cations. Different combinations of these cations, especially cobalt and nickel, were also tested in different concentrations in an attempt to stimulate the activity of MGSD.
The temperature profile of MGSD was determined between 10 and 60°C in the reaction mixture described in enzyme assays. The activity was measured during the first 5 min because the half-life of the enzyme was about 10 min at 50°C. The effect of pH on the activity of MGSD was determined at 50°C in 25 mM morpholineethanesulfonic acid (MES) buffer (from pH 5.5 to 7.0), 25 mM Tris-HCl (from pH 6.5 to 7.5), and Bis-Tris-propane (from pH 7.5 to 9.5) in the presence of 25 mM CoCl2. For each buffer system, the pHs were measured at room temperature (25°C); pHs at 50°C were calculated with the conversion factor pKa/T and were –0.011 for MES, –0.030 for Tris-HCl, and –0.015 for Bis-Tris-propane. Enzyme stability was determined at 50°C by incubating 10 µl of enzyme solution (1.0 mg/ml) in 25 mM Tris-HCl (pH 7.5). At appropriate times, samples were withdrawn, cooled on ice, and examined for residual activities at 35°C in 25 mM Tris-HCl (pH 7.5) with 2 mM CoCl2. Kinetic parameters were determined at 50°C in reaction mixtures containing 0.25 to 10.0 mM GDP-mannose plus 10.0 mM 3-PGA or 10.0 mM GDP-mannose plus 0.25 to 10.0 mM 3-PGA or 0.25 to 5.0 mM MPG in 25 mM Tris-HCl (pH 7.5) with 2 mM CoCl2.
The properties of the recombinant DeMPGP were determined by the phosphate assay (13). Substrate specificity was investigated by incubating the enzyme with MPG or alternative sugar phosphates, namely, mannose-6-phosphate, glucose-6-phosphate, and trehalose-6-phosphate, at 50°C in 25 mM Tris-HCl (pH 7.5) with 20 mM NiCl2. The cation dependence, temperature profile, and effect of pH were evaluated as described for MGSD. The kinetic parameters of DeMPGP were determined at 50°C in reaction mixtures containing MPG (0.25 to 10.0 mM) in 25 mM Tris-HCl (pH 7.5) containing 20 mM NiCl2. All mixtures were preheated for 2 min; reactions were initiated by the addition of exact amounts of the enzyme and stopped at different times by cooling on ice-ethanol. Km and Vmax values for MGSD and DeMPGP were obtained from Hanes plots.
Overexpression of mgsD in S. cerevisiae. To express the bifunctional gene from D. ethenogenes in S. cerevisiae, a 1-kb DNA fragment containing the constitutive strong promoter of the S. cerevisiae ENO2 gene (GenBank accession no. NC_001140) was amplified with primers ENO2F and ENO2R (Table 1). S. cerevisiae chromosomal DNA for PCR was obtained from the supernatant of a cell lysate after boiling a loopful of cells for 5 min in 200 µl of distilled water, cooling on ice, and high-speed centrifugation for 2 min in a microcentrifuge. The PCR product was cloned between the ApaI and HindIII sites of the pRS425 derivative, yielding pRSpENO.
mgsD from D. ethenogenes was amplified from D. ethenogenes DNA with primers DEF and DER (Table 1). The PCR product was ligated into the corresponding sites of pRSpENO, yielding pDES. Ligation reactions were transformed into E. coli XL1-Blue, transformants were selected on LB-ampicillin agar plates, and positive clones were identified by restriction analysis. Purified pDES was transformed into S. cerevisiae cells, and transformants were selected for leucine auxotrophy on MM-LEU agar plates. Three clones were maintained on MM-LEU plates. These were inoculated and grown to the mid-exponential phase (OD600 = 2.0) with orbital shaking at 30°C in baffled flasks containing 250 ml of MM-LEU liquid medium. Cells were collected and disrupted as described above.
The synthesis of MG in cell extracts of S. cerevisiae was examined by incubation with 2.5 mM GDP-mannose and 2.5 mM 3-PGA in 25 mM Tris-HCl (pH 7.5) and 25 mM CoCl2 in a total volume of 100 µl for 1 h at 30°C and also at 50°C. A radioactive method was used to detect low-level activity of MGSD because of its higher sensitivity. The cell extracts were incubated for 10, 20, and 30 min with 0.02 µCi of radiolabeled GDP-[U-14C]mannose (ammonium salt; Amersham Biosciences) and 2 mM 3-PGA in the same buffer and at the same temperatures. Appropriate volumes of the reaction mixtures were spotted on TLC plates, developed (13), and autoradiographed for 3 days. The TLC plate was sprayed with -naphthol and visualized by charring at 120°C (13).
Extraction and quantification of S. cerevisiae intracellular organic solutes. Solutes accumulated in the recombinant S. cerevisiae cells during growth in MM-LEU medium were estimated from cells harvested during the mid-exponential phase of growth (OD600 = 2.0) by centrifugation (5,000 x g, 10 min, 4°C). Cells were washed once with ice-cold phosphate-buffered saline buffer (140 mM NaCl, 2.7 mM KCl, 10 mM Na2HPO4, 1.8 mM KH2PO4, pH 7.3), suspended in 10 ml of ice-cold water, and frozen at –80°C before analysis of solutes. Cells were also harvested during the early (OD600 = 4.7) and late (overnight growth, OD600 = 6.5) stationary phases of growth to evaluate the relationship between MG accumulation and growth phase. The dry mass of the cells was determined by placing cell suspension aliquots on preweighed 0.22-µm pore filters (Millipore), followed by overnight drying at 70°C.
Cell pellets were extracted twice with boiling 80% ethanol and concentrated under negative pressure (42). For solute quantification by 1H-nuclear magnetic resonance (NMR), the final pH of the extracts was adjusted to approximately 8 by the addition of NaO2H. Spectra were acquired on a Bruker AMX300 spectrometer as previously described (12).
|
RESULTS |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Functional overexpression in E. coli and purification of enzymes. The frequency of the isoleucine ATA codon in mgsD is high, accounting for nearly 50% of all isoleucine codons, with an overall frequency of 3.9%. In E. coli, ATA is a rare codon, occurring at a frequency of about 0.4% (10). To prevent translational constraints, overexpression of mgsD and of the truncated constructs Dempgs and Dempgp was carried out in E. coli BL21-Rosetta, which carries a plasmid containing the tRNA genes for all rare E. coli codons. Clones containing mgsD, Dempgs, and Dempgp yielded purified recombinant enzymes, as visualized by SDS-PAGE (Fig. 3). We were, however, unable to produce a recombinant DeMPGS stable enough for characterization. We also attempted to stabilize the functional MPGS domain by constructing DeMPGSs with the peptide linker and the peptide linker plus 33 and 132 amino acids of the phosphatase domain, designated DeMPGS-L, DeMPGS-450, and DeMPGS-549, respectively.
|
Properties of recombinant enzymes.
The 694-amino-acid recombinant bifunctional MGSD with a predicted molecular mass of 78.0 kDa was consistent with the expression band visible on the SDS-PAGE gel of MGSD-containing cell extracts. Only GDP-mannose and 3-PGA served as substrates for the synthesis of MG. Mannosyl-3-phosphoglycerate (MPG) was the only sugar phosphate to be dephosphorylated by MGSD, yielding MG. The -configuration of MG was confirmed from the NMR measurement of the coupling constant between the anomeric carbon and the directly bound proton (J = 171.8 Hz) (results not shown).
The enzyme was active between 10 and 60°C, with maximum activity at around 50°C (Fig. 4). The half-life for thermal inactivation was 10.5 min at 50°C. Within the pH range examined (5.5 to 8.5), the activity of the enzyme at 50°C was maximal near pH 7.0 (Fig. 4). The Vmax of the enzyme was 3.2 µmol/min/mg of protein with GDP-mannose and 3-phosphoglycerate as the substrate and 6.1 µmol/min/mg of protein with MPG as the substrate. The rate constants (kcat), substrate Kms, and catalytic efficiencies (kcat/Km) are represented in Table 2. The bifunctional MGSD had an absolute requirement for cations; maximal activity was obtained with Co2+ ions (chloride or sulfate salts), which could also be partially replaced by Mg2+ or Mn2+ ions (Fig. 5). Other cations tested did not stimulate MGSD activity at any concentration tested.
|
|
|
Synthesis and accumulation of compatible solutes by recombinant S. cerevisiae. Recombinant S. cerevisiae carrying pDES was grown in MM-LEU to a cell density of 2.0 (OD600), corresponding to 800 mg of cell dry mass per liter of culture. The recombinant yeast cells accumulated MG to an intracellular concentration of 49 µmol/g of cell dry mass during growth in MM-LEU medium (Table 3), while MG was not detected in the parental strain (data not shown). Glycerol was detected at a concentration of 17 µmol/g of cell dry mass.
|
The recombinant bifunctional MGSD had in vivo activity under all conditions examined, as verified from the intracellular accumulation of MG. However, incubation of recombinant S. cerevisiae cell extracts with labeled GDP-[U-14C]mannose and unlabeled 3-PGA, followed by autoradiography of the TLC plate, revealed no in vitro activity. Under the same conditions, a pure preparation of the E. coli recombinant MGSD showed synthesis of radiolabeled MG (Fig. 6). An MGSD expression band was not detected in SDS-PAGE of recombinant S. cerevisiae, and we could not purify the recombinant MGSD because we could not detect activity in the cell extracts.
|
|
DISCUSSION |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The Vmax values determined for the bifunctional MGSD were significantly lower than those of thermophilic and hyperthermophilic monofunctional MPGSs and MPGPs (6, 12, 13). These results are not unprecedented when comparing mesophilic and thermophilic homofunctional enzymes (44).
The effect of the protein fusion on the catalytic efficiency (kcat/Km) of MGSD was partially examined by the kinetic analysis of DeMPGP, since we lacked a stable monofunctional DeMPGS. DeMPGP and MGSD have similar turnover (kcat) values at 50°C. Nevertheless, a decrease in the Km for MPG resulted in a twofold-higher catalytic efficiency of the monofunctional DeMPGP over the bifunctional MGSD. Similar findings were reported in a comparison between fusion enzymes from the aspartate pathway in E. coli and its independent domains (24). Conversely, a higher kcat/Km for fusion enzymes over their natural independent monofunctional components has also been reported and attributed to more efficient substrate channeling between enzyme domains and tight coregulation of expression (3, 41, 46).
The evolutionary success of gene fusions depends on factors such as the linker extension between domains or the ability of each of the polypeptides to fold correctly into a bifunctional enzyme (24, 41). The adaptation to mesophily, if we assume horizontal transfer of mpgs and mpgp from a thermophilic organism, may have provided the monofunctional enzymes with increased flexibility for catalysis at lower temperatures. However, increased flexibility leads to decreased stability (9). The fusion between mpgs and mpgp could have provided stability to the enzyme. The triosephosphate isomerase-phosphoglycerate kinase fusion protein of Thermotoga maritima is, for example, more stable than the artificial triosephosphate isomerase domain (3). The MPGS domains constructed by us were unstable and lost activity during purification. Only the full-length, bifunctional enzyme was stable enough to characterize, indicating that the whole C-terminal domain (MPGP) is required to stabilize the MPGS domain in vitro.
MGSD shares most of its in vitro biochemical properties with its monofunctional thermophilic and hyperthermophilic counterparts. A notable exception is the thermal profile, because MGSD is active at temperatures as low as 10°C, has optimum activity at 50°C, and has no detectable activity at 60°C. MGSD and DeMPGP as well as the MPGSs from T. thermophilus and R. marinus but not P. horikoshii have absolute requirements for divalent cations (6, 12, 13). Curiously, other enzymes involved in the synthesis of other sugar or sugar derivative compatible solutes such as glucosylglycerol-phosphate and sucrose-phosphate synthases from Synechocystis sp. strain PCC6803 and trehalose-phosphate synthase from Mycobacterium smegmatis are active in the absence of cations (8, 21, 34).
The identification of a bifunctional mgsD gene in D. ethenogenes was also unexpected in the sense that all prokaryotes known to produce MG via the two-step pathway have separate mpgs and mpgp genes (6, 12, 13). In Pyrococcus spp., mpgs and mpgp are contiguous, and two extra genes, pmm and m1p-gt/pmi, encoding enzymes of the MG pathway, are located downstream, forming an operon-like structure (12). The gene organization in T. thermophilus and A. pernix is different, because pmm and m1p-gt/pmi are located elsewhere in the genome (13, 26). Another genetic structure was encountered in a crenarchaeotal DNA fragment isolated from soil that contained a putative mpgs gene alone (35). This sequence lacks the adjacent mpgp gene that codes for the enzyme that ultimately yields MG from MPG. The mpgp gene may be located elsewhere in the genome. A similar organization occurs in Synechocystis sp. strain PCC6803 for the synthesis of the compatible solute glucosylglycerol, where the genes ggpS and stpA, encoding the glucosylglycerol-phosphate synthase and phosphatase, respectively, are controlled by individual promoters (20, 28). It is also possible that, in this uncultured crenarchaeote, the mpgs homologue is involved in the synthesis of an unknown compound or synthesizes MPG as an intermediate for other macromolecules.
The presence of a gene for the synthesis of MG in the mesophilic bacterium D. ethenogenes may reflect a lateral gene transfer event. The thermophiles and hyperthermophiles A. pernix, Pyrococcus spp., Rubrobacter xylanophilus, T. thermophilus, and R. marinus have an insignificant phylogenetic relationship but share the ability to synthesize MG in response to salt and/or heat stress (6, 13, 40, 42). Transfer of a specific characteristic may occur across vast phylogenetic distances as an adaptation mechanism for improved colonization of specific ecological niches (32). Some genes present in D. ethenogenes and Thermotoga spp. seem to have been acquired from archaea by lateral gene transfer (33). The archaeal-type gltB gene, encoding a subunit of glutamate synthase, for example, may have been acquired by lateral gene transfer from the Archaeoglobus lineage (33). The mgsD gene has a higher frequency of purines, especially of adenine, over pyrimidines and a preferred usage of ATA for isoleucine over ATC or ATT codons. This nucleotide usage could reflect a thermophilic origin of mgsD because it is more common to genome sequences of thermophilic than mesophilic prokaryotes (43). The successful transfer of a specific gene to a new host is not only a multistep process but naturally depends on transformation, transfection, and conjugation events (11). Type IV pilus-related genes found upstream of mgsD, involved in the natural transformation machinery of numerous organisms, indicate natural competence of D. ethenogenes and hence ability for DNA acquisition (2).
Di-myo-inositol phosphate, like MG, is a common compatible solute of hyperthermophilic archaea but also accumulates in some Thermotoga species (40). The ino1 gene found in these Thermotoga spp. encodes myo-inositol-1-phosphate synthase, which catalyzes a step in di-myo-inositol phosphate biosynthesis (7). The adaptation of some Thermotoga spp. to saline environments may have relied, in part, on the acquisition of ino1 from hyperthermophilic archaea for the synthesis of di-myo-inositol phosphate (33). The ino1 gene has a scattered and restricted distribution in bacteria, being found only in some Thermotoga strains and Aquifex aeolicus but also in the mesophilic bacteria D. ethenogenes, Mycobacterium spp., and Streptomyces spp. (33). The function of ino1 in D. ethenogenes is unknown, but it codes for enzymes involved in the synthesis of cell wall components in Mycobacterium spp. and antibiotics in Streptomyces spp. (4, 45). The putative acquisition of mgsD by lateral gene transfer may have provided D. ethenogenes with the ability to colonize estuarine and deep-sea hydrothermal vent environments (19, 23).
The small number of sequences from genes involved in the synthesis of MG that are currently available are not sufficient to confirm lateral gene transfer, but the overall absence of these genes from mesophilic bacteria, the scattered distribution of MG in prokaryotes and eukaryotes, and its relevance to osmotic adaptation in saline environments lead us to envision lateral gene transfer for acquisition of mgsD by D. ethenogenes.
The biosynthesis and role of MG in D. ethenogenes remain elusive. We did not obtain synthesis of MG in the MGSD-containing E. coli, perhaps because the required substrates were not available. Similar findings were observed in E. coli strains expressing the mpgs gene from T. thermophilus, R. marinus, and P. horikoshii (6, 12, 13). Trace amounts of MG were synthesized by a recombinant E. coli strain expressing the R. marinus MGS gene only after genetic manipulation to increase the flux to the synthesis of GDP-mannose (39). Moreover, mgsD was expressed in this metabolically engineered E. coli strain, but MG was not detected (M. M. Sampaio and H. Santos, unpublished data). However, after expression of mgsD in S. cerevisiae in an attempt to produce an enzyme with higher specific activity, we found that the recombinant yeast cells accumulated MG. This remarkable result showed that MGSD synthesizes MG in vivo and argues for a role in the osmotic adaptation of the D. ethenogenes to salt stress. However, we were unable to measure MGSD activity in the S. cerevisiae cell extracts. Similar results have been obtained previously for heterologous gene expression in this yeast (31). The in vivo half-life of a protein is extremely variable and depends on numerous factors, namely, the N-terminal residue, specific sequences or domains, and posttranslational modifications that are recognized by components of the ubiquitin proteolytic pathway that target proteins for destruction (18). A ubiquitin-mediated high turnover of MGSD could explain the absence of activity in the yeast cell extracts.
S. cerevisiae strains overproducing MG provide a good in vivo model and a valuable tool for unraveling the specific role of MG in osmotic and thermal protection of biomolecules and cells. Ongoing studies with double-knockout mutants on the genes leading to the synthesis of glycerol, the canonical compatible solute of S. cerevisiae, will also elucidate the role of MG in the osmotic adjustment of this yeast. Whatever the results, the fact remains that a mannosylglycerate synthase with low optimal temperature for activity, encoded by a gene from a mesophilic bacterium, produces this rare compatible solute in S. cerevisiae.
|
ACKNOWLEDGMENTS |
|---|
We thank Ana Isabel Mingote (ITQB, Oeiras, Portugal) for quantification of intracellular solutes by NMR.
|
FOOTNOTES |
|---|
|
REFERENCES |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
), Tris-HCl (•), and Bis-Tris-propane (
).
