|Home | About | Journals | Submit | Contact Us | Français|
Novel methylene tetrahydromethanopterin (H4MPT) dehydrogenase enzymes, named MtdC, were purified after expressing in Escherichia coli genes from, respectively, Gemmata sp. strain Wa1-1 and environmental DNA originating from unidentified microbial species. The MtdC enzymes were shown to possess high affinities for methylene-H4MPT and NADP but low affinities for methylene tetrahydrofolate or NAD. The substrate range and the kinetic properties revealed by MtdC enzymes distinguish them from the previously characterized bacterial methylene-H4MPT dehydrogenases, MtdA and MtdB. While revealing higher sequence similarity to MtdA enzymes, MtdC enzymes appear to fulfill a function homologous to the function of MtdB, as part of the H4MPT-linked pathway for formaldehyde oxidation/detoxification.
Formaldehyde detoxification is a metabolic function essential to all of life, due to the extreme toxicity of formaldehyde (4, 12). However, in methylotrophic bacteria formaldehyde oxidation is a part of their central metabolism (1). One of the most widespread modes of formaldehyde oxidation in methylotrophs is the pathway that involves tetrahydromethanopterin (H4MPT) as a cofactor (28, 30). Most of the enzymes involved in this pathway are homologous to the enzymes involved in methanogenesis by the Archaea (6, 9). However, one enzyme in the pathway, an NAD(P)-linked methylene-H4MPT dehydrogenase (MtdB), is unique to Bacteria (14, 15). MtdB has evolved independently of the archaeal functional counterparts that are linked to H2 or cofactor F420 (26), based on the lack of sequence similarity (9, 15). Enzyme properties and mutant analyses have demonstrated that MtdB fulfills a dual physiological role in methylotrophic metabolism, in energy generation (in the form of NADH) and in formaldehyde detoxification (9, 14, 15, 21, 28). In some methylotrophs, so far only in methylotrophs employing the serine cycle for formaldehyde assimilation, a paralog of MtdB is present, an NADP-linked methylene-H4MPT/methylene-tetrahydrofolate (H4F) dehydrogenase (MtdA) (7, 10, 28, 29). Experiments with Methylobacterium extorquens AM1, including protein purification and analysis, mutant analysis, and flux analysis, all suggested that in vivo, the main function of MtdA is in reducing methenyl-H4F to methylene-H4F, which feeds into the serine cycle (7, 23, 25, 28, 29). MtdA also has a secondary function, in general metabolism (purine biosynthesis, etc.) in the serine cycle methylotrophs that do not possess FolD (7, 23), the enzyme fulfilling this function in most bacteria and eukaryotes (20). Heterologously expressed folD was shown to complement the function of mtdA in general metabolism, but not in methylotrophy (23). The origin and the evolutionary history of MtdA and MtdB remain poorly understood. While MtdA reveals low levels of sequence similarity to FolD enzymes (15% identity at the amino acid level [7, 29]), MtdB shares no similarity with FolD (15). However, the two paralogs reveal a significant level of similarity to each other (about 30% at the amino acid level), pointing to their common origin (15). In this work we identify and characterize a third class of bacterial methylene H4MPT dehydrogenases, MtdC, that possess higher sequence similarity toward MtdA but appear to be functional homologs of MtdB.
The sequences translated from the mtdA and mtdB genes of Methylobacterium extorquens AM1 (7, 15) were used as queries in BLAST analyses to identify mtdA/mtdB gene orthologs in the genomes of Gemmata sp. strain Wa1-1 (6) and Methylobium petroleophilum (http://genome.jgi-psf.org/finished_microbes/metpe/metpe.home.html). A single ortholog was previously identified in the genomes of Rhodopirellula baltica and Gemmata obscuriglobus (2, 13). A divergent ortholog of mtdA (env97) was identified in our previous study (16). To obtain the complete sequence of this gene and the surrounding DNA region, we analyzed a metagenomic library of Lake Washington (unpublished results) by PCR with primers specific for env97 and identified a positive clone containing env97 (LWBAC10-10). The region surrounding env97 was sequenced from primers originating in env97. Two other clones in the metagenomic library, named LWBAC-L1N9 and LWBAC10-4, were identified by hybridization with, respectively, divergent fae and divergent fhcD genes identified in our previous studies (16, 17). Two additional divergent mtdA-like genes employed in this study were identified via sequencing of the regions surrounding, respectively, the divergent fae gene in clone LWBAC-L1N9 and the divergent fhcD gene in clone LWBAC10-4.
mtdA-like genes were PCR amplified from the Gemmata sp. strain Wa1-1 chromosome and from the DNA of the fosmid clone LWBAC10-10 and cloned into the Methylobacterium expression vector pCM80, as previously described (22). A fae gene was amplified from clone LWBAC-L1N9 and cloned into pCM80 in a similar fashion. The Escherichia coli JM109 (Invitrogen) strains harboring the plasmids expressing mtdA-like genes were used as donors in matings with both mtdA and mtdB mutants of M. extorquens (15, 23), as previously described (22). The plasmid expressing fae from LWBAC-L1N9 was mated into the fae mutant of M. extorquens (31) in a similar fashion. Progenies were selected on succinate medium supplemented with 1 mM methanol or on succinate medium with the precaution of avoiding methanol vapors, respectively, as described before (8, 18, 21). The resulting transconjugants were tested for growth on succinate, methanol, or succinate in the presence of methanol vapors, as described before, to test for complementation of the specific phenotypes of the mtdA and mtdB mutants (8, 18, 21).
mtdA-like genes were PCR amplified from chromosomal DNA of Gemmata sp. strain Wa1-1 and from the DNA of clones LWBAC-L1N9 and LWBAC10-10. The following primers were employed: Gem-NdeF (5′-GGAATTCCATATGTCCGAAAAACCCACGATCC-3′) and Gem-XhoR (5′-CCGCTCGAGCGGGAGCTGCAGCGCGAGACTGTAC-3′) to amplify the mtdA ortholog from Gemmata sp. strain Wa1-1; L1N9-NdeF (5′-GGAATTCCATATGCGACCGCTTCTCCTGCAG-3′) and L1N9-NotR (5′-ATAAGAATGCGGCCGCCACCATCCCCCGCGCGATATC-3′) to amplify the mtdA ortholog from clone LWBAC-L1N9; and LWBAC10-NdeF (5′-GGAATTCCATATGCGAACACTTCTTCTCC-3′) and LWBAC10-NotR (5′-ATAAGAATGCGGCCGCGCCTCGGGGCGGAGACAGCGTG-3′) to amplify the mtdA ortholog from clone LWBAC10-10. The PCR product amplified from Gemmata sp. strain Wa1-1 was digested with NdeI and XhoI and ligated into the pET15b expression vector (Novagen). The PCR products amplified from clones LWBAC-L1N9 and LWBAC10-10 were digested with NdeI and NotI and ligated into pET24a (Novagen). The resulting constructs were transformed into E. coli JM109. Sequencing of the cloned inserts revealed that no sequence errors were generated during PCR, and these constructs were used to transform E. coli BL21(DE3) (Novagen). The construct containing the mtdA ortholog amplified from clone LWBAC-L1N9 resulted in insoluble expressed protein (data not shown). Thus, this insert was cloned into the pET15b vector using NdeI/XhoI restriction sites, and the resulting construct was transformed into E. coli. Cells were grown to an optical density at 600 nm of approximately 0.6 and then induced by 1 mM isopropyl-β-thiogalactoside, followed by additional 3-hour incubation at 24°C (optical density at 600 nm of 1.0 to 1.2). MtdA-like proteins were purified via three chromatographic steps as follows. Frozen cells were resuspended in 50 mM sodium phosphate (pH 8.0) buffer containing 0.5 M NaCl, 10 mM imidazole, and 10 mM β-mercaptoethanol and passed two times through a French pressure cell at 1.2 × 108 Pa. Centrifugation was performed at 15,000 × g for 30 min at 4°C to remove cell debris. The supernatants (25 ml) were mixed with 1 ml Ni-nitrilotriacetic acid-agarose, incubated for 1 h on ice, and purified using a QIAexpress type IV kit as described by the manufacturer (QIAGEN). Proteins were eluted with 5 ml of 250 mM imidazole, 0.5 M NaCl, 1 mM dithiothreitol, and dialyzed against 50 mM HEPES buffer, pH 8.0 (LWBAC-L1N9 and LWBAC10-10) or pH 7.5 (Gemmata sp. strain Wa1-1). The dialyzed extracts were applied onto 1-ml HiTrap DEAE-FF columns (Amersham Biosciences). LWBAC-L1N9 and LWBAC10-10 proteins were eluted with 1 M NaCl in HEPES (pH 8.0). The Gemmata sp. strain Wa1-1 protein was collected in the flowthrough fractions, which were pooled, dialyzed against 50 mM bis-Tris pH 5.0 buffer, and subjected to cation-exchange chromatography on a 1-ml HiTrap SP FF column (Amersham Biociences) equilibrated with the same buffer. The protein was eluted with 1 M NaCl. The fractions containing MtdA-like proteins were pooled, desalted, and concentrated using Amicon Ultra-4 centrifugal filter devices (Millipore), replacing the respective buffers with 25 mM HEPES (pH 7.5), 0.5 M NaCl, 2 mM dithiothreitol, and 5% glycerol. Preparations were frozen in liquid nitrogen and stored at −80°C. Preparations used for determination of kinetic parameters of the enzymes contained at least 95% pure MtdA-like enzymes (Fig. (Fig.1).1). In addition, enzyme activities were tested in pure protein preparations obtained after an additional purification step, size-exclusion chromatography using HiLoad 26/60 Superdex 75 (Amersham Biosciences), to ensure that the H4F-linked activity was not a result of the presence of contaminating FolD (data not shown).
H4MPT was purified from Methanobacterium thermoautotrophicum Marburg (3) and was a gift of R. K. Thauer. H4F was purchased from Sigma. The anoxic 120 mM potassium phosphate buffer (pH 6.0) was prepared as follows. Buffer was boiled, and dithiothreitol was added (final concentration, 2 mM). The gas phase over the buffer was then changed to nitrogen. The anoxic buffer was used to make stock solutions of H4MPT and H4F to prevent the preparations from oxygen damage. Assays were performed routinely at room temperature in 1-ml cuvettes (depth, 1 cm) in a total volume of 0.7 ml (this volume was used due to limited amounts of H4MPT available). The reactions were monitored spectrophotometrically by measuring the increase in absorbance at 340 nm. For the calculations, 340 values of 27 mM−1 cm−1 for methylene H4MPT dehydrogenation with NAD(P) and 27.9 mM−1 cm−1 for methylene H4F dehydrogenation with NAD(P) were used (29). Units of enzyme activities are defined as 1 μmol/min per mg of protein at room temperature. Methylene H4MPT dehydrogenation with NAD(P) was measured in 120 mM potassium phosphate, pH 6.0, at a methylene H4MPT concentration of 40 μM and an NAD(P) concentration of 0.3 mM. Methylene H4F was measured in the same way, whereby the concentration of methylene H4F was 70 μM. Methylene H4F and methylene H4MPT were generated by spontaneous reaction of formaldehyde (3 mM) with H4F and H4MPT, respectively. Km and Vmax values were deduced from reciprocal plots of the initial rates versus the concentration of one substrate at different fixed concentrations of the second substrate.
Translated amino acid sequences were aligned using the ClustalW program (27). For phylogenetic analyses, the Phylip package (11) was used. Maximum likelihood, distance, and parsimony methods were employed, and 1,000 bootstrap analyses were performed.
The sequences of the fosmid inserts used in this study have been deposited with GenBank under accession numbers DQ084247, DQ084250 and DQ084248.
Planctomycetes have recently emerged as the third major microbial group to possess the genes encoding reactions of the H4MPT-linked C1 transfer pathway (2, 6, 13). Analysis of 16 genes conserved between Bacteria and Archaea has revealed the deeply branching nature of the planctomycete sequences, pointing toward the antiquity of this group and to the long evolutionary history of the C1 reactions linked to H4MPT (2, 6). In this work we analyzed the mtdA gene orthologs in the three available planctomycete genomes. A single ortholog was identified in each genome via BLASTP analysis using either MtdB or MtdA sequences as queries. When compared to the data in the nonredundant sequence database (NCBI), translated orthologs from planctomycetes revealed significant divergence from both MtdA and MtdB polypeptides but possessed higher similarity to the former (43 to 53% and 28 to 32% at the amino acid level, respectively). This finding was surprising, as in methylotrophic Proteobacteria MtdB has been shown to serve as the main enzyme in H4MPT-linked C1 transfer (9, 14, 15, 21, 28), while the major function of MtdA seems to be in providing methylene-H4F to the serine cycle (24), in addition to its function in general metabolism (7, 23). The presence of a single ortholog in planctomycete genomes and the presence of traditional folD orthologs in these genomes (reference 13 and unpublished data) suggested that the function of the MtdA protein orthologs in planctomycetes could be more similar to the function of MtdB than to the function of MtdA. Gene orthologs significantly diverging from both mtdB and mtdA were also detected in environmental samples from Lake Washington via direct PCR amplification (16) or via metagenome analysis, as described in Materials and Methods. Phylogenetically, mtdA orthologs from planctomycetes and mtdA orthologs from uncharacterized microbes fell into two distinct groups that were clearly separated from both MtdA and MtdB enzymes (Fig. (Fig.22).
Mutants of M. extorquens that contain lesions in, respectively, mtdB and mtdA possess characteristic phenotypes: an mtdB mutant is negative for growth on methanol and is highly sensitive to methanol vapors (15, 21), while an mtdA mutant is methanol negative and requires an added C1 substrate (methanol or formate) for growth on succinate but is not sensitive to methanol vapors (7, 23). To test for the potential of the proteins encoded by the mtdA orthologs to fulfill either MtdB or MtdA function, genes from Gemmata sp. strain Wa1-1 and from one environmental clone (LWBAC10-10) were overexpressed in these mutants, and the transconjugants were checked for phenotypic complementation. Neither of the novel genes could complement either mtdA or mtdB mutants. For comparison, expression of other genes of Gemmata sp. strain Wa1-1 (6) or the divergent fae from clone LWBAC-L1N9 (this work) using the same expression system resulted in complementation of the respective mutants of M. extorquens. We tested whether the failure in complementation was caused by the lack of expression in the system used or by different substrate specificities of the enzymes encoded by the novel genes. Activities were measured in cell extracts of E. coli and in M. extorquens MtdA and MtdB mutant backgrounds, using different combinations of cofactors. Considerable activity of methylene-H4MPT dehydrogenase with NADP as a cofactor was measured in E. coli (0.6 U/mg for the LWBAC10-10 gene and 0.9 U/mg for the Gemmata sp. strain Wa1-1 gene), while none was measured with NAD as a cofactor. The LWBAC10-10 gene was responsible for an increase over the background in the methylene-H4F dehydrogenase (NADP) activity (0.01 U/mg), but no activity over the background was detected with the Gemmata sp. strain Wa1-1 gene construct. Similarly, an increase in the NADP-linked methylene-H4MPT dehydrogenase activity was observed in cell extracts of M. extorquens background (up to 0.1/mg U for the LWBAC10-10 gene and up to 5.2 U/mg for the Gemmata sp. strain WA-1-1 gene). None to a very low activity increase was observed with NAD as a cofactor (<0.001/mg and 0.003 U/mg, respectively), and only a small increase of NADP-dependent methylene-H4F dehydrogenase activity of 0.002 U/mg was observed upon complementation with the Gemmata sp. strain WA-1-1 gene, while no activity was detected with the LWBAC10-10 gene. These data point toward the novel enzymes having substrate specificities differing from those of MtdA or MtdB. Likely, these differences and not the lack of expression were responsible for negative complementation results.
We expressed in E. coli three novel MtdA orthologs, one from Gemmata sp. strain Wa1-1 and two from the yet-unidentified species present in Lake Washington. Each was purified and tested for the substrate specificity and kinetic properties. The purified enzymes were highly active in catalyzing the methylene-H4MPT dehydrogenase reaction using NADP as a cofactor but were much less efficient in catalyzing methylene-H4MPT dehydrogenation with NAD (Table (Table1).1). They also revealed low efficiency in catalyzing the reaction of dehydrogenation of methylene-H4F with NADP as a cofactor (Table (Table1).1). The kinetic parameters (Km and Vmax) were determined for the new enzymes for different combinations of cofactors, and these, along with the Vmax/Km ratios (the indicator of enzyme efficiency), are shown in Table Table11 in comparison to the values previously reported for MtdA and MtdB enzymes from M. extorquens. Based on these values, the catalytic properties of the novel enzymes are different from the properties of the previously characterized MtdA and MtdB enzymes. While the Km and Vmax values varied significantly between the three new enzymes, all three possessed much lower efficiencies in catalyzing NAD-linked dehydrogenation of methylene-H4MPT (52- to 55,000-fold lower than for MtdB), and they showed very low efficiencies in catalyzing NADP-linked dehydrogenation of methylene-H4F (91- to 448-fold lower than for MtdA). The novel enzymes thus differed from MtdA and MtdB enzymes by their broader substrate specificities. While MtdA revealed no detectable affinity for NAD and MtdB revealed no detectable affinity for H4F, the novel enzymes revealed affinities for NAD, NADP, H4F, and H4MPT, with the highest affinity for the H4MPT/NADP couple. It seems likely that the latter represents the physiologically relevant combination of cofactors for these enzymes, in agreement with the failure to complement either mtdA or mtdB mutants (see above). Based on the properties distinguishing them from both MtdA and MtdB, we named the new enzymes MtdC. As is the case with MtdA and MtdB, MtdC enzymes did not catalyze detectable dehydrogenation of methylene-H4F with NAD.
The sequences of the newly characterized MtdC enzymes, as well as representative sequences of MtdB enzymes from proteobacteria, were aligned with the sequence of MtdA from M. extorquens, whose crystal structure has been determined with and without bound NADP (10). In general, the MtdA active site residues showed more conservation with MtdC than with MtdB. For example, arginine 152 of MtdA, which is involved in binding of the 2′ phosphate of NADP, is conserved in MtdC but not in MtdB (data not shown). This may contribute to the strong preference of MtdC enzymes for NADP as opposed to NAD. In MtdB enzymes, the arginine is replaced by a histidine, which likely enables interaction with both NADP and NAD. Since no crystal structure of MtdA has been determined with H4MPT bound, the amino acid residues involved in H4F versus H4MPT specificity remain unknown.
Here we describe a novel class of bacterial methylene-H4MPT dehydrogenases, named MtdC, possessing high specificity toward methylene-H4MPT and NADP. This specificity distinguishes MtdCs from the previously characterized bacterial methylene-H4MPT dehydrogenases, MtdB and MtdA, with MtdB being specific to methylene-H4MPT in combination with either NAD or NADP and MtdA being specific to NADP in combination with either H4MPT or H4F. So far, genes encoding MtdA, whose main function in M. extorquens is in supplying methylene-H4F into the serine cycle (25), have only been found in bacterial genomes encoding the serine cycle enzymes (Methylobacterium species M. capsulatus and M. petroleophilum) (7, 19, 32; this study). Moreover, in those genomes, mtdA genes are physically linked to fch genes encoding methenyl-H4F cyclohydrolase and, in Methylobacterium species and M. petroleophilum, these are parts of larger serine cycle gene clusters (5, 19, 32; this study). On the contrary, genes encoding MtdB, whose function in M. extorquens is in the oxidation of methylene-H4MPT with energy generation and also in formaldehyde detoxification (9, 15, 21, 28), are widely distributed within the Proteobacteria possessing the H4MPT-linked formaldehyde oxidation pathway (16, 18), and in many cases they are physically linked to other genes specifically involved in the pathway (5, 18, 19, 24).
MtdC genes so far have been found in planctomycetes (2, 6, 13; this study) and in yet-unidentified species residing in Lake Washington. Sequencing of fosmid inserts in clones LWBAC-L1N9, LWBAC10-4, and LWBAC10-10 (25, 40, and 34 kbp, respectively) revealed a low degree of similarity between the genes on the fosmids and the sequences of known microbes (data not shown), suggesting that these species may represent deeply branching phyla as well. In all known cases, mtdC genes are linked on the chromosomes to other genes involved in the H4MPT-linked formaldehyde oxidation pathway and most consistently to fae, the gene encoding a reaction preceding the methylene-H4MPT dehydrogenase reaction (31) (Fig. (Fig.3).3). On the other hand, no homologs for the serine cycle genes were found in the planctomycete genomes. These data point toward MtdC being a part of the H4MPT-linked formaldehyde oxidation/detoxification pathway, i.e., fulfilling a function homologous to the function of MtdB, with which it shares lower sequence similarity, as opposed to the function of MtdA, with which it shares higher similarity (Fig. (Fig.2).2). Confirmation of this hypothesis will require mutant analysis, which is not feasible at this time either with planctomycetes or of course with the uncultured environmental species.
Compared to MtdA and MtdB, MtdC enzymes reveal broader substrate specificity. This substrate “promiscuity” is consistent with an ancestral role of MtdC with respect to both MtdA and MtdB. Testing this hypothesis will require identification of more divergent mtdA/B sequences and possibly identification of species carrying mtdC in combination with mtdA or mtdB.
In conclusion, we here describe a fifth class of microbial methylene-H4MPT dehydrogenases, of which two are found in the Archaea (Mtd and Hmd) and three are found in the Bacteria (MtdA, MtdB, and MtdC). While archaeal and bacterial analogs are not evolutionarily related, the three bacterial enzymes must originate from a common ancestor.
M.E.L. and L.C. acknowledge support from the Microbial Observatories program funded by the National Science Foundation (MCB-0131957). J.A.V. acknowledges support from the Centre National de la Recherche Scientifique and the Max-Planck-Gesellschaft.
We are grateful to R. K. Thauer of Max-Planck-Institute for Terrestrial Microbiology, Marburg, Germany, for the generous gift of H4MPT. O. Kalyuzhniy, C. Mehlin, and E. Boni of the University of Washington, Seattle, are acknowledged for technical assistance.