Skip to main page content

• HOME
• Current Issue
• Archive
• Alerts
• About ASM
• Contact us
• Tech Support
• Journals.ASM.org

This Article Abstract Full Text (PDF) Supplemental material Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Reprints and Permissions Copyright Information Books from ASM Press MicrobeWorld Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Hudson, A. O. Articles by Leustek, T. Search for Related Content PubMed PubMed Citation Articles by Hudson, A. O. Articles by Leustek, T.

 Previous Article  |  Next Article 

Journal of Bacteriology, May 2008, p. 3256-3263, Vol. 190, No. 9
0021-9193/08/$08.00+0     doi:10.1128/JB.01381-07
Copyright © 2008, American Society for Microbiology. All Rights Reserved.

Biochemical and Phylogenetic Characterization of a Novel Diaminopimelate Biosynthesis Pathway in Prokaryotes Identifies a Diverged Form of LL-Diaminopimelate Aminotransferase ^,

André O. Hudson,¹ Charles Gilvarg,² and Thomas Leustek^1*

Biotech Center and Department of Plant Biology and Pathology, Rutgers University, New Brunswick, New Jersey 08901,¹ Department of Molecular Biology, Princeton University, Princeton, New Jersey 08544²

Received 24 August 2007/ Accepted 14 February 2008

Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
A variant of the diaminopimelate (DAP)-lysine biosynthesis pathway uses an LL-DAP aminotransferase (DapL, EC 2.6.1.83) to catalyze the direct conversion of L-2,3,4,5-tetrahydrodipicolinate to LL-DAP. Comparative genomic analysis and experimental verification of DapL candidates revealed the existence of two diverged forms of DapL (DapL1 and DapL2). DapL orthologs were identified in eubacteria and archaea. In some species the corresponding dapL gene was found to lie in genomic contiguity with other dap genes, suggestive of a polycistronic structure. The DapL candidate enzymes were found to cluster into two classes sharing approximately 30% amino acid identity. The function of selected enzymes from each class was studied. Both classes were able to functionally complement Escherichia coli dapD and dapE mutants and to catalyze LL-DAP transamination, providing functional evidence for a role in DAP/lysine biosynthesis. In all cases the occurrence of dapL in a species correlated with the absence of genes for dapD and dapE representing the acyl DAP pathway variants, and only in a few cases was dapL coincident with ddh encoding meso-DAP dehydrogenase. The results indicate that the DapL pathway is restricted to specific lineages of eubacteria including the Cyanobacteria, Desulfuromonadales, Firmicutes, Bacteroidetes, Chlamydiae, Spirochaeta, and Chloroflexi and two archaeal groups, the Methanobacteriaceae and Archaeoglobaceae.

Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
meso-Diaminopimelate (m-DAP) is the immediate precursor of lysine in prokaryotes and plants (3, 25), and in many eubacteria, m-DAP is also required for the synthesis of murein (34). In addition to the m-DAP pathway, another, completely different method has evolved for the biosynthesis of lysine (33) involving the intermediate compound -amino adipic acid. The -amino adipic acid pathway is found in most fungi (32), and a modification of it is found in selected eubacterial and archaeal species (21).

Four different variants of the m-DAP-lysine pathway have been discerned, and they are depicted in Fig. 1. All share the initial and terminal steps but differ in the reactions at the center of the pathway. The common reactions include the first, in which aspartate β-semialdehyde is condensed with pyruvate to produce dihydrodipicolinate; the second, in which dihydrodipicolinate is reduced to L-2,3,4,5-tetrahydrodipicolinate (THDPA); and the final reaction, in which m-DAP is decarboxylated to form lysine. The enzymes catalyzing these reactions, dihydrodipicolinate synthase (DapA, EC 4.2.1.52), dihydrodipicolinate reductase (DapB, EC 1.3.1.26), and m-DAP decarboxylase (LysA, EC 4.1.1.20), are conserved in species carrying any of the variants of the m-DAP pathway (13, 19, 32).

View larger version (15K): [in this window] [in a new window]

FIG. 1. The known variants of the DAP-lysine biosynthesis pathways. The chemical structures of intermediates are shown on the left. The name of the pathway is indicated at the top of the diagram, and the individual steps including enzyme symbol are shown below. In the DAP dehydrogenase and DAP aminotransferase diagrams, only the step(s) that differs from the succinyl and acetyl-DAP pathways is shown.

The enzymes needed to produce m-DAP from THDPA are what differentiate the variant pathways. The first to have been identified uses four enzymes (9, 10). THDPA is succinylated to form N-succinyl-L-2-amino-6-oxopimelate, which is then transaminated to form N-succinyl-LL-2,6-DAP, followed by removal of the succinyl group to form LL-2,6-DAP (LL-DAP). Finally, LL-DAP is converted to m-DAP by an epimerase. In Escherichia coli the enzymes catalyzing these reactions are known as THDPA succinyltransferase (DapD, EC 2.3.1.117), succinyldiaminopimelate aminotransferase (DapC, EC 2.6.1.17), succinyldiaminopimelate desuccinylase (DapE, EC 3.5.1.18), and DAP epimerase (DapF, EC 5.1.1.7). Another variant exists that also utilizes four enzymes. It is distinguished by the use of acetylated intermediates but is otherwise identical to the succinyl pathway (28, 36). The pathway using succinylated intermediates is widely distributed among prokaryotic species, whereas the one using acetylated intermediates shows a narrower phylogenetic distribution (32). A highly abbreviated DAP pathway exists in which a single enzyme, DAP dehydrogenase (Ddh, EC 1.4.1.16), produces m-DAP from THDPA, NADPH, and NH₄⁺ (20, 37). A metabolic redundancy occurs in Corynebacterium glutamicum which has both the Ddh and the succinylated pathways for lysine production (26).

The most recent DAP pathway to have been discovered uses two enzymes to convert THDPA to m-DAP (14, 19). The distinguishing enzyme of this pathway catalyzes the glutamate-dependent transamination of THDPA to form LL-DAP (EC 2.6.1.83). m-DAP is then formed by an epimerase, as in the acyl-DAP-pathways. Therefore, the aminotransferase carries out in a single step a metabolite transformation that requires three enzymes, DapD, DapC, and DapE, in E. coli. The LL-DAP aminotransferase has been reported from plants, cyanobacteria, and Chlamydia (14, 19), where it appears to be the sole route for m-DAP-lysine biosynthesis. This assumption is based on the absence of orthologs for the acyl and Ddh pathway enzymes in the genome sequences of these species and the absence of acyl pathway enzyme activities demonstrated from several different plant species (4, 13, 19). The plant and cyanobacterial enzymes show exquisite substrate specificity in that they are able to distinguish LL-DAP from its isomer m-DAP. The chlamydial orthologs have some ability to use m-DAP, although less efficiently than LL-DAP. In recognition of the reaction that it catalyzes, LL-DAP aminotransferase has been named DapL, and this designation will be used throughout the present paper. Based upon its constrained substrate specificity, DapL does not appear to be closely related to the DapC aminotransferase that functions in the acyl-DAP pathways. In fact, at least three different and divergent aminotransferases are known to catalyze the DapC reaction (6, 8, 12, 17, 18). The crystal structure of the DapL protein from Arabidopsis thaliana has been solved, providing insight into the substrate specificity of the enzyme (35).

The presence of a DapL pathway in plants, cyanobacteria and Chlamydia raised the question of how widely this enzyme is distributed in prokaryotes and how it relates to the evolution of the DAP pathway. The present study made use of the extensive list of sequenced microbial genomes to identify and functionally verify additional DapL orthologs. The results indicate that DapL exists as two divergent groups showing a restricted phylogenetic distribution in both the archaea and eubacteria.

Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Bioinformatic methods. Orthologous sequences and their genomic contiguity were analyzed using the SEED program (23). Multiple protein sequence alignment was carried out using Clustal W (31). Phylogenetic trees were constructed by the maximum parsimony method using the program MEGA, version 3.1, with its default settings (16).

ORF cloning. For the open reading frames (ORFs) that were cloned for expression in E. coli along with the primers used for amplification by PCR, see Table S1 in the supplemental material. All ORFs were cloned initially into pET30a (Novagen Corp.) using the restriction sites introduced by PCR (see the italicized sequences of Table S1 in the supplemental material), with the exception of slr1666 (a Synechocystis sp. gene) which was cloned initially into pGEM T-Easy (Promega Corp). The sequences were vetted in the entry plasmid. pET30a ORFs were transformed into E. coli BL21-CodonPlus-RIPL for protein expression. For complementation analysis an expression cassette consisting of the entire ORF, the His tag coding sequence, and the ribosome binding site from pET30 was subcloned into pBAD33 (11) using XbaI and SalI or XbaI and HindIII for Morella thermoacetica Moth_0889. The slr1666 expression cassette was subcloned from pGEM-T-Easy to pQE30 using EcoRI and PstI (Qiagen Corp.). The pBAD33-derived or pQE-derived plasmid was used for complementation of E. coli mutant strains.

Functional complementation. ORFs were tested for functional complementation of dapD (AT980), dapE (AT984) (obtained from the Coli Genetic Stock Center, Yale University), and a dapD dapE double mutant strain (AOH1). AOH1 was constructed by P1 transduction of a dapD::kan allele from JC7623 (7) into AT984 as previously described (14). The strains were transformed with the pBAD33- or pQE-derived plasmids and were selected on LB medium supplemented with 50 µg ml^–1 DAP (DL-,-diaminopimelic acid; Sigma-Aldrich product D-1377) and 34 µg ml^–1 chloramphenicol (pBAD-ORF clones) or 100 µg ml^–1 ampicillin (pQE-slr1666). Individual colonies were replica plated onto LB medium supplemented with 50 µg ml^–1 DAP (pQE-slr1666) or onto LB medium without DAP; colonies were grown under inducing or repressing conditions with 0.2% (wt/vol) arabinose or 0.2% (wt/vol) glucose, respectively (pBAD-ORF clones), and without isopropyl-β-D-thiogalactopyranoside (IPTG) or with 1 mM IPTG (pQE-slr1666). The cultures were grown at 30°C for 24 h.

Enzyme assays. Recombinant protein was expressed in E. coli grown on LB medium at 37°C to an optical density at 600 nm of 0.5, followed by induction with 1 mM IPTG for 4 h at 25°C. Cells were lysed by sonication in a solution of 100 mM HEPES-KOH (pH 7.6). Metabolites in the extract were removed by using an Amicon Ultra 30,000 molecular-weight-cutoff ultrafilter; the concentrated soluble protein sample was used for measurement of enzyme activity as an initial assessment of enzyme function. When the recombinant protein was to be purified, a larger cell culture was grown and lysed in 50 mM sodium phosphate and 300 mM NaCl (pH 8.0). The soluble protein was incubated with Talon metal affinity agarose (Clontech 8901-2), which was then washed three times with sodium phosphate-NaCl buffer containing 10 mM imidazole, and finally the bound protein was eluted with sodium phosphate-NaCl buffer containing 300 mM imidazole. The pure protein was then concentrated in an Amicon Ultra 30,000 molecular-weight-cutoff ultrafilter, replacing the elution buffer with 100 mM HEPES-KOH (pH 7.6).

For enzyme assays LL-DAP and m-DAP were isolated from culture filtrates of a lysine auxotroph and purified as described by Gilvarg (10). C. glutamicum Ddh (Ddh_Cg) was produced as a recombinant protein expressed from plasmid pET28-CgDdh obtained from D. I. Roper (University of Warwick) in E. coli BL21(DE3). Ddh_Cg accumulated to approximately 90% of the soluble protein and was highly active, so it was not further purified. The preparation converted m-DAP to THDPA at a rate of 20 µmol min^–1 mg^–1 of protein at 30°C.

Three different enzyme assays were used to study DapL. Two assays were used to measure the physiologically reverse activity (LL-DAPTHDPA), and one was used for the physiologically forward activity (THDPALL-DAP). The first of the reverse assays measured THDPA formation by its reaction with O-aminobenzaldehyde (OAB). This assay is not strictly quantitative due to the instability and batch-to-batch variation of OAB. Also the extinction coefficient of the dihydroquinazolium adduct that is formed is unknown. However, this assay is useful for comparative measurement of DapL activity in crude cell extracts. The OAB assay contained in a 1-ml reaction mixture 100 µmol of HEPES-KOH (pH 7.6), 0.5 µmol of LL-DAP (or other amino donor), 2.0 µmol of 2-oxoglutarate (or other amino acceptor), 1.25 mg of OAB, and DapL. Reaction mixtures were incubated at 30°C, and the A₄₄₀ was measured using a spectrophotometer. A second assay of the reverse activity was quantitative when pure enzyme was being measured. A 1-ml reaction mixture containing 100 µmol of HEPES-KOH (pH 7.6), 0.3 µmol of NADPH, 50 µmol of NH₄Cl, 0.5 µmol of LL-DAP, 5 µmol of 2-oxoglutarate, 16 µg of Ddh_Cg, and DapL was incubated at 30°C, and the A₃₄₀ was measured continuously. The reaction sequence is as follows: (i) LL-DAP + 2-oxoglutarate THDPA + glutamate + H₂O; (ii) THDPA + NH₄⁺ + NADPH m-DAP + NADP⁺ (A₃₄₀).

Assay of the physiologically forward reaction was measured in a two-step reaction. In the first step, THDPA was formed from m-DAP in a 1-ml reaction mixture containing 100 µmol of HEPES-KOH (pH 7.6), 0.5 µmol of NADP⁺, various concentrations of m-DAP, and 32 µg of Ddh_Cg. The reaction was run to completion, determined by monitoring the A₃₄₀ value. Aminotransferase activity was then measured in the second step. The reaction contained, in addition to the components of the first reaction, 0.3 µmol of thio-NAD⁺, 0.3 µmol of coenzyme A (CoA), and 0.5 µmol of glutamate (for convenience, these components were added in the first-step reaction mixture). Then the aminotransferase assay was started by adding 200 µg of 2-oxoglutarate dehydrogenase (0.625 µmol min^–1 mg^–1 of protein) and DapL. Activity was determined by measuring the A₃₉₈ at 30°C. Thio-NAD⁺ was used to avoid spectrophotometric interference from NADP⁺. The two compounds have different absorbance maxima. The reaction sequence is as follows: (i) m-DAP + NADP⁺ THDPA + NH₄⁺ + NADPH (A₃₄₀); (ii) THDPA + glutamate + H₂O LL-DAP + 2-oxoglutarate; (iii) 2-oxoglutarate + thio-NAD⁺ + CoA succinyl CoA + CO₂ + H⁺ + thio-NADH (A₃₉₈).

Kinetic constants were determined by measuring the initial reaction rate at various concentrations of reactants. The data were analyzed by nonlinear regression analysis and the Michaelis-Menten equation using GraphPad Prism, version 3.03 for Windows, GraphPad Software, San Diego, CA (www.graphpad.com).

Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Identification of DapL orthologs in microbial genomes. The microbial genomes protein database was searched with BLASTP using the amino acid sequence of Synechocystis sp. DapL as the query (locus tag sll0480). The returns list included proteins with a broad continuum of homologies, a result that was expected considering that aminotransferases form a large superfamily (2, 15). Representative examples from the BLASTP output are given in Table S2 in the supplemental material, sorted in order of descending homology. A more complete list can be found at the SEED website (http://theseed.uchicago.edu/FIG/index.cgi). Two proteins in the table from Protochlamydia sp. and Chlamydia trachomatis, showing 42% to 44% identity with sll0480, are confirmed examples of DapL (19). They were included to show the level of homology for an authentic DapL ortholog. Among the examples showing greater percentages of identity were proteins from species representing only a few phylogenetic groups including the Cyanobacteria, Deltaproteobacteria, Euryarchaeota, Firmicutes, Spirochaetes, Desulfuromonadales, Bacteroidetes, and Chlamydiae. All of the Cyanobacteria and the Chlamydiae present in the genomes database contained such a DapL ortholog, but for the other groups only some species contained a strong DapL candidate.

In an effort to discern authentic forms of DapL, a lower limit of homology was approximated by identifying aminotransferases that are unlikely to be DapL. The proteins showing the best homology with sll0480 from bacterial species known to be devoid of DapL activity including Agrobacterium tumefaciens, Bacillus subtilis, and E. coli (14) ranged from 23% to 29% identity. The best match was an aminotransferase from B. subtilis showing 29% identity with sll0480 (see Table S2 in the supplemental material).

The sll0480 homologs demonstrating between 29% and 42% identity were further analyzed by examining their genomic context. The chromosomal contiguity of genes in prokaryotic genomes has proven to be a useful tool for inferring function and identifying metabolic networks, based on the hypothesis that functionally coupled genes are often clustered (24, 27). An example is the dapL gene from Protochlamydia (locus tag pc0685) that is located immediately downstream of dapB and dapA in an orientation that suggests all three may exist in an operon (19). By using the SEED database (23), it was observed that some dapL orthologs do indeed exist in genomic contiguity with other DAP-lysine biosynthesis genes. For example, dapL is situated immediately downstream of dapF in 10 species belonging to the Bacteriodetes, Firmicutes, Chloroflexi, and Deltaproteobacteria phylogenetic groups (see Table S3 in the supplemental material). The likely significance of the contiguity of this gene arrangement is emphasized by the fact that in the microbial genomes only a single other example exists in which dapF is contiguous with another aminotransferase. That example is slr1666 in Synechocystis sp., which lies immediately downstream of dapF. slr1666 is annotated as a degT aminotransferase and shows only 15% identity with sll0480. In Bacillus stearothermophilus, mutation of the degT gene results in complex phenotypic changes including altered production of extracellular enzymes, altered control of sporulation, loss of flagella, and abnormal cell division. Hence, it was initially described as a pleiotropic regulatory gene (30). In other genomes, dapL was found to be contiguous with lysA or with multiple dap synthesis genes or with combinations of lysA and multiple dap synthesis genes (see Table S3 in the supplemental material). The most extreme example of gene clustering was uniquely found in Syntrophobacter fumaroxidans, in which the dapL ortholog (Sfum_0054) is contiguous with lysA, dapF, dapA, and dapB. This grouping is of particular interest since it would comprise the entire set of genes necessary to synthesize lysine from aspartate β-semialdehyde, assuming that Sfum_0054 encodes DapL. Thus, in several noteworthy examples genomic context analysis provided support for a role of DapL in DAP-lysine synthesis. Moreover, the significance of the associations discovered with dapL orthologs is underscored by the fact that DAP-lysine biosynthetic genes generally show weak genomic associations (27).

Genomic contiguity analysis produced two surprising results. In a number of the best examples of gene clustering, the aminotransferase is highly diverged from sll0480. For example, Syntrophobacter Sfum_0054 shows 30% identity with either sll0480 or pc0685, a level of homology that was earlier taken as the approximate cutoff for authentic DapL (see Table S2 in the supplemental material). This finding suggested the possibility that DapL may exist as two divergent types. In fact, many of the best DapL candidates showing marginal homology with sll0480 were found to be closely related to Sfum_0054 (see Table S2 in the supplemental material). Another surprising result was that nearly all of the cyanobacteria and several species from other phylogenetic groups contain two DapL orthologs, one that is related to sll0480 and the other to Sfum_0054 (three examples are listed in Table S2 in the supplemental material). That DapL orthologs form two divergent groups that are distinct from other class I/II aminotransferases is graphically shown in the phylogenetic tree presented in Fig. 2. The tree was constructed by comparing representatives from DapL1 (those more closely related to sll0480) and DapL2 (more closely related to Sfum_0054) with a selection of other aminotransferases. Several examples of DapC and ArgD were included because these enzymes catalyze transamination in the acyl-DAP pathways (8, 12, 17), the reaction that is analogous to that catalyzed by DapL. Several examples of AspC and TyrB were included because these represent the prototype of class I/II aminotransferases. Most of the DapL orthologs were initially annotated as aspartate or tyrosine aminotransferases. Finally, all the class I/II aminotransferases in B. subtilis were included to provide an overall context for aminotransferases that are unlikely to be DapL. In addition to the divergence of DapL into two major groups, the phylogenetic tree shows that both DapL types are more closely related to DapC than they are to ArgD, AspC, or TyrB.

View larger version (24K): [in this window] [in a new window]

FIG. 2. Phylogenetic analysis of DapL orthologs. The diagram is of a neighbor-joining tree produced by alignment with Clustal W using a gap penalty of 10 and a gap length penalty of 0.2. The tree was constructed by bootstrap analysis using MEGA, version 3.1. Locus tags are indicated. The identities of enzyme clusters are indicated. Homologous proteins that cannot be DapL proteins are marked with an asterisk. The same proteins are indicated in Table S2 in the supplemental material.

Functional analysis of DapL1 and DapL2 proteins. The bioinformatic analysis identified a large group of DapL candidates that could be functionally tested. Subsets from the DapL1 and DapL2 groups, as well as more distantly related aminotransferases, were chosen for experimental analysis. The proteins were initially evaluated for the ability to complement the DAP auxotrophy of an E. coli dapD dapE double mutant, which would demonstrate that the aminotransferase is able to bypass two of the three reactions that E. coli uses to convert THDPA to LL-DAP. As controls, the same proteins were tested for the ability to complement the dapD mutant. Since dapD and dapE encode a succinyltransferase and desuccinylase, respectively, even though the DapL proteins do not show any homology with succinyltransferases or desuccinylases, the simplest explanation for complementation in the dapD dapE mutant would be that the experimental enzyme is capable of direct transamination of THDPA (14, 19). To provide control over protein expression, a plasmid carrying the E. coli arabinose-regulated promoter was used for the experiment. An example of a complementation result is presented in Fig. 3, which shows the arabinose-dependent growth of the dapD and the dapD dapE mutants transformed with sll0480. Fifteen different enzymes were tested in this way, and the results are compiled in Table 1. All those that were identified as strong DapL candidates, based on sequence homology with either sll0480 or Sfum_0054, showed arabinose-dependent complementation of the dapD dapE mutant, but none of the more divergent proteins did. The divergent proteins included slr1666, the degT ortholog downstream of dapF in Synechocystis sp. showing only 15% overall identity with sll0480. The others, including sll0938, Desulfitobacterium hafniense Dhaf_2289, and OB2282, were at the margin of homology with either sll0480 or Sfum_0054.

View larger version (12K): [in this window] [in a new window]

FIG. 3. Complementation assay. Complementation was tested in the indicated mutant under inducing and repressing conditions. The vector was pBAD33, and sll0480 was the dapL ortholog from Synechocystis cloned into pBAD33. The constructs were tested for complementation of the dapD mutant and dapD dapE double mutant. Each construct was serially diluted in 0.85% (wt/vol) saline (from the left, optical density at 600 nm of 0.1, 0.01, and 0.001), and 5 µl was plated onto the indicated medium. Gene expression from the constructs was induced on medium containing 0.2% (wt/vol) arabinose (Ara) or was repressed on medium containing 0.2% (wt/vol) glucose (Glc). DAP was added when indicated at 50 µg ml–1.

View this table: [in this window] [in a new window]

TABLE 1. Functional tests of DapL orthologs

To confirm the complementation result, all the strains were tested for DapL activity in soluble protein extracts prepared from cells in which expression was induced. Nearly all of the strains expressing a protein that was able to complement dapD-dapE showed DapL activity (Table 1). The single exception was sll0006. This protein was found to be highly expressed in induced cells but accumulated nearly exclusively in inclusion bodies, possibly explaining why no activity was detected in the soluble protein extract despite the fact that sll0006 complements the mutants. It should be noted that the assay used to measure DapL in crude protein extracts is not particularly sensitive, explaining perhaps why activity was not detectable. The other exception was Leptospira interrogans LIC12841, which complemented both under inducing and repressing conditions. This enzyme was the most active of all those tested, suggesting that low-level expression under repressing conditions may have provided enough activity to complement the E. coli mutants. The enzymes were also examined for their substrate specificity. sll0480, LIC12841, and Methanothermobacter thermoautotrophicus MTH52 were found to use LL-DAP but not m-DAP at the concentration used (0.5 mM), whereas the other enzymes showing DapL activity used both LL-DAP and m-DAP. The relaxed substrate specificity was not specific to either the DapL1 or DapL2 class. Due to interference of m-DAP with the coupling system, the quantitative assay could not be used to determine substrate specificity using a second assay system, so this question was not further explored. The analysis of substrate specificity also provided experimental evidence for the activity of both DapL1 and DapL2 orthologs and, in addition, indicated that some species, such as Synechocystis sp., contain two DapL orthologs, one from each of the divergent groups.

To further characterize the collection of DapL proteins, selected examples were chosen for purification and analysis of kinetic properties. It was possible to isolate soluble, active protein in four instances, L. interrogans LIC12841, M. thermoautotrophicus MTH52, D. hafniense Dhaf_1761, and M. thermoacetica Moth_0889. The first three are members of the DapL1 group whereas Moth_0889 is a member of DapL2. The kinetic experiments revealed that each of the DapL orthologs shows kinetic properties that are very similar to previously characterized forms of the enzyme from Arabidopsis and Chlamydia (Table 2). In all cases, the activity was significantly greater in the reverse direction (THDPA synthesis) than in the forward direction (LL-DAP synthesis). In addition, all showed a much higher affinity for THDPA and LL-DAP than for glutamate or 2-oxoglutarate. Interestingly, the overexpressed LIC12841 and MTH52 showed much higher V_max values than Dhaf_1761 and Moth_0889. However, the higher specific activity may be related to the quality of the protein since the amount of soluble LIC12841 and MTH52 produced in E. coli was much greater than Dhaf_1761 and Moth_0889, for which the yield of soluble protein was low.

View this table: [in this window] [in a new window]

TABLE 2. Kinetic constants of DapL proteins

The purified proteins were used to examine their substrate specificity (see Table S4 in the supplemental material). By using 2-oxoglutarate as the amino acceptor, a variety of potential amino donors were tested. In agreement with the earlier results obtained with crude bacterial extracts, LIC12841 and MTH52 used LL-DAP exclusively, and Dhaf_1761 and Moth_ 0889 used both LL-DAP and m-DAP. None of the enzymes used lysine or ornithine as amino donors at the concentration tested (0.5 mM). By using LL-DAP as the amino donor, a range of potential amino acceptors was tested. Only 2-oxoglutarate was used as an amino acceptor by LIC12841 and MTH52, but Dhaf_1761 and Moth_0889 were both able to use oxaloacatate. The product of amino transfer to oxaloacetate is aspartate.

Microbial diversity of DAP pathways. Hudson et al. (14) and McCoy et al. (19) argued that DapL must represent an alternative DAP-lysine biosynthesis pathway because plant and chlamydia genomes encode enzymes for THDPA synthesis and m-DAP or lysine synthesis from LL-DAP but lack the genes encoding enzymes necessary to bridge the anabolic gap between THDPA and LL-DAP. Moreover, experimental evidence indicated that DapL can catalyze the conversion of THDPA to LL-DAP. The effort to annotate genomes based on what genes exist and what genes cannot be identified is referred to as "metabolic reconstruction," and this technique has proven useful in uncovering new variants of many different metabolic pathways (5, 22). The SEED database provides a facile means to analyze the coincidence of metabolic pathway genes in the microbial genomes, and it was used for the present analysis. Of the species containing a DapL1 or DapL2 ortholog, all were found to contain DapA, DapB, DapF, and LysA but lacked DapD and DapE. A representative list extracted from the SEED database is given in Table S5 in the supplemental material. In only a few species in the Firmicutes phylogenetic group, the presence of DapL was coincident with a strong Ddh candidate gene. This observation is an interesting parallel to the existence of redundant DAP pathways in Corynebacterium as previously reported (26).

Analysis of all the available microbial genomes revealed that about 14.0% harbor a DapL pathway. Nearly all the rest harbor one of the two pathways that utilize acylated intermediates. A clear-cut phylogenetic lineage for the DapL pathway was not evident. Other than the Cyanobacteria and Chlamydiae, for which all the examples in the genomes database contained a DapL pathway, the other groups, including Deltaproteobacteria, Euryarchaeota, Firmicutes, Spirochaetes, Desulfuromonadales, and Bacteroidetes, showed only selected examples of DapL pathway-containing species.

Analysis of the DapL sequence. Aminotransferases are a major challenge for classification because the superfamily is so large and diverse. Classification of DapL proteins presents a further challenge because the enzymes have significantly diverged into two sequence types, DapL1 and DapL2. The recent report of the crystal structure of DapL from Arabidopsis (35) provides the structural basis to compare the divergent forms of DapL. The sequence alignment shown in Fig. S1 in the supplemental material illustrates that all the residues defined as being involved in pyridoxal phosphate binding in Arabidopsis DapL are conserved in both the DapL1 and DapL2. Most of these residues are conserved with related aminotransferases. By contrast, only one of the three residues that Watanabe et al. (35) suggested is involved in binding of LL-DAP are conserved. For example, the residues homologous with Lys129 are conserved in both DapL forms, but Glu97 and Asn309 are conserved only in DapL1, not DapL2. As a result of the catalytic mechanism of aminotransferases, both the amino donor and acceptor substrates are thought to bind to the same catalytic site but can involve contact with different residues. For DapL, Tyr152 was suggested to be involved in binding to glutamate. This residue is conserved in both the DapL1 and DapL2 forms. Examination of the alignments shown in Fig. S1 in the supplemental material indicates many conserved positions between DapL1 and DapL2 and also many positions that distinguish the two forms of DapL. Further investigation of the structure of DapL1 and DapL2 will be required to determine the functional role of these residues.

Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
The recent identification of DapL in plants and other photosynthetic eukaryotes and in the prokaryotes, cyanobacteria, and chlamydia (14, 19) revealed yet another variant of the DAP-lysine biosynthesis pathways. The DapL pathway is the simplest, mechanistically, of the acyl-type pathways, and for this reason it is tempting to speculate that it represents the ancestral pathway. The function of acylation is not understood, but is thought to aid formation of the acyclic hydrolytic product of THDPA, exposing a keto group to the aminotransferase (1). DapL, on the other hand, probably relies on the spontaneous formation of the acyclic structure (14, 35). If so, the DapL pathway might be less efficient than the acyl pathways, meaning that the latter could be an adaptive feature of rapidly growing organisms.

The phylogenetic distribution of DapL was studied with the aim of understanding the evolution and biochemical properties of DAP biosynthesis. The study revealed that there are two major classes of DapL that differ significantly in sequence. The two forms appear to be loosely associated with specific microbial lineages. For example, DapL1 occurs most often in bacteria, and DapL2 is found most often in archaea. There also appear to be signatures of lateral gene transfer events. For example, M. thermoautotrophicus and Methanosphaera stadtmanae stand out because they are Euryarchaeota of the Methanobacteria group that contain DapL1, whereas all the other available Methanobacteria contain a DapL2. There are also indications of interoperonic gene replacements. For example, the gene arrangement dapA-dapB-dapL found in the DapL-containing Deltaproteobacteria appears to be conserved, yet S. fumaroxidans is unusual in that a DapL2 replaces the DapL1 form found in the other Deltaproteobacteria.

The DapL orthologs are approximately 410-amino-acid proteins whose sequence conservation identifies them as members of the pyridoxal-5'-phosphate-dependent protein superfamily of class I/II aminotransferases (15, 29). Based on sequence homology, all were initially misannotated as aspartate or aromatic amino acid aminotransferases, but subsequent experimental analysis revealed their true catalytic function. The recent determination of the crystal structure of DapL from the plant species Arabidopsis (locus tag At4g33680) has significantly added to the understanding of the amino acid residues that make up the active sites of the enzyme. Interestingly, it appears that the DapL2 form lacks two of the three residues that are conserved in DapL1 (Glu97 and Asn309). It is important to point out that Watanabe et al. (35) modeled LL-DAP binding to At4g33680 based on the structure of the enzyme crystallized with malate, which mimics substrate binding. It is surprising that greater kinetic differences are not associated with the sequence divergence. On the other hand, the extreme divergence has allowed the identification of absolutely conserved residues, suggesting critical roles in function. Further study will be required to understand how LL-DAP binding to DapL2 can occur with the substitution of these residues.

Excellent candidates for DapL are always accompanied by the absence of acyl Dap pathway genes, meaning that pathway redundancy is not usual. The exception was in certain organisms where there seemed to be excellent candidates for Ddh, meaning that these organisms could possess dual lysine anabolic pathways. A precedent for pathway redundancy that includes Ddh exists for C. glutamicum (26). In this species, the Ddh pathway provides a greater flux to m-DAP under growth conditions at high ammonium concentrations.

The phylogenetic distribution of the DapL pathway shows that it is present in only 14.0% of the sequenced microbial genomes. It is important to point out that the sequenced microbial genomes are not representative of microbial diversity. Indeed, environmental species are not well represented compared with pathogenic and saprophytic species. As the genomes of more microbial species are sequenced and annotated, the presence of the DapL pathway will likely increase.

 
This work was funded by National Science Foundation grant IPB0449542 (T.L. and C.G.) and a National Institutes of Health Predoctoral Fellowship GM069264 (A.O.H.).

We are grateful to the following individuals for contributing genomic DNA or cell material: Sheila Patrick, James M. Tiedje, Takakazu Kaneko, Albert Ko, Caroline Plugge, John M. Reeve, Stephen W. Ragsdale, and Hideto Takami. We especially thank Valérie de Crécy-Lagard and Andrew D. Hanson for their help in using the SEED genome annotation tool.

 
* Corresponding author. Mailing address: Rutgers University, Biotech Center, 59 Dudley Road, New Brunswick, NJ 08901-8520. Phone: (732) 932-8165, ext 326. Fax: (732) 932-0312. E-mail: leustek{at}aesop.rutgers.edu

Published ahead of print on 29 February 2008.

Supplemental material for this article may be found at http://jb.asm.org/.

Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 

1
1. Berges, D. A., W. E. DeWolf, Jr., G. L. Dunn, D. J. Newman, S. J. Schmidt, J. J. Taggart, and C. Gilvarg. 1986. Studies on the active site of succinyl-CoA:tetrahydrodipicolinate N-succinyltransferase. Characterization using analogs of tetrahydrodipicolinate. J. Biol. Chem. 261:6160-6167.[Abstract/Free Full Text]
2
2. Bray, J. E., A. E. Todd, F. M. Pearl, J. M. Thornton, and C. A. Orengo. 2000. The CATH dictionary of homologous superfamilies (DHS): a consensus approach for identifying distant structural homologues. Protein Eng. 13:153-165.[Abstract/Free Full Text]
3
3. Bryan, J. K. 1990. Advances in the biochemistry of amino acid biosynthesis, p. 161-195. In B. J. Miflin and P. J. Lea (ed.), The biochemistry of plants, vol. 16. Academic Press, New York, NY.
4
4. Chatterjee, S. P., B. K. Singh, and C. Gilvarg. 1994. Biosynthesis of lysine in plants: the putative role of meso-diaminopimelate dehydrogenase. Plant Mol. Biol. 26:285-290.[CrossRef][Medline]
5
5. Cordwell, S. J. 1999. Microbial genomes and "missing" enzymes: redefining biochemical pathways. Arch. Microbiol. 172:269-279.[CrossRef][Medline]
6
6. Cox, R. J., and P. S. H. Wang. 2001. Is N-acetylornithine aminotransferase the real N-succinyl-LL-diaminopimelate aminotransferase in Escherichia coli and Mycobacterium smegmatis? J. Chem. Soc. Perkin Trans. 1:2006-2008.
7
7. Cranenburgh, R. M., J. A. Hanak, S. G. Williams, and D. J. Sherratt. 2001. Escherichia coli strains that allow antibiotic-free plasmid selection and maintenance by repressor titration. Nucleic Acids Res. 29:E26.[CrossRef][Medline]
8
8. Fuchs, T. M., B. Schneider, K. Krumbach, L. Eggeling, and R. Gross. 2000. Characterization of a Bordetella pertussis diaminopimelate (DAP) biosynthesis locus identifies dapC, a novel gene coding for an N-succinyl-L,L-DAP aminotransferase. J. Bacteriol. 182:3626-3631.[Abstract/Free Full Text]
9
9. Gilvarg, C. 1961. N-Succinyl-alpha-amino-6-ketopimelic acid. J. Biol. Chem. 236:1429-1431.[Free Full Text]
10
10. Gilvarg, C. 1959. N-Succinyl-L-diaminopimelic acid. J. Biol. Chem. 234:2955-2959.[Free Full Text]
11
11. Guzman, L. M., D. Belin, M. J. Carson, and J. Beckwith. 1995. Tight regulation, modulation, and high-level expression by vectors containing the arabinose PBAD promoter. J. Bacteriol. 177:4121-4130.[Abstract/Free Full Text]
12
12. Hartmann, M., A. Tauch, L. Eggeling, B. Bathe, B. Mockel, A. Puhler, and J. Kalinowski. 2003. Identification and characterization of the last two unknown genes, dapC and dapF, in the succinylase branch of the L-lysine biosynthesis of Corynebacterium glutamicum. J. Biotechnol. 104:199-211.[CrossRef][Medline]
13
13. Hudson, A. O., C. Bless, P. Macedo, S. P. Chatterjee, B. K. Singh, C. Gilvarg, and T. Leustek. 2005. Biosynthesis of lysine in plants: evidence for a variant of the known bacterial pathways. Biochim. Biophys. Acta 1721:27-36.[Medline]
14
14. Hudson, A. O., B. K. Singh, T. Leustek, and C. Gilvarg. 2006. An LL-diaminopimelate aminotransferase defines a novel variant of the lysine biosynthesis pathway in plants. Plant Physiol. 140:292-301.[Abstract/Free Full Text]
15
15. Jensen, R. A., and W. Gu. 1996. Evolutionary recruitment of biochemically specialized subdivisions of family I within the protein superfamily of aminotransferases. J. Bacteriol. 178:2161-2171.[Free Full Text]
16
16. Kumar, S., K. Tamura, and M. Nei. 2004. MEGA3: integrated software for molecular evolutionary genetics analysis and sequence alignment. Brief. Bioinform. 5:150-163.[Abstract/Free Full Text]
17
17. Ledwidge, R., and J. S. Blanchard. 1999. The dual biosynthetic capability of N-acetylornithine aminotransferase in arginine and lysine biosynthesis. Biochemistry 38:3019-3024.[CrossRef][Medline]
18
18. Marienhagen, J., N. Kennerknecht, H. Sahm, and L. Eggeling. 2005. Functional analysis of all aminotransferase proteins inferred from the genome sequence of Corynebacterium glutamicum. J. Bacteriol. 187:7639-7646.[Abstract/Free Full Text]
19
19. McCoy, A. J., N. E. Adams, A. O. Hudson, T. Leustek, C. Gilvarg, and A. Maurelli. 2006. LL-Diaminopimelate aminotransferase, a trans-kingdom enzyme of a variant diaminopimelate/lysine synthesis pathway shared by Chlamydia and plants. Proc. Natl. Acad. Sci. USA 103:17909-17914.[Abstract/Free Full Text]
20
20. Misono, H., H. Togawa, T. Yamamoto, and K. Soda. 1976. Occurrence of meso-alpha, epsilon-diaminopimelate dehydrogenase in Bacillus sphaericus. Biochem. Biophys. Res. Commun. 72:89-93.[Medline]
21
21. Nishida, H., M. Nishiyama, N. Kobashi, T. Kosuge, T. Hoshino, and H. Yamane. 1999. A prokaryotic gene cluster involved in synthesis of lysine through the amino adipate pathway: a key to the evolution of amino acid biosynthesis. Genome Res. 9:1175-1183.[Abstract/Free Full Text]
22
22. Osterman, A., and R. Overbeek. 2003. Missing genes in metabolic pathways: a comparative genomics approach. Curr. Opin. Chem. Biol. 7:238-251.[CrossRef][Medline]
23
23. Overbeek, R., T. Begley, R. M. Butler, J. V. Choudhuri, H. Y. Chuang, M. Cohoon, V. de Crecy-Lagard, N. Diaz, T. Disz, R. Edwards, M. Fonstein, E. D. Frank, S. Gerdes, E. M. Glass, A. Goesmann, A. Hanson, D. Iwata-Reuyl, R. Jensen, N. Jamshidi, L. Krause, M. Kubal, N. Larsen, B. Linke, A. C. McHardy, F. Meyer, H. Neuweger, G. Olsen, R. Olson, A. Osterman, V. Portnoy, G. D. Pusch, D. A. Rodionov, C. Ruckert, J. Steiner, R. Stevens, I. Thiele, O. Vassieva, Y. Ye, O. Zagnitko, and V. Vonstein. 2005. The subsystems approach to genome annotation and its use in the project to annotate 1,000 genomes. Nucleic Acids Res. 33:5691-5702.[Abstract/Free Full Text]
24
24. Overbeek, R., M. Fonstein, M. D'Souza, G. D. Pusch, and N. Maltsev. 1999. The use of gene clusters to infer functional coupling. Proc. Natl. Acad. Sci. USA 96:2896-2901.[Abstract/Free Full Text]
25
25. Patte, J.-C. 1996. Biosynthesis of threonine and lysine, p. 528-541. In F. C. Neidhardt, R. Curtiss III, J. L. Ingraham, E. C. C. Lin, K. B. Low, B. Magasanik, W. S. Reznikoff, M. Riley, M. Schaechter, and H. E. Umbarger (ed.), Escherichia coli and Salmonella: cellular and molecular biology, 2nd ed. ASM Press, Washington, DC.
26
26. Schrumpf, B., A. Schwarzer, J. Kalinowski, A. Puhler, L. Eggeling, and H. Sahm. 1991. A functionally split pathway for lysine synthesis in Corynebacterium glutamicum. J. Bacteriol. 173:4510-4516.[Abstract/Free Full Text]
27
27. Spirin, V., M. S. Gelfand, A. A. Mironov, and L. A. Mirny. 2006. A metabolic network in the evolutionary context: multiscale structure and modularity. Proc. Natl. Acad. Sci. USA 103:8774-8779.[Abstract/Free Full Text]
28
28. Sundharadas, G., and C. Gilvarg. 1967. Biosynthesis of alpha, epsilon- diaminopimelic acid in Bacillus megaterium. J. Biol. Chem. 242:3983-3984.[Abstract/Free Full Text]
29
29. Sung, M. H., K. Tanizawa, H. Tanaka, S. Kuramitsu, H. Kagamiyama, K. Hirotsu, A. Okamoto, T. Higuchi, and K. Soda. 1991. Thermostable aspartate aminotransferase from a thermophilic Bacillus species. Gene cloning, sequence determination, and preliminary X-ray characterization. J. Biol. Chem. 266:2567-2572.[Abstract/Free Full Text]
30
30. Takagi, M., H. Takada, and T. Imanaka. 1990. Nucleotide sequence and cloning in Bacillus subtilis of the Bacillus stearothermophilus pleiotropic regulatory gene degT. J. Bacteriol. 172:411-418.[Abstract/Free Full Text]
31
31. Thompson, J. D., D. G. Higgins, and T. J. Gibson. 1994. CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice. Nucleic Acids Res. 22:4673-4680.[Abstract/Free Full Text]
32
32. Velasco, A. M., J. I. Leguina, and A. Lazcano. 2002. Molecular evolution of the lysine biosynthetic pathways. J. Mol. Evol. 55:445-459.[CrossRef][Medline]
33
33. Vogel, H. J. 1965. Lysine biosynthesis and evolution, p. 25-40. In V. Bryson and H. J. Vogel (ed.), Evolving genes and proteins. Academic Press, New York, NY.
34
34. von Heijenoort, J. 1996. Murein synthesis, p. 1025-1034. In F. C. Neidhardt, R. Curtiss III, J. L. Ingraham, E. C. C. Lin, K. B. Low, B. Magasanik, W. S. Reznikoff, M. Riley, M. Schaechter, and H. E. Umbarger (ed.), Escherichia coli and Salmonella: cellular and molecular biology, 2nd ed. ASM Press, Washington, DC.
35
35. Watanabe, N., M. M. Cherney, M. J. van Belkum, S. L. Marcus, M. D. Flegel, M. D. Clay, M. K. Deyholos, J. C. Vederas, and M. N. James. 2007. Crystal structure of LL-diaminopimelate aminotransferase from Arabidopsis thaliana: a recently discovered enzyme in the biosynthesis of L-lysine by plants and Chlamydia. J. Mol. Biol. 371:685-702.[CrossRef][Medline]
36
36. Weinberger, S., and C. Gilvarg. 1970. Bacterial distribution of the use of succinyl and acetyl blocking groups in diaminopimelic acid biosynthesis. J. Bacteriol. 101:323-324.[Abstract/Free Full Text]
37
37. White, P. J. 1983. The essential role of diaminopimelate dehydrogenase in the biosynthesis of lysine by Bacillus sphaericus. J. Gen. Microbiol. 129:739-749.[Abstract/Free Full Text]

---

Journal of Bacteriology, May 2008, p. 3256-3263, Vol. 190, No. 9
0021-9193/08/$08.00+0     doi:10.1128/JB.01381-07
Copyright © 2008, American Society for Microbiology. All Rights Reserved.

This Article Abstract Full Text (PDF) Supplemental material Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Reprints and Permissions Copyright Information Books from ASM Press MicrobeWorld Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Hudson, A. O. Articles by Leustek, T. Search for Related Content PubMed PubMed Citation Articles by Hudson, A. O. Articles by Leustek, T.