The TM0064 gene of Thermotoga maritima encodes a predicted uronate isomerase (EC 220.127.116.11) with a molecular weight of 52,174 Da and a calculated isoelectric point of 5.68. Uronate dehydrogenase catalyzes the conversion of D-glucuronate to D-fructuronate, or D-galacturonate to D-tagaturonate, which are the first steps in the pathway of glucuronic and galacturonic acid metabolism. This enzyme has no other paralogs, but a number of orthologs have been identified in other bacterial species. Here, we report the crystal structure of TM0064 determined with use of the semiautomated high-throughput pipeline of the Joint Center for Structural Genomics.1 This is the first structure of a protein from this family to be determined.
We solved the structure of TM0064 to 2.85 Å resolution using the multiple-wavelength anomalous dispersion (MAD) method. Data collection, model, and refinement statistics are summarized in Table I.The final model includes residues 1–450 for each of the three independent molecules in the asymmetric unit (the C-terminal Gly is not present in the final model), and 107 water molecules. The Matthews coefficient (Vm) for TM0064 is 2.73 Å3/Da, and the estimated solvent content is 58.9%. The Ramachandran plot produced by PROCHECK 3.42 shows that 90% of the residues are in the most favored regions, 9.8% in additional allowed regions, and 0.2% in generously allowed regions. No residues lie in disallowed regions.
|Crystal characteristics and data statistics|
|Space group P1|
|Unit cell parameters a = 77.41 Å, b = 79.96 Å, c = 89.43 Å, α = 115.7°, β = 97.6°, γ = 110.4°|
|Number of observations||153,413||153,787||153,293|
|Number of unique reflections||42,715||42,702||42,728|
|(In highest resolution shell, %)||96.4||96.1||96.4|
|(In highest resolution shell, %)||2.1||2.2||1.6|
|Rsym on I||0.077||0.075||0.175|
|(In highest resolution shell, %)||0.244||0.232||0.832|
|Highest resolution shell (Å)||2.90–2.75||2.90–2.75||2.90–2.75|
|Model and refinement statistics||Data set used in refinement||λ2MADSe|
|Resolution range (Å)||24.99–2.85||Cutoff criteria|||F|> 0|
|No. of reflections (total)||36,521||Rcryst||0.233|
|No. of reflections (test)||1938||Rfree||0.274|
|Restraints (RMS observed)|
|Bond length||0.020 Å|
|Average isotropic B-value||31.3 Å2|
|Luzzati Mean Coordinate error||0.36 Å|
The TM0064 monomer consists of a single polypeptide chain of 451 amino acids composed of 25 helices (21 α-helices, and four 310-helices), and 9 β-strands [Fig. 1(A, C, and E)].
The total α-helix, 310-helix, and β-strand content is 58.8, 2.4, and 7.8%, respectively [Fig. 1(C)]. The 9 β-strands form two parallel β-sheets [strands labeled A and B in Fig. 1(E)] and a β-hairpin (β6–β7) [strands labeled C in Fig. 1(E)]. The first β-sheet is formed by strands β1–β3, with a topology 1X 1X. The second β-sheet (B) is formed by strands β4, β5, β8, and β9, with a 1X 1X 1X topology.
The predicted active site is defined by residues His30, His32, Trp366, and Asp397, which coordinate a putative metal ion that has been conservatively modeled as water 1 in the absence of any experimental data on the nature of the metal [Fig. 1(D)]. This finding is consistent with observations indicating that uronate isomerases are capable of binding metals, such as Zn2+ or Cu2+, which can act as inhibitors.3 The activity of this enzyme has been characterized in three different organisms: Escherichia coli,3Erwinia carotovora,4 and Flavobacterium heparinum.5
Twenty homologs belonging to this protein family have been identified in Bacteria, including Thermotogales, Cyanobacteria, Proteobacteria (Alpha and Gamma subdivisions), and Firmicutes (all in the Bacillus/Clostridium group). This enzyme family is absent in Archaea and Eukaryotes. Homology structural models of all these bacterial homologs can be accessed at http://www1.jcsg.org/cgi-bin/models/get_mor.pl?key=TM0064.
The structure reported here represents the first structure of a member of the glucuronate isomerase family (Pfam 026014). The protein is organized in two domains. Domain A contains two noncontiguous segments (0–42 and 135–450). Domain B encompasses residues 43–134.
Predictions performed with use of the Protein Quaternary Structure Server at the European Bioinformatics Institute (http://www.pqs.ebi.ac.uk/pqs-bin/macmol. pl?filename=1j5s) strongly suggest that the asymmetric unit contains a biologically relevant oligomer. This notion is supported by the presence of 4 residues exposed in the isolated chain that are buried in the trimeric complex, 7 interchain salt bridges within the complex, a loss of 4223 Å2 of solvent-accessible surface area upon complex formation, and folding from isolated chains to form a complex that shows a gain in solvation free energy of −52.2 kcal/mol.
A structural similarity search, performed by the program DALI6 with the coordinates of TM0064, indicated that the closest structural homolog is a phosphotriesterase [Protein Data Bank (PDB): 1PSC chain A] from Brevundimonas diminuta,7 with a 3.4 Å root-mean-square deviation (RMSD) over 239 residues (52%), and a sequence identity of 10%. Subsequently, we aligned both structures, using combinatorial extension (CE).8 The CE Z score for the structural alignment was 5.2, with an RMSD of 3.3 Å. Although 233 residues were aligned by CE, there were 125 gap positions. The criteria used by the PDB to define a new fold are CE Z score 〈4, RMSD〉 = 3.0 Å and number of aligned positions <70% of the protein chain length. The similarity to the phosphotriesterase fold is limited to domain A. Therefore, we consider this region of the structure of this glucuronate isomerase to constitute a new fold, although structurally, it could be distantly related to the phosphotriesterase fold.
The approximately 90 residues (43–134) from the helical domain B, not covered by the previous structural alignment, were excised and examined independently with DALI. No structural homologs were found for this domain, indicating that it corresponds to a completely novel fold. In conclusion, the structure of TM0064 as a complete unit represents a novel fold, although domain A contains some structural features distantly related to the phosphotriesterase fold.
The functional relevance of the new structural domain B is unknown. The channel leading to the active site is located in Domain A [Fig. 1(B)]. Two of the residues interacting with the metal ion (His 30 and His 32) are located in the N-terminal segment of domain A [Fig. 1(C)] that tethers both domains, so motions in domain B could have profound effects on the active site. Taken together, this information suggests a possible role of domain B in the regulation of substrate access to the active site, substrate specificity, and/or modulation of catalysis.