ORF tte0206 in the T. tengcongensis genome sequence 29 was postulated to be a ribose-binding protein homolog (tteRBP) based on its sequence similarity to the known E. coli RBP (57% identity, 76% similarity) (Figure 1) and its position within a putative operon containing ORFs homologous to ABC transporters characteristic of solute transport. The DNA for ORF tte0206, lacking a putative periplasmic signal sequence 30 (residues 1–39), was amplified from T. tengcongensis genomic DNA by the polymerase chain reaction. The resulting DNA fragment was cloned into a pET21a vector in-frame with a C-terminal hexa-histidine tag preceded by a glycine-serine linker. The nucleotide sequence was confirmed by DNA sequencing of the resulting vector. Over-expression of tte0206 produced ~30 mg of pure protein per liter of medium, which was purified by immobilized metal affinity chromatography followed by gel filtration chromatography. tteRBP eluted from the gel filtration column as a broad peak immediately following the void volume of the column (data not shown). For subsequent crystallization and characterization of tteRBP fractions of the broad peak from the gel filtration column, that were consistent with monomeric tteRBP, (fractions with a calculated hydrodynamic radius of 30 kDa ± 15 kDa) were pooled and concentrated to ~15 mg/mL (see Materials and Methods).

The thermal stability of tteRBP was determined by thermal denaturation using circular dichroism (CD). In the absence of denaturant no significant temperature-dependent change in the CD signal was observed up to 100°C; consequently, heat denaturations were carried out in the presence of varying concentrations of guanidine hydrochloride (GdCl) (Figure 2). Melting curves were found to fit a two-state model 3132. The apparent thermal transition midpoint (^appT_m) of 102°C in the absence of GdCl was determined by linear extrapolation of a series of melting point determinations carried out at different GdCl concentrations 33 (Figure 2). tteRBP is therefore significantly more stable than the mesophilic ecRBP (^appT_mvalue is 56°C) (Figure 2).

Binding of ribose to tteRBP was confirmed by observing ligand-mediated changes in the ^appT_m value in the presence of 4.0 M GdCl. Under these conditions in the absence of ribose, the ^appT_m value is 74°C; and 92°C in the presence of 1 mM ribose (Figure 2). The ^appT_m value of the ribose complex in the absence of GdCl is 114°C; the ^appT_m value for ecRBP under equivalent conditions is 72°C (Figure 2).

The tteRBP crystal structure was solved to 1.9 Å resolution by molecular replacement 34 using the ribose-bound form of ecRBP as the search model 23. The tteRBP structure adopts the overall fold and topology that is characteristic of periplasmic ribose-binding proteins (Figure 3). The asymmetric unit contains 346 water molecules and two tteRBP molecules (residues 40–313) in essentially identical conformations (0.12 Å RMSD of backbone atoms) complexed with ribose. Data collection, stereochemistry, and refinement statistics are summarized in Table 1.

Analysis of main-chain and side-chain geometry of the aligned structures indicates there are few differences in the main-chain geometries of ecRBP and tteRBP (0.4 Å RMSD of 235/271 C_α positions and 0.65 Å RMSD of 270/271 C_α positions and distance between aligned C_α positions range from 0.03–3.1 Å over 270 Cα positions). The loops and turns in the binding pocket retain near-identical conformations. Modest backbone conformational heterogeneity is observed in loops and turns that connect alternating β-strands and α-helices in tteRBP and ecRBP (RMSD of C_α positions for residues 55–61 is 0.9 Å, 117–126 is 1.6 Å and 149–156 is 3.02 Å) (Figure 4). Proline 153 in tteRBP corresponds to a single-residue insertion relative to ecRBP; small structural perturbations associated with this insertion are contained within five amino acids preceding and following this residue (3.1 Å RMSD of C_α positions). tteRBP also contains an additional three amino acids at the C-terminus that are not present in ecRBP.

The amino acid side-chain conformations are also remarkably well conserved (Table 2). Only 24 residues show a significant change in the χ₁ torsion, resulting in the adoption of a different side-chain rotamer. Of these, 19 correspond to substitutions, including non-conservative changes; there is therefore a significant bias for non-conservative mutations in the population of residues that exhibit rotameric changes. 17 of the rotamer changes occur in the surface, three in the boundary, and four in the core (Table 2). The non-conservative changes occur mostly on the surface (seven residues). Two of the surface rotamer changes (Q24/E24, D182/D183) involve charged amino acids which results in the formation of two salt bridges, two involve the loss of a salt bridge (K110/E110, K243/L244), one involves the loss of a hydrogen-bond (T178/Q179) and three positions are involved gain of five additional hydrogen-bonds (D52, Q80/S80, R139) relative to ecRBP. The four core changes are conservative substitutions and involve β-branched amino acids, altering packing of the core (V8/I8, I60/V60, T66/V66, V183/I184), and in one case (T66/V66) increasing the hydrophobicity and removing an unsatisfied core hydrogen-bond.

Polar amino acids, non-polar amino acids, waters, and the hydrogen-bonding interactions are identical in both tteRBP and ecRBP sugar-binding pockets (Figure 4). The total number of hydrogen-bonding interactions 35 is also well conserved among tteRBP and ecRBP (Table 3). Overall, tteRBP has a total of 264 hydrogen-bonds, ecRBP has 257. The hydrogen-bonding pattern outside of the binding pocket varies slightly among tteRBP and ecRBP. tteRBP has an additional three side-chain/main-chain and nine main-chain/main-chain hydrogen-bonds, but has lost five side-chain/side-chain hydrogen-bonds relative to ecRBP (Table 3). Five of the additional seven hydrogen-bonds observed in tteRBP (two main-chain/side-chain, three main-chain/main-chain) are accounted for by the four-residue insertion in tteRBP. There is therefore a net gain of two hydrogen-bonds in tteRBP, which arise from the slight differences in the hydrogen bonding pattern of the side-chain/side-chain and main-chain/main-chain residues. It is also observed that tteRBP has lost two salt bridges relative to ecRBP.

The two RBPs share 57% amino acid identity and 76% similarity (as defined in 36; in our study charge inversions are scored as non-conservative) (Figure 1). The structures can be divided into core (C), boundary (B), and surface (S) regions using an objective, structure-based classification scheme 37 (Figure 1 and Table 4). The conditional probability of a substitution occurring in a particular region (R) of the protein, p(M|R), is strongly biased (g(M|R) = p(M|R)|p(R)), with g(M|C) = 0.53, g(M|B) = 2.7 and g(M|S) = 1.6 for all interactions and 0.21, 1.36, and 1.74 respectively for non-conservative mutations (g<1, anticorrelated; g = 1, uncorrelated; g>1, positively correlated). The pattern of sequence divergence is also correlated with the distance from the ribose-binding site, as measured by the sequence identity and similarity, in a series of concentric shells centered on the bound ribose (Figure 5). Not surprisingly, the residues in the shell forming the ribose contacts are identical. With increasing distance, there is an approximately monotonic decrease in sequence identity and similarity, with the farthest shell having 73% and 34% similarity and identity respectively (Figure 5).

Analysis of the amino acid diversity among the core, boundary and surface of ecRBP and tteRBP allows identification of possible determinants of thermal stability in tteRBP (Table 5). A bias is observed for a gain of polar and charged amino acids on the surface of tteRBP (net of twelve charged and three polar substitutions), while the opposite is observed for the tteRBP core, where there is a bias for the loss of polar amino acids (seven net substitutions). There are also a significant number of substitutions of non-β-branched amino acids for β-branched amino acids in the core and boundary of tteRBP (five net substitutions) and a loss of β-branched amino acids in the surface of tteRBP (four net substitutions). Interestingly a large number of β-branched amino acids are conserved in the core and boundary of tteRBP and ecRBP, there is however a bias for the substitution of valine for isoleucine in the thermophile (eight net substitutions).

The tte0206 gene was amplified from T. tengcongensis genomic DNA by the sticky-end PCR method using the following primers: PO₄-TATGA AAACTATAGG ATTAGTGATATCTACTCTTAACAATCC, and TATGAAAACTATAGG ATTAGTGATATCTACTCTTAACAATCC for the 5' end of the gene; PO₄- AATTCTAATGGTGATGGTGATGGTGTGATCCCTGTACATTTTCTTTTGTTATGAGTTTAAGTTCTGC, and CTAATGGTGATGGTGATGGTGTGATCCCTGTACATTTTCTTTTGTTATGAGTTTAAGTTCTGC for the 3' end of the gene 51. The resulting fragment was cloned into the NdeI/EcoRI sites of a pET21a (Novagen) plasmid for over-expression in E. coli. This ORF lacks the putative periplasmic signal sequence 30. The coding sequence starting at lysine 40 was cloned in-frame with an ATG start codon. A hexahistidine affinity tag and a glycine-serine linker was fused in-frame at the carboxy terminus to facilitate purification by immobilized metal affinity chromatography (IMAC). Protein concentration was determined spectrophotometrically (ε₂₈₀ = 3800 M^-1cm^-1) 52. The resulting gene product was expressed and purified by IMAC as described 33. Pooled IMAC fractions were concentrated to 12 mL and were loaded onto a Superdex 26/60 S75 (Amersham) gel filtration column that was previously calibrated with blue dextran, bovine serum albumin, chicken serum albumin, chymotrypsin and lysozyme. tteRBP eluted from the column beginning at the void volume and ending at a calculated hydrodynamic radius corresponding to ~20 KDa. For crystallization and characterization, 10 mL fractions corresponding to a calculated hydrodynamic radius corresponding to an apparent molecular weight of 30 KDa ± 15 kDa, were collected and concentrated to 0.5 mM and dialyzed in 10 mM Tris pH7.8, 20 mM NaCl. An average of 30 mg of pure protein produced per liter of medium.

Circular dichroism (CD) measurements were determined on an Aviv Model 202 circular dichroism spectrophotometer. Thermal denaturations were determined by measuring the CD signal at 222 nm (1 cm path length) as a function of temperature, using 1 μM protein (10 mM Tris-HCl pH7.8, 150 mM NaCl), GdCl at various concentrations, in the presence or absence of 1 mM ribose. Protein samples were incubated for 15 minutes prior to collecting data. Each measurement includes a 3-second averaging time for data collection and a 60 second equilibration period at each temperature. Data was fit to a two-state model which accounts for the native and denatured baseline slopes, to determine the apparent T_m values 3132. It is not known whether equilibrium was achieved under these conditions; denaturation midpoint temperatures are therefore reported as apparent values (^appT_m). The ^appT_m values in the absence of denaturant were determined by linear extrapolation 33.

Ribose was added to tteRBP in 3-fold stoichiometric excess prior to crystallization. tteRBP crystals were grown by micro-batch under paraffin oil in drops that contained 2 μl of the protein solution (0.5 mM) mixed with 2 μl of 0.1 M sodium citrate pH 4.0, 50% (w/v) PEG 1000 and 0.1 M potassium phosphate monobasic. The tteRBP crystals diffract to 1.9 Å resolution, belong to the C2 space group (a = 123.18 Å, b = 35.8 Å, c = 118.03 Å, β = 107.02) and typically grew within three weeks at 17°C (Table 1). No stabilizing cryoprotectant was used and crystals were frozen directly in precipitant solution, mounted in a nylon loop and flash frozen in liquid nitrogen. All data were collected at 100 K at the SER-CAT 22 BM beam line at the Advanced Photon Source. The diffraction data were scaled and indexed using SCALA and XDS 5354.

The tteRBP structure was determined by molecular replacement using the ribose-bound form of the ribose binding protein from E. coli 23 as the search model 34. Rotation, translation, and fitting functions revealed a clear solution yielding higher correlation coefficients and a lower R factor than all the others. Manual model building was carried out in the programs O and COOT and refined using REFMAC5 555657. The final model for the tteRBP complex includes two intact tteRBP monomers (residues 2–275), two ribose molecules, and 346 water molecules. The model exhibits good stereochemistry as determined by PROCHECK and MolProbity; final refinement statistics are listed in Table 15859. PDB coordinates and structure factors have been deposited in the RCSB Protein Data Bank under the accession code 2IOY.