Biochemical and structural characterization of alanine racemase from Bacillus anthracis (Ames)
© Couñago et al; licensee BioMed Central Ltd. 2009
Received: 6 January 2009
Accepted: 20 August 2009
Published: 20 August 2009
Bacillus anthracis is the causative agent of anthrax and a potential bioterrorism threat. Here we report the biochemical and structural characterization of B. anthracis (Ames) alanine racemase (Alr Bax ), an essential enzyme in prokaryotes and a target for antimicrobial drug development. We also compare the native Alr Bax structure to a recently reported structure of the same enzyme obtained through reductive lysine methylation.
B. anthracis has two open reading frames encoding for putative alanine racemases. We show that only one, dal1, is able to complement a D-alanine auxotrophic strain of E. coli. Purified Dal1, which we term Alr Bax , is shown to be a dimer in solution by dynamic light scattering and has a Vmax for racemization (L- to D-alanine) of 101 U/mg. The crystal structure of unmodified Alr Bax is reported here to 1.95 Å resolution. Despite the overall similarity of the fold to other alanine racemases, Alr Bax makes use of a chloride ion to position key active site residues for catalysis, a feature not yet observed for this enzyme in other species. Crystal contacts are more extensive in the methylated structure compared to the unmethylated structure.
The chloride ion in Alr Bax is functioning effectively as a carbamylated lysine making it an integral and unique part of this structure. Despite differences in space group and crystal form, the two Alr Bax structures are very similar, supporting the case that reductive methylation is a valid rescue strategy for proteins recalcitrant to crystallization, and does not, in this case, result in artifacts in the tertiary structure.
Bacillus anthracis is a soil-dwelling, spore-forming, Gram-positive bacterium that is the causative agent of the zoonotic disease anthrax. Although the disease is most common in wild and domestic mammals, it can also occur in humans when exposed to infected animals or living spores . The severity of anthrax in humans depends on the route of infection. Inhalation of B. anthracis spores can lead to the most severe form of the disease, historically associated with a case-fatality rate as high as 85% [2, 3]. The high mortality rate, the existence of a respiratory route of infection and the great resistance of its spores has made B. anthracis the subject of biological warfare research programs in many countries for over 60 years . The United States Centers for Disease Control and Prevention (CDC) has classified anthrax as a category A bioterrorism agent, posing the greatest possible threat to public health and with the ability to spread across large areas . In 2001, the Ames strain of B. anthracis was used in a series of bioterrorist attacks that resulted in five fatalities and cost billions of dollars to the US economy [6, 7]. As B. anthracis spores are resilient, remaining viable and infective for many years, efforts to decontaminate affected facilities are time-consuming and costly. Therefore, it would be of significant importance to public health and security to develop new strategies aimed at containing B. anthracis spores upon their release into the environment.
Alanine racemase (EC 220.127.116.11) is an essential enzyme in prokaryotes. The enzyme utilizes a pyridoxal 5'-phosphate (PLP) cofactor to catalyze the racemization of L-alanine to D-alanine, an essential component of the peptidoglycan layer in bacterial cell walls. The lack of alanine racemase function in eukaryotes has made this enzyme an attractive target for antimicrobial drug development [8, 9]. In B. anthracis, the gene coding for alanine racemase, dal1, is one of only four genes up-regulated during sporulation . The dal1 gene product (Alr Bax ) is found on the spores' outermost layer  and addition of alanine racemase inhibitors has been shown to promote germination of B. anthracis' spores  while endogenous production of D-alanine, mediated by alanine racemase, inhibits germination . Triggering the premature germination of B. anthracis spores by spraying alanine racemase inhibitors on affected areas may therefore be a strategy to speed decontamination efforts and reduce the risk of infection in humans. Further, Alr Bax has recently been reported as an immunodominant protein in a proteomic analysis of the B. anthracis spore induced immunome [13, 14]. Given the importance of three-dimensional information in structure-aided inhibitor design [15–17] and its growing role in vaccine development [18–20], structural studies on Alr Bax are crucial to both of these goals.
The first structural studies of alanine racemases, which were performed on the enzyme isolated from Geobacillus (then Bacillus)stearothermophilus, revealed a homodimeric enzyme with each monomer consisting of an α/β barrel domain at the N-terminus and a C-terminal domain composed mainly of β strands. The active site is located at the interface of the α/β barrel and β domain near a PLP cofactor forming an internal aldimine linkage to a lysine residue [21, 22]. Structural studies performed in the presence of substrate analogs have identified the residues involved in catalysis and shed light onto the enzyme's catalytic mechanism . Moreover, structures of G. stearothermophilus alanine racemase with the covalent inhibitors alanine phosphonate and D-cycloserine (DCS) have shown that enzyme inhibition is due to the covalent link of these compounds to the PLP cofactor and helped explain their non-specific inhibition of eukaryotic PLP-containing enzymes [24, 25].
Alanine racemase structures from the human pathogens Pseudomonas aeruginosa and Mycobacterium tuberculosis have been solved  and both revealed further insights into the enzyme's catalytic site that may lead to identification of new, more specific inhibitors. The high-resolution (1.45 Å) structure of alanine racemase (DadX) from P. aeruginosa showed evidence for an external aldimine linkage of an unanticipated guest substrate in the active site , while the M. tuberculosis structure revealed that the narrow entryway to the enzyme's active site is composed of highly conserved residues distributed in layers beginning at the PLP site . The structure of the DCS-producing Streptomyces lavendulae has also been determined . S. lavendulae can grow in the presence of DCS, and the structural basis for the slower inhibition rate of DCS on S. lavendulae Alr has been attributed to the enzyme's larger and more rigid active site .
Here we report the cloning and characterization of the two genes, dal1 and dal2, from the B. anthracis genome with sequence similarities to other bacterial alanine racemase genes. Although expression of dal2 in a heterologous system failed, we have successfully expressed and purified the gene product of dal1, which we term Alr Bax , and performed its kinetic and structural characterization.
Recently another group has reported that B. anthracis alanine racemase crystallization required reductive methylation . Interestingly we have not found this to be the case. However, the availability of both structures, one with and one without methylation, allows for a careful comparison to be performed. Reductive methylation has been employed previously to obtain atomic structures for proteins recalcitrant to crystallization [29–33]. Due to its reported successes, this method is becoming more utilized [28, 34, 35]. Nevertheless, there may be concerns as to how methylation impacts protein structure. Our analyses of both structures suggest that despite differences in space group and crystal lattice, reductive methylation does not significantly alter the structure of alanine racemase from B. anthracis.
Results and Discussion
Sequence analysis of the Bacillus Dal proteins
In order to confirm that both dal1 and dal2 genes encode functional alanine racemases we expressed these genes in a D-alanine auxotrophic strain of E. coli, MB2795 . Expression of the dal1 gene from B. anthracis or dadX from P. aeruginosa fully restored the wild-type phenotype. Cells transformed with pET28-TEV failed to grow, as did those transformed with the B. anthracis dal2 gene (data not shown).
Overexpression, purification and biochemical characterization of Dal proteins
We used strain BL21(DE3), pLysS of E. coli to express dal1 and dal2 recombinant gene products. While dal1 was expressed successfully, the expression of dal2 failed repeatedly, even when conditions such as temperature, IPTG concentration, or strain background were changed (data not shown). Sequencing of the plasmid construct revealed no obvious errors, and our expression system has been successfully used for numerous proteins in the past. We have no conclusive explanation for our inability to express dal2 at measurable levels in E. coli. While the orf appears to encode an alanine racemase enzyme, it clearly is not expressed in the T7 overexpression system, possibly also explaining the lack of complementation observed in our earlier experiments. Given that, to our knowledge, there are no reports characterizing dal2 in the literature, we are led to believe that this gene is not usually expressed in its homologous host. As overproduction of Dal2 Bax failed, all subsequent work was performed with dal1 to yield a product, which we term Alr Bax . Alr Bax was purified to homogeneity and displayed a single peak on molecular sieve chromatography.
Previous studies have suggested Alr Bax might exist partly as a tetramer in solution . We have used dynamic light scattering (DLS) to determine that Alr Bax has a hydrodynamic radius of 3.7 nm, corresponding to a molecular weight of 93 kDa. As the calculated molecular weight of Alr Bax is 43.7 kDa, this enzyme is unambiguously a dimer (ca. molecular weight 87.4 kDa) in solution under the conditions of this experiment. These measurements were made on a monodisperse solution of Alr Bax in which 99.9% of the mass was accounted for by the single peak at 3.7 nm.
We find that purified Alr Bax has a K m for D-alanine of 2.8 mM and a V max of 101 U mg-1, where one unit was defined as the amount of enzyme that catalyzed racemization of 1 μmol of substrate per minute. These kinetic parameters for racemization of L- to D-alanine of Alr Bax fall in the range of what has been observed before for other bacterial alanine racemases [38–40]. Interestingly, despite the high identity levels observed for residues in the active site of Alr Bax and Alr Gst (Figure 1), the V max for the anthrax enzyme is one order of magnitude lower than the one reported for the G. stearothermophilus enzyme and closer to that observed for alanine racemases isolated from other pathogenic organisms. Our kinetic characterization reinforces previous observations that there is a very wide dynamic range in kinetic constants for alanine racemase, despite the sequence and structural similarities of their active sites.
Description of the Overall Structure of Alr Bax from B. anthracis
Data-collection and structure-refinement statistics.
a = 49.62 Å, b = 141.27 Å, c = 60.12 Å
α = γ = 90.00°, β = 103.11°
Rfree (%) (for 1627 reflections)
Average B factor (Å2)
bond length (Å)
bond angles (deg.)
no. of residues
no. of protein atoms
no. of PLP atoms
no. of acetate atoms
no. of water atoms
no. of chloride ions
Structural comparisons of Alr Bax with closely related enzymes
Average r.m.s.d. (Å) between the Cα atoms of AlrBax and other Alrs
As noted, the B. anthracis Alr Bax secondary structure and general fold closely resembles that seen for other alanine racemases . However, there are a few small topological differences between the structures of Alr Gst and Alr Bax . Alr Bax is five residues longer than Alr Gst ; three of the five extra residues in Alr Bax extend Helix 8 by one turn; while the remaining two extra residues locate to the very N-terminus of Alr Bax . Helix 8 does not take part in the enzyme's active site nor does it make intermonomer contacts, therefore, we do not anticipate this secondary structure to play a critical role in Alr Bax function.
Least-squares superposition of Cα atoms from N- and C-terminal domains from Alr Gst , Alr Sla , Alr Mtb and DadX Pao to the equivalent domains in Alr Bax reveals average rms differences ranging from 1.10 to 2.30 Å. The rms differences correlate with sequence identity levels (Table 2). Superposition of the N-terminal domains of Alr Bax and Alr Gst reveals significant Cα deviations (≥ 1.8 Å) for residues in three loops (residues 121–125, between H6 and S6; residues 198–202, between H8 and S8; residues 215–219, between H9 and S9) and residues 148–158 on H7. These regions all locate to the protein surface and have no reported role in homodimer formation or substrate binding and catalysis. On the other hand, superposition of the C-terminal domains of Alr Bax and Alr Gst shows no regions with Cα rms differences greater than 1.4 Å.
Alr Bax and Alr Gst have a similar hinge angle between N- and C-terminal domains
This analysis notwithstanding, the relevance of the hinge angle to enzyme catalysis in alanine racemase remains unclear. Although the V max values reported for all alanine racemases studied to date vary by over three orders of magnitude, it is not straightforward to attribute these differences to hinge angle. Differences in hinge angles certainly alter the relative orientation of the two active sites in the dimer, but affect very little the geometry of each active site as indicated in Table 2. Further, the V max for Alr Gst and Alr Bax enzymes varies by more than 10 fold, despite having similar hinge angles. Altering the hinge angle of this enzyme experimentally through mutation or cassette swapping may resolve this issue.
Intermonomer contacts and surface area
The dimer interface in alanine racemase is an important area for structural analysis. Only the dimeric form of the enzyme is catalytically active . Therefore, interface residues are critical in forming a functional active site. Certainly the interface functions to correctly position the second catalytic tyrosine residue from the opposite monomer on top of the active site. In addition, both monomers contribute to the overall composition of the alanine entryway and binding pocket. Loss of interface contacts would alter this arrangement and could be used as a strategy to inhibit Alr activity. Disruption of dimer interfaces is becoming more common and has been successfully used recently for drug targets in HIV and HCV [42–44].
Intermonomer interactions for alanine racemases
Intermonomer surface area (Å2)
Alr Bax PLP-binding and active site
The identity and position of active site residues is strongly conserved among various Alrs (Table 2). As a result, the hydrogen bonding network found for the PLP molecule in the active site of Alr Bax is similar to the one observed for other Alrs. In Alr Bax , side chain atoms from Tyr45, Arg138, Arg24, His168, Ser209 and Tyr359 establish hydrogen bonds to atoms in the PLP cofactor (Figure 6). These residues are strictly conserved for Alr Gst , Alr Mtb , Alr Sla and DadX Pao and have similar orientations in the PLP-binding site of their respective enzymes. The PLP in Alr Bax also hydrogen bonds with main chain atoms from Ser209, Gly226 and Ile227. The first two of these residues is strictly conserved in Alr Gst , Alr Mtb , Alr Sla and DadX Pao . The third would be as well but in Alr Sla , the Ile227 is replaced by a leucine residue. Perhaps a more significant difference is the presence in Alr Sla and Alr Mtb of a tryptophan residue in place of Alr Bax Leu87. A tryptophan residue at this position is one of the differences found between the active sites of the slower enzymes from M. tuberculosis and S. lavendulae and the faster Alr from G. stearothermophilus. In Alr Sla and Alr Mtb , the Nε atom of this tryptophan makes a water-mediated hydrogen bond to O3 from PLP. Although this extra interaction may have a role in catalysis it does not seem to reduce the size of the Alr Sla and Alr Mtb active sites as the loop that harbors this tryptophan residue is shifted away (~2.1 Å) from the PLP cofactor when compared to the same loop in Alr Bax . Mutagenesis studies could thus be performed in order to evaluate the impact of this tryptophan residue for enzyme catalysis.
One striking difference in the active site involves Asn131, which in other alanine racemases is generally a carbamylated lysine that participates in a hydrogen bond with the residue homologous to Arg138. In Alr Bax , however, we note a prominent chloride ion that is located near Arg138 in the active site (Figure 6). This chloride ion has not been described in Alr structures from other species and it was originally modeled by us as a water molecule. However, the resulting low B-factor (~10 Å2) and its hexa-coordination with three water molecules and atoms Nδ2 from Asn131 and Nε and Nη2 from Arg138 suggested the presence of a chloride ion. Notably, there is no chloride present in the crystallization buffer and we can only assume that the enzyme binds so tightly to this halide that it is carried over from the enzyme's purification. The chloride ion is also observed on the Alr Bax structure obtained following lysine reductive methylation . The presence of a chloride ion in two independent structures reinforces the idea that this ion plays an important structural role in Alr Bax . Other Alrs have a negative charge at the same position, but the charge has always been from a carbamylated lysine residue (Figure 6). In the Alr Mtb structure a carbamylated lysine was not noted but the side chain density for this lysine was poor. Like the chloride ion in Alr Bax , the carbamyl group found in other Alrs hydrogen bonds with Nε and Nη2 from the active site arginine (Arg138 in Alr Bax ), thus positioning this residue in the active site. The general conservation of the modified lysine residue among various Alrs and its role in positioning the active site arginine indicates that the presence of a negative charge at this position is critical for enzyme catalysis. As Alr Bax lacks the conserved lysine residue necessary for carbamylation it has apparently drafted a chloride ion to fill the same role for this species. It is open to speculation whether the addition of chloride chelators like SPQ (6-methoxy-N-(3-sulfopropyl)-quinolinium) would affect the enzyme activity and whether it might be possible to design specific inhibitors for Alr Bax based on this unique interaction.
In Alr Bax in addition to the interactions facilitated by the chloride ion, Arg138 is further positioned by the side chain oxygen of Thr316'. Further, an alignment of 105 Alrs, having between 24% to 99% sequence identity to Alr Bax , revealed that the presence of an asparagine at the equivalent position to Asn131 in Alr Bax is always accompanied by the presence of a threonine residue equivalent to Thr316' (data not shown) suggesting that this interaction with Arg138 would be a conserved feature of alanine racemases with active site structural chlorides. Sequences of Alrs that contain a lysine in position 131 almost always have an accompanying serine or a cysteine residue in position equivalent to Alr Bax Thr316'. In the case of Alr Pao this serine is involved in an equivalent active site arginine interaction. The exception to this latter observation is Alr Sla which has an alanine at this position. It is important to note that there is not really a specific chloride-binding motif as the residues that interact with Cl- in Alr Bax are the same that interact with the carbamylated lysine in the other structures.
The active site entryway of Alr Bax
Residues from loops in the α/β barrel domain of one monomer and residues from the C-terminal domain of the second monomer make up an entryway to the active site and the PLP binding site. The active site entryway of Alr has been previously divided in inner, middle and outer layers, starting from the PLP binding pocket and moving towards the protein surface . Residues in the inner and middle layers show strong conservation among various Alrs . For Alr Bax , residues Tyr270', Tyr359, Tyr289' and Ala172 constitute the inner layer, while residues Arg314', Ile357, Arg295' and Asp173 make up the middle layer. These residues are absolutely conserved between Alr Bax and Alr Gst , Alr Mtb , Alr Sla and DadX Pao . The outer layer for the active site entryway of various Alrs displays less conservation, but in this region Alr Bax contains an Asn271' while Alr Gst , Alr Mtb , Alr Sla and DadX Pao contain a glycine. As a result of this substitution, the entryway is somewhat more restricted than the ones observed for other alanine racemases. Finally, for Alr Bax a conserved pair of acidic residues (Asp-Glu) is found at positions 173 and 174, which are located in the middle and outer layers of the entryway. Identical residues are found in the same position for Alr Gst , Alr Sla and DadX Pao , but for Alr Mtb , a much slower alanine racemase, these two residues are (Asp-Lys). This site has recently been shown to be important catalytically, as making this Asp-Glu to Asp-Lys change at the same position in E. coli alanine racemase has been shown to significantly decrease its catalytic rate .
Structural comparison of native and reductively methylated alanine racemases from B. anthracis
Recently, the structure of Alr Bax after reductive methylation of its lysine residues (Alr BaxRM ) has been reported . In that report, the unmodified protein failed to crystallize. Scientists at the Oxford Protein Production Facility (OPPF) and the York Structural Biology laboratory reported that extensive crystallization trials (approximately 800 conditions) with native Alr Bax proved unsuccessful and that reductive lysine methylation was essential for crystallization of the protein [34, 46]. Based on data from mass spectroscopy and on the methylated crystal structure of Alr Bax , Au and colleagues concluded that the N terminus and 18 out of the 20 lysines in Alr Bax were methylated after the protein was treated with dimethyl-amine-borane complex and formaldehyde.
Reductive methylation modifies all free primary amines in a protein molecule (NH groups from lysine residues and the N terminus) to tertiary amines. This modification of lysine residues, especially those found on the protein surface, offers an opportunity to change a protein's crystallization properties and is a proven method to rescue proteins recalcitrant to crystallization [28–33, 35, 47]. However, there are few structural studies showing that reductive alkylation does not alter a protein's structure, especially of proteins that do not readily crystallize. One study  reporting on the effects of reductive lysine methylation on HEW lysozyme found that crystals were formed under different conditions and with a different crystalline lattice than observed for the unmodified enzyme. Nevertheless, the structures of both modified and unmodified enzymes showed no significant structural differences and their superimposed Cα atoms had an rms difference of only 0.4 Å . The availability of native and modified structures for Alr Bax , therefore, offers another opportunity to evaluate the impact of reductive lysine methylation, this time on a protein more recalcitrant to crystallization.
In our hands, Alr Bax protein readily formed small crystals using commercially available crystallization screens. Notably our form contains eight additional residues at the C-terminus that remain following cleavage of a C-terminal His-tag using TEV protease. These residues are not involved in crystal contacts, but still could have an influence on crystallization. Our initial crystallization conditions required extensive fine-tuning, and the addition of the glutathione additive proved important for obtaining diffraction quality crystals. Moreover, finding the proper conditions for freezing Alr Bax crystals without compromising diffraction quality proved challenging. For simplicity's sake we have referred to this form of Alr Bax as unmethylated or native. Our review of the expression and purification protocols for both native and alkylated enzymes suggests that they were very similar. Also, modified and unmodified Alr Bax crystallize under similar conditions, despite a reported small reduction in the isoelectric point and the expected changes in the surface properties of Alr Bax . Both proteins were crystallized in the presence of PEG (18% PEG 8000 for the native and 25% PEG 3350 for the modified protein), high salt concentrations (0.2 M sodium acetate for the native and 0.2 M magnesium chloride for the modified protein) and at the same pH, 6.5. Interestingly, the modified enzyme was crystallized at 60 mg/ml while the native structure was obtained from crystals grown at 15 mg/ml.
Despite similar crystallization conditions native and modified Alr Bax crystals show different crystalline lattices and solvent content. Native Alr Bax crystals are monoclinic with space group P21 and unit cell parameters a = 49.6 Å, b = 141.3 Å, c = 60.1 Å and β = 103.11°. On the other hand, crystals for the methylated enzyme are orthorhombic in space group P212121 with cell dimensions of a = 57.6 Å b = 88.4 Å and c = 139.0 Å. Crystals for the modified enzyme display a lower solvent content (38% vs. 48%) and a higher packing density (1.99 Å3/Da vs. 2.35 Å3/Da) than native crystals.
Crystal contacts comparison
The surfaces of the modified and native Alr Bax crystals are also different in terms of metal and halide content. Four magnesium and three chloride ions were found on the surface of modified Alr Bax and take part in crystal contacts. For the native Alr Bax structure we did not identify any metal or halide ions at equivalent positions. Furthermore, the temperature factors for these surface ions are quite low, with four less than 20 Å2, and many are involved in extensive electrostatic interactions. Perhaps the presence of additional metal ions observed exclusively for the methylated crystal form of Alr Bax acts to compensate for the loss of positive charges at the protein surface.
Most importantly, reductive methylation did not alter the overall fold of Alr Bax . Structural alignment of methylated and native Alr Bax shows no significant difference in their overall structures. For the individual monomers the rms difference between their Cα atoms is just under 0.4 Å. Alignment of the active site residues from the two Alr Bax structures shows that reductive alkylation of the enzyme did not result in any significant changes in the position and hydrogen bond pattern of active site residues and the PLP co-factor. Moreover, the hinge angle between N- and C-terminal domains is very similar for both modified and unmodified Alr Bax . Thus, the hinge angles observed for Alr Bax are inherent to this particular enzyme and not an artifact of crystallization. As an aside, this observation makes a strong argument that the disparate hinge angles observed for other Alrs are not a consequence of divergent crystal packing.
Reductive methylation also did not significantly alter the dimer interface, which is found to be comparable between methylated and unmethylated structures (3600 Å2 vs. 3500 Å2, respectively). For the modified structure, two methylated lysines contribute atoms from their methyl groups to the interface; Mly182 and Mly255. The corresponding lysines in the native structure are not considered to be part of the interface; Lys182 displays poor density and did not have its complete side chain modeled in the native structure and no atoms from Lys255 in one monomer are in contact distance to atoms in the other monomer.
Only two lysines escaped methylation in the modified crystal structure, Lys41 and Lys260 . Lys41 is found covalently bound to the PLP co-factor. Thus its NH group is not a primary amine and is not surprising that this residue is unaffected by the reductive methylation protocol. Lys260 is the lysine residue least exposed to the solvent and it makes hydrogen bonds to Gly137 and Arg138 which, in turn, hydrogen bonds to the phenolic oxygen of the PLP cofactor and to the substrate (see above). These two residues are, therefore, involved in either a covalent bond or a strong polar interaction in the present structure and thus predictably escaped reductive methylation.
In conclusion, we report the high-resolution crystal structure of alanine racemase from the dal1 gene of B. anthracis and characterize it kinetically and in an E. coli complementation system. This structure contains some unique features in its active site including a structural chloride atom. It shares a similar hinge angle to its close relative from Geobacillus and has an active site and topology much like other members of this family. Based on the results shown here the active site of Alr Bax is as accessible for inhibitor binding as other alanine racemases studied to date. Furthermore, it is very likely that alanine racemase inhibitors like D-cycloserine or alanine phosphonate will be effective as modulators of sporulation. Finally, as treatment of spores will take place in the environment and not internally, the problems associated with non-specific PLP inhibition ascribed to these inhibitors should not detract from their usefulness in bioremediation. We look forward to exploring more structural studies on these inhibitors as they become available.
Amplification and cloning of the B. anthracis alr genes
Two putative open reading frames, dal1 and dal2, for alanine racemase from B. anthracis were identified through sequence comparisons using the known alanine racemase sequence from G. stearothermophilus  as a probe against the B. anthracis genome deposited in GenBank . Two sets of primers were used in PCR to amplify the two putative alr genes from genomic DNA of B. anthracis (Ames), dal1–5' (5'-GGG GCC ATG GAA GAA GCA CCA TTT TAT CGT G-3')/dal1–3' (5'-CCC CCT CGA GTA TAT CGT TCA AAT AAT TAA TTA C-3') and dal2–5' (5'-GGG GCA TAT GAG TTT GAA ATA TGG AAG AG-3')/dal2–3' (5' CCCCCTGCAGAATCCGTAGGTTTTAAGGAC 3'), resulting in amplicons of 1169 bp and 1175 bp, respectively. The PCR products were sequenced, inserted into a modified pET28 vector (pET28-TEV) containing a C-terminal His-tag and a TEV protease cleavage sequence, LEENLYFQ/SLQVEH6 and cloned in E. coli MB1547. (/) denotes the location of the cleavage site.
Characterization of the two cloned genes continued with their transformation into the D-alanine auxotrophic E. coli strain MB2795 . A plasmid encoding the cloned P. aeruginosa DadX alanine racemase, pMB1921 , was used as a positive control. Plasmid pET28-TEV without any inserts served as the negative control. Cells were grown on solid LB medium with and without D-alanine supplementation, and scored for colony growth after 16 h at 37°C as described previously .
Dal1 overexpression and purification
Cultures of E. coli BL21(DE3), pLysS containing the pET28-TEV-dal expression plasmids were grown at 37°C in LB medium containing 100 μg/ml kanamycin and 30 μg/ml chloramphenicol to an OD600 of 0.8. Expression of recombinant proteins was induced by addition of 0.5 mM IPTG and carried at 30°C for 19 hours. Cells were harvested by centrifugation and the cell lysate was cleared and loaded onto a Hi Trap affinity (Ni2+) column (GE Healthcare Life Sciences). The column was washed and Alr Bax eluted with a stepwise imidazole gradient. The C-terminal 6xHis tag was removed by treatment with His-tagged TEV protease (1 mg TEV protease per 10 mg of protein for 16 hours at 4°C). Alr Bax without the 6xHis tag was purified from the reaction mixture using the same chromatography strategy described above. Following concentration, Alr Bax was loaded onto a Pharmacia Superdex 200 Preparative Grade column; sample purity was assessed by SDS-PAGE to be greater than 95%.
Dynamic light scattering
Purified Alr Bax was dialyzed against 20 mM Tris pH 8.0. Protein samples (1 mg/ml) were centrifuged (10 min. at 14,000 rpm) and filtered using 0.02 μm Whatman Anotop filters prior to recording data. All measurements were made at 298 K using the DynaPro system according to the manufacturer's instructions (Wyatt Technology).
Enzyme Kinetics and Crystallization
The kinetic parameters (K m and V max ) for the racemization reaction (D- to L-alanine) catalyzed by Alr Bax were estimated using the spectrophotometric alanine racemase assay as described previously . Alr Bax crystallization screening trials were performed using the vapor diffusion method with sitting drops (5 μl of protein at 15 mg/ml and 5 μl of mother liquor) in 24-well plates incubated at 4°C. Initial screens revealed thin needle crystals growing in 20% PEG 8000, 0.2 M sodium acetate, 0.1 M sodium cacodylate, pH 6.5 . Crystals were optimized using streak-seeding with crushed crystals and further optimized using additive screening resulting in rectangular, deep yellow crystals suitable for data collection. The final crystallization condition was 18% PEG 8000, 0.2 M sodium acetate, 0.1 M sodium cacodylate, pH 6.5, 0.01 M GSH (L-glutathione reduced), 0.01 M GSSG (L-glutathione oxidized).
Data Collection and Processing
Crystals were passed through cryoprotectant solutions consisting of 20.7% PEG 8000, 0.2 M sodium acetate, 0.1 M sodium cacodylate supplemented with 3, 6, 9, 12, 15 and 18% (v/v) ethylene glycol, mounted into a nylon loop and flash frozen in liquid nitrogen at 110 K. A native data set was collected at 110 K on a Micromax 007 HF rotating-anode X-ray generator equipped with a copper anode, Hi-res optics, an RAXIS IV++ image-plate detector (Rikagu) using a frame width of 0.5° and an exposure time of 600 s. Images were integrated using MOSFLM , processed with SCALA  and analyzed using programs from the CCP4 suite . Data collection and processing statistics for the native data set can be found in Table 1. Alr Bax crystallized in space group P21 with unit cell parameters a = 49.62 Å, b = 141.27 Å, c = 60.12 Å and β = 103.11. There is one Alr Bax dimer per asymmetric unit.
Structure Determination and Refinement
Molecular replacement was carried out with MolRep  using the G. stearothermophilus Alr (PDB entry – 1SFT) atomic coordinates . Molecular replacement was performed assuming two monomers per asymmetric unit as suggested by a Matthew's coefficient of 2.35  and resulted in the proper orientation of the search model in the crystal lattice (Rfac 43.6%; score 0.699). The primary sequence of the search model was changed to that of Alr Bax using Coot . All structural refinements (32.79 – 1.95 Å) were carried in Refmac5  using standard restraints and were followed by visual inspection of protein models and density maps in Coot. Ten cycles of positional refinement, performed using NCS restraints, resulted in R and R free of 23.9 and 27.2%, respectively. Waters were added using the arp_water function on Refmac5, and when the active site density was clearly interpretable, PLP was added to both active sites. A further 10 cycles of positional and Biso refinements brought R and R free to 19.6 and 23.7%, respectively. Water molecules with B-factors higher than 55.0 Å2 and electron density lower than 1.0 σ on a 2F obs – Fcalc map were then deleted.
B. anthracis crystals displayed somewhat anisotropic x-ray diffraction and previous alanine racemase structures have shown indication of subdomain movement. This encouraged us to try TLS refinement . TLS analyses were carried on with different domains of the protein acting as a rigid body. All models resulted in similar improvements in R and R free and in the end we adopted the most parsimonious one, which treated all protein atoms found in the asymmetric unit as a rigid body. After TLS refinement, the R and R free were 16.0 and 20.1% with root-mean-square deviations from ideality for bond lengths of 0.017 and angles of 1.46° (Table 1). As noted above, inclusion of the C-terminal His-tag has resulted in eight additional residues in our sequence. In the final map we attempted to build some of these residues into extra density at the C-terminus, but as we did not gain anything in terms of R or R free we have elected to leave out the extra residues from this region in the final structure. Structure factors and final atomic coordinates for Alr Bax have been deposited in the Protein Databank (PDB ID 3ha1).
The structure of Alr Bax was compared to other closely related enzymes; their accession numbers are: 1sft – Alr Gst bound with acetate ; 1vfh – Alr Sla with no ligand ; 2vd8 – methylated Alr Bax , 1rcq – DadX Pao  and 1xfc – Alr Mtb . Structural alignments were performed using SSM . Interface surface area was calculated using PISA . The number of polar contacts (hydrogen bonds and salt bridges) was determined using WHAT IF [61, 62].
List of Abbreviations
root mean square
This work was supported by funding from the University of Otago, the Robert A. Welch Foundation, the National Institutes of Health, the Thrash Foundation and the Foundation for the Centers for Molecular Research in Infectious Diseases. We would like to dedicate this manuscript to John Francis Thrash, MD, a noted philanthropist, physician and businessman from Houston, Texas who generously aided in the establishment of our structural biology laboratory in New Zealand.
- Dixon TC, Meselson M, Guillemin J, Hanna PC: Anthrax. N Engl J Med 1999, 341(11):815–826. 10.1056/NEJM199909093411107View ArticlePubMedGoogle Scholar
- Dahlgren CM, Buchanan LM, Decker HM, Freed SW, Phillips CR, Brachman PS: Bacillus anthracis aerosols in goat hair processing mills. Am J Hyg 1960, 72: 24–31.PubMedGoogle Scholar
- Meselson M, Guillemin J, Hugh-Jones M, Langmuir A, Popova I, Shelokov A, Yampolskaya O: The Sverdlovsk anthrax outbreak of 1979. Science 1994, 266(5188):1202–1208. 10.1126/science.7973702View ArticlePubMedGoogle Scholar
- CDC: Use of anthrax vaccine in the United States: recommendations of the Advisory Committee on Immunization Practices (ACIP). MMWR 2000, 49(RR15):1–20.Google Scholar
- Radosavljevic V, Jakovljevic B: Bioterrorism – types of epidemics, new epidemiological paradigm and levels of prevention. Public Health 2007, 121(7):549–557. 10.1016/j.puhe.2006.12.010View ArticlePubMedGoogle Scholar
- Jernigan JA, Stephens DS, Ashford DA, Omenaca C, Topiel MS, Galbraith M, Tapper M, Fisk TL, Zaki S, Popovic T, et al.: Bioterrorism-related inhalational anthrax: the first 10 cases reported in the United States. Emerg Infect Dis 2001, 7(6):933–944.PubMed CentralView ArticlePubMedGoogle Scholar
- Read TD, Salzberg SL, Pop M, Shumway M, Umayam L, Jiang L, Holtzapple E, Busch JD, Smith KL, Schupp JM, et al.: Comparative genome sequencing for discovery of novel polymorphisms in Bacillus anthracis . Science 2002, 296(5575):2028–2033. 10.1126/science.1071837View ArticlePubMedGoogle Scholar
- Lambert MP, Neuhaus FC: Mechanism of D-cycloserine action: alanine racemase from Escherichia coli W. J Bacteriol 1972, 110(3):978–987.PubMed CentralPubMedGoogle Scholar
- Silverman RB: The potential use of mechanism-based enzyme inactivators in medicine. J Enzyme Inhib 1988, 2(2):73–90. 10.3109/14756368809040714View ArticlePubMedGoogle Scholar
- Todd SJ, Moir AJ, Johnson MJ, Moir A: Genes of Bacillus cereus and Bacillus anthracis encoding proteins of the exosporium. J Bacteriol 2003, 185(11):3373–3378. 10.1128/JB.185.11.3373-3378.2003PubMed CentralView ArticlePubMedGoogle Scholar
- Titball RW, Manchee RJ: Factors affecting the germination of spores of Bacillus anthracis . J Appl Bacteriol 1987, 62(3):269–273.View ArticlePubMedGoogle Scholar
- McKevitt MT, Bryant KM, Shakir SM, Larabee JL, Blanke SR, Lovchik J, Lyons CR, Ballard JD: Effects of endogenous D-alanine synthesis and autoinhibition of Bacillus anthracis germination on in vitro and in vivo infections. Infect Immun 2007, 75(12):5726–5734. 10.1128/IAI.00727-07PubMed CentralView ArticlePubMedGoogle Scholar
- Delvecchio VG, Connolly JP, Alefantis TG, Walz A, Quan MA, Patra G, Ashton JM, Whittington JT, Chafin RD, Liang X, et al.: Proteomic profiling and identification of immunodominant spore antigens of Bacillus anthracis , Bacillus cereus , and Bacillus thuringiensis . Appl Environ Microbiol 2006, 72(9):6355–6363. 10.1128/AEM.00455-06PubMed CentralView ArticlePubMedGoogle Scholar
- Huang CM, Foster KW, DeSilva TS, Van Kampen KR, Elmets CA, Tang DC: Identification of Bacillus anthracis proteins associated with germination and early outgrowth by proteomic profiling of anthrax spores. Proteomics 2004, 4(9):2653–2661. 10.1002/pmic.200400831View ArticlePubMedGoogle Scholar
- Veerapandian B: Three Dimensional Structure-Aided Drug Design. In Burger's Medicinal Chemistry and Drug Discovery. Volume 1. 5th edition. Edited by: Wolff ME. New York, New York: John Wiley & Sons, Inc.; 1995:303–348.Google Scholar
- Marrone TJ, Briggs JM, McCammon JA: Structure-based drug design: computational advances. Annu Rev Pharmacol Toxicol 1997, 37: 71–90. 10.1146/annurev.pharmtox.37.1.71View ArticlePubMedGoogle Scholar
- Blundell TL: Structure-based drug design. Nature 1996, 384(Supp):23–26.PubMedGoogle Scholar
- Blythe MJ, Flower DR: Benchmarking B cell epitope prediction: underperformance of existing methods. Protein Sci 2005, 14(1):246–248. 10.1110/ps.041059505PubMed CentralView ArticlePubMedGoogle Scholar
- Pizarro JC, Vulliez-Le Normand B, Chesne-Seck ML, Collins CR, Withers-Martinez C, Hackett F, Blackman MJ, Faber BW, Remarque EJ, Kocken CH, et al.: Crystal structure of the malaria vaccine candidate apical membrane antigen 1. Science 2005, 308(5720):408–411. 10.1126/science.1107449View ArticlePubMedGoogle Scholar
- Tolia NH, Enemark EJ, Sim BK, Joshua-Tor L: Structural basis for the EBA-175 erythrocyte invasion pathway of the malaria parasite Plasmodium falciparum . Cell 2005, 122(2):183–193. 10.1016/j.cell.2005.05.033View ArticlePubMedGoogle Scholar
- Morollo AA, Petsko GA, Ringe D: Structure of a Michaelis complex analogue: propionate binds in the substrate carboxylate site of alanine racemase. Biochemistry 1999, 38(11):3293–3301. 10.1021/bi9822729View ArticlePubMedGoogle Scholar
- Shaw JP, Petsko GA, Ringe D: Determination of the structure of alanine racemase from Bacillus stearothermophilus at 1.9 Å resolution. Biochemistry 1997, 36(6):1329–1342. 10.1021/bi961856cView ArticlePubMedGoogle Scholar
- LeMagueres P, Im H, Dvorak A, Strych U, Benedik M, Krause KL: Crystal structure at 1.45 Å resolution of alanine racemase from a pathogenic bacterium, Pseudomonas aeruginosa , contains both internal and external aldimine forms. Biochemistry 2003, 42(50):14752–14761. 10.1021/bi030165vView ArticlePubMedGoogle Scholar
- Fenn TD, Stamper GF, Morollo AA, Ringe D: A side reaction of alanine racemase: transamination of cycloserine. Biochemistry 2003, 42(19):5775–5783. 10.1021/bi027022dView ArticlePubMedGoogle Scholar
- Stamper GF, Morollo AA, Ringe D: Reaction of alanine racemase with 1-aminoethylphosphonic acid forms a stable external aldimine. Biochemistry 1998, 37(29):10438–10445. 10.1021/bi980692sView ArticlePubMedGoogle Scholar
- LeMagueres P, Im H, Ebalunode J, Strych U, Benedik MJ, Briggs JM, Kohn H, Krause KL: The 1.9 Å crystal structure of alanine racemase from Mycobacterium tuberculosis contains a conserved entryway into the active site. Biochemistry 2005, 44(5):1471–1481. 10.1021/bi0486583View ArticlePubMedGoogle Scholar
- Noda M, Matoba Y, Kumagai T, Sugiyama M: Structural evidence that alanine racemase from a D-cycloserine-producing microorganism exhibits resistance to its own product. J Biol Chem 2004, 279(44):46153–46161. 10.1074/jbc.M404605200View ArticlePubMedGoogle Scholar
- Au K, Ren J, Walter TS, Harlos K, Nettleship JE, Owens RJ, Stuart DI, Esnouf RM: Structures of an alanine racemase from Bacillus anthracis (BA0252) in the presence and absence of (R)-1-aminoethylphosphonic acid (L-Ala-P). Acta Crystallogr Sect F Struct Biol Cryst Commun 2008, 64(Pt 5):327–333. 10.1107/S1744309108007252PubMed CentralView ArticlePubMedGoogle Scholar
- Kobayashi M, Kubota M, Matsuura Y: Crystallization and improvement of crystal quality for x-ray diffraction of maltooligosyl trehalose synthase by reductive methylation of lysine residues. Acta Crystallogr D Biol Crystallogr 1999, 55(Pt 4):931–933. 10.1107/S0907444999002115View ArticlePubMedGoogle Scholar
- Kurinov IV, Mao C, Irvin JD, Uckun FM: X-ray crystallographic analysis of pokeweed antiviral protein-II after reductive methylation of lysine residues. Biochem Biophys Res Commun 2000, 275(2):549–552. 10.1006/bbrc.2000.3329View ArticlePubMedGoogle Scholar
- Rayment I, Rypniewski WR, Schmidt-Base K, Smith R, Tomchick DR, Benning MM, Winkelmann DA, Wesenberg G, Holden HM: Three-dimensional structure of myosin subfragment-1: a molecular motor. Science 1993, 261(5117):50–58. 10.1126/science.8316857View ArticlePubMedGoogle Scholar
- Saxena AK, Singh K, Su HP, Klein MM, Stowers AW, Saul AJ, Long CA, Garboczi DN: The essential mosquito-stage P25 and P28 proteins from Plasmodium form tile-like triangular prisms. Nat Struct Mol Biol 2006, 13(1):90–91. 10.1038/nsmb1024View ArticlePubMedGoogle Scholar
- Schubot FD, Waugh DS: A pivotal role for reductive methylation in the de novo crystallization of a ternary complex composed of Yersinia pestis virulence factors YopN, SycN and YscB. Acta Crystallogr D Biol Crystallogr 2004, 60(Pt 11):1981–1986. 10.1107/S0907444904023005View ArticlePubMedGoogle Scholar
- Walter TS, Meier C, Assenberg R, Au KF, Ren J, Verma A, Nettleship JE, Owens RJ, Stuart DI, Grimes JM: Lysine methylation as a routine rescue strategy for protein crystallization. Structure 2006, 14(11):1617–1622. 10.1016/j.str.2006.09.005View ArticlePubMedGoogle Scholar
- Kim Y, Quartey P, Li H, Volkart L, Hatzos C, Chang C, Nocek B, Cuff M, Osipiuk J, Tan K, et al.: Large-scale evaluation of protein reductive methylation for improving protein crystallization. Nat Methods 2008, 5(10):853–854. 10.1038/nmeth1008-853PubMed CentralView ArticlePubMedGoogle Scholar
- Read TD, Peterson SN, Tourasse N, Baillie LW, Paulsen IT, Nelson KE, Tettelin H, Fouts DE, Eisen JA, Gill SR, et al.: The genome sequence of Bacillus anthracis Ames and comparison to closely related bacteria. Nature 2003, 423(6935):81–86. 10.1038/nature01586View ArticlePubMedGoogle Scholar
- Huang CM, Elmets CA, Tang DC, Li F, Yusuf N: Proteomics reveals that proteins expressed during the early stage of Bacillus anthracis infection are potential targets for the development of vaccines and drugs. Genomics Proteomics Bioinformatics 2004, 2(3):143–151.PubMedGoogle Scholar
- Strych U, Penland RL, Jimenez M, Krause KL, Benedik MJ: Characterization of the alanine racemases from two mycobacteria. FEMS Microbiol Lett 2001, 196(2):93–98. 10.1111/j.1574-6968.2001.tb10547.xView ArticlePubMedGoogle Scholar
- Inagaki K, Tanizawa K, Badet B, Walsh CT, Tanaka H, Soda K: Thermostable alanine racemase from Bacillus stearothermophilus : molecular cloning of the gene, enzyme purification, and characterization. Biochemistry 1986, 25(11):3268–3274. 10.1021/bi00359a028View ArticlePubMedGoogle Scholar
- Strych U, Huang HC, Krause KL, Benedik MJ: Characterization of the alanine racemases from Pseudomonas aeruginosa PAO1. Curr Microbiol 2000, 41(4):290–294. 10.1007/s002840010136View ArticlePubMedGoogle Scholar
- Strych U, Benedik MJ: Mutant analysis shows that alanine racemases from Pseudomonas aeruginosa and Escherichia coli are dimeric. J Bacteriol 2002, 184(15):4321–4325. 10.1128/JB.184.15.4321-4325.2002PubMed CentralView ArticlePubMedGoogle Scholar
- Boggetto N, Reboud-Ravaux M: Dimeric Inhibitors of HIV-1 Protease. Biol Chem 2002, 383(9):1321–1324. 10.1515/BC.2002.150View ArticlePubMedGoogle Scholar
- Song M, Rajesh S, Hayashi Y, Kiso Y: Design and Synthesis of New Inhibitors of HIV-1 Protease Dimerization with Conformationally Constrained Templates. Bioorg Med Chem Lett 2001, 11(18):2465–2468. 10.1016/S0960-894X(01)00468-1View ArticlePubMedGoogle Scholar
- Strosberg AD: Breaking the spell: drug discovery based on modulating protein-protein interactions. Expert Rev Proteomics 2004, 1(2):141–143. 10.1586/14789418.104.22.168View ArticlePubMedGoogle Scholar
- Wu D, Hu T, Zhang L, Chen J, Du J, Ding J, Jiang H, Shen X: Residues Asp164 and Glu165 at the substrate entryway function potently in substrate orientation of alanine racemase from E . coli : Enzymatic characterization with crystal structure analysis. Protein Sci 2008, 17(6):1066–1076. 10.1110/ps.083495908PubMed CentralView ArticlePubMedGoogle Scholar
- Au K, Berrow NS, Blagova E, Boucher IW, Boyle MP, Brannigan JA, Carter LG, Dierks T, Folkers G, Grenha R, et al.: Application of high-throughput technologies to a structural proteomics-type analysis of Bacillus anthracis . Acta Crystallogr D Biol Crystallogr 2006, 62(Pt 10):1267–1275. 10.1107/S0907444906033555View ArticlePubMedGoogle Scholar
- Rauert W, Eddine AN, Kaufmann SH, Weiss MS, Janowski R: Reductive methylation to improve crystallization of the putative oxidoreductase Rv0765c from Mycobacterium tuberculosis . Acta Crystallogr Sect F Struct Biol Cryst Commun 2007, 63(Pt 6):507–511. 10.1107/S1744309107022506PubMed CentralView ArticlePubMedGoogle Scholar
- Rypniewski WR, Holden HM, Rayment I: Structural consequences of reductive methylation of lysine residues in hen egg white lysozyme: an X-ray analysis at 1.8-A resolution. Biochemistry 1993, 32(37):9851–9858. 10.1021/bi00088a041View ArticlePubMedGoogle Scholar
- Strych U, Davlieva M, Longtin JP, Murphy EL, Im H, Benedik MJ, Krause KL: Purification and preliminary crystallization of alanine racemase from Streptococcus pneumoniae . BMC Microbiol 2007, 7: 40. 10.1186/1471-2180-7-40PubMed CentralView ArticlePubMedGoogle Scholar
- Jancarik J, Kim S-H: Sparse matrix sampling: a screening method for crystallization of proteins. J Appl Crystallogr 1991, 24: 409–411. 10.1107/S0021889891004430View ArticleGoogle Scholar
- Leslie AGW: Recent changes to the MOSFLM package for processing film and image plate data. Joint CCP4 + ESF-EAMCB Newsletter on Protein Crystallography 1992, 26: 27–33.Google Scholar
- Evans PR: SCALA, version 3.1.9. Medical Research Council Laboratory of Molecular Biology, Cambridge, UK 1993.Google Scholar
- CCP4 suite: The CCP4 suite: programs for protein crystallography. Acta Crystallogr D Biol Crystallogr 1994, 50(Pt 5):760–763.Google Scholar
- Vagin A, Teplyakov A: MOLREP: an automated program for molecular replacement. J Appl Cryst 1997, 30: 1022–1025. 10.1107/S0021889897006766View ArticleGoogle Scholar
- Matthews BW: Solvent Content of Protein Crystals. J Mol Biol 1968, 33: 491–497. 10.1016/0022-2836(68)90205-2View ArticlePubMedGoogle Scholar
- Emsley P, Cowtan K: Coot: model-building tools for molecular graphics. Acta Crystallogr D Biol Crystallogr 2004, 60(Pt 12 Pt 1):2126–2132. 10.1107/S0907444904019158View ArticlePubMedGoogle Scholar
- Murshudov GN, Vagin AA, Dodson EJ: Refinement of macromolecular structures by the maximum-likelihood method. Acta Crystallogr D Biol Crystallogr 1997, 53(Pt 3):240–255. 10.1107/S0907444996012255View ArticlePubMedGoogle Scholar
- Schomaker V, Trueblood KN: On the rigid-body motion of molecules in crystals. Acta Cryst, Section B 1968, 24(1):63–76. 10.1107/S0567740868001718View ArticleGoogle Scholar
- Krissinel E, Henrick K: Secondary-structure matching (SSM), a new tool for fast protein structure alignment in three dimensions. Acta Crystallogr D Biol Crystallogr 2004, 60(Pt 12 Pt 1):2256–2268. 10.1107/S0907444904026460View ArticlePubMedGoogle Scholar
- Krissinel E, Henrick K: Inference of macromolecular assemblies from crystalline state. J Mol Biol 2007, 372(3):774–797. 10.1016/j.jmb.2007.05.022View ArticlePubMedGoogle Scholar
- Vriend G: WHAT IF: a molecular modeling and drug design program. J Mol Graph 1990, 8(1):52–56. 29. 10.1016/0263-7855(90)80070-VView ArticlePubMedGoogle Scholar
- Rodriguez R, Chinea G, Lopez N, Pons T, Vriend G: Homology modeling, model and software evaluation: three related resources. CABIOS 1998, 14: 523–528.Google Scholar
- O'Sullivan O, Suhre K, Abergel C, Higgins DG, Notredame C: 3DCoffee: combining protein sequences and structures within multiple sequence alignments. J Mol Biol 2004, 340(2):385–395. 10.1016/j.jmb.2004.04.058View ArticlePubMedGoogle Scholar