Browse Subject Areas

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

The Functional Unit of Neisseria meningitidis 3-Deoxy-ᴅ-Arabino-Heptulosonate 7-Phosphate Synthase Is Dimeric

  • Penelope J. Cross ,

    Contributed equally to this work with: Penelope J. Cross, Logan C. Heyes

    Affiliation Biomolecular Interaction Centre and Department of Chemistry, University of Canterbury, Christchurch, New Zealand

  • Logan C. Heyes ,

    Contributed equally to this work with: Penelope J. Cross, Logan C. Heyes

    Affiliation Biomolecular Interaction Centre and Department of Chemistry, University of Canterbury, Christchurch, New Zealand

  • Shiwen Zhang,

    Affiliation Biomolecular Interaction Centre and Department of Chemistry, University of Canterbury, Christchurch, New Zealand

  • Ali Reza Nazmi,

    Affiliation Biomolecular Interaction Centre and Department of Chemistry, University of Canterbury, Christchurch, New Zealand

  • Emily J. Parker

    Affiliation Biomolecular Interaction Centre and Department of Chemistry, University of Canterbury, Christchurch, New Zealand

The Functional Unit of Neisseria meningitidis 3-Deoxy-ᴅ-Arabino-Heptulosonate 7-Phosphate Synthase Is Dimeric

  • Penelope J. Cross, 
  • Logan C. Heyes, 
  • Shiwen Zhang, 
  • Ali Reza Nazmi, 
  • Emily J. Parker


Neisseria meningitidis 3-deoxy-d-arabino-heptulosonate 7-phosphate synthase (NmeDAH7PS) adopts a homotetrameric structure consisting of an extensive and a less extensive interface. Perturbation of the less extensive interface through a single mutation of a salt bridge (Arg126-Glu27) formed at the tetramer interface of all chains resulted in a dimeric DAH7PS in solution, as determined by small angle X-ray scattering, analytical ultracentrifugation and analytical size-exclusion chromatography. The dimeric NmeDAH7PSR126S variant was shown to be catalytically active in the aldol-like condensation reaction between d-erythrose 4-phosphate and phosphoenolpyruvate, and allosterically inhibited by l-phenylalanine to the same extent as the wild-type enzyme. The dimeric NmeDAH7PSR126S variant exhibited a slight reduction in thermal stability by differential scanning calorimetry experiments and a slow loss of activity over time compared to the wild-type enzyme. Although NmeDAH7PSR126S crystallised as a tetramer, like the wild-type enzyme, structural asymmetry at the less extensive interface was observed consistent with its destabilisation. The tetrameric association enabled by this Arg126-Glu27 salt-bridge appears to contribute solely to the stability of the protein, ultimately revealing that the functional unit of NmeDAH7PS is dimeric.


The shikimate pathway is responsible for the biosynthesis of important aromatic compounds including the aromatic amino acids phenylalanine (Phe), tyrosine (Tyr) and tryptophan (Trp) [1]. The presence of this pathway in plants, microorganisms and apicomplexan parasites and its absence in mammals has drawn attention to the enzymes of this pathway as antimicrobial drug design targets [2]. 3-Deoxy-d-arabino-heptulosonate 7-phosphate synthase (DAH7PS) catalyses the first committed step of the shikimate pathway, which involves the aldol-like condensation of d-erythrose 4-phosphate (E4P) and phosphoenolpyruvate (PEP) to form 3-deoxy-d-arabino-heptulosonate 7-phosphate (DAH7P). Entry into the pathway is controlled by the allosteric regulation of DAH7PS by the pathway end products Phe, Tyr and Trp, or other intermediates of the pathway [3].

Despite significant sequence diversity, all known DAH7PS enzymes share a common (β/α)8-barrel core and similar active site architectures formed by conserved residues that support a similar catalytic mechanism [4]. On the basis of sequence and size, DAH7PS enzymes are classified into groups (type Iα, type Iβ and type II). Each group features distinct structural elements appended to the core catalytic barrel that are associated with the allosteric inhibition of the enzymes and different quaternary structure associations [3,510]. The type I groups, type Iα and type Iβ, share approximately 30% identity, and although a common dimeric unit is shared between by enzymes of these groups, they form different homotetrameric structures [11].

The DAH7PS isolated from N. meningitidis (NmeDAH7PS), the causative agent of pyogenic meningitis and meningococcal septicaemia, belongs to the type Iα group. This protein adopts a homotetrameric assembly, and contains similar structural features to the other characterised type Iα DAH7PSs from Saccharomyces cerevisiae and Escherichia coli. S. cerevisiae and E. coli both express multiple isozymes of DAH7PS and the catalytic activity of each of these isozymes is sensitive to the presence of a different aromatic amino acid, however, the genome of N. meningitidis, encodes a single DAH7PS (NmeDAH7PS) which is most sensitive towards feedback inhibition by Phe [12].

A comparison of the structure of the S. cerevisiae and E. coli enzymes with NmeDAH7PS reveals that although the single chain and dimeric units are superimposable, the two dimeric units that generate the homotetramer are oriented at different angles. The homotetrameric NmeDAH7PS comprises two interfaces, the more extensive (tight dimer) and the less extensive (tetramer interface) (Fig 1). The tight dimer interface involves interactions between 79 residues per chain and buries ~ 2800 Å2 or 18% of the surface area between chains. The tetramer interface is far weaker and consists of 17 residues per chain and buries only ~510 Å2 or 3.8% of the NmeDAH7PS chain. Both of these interfaces are comparable to the other structurally characterised type Iα DAH7PS enzymes [13,14]. In the EcoDAH7PS structure (PDB code: 1QR7) the angle between tight dimers is ~22° less than that of NmeDAH7PS (PDB code: 4HSN). Conversely, the arrangement of tight dimers in SceDAH7PS (PDB code: 1OAB) is ~26° greater than that of NmeDAH7PS. This illustrates the variability at the tetramer interface between members of the same DAH7PS subfamily.

Fig 1. A). Quaternary structure of NmeDAH7PS (PDB code: 4HSN) indicating the tight dimer interface and tetramer interface. The core catalytic (β/α)8 barrel is coloured blue and the N-terminal and internal extensions are coloured red and yellow, respectively. Manganese metal ion (purple), ligand (PEP) and Glu27-Arg126 salt bridge (green-sticks) A schematic of secondary structural features of monomeric NmeDAH7PS according to residue number, circles denote β-sheets and rectangles denote α-helices B).

Hydrogen bonding (red) and salt bridge (blue) interactions identified by analysis of each interface. Top. Tight-dimer. Bottom. Tetramer C) Hydrophobic interactions (green) at each interface of NmeDAH7PS. Top. Tight-dimer. Bottom. Tetramer. Chain B, C and D are denoted `, “and “`respectively.

Herein, we describe the disruption of the quaternary structure of NmeDAH7PS via mutagenesis of a single salt bridge at the tetramer interface, generating a dimeric species that we show to be the functional unit of NmeDAH7PS.

Experimental Procedures

Bacterial Strains, Plasmids, Media and Growth Conditions

Wild-Type NmeDAH7PS.

Purification procedures and kinetic measurements were performed as previously described [12].


The variant R126S (NmeDAH7PSR126S) was constructed using a Quik Change Lightning Mutagenesis Kit using the pT7-7 wild-type plasmid as template. Primers used to introduce the mutation were FOR: 5’-CATCAATTTCCGTTTGAGCCAGGAGAGGCAGCCTG-3’ and REV: 5’- CAGGCTGCCTCTCCTGGCTCAAACGGAAATTGATG -3’. The PCR product was transformed into Top10 cells. Plasmids of the expected size, as determined by agarose gel, were DNA sequenced to identify those containing the successful mutation and transformed into BL21 StarTM (DE3) cells. Expression and purification were carried out as previously described for the wild-type enzyme [12].

Interface Analysis

Resmap [15] and PISA [16] were used to analyse and visualise the interfacing residues. Free energy calculations for the formation of interfaces were calculated from PISA.

Differential Scanning Calorimetry

DSC was performed on a NanoDSC (TA instruments). Protein solutions were dialysed in SEC buffer (10 mM BTP, pH 7.3, 100mM KCl and 200μM PEP) and the dialysis buffer was used as reference. A final protein concentration of 26 μM (1 mg/mL) was used for the DSC experiments, and prior to scanning, all solutions were degassed by stirring under vacuum. An excess pressure of 3 atm was applied to the cells during scanning. Buffer scans were run prior to protein scans. DSC scans were performed using a heating rate of 1°C/min, and the data was analysed using the Nanoanalyze™ software supplied with the instrument.

Analytical Size Exclusion Chromatography

To determine the molecular mass of NmeDAH7P synthase in solution, protein standards of known molecular mass were applied to a Superdex S200 column (GE Healthcare). β-Amylase (200 kDa), Alcohol dehydrogenase (150 kDa), ovalbumin (75 kDa), conalbumin (66 kDa) and carbonic anhydrase (29 kDa) were used as molecular weight standards.

Analytical Ultracentrifugation

Sedimentation velocity experiments were performed in a Beckman Coulter XL-1 analytical ultracentrifuge equipped with UV/Vis scanning optics at 20°C. Reference buffer solution (10 mM BTP, pH 7.3, 100 mM KCl, 200 μM PEP) and sample solutions were loaded into 12-mm double-sector cells with quartz windows, and mounted in an An-50 Ti 8-hole rotor.

Protein samples at concentrations of 0.2, 0.4 and 1.2 mg/mL were centrifuged at 50,000 g and the absorbance data was collected at 277 nm or 276 nm for NmeDAH7PS and NmeDAH7PSR126S respectively. Radial scans collected between 5.8 cm and 7.2 cm with 0.003 cm increments were then analysed with the continuous size distribution model using SEDFIT [17]. The partial specific volume (v) of the protein samples (0.737734 mL/g), buffer density (1.005 g/mL) and buffer viscosity (0.01002 cp) were calculated using the program SEDNTERP [18].

Small Angle X-Ray Scattering

SAXS measurements.

Measurements were performed at the Australian Synchrotron SAXS/WAXS beamline equipped with a Pilatus detector (1M, 170 mm x 170 mm, effective pixel size, 172 x 172 μm). The wavelength of the x-rays was 1.0332 Å. The sample–detector distance was 1600 mm, which provided an q range of 0.01–0.565 Å−1 [where q is the magnitude of the scattering vector, which is related to the scattering angle (2θ) and the wavelength (λ) as follows: q = (4π/λ) sinθ].

Data were collected from a 2 mm glass capillary (Hampton Research) and the temperature was 25°C. Scattering data was collected from wild-type NmeDAH7PS and NmeDAH7PSR126S (initial concentration of 7 and 12 mg/mL, respectively, prior to SEC) in the buffer (10 mM BTP, pH 7.5, 100 mM KCl, 200 μM PEP). Scattering from glucose isomerase at an initial concentration of 10 mg/ml was collected to calibrate the mass of scattering particles. Data were collected in 2 second intervals. 2D intensity plots from the peak of the SEC run were radially averaged, normalised to sample transmission and background subtracted.

SAXS data analysis.

Scattered intensity (I) was plotted versus q. Extrapolation of the DAH7PS I(q) profiles to zero angle [I(0)] and comparison with that of glucose isomerase standards indicated a molecular mass consistent with the NmeDAH7PS tetramer or dimer. All samples were devoid of an increase in intensity at low q (increase is indicative of aggregation). Radius of gyration (Rg) did not vary significantly over the concentrations measured, and all Guinier plots were linear for q·Rg < 1.3. The data sets for structural analyses were recorded with 302 data points over the range 0.0110 ≤ q ≤ 0.4 Å−1. 1D profiles were background subtracted and Guinier fits were made using PRIMUS. Indirect Fourier transform was performed using GNOM [19] to yield the function P(r), which gives both the relative probabilities of distances between scattering centers and the maximum dimension of the scattering particle Dmax. Theoretical scattering curves were generated from atomic coordinates and compared with experimental scattering curves using CRYSOL [20].


The assay system for NmeDAH7PS was performed as previously described by monitoring the disappearance of PEP at 232 nm [12]. Standard assay buffer was 50 mM BTP (pH 6.8) containing MnSO4 (100 μM) and variable concentrations of substrates PEP and E4P. Assays were carried out at 25°C


The reaction mixtures for the determination of the Km of E4P consisted of PEP (126 μM), MnSO4 (100 μM) and E4P (9–140 μM), in 50 mM BTP, pH 6.8 buffer. The reaction mixtures for the determination of the Km of PEP consisted of E4P (232 μM) and PEP (12–315 μM) in assay buffer. The reaction was initiated by the addition of NmeDAH7PSR126S (2 μL, 1.3 mg/mL).


The reaction mixtures for the determination of the Km of E4P consisted of PEP (100 μM) and E4P (10–320 μM), in assay buffer. The reaction mixtures for the determination of the Km of PEP consisted of E4P (300 μM) and PEP (5–160 μM) in assay buffer. The reaction was initiated by the addition of NmeDAH7PSWT (2 μL, 0.6 mg/mL). Apparent Km and kcat values were determined by fitting the data to the Michaelis-Menten equation using Grafit (Erithicus).

Time Dependent Loss of Activity

Standard assays were performed which consisted of E4P (230 μM) and PEP (100 μM) in assay buffer. Assays were initiated by the addition of 2μl of ~1 mg/mL NmeDAH7PSWT or NmeDAH7PSR126S. Assays were performed in triplicate and averaged at timepoints (0 hr, 2 hr, 4 hr, 6 hr, 18 hr and 24 hr). Both NmeDAH7PSWT and NmeDAH7PSR126S were diluted to 1 mg/mL at time 0 hr and kept at ~4°C for the experiment duration. There was no loss of protein after 24 hr at 1 mg/mL calculated via absorbance at 280 nm using the calculated molar extinction coefficient of 31650 M-1 cm-1 post experiment [21].


Standard assays were performed which consisted of E4P (74 μM), PEP (162 μM) and the amino acid; Phe, Tyr or Trp (25 μM– 1 mM) in assay buffer. Assays were were initiated by the addition of NmeDAH7PSR126S. Assays were performed in triplicate.

Isothermal Titration Calorimetry

Binding of NmeDAH7PSWT and NmeDAH7PSR126S to Phe at pH 7 was measured by ITC using a VP- ITC unit operating at 298 K (Microcal, GE Health- care). Before use, the protein was buffer exchanged against binding buffer [0.5 mM MnSO4 in 50 mM BTP buffer (pH 7)] and all solutions were degassed in a vacuum. Protein concentration was measured by UV absorption immediately before titrations were started. The titrations were comprised of 28 injections, one 2 μL injection followed by 27 x 10 μL injections of Phe. The initial datum point was routinely deleted to allow for diffusion of ligand across the needle tip during the equilibration period. A heat of dilution experiment was measured independently and subtracted from the integrated data before curve fitting in Origin 7.0. For the binding of NmeDAH7PS to Phe, 15 μM of NmeDAH7PSWT and NmeDAH7PSR126S were used and the syringe contained 1.2 mM Phe; and the data were fitted with the two-site sequential-binding model supplied by Micro-Cal.

Crystallisation of NmeDAH7PSR126S

NmeDAH7PSR126S was expressed and purified as previously described for the wild-type enzyme [12]. The resulting protein solution was concentrated to approximately 10 mg/ mL. 1 μL of enzyme solution (9–11 mg/mL) was mixed with 1 μL of crystallisation buffer containing 0.1 M Tris HCl (pH 7.3), 0.2 M trimethyl-amino-N-oxide (TMAO), 600 μM MnSO4 and 15–20% (w/v) PEG 2000MME. Crystals were grown by hanging drop vapour diffusion over 500 μL of crystallisation buffer and the crystallisation trays were incubated at 20°C. Crystals began to form in 48 hours and were fully formed within seven days. Crystals were flash frozen using liquid nitrogen in a cryoprotectant solution containing reservoir solution and 20% (v/v) PEG400.

Crystallography and Structure Determination

An X-ray diffraction dataset was collected at the Australian Synchrotron using the MX1 beamline [22]. The datasets were integrated and processed using XDS and Aimless [23,24]. Appropriate cut-off resolution was determined via CC1/2≥0.5 ensuring the data was complete in the highest resolution shell [25]. Space group and unit cell parameters for NmeDAH7PSR126S were the same as those previously identified for the wild-type enzyme (PDB code 4HSN [12]) indicating that initial phases could be obtained via molecular replacement using the original structure as a search model in Phaser MR [26]. All ligands and waters were removed from the search model (PDB code 4HSN) before molecular replacement was carried out. Refmac5 was used to generate the electron density map and this was manually analysed and refined in COOT (Table 1) [2729]. The quality of the model was optimised by consecutive model building in COOT and refinement with Refmac5. Water molecules were added manually via interpretation of the │2Fo-Fc│map ensuring that they had the ability to hydrogen bond to at least one acceptor or donor. In all structures, no electron density for the 14 amino acids at the N-terminus of each NmeDAH7PSR126S chain could be found, which is attributed to these residues being part of a highly flexible allosteric region of the protein. Molprobity was used to assess structure quality during refinement cycles and before deposition [30].

Table 1. Data collection and refinement statistics of NmeDAH7PSR126S variant (PDB code: 4UCG).


Interface Analysis of Interface and Choice of Mutation Site

The minor interface through which the tetramer is formed comprises primarily hydrophilic interactions, in contrast to the extensive hydrophobic interactions that contribute to the tight dimer interface (Fig 1). Of note is the single salt bridge between Arg126 and Glu27 from opposing tight dimers, which may play a role in determining the twist between tight dimers (Fig 1). E. coli DAH7PS shares a similar tetramer interface to that of NmeDAH7PS, and contains the same Glu-Arg salt-bridge. This salt bridge is not observed for S. cerevisiae DAH7PS, whose tetramer interface consists solely of hydrophobic interactions between four residues from each chain. This salt bridge was disrupted by substitution of Arg126 to a serine residue and this change was chosen to maintain the hydrophilicity of the protein surface.

The variant protein NmeDAH7PSR126S was created using site directed mutagenesis and purified following the protocol developed for the wild-type protein (NmeDAH7PSWT). NmeDAH7PSR126S had an expected mass of 77,300 Da, and showed similar thermal stability properties to the wild-type protein, with a melting temperature measured by differential scanning calorimetery of 47°C (NmeDAH7PSWT displayed a melting temperature of 49°C) (S1 Fig).

NmeDAH7PSR126S Is Dimeric in Solution

Analysis of the quaternary structure assembly of the NmeDAH7PSWT and NmeDAH7PSR126S enzymes was carried out using both analytical size-exclusion chromatography (SEC) and analytical ultracentrifugation (AUC). SEC analysis revealed a clear difference in the solution behaviour of the two enzymes (Fig 2A). The wild-type protein gave a clear symmetrical peak shape corresponding to a mass of 150 kDa. Based on the theoretical molecular weight of a single wild type enzyme chain (38.7 kDa), this observation is consistent with a homotetrameric protein. Conversely, for NmeDAH7PSR126S variant no tetrameric species was observed. An asymmetrical peak corresponding to a mass of ~66 kDa is consistent with the presence of a dimer under the conditions used for analysis. The same profile for each enzyme was observed when the samples were prepared for analysis and left at room temperature for 18 hours, indicating that the two enzymes observed form stable quaternary forms.

Fig 2. (A) Analytical SEC traces of NmeDAH7PSWT and NmeDAH7PSR126S at two different concentrations. NmeDAH7PSWT 1 mg/mL (red) and 0.2 mg/mL (orange). NmeDAH7PSR126S 1 mg/mL (blue) and 0.2 mg/mL (light blue). B. Calculated sedimentation coefficient c(s) distribution plot of wild-type NmeDAH7PS (top) and NmeDAH7PSR126S (bottom). For both enzymes, the distribution of species for three different concentrations are shown: 0.2 mg/mL (red), 0.4 mg/mL (green) and 1.2 mg/mL (blue). The sedimentation velocity data, fit data and residuals from the fits are shown in S2 Fig.

Sedimentation velocity experiments were carried out on NmeDAH7PSWT and NmeDAH7PSR126S enzymes. The absorbance versus radial position data for the NmeDAH7PSWT and NmeDAH7PSR126S variant are shown in Fig 2B. Fitting the data for NmeDAH7PSWT to a continuous size distribution model indicates the presence of a major species with a standardised sedimentation coefficient of 8.1 S (Fig 2B). The calculated mass of this species is 141 kDa, which is consistent with the expected mass of a homotetrameric species (theoretical mass of 154.6 kDa). For NmeDAH7PSR126S, the fit indicated the presence of a major species with a standardised sedimentation coefficient of 5.7 S (Fig 2B). This predominant species has a calculated mass of 69 kDa, consistent with the expected molecular weight of the homodimeric assembly (77.3 kDa). At all three concentrations measured (0.4, 0.8, 1.2 mg/mL), a minor species with a sedimentation coefficient of 3.7 S, corresponding to an expected molecular weight of 36 kDa was observed, which is consistent with the molecular weight of monomeric NmeDAH7PS (38.7 kDa).

Small-angle X-ray scattering (SAXS) was employed to confirm the quaternary structure of NmeDAH7PS R126S in solution. The scattering profile collected for the NmeDAH7PSR126S variant was compared to that collected for the wild-type enzyme (Fig 3, Table 2). The difference between the scattering intensities at low angles is indicative of a change in molecular weight and suggests these two enzymes adopt distinct oligomeric assemblies. The radius of gyration (Rg) calculated from the scattering pattern of NmeDAH7PSR126S variant using the indirect fourier transform function, (P(r)), was ~ 26 Å and the maximum diameter of the scattering particle, Dmax, was calculated as ~ 95 Å using GNOM [19]. These values are in close agreement with the tight-dimer generated from the NmeDAH7PSR126S crystal structure (Rg of ~ 25 Å and Dmax of ~ 86 Å). Similarly, CRYSOL [20] was used to directly compare the collected data to the hypothetical scattering from the NmeDAH7PSR126S tetramer and dimer generated from the crystal structure. These data clearly show the NmeDAH7PSR126S variant adopts a dimeric structure in solution comparable to that of the NmeDAH7PSR126S tight-dimer (Fig 3B).

Fig 3. SAXS scattering profiles collected for NmeDAH7PSWT and NmeDAH7PSR126S.

(A) Overlay of the scattering data of NmeDAH7PSWT (closed circles) and NmeDAH7PSR126S variant (open circles) and (B) CRYSOL fits of the NmeDAH7PSR126S variant data to the theoretical scattering data generated from the R126S tetramer (dotted line) and tight-dimer (dashed line). The discrepancies of the fit of the theoretical scattering of tetrameric and dimeric forms to the experimentally determined data (χ) were calculated as 22.80 and 0.67 respectively.

NmeDAH7PSR126S Is Active

NmeDAH7PSR126S was observed to be catalytically active with parameters that are comparable with those of the wild-type enzyme (Table 3). The apparent KmPEP of NmeDAH7PSR126S is 25 μM, which is similar to that observed for the wild-type enzyme (15 μM). The NmeDAH7PSR126S displays a lower KmE4P (13.4 μM) than the wild-type enzyme (37 μM). However, there is a noticeable decrease in specific activity for the dimeric variant. This implies that tetramerisation may have some influence on catalysis by NmeDAH7PS.

NmeDAH7PSR126S Shows Time-Dependent Loss of Activity

Comparison of stability of NmeDAH7PSWT and the NmeDAH7PSR126S was carried out by measuring the activity of the enzymes over a 24 hour period. NmeDAH7PSWT lost ~10% catalytic activity, whereas the dimeric variant was far less stable and was observed to retain less than 60% of its activity after 24 hours (Fig 4).

Fig 4. Catalytic activity of NmeDAH7PSR126S (blue) NmeDAH7PSWT (red) over 24 hour time period relative to maximum activity observed for each enzyme at the beginning of the experiment.

NmeDAH7PSR126S Binds and Is Inhibited by Phe

As for the wild-type enzyme, NmeDAH7PSR126S is inhibited to the largest extent by Phe and to a lesser extent by Tyr and Trp (Fig 5A). Isothermal titration calorimetry experiments were employed to illustrate that the affinity of the dimeric enzyme for Phe, was not altered upon mutation (Fig 5B). For NmeDAH7PSR126S cooperativity was observed in the binding isotherm. The data were fitted to a two-site sequential model giving KD values for NmeDAH7PSR216S of 1.2 ± 0.7 μM for the first binding event and 33 ± 3 μM for the second binding event. This indicates similar Phe binding characteristics to those observed for NmeDAH7PSWT [12].

Fig 5. A. Effect of the aromatic amino acids, Phe (black), Tyr (red) and Trp (blue) on the catalytic activity of NmeDAH7PSR126S (open circles) in comparison to data previously reported for the wild-type enzyme (closed circles) [12]. B. Isothermal titration calorimetry binding isotherm for NmeDAH7PSR126S fit to a two-site sequential binding model.

NmeDAH7PSR126S Crystallises as a Tetramer

The crystal structure of the NmeDAH7PSR126S variant was determined by X-ray diffraction. This structure was refined to a resolution of 2.0 Å and crystallised in an orthorhombic space group (P212121). Data collection and refinement statistics are shown in Table 1. This space group is different to NmeDAH7PSWT (PDB code 4HSN), which crystallised in the monoclinic (P1211) space group [12].

The asymmetric unit for NmeDAH7PSR126S structure (PDB code: 4UCG) contains four chains interacting to provide a homotetramer, resembling the arrangement of wild-type NmeDAH7PS.

Like the wild-type enzyme, the NmeDAH7PSR126S tetramer has two distinct interfaces between the chains. The tight dimer interface remains extensive, burying 14% of the surface area of the monomeric unit, but is smaller than NmeDAH7PSWT (18%). The tetramer interface, on which the R126S substitution is located, is far smaller, burying 3.8% of the surface of the monomeric unit in NmeDAH7PSWT and this is reduced to 3.4% for the NmeDAH7PSR126S. Superposition of the NmeDAH7PSR126S homotetramer on the NmeDAH7PSWT protein yields an RMSD of 0.61 Å (1328 Cα atoms), whereas superposition of NmeDAH7PSR126S dimeric units on those of the wild-type protein gives RMSD values between 0.18–0.25 Å. These values are consistent with the small decrease in rotation of ~2°of the dimeric units with respect to each other at the tetramer interface, and this change may also account for the difference in space group between the NmeDAH7PSR126S and NmeDAH7PSWT structures.

According to PISA server calculations [16], the NmeDAH7PSR126S crystal structure displays a theoretical solvation free energy gain (ΔG) for the formation of the tight dimer interface of -16.1 and -16.5 kcal/mol, correlating to the chain A-chain B and chain B-chain D interfaces respectively [16]. This value is less that what is observed for the NmeDAH7PSWT enzyme, where the ΔG values are -20.1 kcal/mol and -18.1 kcal/mol. The lower values are indicative of a weakening of the tight dimer interface in NmeDAH7PSR126S.

The ΔG for formation of the tetramer interface in NmeDAH7PSR126S are -6.5 and -9.1 kcal/mol for the chain A-chain C and chain B-chain D interfaces respectively (interface interactions are shown in S3 Fig). Comparison of this with the ΔG for formation of the tetramer interface in NmeDAH7PSWT of -6.9 and -7.0 kcal.mol, provides interesting insight into the structural asymmetry at the tetramer interface in NmeDAH7PSR126S. The structural asymmetry observed in the crystal structure of NmeDAH7PSR126S may contribute to the in-solution dimerisation of this variant. This directly implicates the Arg126-Glu27 salt bridge in the formation of a symmetrical tetramer interface in NmeDAH7PSWT. Comparison of the surrounding residues in the crystal structure of NmeDAH7PSR126S with that of NmeDAH7PSWT illustrates this structural asymmetry. Tyr26, adjacent to Glu27, adopts a new, or alternate conformation in three of the four chains and in this new conformation, it forms a hydrogen bond with the backbone amine of Glu17. This position was previously occupied by a water molecule in the NmeDAH7PSWT structure (Fig 6). Furthermore, His219, located at the centre of the tetramer, adopts several alternative conformations in the NmeDAH7PSR126S structure as a result of this asymmetry.

Fig 6. Tetramer interface of NmeDAH7PSWT (green) (PDB code:4HSN) and NmeDAH7PSR126S (cyan) indicating the variable positioning of Tyr26 in place of the conserved water molecule from the NmeDAH7PSWT.

The loss of salt bridge between Glu27 and Arg126 for the NmeDAH7PSR126S variant is also illustrated.


A key feature of proteins is their tendency to form multimeric structures, usually homomers with even numbers of subunits all derived from the same gene, generating symmetry related interfaces, resulting in a functional biological assembly [31]. The DAH7PS family of proteins adopts a variety of homotetrameric forms, and like many other homotetramers, these tetramers are dihedral in nature and comprise dimers of dimers. Whereas the type Iα proteins share a common dimer subunit with the type Iβ group, they have a unique, small tetramer interface.

A single residue exchange in the small tetramer interface resulted in a solution- state dimer. The presence of this dimer and the absence of tetramer were confirmed by three independent solution-state techniques. The SAXS measurements also confirm that, as expected, the dimeric form adopts the assembly of the dimeric assembly of the parent protein. In contrast to these-solution state measurements, the protein crystallises as a tetramer. Some minor disruption and asymmetry has been introduced by the substitution indicating that this interface may be less stable. It should be noted that the dimeric form was observed by SAXS measurements even at relatively high concentrations, suggesting the crystal packing and crystallisation conditions favour the tetrameric form. Similar observations have been made with other related systems. The equivalent salt bridge of the E. coli Phe sensitive enzyme was disrupted by the mutation of Glu24 to Gln. This also resulted in a tetrameric crystalline form, however the enzyme was reported to be dimeric in solution, although no functional studies of this protein were provided [32]. Likewise, the tetrameric type Iβ enzyme from Pyrococcus furiosus has also been disrupted by mutation to form a mixture of dimeric and tetrameric forms in solution [11]. In this case the protein also crystallised as the tetramer.

The dimeric NmeDAH7PS retained its catalytic and allosteric properties. Catalysis was observed and both Phe binding and response was measured, revealing that the primary functional unit can be considered to be dimeric. The tetrameric form appears to provide advantage by a boost in catalytic rate and increased protein stability; the NmeDAH7PSR126S enzyme lost significantly more catalytic activity over a long period than the wild-type tetrameric protein, and this instability may contribute to the reduction in kcat. Once again parallels can be drawn here from the studies with the P. furiosus DAH7PS. The thermal stability of this protein, derived from a hyperthermophilic source, was significantly reduced by disruption of the tetramer [11].

These studies may help shed light on the common evolution and evolutionary divergence of the DAH7PS proteins into the different classes. There is some evidence that dihedral homotetramers become established through the interaction of symmetrical dimers for which the largest interface is included. It has been shown that the symmetrical nature of the tetramer interface in NmeDAH7PS may be a contributing factor to tetramerisation. It is becoming increasingly clear is that both type Iα and type Iβ DAH7PS proteins are functional through a similar dimeric structure. Type Iα proteins differ in the extent of this interface, as the allosteric machinery including an N-terminal extension and a β-hairpin contributes to the interactions that form the extensive dimer interface and provide the binding site for Phe. Furthermore, it is notable that the binding of Phe to both the tetrameric and dimeric forms of NmeDAH7PS is inhibitory and follows a two-site sequential model suggesting that both the allosteric and catalytic function can be provided by the dimeric unit.

This investigation clearly illustrates the importance of a single salt-bridge at tetramer interface of NmeDAH7PS and has implications for the evolution of quaternary structure in this family of enzymes.

Supporting Information

S1 Fig. Differential scanning calorimetry showing thermal stability of NmeDAH7PSWT and NmeDAH7PSR126S.



S2 Fig. Sedimentation velocity data for NmeDAH7PSWT and NmeDAH7PSR126S.



S3 Fig. Interface analysis of NmeDAH7PSWT and NmeDAH7PSR126S.




Part of this research was undertaken on the MX and SAXS/WAXS beamlines at the Australian Synchrotron, Victoria, Australia with access administered by the New Zealand Synchrotron Group. L.C.H is grateful for a BIC administered Marsden funded PhD Doctoral scholarship. This research was funded, in part, by the New Zealand Marsden Fund (UOC1105).

Author Contributions

Conceived and designed the experiments: EJP PJC LCH. Performed the experiments: PJC LCH SZ ARN. Analyzed the data: PJC LCH SZ ARN EJP. Contributed reagents/materials/analysis tools: PJC LCH SZ ARN. Wrote the paper: EJP LCH PJC ARN.


  1. 1. Bentley R. The Shikimate Pathway—A Metabolic Tree with Many Branches. Crit Rev Biochem Mol Biol. 1990;25: 307–384. pmid:2279393
  2. 2. Maeda H, Dudareva N. The shikimate pathway and aromatic amino Acid biosynthesis in plants. Annu Rev Plant Biol. 2012;63: 73–105. doi: 10.1146/annurev-arplant-042811-105439. pmid:22554242
  3. 3. Light SH, Anderson WF. The diversity of allosteric controls at the gateway to aromatic amino acid biosynthesis. Protein Sci. 2013;22: 395–404. doi: 10.1002/pro.2233. pmid:23400945
  4. 4. Webby C, Baker H, Lott S, Baker E, Parker E. The Structure of 3-Deoxy-D-arabino-heptulosonate 7-phosphate Synthase from Mycobacterium tuberculosis Reveals a Common Catalytic Scaffold and Ancestry for Type I and Type II Enzymes. J Mol Biol. 2005;354: 927–939. pmid:16288916
  5. 5. Wu J, Woodard RW. New insights into the evolutionary links relating to the 3-deoxy-D-arabino-heptulosonate 7-phosphate synthase subfamilies. J Biol Chem. 2006;281: 4042–4048. pmid:16339761
  6. 6. Webby CJ, Jiao W, Hutton RD, Blackmore NJ, Baker HM, Baker EN, et al. Synergistic allostery, a sophisticated regulatory network for the control of aromatic amino acid biosynthesis in Mycobacterium tuberculosis. J Biol Chem. 2010;285: 30567–30576. doi: 10.1074/jbc.M110.111856. pmid:20667835
  7. 7. Shumilin IA, Zhao C, Bauerle R, Kretsinger RH. Allosteric Inhibition of 3-Deoxy-D-arabino-heptulosonate 7-phosphate Synthase Alters the Coordination of Both Substrates. J Mol Biol. 2002;320: 1147–1156. pmid:12126632
  8. 8. Cross PJ, Dobson RCJ, Patchett ML, Parker EJ. Tyrosine latching of a regulatory gate affords allosteric control of aromatic amino acid biosynthesis. J Biol Chem. 2011;286: 10216–10224. doi: 10.1074/jbc.M110.209924. pmid:21282100
  9. 9. Blackmore NJ, Reichau S, Jiao W, Hutton RD, Baker EN, Jameson GB, et al. Three Sites and You Are Out: Ternary Synergistic Allostery Controls Aromatic Amino Acid Biosynthesis in Mycobacterium tuberculosis. J Mol Biol. 2013;425: 1582–1592. doi: 10.1016/j.jmb.2012.12.019. pmid:23274137
  10. 10. Cross PJ, Allison TM, Dobson RCJ, Jameson GB, Parker EJ. Engineering allosteric control to an unregulated enzyme by transfer of a regulatory domain. PNAS. 2013;110: 2111–2116. doi: 10.1073/pnas.1217923110. pmid:23345433
  11. 11. Nazmi AR, Schofield LR, Dobson RCJ, Jameson GB, Parker EJ. Destabilization of the homotetrameric assembly of 3-deoxy-D-arabino-heptulosonate-7-phosphate synthase from the hyperthermophile Pyrococcus furiosus enhances enzymatic activity. J Mol Biol. 2014;426: 656–673. doi: 10.1016/j.jmb.2013.11.008. pmid:24239948
  12. 12. Cross PJ, Pietersma AL, Allison TM, Wilson-Coutts SM, Cochrane FC, Parker EJ. Neisseria meningitidis expresses a single 3-deoxy-D-arabino-heptulosonate 7-phosphate synthase that is inhibited primarily by phenylalanine. Protein Sci. 2013;22: 1087–1099. doi: 10.1002/pro.2293. pmid:23754471
  13. 13. Shumilin I, Bauerle R, Kretsinger R. Crystal structure of phenylalanine-regulated 3-deoxy-D-arabino-heptulosonate 7-phosphate synthase from Escherichia coli. Structure. 1999;7: 865–875. pmid:10425687
  14. 14. Konig V, Braus G, Schneider T, Pfeil A. Substrate and metal complexes of 3-Deoxy-D-arabino-heptulosonate 7-phosphate synthase from Saccharomyces cerevisiae provide new insights into the catalytic mechanism. J Mol Biol. 2004;337: 675–690. pmid:15019786
  15. 15. Swint-Kruse L, Brown CS. Resmap: Automated representation of macromolecular interfaces as two-dimensional networks. Bioinformatics. 2005;21: 3327–3328. pmid:15914544
  16. 16. Krissinel E, Henrick K. Inference of Macromolecular Assemblies from Crystalline State. J Mol Biol. 2007;372: 774–797. pmid:17681537
  17. 17. Schuck P. Size-Distribution Analysis of Macromolecules by Sedimentation Velocity Ultracentrifugation and Lamm Equation Modeling. Biophys J. 2000;78: 1606–1619. pmid:10692345
  18. 18. Lebowitz J, Lewis MS, Schuck P. Modern analytical ultracentrifugation in protein science : A tutorial review. Protein Sci. 2002;11: 2067–2079. pmid:12192063
  19. 19. Svergun DI. Determination of the regularization parameter in indirect-transform methods using perceptual criteria. J Appl Crystallogr. 1992;25: 495–503.
  20. 20. Svergun DI, Barberato C, Koch MHJ. CRYSOL–a Program to Evaluate X-ray Solution Scattering of Biological Macromolecules from Atomic Coordinates. J Appl Crystallogr. 1995;28: 768–773.
  21. 21. Wilkins M, Gasteiger E, Bairoch A, Sanchez J, Williams K, Appel R, et al. Protein identification and analysis tools in th ExPASY server. Methods Mol Biol. 1999;112: 531–552. pmid:10027275
  22. 22. Mcphillips TM, McPhillips SE, Chiu H-J, Cohen AE, Ashley M, Ellis PJ, et al. Blu-Ice and the Distributed Control System: software for data acquisition and instrument control at macromolecular crystallography beamlines. J Synchrotron Radiat. 2002;9: 401–406. pmid:12409628
  23. 23. Evans P. Scaling and assessment of data quality. Acta Crystallogr. 2006;62: 72–82.
  24. 24. Evans P. An introduction to data reduction: space-group determination, scaling and intensity statistics. Acta Crystallogr. 2011;67: 282–292.
  25. 25. Karplus PA, Diedrichs K. Linking Crystallographic Model and Data Quality. Science. 2012;336: 1030–1033. doi: 10.1126/science.1218231. pmid:22628654
  26. 26. McCoy AJ, Grosse-Kuntslave RW, Adams PD, Winn MD, Storoni LC, Read RJ. Phaser Crystallographic software. J Appl Crystallogr. 2007;40: 658–674. pmid:19461840
  27. 27. Emsley P, Lohkamp B, Scott W, Cowtan K. Features and Development of Coot. Acta Crystallogr. 2010;66: 486–501.
  28. 28. Murshudov GN, Skubák P, Lebedev AA, Pannu NS, Steiner RA, Nicholls RA, et al. REFMAC5 for the refinement of macromolecular crystal structures. Acta Crystallogr. 2011;67: 355–367.
  29. 29. Murshudov GN, Vagin AA, Dodson ED. Refinement of Macromolecular Structures by the Maximum-Likelihood Method. Acta Crystallogr. 1997;53: 240–255.
  30. 30. Chen VB, Arrendall BW III, Headd JJ, Keedy DA, Immormino RM, Kapral GJ, et al. MolProbity: all-atom structure validation for macromolecular crystallography. Acta Crystallogr. 2010;66: 12–21.
  31. 31. Lynch M. Evolutionary diversification of the multimeric states of proteins. PNAS. 2013;110: 2821–2828. doi: 10.1073/pnas.1310980110.
  32. 32. Shumulin IA, Bauerle R, Kretsinger RH. The High-Resolution Structure of 3-Deoxy-D-arabino-heptulosonate-7-phosphate Synthase Reveals a Twist in the Plane of Bound Phosphoenolpyruvate. Biochemistry. 2003;42: 3766–3776. pmid:12667068