Crystal structure of chorismate synthase from <i>Aquifex aeolicus</i> reveals a novel beta alpha beta sandwich topology

Chorismate synthase (CS) (EC 4.6.1.4) catalyzes the last enzymatic step of the shikimate biosynthetic pathway involving the conversion of 5-enolpyruvylshikimate 3-phosphate (EPSP) to chorismate.1 Chorismate serves as a precursor for the biosynthesis of a large number of aromatic compounds, including the aromatic amino acids, and secondary metabolites in plants. This pathway is essential in microbes, fungi, parasites and plants, and is absent in mammals, which makes it an excellent target for the design of drug compounds and herbicides. The potential of such compounds has been exemplified by the herbicide glyphosate, which inhibits EPSP synthase, the enzyme preceding chorismate synthase in the pathway.2 The shikimate pathway has also recently been discovered in apicomplexan parasites such as Plasmodium falciparum, thus providing another avenue to improve the present treatment for devastating diseases such as malaria or toxoplasmosis.3 The CS-catalyzed reaction requires reduced FMN and substrate EPSP to produce chorismate, a phosphate molecule and reduced FMN. It is an unusual reaction in that the redox state of the functional cofactor (FMN) remains unchanged. Another unusual feature the of CS catalyzed reaction is the nonconcerted anti-1,4-elimination of the 3-phosphate and the C-(6-pro-R) hydrogen.4 It is expected that the three-dimensional structure of CS can significantly enhance our understanding of its enzymatic mechanism for this unusual chemistry. The gene encoding CS from Aquifex aeolicus (AA) (gi16081332) was cloned from genomic AA DNA and its gene product was expressed, selenomethionine (SeMet)-labelled, and purified from a bacterial system as described elsewhere.5, 6 Screening for crystallization conditions was also performed as described elsewhere.5 The final crystallization condition consists of 30% isopropanol, 0.2 M magnesium chloride and 0.1 M sodium HEPES at pH 7.5. The crystals chosen for X-ray data collection were flash-frozen in this solution containing 25% glycerol as cryoprotectant. Crystals of the AACS protein belong to the monoclinic space group C2 with the unit cell dimensions a = 153.3 Å, b = 95.8 Å, c = 113.4 Å, β = 93.1 °. A single wavelength anomalous dispersion (SAD) experiment with selenium as the anomalous scatterer was carried out at 100 K at beamline ID14B, APS, using a MAR CCD detector. The 2.05 Å dataset was collected at 100 K at beamline BM14C, APS using a Q4 CCD detector. Both, the SAD and remote data were processed and scaled using the HKL2000 software package.7 The Selenium sites were identified using SOLVE8 and density modification and initial model building was carried out with RESOLVE.9 Automated model building and refinement were carried out with RESOLVE and Refmac 5.0 of the CCP4 package.9, 10 With this approach, greater than 60% of the initial protein model was built. Further model building was done manually with the O software package11 and CNS12 was used for refinement of the model. All four molecules were built and refined using this approach without applying NCS constraints. Once both R values were below 30%, water molecules were identified with CNS and manually verified in O with the criteria: A peak of at least 2.5 σ in an Fo-Fc map and with expected interacting distances. The crystallographic and refinement data are shown in Table I. The crystal structure of CS from AA was solved by the SAD method using the anomalous signal from Se at its peak energy. The atomic models are of good quality as indicated by the R factors and small deviations from ideal bond length and bond angle parameters. The final model contains 331 × 4 amino acids (residues 2–49, 56–86, 118–325 and 351–398) and 559 water molecules. Refinement at 2.05 Å resulted in an R of 0.209 and an Rfree of 0.250. According to Procheck, 99.4 % of the residues are in the most favored and additionally allowed regions of the Ramachandran plot.13 Two residues (Arg84A and Leu116D) are in the disallowed regions. The dominant structural topology of CS is a beta-alpha-beta sandwich, in which each monomer of CS consists of a central helical core, helix order α1, α4, α9 and α7, sandwiched between two four-stranded antiparallel beta sheets, strand order β1, β2, β7 and β4 for sheet “one” and strand order β8, β10, β17 and β12 for sheet “two” (Fig. 1). The core helices are staggered in a cris-cross manner about 45° away from each other. The two sheets are further characterized by a golf club topology, in which the four-stranded sheet is characterized as the handle and the club is characterized by a short three-stranded antiparallel sheet which is a continuation of the individual strands separated by short loops (Fig. 1). This golf club topology is also seen for EPSP synthase, the enzyme preceding CS in the shikimate pathway.2 In addition the monomer of CS is composed of other structural features. A beta hairpin is formed by β13 and β14 and is observed to extend away from the core of the monomer. Two helices, α5 and α6, are observed to flank sheet “two” of the monomer. Packing within the monomer is stabilized mainly by the sandwich topology, which produces a tightly packed molecule. In addition the long C-terminal helix, α9, and the beta hairpin, β13 and β14, are stabilized by H-bonding with each other and neighboring residues within the protein. a: Ribbon diagram of the CS monomer. β strands and α helices are colored green and gold, respectively. β sheet “one” of the β-α-β sandwich is located on the left of the helical core and β sheet “two” is on the right. b: Topology diagram for CS. β strands (β 1–β17) and alpha helices (α1–α10) are shown as dark gray arrows and light gray cylinders, respectively. The diagram reveals an internal 2-fold symmetry involving two equivalent major 4-stranded anti-parallel β sheets and two minor 3-stranded β sheets surrounding a helical core. The symmetry is disrupted by an extension in β sheet “two” (between β8 and β10) and by a β hairpin (β13, β14). The hairpin interacts with α10. c: The tetramer of CS with each monomer colored separately. The tetramer is constructed from a dimer of dimers with AD and BC staggered by 90°. The putative binding site is located at the interface of dimers AD and BC. d: A molecular surface representation showing electrostatic potential (same orientation as part c). Acidic regions are colored red, basic regions blue and neutral regions grey. Two pockets of positive electrostatic potential (blue) on the front of the molecule represent the putative substrate binding site. CS was determined to be a tetramer in solution by gel filtration chromatography, which is consistent with the crystal structure obtained for AACS. The crystal structure of AACS, depicts a tetramer characterized as a dimer of dimers such that intermolecular interactions within A and D form the first dimer and B and C form the second dimer (Fig. 1). The dimer is characterized by a large antiparallel eight-stranded β sheet which is formed by combining sheet “two” of each subunit, strand order β8A, β10A, β17A, β12A, β12D, β17D, β10D and β8D. In each case, 6116 A2 out of a total of 25592 A2 molecular surface is buried upon dimerization (A with D and B with C). The interactions at the core of the dimer interface are mainly hydrophobic; ionic interactions are more prominent in the solvent-exposed regions. The tetramer is formed by two dimers interacting via their N-terminal faces and staggered by a 90° rotation, which results in sheet “one” of each subunit interacting on a perpendicular axis to produce a meshed pattern (Fig. 1). Two prominent sets of interactions are seen involving residues of sheet “one” from molecule A with sheet “one” of molecule C and residues of helix 9 from molecule A interacting with helix 9 from molecule C. The same interactions are seen between molecules B and D. The interactions with sheet “one” involve a combination of hydrophobic, polar and ionizable residues. The core of the tetramer interface is characterized by two ionic networks, the first involving R7 and E24 (from A) with D246 and R247(from B) and the second involving D379 (from A) with K375 (from B). The same interactions are seen between molecules C and D. Such ionic interactions at the hydrophobic core are known to have a stabilizing effect and thus may provide added stability to the oligomeric state of CS. Comparison of the crystal structure of CS with other proteins in the protein database using the software DALI identified seven proteins with Z scores from 3.2 to 2.0, with Bira functional protein (PDB ID 1BIA) being the closest structural match. Structure superposition of these proteins with that of CS revealed that the central 4-stranded β-sheet is the only conserved structural element. Further analysis indicated that the structure of CS is different from any known flavin binding proteins. Visual inspection of other proteins containing a βαβ-sandwich motif confirmed that the structural topology of CS is indeed unique. The active site of CS was identified by searching for a cavity with the fingerprint sequence signatures observed for CS, 1) R-P-[GS]-H-[AG]-D-x(5)-K and 2) R-x-S-[AG]-R-[EV]-[ST]-x(3)-V-x(2)-G-x(6)-L.14 A large pocket with approximate dimension 28 × 28 × 10 A3 containing residues from this sequence signature in addition to other conserved residues was identified. A number of these conserved residues (S14, S135, H15, R44, R137, D59, D85, K55, E57, N261, Y48, Q45, and N83) were identified in this pocket. The disordered loops (residues 50–56, 87–117 and 326–350) also map to this location of the protein. This binding site is shared with two other subunits; the binding site of molecule A is shared with molecules C and D. Using the fingerprint motif, we can propose that this identified pocket is the active site of CS. Additional structural information, complexes with substrate analogues and FMN analogs are required to further characterize this binding site and the role of active site residues in CS reaction.

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.