|Home | About | Journals | Submit | Contact Us | Français|
AcsD, a type A siderophore synthetase with a molecular weight of 71 140 Da from Pectobacterium chrysanthemi, has been expressed, purified and crystallized at 293 K. The protein crystallized in the primitive orthorhombic space group P212121, with unit-cell parameters a = 80.3, b = 95.7, c = 161.1 Å, α = β = γ = 90°. Systematic absences were consistent with space group P212121. A complete data set has been collected to 2.25 Å resolution on BM14 at the ESRF. Consideration of the likely solvent content suggested that the asymmetric unit contained two molecules. Gel-filtration experiments indicated that the protein was a dimer, although self-rotation analyses did not detect a convincing twofold symmetry axis in the asymmetric unit. The protein has no convincing sequence match to any known structure and thus solution is likely to require experimental phasing.
Iron is an essential cofactor for both microorganisms and pathogens. In microorganisms it is required for ATP biosynthesis, whilst in pathogens it is often essential in establishing virulence. Unfortunately, iron is not freely available in the environment or higher eukaryotes as it is either insoluble (Fe3+) or already being utilized by other proteins (Bullen & Griffiths, 1999 ). As a consequence of this, iron assimilation is a significant challenge for both saprophytic microorganisms and invading pathogens. To overcome this problem, many pathogenic microorganisms synthesize and excrete iron-sequestering molecules called siderophores (Wandersman & Delepelaire, 2004 ). Siderophores scavenge ferric iron from the host and are readsorbed into the cell via a selective membrane-associated ATP transport system, where Fe3+ is then reduced to Fe2+ for release from the siderophore complex, storage and utilization (Koster, 2001 ).
Two biosynthetic routes exist for siderophores that yield structurally distinct molecules. Most polypeptide-based siderophores, for example pyoverdins, are biosynthesized via the well characterized nonribosomal peptide synthetase (NRPS) multienzyme family (Challis & Naismith, 2004 ), whereas many nonpolypeptide siderophores are biosynthesized via the much less well understood NRPS-independent siderophore (NIS) pathway (Challis, 2005 ). The Escherichia coli siderophore aerobactin was the first siderophore identified to be assembled by an NIS pathway (De Lorenzo et al., 1986 ). Over the last 10 y, several more NIS pathways for biosynthesis of a structurally diverse group of siderophores have been discovered. A superfamily of enzymes termed the NIS synthetases is common to all these pathways. NIS synthetases have been subdivided into A, B and C subfamilies according to their predicted substrate specificities. The NIS synthetase superfamily encompasses over 80 proteins from more than 40 bacterial species (Challis, 2005 ). Very recently, the first biochemical studies of AsbA, a type A NIS synthetase, and DesD, a type C NIS synthetase, have been reported (Oves-Costales et al., 2007 ; Kadi et al., 2007 ), confirming the predicted substrate specificities of these enzymes (Challis, 2005 ).
Pectobacterium chrysanthemi is a Gram-negative plant pathogen that is known to cause soft rot in various hosts. Two siderophores, chrysobactin and achromobactin, are required for full virulence of this pathogen. While chrysobactin is assembled by an NRPS-dependent pathway, achromobactin biosynthesis is NRPS-independent and involves type A, B and C NIS synthetases (Franza et al., 2005 ). The gene cluster that directs achromobactin production comprises eight genes, five of which (acsF, acsE, acsD, acsC and acsA) are proposed to be required for biosynthesis. Its expression is regulated by iron availability via direct interaction of iron with the ferric uptake regulatory protein Fur (Franza et al., 2005 ).
AcsD (MW 71 140 Da, pI 6.7) has been classified as a type A siderophore synthetase on the basis of sequence comparisons (Challis, 2005 ). According to the model for NIS synthetase substrate specificity, it is likely that AcsD catalyses the condensation of one of the prochiral carboxyl groups of citric acid with either the hydroxyl group of ethanolamine, the hydroxyl group of d- or l-serine or the N4-amino group of l-2,4-diaminobutyric acid (Fig. 1 ; Challis, 2005 ). However, the order of the steps in achromobactin biosynthesis remains to be clarified and it is possible that AcsA catalyzes the condensation of the δ-carboxyl group of α-ketoglutarate with the amino group of ethanolamine or the 2-amino group of 2,4-diaminobutyric acid prior to AcsD-catalysed condensation of the product of one of these reactions with citric acid.
The achromobactin-biosynthetic pathway offers potential targets for the design of new drugs against P. chrysanthemi and other pathogens containing related siderophore-biosynthetic pathways. As yet, there are no known three-dimensional structures of NIS synthetases. As part of our program on deciphering NIS siderophore biosynthetic pathways, we have cloned, expressed, purified and crystallized AcsD.
The gene encoding AcsD was amplified from cosmid pL9G1 (Franza et al., 2005 ) and cloned into the pET151/D-TOPO vector (Invitrogen). The expressed construct contains an N-terminal tobacco etch virus protease-cleavable site between the protein and a hexahistidine tag. The resulting plasmid was used to transform E. coli BL21 (DE3) (Novagen). 10 l of cells was grown in tryptone phosphate broth with ampicillin (final concentration 100 µg ml−1) at 310 K. When the cells reached an A 600 of 0.6, 0.4 mM isopropyl β-d-1-thiogalactopyranoside (IPTG) was added and the temperature was lowered to 288 K. High temperatures led to insoluble overexpression. After a further 18 h incubation period, cells were harvested by centrifugation at 2500g and 277 K for 30 min and resuspended in phosphate-buffered saline (PBS). The mixture was centrifuged as before and cell pellets were stored at 193 K.
Cell pellets were resuspended in lysis buffer (50 mM sodium phosphate pH 7.5, 500 mM NaCl, 30 mM imidazole, 10% glycerol, 1 mg ml−1 lysozyme) with protease inhibitors and lysed on ice using a Constant Systems cell disruptor at 207 MPa. The crude lysate was clarified by centrifugation (15 000g, 30 min, 277 K) and passed through a 0.22 µm filter. The cell-free extract was loaded onto a Nickel Sepharose 6 Fast Flow column (GE Healthcare) and the column was washed extensively with lysis buffer containing 50 mM imidazole before elution of AcsD in lysis buffer containing 500 mM imidazole. We found that AcsD precipitates if left in imidazole-containing buffer for any significant time. Fractions containing AcsD were immediately passed through a desalting (HiPrep 26/10) column (GE Healthcare), eluting with 50 mM Tris, 500 mM NaCl, 10% glycerol pH 7.5. The His tag was removed by incubation with TEV protease (mass ratio 1:50) overnight for 15 h at 293 K. The protease and uncleaved protein were removed by passage through a column in 50 mM Tris, 500 mM NaCl, 10% glycerol pH 7.5 containing Nickel Sepharose 6 Fast Flow medium. AcsD was further purified by gel filtration on an S-200 column (GE Healthcare) in 10 mM Tris, 150 mM NaCl, 10% glycerol pH 7.5. The purified protein (Fig. 2 ), which contains an additional N-terminal alanine residue when compared with the native protein, was characterized by SDS–PAGE and mass spectrometry before being concentrated to 5 mg ml−1 in gel-filtration buffer for crystallization.
Initial crystallization conditions for AcsD were identified using screening by sitting-drop vapour diffusion with a variety of commercial crystallization kits (The Classics and JCSG crystallization screens from Qiagen) and one custom PEG-based screen. Preliminary trials with these precipitants screened three different final protein concentrations. These were 0.1 µl AcsD (2.5 mg ml−1) mixed with 0.1 µl precipitant, 0.2 µl AcsD (2.5 mg ml−1) mixed with 0.1 µl precipitant and 0.1 µl AcsD (5 mg ml−1) mixed with 0.1 µl precipitant. Drops were set up at 293 K and were prepared using a nanodrop crystallization robot (Cartesian HoneyBee) as a part of a Hamilton–Thermo Rhombix system in Greiner Bio-One sitting-drop square-well plates with 100 µl of precipitant per reservoir. Small crystals (0.1 × 0.05 × 0.05 mm) were observed in various conditions throughout the screens, but the best crystals, as judged morphologically, were obtained using 0.1 M Tris–HCl pH 8.5 and 1.0 M sodium tartrate. Systematic optimization of this condition was performed in conventional sitting-drop plates by mixing 1 µl AcsD (2.6 mg ml−1) and 1 µl precipitant and equilibrating over 500 µl precipitant at 293 K. This resulted in slightly larger crystals (0.2 × 0.1 × 0.1 mm) that grew to full size in one week (Fig. 2 ) from 0.1 M Tris–HCl pH 8.5 and 1.5 M sodium tartrate.
The crystals of AcsD were soaked for 5–10 s in a cryoprotectant solution which consisted of 1.5 M sodium tartrate, 0.1 M Tris–HCl pH 8.5 and 20% glycerol. The soaked crystal was mounted in a cryo-loop (Hampton) and placed into a stream of nitrogen at 100 K. Diffraction data were collected on a MAR 225 CCD on BM14 at the European Synchrotron Radiation Facility (ESRF), Grenoble. The incident X-ray beam had a wavelength of 0.976 Å. 200 frames were recorded with a crystal-to-detector distance of 220 mm as non-overlapping 0.5° oscillations with 60 s exposure per image. A typical image is shown in Fig. 3 . The data were integrated and reduced with the program XDS (Kabsch, 1993 ). Initial indexing showed the crystals to be primitive orthorhombic, with unit-cell parameters a = 80.3, b = 95.7, c = 161.1 Å, α = 90, β = 90, γ = 90°. A study of the systematic absences revealed absences consistent with space group P212121 and this was supported by the CCP4 program POINTLESS (Collaborative Computational Project, Number 4, 1994 ). Data-collection statistics are given in Table 1 .
Calculation of the Matthews coefficient (V M = 2.17 Å3 Da−1) suggested the presence of two molecules of AcsD in the asymmetric unit and a solvent content of 43.3%. A monomer or trimer in the asymmetric unit gave Matthews coefficients of 4.3 and 1.45 Å3 Da−1 with solvent contents of 72% and 15%, respectively. When compared with protein standards of known molecular weight, the elution profile of AcsD from S-200 gel-filtration columns is consistent with the weight of a dimer (~150 kDa; data not shown). This evidence would suggest that there are two molecules in the asymmetric unit. Self-rotation analysis does not identify a threefold rotation axis, but there is no strong peak on the κ = 180° section to suggest the presence of a twofold axis. There is no significant sequence match between AcsD and any structure deposited in the Protein Data Bank (PDB). We will therefore have to rely on experimental phasing strategies to solve the structure. We have expressed and crystallized selenomethionine-derivatized AcsD using the procedures outlined above, modified only to prevent methionine biosynthesis. Although the selenomethionine-derivatized AcsD crystals are small, we expect to be able to improve their quality and then utilize the anomalous signal from Se to solve the structure by multi-wavelength anomalous dispersion methods shortly.
The structure of AcsD will provide an insight into the molecular mechanism of citrate recognition and whether the enzyme discriminates between the two prochiral carboxymethyl groups of citrate. Additionally, a structure will facilitate the design of inhibitors of AcsD and homologues, e.g. AsbA, which catalyses the ATP-dependent condensation of citric acid with spermidine as the first committed step in the biosynthesis of the anthrax ‘stealth siderophore’ petrobactin (Oves-Costales et al., 2007 ). Such inhibitors may form the basis for the development of new antibiotics. The structure of AcsD may also aid in the elucidation of the reaction it catalyses in achromobactin biosynthesis. Based on the available data, AcsD could catalyse the ATP-dependent condensation of one of the prochiral carboxyl groups of citric acid with one of several different diamine, amino alcohol, amido amine or amido alcohol substrates (Fig. 1 ).
NK and GLC thank Professor Dominique Expert for the gift of cosmid pL9G1. Mark Dorward and Michal Zawadzki are thanked for valuable technical support. The protein was targeted as part of the Scottish Structural Proteomics Facility (SSPF), which is funded by the Scottish Higher Education Funding Council (SHEFC) and the Biotechnology and Biological Sciences Research Council (BBSRC), UK.