Structural and Physiological Analyses of the Alkanesulphonate-Binding Protein (SsuA) of the Citrus Pathogen Xanthomonas citri

Background The uptake of sulphur-containing compounds plays a pivotal role in the physiology of bacteria that live in aerobic soils where organosulfur compounds such as sulphonates and sulphate esters represent more than 95% of the available sulphur. Until now, no information has been available on the uptake of sulphonates by bacterial plant pathogens, particularly those of the Xanthomonas genus, which encompasses several pathogenic species. In the present study, we characterised the alkanesulphonate uptake system (Ssu) of Xanthomonas axonopodis pv. citri 306 strain (X. citri), the etiological agent of citrus canker. Methodology/Principal Findings A single operon-like gene cluster (ssuEDACB) that encodes both the sulphur uptake system and enzymes involved in desulphurisation was detected in the genomes of X. citri and of the closely related species. We characterised X. citri SsuA protein, a periplasmic alkanesulphonate-binding protein that, together with SsuC and SsuB, defines the alkanesulphonate uptake system. The crystal structure of SsuA bound to MOPS, MES and HEPES, which is herein described for the first time, provides evidence for the importance of a conserved dipole in sulphate group coordination, identifies specific amino acids interacting with the sulphate group and shows the presence of a rather large binding pocket that explains the rather wide range of molecules recognised by the protein. Isolation of an isogenic ssuA-knockout derivative of the X. citri 306 strain showed that disruption of alkanesulphonate uptake affects both xanthan gum production and generation of canker lesions in sweet orange leaves. Conclusions/Significance The present study unravels unique structural and functional features of the X. citri SsuA protein and provides the first experimental evidence that an ABC uptake system affects the virulence of this phytopathogen.


Introduction
All living organisms require sulphur for the biosynthesis of amino acids (cysteine and methionine) and cofactors such as glutathione, coenzyme A and coenzyme M [1]. Bacteria must either obtain these molecules directly from the environment or synthesise them using inorganic (e.g., sulphate) or organic (e.g., sulphonates) sulphur sources [2,3]. In aerobic soils, the sulphur content is almost entirely represented by sulphonates and sulphate esters of various organic compounds, with inorganic sulphur representing less than 1-5% of the available element [4]. Under such conditions, several bacterial species are known to express proteins required for the uptake of organic sulphur-containing substances such as sulphate esters, sulphamates, sulphonates and alkanesulphonates [1,5,6].
In Escherichia coli, the uptake of sulphur-containing organic compounds requires the expression of two ATP binding cassette (ABC) transport systems: the Tau system, which is specifically involved in the uptake of taurine, and the Ssu system, which mediates the uptake of other aliphatic sulphonates [5,7]. The Ssu system is encoded by a single operon encompassing five cistrons, ssuEADCB, that encode the three components of the uptake system, which are the periplasmic alkanesulphonate-binding protein (SsuA), the membrane permease protein (SsuC) and the nucleotide-binding protein (SsuB), and two cytoplasmic proteins involved in sulphur release, NAD(P)H flavin mononucleotide oxidoreductase (SsuD) and the FMNH2-dependent sulphonate monooxygenase (SsuE) [7]. Structural analysis of the Ssu uptake system has been restricted to the recent description of the substrate-free E. coli SsuA protein [8]. The SsuA protein is similar to other periplasmic ABC transporters in which two globular domains form a cleft in which the ligand binds and from which it is subsequently translocated to the membrane-bound compartment. However, no information is presently available about the structure of the SsuA protein-ligand complex, the interactions of specific amino acid residues with the alkanesulphonate substrates or the molecular features that allow various alkanesulphonate molecules to fit into the protein's ligand-binding pocket.
In addition to the Ssu system of E. coli, functional Ssu uptake systems have also been reported in Bacillus subtilis and Pseudomonas putida, further supporting the relevant physiological role of this nutrient uptake system in soil-inhabiting bacteria [6,9]. Despite the apparently relevant physiological role of the Ssu system, no information is available regarding the uptake of organic sulphurcontaining compounds by bacterial plant pathogens. Indeed, once absorbed by plants as inorganic sulphate, sulphur is rapidly converted into complex organic molecules such as proteins and sulphonates or to sulphoquinovose in sulpholipids of thylakoid membranes [7]. However, no information regarding the role of aliphatic sulphonate uptake in the growth and virulence of phytopathogens is presently available.
The Xanthomonas genus encompasses 27 different bacterial species and over 140 pathovars that interact with more than 400 plant species, including several economically relevant species [10,11]. Xanthomonas citri, the causative agent of citrus canker, is capable of infecting all citrus cultivars, although different citrus species may show distinct susceptibility to the disease, as illustrated by the high susceptibility of sweet orange (Citrus sinensis) and the lower susceptibility of mandarin species [12]. Canker is one of the most economically damaging diseases of citrus plants; it begins with the epiphytic colonisation of the leaf surface followed by the entrance of the pathogen into leaf tissue through stomata or wounds. Upon entrance into the plant mesophyll, the bacterium induces cell enlargement and multiplication (hyperplasia) followed by a water-soaking phenotype and the formation of blister-like lesions approximately 4 days after infection. The production of highly hygroscopic xanthan gum helps bacteria increase the water adsorption of the plant through the capillary effect from xylem, leading to disruption of the plant epidermis and formation of yellow spongy pustules that become brown and corky with time [13].
In the present study, we investigated for the first time the presence of alkanesulphonate uptake systems in the Xanthomonas genus, with particular emphasis on the system encoded by X. citri. A single operon-like gene cluster (ssuEDACB) encoding both the uptake system and the intracellular enzymes involved in sulphur release was detected in the genome of the X. citri 306 strain but not in other Xanthomonas species with the exception of the closely related X. fuscans. The structure of SsuA in complex with MOPS, MES and HEPES was determined, showing the importance of a conserved dipole at the substrate cleft for sulphate group coordination and the presence of a rather spacious ligand pocket, partially filled with water molecules, that permits the binding of alkanesulphonates of quite different molecular sizes, charges and shapes. We also showed that an active alkanesulphonate uptake system is required for the growth of X. citri under sulphate-restricted conditions. Finally, generation of an ssuA-knockout mutant of the X. citri 306 strain showed that a defective alkanesulphonate uptake system affects both xanthan gum production and the generation of canker lesions in a susceptible citrus host. Altogether, the present work represents the first structural and functional characterisation of an alkanesulphonate uptake system of a bacterial plant pathogen and demonstrates that this nutrient uptake system plays a role in the pathogenesis of X. citri.

The alkanesulphonate uptake system in Xanthomonas
A search of the available Xanthomonas genomes for genes involved in the uptake of alkanesulphonates revealed 4 species (X. citri, X. fuscans, X. gardneri, and X. campestris) in which a single operon-like gene cluster sharing similarity with E. coli ssu genes was found ( Figure 1). The X. citri ssu genes showed a similar genetic organisation to that of the ssu operon found in E. coli with the exception that ssuD and ssuA were present in inverted positions in the two species. The fact that the putative functional organisations of these genes have been maintained in two citrus pathogens indicates that the ssu genes have a recent evolutionary history in the genus. Further structural and functional analyses of the alkanesulphonate uptake genes concentrated on X. citri SsuA, the alkanesulphonate-binding protein. The deduced amino acid sequence of the X. citri SsuA protein shared 59% identity with the E. coli orthologue, and the X. citri SsuA gene was chosen for subsequent cloning, expression and purification of a recombinant form of the protein.
Interaction of recombinant X. citri SsuA with various alkanesulphonate substrates The recombinant X. citri SsuA protein was expressed as a soluble cytosolic protein genetically fused with a histidine tag and a thrombin cleavage site at the N-terminal end. Maximum soluble protein yields (,80 mg/L) were achieved after expression under optimum inducing conditions. The protein was purified by singlestep affinity chromatography and subsequently cleaved with thrombin to remove the vector-encoded histidine ( Figure 2). The purified recombinant protein remained soluble and stable at high concentrations (6-12 mg/mL) even after extended storage at -20uC. Thermal shift experiments showed increased stability of recombinant X. citri SsuA in the presence of MOPS, MES, CHES, or HEPES but not in the presence of taurine, sulphate, thiosulphate or hydroxylamine ( Figure 3A). These results indicate that the recombinant protein binds specifically to alkanesulphonate ligands. The Tm of SsuA varied from 39.5uC for the unbound form to 43.5uC (CHES), 45.8uC (HEPES), 47.1uC (MES) and 51uC (MOPS) for the putative ligand-bound forms. CD and fluorescence analyses showed that the recombinant SsuA was stable at pH values of 5, 7 and 9 ( Figure 3B and 3C) and that it undergoes small changes in secondary structure content upon ligand binding ( Figure 3D and 3E). The quenching of fluorescence observed after the addition of alkanesulphonates corroborated the CD results and suggested the possible presence of tryptophan residues close to the ligand-binding pocket ( Figure  3E). Taken together, these results show that recombinant SsuA undergoes conformational changes on binding to alkanesulphonates.
The structure of ligand-bound X. citri SsuA SsuA crystallised under a variety of conditions in which ammonium sulphate and alkanesulphonates (HEPES and MES) were present (Table S1). Initial crystallisation trials of the ligandbound form of recombinant SsuA resulted in needle-like crystals that grew after one week in the presence of ammonium sulphate, MES and PEG. Refinement of these conditions produced better crystals in the absence of PEG and the presence of sodium chloride. The best diffraction patterns (1.7 and 1.9 Å ) were obtained from crystals grown in 0.1 M NaCl, 1.6 M ammonium sulphate, and 0.1 M HEPES, pH 7.5 ( Figures 4A and 4B). These crystals showed symmetry and systematic absence of the orthorhombic space group P2 1 . Table 1 summarises the datacollection statistics. The Matthews coefficient was calculated to be 1.67, and a solvent content of 26%, corresponding to 1 molecule in the asymmetric unit, was calculated. The SsuA crystal structure revealed the characteristic folding of periplasmic-binding proteins, which consists of an a/b sandwich pattern organised into two domains (I and II) separated by a cleft at the binding site, with the ligand remaining hidden inside the pocket ( Figure 4C). The N-and C-termini are present in domain I, which is connected by a hinge (Pro 107 to Thr 109 and Gly 204 to Gly 211 ) to domain II. The tertiary structures of SsuA bound to MOPS and MES (PDB codes 3KSJ and 3KSX, respectively) were generated by molecular replacement using the structural coordinates of SsuA bound to HEPES (PDB code 3E4R). All crystal structures presented one molecule in the asymmetric unit. The refinement statistics, including statistics from the processing of the collected data and structure refinement, are presented in Table 1. The SsuA ligand-binding pocket and interactions with alkanesulphonates The crystal structure of SsuA bound to each of the three different alkanesulphonates tested showed that ligand binding is stabilised primarily through a range of polar interactions between the sulphonic acid oxygen atoms and the NH groups of mainchain peptide hydrogen bonds (Gly 68 , Gly 86 , Ser 141 ), one side chain NH group of Gln 36 and one hydroxyl group of Ser 141 ( Figure 5A and Table S2). Water molecules are responsible for the stability of the alkane chains. Residues in the pocket that interacts with the alkanesulphonates are highly conserved among orthologues from phytopathogens, plant-associated and soil bacteria, and enterobacteria ( Figure 5B). Some residues exclusively found in X. citri SsuA confer a more apolar local environment than that observed in the other bacterial orthologues ( Figure 5B, green colour).
The SsuA ligand-binding pocket has a volume of 1635+-26 Å 3 and an area of 2951+-12 Å 2 , but only 14% of the volume is occupied by the ligand molecule ( Figure 5C). The large binding pocket and the participation of at least 12 hydrophobic residues in forming this site reveal an adaptation that could explain the binding of a rather wide range of alkanesulphonate molecules of different sizes and shapes by this protein. Corroborating the intrinsic fluorescence data, W 185 faced the binding pocket, stabilising the alkane chain. Interestingly, all of the residues responsible for coordination of the sulphate (Gln 36 , Gly 68/86 and Ser 141 ) belong to two helices and one strand (a2, a5 and b3) ( Figure 5C, in green); together, these residues define a positively charged cluster ( Figure 5D). A similar structural organisation is found in other periplasmic ion-binding proteins such as the E. coli aliphatic sulphonate-binding protein, the Sinechocystis sp nitratebinding protein, the E. coli phosphate-binding protein, the Thermus thermophilus glutamate/glutamine-binding protein and the X. citri molybdate-binding protein ( Figure 5E).

Functional roles of X. citri SsuA
To evaluate the physiological role of SsuA during in vitro and in vivo growth of X. citri, a ssuA knockout strain (Xac::ssuA) was generated by site-specific mutagenesis ( Figure 6A and B). A complementary strain (Xac::ssuAc) was generated by transformation of Xac::ssuA with the pKX33-pssuA vector, which encoded the full length ssuA gene under control of the native ssu promoter, giving raise to the Xac::ssuAc strain ( Figure 6C). The isogenic X. citri ssuAdeleted strain was unable to grow in minimal medium containing alkanesulphonates (HEPES, MOPS, or MES) as the sole sulphur source but grew well in the presence of sulphate ( Figures 7A and  7B). This result indicated that the ssu operon is functional and is required for alkanesulphonate uptake in X. citri. In addition, the Xac::ssuA mutant formed small bright yellow colonies after growth in LB plates for 24 h ( Figure 7C). In both situations the altered phenotypes were reverted in the Xac::ssuAc strain ( Figure 7C). The altered colony morphology of the Xac::ssuA strain suggests a decrease in xanthan gum production, a feature previously observed by other groups [14,15,16]. Indeed, we determined that xanthan gum production by the Xac::ssuA strain during growth in LB medium was at least three-fold lower than that of the parental strain ( Figure 7D). Complementation with the pKX33-pssuA restored in a great extent production the colony morphology and production of xantham gum by the Xac::ssuA strain ( Figure 7D). To evaluate the effects of alkanesulphonate uptake on the The SsuA Protein from Xanthomonas citri PLOS ONE | www.plosone.org pathogenicity of X. citri, we monitored the behaviour of the wildtype, the Xac::ssuA and Xac::ssuAc strains after in vivo inoculation into a susceptible citrus host (C. sinensis). As shown in Figure 8A, the Xac::ssuA strain showed defective growth in leaf tissues 6 days after inoculation in comparison with the parental and complemented strains. In addition, the leaf lesions formed by the Xac::ssuA strain were smaller than those formed by the parental strain (data not shown). Leaves infected with the X. citri 306 strain showed typical blister-like lesions characteristic of the hyperplasia and the water-soaking phenotypes ( Figure 8B, I to III). Complementation of the Xac::ssuA with pKX33-pssuA restored the canker lesions symptoms to the wild-type levels ( Figure 8B, IV to VI). In contrast, much reduced lesions were observed in leaves inoculated with the Xac::ssuA strain, and the lesions that formed were not blister-like ( Figure 8B, VII to IX). Collectively, these results indicate that the alkanesulphonate uptake system also affects the in vivo behaviour of X. citri in C. sinensis plants.

Discussion
Although sulphonates and sulphur-containing organic compounds represent the most abundant sulphur source in most soils, there are no data regarding the role of alkanesulphonate uptake in the behaviour of plant pathogens, particularly those belonging to the Xanthomonas genus. In this work, we showed that a complete set of genes required for the uptake and metabolism of alkanesulphonates is found in the genomes of X. citri and X. fuscans, two closely related citrus pathogens. We expressed and purified a recombinant form of X. citri SsuA that specifically binds to various alkanesulphonates but not to sulphate, sulphate esters or a number of other sulphur-containing compounds. Using the recombinant protein, we solved for the first time the three-dimensional structure of a bacterial SsuA protein bound to three different alkanesulphonates (MES, HEPES and MOS). Determination of the crystal structure of the protein permitted the identification of the SsuA protein domains that interact with alkanesulphonate ligands and of the amino acid residues that coordinate these ligands within the ligand-binding pocket of the protein. The rather large binding pocket and the presence of several water molecules clearly accounts for how X. citri SsuA accommodates different alkanesulphonates molecules. Finally, we presented evidence that the Ssu system is functional in the X. citri 306 strain. Using an isogenic mutant carrying a knockout copy of the ssuA gene generated using gene replacement techniques we showed that the Ssu system affects both the in vitro and in vivo behaviour of the strain. In addition, complementation of the ssuA mutant strain with a plasmid encoding the wild type gene corrected the altered in vitro and in vivo phenotypes including the capability of the strain to induce disease symptoms in a susceptible citrus host. The present study offers new and relevant information regarding the structural and functional aspects of the alkanesulphonate uptake system of X.
citri and raises interesting questions regarding the metabolism of alkanesulphonates in the physiology and pathogenicity of this economically relevant phytopathogen.
The search for ssu genes homologous to those found in E. coli and other bacterial species revealed that, in the Xanthomonas genus, only X. citri, X. fuscans, X. campestris pv. musacearum and X. gardneri carry a complete set of ssu genes including the ABC transporter and enzymes involved in sulphonate desulphurisation (ssuD and ssuE). This finding contrasts with the widespread occurrence of functional genes involved with the uptake and metabolism of alkanesulphonates in many bacterial genera such as Bacillus, Pseudomonas, Ralstonia, Burkeholderia, Shingomonas and Nostoc; these bacteria are present in soil and live in environments in which sulphate esters and carbon-bonded sulphur (sulphonates or amino acid sulphur) represent most of the available sulphur [17,1]. The paucity of ssu genes in Xanthomonas species suggests that these genes have been recently acquired in the evolutionary history of the genus, most likely by a horizontal gene transfer event. This finding also raises questions about the role of alkanesulphonate uptake in the physiology of the species in which these genes are found.
The remainder of our study focused on the investigation of the Ssu uptake system in X. citri, the most economically relevant citrus pathogen. For that purpose, we determined the crystal structure and characterised the physiological role of the SsuA protein, the periplasmic component that confers specificity and affinity to the uptake system [1]. As a first step, we generated a recombinant form of X. citri SsuA that retained alkanesulphonate-binding properties similar to those expected for the native protein. As Table 1. Data processing statistics of the X. citri SsuA crystals and refinement data from its structures.  experimentally demonstrated, the recombinant protein exhibited changes in secondary structure and increased thermal stability after specific binding to MOPS, MES, CHES and HEPES but not after exposure to taurine, sulphate, thiosulphate or hydroxylamine.
These results indicate that the X. citri Ssu system is apparently specific for alkanesulphonates, as previously reported for Ssu orthologues found in P. putida, E. coli and B. subtilis [18,7,6]. Previous description of the crystal structure of Ssu components has been restricted to the unbound form of E. coli SsuA [8]. The availability of large amounts of recombinant X. citri SsuA capable of binding to various alkanesulphonates permitted us to solve for the first time the crystal structure of a bacterial SsuA protein in complex with HEPES, MOPS and MES. The size of the ligand cavity and the presence of 12 hydrophobic residues within it reveal an adaptation of the SsuA protein for interaction with long-chain alkane groups. Similarly relevant is the presence of water molecules in the binding pocket, which may offer flexibility to accommodate the ligands. In agreement with the results of spectroscopic assays, the crystal structure results indicate that SsuA-ligand interactions involve conformational changes associated with movement of the protein domains. According to the threedimensional structure of SsuA, W 185 is the most probable residue involved in the observed quenching of fluorescence after ligand binding because it is in close contact with the bound alkanesulphonate and stabilises the aromatic ring of the apolar chains of the molecule. The higher thermal stability evidenced when SsuA is bound to MOPS is due to the presence of shorter hydrogen bonds with the residues belonging to the dipole. Indeed, we showed that alkanesulphonate stabilisation of SsuA is centred in residues that are conserved in different SsuA orthologues, primarily those that form a positive dipole that specifically attracts and interacts with the sulphate group of the alkanesulphonate molecule [19][20][21]. The mechanism of binding and the very positive electrostatic potential of the entrance of the binding pocket suggest that the ligand is attracted to the protein by the sulphate group and then stabilised inside the pocket by the water molecules and the apolar environment. The X. citri SsuA structure reported here represents the first report of a bacterial SsuA orthologue bound to different helices and b-strands belong to the E. coli aliphatic sulphonate-binding protein (PDB 2X26), Sinechocystis sp. 6856 nitrate-binding protein (PDB 2G29), E. coli phosphate-binding protein (PDB 1IXH), Thermus thermophilus glutamate/glutamine-binding protein (PDB 1US5) and X. citri molybdate-binding protein (PDB 2H5Y). doi:10.1371/journal.pone.0080083.g005 alkanesulphonates and provides a reference for future studies of alkanesulphonate ABC transporters in different bacterial species.
The results obtained in vitro showed that a bacterial strain carrying an inactive copy of the ssuA gene failed to grow in the presence of alkanesulphonates such as HEPES, MOPS and MES under inorganic sulphate starvation conditions. This observation indicates that alkanesulphonate uptake is active in X. citri and that it may be required under environmental conditions in which the availability of inorganic sulphate is restricted. Xenobiotic and naturally occurring sulphonates, including taurine, isethionate, cysteic acid, methanesulphonate, and several undefined sulphonates, make up a large part of the sulphur present in soil humus and marine sediments [2,17]. In plants, in addition to sulphonolipids found in thylakoid membranes, a variety of S-containing secondary metabolites that often play important roles in defence against pathogens are synthesised [22]. Our results demonstrated that disruption of the ssuA gene resulted in an attenuated phenotype following infection of a susceptible citrus host (C. sinensis). This phenotype was attributable at least in part to the reduced production of xanthan gum, which is required for the water soaking phenotype and for the plant's resistance to various forms of environmental stress [23,24,25]. The biosynthesis of xanthan gum involves the coordinate expression of 12 genes that are part of the gumB-gumM operon [25,26]. Twelve of these genes encode proteins with 4 or more cysteine residues. A reduction in the sulphur supply after disruption of the ssuA gene may have affected the synthesis of cysteine-enriched proteins and, thus, indirectly reduced xanthan gum production.
Although the precise role of alkanesulphonates during the in vivo growth of X. citri is not presently known, our results indicate that alkanesulphonates may have a previously unsuspected relevance to plant-bacterial interactions. Future studies directed at achieving a better understanding of the nutritional and physiological roles of alkanesulphonates in X. citri are warranted.

Bacterial strains, plasmids and culture conditions
The X. citri 306 strain and the E. coli DH5a and BL21 strains were grown in LB medium [27] at 30uC and 37uC, respectively. The X. citri 306 strain, the isogenic ssuA-knockout mutant and complemented strain Xac::ssuAc were also grown in M9 minimal medium [27] and sulphate-free M9 minimal medium supplemented with sulphate or alkanesulphonates as described in the Results section at 28uC under aerobic conditions. When required, kanamycin (25 or 50 mg/ml), spectinomycin (50 mg/ml), and/or ampicillin (100 mg/ml) were added to selective media.
Cloning of the gene encoding X. citri SsuA protein. All gene cloning steps were carried out with the E. coli DH5a strain; expression of the recombinant SsuA protein was carried out with the E. coli BL21 (DE3) strain (Novagen). The DNA fragment with the X. citri ssuA gene sequence, encoding the mature protein without the first 126 base pairs (42 amino acids) corresponding to the signal peptide, was amplified by PCR (forward primer 59 gCgCATATggCCgAgCCggCgCA 39; reverse primer 59 gCgAAgCTTTCATTTgCTCAC 39) with Platinum High Fidelity Taq polymerase (Invitrogen) under standard amplification conditions. The amplicon, corresponding to 972 base pairs, was cloned into the vector pGEM T-Easy (Promega) and subcloned into the vector pET28a (Novagen) for expression in E. coli BL21 (DE3).
Expression and purification of recombinant X. citri SsuA. Cultures of the recombinant E. coli BL21 (DE3) strain transformed with pETSsuA were prepared aerobically in LB medium supplemented with 50 mg/mL kanamycin until mid-log phase (OD 600 0.5-0.6); IPTG was then added to a final concentration of 0.1 mM. The cultures were induced aerobically (200 rpm) for 2.5 h at 28uC. Cells were collected by centrifugation at 8,000 g for 15 min at 4uC and stored at -20uC for approximately 16 h before lysis. The cell pellets were suspended in buffer 1 (50 mM sodium phosphate buffer, pH 7.2, containing 100 mM NaCl, 5% glycerol and 20 mM imidazole) and incubated with lysozyme (final concentration of 100 mg/mL) and PMSF (1 mM) for 1 h in an ice bath. The cells were sonically disrupted in a Branson Digital Sonifier (Model 450), and soluble fractions were separated from the non-soluble material by centrifugation at 16,000 g for 30 min at 4uC. The recombinant SsuA protein was purified from the soluble fraction by immobilised metal affinity chromatography using a HistrapHP column (GE Healthcare) according to the manufacturer's instructions. The charged resin was washed with buffer 1 (30 bed volumes) followed by step gradient elution with buffers containing increasing concentrations of imidazole (50-500 mM). The eluted fractions were dialysed once against 20 mM Tris-HCl at pH 7 containing 50 mM NaCl and concentrated with Ultrafree MWCO 10,000 centrifugal filters (Amicon Millipore) to a final concentration of 6 mg/mL. The eluted fractions were analysed by SDS-PAGE using 12% acrylamide gels and stained with Coomassie Blue. Protein concentration was determined spectrophotometrically using the Edelhoch method [31]. Recombinant SsuA without the His6-tag was obtained by cleavage of the purified protein with thrombin (Sigma Aldrich, USA) at room temperature. Samples of 20 mg protein were incubated for 2 h with 10 units thrombin in 50 mM Tris, pH 7.5, 10 mM NaCl and 1 mM DTT.
Spectroscopic analyses of the recombinant SsuA protein. The stability of the protein in solution at various pH values (20 mM Tris, pH 7.0, 50 mM NaCl; 20 mM sodium acetate, pH 5.0, 50 mM NaCl; 20 mM sodium citrate, pH 3.0, 50 mM NaCl; 20 mM glycine, pH 10.0, 50 mM NaCl) and its behaviour in the presence of aliphatic sulphonates was monitored by following the intrinsic fluorescence of the tryptophan residues using an Aminco BOWMAN series 2 spectrofluorometer. The excitation and emission bandwidths were 4 and 8 nm, respectively. The fluorescence cell (161 cm) was mounted on a thermostatic holder. Tryptophan fluorescence was measured at an excitation wavelength of 295 nm, and emission spectra were recorded between 340 and 420 nm. All measurements were performed using 10 mM protein and 20 mM sulphonates in 20 mM Tris, pH 8.0. Circular dichroism measurements were carried out on a JASCO J-810 spectropolarimeter equipped with a Peltier-type temperature controller and a thermostatic cell holder interfaced with a thermostatic bath. Spectra were recorded in quartz cells with a 0.1 cm path length at a protein concentration of 10 mM in various buffers. Twenty consecutive scans were accumulated and averaged. The data were corrected for the baseline contribution of the buffer, and the observed ellipticity was converted into the mean residue ellipticity [h] based on a mean residue molecular mass of 34,000 Da. Secondary structure was estimated from fitted Far-UV CD spectra using the DICROPROT software package [32]. Thermal shift assays were performed with the dye SYPRO orange and conducted in the iCycler iQ Real Time Detection System (Bio-Rad, Hercules, CA). The Tm represents the temperature at the midpoint of the unfolding transition. Various concentrations of protein, ligand and buffer were evaluated in 96well iCycler iQ PCR plates for determination of the appropriate conditions for the assay. For the final experiments, solutions of 100 ml were prepared with 6 mM SsuA, 2.7X SYPRO orange, and 5 mM ligand in 20 mM Tris, pH 7.4, 50 mM NaCl. The plate was heated from 25uC to 85uC at a heating rate of 0.5uC/min. Crystallisation of the X. citri SsuA protein. Crystallisation conditions were screened using sparse-matrix screens in 96well plates with protein at a concentration of 6 mg/ml in 20 mM Tris buffer, pH 7.0, containing 50 mM NaCl at 18uC. SsuA crystallisation was performed with the sitting-drop vapourdiffusion method by mixing equal volumes (3 ml) of protein solution and crystallisation solutions. Data from the first crystals obtained in HEPES were collected at 100 K at the D03B-MX1 beam line Brazilian Synchrotron Light Laboratory (LNLS) [33] using 1.433 Å radiation and recorded on a MARCCD165 detector (oscillation data with Dw = 1.0 o ). The crystals were cooled to 110 K in a stream of nitrogen gas to minimise radiation damage. Cryoprotection was obtained by soaking crystals into drops of crystal buffer plus 10% glycerol. Data collection was performed at 100 K at the D03B-MX2 beam line Brazilian Synchrotron Light Laboratory (LNLS) using 1.6 Å radiation and recorded on a MARCCD325 detector (oscillation data with Dw = 1.0 o ). Structure solution, model building and refinement. Diffraction data were indexed using the XDS package [34] and processed using SCALA and TRUNCATE [35]. NaI and CsCl 3 derivative crystals were obtained by quick cryo-soaking for 15 s in a solution composed of the mother liquor supplemented with 0.5 M NaI or CsCl 3 and 20% glycerol, yielding a data resolution of 2.4 Å . The SsuA 3D structure was solved by multiple isomorphous replacement and anomalous scattering (MIRAS) using the NaI and CsCl 3 derivative datasets. The program AutoSHARP [36] was used to calculate the phases and to determine the incorporation of iodine and caesium chloride sites using SHELX [37] and density modification procedures. The ArpwArp [38] program was used to automatically build approximately 80% of the residues. Construction of the model was performed using COOT, and 14 cycles of refinement were realised using REFMAC [39]. The SsuA structure in the presence of MOPS and MES was obtained by molecular replacement using the structural coordinates of SsuA bound to HEPES (PDB code 3E4R).
Construction of the Xac::ssuA mutant. Site-specific inactivation of the ssuA gene was carried out with the X. citri 306 strain using a suicide plasmid carrying an inactivated copy of the ssuA gene for the selection of a gene replacement event, according to previously published procedures [40]. The first step involved cloning of the ssuA gene into the pUC4 plasmid after PCR amplification with primers FssuA2Nde28a (59 GCGCA-TATGGCCGAGCCGGCGCA 39) and RssuA2Hind28a (59GCGAAGCTTTCATTTGCTCACC 39) (the NdeI and NotI restriction sites are underlined). Inactivation of the ssuA gene was achieved by cloning a 2 kb gene cassette derived from SmaIcleaved pHP45omega and encoding resistance to spectinomycin/ streptomycin [41] into the single KpnI cleavage site within the ssuA gene. A clone carrying an inactivated copy of the ssuA gene (ssuA::sp/sm) was selected and confirmed by restriction analysis and automatic DNA sequencing. The mutated copy of the ssuA gene was released for the next cloning step by cleavage with NdeI and NotI restriction enzymes. The mutated ssuA::sp/sm gene was cloned into the suicide pNPTS138 vector (originally constructed by D. Alley and kindly supplied by Dr. M. V. Marques of the Department of Microbiology, University of São Paulo). The pNPTS138 plasmid permits selection of single chromosomal integrations after plating transformed cells on plates containing kanamycin and subsequent selection of gene replacement events after a second selection round in media containing sucrose (because the plasmid carries a sacB gene that confers sensitivity to sucrose) and spectinomycin [41]. The recombinant plasmid was named pNssuA and introduced by electroporation into the X.citri 306 strain as previously described [40]. Initial selection for the gene replacement event was carried out on plates containing kanamycin, followed by overnight growth of selected colonies in nonselective media and subsequent plating on 3% sucrose and 50 m/ml spectinomycin to screen for cells that had undergone a second crossover event leading to excision of the plasmid carrying the wild type ssuA copy. PCR with primers FssuA2Nde28a and RssuA2Hind28a was used to confirm chromosomal deletions.
Construction of the expression vector for complementation of the Xac::ssuA mutant. A fragment containing a 900 base pairs upstream of the XAC3200 gene from ssu operon was amplified by PCR using primers FssuSalI-p (59 CCAGAAACG-CATTCATGACCTC 39) and RssuEcoRI-p (59 CAGCCCATA-CAAAAGGACGTCA 39). This fragment contained the promotor sequence, which was cloned into the pKX33 vector [42] for construction of the pKX33-p. The full ssuA gene (1026 bp) was amplified from X. citri genomic DNA by PCR using the primers FssuA-EcoRIc (59 GAATTCATGCGGGCAACGGGCAGG 39) e RssuA-XbaIc (59 TCTAGATCATTTGCTCACCGCCTGCG 39) and cloned in to the pKX33 with the promoter sequence to build the pKX33-pssuA vector for expression in X. citri cells. The final plasmid was confirmed after digestion with SalI and XbaI ( Figure 6C) and automatic DNA sequencing. Competent cells of Xac::ssuA strain were transformed with the final vector pKX33-pssuA by electroporation followed by selection on LB plates with 50 mg/ml kanamycin. The cells are described as Xac::ssuAc.
Xanthan gum production. X. citri 306 strain, the isogenic Xac::ssuA mutant and the mutant complemented (Xac::ssuAc) were grown in LB medium for 24 h at 30uC under aerated conditions (200 rpm) and subjected to the protocol described by Vojnov and collaborators [14]. Kanamycin (50 mg/ml) was added to the cultures of the mutant and complementary strains.
Growth curves and plant infection experiments. In vitro growth of wild-type and mutant strains was performed in 10 ml M9 minimal medium and sulphur-free M9 minimal medium supplemented with different alkanesulphonates (MES, CHES, HEPES) as sulphur sources at 28uC. Samples were taken every 2 h for determination of the total number of viable cells. In vivo assays were performed as previously described [43]. Sixmonth-old plants of sweet orange (C. sinensis) were obtained from certified nurseries and kept in a growth room at 25-28uC with fluorescent light illumination. Leaf sectors were infiltrated with approximately 7610 6 viable bacteria in a total volume of 0.3 ml. Bacterial cultures were prepared in LB without NaCl for 48 h at 28uC with shaking at 200 rpm and subsequently suspended in sterile water. The appearance of canker pustules and lesion phenotypes was monitored daily for 20 days. To follow the growth of the bacteria inside the plant tissue, leaves were infiltrated with 7610 6 viable bacteria in 6 different sectors. Every other day, three circular sections 1 cm in diameter containing the bacterial infiltrates were macerated in water and diluted aliquots were plated on LB agar plates for determination of the number of viable bacteria.