Genome-Wide Mutagenesis of  Xanthomonas axonopodis  pv.  citri  Reveals Novel Genetic Determinants and Regulation Mechanisms of Biofilm Formation

doi:10.1371/journal.pone.0021804

Research Article

Genome-Wide Mutagenesis of Xanthomonas axonopodis pv. citri Reveals Novel Genetic Determinants and Regulation Mechanisms of Biofilm Formation

Abstract

Xanthomonas axonopodis pv. citri (Xac) causes citrus canker disease, a major threat to citrus production worldwide. Accumulating evidence suggests that the formation of biofilms on citrus leaves plays an important role in the epiphytic survival of this pathogen prior to the development of canker disease. However, the process of Xac biofilm formation is poorly understood. Here, we report a genome-scale study of Xac biofilm formation in which we identified 92 genes, including 33 novel genes involved in biofilm formation and 7 previously characterized genes, colR, fhaB, fliC, galU, gumD, wxacO, and rbfC, known to be important for Xac biofilm formation. In addition, 52 other genes with defined or putative functions in biofilm formation were identified, even though they had not previously reported been to be associated with biofilm formation. The 92 genes were isolated from 292 biofilm-defective mutants following a screen of a transposon insertion library containing 22,000 Xac strain 306 mutants. Further analyses indicated that 16 of the novel genes are involved in the production of extracellular polysaccharide (EPS) and/or lipopolysaccharide (LPS), 7 genes are involved in signaling and regulatory pathways, and 5 genes have unknown roles in biofilm formation. Furthermore, two novel genes, XAC0482, encoding a haloacid dehalogenase-like phosphatase, and XAC0494 (designated as rbfS), encoding a two-component sensor protein, were confirmed to be biofilm-related genes through complementation assays. Our data demonstrate that the formation of mature biofilm requires EPS, LPS, both flagellum-dependent and flagellum-independent cell motility, secreted proteins and extracellular DNA. Additionally, multiple signaling pathways are involved in Xac biofilm formation. This work is the first report on a genome-wide scale of the genetic processes of biofilm formation in plant pathogenic bacteria. The report provides significant new information about the genetic determinants and regulatory mechanism of biofilm formation.

Citation: Li J, Wang N (2011) Genome-Wide Mutagenesis of Xanthomonas axonopodis pv. citri Reveals Novel Genetic Determinants and Regulation Mechanisms of Biofilm Formation. PLoS ONE 6(7): e21804. doi:10.1371/journal.pone.0021804

Editor: Indra Neil Sarkar, University of Vermont, United States of America

Received: March 31, 2011; Accepted: June 7, 2011; Published: July 5, 2011

Copyright: © 2011 Li, Wang. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Funding: This work was supported by the United States Department of Agriculture, Cooperative State Research, Education, and Extension Service, Special Citrus Canker Grant Project 88811 (www.csrees.usda.gov). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Competing interests: The authors have declared that no competing interests exist.

Introduction

Gram-negative plant pathogenic bacteria belonging to the genus Xanthomonas cause severe diseases in many economically important crop plants around the world and exhibit extremely high host-pathogen specificity [1]. Xanthomonas spp., including Xanthomonas axonopodis pv. citri (Xac) (syn. Xanthomonas citri, Xanthomonas campestris pv. citri or Xanthomonas citri subsp. citri) [2], [3], [4], have been used as model organisms for studying host-bacterium interactions [1], [5], [6], [7]. Xac causes citrus canker, a destructive disease in citrus, and affects most commercial varieties of citrus, limiting citrus production worldwide [8], [9]. Xac is typically spread by windblown rain and invades host plants directly through natural openings, such as stomata, and through wounds. The pathogen multiplies in intercellular spaces to cause canker disease [8]. Typical symptoms include raised corky lesions surrounded by a water or oil-soaked margin on leaves, stems, and fruits, inducing defoliation, twig dieback, general tree decline, blemished fruit, and premature fruit drop in severely infected trees [9]. Control is difficult in areas where the disease is already established and is based on the heavy use of copper-containing compounds. Recurrent and severe attacks of the disease are responsible for serious economic losses in citrus groves [9].

Early studies have shown that Xac forms biofilms on both abiotic and biotic surfaces [10], [11], [12]. Biofilms exhibit complex structures that involve groups of microcolonies attached by a bacterially produced matrix. Biofilms are of great medical, industrial, and agricultural interest because of their prevalence and general resistance to adverse conditions such as environmental stresses, host defense mechanisms, and antimicrobial treatment [13], [14]. In plant-bacteria interactions, biofilm formation has been reported to be implicated in the virulence of diverse bacterial pathogens [13]. The ability of the citrus canker pathogen to form biofilms enhances epiphytic persistence on host leaves, which plays an important role in the early stages of infection [12], [15].

Biofilm formation is a dynamic and complex process that generally includes the initial surface attachment of cells to the substratum, physiological changes within the organism, multiplication of the cells to form microcolonies, and eventually maturation of the biofilm [14]. Because of this complexity, knowledge regarding the process of biofilm formation and its regulation is limited. Thus, elucidation of the underlying genetic determinants and regulatory processes are prerequisites for further understanding the mechanism of biofilm formation [16]. The genome sequences of Xac and multiple close relatives have been sequenced [5], [17], which has facilitated research in this area. Tremendous progress has been made in identifying the genetic determinants and regulatory processes of Xac. These genes include the two-component signal transduction system (TCSTS) encoded by colS/colR [18], the filamentous hemagglutinin-like adhesin encoded by fhaB [10], the flagellin gene encoded by fliC [19], the UTP-glucose-1-phosphate uridylyltransferase encoded by galU [11], the xanthan EPS gene cluster encoded by gumB and gumD [12], [20], and the LPS biosynthesis associated genes encoded by wxacO and rbfC [15]. However, a comprehensive understanding of Xac biofilm formation has not yet been achieved because previous studies have focused only on individual genes and the specific genetic pathways responsible for biofilm formation. Transposon mutagenesis has been widely used for a global identification of biofilm-related genes in human bacterial pathogens [21], [22] and has the potential to identify novel Xac biofilm-related genes.

The goal of the present study was to advance our understanding of the underlying genetic determinants and regulatory processes of biofilm formation by the citrus canker bacterium Xac. An EZ-Tn5 library containing 22,000 Xac strain 306 mutants was used to study the process of biofilm formation. Screening for biofilm-defective mutants led to the identification of novel genetic determinants and an understanding of the regulatory mechanisms of Xac biofilm formation. Importantly, this work is the first report of the genetic process of biofilm formation by plant pathogenic bacteria on a genome-wide scale, and novel insights into the underlying mechanisms of biofilm formation revealed by this work hold the potential for exploitation to improve control strategies of citrus canker.

Results

Isolation and characterization of biofilm-defective mutants

To further understand citrus canker bacterium biofilm formation, an EZ-Tn5 mutant library containing 22,000 Xac strain 306 mutants [11] was screened for biofilm-defective mutants to identify genes involved either directly or indirectly in biofilm formation. During the initial screening, 327 mutants were identified as biofilm-defective mutants based on the reduced ability of the bacteria to adhere to the surface of the wells of polystyrene 96-well microtiter plates, as visualized by staining with crystal violet. Mutants that lacked a violet ring formation on the side of each well or a generalized staining of the well were scored as having a biofilm-defective phenotype. Following further confirmation of these phenotypes by quantitative biofilm tests in borosilicate glass tubes, the growth of each strain was assessed. After the elimination of mutants with inconsistent biofilm formation or general growth deficiencies, 292 mutants were confirmed as being biofilm-defective mutants (Fig. 1 and data not shown) and subjected to further investigation. All 292 mutants exhibited significant (P<0.01 by Student's t-test) and strong (from 20% to 80%) defects in biofilm formation compared to wild-type Xac strain 306 (Fig. 1 and data not shown). For convenience, these mutants were designated as biofilm-defective phenotype (bdp) mutants. These 292 bdp mutants represented 1.32% of the 22,000 mutants tested.

thumbnail — Figure 1. Quantitative biofilm assay in borosilicate glass tubes showing the ability to form biofilm by a proportion of the EZ-Tn5 mutants of *Xanthomonas axonopodis* pv. *citri* strain 306.
The biofilm was determined using crystal violet staining coupled with measurements of absorbance at 590 nm (A₅₉₀). The experiment was repeated three times with five replicates. Averages and standard errors from one of three representative experiments with similar results are shown. Wild-type strain 306 was set as equal to 100%. Statistical significance was tested by Student's t-test (* indicates significant differences from wild-type strain 306 at P<0.01; ** indicates significant differences from wild-type strain 306 at P<0.001).

doi:10.1371/journal.pone.0021804.g001
More »

To test the EZ-Tn5 transposon copy number in each mutant, Southern blotting was performed with the kanamycin resistance gene as a probe. Southern hybridization analysis revealed a single band in 16 out of 17 bdp randomly selected EZ-Tn5 insertion mutants in novel genes (Table 1 and see below), but no bands were detected in the wild-type Xac strain 306 (Supporting Information Figure S1); a double insertion was found in only one mutant, suggesting a high ratio of single transposon insertions in the genome as described by the EZ-Tn5™ <R6Kγori/KAN-2>Tnp Transposome™ Kit (Epicentre, Madison, WI).

Genomic distribution of transposon-tagged biofilm related genes

The precise EZ-Tn5 insertion sites of the bdp mutants were determined by sequencing analysis based on the random amplification of transposon ends polymerase chain reactions (RATE-PCR) [23]. Sequences flanking transposon insertions were identified by a homology search of the entire Xac strain 306 genome sequence (GenBank accession no. AE008923) using the BLASTN search algorithm. To verify the accuracy of transposon insertions in the sequences, PCR amplification was performed using primers designed from the sequences flanking the ORF of _target genes. The size of each PCR product from the mutants was increased by the insertion of a transposon (1980 bp), indicating that the transposons were precisely inserted into _target genes in the Xac genome (data not shown).

In some cases (approximately 40% of the identified mutants), two or more independently derived mutants mapping to the same gene coding region were found, but at different locations. In these cases, only a single mutant from each gene was chosen for further analyses. In total, ninety-two distinct genes and four intergenic regions were identified (Supporting Information Table S1), representing 2.1% of the 4427 predicted open reading frames (ORFs) in the chromosomal genome (4312 ORFs) and in plasmids (115 ORFs) [5]. Ninety-three insertions occurred in the chromosome of Xac strain 306, two insertions were on the pXAC66 plasmid, and one insertion was on the pXAC33 plasmid. The transposons were widely inserted in diverse Xac genes, suggesting genome-wide insertions, and transposon _target genes were randomly distributed throughout the Xac strain 306 genome (Supporting Information Figure S2).

Transposon insertion may cause a polar effect that affects the normal transcription of downstream genes within the same operon. The availability of Xac strain 306 genomic sequences made it possible to predict polar effects in the obtained mutants. Based on the results of transcription-unit prediction by BioCyc [24] and the insertion direction of the transposon, all biofilm-formation-related genes identified in this study could be divided into two groups. One contains 66 genes that are organized as a monocistron or at the 3′ end of operons, or at the 5′ end of operons and transposon insertions with the same transcription direction as the operons. In this group of 66 genes, transposon insertions may not cause polar effects (Supporting Information Table S1). The other group of biofilm-formation-related genes contained 26 genes with insertions that at the 5′ end of operons but with the insertion occurring in the reverse transcriptional direction from the operons. Transposon integration in these genes may lead to polar effects on downstream gene expression that complicate phenotypic characterization.

Functional classification of biofilm-formation-related genes

As suggested by the Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway[25], the 92 biofilm-formation-related genes are involved in carbohydrate, amino acid, nucleotide, or energy metabolism; bacterial chemotaxis and motility; flagellum or pilus assembly; DNA replication and repair; transcription; membrane transport; signal transduction; signaling molecule biosynthesis and interaction; not well characterized and unknown functions (Supporting Information Table S1). Interestingly, 17 out of the 92 genes were previously unknown genes encoding hypothetical proteins whose functions have yet to be assigned. There were 16 other genes assigned putative functions, including 11 without gene names that belonged to the “not well characterized” gene family based on the KEGG pathway [25]. Therefore, these 28 genes, which included the 17 hypothetical genes and the 11 genes without gene names, were designated as bdp genes for their biofilm-defective phenotype (Supporting Information Table S1). Among the 64 other genes, 7 (colR, fhaB, fliC, galU, gumD, wxacO, and rbfC) were previously characterized as being important for Xac biofilm formation, and the 57 other genes were assigned putative functions and may be involved in biofilm formation (Supporting Information Table S1).

Transposon insertions in known loci related to biofilm formation

As mentioned above, the 92 biofilm-formation-related genes identified in this study contain both genes that were known to be linked to biofilm formation and genes that were not known to be associated with biofilm. In the category of known loci, a subset of 7 genes (Supporting Information Table S1), i.e., colR [18], fhaB [10], fliC [19], galU [11], gumD [20], wxacO, and rbfC [15], were previously characterized to be important for Xac biofilm formation. In addition to the 7 genes mentioned above, we uncovered 52 genes with defined or putative functions involved in biofilm formation, although they had not been previously reported to be involved in Xac biofilm formation (Supporting Information Table S1; Supporting Information Figure S3). These genes included thirteen flagellum biosynthesis genes (fleN, flgABFKL, flhB, fliCFMQR); six bacterial chemotaxis and motility genes (cheA, cheY, mcpA, motB, motD and tsr); one pilus biosynthesis protein (pilB); five gum genes (gumCEFJK); three LPS biosynthesis genes (ipsJ, rmlA and rmlB); multiple enzymes involved in carbohydrate, amino acid, nucleotide, or energy metabolisms (bioB, ilvE, ldp, metB, nuoM, rpfF, sahH, thiE, thiG, trpE, ugd, xanA, and xanB); DNA replication and repair (mrdB and nrdF); signal transduction and transcriptional regulators (clp, opsX, rfpC and rpoN); and transporters (iroN, msbA, sbp, wzm and wzt). Inactivation of orthologs of these genes led to biofilm-defective phenotypes in a variety of bacteria [14].

Novel genes associated with EPS production contribute to biofilm formation

In addition to the known loci associated previously with biofilm formation, we found 33 genes that were either hypothetical genes or not well-characterized genes assigned putative functions based on homology analyses, and of these, 28 genes were designated as bdp genes (see above). Notably, 26 of the 28 bdp genes contained transposon insertions that did not result in polar effects (Supporting Information Table S1). As an initial step in investigating the roles of these novel genes in biofilm formation, we first characterized genes that were potentially associated with polysaccharide production because EPS is the major structural constituent of many bacterial biofilms [14], [16].

Our data revealed that a total of 13 bdp gene mutants, including 332D5 (XAC0482/bdp1), 301B4 (XAC0494/bdp18/rbfS), 285E8 (XAC0721/bdp2), 270H4 (XAC1469/bdp3), 319B1 (XAC1499/bdp20), 411F5 (XAC1509/bdp4), 296G1 (XAC1778/bdp21), 295H3 (XAC2848/bdp7), 223G4 (XAC3110/bdp24), 228F10 (XAC4203/bdp14), 295D8 (XAC4264/bdp15), 288C6 (XACa0007/bdp16), and 419F10 (XACb0050/bdp28), were significantly reduced in total EPS production compared to wild-type strain 306 (P<0.01, Student's t-test). The 13 other mutants that were tested produced similar levels of EPS to those produced by wild-type strain 306 (Table 1).

Based on the putative functions of their corresponding gene products (Table 1), these EPS-related genes could be classified into three groups. The first group consisted of those whose corresponding gene products were probably enzymes directly involved in carbohydrate biosynthesis and/or metabolism, which included the XAC0482/bdp1, XAC3110/bdp24, and XAC4264/bdp15 genes. The protein encoded by XAC0482/bdp1 contains a haloacid dehalogenase (HAD)-like hydrolase domain and shares significant similarity (>95% amino acid identity) with members of an HAD-like hydrolase/phosphatase family that is conserved in most Xanthomonas species. Proteins of this family have been found in various bacteria, but most remain uncharacterized [26]. The putative XAC3110 protein contains a glycosyl transferase family 2 domain and a UDP-Glycosyltransferase/glycogen phosphorylase domain. The glycosyl transferase family 2 plays a general role in polysaccharide biosynthesis [27]. The XAC4264/bdp15 gene encodes a 279 amino acid hypothetical protein with 40% similarity in amino acid sequence to the PalI sucrose isomerase from Erwinia tasmaniensis Et1/99, a non-pathogenic epiphytic plant bacterium [28]. Sucrose isomerase has been reported in a wide range of bacterial species, and in addition to catalyzing the isomerization of sucrose to isomaltulose, it also produces another sucrose isomer, trehalose, and glucose and fructose byproducts [29]. Earlier studies suggested that the XAC4264 protein interacts with VirD4, a component of the type IV secretion system (T4SS), and XAC4264 may act as a cofactor or substrate of the Xanthomonas T4SS [30]. Our data suggest that the XAC4264 protein is involved in EPS production (Table 1); however, whether it encodes a sucrose isomerase remains to be elucidated.

A second group of EPS-related genes contained five genes, XAC0494/bdp18/rbfS, XAC0721/bdp2, XAC1499/bdp20, XAC1778/bdp21 and XACa0007/bdp16, whose corresponding gene products are potentially associated with signaling and regulation pathways, which are described below (Table 1).

A third group of EPS-related genes were those whose corresponding gene products could not be attributed to the above two groups. This group included the XAC1469/bdp3, XAC1509/bdp4, XAC2848/bdp7, XAC4203/bdp14, and XACb0050/bdp28 genes. The putative XAC1469/Bdp3 protein is a hypothetical protein with a TatD_DNase domain. The TatD-related DNase family is a family of Mg-dependent DNases that participate in DNA replication, recombination and repair [31]. The XAC1509/bdp4 gene was annotated as a hypothetical protein without any conserved domains. The closest homolog (80% amino acid identity) to XAC1509 was found in X. campestris pv. campestris strain B100 with a protein of unknown function. The putative 363 amino acid XAC2848 protein contains a predicted DUF3616 domain (Pfam entry PF12275). This family of proteins is mainly found in bacteria with unknown functions. The XAC4203 protein is a 1284 amino acid hypothetical protein that has a predicted signal peptide domain and a DUF0496 domain (Pfam entry PF04357) with an uncharacterized function at its C-terminal. Interestingly, the XAC4203 homolog of PXO_03725 in X. oryzae pv. oryzae strain PXO99A was annotated as a pathogenicity protein [7]. However, the role of PXO_03725 in pathogenesis remains unknown. The deduced XACb0050 protein is an ISxac2 transposase [5]. Bacterial transposase has been reported to be associated with biofilm formation. For example, the transposase encoded by IS256 in Staphylococcus epidermidis has the capacity to influence biofilm formation, either by insertion into regulatory genes or by modulating biofilm gene expression [32].

Novel genes involved in LPS biosynthesis contribute to biofilm formation

Because LPS has been demonstrated to be directly involved in biofilm formation in a number of bacteria, including Escherichia coli [22] and Pseudomonas aeruginosa [33], we were interested in determining whether the above-mentioned 26 bdp genes were associated with LPS biosynthesis in Xac. Sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) analysis revealed that LPS patterns, especially O-antigen profiles, were altered in seven bdp mutants, including 285E8 (XAC0721/bdp2), 296G1 (XAC1778/bdp21), 295H3 (XAC2848/bdp7), 23G4 (XAC3110/bdp24), 292G4 (XAC3364/bdp8), 207A6 (XAC3576/bdp9), and 228F10 (XAC4203/bdp14), when compared with wild-type strain 306 (Fig. 2).

Interestingly, five mutants, including 285E8 (XAC0721/bdp2), 296G1 (XAC1778/bdp21), 295H3 (XAC2848/bdp7), 223G4 (XAC3110/bdp24), and 228F10 (XAC4203/bdp14), were also altered for EPS production (see above). Thus, these genes may play a role in the biosynthesis or transportation of certain precursors of EPS and LPS.

The product of the XAC3364/bdp8 locus is a putative acetyl-CoA hydrolase. In some bacteria, e.g. Pseudomonas reinekei, the acetyl-CoA hydrolase is able to catalyze the hydrolysis of acetyl-CoA, glutaryl-CoA, and 3-butenoyl-CoA and is presumably involved in both central metabolism and energy metabolism [34], [35]. However, it is unknown how acetyl-CoA hydrolase is involved in LPS production or how it contributes to biofilm formation in bacteria.

The XAC3576/bdp9 gene encodes a hypothetical protein containing a signal peptide and a transmembrane segment at the N-terminal region in addition to an RgpF (rhamnose-glucose polysaccharide assembly protein F) domain (Table 1). Members of the RgpF family have been suggested to be involved in the assembly of the LPS O-polysaccharides [36], [37].

Flagellum-independent motility plays an important role in biofilm formation

Given that both flagellum-dependent and flagellum-independent motility are required for the formation of mature biofilm by a variety of bacteria, including Xac [19; for review see 14], we attempted to determine whether the 26 bdp genes are involved in Xac cell motility. A significant reduction (P<0.01, Student's t-test) in both swimming and swarming motility was observed in the following seven mutants: 332D5 (XAC0482/bdp1), 301B4 (XAC0494/bdp18/rbfS), 296G1 (XAC1778/bdp21), 290E7 (XAC2018/bdp6), 291G5 (XAC2670/bdp23), 223G4 (XAC3110/bdp24), and 292G4 (XAC3364/bdp8) (Table 1). In addition, the 285E8 (XAC0721/bdp2), 291A9 (XAC3591/bdp25) and 225E4 (XAC3597/bdp11) mutants showed decreased swimming motility, whereas the 411F5 (XAC1509/bdp4), 319B1 (XAC1499/bdp20), 228F10 (XAC4203/bdp14), and 295D8 (XAC4264/bdp15) mutants showed differences in swarming motility. The other bdp mutants had a similar motility to the wild-type 306 strain (Table 1).

The motility-impaired mutants could be grouped into two categories: 1) mutations potentially associated with signaling and regulation pathways, including the 301B4 (XAC0494/bdp18/rbfS), 285E8 (XAC0721/bdp2), 319B1 (XAC1499/bdp20), 296G1 (XAC1778/bdp21), 290E7 (XAC2018/bdp6), and 291G5 (XAC2670/bdp23) mutants, and 2) mutants that were also impaired in EPS and/or LPS production. This group included the 332D5 (XAC0482/bdp1), 411F5 (XAC1509/bdp4), 223G4 (XAC3110/bdp24), 292G4 (XAC3364/bdp8), 291A9 (XAC3591/bdp25), 225E4 (XAC3597/bdp11), 228F10 (XAC4203/bdp14), and 295D8 (XAC4264/bdp15) mutants. The disrupted genes in these mutants shared no significant similarity to any motility-related genes or regulators (see above), suggesting that their impaired motility may be mainly due to their altered EPS and/or LPS production. The gumD mutant revealed significant reduction both in swimming and swarming motility (Supporting Information Figure S4), which is consistent with the reports that xanthan EPS is required for a type of flagellum-independent motility in Xac [19] and X. campestris pv. campestris [38]. Moreover, the reduced swimming phenotype in these mutants may be caused by an altered LPS. This result is consistent with our recent observations in the wxacO and rfbC mutants [15].

Novel signaling and regulatory factors in biofilm formation

As mentioned above, we uncovered a total of seven bdp genes, XAC0494/bdp18/rbfS, XAC0721/bdp2, XAC1499/bdp20, XAC1778/bdp21, XAC2018/bdp6, XAC2670/bdp22 and XACa0007/bdp16, associated with certain signaling and regulation pathways. Notably, both the XAC0721/bdp2 and XAC1778/bdp21 genes were involved in EPS and LPS production as well as cell motility (Table 1; Fig. 2), both the XAC0494/bdp18/rbfS and XAC1499/bdp20 genes were involved in EPS production and cell motility (Table 1), and both the XAC2018/bdp6 and XAC2670/bdp22 genes were associated with cell motility (Table 1). The XACa0007/bdp16 gene was involved only in EPS production (Table 1).

Interestingly, XAC0494/bdp18/rbfS (designated as rbfS, for regulation of biofilm formation, sensor) encodes a two-component system sensor protein of 769 amino acids with unknown function [5]. Based on BioCyc Transcription-Units prediction [25], XAC0495 and rbfS form one transcription unit (Fig. 3A), suggesting that they may be functionally related. Further analysis revealed that RbfS contains a 23 amino acid signal peptide, a transmembrane domain, a histidine kinase A domain, a histidine kinase-like ATPase domain, a cheY-homologous receiver domain, and a histidine phosphotransfer domain (Fig. 3A). XAC0495 encodes a two-component regulator protein containing an N-terminal receiver domain, a GGDEF domain and an EAL domain at the C-terminal (Fig. 3A). These findings strongly suggest that rbfS and XAC0495 constitute a two-component signaling system. Therefore, for the convenience of discussion and for consistency with rbfS, XAC0495 was designated as rbfR (for regulation of biofilm formation, regulator). Orthologs of the RbfS/RbfR system were found in most Xanthomonas species in the GenBank database and are highly homologous, with >90% amino acid identity (data not shown). The functions of RbfS/RbfR homologs have not yet been determined, with the exception of one rbfR ortholog, XCC0484 in X. campestris pv. campestris strain ATCC33913, which has 92% amino acid identity and has been suggested to be involved in the general stress response of this strain [6].

The XAC0721/bdp2 gene encodes an endonuclease/exonuclease/phosphatase domain-containing protein and is a member of a family of proteins that include Mg-dependent endonucleases and a number of phosphatases putatively involved in intracellular signaling [39]. The product of XAC1778/bdp21 is a MASE1 domain-containing sensor kinase. MASE1-containing sensor kinases were found in various bacteria, including E. coli [40]. Interestingly, the homologous UhpB in E. coli was characterized to be involved in regulating the expression of the UhpT glucose-6-phosphate transporter, and it coordinately functions in sugar transportation [40]. No XAC1778 homolog in Xanthomonas spp. has been characterized, and this needs further research. The XACa0007/bdp16 locus is located on the pXAC33 plasmid [5] and was annotated as a 139 amino acid hypothetical protein with similarity to protein kinase C (PKC) proteins. PKC is a ubiquitous phospholipid-dependent serine/threonine kinase that plays a key role in signal transduction and is involved in the regulation of numerous cellular processes, including a wide variety of biological responses to stimuli [41]. However, its role has not been reported in EPS production or biofilm formation.

In addition, both XAC2018/bdp6 and XAC1499/bdp20 gene products contain a predicted XRE-like HTH DNA-binding domain (Table 1). The XRE DNA-binding protein family is a large family of transcriptional regulators that may act as activators and/or repressors in a variety of bacteria [42]. For example, mqsA in E. coli encodes an antitoxin containing an XRE HTH domain that positively regulates the transcription expression of mqsR (b3022), which is a motility and quorum-sensing regulator involved in biofilm formation [43], [44]. XAC1499/bdp20 was found to be involved in EPS production and cell motility, while XAC2018/bdp6 is only involved in cell motility (Table 1). The putative XAC2018/Bdp6 and XAC1499/Bdp19 proteins exhibited only a slight similarity (12% identity in amino acid sequence), which may explain the functional differences between the two genes. The product of XAC2670/bdp23 was previously annotated as an alginate biosynthesis protein [5]. A BLASTP analysis revealed that this protein was moderately similar (30–40% identity in amino acid sequence) to two-component system sensor proteins in other bacteria, and domain analysis suggested that it contained a signal peptide, three transmembrane domains, and a histidine kinase domain followed by a histidine kinase-like ATPase domain. This evidence strongly suggests that XAC2670/Bdp23 is a two-component system sensor protein. Interestingly, our data suggest that XAC2670/Bdp23 is involved in the regulation of cell motility but not the production of EPS (Table 1). In the Xac strain 306 genome [5], no adjunct cognate response regulator of XAC2670 occurred. The adjacent genes upstream from XAC2670 make up a pilus gene cluster encoding type IV pilus assembly proteins [5]. Taken together, these findings suggest that XAC2670 might play a role in the regulation of expression of genes responsible for type IV pilus biosynthesis and are thus involved in cell motility in Xac strain 306.

Genetic complementation confirms the role of XAC0482/bdp1 and rbfS in biofilm formation

To further validate our results, complementation assays were conducted for the mutants of the XAC0482/bdp1 and rbfS genes. We were able to complement the XAC0482/bdp1 and XAC0494/rbfS mutants with wild-type bdp1 and rbfS, respectively (Fig. 3; Fig. 4). The altered phenotypes, including biofilm formation, EPS production and motility, were restored to levels comparable to those for the wild type (Fig. 3 B; C; D; Fig. 4). These findings further confirmed the linkage between the XAC0482/bdp1 and rbfS genes and Xac biofilm formation.

Discussion

Isolation of Xac mutants defective in biofilm formation

The isolation of a large number of mutants defective in EPS production, flagellum biosynthesis, or motility validates our screening procedure for uncovering mutants because these factors have been implicated in biofilm formation on abiotic and/or biotic surfaces in other model systems [14]. This view has been further verified by the isolation of a set of 7 genes, colR, fhaB, fliC, galU, gumD, wxacO and rbfC, which were previously characterized to be important for Xac biofilm formation [10], [11], [15], [18], [19], [20]. Thus, our procedure will mostly likely validate the 33 novel biofilm-related genes identified in this study. Our approach was not complete, however, because we also missed some genes already shown to be involved in Xac biofilm formation, such as rhsd, XAC3263, XAC3285 and XAC3294 [45]. Twenty-two thousand random mutants represent a roughly fourfold coverage for a _targeted insertion rate of 1 per 1,000 bp (the genome is 5.1×10⁶ bp); therefore, the screen is likely to be saturating. The variability between polystyrene and borosilicate glass and the visual inspection method used for scoring the initial screen might have precluded the capture of subtle biofilm formation deficiencies.

Novel biofilm-related genes uncovered in this work

To the best of our knowledge, XAC1469/bdp3, XAC1509/bdp4, XAC1923/bdp5, XAC2848/bdp7, XAC4024/bdp13 and their homologs have not previously been reported to play a role in biofilm formation in Xanthomonas and other bacteria. Thus, this is the first report demonstrating that these five genes are involved in biofilm formation. XAC1469/Bdp3 is a putative TatD_DNase domain-containing protein similar to members of the TatD-related DNase family (Table 1), which is a family of Mg-dependent DNases that participate in DNA replication, recombination and repair [31]. XAC1509/Bdp4 is a hypothetically cytoplasmic protein of unknown function and is only moderately conserved in X. campestris pv. campestris strain B100 and X. campestris pv. vesicatoria strain 85-10, with 80% and 40% amino acid identity, respectively. XAC1923/Bdp5 is a 334 aa hypothetical protein that contains a 41 aa signal peptide and no other conserved domains, which suggests that it may be a secreted protein. XAC1923/Bdp5 is unique to Xac; there are no significant homologs of XAC1923/Bdp5. The XAC2848/Bdp7 protein has a predicted DUF3616 domain (Pfam entry PF12275). This family of proteins is mainly observed in bacteria, and most have unknown functions. MpeA1724 in Methylibium petroleiphilum PM1, which shares 72% amino acid identity to XAC2848/Bdp7, is a lipocalin-like protein of unknown function [46]. In E. coli, both lipocalin-like proteins, Blc and YodA, were suggested to play a role in the adaptation of cells to certain types of stress, including cadmium stress [47], [48]. However, lipocalin-like proteins have not been reported to be involved in biofilm formation. XAC4024/bdp13 encodes a 119 aa hypothetical protein containing a 27 aa signal peptide and a transmembrane segment at the N- terminal. This protein has a high level of similarity to outer membrane proteins with unknown functions (data not shown). Our results suggest that both XAC1469/Bdp3 and XAC1509/Bdp4 contribute to EPS production and cell motility in Xac and that XAC2848/Bdp7 contributes only to EPS production (Table 1). However, neither XAC1923/Bdp5 nor XAC4024/Bdp13 is associated with EPS production, LPS biosynthesis or cell motility in Xac (Table 1 and data not shown). Overall, this is the first report of a TatD-related DNase, a lipocalin-like protein, the hypothetically cytoplasmic protein XAC1509/Bdp4, the putative secreted protein XAC1923/Bdp5, and the outer membrane protein XAC4024/Bdp13 playing individual roles in biofilm formation.

The first characterization of a HAD-like phosphatase gene in plant pathogenic bacteria

The XAC0482/bdp1 gene encodes a hypothetical protein that possesses significant similarity to phosphatases belonging to the HAD-like hydrolase superfamily. Proteins of this family possess a variety of enzymatic activities and are widespread among the three primary kingdoms of bacteria, archaea, and eukaryotes. In E. coli, HAD-like phosphatases have been suggested to play important roles in carbohydrate metabolism by hydrolyzing the CO-P bond [26]. However, the vast majority of HAD-like hydrolases found in bacteria remain uncharacterized [26]. No information is available on the function of XAC0482 homologs in plant pathogenic bacteria. Our data from genetic complementation assays confirm that XAC0482 is involved in EPS production and subsequent cell motility (Table 1; Fig. 3) but not in LPS biosynthesis (Fig. 2, Lane 3). Thus, this protein may contribute to Xac biofilm formation through promoting the production of EPS and subsequent cell motility via unknown mechanisms. To our knowledge, this is the first analysis of a HAD-like phosphatase in plant pathogenic bacteria.

EPS and LPS are directly involved in Xac biofilm formation

Seven insertions that occurred in the gum gene cluster (gumCDEFJK) required for xanthan EPS production (Supporting Information Table S1; Supporting Information Figure S3) and insertions in 13 bdp genes (bdp1-4, bdp7, bdp14-16, bdp18, bdp20, bdp21, bdp24 and bdp28) that individually contribute to the production of EPS (Table 1) were identified in this study. In addition, an insertion was mapped in the galU gene, which blocked EPS and CPS biosynthesis and led to a dramatic reduction in Xac biofilm formation [11]. Moreover, an additional insertion was found in the XAC0655 locus, and it affected EPS production (Supporting Information Figure S4). The XAC0655 gene was annotated as a sugar kinase [5], and it has a predicted PfkB domain (Pfam entry PF00294) (data not shown). The predicted XAC0655 protein reveals 92% identity in amino acid sequence to the ADKXcc (XC_0690) of X. campestris pv. campestris strain 8004 [38]. Adenosine kinase (ADK), which belongs to the PfkB family of carbohydrate and nucleoside kinases, is a purine salvage enzyme that catalyzes the phosphorylation of adenosine to generate AMP [49]. It was demonstrated that a mutation in adk_Xcc affects EPS production in strain 8004 [38], which is consistent with our observation in the XAC0655/adk_Xac gene mutant (Supporting Information Figure S4). These findings suggest that XAC0655 may contribute to biofilm formation by acting as an ADK in Xac. Therefore, this gene was named adk_Xac. Taken together, these findings are consistent with the idea that EPS is required for structured biofilm formation in a large number of bacteria [14], [16]. Given that X. campestris pv. campestris produces at least two types of EPS in rich media, including xanthan gum and others [50], and that a mutation in gumD abolished Xac production of xanthan [20], we hypothesized that Xac produces at least two types of EPS in rich growth media. To investigate this possibility, we examined the total EPS production of a gumD mutant. The result showed that the gumD mutant, compared with wild-type strain 306, has an 80% reduction in EPS production, but EPS was not abolished when cultured in nutrient broth (NB, Difco; Detroit, IL) supplemented with 2% glucose (Supporting Information Figure S4). This observation suggests that Xac, under the present experimental conditions, may produce at least two types of EPS. The major type of EPS is xanthan, which is encoded by the gum genes, and other types of EPS are encoded by one or more bdp genes. These other types of EPS might be important constituents of the biofilms formed by Xac because xanthan is not the sole polysaccharide implicated in biofilm formation in X. campestris pv. campestris [50].

As one of the major polysaccharide components on cell surfaces of Gram-negative bacteria, LPS has been found to be directly involved in biofilm formation in diverse bacteria, including E. coli O157:H7 [22] and Pseudomonas aeruginosa [33]. We uncovered seven bdp genes associated with LPS synthesis in Xac, including XAC0721/bdp2, XAC1778/bdp21, XAC2848/bdp7, XAC3110/bdp24, XAC3364/bdp8, XAC3576/bdp9, and XAC4203/bdp14 (Fig. 2). We also identified nine EZ-Tn5 insertions in the putative LPS biosynthesis gene cluster in Xac strain 306 (XAC3591, XAC3593, XAC3595, XAC3596/wxacO, XAC3597, XAC3598/rfbC, wzt, wzm and metB) [5] (Supporting Information Figure S3). These findings strongly suggest that LPS is directly involved in Xac biofilm formation; more recently, we have more accurately described the role of LPS in structured Xac biofilm formation [15].

Extracellular proteins and eDNA participate in biofilm formation

The structure and physiology of bacterial biofilms reveal a dramatic diversity of different species and environments. Typically, biofilms contain extracellular matrices consisting of polysaccharides, proteins, membrane vesicles, and extracellular DNA (eDNA) [14], [16]. Many different strains of Xanthomonas spp. and related genera are able to produce biofilms under certain culture conditions [10], [11], [12], [13], [51], [52], but little is known about their composition or structure except that xanthan EPS is commonly a major component of biofilm matrices and that the Xag-type EPS is implicated in biofilm formation by X. campestris pv. campestris [50]. Our most recent report demonstrated that LPS is another essential constituent of the Xac biofilm [15]. Interestingly, Cheng et al. [51] reported that eDNA is associated with biofilm formation in Xylella fastidiosa, a close relative of Xanthomonas spp.. Similarly, we observed a significantly decreased biofilm formation in Xac strain 306 after either DNase I or proteinase K treatment (Supporting Information Figure S5), implying that both eDNA and extracellular proteins are likely to be essential constituents of the Xac biofilm. eDNA can be released by live cells either via membrane vesicles composed of bacterial lipids or from auto-lysed cells [53]. Moreover, this work identified several genes whose products are predicted surface-associated or secreted proteins (fhaB, vacJ and XAC1923/bdp5), outer membrane proteins (iroN and XAC4024), or transporter components (XAC1017 and XAC1459) (Supporting Information Table S1), which at least partly supports the view that both extracellular proteins and eDNA are likely to be essential constituents of Xac biofilms.

Flagellum, type IV pili, bacterial chemotaxis and motility are critical for biofilm formation

Flagella and type IV pili have been implicated in biofilm formation in various bacteria by performing three potentially exclusive roles individually or collectively. They (1) enable planktonic cells to swim toward nutrients associated with a surface or toward signals generated by cells attached to a surface via flagellar-mediated chemotaxis; (2) enable bacteria to overcome repulsive forces and to initially reach a surface and allow attached, dividing bacteria to spread along a surface; and (3) function in a direct manner by physically adhering to a surface [for review, see 14]. In this work, we obtained thirteen different insertions in gene clusters required for flagellum biogenesis (fleN, flgABFKL, flhB, and fliCFMQR), one insertion in the type IV pili biogenesis gene cluster (pilB), four insertions in bacterial chemotaxis-related genes (cheA, cheY, mcpA, and tsr), and two insertions in the motility-related gene, motB (two different copies) (Supporting Information Table S1; Supporting Information Figure S3). Moreover, the filC gene mutant demonstrated severely weakened motility compared to the wild-type strain 306 (Supporting Information Figure S4). Although it is not yet clear which aspect(s) of flagellar structure/function are important in biofilm development for this bacterium, these observations, along with others [19], demonstrate that flagellum- and/or pili-related motility is indeed required for Xac biofilm formation, which is consistent with observations in other model systems [14].

Interestingly, our results show that the 332D5 (XAC0482/bdp1), 411F5 (XAC1509/bdp4), 223G4 (XAC3110/bdp24), 292G4 (XAC3364/bdp8), 291A9 (XAC3591/bdp25), 225E4 (XAC3597/bdp11), 228F10 (XAC4203/bdp14), and 295D8 (XAC4264/bdp15) mutants were not only reduced in motility but were also impaired in EPS and/or LPS production (Table 1). No homologs of these eight genes share significant similarity to any motility-related gene or regulator (Table 1), suggesting that the impaired motility in these mutants may result mainly from their altered EPS and/or LPS production. Indeed, the gumD mutant revealed significantly decreased swimming and swarming motility (Supporting Information Figure S4), which is consistent with the idea that xanthan EPS is required for a flagellum-independent type of motility in Xac [19] and X. campestris pv. campestris [38]. Moreover, we recently observed that altered LPS leads to a reduced swimming phenotype in the wxacO or rfbC gene mutants [15]. However, further characterization is necessary for examining whether this flagellum-independent type of motility contributes to Xac biofilm formation and the exact role that it plays in biofilm formation.

Plasmid genes involved in biofilm formation

In our mutational analyses, we identified three independent insertions in the XACa0007/bdp16, XACb0001/bdp17 and XACb0050/bdp28 genes that are located on the pXAC33 (XACa0007/bdp16) and pXAC64 (XACb0001/bdp17 and XACb0050/bdp28) plasmids. The product of the XACa0007/bdp16 gene is a hypothetical protein with similarity to PKC proteins. PKC is a widespread phospholipid-dependent serine/threonine kinase that plays an important role in signal transduction and is involved in the regulation of numerous cellular processes, such as cellular growth, migration and proliferation, and a wide variety of biological responses to stimuli [41]. In the human bacterial pathogen, E. coli, the PKC-related signal transduction pathway plays a significant role in invasion [41] and is required for bacterial adherence to host cells [54]. Results from this work suggest that XACa0007/bdp16 is associated with the regulation of EPS production (Table 1). The XACb0001/bdp17 gene was annotated as a 68 aa hypothetical protein that shares similarity to Radical S-adenosylmethionine (SAM) domain-containing proteins (data not shown). Radical SAM proteins catalyze a diversity of reactions, including unusual methylation, isomerization, sulfur insertion, ring formation, anaerobic oxidation, and protein radical formation. Additionally, these proteins function as DNA precursors in vitamin, cofactor, antibiotic and herbicide biosynthesis as well as in biodegradation pathways [55]. Our data reveal that XACb0001/Bdp17 is not associated with EPS production or cell motility; thus, the role of XACb0001/Bdp17 in biofilm formation remains unknown. The XACb0050/bdp28 locus encodes an ISxac2 transposase [5]. Previous investigations have demonstrated that bacterial transposases are involved in biofilm formation. The IS256 transposase in Staphylococcus epidermidis influences biofilm formation, either by insertion into regulatory genes or by modulating biofilm gene expression [32]. The results from our study suggest that XACb0050/bdp28 is involved in EPS production (Table 1) and thus contributes to biofilm formation. Both pXAC33 and pXAC64 plasmids are known to be critical for the virulence of Xac because they encode the PthA type III effector protein, which is a required pathogenicity determinant in Xac [1]. The results from the present study indicate that the pXAC33 and pXAC64 plasmids are associated with biofilm formation, suggesting the importance of these plasmids for the adaptation of Xac to adverse environmental conditions.

Multiple signaling and regulation pathways participate in biofilm formation

As indicated above, a total of seven bdp genes, XAC0494/bdp18/rbfS, XAC0721/bdp2, XAC1499/bdp20, XAC1778/bdp21, XAC2018/bdp6, XAC2670/bdp22 and XACa0007/bdp16, were found to be associated with signaling and regulation pathways. In addition to the seven bdp genes, five previously known signaling factors or regulators, including the quorum-sensing RpfC/RpfG system components of RpfC and RpfF, the ColS/ColR two-component system regulator ColR, the global regulator Clp (XAC0483), and the transcriptional activator sigma-54 factor RpoN (XAC1969), were found to be involved in biofilm formation (Supporting Information Table S1). These data show that multiple signaling and regulation pathways are involved in biofilm formation, and these pathways can be induced by various specific extracellular or intracellular signals.

Global regulatory systems, including TCSTSs, have been reported to regulate biofilm formation in a wide variety of bacteria [14], [16]. In addition to the RpfC/RpfG, RbfS/RbfR, and ColS/ColR systems, our work identified an insertion in the XAC1994 locus, which is a homolog of RavS. RavS is the sensor in the RavS/RavR two-component system in X. campestris pv. campestris strain ATCC33913 [56]. The product of XAC1994 was annotated as a HrpX-related protein [5]. Further analyses revealed that XAC1994 contains one transmembrane domain, two PAS domains, one histidine kinase domain, and one histidine kinase-like ATPase domain. Its immediate downstream locus, XAC1992, encodes a cyclic di-GMP (c-di-GMP) phosphodiesterase A that has an N-terminal GGDEF domain, an EAL domain and a receiver domain at the C-terminal (Supporting Information Figure S6). Moreover, the XAC1994 and XAC1992 proteins revealed 90% and 99% identity in amino acid sequences to RavS (Xcc1960) and RavR (Xcc1958) of X. campestris pv. campestris strain ATCC33913, respectively. The RavS/RavR system positively regulates the production of virulence factors, including EPS, through the Clp global regulator in X. campestris pv. campestris strain ATCC33913 in response to limited levels of oxygen [56]. Notably, our study also identified a mutant with an insertion in the clp (XAC0483) locus (Supporting Information Table S1), and phenotype assays showed that both the XAC1994 and clp mutants demonstrated reduced EPS production and cell motility (Supporting Information Figure S4). These findings suggest that XAC1994 and XAC1992 may constitute another two-component system in Xac that regulates biofilm formation through the regulation of EPS production and cell motility via Clp, similar to the process found in RavS/RavR, in X. campestris pv. campestris. Thus, XAC1994 and XAC1992 were renamed as ravS_Xac and ravR_Xac, respectively. In addition, the XAC2670/bdp23 gene encodes a two-component system sensor protein that plays a role in the regulation of cell motility.

We also uncovered five individual signaling factors or regulators that are involved in biofilm formation. The XAC0721/Bdp2 protein is a member of a family of proteins, including Mg-dependent endonucleases and a number of phosphatases putatively involved in intracellular signaling [39]. XAC1778/Bdp21 is a MASE1 domain-containing sensor kinase involved in sugar transportation. XAC2018/Bdp6 and XAC1499/Bdp20 have XRE DNA-binding domains containing transcriptional regulators that are involved in cell motility and EPS production (Table 1), and XACa0007/Bdp16, which is a PKC protein, plays a role in the signal transduction pathway(s) related to EPS production.

Taken together, our results significantly advance our understanding of the regulatory network of biofilm formation in the citrus pathogen, Xac (Fig. 5). This network contains at least four different regulatory systems that are essential for the regulation of biofilm formation. In the diffusible signal factor (DSF)-dependent RpfC/RpfG signaling system, the RpfC/RpfG proteins may play a role similar to their homologs in X. campestris pv. campestris [52], [57], [58], [59]. In Xac, the RpfC sensor kinase interacts with RpfF to control DSF signal generation [60], and upon reaching a threshold concentration (a tightly controlled or balanced amount) in the extracellular environment, a DSF signal is transferred through RpfC to the RpfG response regulator by a conserved phosphorelay mechanism [60]. Activated RpfG modulates the physiological concentrations of the c-di-GMP second messenger via the HD-GYP domain and its interactions with some GGDEF proteins [60]. Relatively higher levels of c-di-GMP promote biofilm formation and inhibit DNA binding by the Clp transcriptional activator [61], but decreases in levels of c-di-GMP promote motility [62], [63], [64] and activate Clp, which directly or indirectly regulates the expression of genes involved in EPS production, extracellular enzyme synthesis, and protein metabolism [58]. These gene products are suggested to be required for mature biofilm formation. Interestingly, it was observed elsewhere that both DSF overproduction and non-production adversely affect the formation of a structured biofilm in X. campestris pv. campestris [52]. One possible reason for this adverse effect is that RpfG negatively regulates the glycosyltransferase system involved in synthesizing polysaccharide adhesins that are essential for biofilm formation [50]. Additionally, RpfG may control another set of genes implicated in biofilm dispersal [57], [65]. Moreover, RpfG may interact with NtrC, a response regulator in the NtrB/NtrC two-component system, and, in this way, contribute to the NtrC-mediated regulation of sigma 54 factor-dependent promoters [60] and subsequent regulation of genes involved in biofilm formation. Another signaling pathway is the RavS_Xac/RavR_Xac two-component regulatory system. The RavS/RavR-mediated low-oxygen tension-sensing pathway in X. campestris pv. campestris positively regulates the production of virulence factors, including EPS, through modulation of the intracellular levels of c-di-GMP and Clp in a manner similar to the DSF-dependent RpfC/RpfG signaling system [56]. Similarly, a limited level of oxygen in the environment, such as in a growing biofilm microcolony, may activate the RavS/RavR regulatory pathway to promote EPS production and thus help biofilm maturation in Xac. A third signaling pathway is the novel RbfS/RbfR regulation pathway. In this system, the RbfS sensor detects certain extracellular or intracellular signals and transfers them to the RbfR response regulator. The RbfR protein contains N-terminal REC and GGDEF domains and a C-terminal EAL domain (Fig. 4A). The activated GGDEF and/or EAL domains interact with c-di-GMP, modulating the intracellular levels of c-di-GMP and Clp and subsequently affecting the expression of genes related to biofilm formation. Additionally, the ColS/ColR two-component system regulates biofilm formation. Our recent work indicated that the ColS/ColR functions in biofilm formation by modulating LPS biosynthesis in response to various environmental stimuli [18]. Taken together, the data suggest cumulative regulation mechanisms whereby Xac may coordinate RpfC/RpfG-dependent quorum-sensing signaling, RavS/RavR-mediated low-oxygen tension sensing and the novel RbfS/RbfR system to regulate biofilm development through modulating intracellular levels of the Clp global regulator and the c-di-GMP second messenger. In addition, the data suggest that there are individual signaling factors or regulators that take part in some unknown signaling pathways to regulate biofilm formation in response to various specific extracellular or intracellular stimuli.

A model for biofilm formation in Xanthomonas

This study also dramatically enhances our understanding of the genetic process of Xanthomonas biofilm formation, as described in Fig. 6. In this model (Fig. 6), bacteria respond to certain extracellular or intracellular signals, such as the quorum-sensing DSF signals and/or nutrient elements, by using flagellum-mediated chemotaxis and motility to move toward, reach and attach to appropriate surfaces with the help of adhesin proteins, such as FhaB. The presence of the FhaB adhesin promotes a stable adherence to the surface. With access to sufficient levels of nutrients and oxygen, genes such as minD, mrdB and nrdF (Supporting Information Table S1), which encode proteins involved in DNA replication, modification, processing and degradation and are linked to adaptation and differentiation, may be activated, and cells can begin to divide on the surface. Additionally, through cell proliferation and division, bacteria spread along the surface through flagellum- and/or type IV pili-mediated chemotaxis/motility, and a monolayer consisting of cells attached to each other emerges and attaches firmly to the surface through the generation of more adhesin proteins. As cells continue to grow on top of one another, a microcolony is formed, and the metabolism of cells near the surface may restrict the diffusion of oxygen and nutrients to the cells at the bottom of the microcolony. Some of these cells may adapt to these changes in local environment, and this leads to phenotypic heterogeneity within this microcolony. The relatively low levels of oxygen and nutrients activate the response regulators of two-component regulatory systems, such as RavS_Xac/RavR_Xac, RbfS/RbfR and ColS/ColR, and consequently, a proportion of cells are activated to express higher levels of EPS, including both xanthan gum and the unknowns. Additionally, LPS, secreted and/or extracellular proteins, and eDNA are all secreted from the cell. These compounds help cells attach to each other and promote the vertical growth of the microcolony. The increased vertical topology may improve access to oxygen and nutrients, which in turn promotes the topological growth of the microcolony and thus the establishment of macrocolonies and a mature biofilm matrix. Indeed, our recent observation of confocal scanning laser micrographs of Xac strain 306 examined the maturation of this bacterium from early-attached cells to the formation of a mature biofilm [15]. Bacterial cells in the biofilm matrix adapt to this new local environment, creating even further heterogeneity within the biofilm. Finally, a proportion of cells or fragments of a microcolony detach and disperse from the biofilm matrix in response to a programmed set of events directed by the DSF signals in a cell-to-cell manner [57, 65, Y Guo and N Wang, unpublished data in this lab) or as a result of physical shear forces.

Conclusions

In conclusion, we report a genome-wide identification of genes related to Xac biofilm formation. We uncovered 92 biofilm-related genes, of which 17 are previously unknown, novel genes and 16 have only putative functions. The putative products of 5 novel genes, including a TatD-related DNase, a lipocalin-like protein, a hypothetically cytoplasmic protein XAC1509/Bdp4, a secreted protein XAC1923/Bdp5, and an outer-membrane protein XAC4024/Bdp13, were described for the first time as being involved in biofilm formation. Furthermore, two of these 92 biofilm-related genes, XAC0482/bdp1 and rbfS, were experimentally confirmed to contribute to biofilm formation. Our findings indicate that EPS, LPS and flagellum-dependent and flagellum-independent motility are required for the formation of a mature biofilm. Additionally, extracellular proteins and eDNA also contribute to biofilm formation. Based on our data, and in conjunction with others, a comprehensive developmental process and a global regulatory network of Xac biofilm formation has been postulated (Figs. 5 & 6). This work is the first report on a genome-wide scale of the genetic processes of biofilm formation in a plant pathogenic bacterium, and it provides new insight into the genetic determinants and regulation mechanisms of biofilm formation. Further analyses of these biofilm formation-related genes and the pathways involved will provide a better understanding of the process of biofilm formation and adaptation in Xac. This, in turn, should aid in the development of more effective control strategies for citrus canker.

Materials and Methods

Bacterial strains, growth conditions and plasmids

The key strains and plasmids used in this work are listed in Supporting Information Table S2. The E. coli strains were cultured at 37°C in Luria-Bertani (LB) medium. Wild-type Xac strain 306 (rifamycin resistant) [66] and mutant strains were grown at 28°C in nutrient broth (NB, Difco; Detroit, IL) or on nutrient agar (NA, Difco; Detroit, IL). When necessary, antibiotics were added at the following concentrations in growth media: ampicillin (Ap), 50 µg/ml; chloramphenicol (Cm), 35 µg/ml; gentamicin (Gm), 5 µg/ml; kanamycin (Km), 50 µg/ml; or rifamycin (Rif), 50 µg/ml.

Mutant generation and screening for biofilm-defective phenotype

A transposon mutant library of Xac strain 306, which contains 22,000 clones, was previously constructed using the EZ-Tn5™ <R6Kγori/KAN-2>Tnp Transposome™ Kit (Epicentre, Madison, WI) [11]. For rapid initial screening, a polystyrene 96-well plate assay was performed as described previously [15]. Candidate bdp mutants that passed the initial screening were subsequently subjected to a stringent screening using a quantitative biofilm assay in borosilicate glass tubes as described by Guo et al. [11]. The test was repeated three times independently. Mutants that revealed biofilm-defective phenotypes were selected for further characterization.

EZ-Tn5 transposon copy number determination

Genomic DNA was extracted from strains of Xac using a Wizard genomic DNA purification kit according to the manufacturer's protocols (Promega, Madison, WI), digested with SacI (which has no _target site in the EZ-Tn5 transposon), subjected to electrophoresis on a 0.8% agarose gel, and transferred to a positively charged nylon membrane (Roche, Indianapolis, IN) using standard procedures [67]. A DNA fragment of the kanamycin-resistance gene of the EZ-Tn5 transposon was amplified with primers Kan-1 and Kan-2 (Supporting Information Table S3) and used as the probe. Probe labeling, hybridization, and immunological detection were performed using a digoxigenin (DIG)-High Prime II DNA labeling and detection starter kit following the manufacturer's instructions (Roche, Indianapolis, IN).

Identification of the EZ-Tn5 flanking sequence

The genomic sequences flanking the EZ-Tn5 transposon in bdp mutants were analyzed using the random amplification of transposon ends (RATE) PCR method [23]. Briefly, mutant genomic DNA was used as a template in a three-step PCR reaction with the Inv-1 or Inv-2 primer (Supporting Information Table S3). The first 30 cycles of the PCR reaction were performed at 55°C with a 30-second extension. The second 30 cycles were conducted at 30°C with a 30-second extension. The last 30 cycles of the amplification were conducted at 55°C with a 2-minute extension. The RATE-PCR products were then sequenced using the forward (KAN-2 FP-1) or reverse (KAN-2 RP-1) primers (Supporting Information Table S3) supplied with the EZ-Tn5™ <R6Kγori/KAN-2> Insertion Kit and analyzed using BLASTN against the Xac strain 306 genome sequence [5].

Construction of plasmids for genetic complementation

Plasmids used for genetic complementation were constructed as follows. The corresponding primers (Supporting Information Table S3) were designed to incorporate upstream BamHI and downstream HindIII restriction sites flanking the coding sequence and were used to amplify XAC0482 (bdp1) and XAC0494 (rbfS) from Xac wild-type strain 306 genomic DNA. The PCR product was cloned into a pGEM®-T Easy vector following the manufacturer's instructions (Promega, Madison, WI). Following this, the BamHI-HindIII fragment was isolated and ligated into the pUFR053 complementary vector [68] to construct the pUF-0482 and pUF-rbfS plasmids (Supporting Information Table S2) for genetic complementation. Plasmids were introduced into the appropriate biofilm mutants by triparental mating as described previously [15].

EPS quantitative determination

To estimate total EPS production, bacterial strains were cultured in 50 ml of NB liquid medium containing 2% (wt/vol) glucose at 28°C with shaking at 200 rpm for 24 hours. EPS was precipitated from the culture supernatant with ethanol, dried, and weighed, as described previously [50]. The experiment was repeated three times with three replicates.

Cell motility assay

To test cell motility, 2 µL of overnight cultures in NB (OD600 of 1.5) of each Xac strain was spotted onto NA plates containing 0.3% (wt/vol) agar (Difco, Franklin Lakes, NJ) for the swimming motility assay or 0.7% (wt/vol) agar for the swarming motility assay. Plates were incubated at room temperature (22–23°C) for 7 days. The diameters of the areas occupied by the strains were measured, and the values were used to indicate the motility of Xac strains. The experiment was repeated three times with three replicates.

LPS analysis

Xac strains were cultured overnight at 28°C in NB liquid medium with shaking at 200 rpm. Five mL of cultures at the exponential stage was collected, and the LPS samples were extracted as previously described [15]. LPS was separated using sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) and visualized using silver staining following the manufacturer's instructions (Bio-Rad Laboratories, Inc., Hercules, CA). Standard LPS from Salmonella enterica serovar Typhimurium was obtained from Sigma. The test was independently performed three times.

Supporting Information

Figure S1.

Southern hybridization analysis of genomic DNA extracted from biofilm-defective mutant strains of Xanthomonas axonopodis pv. citri (Xac). M: size marker, 1 Kb plus (Promega); Lane 1: wild-type strain 306; Lanes 2–18: mutant strains; Lane 12: double-transposon insertion mutant. The genomic DNA of Xac strains was digested with SacI, blotted onto a nylon membrane and hybridized with a kanamycin-resistant gene as the probe.

doi:10.1371/journal.pone.0021804.s001

(TIF)

Figure S2.

Distribution of biofilm-formation-related genes identified in this work in the chromosome genome and plasmids of Xanthomonas axonopodis pv. citri strain 306.

doi:10.1371/journal.pone.0021804.s002

(TIF)

Figure S3.

Genetic organization of biofilm-formation-related gene clusters in Xanthomonas axonopodis pv. citri strain 306. (a) Gum genes cluster, (b) LPS biosynthesis genes cluster, (c) EPS and LPS precursors biosynthesis gene cluster, (d) hemagglutinin coding genes cluster, (e) Chemotaxis/flagellum/motility genes cluster, and (f) pili gene cluster. Black triangles indicate insertion sites of the EZ-Tn5 transposon.

doi:10.1371/journal.pone.0021804.s003

(PDF)

Figure S4.

Assays for EPS production (A) and cell motility (B) of biofilm-defective mutants of Xanthomonas axonopodis pv. citri strain 306. For the EPS assay, a modified ethanol deposit method was applied. For the motility test, bacterial strains were inoculated at a central point on NA plates (0.3% agar for swimming assays and 0.7% agar for swarming assays) and incubated at room temperature (approximately 23°C) for 7 days, after which colony diameters were measured. All experiments were repeated three times with three replicates. Averages and standard errors from one of three representative experiments with similar results are presented. Wild-type strain 306 was set as equal to 100%. Significance was tested by Student's t-test (* indicates significant difference from wild-type strain 306 at P<0.01). Wt306: wild type strain 306; 294E2(XAC0144/iroN): EZ-Tn5 insertion in XAC0144; 332D5(XAC0482/bdp1): EZ-Tn5 insertion in XAC0482; 283C5(XAC0483/clp); EZ-Tn5 insertion in XAC0483; 301B4(XAC0494/bdp18/rbfS): EZ-Tn5 insertion in XAC0494; 421F4(XAC0655/adk): EZ-Tn5 insertion in XAC0655; 285E8(XAC0721/bdp2): EZ-Tn5 insertion in XAC0721; 294E4(XAC0929/bdp19): EZ-Tn5 insertion in XAC0929; 270H4(XAC1469/bdp3): EZ-Tn5 insertion in XAC1469; 319B1(XAC1499/bdp20): EZ-Tn5 insertion in XAC1499; 411F5(XAC1509/bdp4): EZ-Tn5 insertion in XAC1509; 296G1 (XAC1778/bdp21): EZ-Tn5 insertion in XAC1778; 288D11(XAC1923/bdp5): EZ-Tn5 insertion in XAC1923; 212C9(XAC1994/ravS):EZ-Tn5 insertion in XAC1994; 290E7(XAC2018/bdp6): EZ-Tn5 insertion in XAC2018; 295H3(XAC2848/bdp7): EZ-Tn5 insertion in XAC2848; 291G5(XAC2670/bdp23): EZ-Tn5 insertion in XAC2670; 223G4(XAC3110/bdp24): EZ-Tn5 insertion in XAC3110; 292G4(XAC3364/bdp8): EZ-Tn5 insertion in XAC3364; 207A6(XAC3576/bdp9): EZ-Tn5 insertion in XAC3576; 291A9(XAC3591/bdp25): EZ-Tn5 insertion in XAC3591; 225E4(XAC3597/bdp11): EZ-Tn5 insertion in XAC3597; 409D8(XAC4024/bdp13): EZ-Tn5 insertion in XAC4024; 228F10(XAC4203/bdp14): EZ-Tn5 insertion in XAC4203; 302B11(XAC4249/xynA): EZ-Tn5 insertion in XAC4249; 295D8 (XAC4264/bdp15): EZ-Tn5 insertion in XAC4264; 347D10(XAC4344/vacJ): EZ-Tn5 insertion in XAC4344; 288C6(XACa0007/bdp16): EZ-Tn5 insertion in XACa0007; 423H3(XACb0001/bdp17): EZ-Tn5 insertion in XACb0001; 419F10(XACb0050/bdp28): EZ-Tn5 insertion in XACb0050; 276B8(XAC1975/fliC): EZ-Tn5 insertion in XAC1975/fliC, and 257H6(XAC2583/gumD): EZ-Tn5 insertion in XAC2583/gumD.

doi:10.1371/journal.pone.0021804.s004

(PDF)

Figure S5.

DNase I and Proteinase K effect on bifilm formation by Xanthomonas axonopodis pv. citri strain 306. Biofilm assay showing the repartition of the strains with different treatment. W, Wild-type strain 306; W+D, Wild-type strain 306 with DNase I (20 µg/mL Sigma, St. Louis, MO); W+P, Wild-type strain 306 with Proteinase K (10 ug/ml; Sigma); CK-, NB medium without bacteria inoculation. The experiment was repeated three times with eight replicates each time. Averages and standard errors from one representative experiment of three with similar results are presented. Wild-type strain 306 was set equal to 100%. Significance was tested by Student's t test (* indicates significant difference from wild-type strain 306 at P<0.01). The inset shows a typical polystyrene microtiter dish assay plate stained with crystal violet.

doi:10.1371/journal.pone.0021804.s005

(TIF)

Figure S6.

Genetic organization and domain structures of the RavS_xac/RavR_xac system in Xanthomonas axonopodis pv. citri strain 306. RavS_xac is a sensor kinase and RavR_xac is the response regulator. Gene orientation is indicated by arrow. Domain structure prediction was done using the SMART program (http://smart.embl-heidelberg.de). Domain Symbol: T, transmembrane domain; PAS, PAS domain; HKA, histidine kinase A domain; HATP, histidine kinase-like ATPase domain; GGDEF, GGDEF domain; EAL, EAL domain; REC, receiver domain.

doi:10.1371/journal.pone.0021804.s006

(TIF)

Table S1.

Biofilm related genes identified from Xanthomonas axonopodis pv. citri strain 306 in this study.

doi:10.1371/journal.pone.0021804.s007

(DOC)

Table S2.

Bacterial strains and plasmids used in this study^a.

doi:10.1371/journal.pone.0021804.s008

(DOC)

Table S3.

Primers used in this study.

doi:10.1371/journal.pone.0021804.s009

(DOC)

Acknowledgments

We thank Lin Yang for assistance in this work.

Author Contributions

Conceived and designed the experiments: JL NW. Performed the experiments: JL NW. Analyzed the data: JL NW. Contributed reagents/materials/analysis tools: JL NW. Wrote the paper: JL NW.

References

1. Boch J,Bonas U (2010) Xanthomonas AvrBs3 Family-Type III Effectors: Discovery and Function. Annu Rev Phytopathol 48: 419–436.
2. Cubero J,Graham JH (2002) Genetic relationship among worldwide strains of Xanthomonas causing canker in citrus species and design of new primers for their identification by PCR. Appl Environ Microbiol 68: 1257–1264.
3. Schaad N,Postnikova E,Lacy G,Sechler A,Agarkova I,et al. (2006) Emended classification of xanthomonad pathogens on citrus. Syst Appl Microbiol 29: 690–695.
4. Vauterin L,Hoste B,Kersters K,Swings J (1995) Reclassification of Xanthomonas. Int J Syst Bacteriol 45: 472–489.
5. da Silva AC,Ferro JA,Reinach FC,Farah CS,Furlan LR,et al. (2002) Comparison of the genomes of two Xanthomonas pathogens with differing host specificities. Nature 417: 459–463.
6. Qian W,Han Z,Tao J,He C (2008) Genome-scale mutagenesis and phenotypic characterization of two-component signal transduction systems in Xanthomonas campestris pv. campestris ATCC 33913. Mol Plant-Microbe Interact 21: 1128–1138.
7. Salzberg SL,Sommer DD,Schatz MC,Phillippy AM,Rabinowicz PD,et al. (2008) Genome sequence and rapid evolution of the rice pathogen Xanthomonas oryzae pv. oryzae PXO99^A. BMC Genomics 9: 534.
8. Gottwald TR,Graham JH,Schubert TS (2002) Citrus canker: the pathogen and its impact. Plant Health Progress Online. doi:10.1094/PHP-2002-0812-01-RV.
9. Graham JH,Gottwald TR,Cubero J,Achor DS (2004) Xanthomonas axonopodis pv. citri: factors affecting successful eradication of citrus canker. Mol Plant Pathol 5: 1–15.
10. Gottig N,Garavaglia BS,Garofalo CG,Orellano EG,Ottado J (2009) A filamentous hemagglutinin-like protein of Xanthomonas axonopodis pv. citri, the phytopathogen responsible for citrus canker, is involved in bacterial virulence. PLoS ONE 4: e4358.
11. Guo Y,Sagaram US,Kim JS,Wang N (2010) Requirement of the galU gene for polysaccharide production by and pathogenicity and growth In planta of Xanthomonas citri subsp. citri. Appl Environ Microbio 76: 2234–2242.
12. Rigano LA,Siciliano F,Enrique R,Sendin L,Filippone P,et al. (2007) Biofilm formation, epiphytic fitness, and canker development in Xanthomonas axonopodis pv. citri. Mol. Plant-Microbe Interact 20: 1222–1230.
13. Danhorn T,Fuqua C (2007) Biofilm formation by plant-associated bacteria. Annu Rev Microbiol 61: 401–22.
14. Karatan E,Watnick P (2009) Signals, regulatory networks, and materials that build and break bacterial biofilms. Microbiol Mol Biol Rev 73: 310–347.
15. Li J,Wang N (2011) The wxacO gene of Xanthomonas citri ssp. citri encodes a protein with a role in lipopolysaccharide biosynthesis, biofilm formation, stress tolerance and virulence. Mol Plant Pathol 12: 381–396.
16. Flemming H,Wingender J (2010) The biofilm matrix. Nature Rev 8: 623–633.
17. Moreira LM,Almeida NF,Potnis N,Digiampietri LA,Adi SS,et al. (2010) Novel insights into the genomic basis of citrus canker based on the genome sequences of two strains of Xanthomonas fuscans subsp. aurantifolii. BMC Genomics 11: 238.
18. Yan Q,Wang N (2011) The ColR/ColS two-component system plays multiple roles in the pathogenicity of the citrus canker pathogen Xanthomonas citri subsp. citri. J Bacteriol 193: 1590–1599.
19. Malamud F,Torres PS,Roxana R,Rigano LA,Enrique R,et al. (2011) Xanthomonas axonopodis pv. citri flagellum is required for mature biofilm and canker development. Microbiology 157: 819–829.
20. Dunger G,Relling VM,Tondo ML,Barreras M,Ielpi L,et al. (2007) Xanthan is not essential for pathogenicity in citrus canker but contributes to Xanthomonas epiphytic survival. Arch Microbiol 88: 127–135.
21. Tu Quoc PH,Genevaux P,Pajunen M,Savilahti H,Georgopoulos C,et al. (2007) Isolation and characterization of biofilm formation-defective mutants of Staphylococcus aureus. Infect Immun 75: 1079–1088.
22. Puttamreddy S,Cornick NA,Minion FC (2010) Genome-wide transposon mutagenesis reveals a role for pO157 genes in biofilm development in Escherichia coli O157:H7 EDL933 Infect Immun 78: 2377–2384.
23. Ducey TF,Dyer DW (2002) Rapid identification of EZ::TN™transposon insertion sites in the genome of Neisseria gonorrhoeae. EPICENTRE Forum 9: 6–7.
24. Caspi R,Altman T,Dale JM,Dreher K,Fulcher CA,et al. (2010) The MetaCyc database of metabolic pathways and enzymes and the BioCyc collection of pathway/genome databases. Nucleic Acids Res 38(Database issue): D473–D479.
25. Kanehisa M,Goto S,Furumichi M,Tanabe M,Hirakawa M (2010) KEGG for representation and analysis of molecular networks involving diseases and drugs. Nucleic Acids Res 38(Database issue): D355–D360.
26. Kuznetsova E,Proudfoot M,Gonzalez CF,Brown G,Omelchenko M,et al. (2006) Genome-wide analysis of substrate specificities of the Escherichia coli haloacid dehalogenase-like phosphatase family. J Biol Chem 281: 36149–36161.
27. Breton C,Snajdrova L,Jeanneau C,Koca J,Imberty A (2006) Structures and mechanisms of glycosyltransferases. Glycobiology 16: 29R–37R.
28. Kube M,Migdoll AM,Muller I,Kuhl H,Beck A,et al. (2008) The genome of Erwinia tasmaniensis strain Et1/99, a non-pathogenic bacterium in the genus Erwinia. Environ Microbiol 10: 2211–2222.
29. Wu L,Birch RG (2005) Characterization of the highly efficient sucrose isomerase from Pantoea dispersa UQ68J and cloning of the sucrose isomerase gene. Appl Environ Microbiol 71: 1581–1590.
30. Alegria MC,Souza DP,Andrade MO,Docena C,Khater L,et al. (2005) Identification of new protein-protein interactions involving the products of the chromosome- and plasmid-encoded type IV secretion loci of the phytopathogen Xanthomonas axonopodis pv. citri. J Bacteriol 187: 2315–2325.
31. Wexler M,Sargent F,Jack RL,Stanley NR,Bogsch EG,et al. (2000) TatD is a cytoplasmic protein with DNase activity. No requirement for TatD family proteins in Sec-independent protein export. J Biol Chem 275: 16717–16722.
32. Hennig S,Ziebuhr W (2010) Characterization of the transposase encoded by IS256, the prototype of a major family of bacterial insertion sequence elements. J Bacteriol 192: 4153–4163.
33. Lau PCY,Lindhout T,Beveridge TJ,Dutcher JR,Lam JS (2009) Differential lipopolysaccharide core capping leads to quantitative and correlated modifications of mechanical and structural properties in Pseudomonas aeruginosa biofilms. J Bacteriol 191: 6618–6631.
34. Mack M,Buckel W (1997) Conversion of glutaconate CoA-transferase from Acidaminococcus fermentans into an acyl-CoA hydrolase by site-directed mutagenesis. FEBS Letters 405: 209–212.
35. Marin M,Perez-Pantoja D,Donoso R,Wray V,Gonzalez B,et al. (2010) Modified 3-oxoadipate pathway for the biodegradation of methylaromatics in Pseudomonas reinekei MT1. J Bacteriol 192: 1543–1552.
36. Yamashita Y,Tsukioka Y,Tomihisa K,Nakano Y,Koga T (1998) Genes involved in cell wall localization and side chain formation of rhamnose-glucose polysaccharide in Streptococcus mutans. J Bacteriol 180: 5803–5807.
37. Wang L,Makino S,Subedee A,Bogdanove AJ (2007) Novel candidate virulence factors in rice pathogen Xanthomonas oryzae pv. oryzicola as revealed by mutational analysis. Appl Environ Microbiol 73: 8023–8027.
38. Lu G,Tang Y,Li C,Li R,An S,et al. (2009) An adenosine kinase exists in Xanthomonas campestris pathovar campestris and is involved in extracellular polysaccharide production, cell motility, and Virulence. J Bacteriol 191: 3639–3648.
39. Finn RD,Mistry J,Schuster-Bockler B,Griffiths-Jones S,Hollich V,et al. (2006) Pfam: clans, web tools and services. Nucleic Acids Res 34: D247–D251.
40. Mascher T,Helmann JD,Unden G (2006) Stimulus perception in bacterial signal-transducing histidine kinases. Microbiol Mol Biol Rev 70: 910–938.
41. Sukumaran SK,Prasadarao NV (2002) Regulation of protein kinase C in Escherichia coli K1 invasion of human mrain microvascular endothelial cells. J Biol Chem 277: 12253–12262.
42. Gerstmeir R,Cramer A,Dangel P,Schaffer S,Eikmanns BJ (2004) RamB, a novel transcriptional regulator of genes involved in acetate metabolism of Corynebacterium glutamicum. J Bacteriol 186: 2798–2809.
43. Gonzalez-Barrios AF,Zuo R,Hashimoto Y,Yang L,Bentley W E,et al. (2006) Autoinducer 2 controls biofilm formation in Escherichia coli through a novel motility quorum-sensing regulator (MqsR, B3022). J Bacteriol 188: 305–316.
44. Brown BL,Wood TK,Peti W,Page R (2011) Structure of the Escherichia coli antitoxin MqsA (YgiT/b3021) bound to its gene promoter reveals extensive domain rearrangements and the specificity of transcriptional regulation. J Biol Chem 286: 2285–2296.
45. Laia ML,Moreira LM,Dezajacomo J,Brigati JB,Ferreira CB,et al. (2009) New genes of Xanthomonas citri subsp. citri involved in pathogenesis and adaptation revealed by a transposonbased mutant library. BMC Microbiol 9: 12.
46. Kane SR,Chakicherla AY,Chain PS,Schmidt R,Shin MW,et al. (2007) Whole-genome analysis of the methyl tertbutyl ether-degrading beta-proteobacterium Methylibium petroleiphilum PM1. J Bacteriol 189: 1931–45.
47. Gabriel D,Karine B,Marc S,Simon P,Anita L (2003) YodA from Escherichia coli is a metal-binding, lipocalin-like Protein. J Biol Chem 278: 43728–43735.
48. Stojnev T,Harichová J,Ferianc P,Nyström T (2007) Function of a novel cadmium-induced YodA protein in Escherichia coli. Curr Microbiol 55: 99–104.
49. Spychala J,Datta NS,Takabayashi K,Datta M,Fox IH,et al. (1996) Cloning of human adenosine kinase cDNA: sequence similarity to microbial ribokinases and fructokinases. Proc Natl Acad Sci USA 93: 1232–1237.
50. Tao F,Swarup F,Zhang LH (2010) Quorum sensing modulation of a putative glycosyltransferase gene cluster essential for Xanthomonas campestris biofilm formation. Environ Microbiol 12: 3159–3170.
51. Cheng DW,Lin H,Civerolo EL (2010) Extracellular genomic DNA mediates enhancement of Xylella fastidiosa biofilm formation in Vitro. J Plant Pathol 92: 405–410.
52. Torres PS,Malamud F,Rigano LA,Russo DM,Marano MR,et al. (2007) Controlled synthesis of the DSF cell-cell signal is required for biofilm formation and virulence in Xanthomonas campestris. Environ Microbiol 9: 2101–2109.
53. Rice KC,Mann EE,Endres JL,Weiss EC,Cassat JE,et al. (2007) The cidA murein hydrolase regulator contributes to DNA release and biofilm development in Staphylococcus aureus. Proc Natl Acad Sci USA 104: 8113–8118.
54. Rosenshine I,Ruschkowski S,Stein M,Reinscheid DJ,Mills SD,et al. (1996) A pathogenic bacterium triggers epithelial signals to form a functional bacterial receptor that mediates actin pseudopod formation. EMBO J 15: 2613–2624.
55. Sofia HJ,Chen G,Hetzler BG,Reyes-Spindola JF,Miller NE (2001) Radical SAM, a novel protein superfamily linking unresolved steps in familiar biosynthetic pathways with radical mechanisms: functional characterization using new analysis and information visualization methods. Nucl Acids Res 29: 1097–1106.
56. He Y,Boon C,Zhou L,Zhang L (2009) Co-regulation of Xanthomonas campestris virulence by quorum sensing and a novel two-component regulatory system RavS/RavR. Mol Microbiol 71: 1464–1476.
57. He Y,Xu M,Lin K,Ng Y,Wen C,et al. (2006) Genome scale analysis of diffusible signal factor regulon in Xanthomonas campestris pv. campestris: identification of novel cell-cell communication-dependent genes and functions. Mol Microbiol 59: 610–622.
58. He Y,Ng Y,Xu M,Lin K,Wang L,et al. (2007) Xanthomonas campestris cell-cell communication involves a putative nucleotide receptor protein Clp and a hierarchical signaling network. Mol Microbiol 64: 281–292.
59. Ryan RP,McCarthy Y,Andrade M,Farah CS,Armitage JP,et al. (2010) Cell–cell signal dependent dynamic interactions between HD-GYP and GGDEF domain proteins mediate virulence in Xanthomonas campestris. Proc Natl Acad Sci USA 107: 5989–5994.
60. Andrade MO,Alegria MC,Guzzo CR,Docena C,Pareda-Rosa MC,et al. (2006) The HD-GYP domain of RpfG mediates a direct linkage between the Rpf quorum-sensing pathway and a subset of diguanylate cyclase proteins in the phytopathogen Xanthomonas axonopodis pv. citri. Mol Microbiol 62: 537–551.
61. Leduc JL,Roberts GP (2009) Cyclic di-GMP allosterically inhibits the CRP-like protein (Clp) of Xanthomonas axonopodis pv.citri. J Bacteriol 191: 7121–7122.
62. Simm R,Morr M,Kader A,Nimtz M,Römling U (2004) GGDEF and EAL domains inversely regulate cyclic di-GMP levels and transition from sessility to motility. Mol Microbiol 53: 1123–1134.
63. Tischler AD,Camilli A (2004) Cyclic diguanylate (c-di-GMP) regulates Vibrio cholera 3 biofilm formation. Mol Microbiol 53: 857–869.
64. Römling U,Gomelsky M,Galperin MY (2005) C-di-GMP: the dawning of a novel bacterial signalling system. Mol Microbiol 57: 629–639.
65. Dow JM,Crossman L,Findlay K,He YQ,Feng JX,et al. (2003) Biofilm dispersal in Xanthomonas campestris is controlled by cell–cell signaling and is required for full virulence to plants. Proc Natl Acad Sci USA 100: 10995–11000.
66. Rybak M,Minsavage GV,Stall E,Jones JB (2009) Identification of Xanthomonas citri subsp. citri host specificity genes in a heterologous expression host. Mol Plant Pathol 10: 249–262.
67. Sambrook J,Russell DW (2001) Molecular cloning: a laboratory manual, 3rd ed. NY: Cold Spring Harbor Laboratory Press. pp. 633–638.
68. El Yacoubi B,Brunings A,Yuan Q,Shankar S,Gabriel D (2007) In planta horizontal transfer of a major pathogenicity effector gene. Appl Environ Microbiol 73: 1612–1621.

[1] 1. Boch J,Bonas U (2010) Xanthomonas AvrBs3 Family-Type III Effectors: Discovery and Function. Annu Rev Phytopathol 48: 419–436.

[2] 2. Cubero J,Graham JH (2002) Genetic relationship among worldwide strains of Xanthomonas causing canker in citrus species and design of new primers for their identification by PCR. Appl Environ Microbiol 68: 1257–1264.

[3] 3. Schaad N,Postnikova E,Lacy G,Sechler A,Agarkova I,et al. (2006) Emended classification of xanthomonad pathogens on citrus. Syst Appl Microbiol 29: 690–695.

[4] 4. Vauterin L,Hoste B,Kersters K,Swings J (1995) Reclassification of Xanthomonas. Int J Syst Bacteriol 45: 472–489.

[5] 5. da Silva AC,Ferro JA,Reinach FC,Farah CS,Furlan LR,et al. (2002) Comparison of the genomes of two Xanthomonas pathogens with differing host specificities. Nature 417: 459–463.

[6] 6. Qian W,Han Z,Tao J,He C (2008) Genome-scale mutagenesis and phenotypic characterization of two-component signal transduction systems in Xanthomonas campestris pv. campestris ATCC 33913. Mol Plant-Microbe Interact 21: 1128–1138.

[7] 7. Salzberg SL,Sommer DD,Schatz MC,Phillippy AM,Rabinowicz PD,et al. (2008) Genome sequence and rapid evolution of the rice pathogen Xanthomonas oryzae pv. oryzae PXO99^A. BMC Genomics 9: 534.

[8] 8. Gottwald TR,Graham JH,Schubert TS (2002) Citrus canker: the pathogen and its impact. Plant Health Progress Online. doi:10.1094/PHP-2002-0812-01-RV.

[9] 9. Graham JH,Gottwald TR,Cubero J,Achor DS (2004) Xanthomonas axonopodis pv. citri: factors affecting successful eradication of citrus canker. Mol Plant Pathol 5: 1–15.

[10] 10. Gottig N,Garavaglia BS,Garofalo CG,Orellano EG,Ottado J (2009) A filamentous hemagglutinin-like protein of Xanthomonas axonopodis pv. citri, the phytopathogen responsible for citrus canker, is involved in bacterial virulence. PLoS ONE 4: e4358.

[11] 11. Guo Y,Sagaram US,Kim JS,Wang N (2010) Requirement of the galU gene for polysaccharide production by and pathogenicity and growth In planta of Xanthomonas citri subsp. citri. Appl Environ Microbio 76: 2234–2242.

[12] 12. Rigano LA,Siciliano F,Enrique R,Sendin L,Filippone P,et al. (2007) Biofilm formation, epiphytic fitness, and canker development in Xanthomonas axonopodis pv. citri. Mol. Plant-Microbe Interact 20: 1222–1230.

[13] 13. Danhorn T,Fuqua C (2007) Biofilm formation by plant-associated bacteria. Annu Rev Microbiol 61: 401–22.

[14] 14. Karatan E,Watnick P (2009) Signals, regulatory networks, and materials that build and break bacterial biofilms. Microbiol Mol Biol Rev 73: 310–347.

[15] 15. Li J,Wang N (2011) The wxacO gene of Xanthomonas citri ssp. citri encodes a protein with a role in lipopolysaccharide biosynthesis, biofilm formation, stress tolerance and virulence. Mol Plant Pathol 12: 381–396.

[16] 16. Flemming H,Wingender J (2010) The biofilm matrix. Nature Rev 8: 623–633.

[17] 17. Moreira LM,Almeida NF,Potnis N,Digiampietri LA,Adi SS,et al. (2010) Novel insights into the genomic basis of citrus canker based on the genome sequences of two strains of Xanthomonas fuscans subsp. aurantifolii. BMC Genomics 11: 238.

[18] 18. Yan Q,Wang N (2011) The ColR/ColS two-component system plays multiple roles in the pathogenicity of the citrus canker pathogen Xanthomonas citri subsp. citri. J Bacteriol 193: 1590–1599.

[19] 19. Malamud F,Torres PS,Roxana R,Rigano LA,Enrique R,et al. (2011) Xanthomonas axonopodis pv. citri flagellum is required for mature biofilm and canker development. Microbiology 157: 819–829.

[20] 20. Dunger G,Relling VM,Tondo ML,Barreras M,Ielpi L,et al. (2007) Xanthan is not essential for pathogenicity in citrus canker but contributes to Xanthomonas epiphytic survival. Arch Microbiol 88: 127–135.

[21] 21. Tu Quoc PH,Genevaux P,Pajunen M,Savilahti H,Georgopoulos C,et al. (2007) Isolation and characterization of biofilm formation-defective mutants of Staphylococcus aureus. Infect Immun 75: 1079–1088.

[22] 22. Puttamreddy S,Cornick NA,Minion FC (2010) Genome-wide transposon mutagenesis reveals a role for pO157 genes in biofilm development in Escherichia coli O157:H7 EDL933 Infect Immun 78: 2377–2384.

[23] 23. Ducey TF,Dyer DW (2002) Rapid identification of EZ::TN™transposon insertion sites in the genome of Neisseria gonorrhoeae. EPICENTRE Forum 9: 6–7.

[24] 24. Caspi R,Altman T,Dale JM,Dreher K,Fulcher CA,et al. (2010) The MetaCyc database of metabolic pathways and enzymes and the BioCyc collection of pathway/genome databases. Nucleic Acids Res 38(Database issue): D473–D479.

[25] 25. Kanehisa M,Goto S,Furumichi M,Tanabe M,Hirakawa M (2010) KEGG for representation and analysis of molecular networks involving diseases and drugs. Nucleic Acids Res 38(Database issue): D355–D360.

[26] 26. Kuznetsova E,Proudfoot M,Gonzalez CF,Brown G,Omelchenko M,et al. (2006) Genome-wide analysis of substrate specificities of the Escherichia coli haloacid dehalogenase-like phosphatase family. J Biol Chem 281: 36149–36161.

[27] 27. Breton C,Snajdrova L,Jeanneau C,Koca J,Imberty A (2006) Structures and mechanisms of glycosyltransferases. Glycobiology 16: 29R–37R.

[28] 28. Kube M,Migdoll AM,Muller I,Kuhl H,Beck A,et al. (2008) The genome of Erwinia tasmaniensis strain Et1/99, a non-pathogenic bacterium in the genus Erwinia. Environ Microbiol 10: 2211–2222.

[29] 29. Wu L,Birch RG (2005) Characterization of the highly efficient sucrose isomerase from Pantoea dispersa UQ68J and cloning of the sucrose isomerase gene. Appl Environ Microbiol 71: 1581–1590.

[30] 30. Alegria MC,Souza DP,Andrade MO,Docena C,Khater L,et al. (2005) Identification of new protein-protein interactions involving the products of the chromosome- and plasmid-encoded type IV secretion loci of the phytopathogen Xanthomonas axonopodis pv. citri. J Bacteriol 187: 2315–2325.

[31] 31. Wexler M,Sargent F,Jack RL,Stanley NR,Bogsch EG,et al. (2000) TatD is a cytoplasmic protein with DNase activity. No requirement for TatD family proteins in Sec-independent protein export. J Biol Chem 275: 16717–16722.

[32] 32. Hennig S,Ziebuhr W (2010) Characterization of the transposase encoded by IS256, the prototype of a major family of bacterial insertion sequence elements. J Bacteriol 192: 4153–4163.

[33] 33. Lau PCY,Lindhout T,Beveridge TJ,Dutcher JR,Lam JS (2009) Differential lipopolysaccharide core capping leads to quantitative and correlated modifications of mechanical and structural properties in Pseudomonas aeruginosa biofilms. J Bacteriol 191: 6618–6631.

[34] 34. Mack M,Buckel W (1997) Conversion of glutaconate CoA-transferase from Acidaminococcus fermentans into an acyl-CoA hydrolase by site-directed mutagenesis. FEBS Letters 405: 209–212.

[35] 35. Marin M,Perez-Pantoja D,Donoso R,Wray V,Gonzalez B,et al. (2010) Modified 3-oxoadipate pathway for the biodegradation of methylaromatics in Pseudomonas reinekei MT1. J Bacteriol 192: 1543–1552.

[36] 36. Yamashita Y,Tsukioka Y,Tomihisa K,Nakano Y,Koga T (1998) Genes involved in cell wall localization and side chain formation of rhamnose-glucose polysaccharide in Streptococcus mutans. J Bacteriol 180: 5803–5807.

[37] 37. Wang L,Makino S,Subedee A,Bogdanove AJ (2007) Novel candidate virulence factors in rice pathogen Xanthomonas oryzae pv. oryzicola as revealed by mutational analysis. Appl Environ Microbiol 73: 8023–8027.

[38] 38. Lu G,Tang Y,Li C,Li R,An S,et al. (2009) An adenosine kinase exists in Xanthomonas campestris pathovar campestris and is involved in extracellular polysaccharide production, cell motility, and Virulence. J Bacteriol 191: 3639–3648.

[39] 39. Finn RD,Mistry J,Schuster-Bockler B,Griffiths-Jones S,Hollich V,et al. (2006) Pfam: clans, web tools and services. Nucleic Acids Res 34: D247–D251.

[40] 40. Mascher T,Helmann JD,Unden G (2006) Stimulus perception in bacterial signal-transducing histidine kinases. Microbiol Mol Biol Rev 70: 910–938.

[41] 41. Sukumaran SK,Prasadarao NV (2002) Regulation of protein kinase C in Escherichia coli K1 invasion of human mrain microvascular endothelial cells. J Biol Chem 277: 12253–12262.

[42] 42. Gerstmeir R,Cramer A,Dangel P,Schaffer S,Eikmanns BJ (2004) RamB, a novel transcriptional regulator of genes involved in acetate metabolism of Corynebacterium glutamicum. J Bacteriol 186: 2798–2809.

[43] 43. Gonzalez-Barrios AF,Zuo R,Hashimoto Y,Yang L,Bentley W E,et al. (2006) Autoinducer 2 controls biofilm formation in Escherichia coli through a novel motility quorum-sensing regulator (MqsR, B3022). J Bacteriol 188: 305–316.

[44] 44. Brown BL,Wood TK,Peti W,Page R (2011) Structure of the Escherichia coli antitoxin MqsA (YgiT/b3021) bound to its gene promoter reveals extensive domain rearrangements and the specificity of transcriptional regulation. J Biol Chem 286: 2285–2296.

[45] 45. Laia ML,Moreira LM,Dezajacomo J,Brigati JB,Ferreira CB,et al. (2009) New genes of Xanthomonas citri subsp. citri involved in pathogenesis and adaptation revealed by a transposonbased mutant library. BMC Microbiol 9: 12.

[46] 46. Kane SR,Chakicherla AY,Chain PS,Schmidt R,Shin MW,et al. (2007) Whole-genome analysis of the methyl tertbutyl ether-degrading beta-proteobacterium Methylibium petroleiphilum PM1. J Bacteriol 189: 1931–45.

[47] 47. Gabriel D,Karine B,Marc S,Simon P,Anita L (2003) YodA from Escherichia coli is a metal-binding, lipocalin-like Protein. J Biol Chem 278: 43728–43735.

[48] 48. Stojnev T,Harichová J,Ferianc P,Nyström T (2007) Function of a novel cadmium-induced YodA protein in Escherichia coli. Curr Microbiol 55: 99–104.

[49] 49. Spychala J,Datta NS,Takabayashi K,Datta M,Fox IH,et al. (1996) Cloning of human adenosine kinase cDNA: sequence similarity to microbial ribokinases and fructokinases. Proc Natl Acad Sci USA 93: 1232–1237.

[50] 50. Tao F,Swarup F,Zhang LH (2010) Quorum sensing modulation of a putative glycosyltransferase gene cluster essential for Xanthomonas campestris biofilm formation. Environ Microbiol 12: 3159–3170.

[51] 51. Cheng DW,Lin H,Civerolo EL (2010) Extracellular genomic DNA mediates enhancement of Xylella fastidiosa biofilm formation in Vitro. J Plant Pathol 92: 405–410.

[52] 52. Torres PS,Malamud F,Rigano LA,Russo DM,Marano MR,et al. (2007) Controlled synthesis of the DSF cell-cell signal is required for biofilm formation and virulence in Xanthomonas campestris. Environ Microbiol 9: 2101–2109.

[53] 53. Rice KC,Mann EE,Endres JL,Weiss EC,Cassat JE,et al. (2007) The cidA murein hydrolase regulator contributes to DNA release and biofilm development in Staphylococcus aureus. Proc Natl Acad Sci USA 104: 8113–8118.

[54] 54. Rosenshine I,Ruschkowski S,Stein M,Reinscheid DJ,Mills SD,et al. (1996) A pathogenic bacterium triggers epithelial signals to form a functional bacterial receptor that mediates actin pseudopod formation. EMBO J 15: 2613–2624.

[55] 55. Sofia HJ,Chen G,Hetzler BG,Reyes-Spindola JF,Miller NE (2001) Radical SAM, a novel protein superfamily linking unresolved steps in familiar biosynthetic pathways with radical mechanisms: functional characterization using new analysis and information visualization methods. Nucl Acids Res 29: 1097–1106.

[56] 56. He Y,Boon C,Zhou L,Zhang L (2009) Co-regulation of Xanthomonas campestris virulence by quorum sensing and a novel two-component regulatory system RavS/RavR. Mol Microbiol 71: 1464–1476.

[57] 57. He Y,Xu M,Lin K,Ng Y,Wen C,et al. (2006) Genome scale analysis of diffusible signal factor regulon in Xanthomonas campestris pv. campestris: identification of novel cell-cell communication-dependent genes and functions. Mol Microbiol 59: 610–622.

[58] 58. He Y,Ng Y,Xu M,Lin K,Wang L,et al. (2007) Xanthomonas campestris cell-cell communication involves a putative nucleotide receptor protein Clp and a hierarchical signaling network. Mol Microbiol 64: 281–292.

[59] 59. Ryan RP,McCarthy Y,Andrade M,Farah CS,Armitage JP,et al. (2010) Cell–cell signal dependent dynamic interactions between HD-GYP and GGDEF domain proteins mediate virulence in Xanthomonas campestris. Proc Natl Acad Sci USA 107: 5989–5994.

[60] 60. Andrade MO,Alegria MC,Guzzo CR,Docena C,Pareda-Rosa MC,et al. (2006) The HD-GYP domain of RpfG mediates a direct linkage between the Rpf quorum-sensing pathway and a subset of diguanylate cyclase proteins in the phytopathogen Xanthomonas axonopodis pv. citri. Mol Microbiol 62: 537–551.

[61] 61. Leduc JL,Roberts GP (2009) Cyclic di-GMP allosterically inhibits the CRP-like protein (Clp) of Xanthomonas axonopodis pv.citri. J Bacteriol 191: 7121–7122.

[62] 62. Simm R,Morr M,Kader A,Nimtz M,Römling U (2004) GGDEF and EAL domains inversely regulate cyclic di-GMP levels and transition from sessility to motility. Mol Microbiol 53: 1123–1134.

[63] 63. Tischler AD,Camilli A (2004) Cyclic diguanylate (c-di-GMP) regulates Vibrio cholera 3 biofilm formation. Mol Microbiol 53: 857–869.

[64] 64. Römling U,Gomelsky M,Galperin MY (2005) C-di-GMP: the dawning of a novel bacterial signalling system. Mol Microbiol 57: 629–639.

[65] 65. Dow JM,Crossman L,Findlay K,He YQ,Feng JX,et al. (2003) Biofilm dispersal in Xanthomonas campestris is controlled by cell–cell signaling and is required for full virulence to plants. Proc Natl Acad Sci USA 100: 10995–11000.

[66] 66. Rybak M,Minsavage GV,Stall E,Jones JB (2009) Identification of Xanthomonas citri subsp. citri host specificity genes in a heterologous expression host. Mol Plant Pathol 10: 249–262.

[67] 67. Sambrook J,Russell DW (2001) Molecular cloning: a laboratory manual, 3rd ed. NY: Cold Spring Harbor Laboratory Press. pp. 633–638.

[68] 68. El Yacoubi B,Brunings A,Yuan Q,Shankar S,Gabriel D (2007) In planta horizontal transfer of a major pathogenicity effector gene. Appl Environ Microbiol 73: 1612–1621.

PLOS