By continuing to browse this site you agree to us using cookies as described in About Cookies
Notice: Wiley Online Library will be unavailable on Saturday 7th Oct from 03.00 EDT / 08:00 BST / 12:30 IST / 15.00 SGT to 08.00 EDT / 13.00 BST / 17:30 IST / 20.00 SGT and Sunday 8th Oct from 03.00 EDT / 08:00 BST / 12:30 IST / 15.00 SGT to 06.00 EDT / 11.00 BST / 15:30 IST / 18.00 SGT for essential maintenance. Apologies for the inconvenience.
Xylan is a major structural component of plant cell wall and the second most abundant plant polysaccharide in nature.
Here, by combining genomic and functional analyses, we provide a comprehensive picture of xylan utilization by Xanthomonas campestris pv campestris (Xcc) and highlight its role in the adaptation of this epiphytic phytopathogen to the phyllosphere.
The xylanolytic activity of Xcc depends on xylan-deconstruction enzymes but also on transporters, including two TonB-dependent outer membrane transporters (TBDTs) which belong to operons necessary for efficient growth in the presence of xylo-oligosaccharides and for optimal survival on plant leaves. Genes of this xylan utilization system are specifically induced by xylo-oligosaccharides and repressed by a LacI-family regulator named XylR.
Part of the xylanolytic machinery of Xcc, including TBDT genes, displays a high degree of conservation with the xylose-regulon of the oligotrophic aquatic bacterium Caulobacter crescentus. Moreover, it shares common features, including the presence of TBDTs, with the xylan utilization systems of Bacteroides ovatus and Prevotella bryantii, two gut symbionts. These similarities and our results support an important role for TBDTs and xylan utilization systems for bacterial adaptation in the phyllosphere, oligotrophic environments and animal guts.
Xylans represent the predominant hemicelluloses in the cell wall of terrestrial plants. They comprise a conserved backbone composed of 1,4-linked β-d-xylose residues which may be substituted with glucuronic acid, 4-O-methyl-glucuronic acid, arabinose or a combination of types of decorations (Burton et al., 2010; Scheller & Ulvskov, 2010; Fig. 1a). Altogether, xylans account for approximately one-third of all renewable organic carbon on earth. They therefore represent a substantial source of nutriment and many bacteria are able to degrade this hemicellulolytic substrate (Kulkarni et al., 1999; Saha, 2003; Dodd & Cann, 2009). These xylanolytic microbes can be found in diverse ecological niches, which typically comprise environments where plant material accumulates and deteriorates, including plant debris, soil, aquatic environments and the digestive tract of animals (Collins et al., 2005). Plant pathogenic bacteria also display xylanolytic activities, which may help them to breach the cell wall obstacle and to release nutrients during the colonization of plants. Bioconversion of xylans has been intensively studied in the past decade because of its potential applications in agro-industrial processes, such as the pulp and paper industry and biofuel production. These studies have shown that xylan bioconversion is mediated by a wide array of enzymes (Collins et al., 2005; Dodd & Cann, 2009). Although the xylanolytic systems of bacteria isolated from soil or from digestive tracts of animals have been studied in detail, there is only limited information regarding the xylanolytic systems of plant pathogenic bacteria. Moreover, little is known about transport into bacterial cells of xylan deconstruction products.
The Xanthomonas genus comprises an important group of plant pathogenic bacteria that together affect c. 400 plant hosts, including agronomically important crops (Buttner & Bonas, 2009; Ryan et al., 2011). Most Xanthomonas species are able to survive on the aerial part of plants (phyllosphere), a feature that plays an important role in the early stages of infection (Rigano et al., 2007; Li & Wang, 2011). Xylanases have been shown to control the virulence of two members of this genus, Xanthomonas oryzae pv oryzae (Xoo) and Xanthomonas campestis pv vesicatoria (Xcv) (Rajeshwari et al., 2005; Szczesny et al., 2010). The aim of this study was to characterize the xylan utilization system of Xanthomonas campestris pv campestris (Xcc) the causal bacterium of black rot disease of Brassica. Xcc harbours CUT systems (Carbohydrate Utilization with TBDT systems) which are involved in plant carbohydrate scavenging (Blanvillain et al., 2007). These systems comprise inner membrane transporters, degrading enzymes, transcriptional regulators and TonB-dependent outer membrane transporters (TBDTs; Blanvillain et al., 2007). In contrast to passive transport mediated by porins, TBDTs allow high-affinity and active transport of bigger substrate molecules (Cornelis, 2010; Krewulak & Vogel, 2011). TBDTs have been shown to transport iron-siderophore complexes, vitamin B12 and, more recently, various carbohydrates (Neugebauer et al., 2005; Blanvillain et al., 2007; Eisenbeis et al., 2008; Schauer et al., 2008).
A global study of Xcc ATCC33913 (LMG568) (Xcc-568) TBDT genes has shown that the expression of two of them, XCC2828 and XCC4120, is specifically induced by xylan and xylose (Blanvillain et al., 2007). In this study we show that they belong to a complex CUT system involved in the uptake and utilization of xylan. This system is important for fitness of Xcc-568 in the phyllosphere.
Materials and Methods
Bacterial strains, plasmids and growth conditions
The Xcc-568 strains and plasmids used in this study are listed in Supporting Information Table S1. Xcc-568 cells were grown at 28°C in MOKA rich medium (Blanvillain et al., 2007) or in minimal medium (MME; Arlat et al., 1991). Escherichia coli cells were grown on Luria–Bertani medium at 37°C.
Antibiotics were used at the following concentrations: for Xcc-568, 50 μg ml−1 rifampicin, 50 μg ml−1 kanamycin, and 5 μg ml−1 tetracycline; for E. coli, 50 μg ml−1 ampicillin, 25 μg ml−1 kanamycin, and 10 μg ml−1 tetracycline.
Construction of Xanthomonas campestris pv campestris mutants
Insertion mutants were constructed using the suicide plasmid pVO155 (Oke & Long, 1999) with a 300- to 500-bp PCR amplicon internal to each open reading frame (ORF). Deletion mutants were constructed by using the cre-lox system adapted by Angot et al. (2006) from the system of Marx & Lidstrom (2002) or by using the sacB system (Schafer et al., 1994). Deleted regions and pVO155 plasmid insertions are indicated in Table S1 and represented on Fig. S1. Oligonucleotide primers used for PCR amplification will be provided upon request.
DNA manipulations were performed using standard procedures (Sambrook et al., 1989).
For complementation studies, PCR amplicons presented in Fig. S1 (oligonucleotide primers used for PCR amplification will be provided upon request) were cloned into pCZ1016, a derivative of pFAJ1700 containing the Ptac promoter, multiple cloning sites and the T7 terminator from pSC150 (Dombrecht et al., 2001; Cunnac, 2004). To perform chromosomal complementations, PCR amplicons were cloned into pCZ1034, a derivative of pK18mobsacB (Schafer et al., 1994) with the MCS replaced by a Ptac promoter, a MCS and a T7 terminator flanked by a 700-bp fragment corresponding to the region upstream from the open reading frame XCC0127 and a 700-bp fragment corresponding to the region downstream from the open reading frame XCC0128.
The XCC4120 promoter region (see Table S1) was PCR amplified with appropriately designed primers. This promoter region was cloned as HindIII–XbaI fragment, into the pCZ962 plasmid, a pFAJ1700 (Dombrecht et al., 2001) derivative containing the KpnI–AscI lacZ gene from the pCZ367 plasmid (Cunnac, 2004), giving pPr-xytB.
Expression studies, RNA isolation and operon mapping
β-galactosidase and β-glucuronidase assays: bacterial cultures in the appropriate medium were harvested at different time points and β-galactosidase and β-glucuronidase (GUS) assays were performed as previously described (Blanvillain et al., 2007).
In order to investigate the transcriptional organization, reverse transcription-PCR (RT-PCR) experiments were performed. Bacterial cultures from xylR mutant of Xcc-568 grown in minimal medium (MME) were harvested after 6 h of incubation at an optical density at 600 nm (OD600) of 0.6. RNAs were extracted using the RNeasy Mini Kit (Qiagen). A total of 5 μg of RNA was reverse transcribed with Transcriptor Reverse Transcriptase enzyme (Roche Diagnostics, Meylan, France) using random hexamers (Biolabs, Evry, France) for 10 min at 25°C and then for 40 min at 55°C. The resulting cDNAs were used as a template for PCR amplification with Taq polymerase using specific primer pairs for each gene (as indicated in Fig. S2) and analysed by agarose-gel electrophoresis.
Quantitative reverse transcription-PCR (qRT-PCR) experiments were performed essentially as previously described (Blanvillain et al., 2007). For qRT-PCR, experiments were performed on bacteria grown on solid medium containing 4-O-Methyl-d-glucurono-d-xylan-Remazol brilliant blue R (RBB-Xylan; Sigma), colonies obtained after 48 h growth were resuspended in 1 ml of water. A 1 μg sample of RNA was treated with RNase-free DNase I (Sigma) for 20 min at room temperature. After DNase inactivation (10 min at 70°C), RNAs were reverse transcribed as indicated above. Oligonucleotide primers used for quantitative PCR amplification will be provided upon request. 16S rRNA was used as a control for real-time PCR (Morales et al., 2005; Blanvillain et al., 2007).
Calculation of maximal growth rate
Growth curves of Xcc-568 strains grown at 28°C in MME liquid culture in the presence of xylose or xylo-oligosaccharides were generated using a FLUOStar Omega apparatus (BMG Labtech, Offenburg, Germany) with four replicates. Growth was monitored by measuring OD600 using 96-well flat-bottom microtiter plates with 200 μl preparations inoculated at OD600 of 0.1 from four independent washed overnight precultures. The microplates were shaken continuously at 700 rpm using the linear-shaking mode. Generation time (G), defined as doubling time, was calculated during the exponential phase of growth using the following formula: G = tf − t0/n where n is equal to (logNf − logN0)/log2 (N0, initial number of bacteria at the initial time point considered (t0); Nf, final number of bacteria at the final time point considered (tf)). The maximum specific growth rate (μmax), defined as the increase in cell mass per time unit, was calculated as follows: μmax = ln 2/G. Statistical analysis was performed using the RGUI software (GNU General Public License; Free Software Foundation Inc., Boston, MA, USA).
[14C] xylose transport experiments
[14C] xylose transport assays were conducted as previously described (Blanvillain et al., 2007; Boulanger et al., 2010). [14C] xylose (Amersham Biosciences, specific activity of 3.15 GBq mmol−1) was added to a final concentration of 0.5 μM. For competition experiments, unlabelled sugars were added to [14C] xylose at final concentrations of 0.5, 5, 50 and 500 μM, and cells were incubated for 1 h before collection. The initial concentration-dependent xylose transport was determined using the rapid dilution method as previously described (Neugebauer et al., 2005; Blanvillain et al., 2007).
Plate assays for detection of xylanase activity
The plate assay for xylanase activity was performed using MME-agar plates containing 0.1% RBB-xylan (Sigma). Overnight cultures of Xcc-568 strains grown in MOKA medium were centrifuged. Pellets were resuspended in MME medium and the OD600 was adjusted to 0.4. Five microlitres of bacterial suspension were spotted on plates that were incubated at 28°C. The detection of xylanase activity was examined periodically by checking the halo against the blue background.
Pathogenicity tests were conducted on Arabidopsis thaliana Sf-2 ecotype as previously described (Meyer et al., 2005).
Dynamics of bacterial population densities in the phyllosphere of cabbage and bean
Experiments on cabbage (Brassica oleracea cv Bartolo) and dry bean (Phaseolus vulgaris cv Flavert) as well as statistical analyses were performed at IRHS as previously described (Darsonval et al., 2008).
Patscan and Predetector software (Dsouza et al., 1997; Hiard et al., 2007) were used to identify xyl-boxes.
XCC2828 and XCC4120 TBDTs genes are located in loci putatively involved in xylan/xylose metabolism
The XCC2828 and XCC4120 TBDT genes, whose expression is specifically induced by xylan and xylose, display significant homologies with TBDTs genes from the aquatic bacterium Caulobacter crescentus CB15 (Cc-CB15), CC0999 and CC2832, respectively, (Fig. 2a, Table S2). Interestingly, a transcriptomic analysis showed that these two Cc-CB15 genes belong to a complex xylose regulon whose repression is mediated by CC3065, a LacI-family regulator named XylR. This repressor was shown to recognize a specific 14 bp-operator motif (Hottes et al., 2004; Stephens et al., 2007a,b; Fig. S3). This motif, found upstream from both CC0999 and CC2832 TBDT genes has several close matches in the genome of Xcc-568, two of which are located upstream from XCC2828 and XCC4120 TBDT genes (Hottes et al., 2004). The screening of the Xcc-568 genome sequence predicted two additional motifs perfectly matching the 14 bp palindromic motif, named xyl-box, located upstream from XCC2828 and XCC4120 (Table S3). One motif is located upstream from the XCC4119 gene, encoding a putative inner membrane transporter of the major facilitator superfamily (MFS) and the other upstream from the XCC4100 gene, encoding a putative xylose isomerase (Table 1). Interestingly, several genes surrounding XCC2828, XCC4100, XCC4119 and XCC4120 genes code for proteins displaying high similarities to proteins of the xylose regulon of Cc-CB15 (Fig. 2a; Table S2). Among these proteins, XCC4101, a putative LacI family regulator, is very well conserved to XylR from Cc-CB15 and was therefore named XylR. Moreover, several genes located in these loci have predicted functions associated with the utilization of xylan, xylose or glucuronic acid (Table 1; Fig. 1). The major enzymes that attack xylan backbone are classified in the carbohydrate-active enzyme (CAZy) database (http://www.cazy.org; Cantarel et al., 2009). They comprise endo-1,4-β-d-xylanases (EC 126.96.36.199), which generate xylo-oligosaccharides (Table S4). The degradation of xylo-oligosaccharides is mediated by β-d-xylosidases, whereas elimination of the side groups is catalysed by α-l-arabinofuranosidases, α-d-glucuronidases, acetylxylanesterases, ferulic acid esterases and p-coumaric acid esterases (Table S4). The genes of Xcc-568 found in these different glycosyl hydrolases (GH) or carboxylesterase (CE) families were named to indicate their activity and CAZy family, as previously described for Cellvibrio japonicus (DeBoy et al., 2008) and according to the nomenclature recently proposed by Potnis et al. (2011; Table 1). This analysis allowed us to define three loci, named xytA, xytB and xylR, containing xyl-boxes and enzymes putatively associated with xylan deconstruction and glucuronic acid metabolism, as well as inner membrane transporters and TBDTs. They might therefore form a xylan CUT system (Fig. 1). Moreover, the xylR locus contains xylaA2 gene encoding a putative xylose isomerase, an enzyme which carries out the first step in xylose metabolism (Lawlis et al., 1984; Fig. 3). The analysis of the Xcc-568 proteome showed that this pathogen possesses a second xylose isomerase gene (named xylA1) which displays high similarity to xylA2 (97% identity at DNA level). xylA1 does not belong to the xylan CUT system defined above. It is located between XCC1759 (xylE) gene, encoding a putative MFS inner membrane transporter and xylB, a putative d-xylulokinase gene, thus suggesting that Xcc-568 possesses a classical two-step xylose utilization pathway (Lawlis et al., 1984; Fig. 3; Table 1). No perfect or even degenerated xyl-box was found in this locus, named xylE (Fig. 1).
Table 1. Identification and properties of the relevant ORFs from Xanthomonas campestris pv campestris ATCC33913 (Xcc-568) xylan CUT system
As determined by using the Conserved Domain Database (Marchler-Bauer et al., 2011) and the Pfam database (Finn et al., 2010).
The reported homologous proteins are those showing the highest score among proteins with an experimentally defined function. In the absence of relevant biochemical data, the most similar protein from bacteria outside the Xanthomonadaceae family was reported.
Start codon prediction revised in this work. All other start codons are from GenBank (da Silva et al., 2002) or from Blanvillain et al. (2007).
PyrH (Streptomyces rugosporus/AAU95674) (Zehner et al., 2005)
Putative Tryptophan halogenase
No conserved domain
Pass1 (Rattus Norvegicus/Q5BKC6) (Liu et al., 2000)
Genes belonging to xytA, xytB and xylR loci are specifically induced by xylo-oligosaccharides
The expression of most genes located in the xytA, xytB or xylR loci was studied in the presence of xylan, xylose or xylo-oligosaccharides (xylobiose, X2; xylotriose, X3; xylotetraose, X4). These experiments were performed by using pVO155 insertion mutants which carry transcriptional fusions between the targeted genes and the uidA reporter gene (Oke & Long, 1999) or by qRT-PCR.
Most genes in these three loci display a similar expression pattern: their expression is specifically and highly induced by xylo-oligosaccharides. They are also induced to a lesser extent by xylan, and xylose (Tables 2, 3).
Table 2. Relative expression ratios measured by using pVO155 reporter plasmid insertions in genes of the xylan utilization system grown in the presence of xylan, xylose or xylo-oligosaccharides
xylR regulatory gene and xypB, that code for a putative inner membrane transporter, showed distinctive expression induction. When monitored in the xylR::pVO insertion mutant, the expression of xylR was not induced by xylan or xylo-oligosaccharides (Table 2), whereas its expression is induced by xylan and X3 when monitored by qRT-PCR in a wild-type background (Table 3; Fig. S4b). Similarly, when monitored in xypB::pVO insertion mutant, the expression of xypB gene was not induced by xylan, X3, X4, or xylose (Table 2) whereas its expression followed the general induction pattern (i.e. high induction by xylo-oligosaccharides and weaker induction by xylose) when monitored by qRT-PCR in a wild-type background (Table 3). These observations suggested that a functional copy of this gene might be required for its own induction by xylan and/or X3 or X4. This hypothesis was confirmed by introducing the (pC-xypB) plasmid, expressing xypB constitutively, into the xypB::pVO mutant (Fig. 4a). Moreover, the induction by X2, X3 and X4 of xytB TBDT promoter fused to the lacZ reporter gene on the pPr-xytB plasmid was abolished in the ΔxypB deletion mutant and recovered by introducing a functional copy of xypB into the ΔxypB chromosome (Fig. 4b).
The expression pattern of the xylA1, xylB and xylE genes of the xylE locus is clearly different from that observed for the CUT xylan utilization system because they are generally equally induced by xylose and xylo-oligosaccharides (Table 3). Finally, the expression of genes coding for other enzymes located outside the xytA, xylR, xytB and xylE loci, including xyn30A putative xylanase, is not induced by xylose, xylo-oligosaccharides or xylans (data not shown).
XylR represses the expression of genes/operons preceded by a xyl-box
The conservation of XylR and xyl-boxes in Cc-CB15 and Xcc-568 prompted us to compare the expression of genes located in xyl E, xytA, xytB and xylR loci in the wild-type strain or in a xylR::pVO insertion mutant by qRT-PCR analysis. XylR represses the expression of all genes located immediately downstream from putative xyl-boxes (i.e. xytA, xytB, xypA and xylA2; Table 4). The expression of the four genes located downstream of xypA (xyn10A to xyn10C) is also repressed by XylR (Table 4), suggesting that they form an operon with xypA. Operon mapping by RT-PCR analysis confirmed this hypothesis (Fig. S2). Similarly, it appeared that xytA, xyaA and xyaB, on the one hand, and xytB, xypB, and gly43F, on the other, form two operons negatively regulated by XylR (Table 4, Fig. S2).
Table 4. Regulation of genes in the xylan CUT system by XylR
The expression status was clearly different in the xylR locus. xylA2 which is the unique gene of this locus displaying a xyl-box is the only one whose expression is repressed by XylR in this locus. The expression of agu67A, axeXA, uxuA, gly43E, xyl3A and xylR itself is not repressed by XylR in MME (Table 4). As the expression of all these genes is specifically induced by xylo-oligosaccharides and to a lesser extent by xylose, we compared their expression by qRT-PCR in wild-type or xylR::pVO genetic backgrounds. Our data clearly show that the induction of xylR expression by X3 depends on a functional copy of xylR (Fig. S4b). Similarly, the induction of agu67A, axeXA and uxuB by X1 and X3 appears to be positively influenced by XylR (Fig. S4). This might also be the case for the uxuA, gly43E and xyl3A genes which are locaded between axeXA and uxuB and that seem to form a large operon with agu67A. Indeed, expression experiments carried out by qRT-PCR analysis with the wild-type strain and the agu67A::pVO mutant suggest that the pVO insertion into agu67A, the first gene of this putative operon, has a polar effect on the transcription of axeXA and uxuB (Fig. S5) as well as uxuA, gly43E and xyl3A (data not shown). This insertion into agu67A has no effect on the transcription of xylR, xyn10A or xytB (Fig. S5).
Finally, the expression of genes of the xylE locus (Fig. S4a and data not shown) or coding for other enzymes putatively involved in xylan deconstruction but located outside the xytA, xylR or xytB loci, including xyn30A putative xylanase, is not controlled by XylR (data not shown).
Xyn10A, Agu67A, Gly43F, XylR and XypB control the production of extracellular xylanase activity
In order to see whether genes belonging to the xytA, xylR, and xytB loci are involved in the production of the extracellular xylanolytic activity produced by Xcc-568, mutants in these loci were tested for the production of extracellular xylanase activity. Most of the studied mutants displayed xylanase activities similar to that of the wild-type strain (data not shown). However, some mutants were significantly affected (Table 5). The level of xylanase activity was increased in the xylR repressor mutant, thus confirming that this gene represses the expression of genes required for xylan degradation. More surprisingly, the activity was also significantly higher in the Δgly43F deletion mutant than in the wild-type strain (Table 5). This mutant was the only mutant of family GH43 to show a modification in xylanase activity. The introduction of the pC-gly43F complementation plasmid into the Δgly43F mutant significantly reduced the level of xylanase activity, confirming the role of this enzyme in the production of xylanase activity.
Table 5. Production of extracellular Xylanase by Xanthomonas campestris pv campestris strains
Xylanase relative activity was estimated by calculating the (H2–C2)/C2 ratio, where H is the diameter of the halo and C the diameter of the bacterial colony, measured 4 d after spotting. The symbols +++, ++, +, +/− or − refer to the production of very high, high, medium, low or nonproduction of xylanase relative activity by the different strains.
pCZ1016 is the empty expression vector for complementation experiments. This empty vector was introduced into all tested mutants without affecting xylanase activity (data not shown) as shown for agu67A:: pVO mutant.
No activity was detected for Δxyn10A mutant, which carries a deletion of xyn10A xylanase gene. Complementation experiments conducted with pC-xyn10A plasmid, confirmed that the extracellular activity detected in these conditions is coded by xyn10A gene (Table 5). Accordingly, we did not observe any significant reduction in extracellular xylanase activity in Δxyn10C or xyn30A::pVO mutants affected in the two other putative xylanase genes of Xcc (Table 5). The level of extracellular xylanase activity was also significantly lower in agu67A::pVO mutant. This reduced phenotype was complemented by the introduction of pC-agu67A plasmid (Table 5). This suggests that the putative α-glucuronidase encoded by this gene is mandatory to get full extracellular xylanase activity. Finally, the activity was severely decreased in ΔxypB inner membrane transporter mutant but not in any other transporter mutants (Table 5). Complementation experiments carried out with pC-xypB plasmid confirmed that the reduction in xylanase activity is due to the mutation in this gene. This result correlates well with expression results suggesting that this putative transporter plays a crucial role in the induction of the system.
Xylose is transported across Xcc-568 inner membrane by XylE
The phenotype of xypB mutants, the presence of TBDT and other inner membrane transporter genes in the XylR regulon prompted us to study the transport of xylose and xylo-oligosaccharides by Xcc.
The initial concentration-dependent [14C]xylose transport, reflecting the dissociation constant (Kd) for xylose uptake was determined using the previously described rapid dilution method (Neugebauer et al., 2005; Blanvillain et al., 2007). The deduced Kd (122 μM) is in a range similar to that of Kd values obtained for passive diffusion through porins (Boulanger et al., 2010). Moreover, the kinetic values showed that the uptake rate was low and monophasic (Fig. 5), suggesting passive diffusion. In agreement with these data the transport of xylose is not depending on XytA and XytB TBDTs (Table 6). Experiments performed with mutants in xypA, xypB or xylE, the inner membrane transporter genes identified in the xylan/xylose CUT system, showed that XylE only is required for xylose transport across the inner membrane. The uptake rate of labelled xylose obtained for xylE mutants represented only c. 20% of the rate obtained for the wild-type strain (Table 6). These results were confirmed by comparing maximum specific growth rates (μmax) of the wild-type strain and mutants in transporter genes in MME supplemented with xylose. Growth of the xylE::pVO mutant was impaired on MME containing xylose, contrary to the xypA, xypB, ΔxytA or ΔxytB1 mutants (Fig. 6a). In the xylE::pVO-complemented strain, xylose transport capacity and growth on MME-xylose were both restored (Table 6; Fig. 6a).
Table 6. Rates of 14C-labelled xylose transport of mutants compared to the rate in Xanthomonas campestris pv campestris ATCC33913 wild-type straina
Transport rates were measured 60 min after addition of 14C-labelled xylose.
Standard deviations were calculated from three independent experiments.
LacI family regulator
Inner membrane transporters
Inner membrane transporters
xytB locus is required for normal growth in presence of xylo-oligosaccharides
Because growth of the wild-type strain and xylE mutants in the presence of xylose corroborated the transport status observed for [14C] xylose uptake by these strains, we speculated that growth rate studies might indirectly allow us to study the transport of xylo-oligosaccharides. We focused this analysis on the xytA and xytB loci because they both contain transporter genes. The wild-type strain and mutants in these loci were grown in MME supplemented with X2, X3 or X4, at a final concentration of 2 mM, a concentration that induced the expression of most genes of the xylan utilization system. Data obtained with X3 are presented (Fig. 6b). Similar results were obtained with X2 and X4 (data not shown).
We noticed that μmax of the wild-type strain was slightly lower in presence of X3 than in the presence of xylose (Fig. 6). The μmax of the xytB, xypB and gly43F mutants was significantly affected in the presence of X3 as compared to that of the wild-type strain, whereas it was not impaired in presence of xylose (Fig. 6). Wild-type growth rate in the presence of X3 was restored to the gly43F::pVO insertion mutant by introducing the pC-gly43F complementation plasmid. Similarly, the ΔxypB deletion mutant could be complemented by introducing the pC-xypB plasmid. However, the xypB::pVO insertion mutant was only partially complemented by the introduction of pC-xypB whereas it was fully complemented by the pC-xypB-gly43F plasmid (Fig. 6b). These results confirm that xypB and gly43F are co-transcribed. In the presence of X3, the μmax of ΔxytB1 and xytB::pVO mutants was also reduced but remained higher than that of the xypB or gly43F insertion mutants (Fig. 6b). Surprisingly, although considered as nonpolar, the deletion introduced into xytB mutants could not be complemented by introducing the complementation plasmid pC-xytB (Fig. 6b). Smaller deletions were constructed (Fig. S1) and similar results were obtained even with the smaller deletion mutant (ΔxytB3) (data not shown). The ΔxytB1 and xytB::pVO mutants were fully complemented by the pC-xypB-gly43F plasmid (Fig. 6b). Previous data obtained with the XylR::pVO mutant showed that xytB, xypB and gly43F form an operon. However, the fact that the μmax of the xytB insertion or deletion mutants is similar and higher than that of the xypB or gly43F insertion mutants in the presence of X3 suggests that xytB is not fully co-transcribed with xypB and gly43F in these conditions. Moreover, complementation experiments suggest a cis-regulatory effect. Altogether, these data show that that xypB and gly43F may play an important role in xylo-oligosaccharide transport and metabolism. The phenotype of xytB deletion mutants renders the study of the role of this TBTD in xylo-oligosaccharide transport difficult to assess.
xytA and xytB loci are important for growth on plant leaves
We studied pathogenicity of pVO155 insertion mutants constructed in this study on cabbage or Arabidopsis thaliana host plants. These experiments were performed using two distinct methods: the wound inoculation method, that allows direct delivery of bacterial cells into the xylem vessels of leaves, or the infiltration method, which delivers bacteria into the plant leaf mesophyll (Meyer et al., 2005). None of the mutants tested, including the xytA, xytB, xypB single mutants and ΔxytA-ΔxytB1 double mutants, as well as mutants altered in the three xylanase genes, were significantly affected in pathogenicity (data not shown). The growth of xytA::pVO and xytB::pVO mutants in Arabidopsis plant tissues was also not significantly different from that of the wild-type (data not shown). We also compared the survival and the multiplication of the wild-type strain and xytA::pVO or xytB::pVO mutants in the phyllosphere of cabbage (host plant) or bean (nonhost plant). The dynamics of bacterial population densities was followed after spray inoculation of the leaves in conditions that do not favour disease expression (Darsonval et al., 2008). The multiplication of the xytB::pVO mutant on cabbage was significantly lower than that of the wild-type strain only during the first 8 d following the inoculation (Fig. 7a). Cell densities measured for the xytA::pVO mutant on host plants were clearly lower than that measured for the wild-type strain and xytB mutant (Fig. 7a). Interestingly, the survival of both xytA and xytB mutants was significantly altered on nonhost plant and the defect of the xytA mutant was again more pronounced than that of the xytB mutant (Fig. 7b).
The phyllosphere represents the aerial parts of terrestrial plants including leaves, stems, buds, flowers and fruits. This habitat has been estimated to cover a global surface of c. 1 billion square kilometres supporting > 1026 bacteria (Morris & Kinkel, 2002; Lindow & Brandl, 2003; Whipps et al., 2008). Although the phyllosphere has been less intensively studied than the rhizosphere, metagenomic approaches have recently given interesting information on bacterial communities colonizing this vast niche (Vorholt, 2012). Recently, a metaproteogenomics analysis performed on leaves of soybean, clover and Arabidopsis identified TBDTs as the most prominent group of transport proteins. These transporters were over-represented among the proteins assigned to Sphingomonas which was one of the predominant genera identified in a study by Delmotte et al. (2009). It was postulated that this over-representation of TBDTs might play a role in the successful adaptation of these bacteria on plant leaves. In this study, by performing functional and genomic analyses of xylan utilization in Xcc, we identified a xylan CUT system which is required for optimal colonization of plant leaves. Therefore, our work seems to confirm the importance of TBDTs for the adaptation of bacteria to the phyllosphere.
The xylan CUT system of Xcc-568 comprises the xytA, xytB and xylR loci which contain enzymes for the degradation of xylan, the metabolism of xylose and glucuronic acid, as well as inner membrane transporters beside TBDTs. We also identified a fourth locus, xylE, involved in xylose utilization (see model Fig. 3).
The expression of most of genes of the xylan CUT system is specifically and highly induced by short xylo-oligosaccharides and to lesser extent by xylose. The expression of a large proportion of these genes is repressed by XylR LacI-type repressor. The regulation mediated by XylR is strictly correlated with the presence of a 14-bp palindromic xyl-box motif in the promoter region of repressed genes or operons. Interestingly, six contiguous genes, agu67A, axeXA, uxuA, gly43E, xyl3A and uxuB, which seem to form an operon, although being induced by xylo-oligosaccharides or xylose, are not under the repression of xylR and no xyl-box was identified in their promoter regions. On the contrary, their induction by X3 as well as that of xylR is positively affected by XylR. This observation shows that the induction by xylo-oligosaccharides or xylose is not solely under XylR control and suggests the existence of other regulators controlling the utilization of xylan and xylose in Xcc-568. Further work is needed to characterize the inducer of XylR in Xcc-568 and to identify other putative regulators of this system.
Among the three xylanase genes identified in the Xcc-568 genome, xyn10A located in the xytB locus was shown to be responsible for the detected extracellular activity produced by this bacterium in our test conditions. No extracellular activity associated with Xyn10C (XCC4118) or Xyn30A (XCC0857) was detected, although both proteins harbour a signal peptide and seem to be secreted like Xyn10A (XCC4115) (see Fig. S6 and Methods S1). Xyn30A is the orthologue of XynC which is responsible for the extracellular xylanase activity detected in Xcv (Szczesny et al., 2010). Despite the high similarity between Xyn30A and XynC (81% amino acid identity), the presence in both cases of a signal peptide and the conservation of GH30 family specific catalytic residues (Hurlbert & Preston, 2001; Larson et al., 2003; St John et al., 2011), we did not detect any extracellular activity associated with the xyn30A gene in Xcc-568. This gene is neither regulated by XylR nor induced by xylan and xylo-oligosaccharides in Xcc-568. We observed that beside Xyn10A, Agu67A, a putative α-glucuronidase involved in the degradation of glucuronic acid decorations, is also required to get full extracellular xylanase activity. This suggests that removal of these side chains from the xylan backbone may potentiate the degradation of xylan. The importance of glucuronic acid liberation during xylan degradation is underscored by the presence of enzymes involved in the metabolism of this carboxylic acid in the xylR and xytB loci (Figs 1, 3).
There might also be a coupling between xylan degradation and xylose metabolism because the xylan CUT system comprises the xylA2 gene which codes for a putative xylose isomerase gene. This gene is duplicated in the Xcc-568 genome and the second copy, xylA1, maps in the xylE locus with xylE inner membrane transporter gene, required for xylose uptake. The three genes forming the xylE locus are not under the regulation of XylR and have a different induction pattern than that of xylan CUT system genes, because they are equally induced by xylose and xyl-oligosaccahrides. These observations suggest that there are different regulators for xylose and xylan utilization pathways, but they also suggest that both pathways are interconnected through the metabolism of xylo-oligosaccharides. Our results on the xytB-xypB-gly43E operon suggest that XypB and Gly43F play essential functions in this metabolism. Indeed, the XypB putative inner membrane transporter is required for the production of extracellular xylanase activity, for the induction of the xylan CUT system by X3, X4 and xylan, and for growth on xylo-oligosaccharides. These convergent data strongly suggest that this transporter plays a major role in the transport of xylo-oligosaccharides across the inner membrane and that this transport is crucial for the induction of the system and for the physiology of Xcc-568. Although gly43F is also required for normal growth in presence of xylo-oligosaccharides, a mutation in this gene led to a large increase in extracellular xylanase activity, unlike what was observed for xypB mutants. Gly43F is closely related to XynB from Prevotella bryantii B14 (Table 1). XynB is an intracellular exoxylanase which was proposed to release xylose progressively from xylo-oligosaccharides, including xylobiose, transported inside the cells (Gasparic et al., 1995). In Xcc-568, Gly43F is the only protein of the GH43 family that has no signal peptide, suggesting that it functions in the cytoplasm. We can speculate that Gly43F degrades XypB-transported xylo-oligosaccharides to xylose, thereby promoting bacterial growth. Gly43F may therefore play a central role in the physiology of Xcc-568 by maintaining a balance between the production of xylose and the maintenance of xylo-oligosaccharides which induce the CUT system (Fig. 3). The role of the xytB TBDT gene located upstream from xypB and gly43F is still elusive. Mutations in this gene have an effect on growth with xylo-oligosaccharides, but we were unable to complement these mutations. Our data showed that the xytB, xypB and gly43F genes form an operon repressed by XylR. However, our growth rate results suggest that the situation is more complex in the presence of xylo-oligosaccharides. They suggest the presence of cis-regulatory sequences into xytB driving the expression of xypB and gly43F. Further work is necessary to fully characterize this locus.
The xylan CUT system encompasses another TBDT gene, xytA. This gene is the first gene of a XylR-regulated operon which comprises two other genes, xyaA and xyaB whose function in xylan degradation remains unknown. Interestingly, this operon and the downstream gene are very well conserved with a quartet of contiguous genes in Cc-CB15 (Fig. 2a, Table S2). The first two genes of this quartet, including CC0999 TBDT gene, belong to the xylose regulon. This conservation suggests that this set of four genes may play an important role in xylan/xylose metabolism in Xcc-568 and Cc-CB15.
The analogy between the Xcc-568 xylan CUT system and the Cc-CB15 xylose-regulon is not restricted to this locus and 10 other genes of the Xcc-568 xylan/xylose CUT system display significant similarities to proteins of Cc-CB15 (Fig. 2a; Table S2). Interestingly, six of these genes belong to the xylose regulon identified in C. cresecentus CB15 (Hottes et al., 2004). This conservation includes proteins involved in the removal of substitutions, xylo-oligosaccharides degradation, glucuronate metabolism and TBDTs. With the exception of Xyn10C, which displays similarities to CC3042, the two other xylanases of Xcc-568 are not conserved in C. cresenctus CB15. We identified another putative xylanase, CC2803, of the GH10 family in the Cc-CB15 genome which is not conserved in Xcc-568. Like CC3042, CC2803 was not detected as induced by xylose, but it is located between two genes induced by this monosaccharide (Hottes et al., 2004). These observations suggest that Cc-CB15 is able to degrade xylan. Interestingly, the xylose regulon of Cc-CB15 comprises nine TBDTs, two of which are highly conserved with XytA and XytB. This high number and the conservation with Xcc-568 TBDTs suggest a crucial role for these outer membrane transporters in the uptake of molecules during xylan/xylose catabolism. Recently, sets of genes specifically induced by xylan have been revealed by transcriptomic studies on P. bryantii B14 (Pbr) (Dodd et al., 2010b), and Bacteroides ovatus ATCC8483 (Bov) (Martens et al., 2011), two Bacteroidetes present in the bovine rumen and human gut, respectively. Interestingly, several genes belonging to Xcc-568 xylan CUT system display significant similarities with xylan-induced genes of Pbr and Bov (Fig. 2b, Table S2). Moreover, the xylan regulon of these latter bacteria share a cluster of conserved genes which contains two TBDTs belonging to the SusC family. This cluster is widely conserved among human- and animal-associated Bacteroides spp. and Prevotella spp. and was proposed to constitute a core set of genes required for xylan fragments uptake by gut-associated Bacteroidetes (Dodd et al., 2011). Therefore, it seems that the association between TBDT and xylan utilization is a common feature shared by bacteria belonging to very different phyla and having apparently different lifestyles. Does this mean that TBDTs play a very important role in natural conditions? The exact role of these outer membrane transporters has yet to be determined. However, the involvement of TBDTs may represent two advantages. First, TBDTs allow the binding and uptake of larger molecules than porins. Therefore, they could transport large xylan hydrolysis products thus preventing release of saccharides in the medium that could be used by other microorganisms. Second, they allow active transport of substrate molecules with a very high affinity (Blanvillain et al., 2007). Therefore, it is possible that TBDTs in these systems play a crucial role in the transport of xylan breakdown products when these molecules are present in scarce amounts. This property may be pivotal for oligotrophs such as Caulobacter species (Hottes et al., 2004). In this study, we showed that XytA and XytB TBDTs belong to operons that are required for optimal growth of Xcc-568 on plant leaves. The phyllosphere corresponds to an oligotrophic environment (Lindow & Brandl, 2003) and we can speculate that XytA and XytB may play a crucial role in the adaptation of Xcc-568 to this niche, which is an important step for Xcc life cycle. Therefore, it seems that Xcc-568 and Caulobacter, which belong to different Proteobacteria families, share similar strategies to survive in niches where nutrients are limited. However, both species most probably possess specific features reflecting their lifestyles. The survival and development of Xcc in the phyllosphere may be important to maintain population sizes sufficient for disease induction. Previous studies on Xoo and Xcv also showed that xylan degradation play a role in virulence of these strains (Rajeshwari et al., 2005; Szczesny et al., 2010). Therefore, together with another study showing the involvement of HrpG and HrpX, the key regulators of type III secretion system, in the phyllosphere colonization of Xanthomonas fuscans ssp. fuscans (Darsonval et al., 2008), this work sheds new light on mechanisms connecting epiphytic colonization to disease induction in plant pathogenic bacteria. It could also have a significant impact on agro-industrial processes.
We thank Annabelle Four-Burgand and Lennart Lessmeier for technical assistance, Laurent Noël for critical comments on the manuscript. G.D., S.B-B. and A.B. were funded by the French Ministry of Research and Technology. We gratefully acknowledge financial support from the Département Santé des Plantes et Environnement-Institut National de la Recherche Agronomique (grant 2007_0441_02) and from the French Agence Nationale de la Recherche (grant ANR-08-BAN-0193-01). This work is part of the ‘Laboratoire d'Excellence’ (LABEX) entitled TULIP (ANR-10-LABX-41).