Notice: Wiley Online Library will be unavailable on Saturday 27th February from 09:00-14:00 GMT / 04:00-09:00 EST / 17:00-22:00 SGT for essential maintenance. Apologies for the inconvenience.
The diffusible factor synthase XanB2, originally identified in Xanthomonas campestris pv. campestris (Xcc), is highly conserved across a wide range of bacterial species, but its substrate and catalytic mechanism have not yet been investigated. Here, we show that XanB2 is a unique bifunctional chorismatase that hydrolyses chorismate, the end-product of the shikimate pathway, to produce 3-hydroxybenzoic acid (3-HBA) and 4-HBA. 3-HBA and 4-HBA are respectively associated with the yellow pigment xanthomonadin biosynthesis and antioxidant activity in Xcc. We further demonstrate that XanB2 is a structurally novel enzyme with three putative domains. It catalyses 3-HBA and 4-HBA biosynthesis via a unique mechanism with the C-terminal YjgF-like domain conferring activity for 3-HBA biosynthesis and the N-terminal FGFG motif-containing domain responsible for 4-HBA biosynthesis. Furthermore, we show that Xcc produces coenzyme Q8 (CoQ8) via a new biosynthetic pathway independent of the key chorismate-pyruvate lyase UbiC. XanB2 is the alternative source of 4-HBA for CoQ8 biosynthesis. The similar CoQ8 biosynthetic pathway, xanthomonadin biosynthetic gene cluster and XanB2 homologues are well conserved in the bacterial species within Xanthomonas, Xylella, Xylophilus, Pseudoxanthomonas, Rhodanobacter, Frateuria, Herminiimonas and Variovorax, suggesting that XanB2 may be a conserved metabolic link between the shikimate pathway, ubiquinone and xanthomonadin biosynthetic pathways in diverse bacteria.
If you can't find a tool you're looking for, please click the link at the top of the page to "Go to old article view". Alternatively, view our Knowledge Base articles for additional help. Your feedback is important to us, so please let us know if you have comments or ideas for improvement.
Xanthomonas are widely distributed phytopathogens that infect nearly 400 different plant hosts, many of which are commercially important crops (Leyns et al., 1984). Originally identified in Xanthomonas campestris pv. campestris (Xcc), the diffusible factor (DF) plays a critical role in diverse biological processes that are crucial for bacterial survival and virulence, including biosynthesis of the yellow pigment xanthomonadin and maintenance of antioxidant activity (Poplawsky and Chun, 1997; He et al., 2011). More than 40 bacteria species within the genera of Xanthomonas, Xylella, Xylophilus, Streptomyces and Burkholderia have DF-like activities (Poplawsky and Chun, 1997; Poplawsky et al., 2005; He et al., 2011). Recent findings have shown that DF is 3-hydroxybenzoic acid (3-HBA) (He et al., 2011). The enzyme XanB2 is key to the production of the DF in the phytopathogen Xcc (Poplawsky et al., 2005; He et al., 2011). XanB2 homologues are found in a range of other bacterial species (He et al., 2011). XanB2 was predicted to be a pteridine-dependent deoxygenase-like protein in the phytopathogens Xanthomonas (da Silva et al., 2002), but its enzymatic substrates and catalytic mechanism have not yet been characterized.
Xanthomonadins are a unique group of halogenated, aryl-polyene, water-insoluble, membrane-bound yellow pigment produced by the phytopathogens Xanthomonas spp., Xylella fastidiosa and Xylophilus ampelinus (Stephens and Starr, 1963; Andrewes et al., 1976; Schaad and Stall, 1988; Bradbury, 1991). They are useful chemotaxonomic and diagnostic markers for Xanthomonas. In the rice bacterial pathogen Xanthomonas oryzae pv. oryzae and crucifer pathogen Xcc, xanthomonadins are known for their roles in protection of the pathogens from photobiological and peroxidation damages, in epiphytic survival on crucifer leaves, and in bacterial systemic infections (Jenkins and Starr, 1982; Rajagopal et al., 1997; Poplawsky et al., 2000; He et al., 2011). A pig cluster, which consists of about 20 open reading frames and may constitute part of a novel type II polyketide synthase pathway, is responsible for the xanthomonadin biosynthesis (Poplawsky and Chun, 1997; Goel et al., 2002; da Silva et al., 2002). Xanthomonadin biosynthesis was originally found to be regulated by an unknown DF (Poplawsky and Chun, 1997). Our recent findings suggest that DF is 3-HBA and it is likely a key intermediate in xanthomonadin biosynthesis (He et al., 2011).
Quinones are widely distributed in nature. Most Gram-positive bacteria and anaerobic Gram-negative bacteria contain only menaquinone, whereas the majority of strictly aerobic Gram-negative bacteria contain exclusively a benzoquinone termed ubiquinone or coenzyme Q (CoQ) ( and Poole, 1999; Meganathan, 2001). CoQ is a naturally occurring coenzyme formed by the conjugation of a benzoquinone ring with a hydrophobic isoprenoid chain of varying length (Cluis et al., 2007). It is an obligatory cofactor in the aerobic respiratory electron transfer for energy generation. It also acts as an antioxidant that protects membrane phospholipids and proteins from lipid peroxidation by scavenging free radicals directly and/or by regenerating levels of tocopherol (Lass and Sohal, 1998; Cluis et al., 2007; Martin et al., 2007). In the biosynthetic pathway the ‘nucleus’ of CoQ is derived from the shikimate pathway via chorismate in bacteria (Meganathan, 2001). The CoQ biosynthetic pathway in Escherichia coli has been studied extensively. It consists of at least nine enzymes, including UbiA, UbiB, UbiC, UbiD, UbiE, UbiF, UbiG, UbiH and IspB ( and Poole, 1999; Meganathan, 2001; Bentinger et al., 2010). The formation of 4-hydroxybenzoic acid (4-HBA) from chorismate is the first committed step, which is catalysed by chorismate-pyruvate lyase encoded by ubiC in E. coli, Salmonella and Pseudomonas aeruginosa, or by Rv2949c in Mycobacteria (Siebert et al., 1994; and Poole, 1999; Stadthagen et al., 2005). Genomic analysis of the current microbial genome sequence database revealed that many bacterial species lack the genes both ubiC and Rv2949c, suggesting that they may depend on an alternative (yet unknown) source of 4-HBA for CoQ production.
The shikimate pathway is widely present in plants, bacteria and fungi. Seven enzymes (AroABCDEFG) of the shikimate pathway catalyse sequential conversion of erythrose 4-phosphate and phosphoenol pyruvate to the pathway end-product, chorismate (Dewick, 1993; Herrmann and Weaver, 1999). Chorismate is the common precursor for the biosynthesis of a range of aromatic compounds, including phenylalanine, tyrosine, tryptophan, folate cofactors, phenazines, siderophores and salicylic acid (Floss, 1997; Herrmann and Weaver, 1999; Dosselaere and Vanderleyden, 2001; Van Lanen et al., 2008). Recently, a new family of enzymes have been found to act on chorismate in Streptomyces and two of them, Hyg5 and Bra8, have been shown to convert chorismate into 3-HBA (Andexer et al., 2011). In this study, we purified XanB2 and showed that it is a novel bifunctional enzyme, which converts chorismate into 3-HBA and 4-HBA using different catalytic domains. We further showed that 3-HBA and 4-HBA play specific and indispensable roles in Xcc metabolism and physiology. We also present evidence that XanB2 and its homologues may act as a conserved metabolic link in bacteria.
XanB2 acts downstream of the shikimate pathway in Xcc
Previous findings showed that XanB2, a putative pteridine-dependent deoxygenase-like protein, is a key enzyme for 3-HBA biosynthesis in Xcc (Poplawsky et al., 2005; He et al., 2011). However, the details of its enzymatic mechanism and the identity of the precursor for 3-HBA biosynthesis remain unclear. As 3-HBA is an aromatic compound, and the shikimate pathway links carbohydrates metabolism to the biosynthesis of aromatic compounds (Dewick, 1993; Herrmann and Weaver, 1999), we reasoned that the shikimate pathway might be required for XanB2 to synthesize 3-HBA. To verify this hypothesis, four key genes of the shikimate pathway, aroA (Xcc1591, 3-phosphoshikimate 1-carboxyvinyltransferase), aroC (Xcc2551, chorismate synthase encoding), aroE (Xcc3953, shikimate 5-dehydrogenase) and aroK (Xcc2841, shikimate kinase), were identified and were separately in-frame deleted in Xcc strain XC1. As expected, the resulting mutants were unable to produce 3-HBA and xanthomonadins (Fig. 1A and B). However, exogenous addition of 3-HBA to these mutants restored the xanthomonadin production (Fig. 1A and B), suggesting that the shikimate pathway is involved in the biosynthesis of 3-HBA and xanthomonadins.
To further explore the relationship between XanB2 and the shikimate pathway in Xcc, we generated the aroC and xanB2 double deletion mutant, ΔxanB2ΔaroC. In trans expression of aroC in the ΔaroC strain fully restored the production of 3-HBA and xanthomonadins (Fig. 1C). However, expression of aroC in the double deletion mutant ΔxanB2ΔaroC had no effect, suggesting that XanB2 acts at the downstream of the shikimate pathway.
XanB2 catalyses the synthesis of 3-HBA and 4-HBA from chorismate
As chorismate is the end-product of the shikimate pathway, we speculated that XanB2 might use chorismate as a precursor to synthesize 3-HBA. To test this hypothesis, XanB2 was expressed in E. coli strain BL21 and purified as described in the Experimental procedures section. Incubation of the purified XanB2 with chorismate led to formation of two products (Fig. 2C and D). Product (a) had the same elution time and UV absorbance as the standard 3-HBA in high-performance liquid chromatography (HPLC) analysis (Fig. 2). Further mass spectrometry (MS) analysis and pigment-induction analysis confirmed that product (a) is 3-HBA (Figs S1A and 2E).
In addition to 3-HBA, a new product (b) was also detected in the reaction mixture by HPLC analysis (Fig. 2C and D). MS analysis showed the molecular weight of product (b) to be 137 (Fig. S1C). Further nuclear magnetic resonance analysis suggested that product (b) is 4-HBA (Fig. S2), and this was further verified by comparing the H1 and C13 spectra of commercially available 4-HBA (Sigma) and purified product (b) (Fig. S2). For determination of kinetic parameters, the XanB2-catalysed reaction was directly quantified by HPLC. The Km value for chorismate was determined to be 95.3 ± 12.0 μM, and the Kcat values for 3-HBA and 4-HBA were calculated as 66.4 ± 4.7 s−1 and 173.3 ± 9.9 s−1 respectively.
To determine whether the 3-HBA and 4-HBA produced in the reaction mixture were directly derived from chorismate or indirectly from each other, only 4-HBA (but not chorismate) was added to the reaction mixture containing purified XanB2. After 60 min of incubation at 37°C, no 3-HBA was detected in the reaction mixture (Fig. S3B). Similarly, no 4-HBA was detected in the reaction mixture containing XanB2 and 3-HBA (Fig. S3C). These data indicate that the 3-HBA and 4-HBA produced in the reaction mixture were directly derived from chorismate.
Chorismate is relatively unstable and can undergo decomposition to 4-HBA non-enzymatically in Tris-HCl buffer at 37°C (Gibson and Gibson, 1964; Siebert et al., 1992). In the absence of XanB2, we consistently detected low levels of 4-HBA in the reaction mixture containing chorismate in 30–60 min (Fig. S4C and D). However, under the same conditions, 4-HBA production in the presence of XanB2 was around 10 times higher than in the absence of XanB2 (Fig. S4A and B), confirming that the majority of 4-HBA in the reaction mixture was derived from XanB2 enzymatic conversion of chorismate.
To further investigate whether XanB2 is responsible for the in vivo 3-HBA and 4-HBA production, we slightly modified a previously described method for extraction and purification of 3-HBA (He et al., 2011). By using the modified method, 3-HBA and 4-HBA were detected in the supernatants of Xcc culture at stationary growth phase (Fig. 3). Deletion of xanB2 completely abolished the production of 3-HBA and 4-HBA, which could be restored by overexpression of xanB2 (Fig. 3). We then analysed the time-course of 4-HBA production during growth, the results showed that Xcc initiated 4-HBA production during the exponential growth phase, and the highest level of production (∼32.0 μM) was observed at stationary phase (Fig. 4A). This is similar to the growth-dependent pattern of 3-HBA production (He et al., 2011).
4-HBA is associated with antioxidant activity
In previous studies, the xanB2 deletion mutant showed pleiotropic phenotypes, including deficiency for xanthomonadin production, decreased antioxidant activity and reduced virulence in Chinese radish (He et al., 2011). Exogenous addition of 3-HBA could fully restore xanthomonadin production and partially restore H2O2 resistance (He et al., 2011). In this study, by using the ΔxanB2 strain as a reporter, and following the assay methods developed previously for 3-HBA (He et al., 2011), we found that lower concentrations of 4-HBA (3.1–12.5 μM) had no obvious effect on xanthomonadin production and higher concentrations of 4-HBA (25.0–50.0 μM) induced darker yellow pigment production in the mutant ΔxanB2 (Fig. 4B). However, when a mixture of 3-HBA (6.2 μM) and 4-HBA (25.0 μM) was applied, the colour of the pigment induced was similar to that of wild-type strain (Fig. 4B), suggesting that, in the presence of both 3-HBA and 4-HBA, Xcc prefers 3-HBA for xanthomonadin biosynthesis.
Previous results showed that the exogenous addition of 3-HBA could partially restore H2O2 resistance of ΔxanB2 strain (He et al., 2011). In this study, addition of 4-HBA at 25 μM also partially restored H2O2 resistance to the mutant ΔxanB2, while addition of 4-HBA and 3-HBA (25.0 and 6.2 μM respectively) together conferred full H2O2 resistance to ΔxanB2 (Fig. 4C).
XanB2 is involved in ubiquinone biosynthesis
Ubiquinones have been found to play multiple roles in respiration, gene regulation and oxidative stress management ( and Poole, 1999). The quinine structure has isoprenoid side-chains of various lengths depending on the species. Humans and tobacco plants have quinines with 10-unit isoprenoid side-chains, whereas rodents, E. coli and yeast mainly have 9-,8-,6-unit side-chains respectively ( and Poole, 1999; Cluis et al., 2007). We purified ubiquinone from Xcc cells following a previously described method (Cheng et al., 2011). HPLC and MS analysis showed that the ubiquinone produced by Xcc is coenzyme Q8 (CoQ8) (Figs 5A and B and S5).
Xcc genome contains the homologues of the genes required for E. coli CoQ8 biosynthesis: ubiA (Xcc0390), ubiB (Xcc0224), ubiE (Xcc3488), ubiF (Xcc0799), ubiG (Xcc2269), ubiH (Xcc0798) and ispB (Xcc2373), but not the homologue of ubiC or Rv2949c encoding chorismate-pyruvate lyase (Fig. 5C). The finding that XanB2 is responsible for in vivo 3-HBA and 4-HBA production (Fig. 3) suggests that Xcc may depend on XanB2 as an alternative 4-HBA source for CoQ8 production. Consistent with this notion, deletion of xanB2 almost abolished CoQ8 production (Fig. 5B) and overexpression of xanB2 or exogenous addition of 4-HBA restored CoQ8 production (Fig. 5B). In contrast, exogenous addition of 3-HBA had no effect on CoQ8 production (Fig. 5B).
XanB2 is a structurally unique chorismate-utilizing enzyme
Chorismate-pyruvate lyase encoded by ubiC or Rv2949c is the only known enzyme that converts chorismate to 4-HBA for CoQ biosynthesis. It is a 165-amino-acid protein with a common Chor_lyase domain (Figs 6A and S6A). XanB2 does not contain a Chor_lyase domain and shares no significant sequence similarities to either UbiC or Rv2949c. Unlike XanB2, which is responsible for the production of 3-HBA and 4-HBA and the concomitant biosynthesis of yellow pigment and CoQ8 (Fig. 3), overexpression of ubiC from E. coli in the mutant ΔxanB2 resulted in only the production of 4-HBA (Fig. S6C). This highlights the functional similarity and dissimilarity between XanB2 and UbiC.
Domain analysis revealed a YjgF-like domain at the C-terminal of XanB2 (221–335 aa), which is present in bacteria, archaea and eukaryotes, and has no definitive function. The YjgF-like domain has recently been found in the chorismate-utilizing proteins Hyg5, RapK and FkbO from Streptomyces (Fig. 6A). Hyg5 has recently been shown to catalyse the conversion of chorismate into 3-HBA, while FkbO hydrolyses chorismate to form (4R,5R)-4, 5-dihydroxycyclohexa-1,5-dienecarboxylic acid (Andexer et al., 2011). In order to understand why XanB2 can synthesize both 3-HBA and 4-HBA while Hyg5 only synthesizes 3-HBA, we did a sequence alignment using FkbO, RapK, Hyg5, XanB2 and the XanB2 homologues from X. oryzae pv. oryzae (XOO) and X. fastidiosa (XF) (Fig. S7). Based on sequence similarity, conserved motifs and functionality analysis as described in the following section, XanB2 and its homologues were divided into three putative domains (Fig. 6A). These enzymes were found to share a common C-terminal domain (221–335 aa, domain III) and a conserved middle region (126–228 aa, domain II) (Figs 6A and S7). However, the most dissimilar region is the N-terminal domain of XanB2 (1–125 aa, domain I), which shares little sequence similarity to Hyg5, RapK or FkbO (Figs 6A and S7). Furthermore, we performed phylogenetic analysis of the above proteins and the results showed that XanB2 is divergent from UbiC, Hyg5, RapK and FkbO (Fig. 6B). Taken together, these findings suggest that XanB2 is structurally and functionally different from the known 3-HBA- or 4-HBA-producing enzymes.
Different domains of XanB2 are required for 3-HBA and 4-HBA production
Domain complementation and point mutation analysis of XanB2 was conducted to understand how XanB2 can synthesize both 3-HBA and 4-HBA. First, single domains I, II and III of XanB2 were respectively overexpressed in the ΔxanB2 strains. No 3-HBA or 4-HBA was detected in the bacterial cells or in the supernatant of the cell culture. Then, two fusion domains (domains I–II and II–III) were respectively expressed in the ΔxanB2 strain. Although 7.2 μM 3-HBA and 35.4 μM 4-HBA were detected in the supernatants of the xanB2-overexpressing strain, only a low level of 3-HBA (0.7 μM) was detected in the supernatant of the ΔxanB2 strain overexpressing the fusion domain II–III (Fig. 7B). Similarly, the ΔxanB2 strain overexpressing the fusion domain I–II produced only 4-HBA (2.9 μM) in the cell culture. Overexpression of the fusion domain II–III was found to restore pigment production, but it took 7 days for this overexpression strain to produce visible yellow pigment; it took only 2 days in the xanB2-overexpressing strain.
Multiple sequence alignment analysis revealed a range of conserved motifs and amino acids in XanB2 and its homologues (Figs 7A and S7). Single point mutations of the conserved G237, L297 or E329 in domain III significantly decreased the enzymatic activity for 3-HBA production, but not for 4-HBA production (Fig. 7C). Point mutations of P169, E198 and F219 in domain II significantly affected the enzymatic activity for both 3-HBA and 4-HBA production (Fig. 7C). An ‘FGFG’ motif in domain I was found to be conserved among XanB2 and its homologues in Xoo and Xf, but not found in Hyg5, RapK or FkbO (Figs 7A and S7). Substitution of either F40 or G43 in the ‘FGFG’ motif with a different amino acid significantly decreased the enzymatic activity for the production of 4-HBA, but not for 3-HBA (Figs 7C and S7), suggesting that the unique domain I of XanB2 is responsible for 4-HBA biosynthesis.
XanB2 and XanB2-associated metabolic pathways are widely conserved
In this study, by blasting the latest Nr database in NCBI, we found that XanB2 homologues (30% amino acid identity as cut-off value) are not only conserved in all species of plant pathogens Xanthomonas spp. and X. fastidiosa, but also widely presented in the bacterial species belonging to the genera of Pseudoxanthomonas, Rhodanobacter, Frateuria, Nitrosococcus, Acidithiobacillus, Pseudoglubenkiania, Variovorax, Streptomyces, Burkholderia, Sideroxydans, Methylotenera, Herminiimonas, Dechloromonas, Thiorhodospira, Candidatus, Methylococcus, Thibacillus, Halothiobacillus, Thioalkalivibrio, Thermobacillus and Methylibium (Table S1). The majority of these bacterial species are aerobic and Gram-negative. Xanthomonas, Xylella, Pseudoxanthomonas, Rhodanobacter and Frateuria are the members of the family Xanthomonadaceae (Gamma-proteobacteria), which have diverse metabolic potentials (Finkmann et al., 2000; Mergaert et al., 2002).
By using the essential genes for CoQ8 biosynthesis in E. coli as templates for blast analysis, we found that all Xanthomonas spp., X. fastidiosa, Pseudoxanthomonas spadix BD-a59, Pseudoxanthomonas suwonensis 11-1, Rhodanobacter sp. 2APBS1, Frateuria aurantia DSM 6220, Nitrosococcus spp., Sideroxydans lithotrophicus ES-1, Thiorhodospira sibiricaATCC700588, Herminiimonas arsenicoxydans and Variovorax paradoxus contain the homologues of xanB2, ubiA, ubiB, ubiE, ubiF, ubiG and ubiH, but not those of ubiC or Rv2949c (Fig. S8), suggesting that these bacterial species are also dependent on XanB2 for ubiquinone biosynthesis.
The pig cluster is responsible for xanthomonadin biosynthesis in Xanthomonas (Poplawsky et al., 1993; Goel et al., 2002). Genomic analysis revealed that pig-like clusters are also present in X. fastidiosa, P. spadix BD-a59, P. suwonensis 11-1, Rhodanobacter sp. 2APBS1, F. aurantia DSM 6220, H. arsenicoxydans and V. paradoxus (Fig. S9). Consistent with the conservation of the pig cluster, these bacterial species are yellow-pigmented (Table S1). Thus, xanthomonadins may play a key role in the yellow pigmentation of these bacterial species.
The results of this study demonstrate that XanB2 is a novel and unique bifunctional chorismatase that hydrolyses chorismate to produce 3-HBA and 4-HBA. This finding adds a new member to the list of chorismate-utilizing enzymes found in nature. A range of chorismate-utilizing enzymes, such as chorismate mutase, anthranilate synthase, salicylate synthase, 4-amino-4-deoxychorismate synthase, 2-amino-2-deoxychorismate synthase, isochorismate synthase, chorismate-pyruvate lyase, and the recently identified enzymes Hyg5, FkbO and RapK, have been documented (Floss, 1997; Dosselaere and Vanderleyden, 2001; Van Lanen et al., 2008; Andexer et al., 2011). XanB2 is functionally related to the following three types of chorismatase. The first type includes UbiC and Rv2949c, which utilize chorismate to produce 4-HBA. UbiC was identified in a range of bacterial species, including E. coli, Salmonella and Shigella, whereas Rv2949c was found in Mycobacterium tuberculosis (Siebert et al., 1994; Stadthagen et al., 2005). The second type of chorismatase is represented by Hyg5 and Bra8 from Steptomyces species. It converts chorismate into 3-HBA (Andexer et al., 2011). The third type of chorismatase, consisting of FkbO and RapK both identified in Streptomyces species, hydrolyses chorismate to form (4R,5R)-4,5-dihydroxycyclohexa-1,5-dienecarboxylic acid (Andexer et al., 2011). XanB2 represents a fourth type of chorismatase that converts chorismate into 3-HBA and 4-HBA simultaneously (Figs 2, 3 and 6). Further analysis suggests that XanB2 is a unique and bifunctional enzyme and that its putative domain I and domain III affect the enzymatic activity for 4-HBA and 3-HBA biosynthesis respectively (Fig. 7). Chorismate-pyruvate lyase is only known characterized enzyme that converts chorismate into 4-HBA. The enzyme cleaves a carbon–oxygen bond of chorismate to release pyruvate and aromatize the ring. The reaction starts with an initial abstraction of the C4-H of chorismate followed by the loss of the C3-enolpyruvyl group (Walsh et al., 1990; Holden et al., 2002). As XanB2 has no sequence homology with chorismate-pyruvate lyase encoded by ubiC or Rv2949c, it remains unclear whether XanB2 uses a similar strategy to convert chorismate into 4-HBA. The mechanic details of 3-HBA conversion remain to be elucidated. Three proteins, including the XanB2 described in this study, the Hyg5 and Bra8 from Streptomyces (Andexer et al., 2011), have been shown to convert chorismate into 3-HBA. The C-terminal domain of these proteins bears a strong structural resemblance to the YjgF/YER057c/UK114 protein superfamily, highly conserved among eubacteria, archaea and eukaryotes. 3D structures have been solved for several of these proteins, but their biochemical functions remain ill-defined (Burman et al., 2007). How the three domains work together to catalyse 3-HBA and 4-HBA biosynthesis will be the subject of further investigations.
XanB2 is widely conserved in a range of bacterial species of different genera (Table S1). These bacteria are either economically important phytopathogens, or have diverse industrial potentials. For example, Xanthomonas spp. and X. fastidiosa infect rice, cotton, soybean, banana and citrus and cause serious losses in yield (Hayward, 1993). Burkholderia ambifaria AMMD is a rhizosphere organism effective in stimulating plant growth and suppressing soil-borne plant pathogenic Pythium spp. and Aphanomyces euteiches (Heungens and Parke, 2001). P. spadix BD-a59 and Dechloromonas aromatica RCB are able to degrade aromatic industrial polluters: benzene, toluene, ethylbenzene and xylene (Coates et al., 2001; Kim et al., 2008). Methylotenera versatilis 301 and Methylibium petroleiphilum PM1 are naturally methylotrophic isolates and are capable of using methyl tert-butyl ether as a sole source of carbon (Hanson et al., 1999; Kalyuzhnaya et al., 2012). Acidithiobacillus ferrooxidans is a Gram-negative rod-shaped acidophilic bacterium that is commonly found in deep caves or acid mine drainage, such as coal waste where they convert insoluble metals to their soluble state (Merson, 1992). These bacteria have been utilized in industrial bioleaching efforts to extract otherwise unobtainable metals (Merson, 1992; Yang et al., 2007). H. arsenicoxydans is a species of ultramicrobacteria. It was isolated from industrial sludge and is able to oxidize the toxic chemical element arsenic (2006). S. lithotrophicus ES-1, Candidatus accumulibacter phosphatis str. UW-1 and Thioalkalivibrio sulfidophilus HL-EbGr7 are able to grow at the extreme environments and have diverse metabolic potentials (Emerson and Moyer, 1997; Fukushima et al., 2007; Janssen et al., 2009). Identification of the biological roles of XanB2 homologues in these bacterial species will advance our understanding of the molecular mechanism of adaptation to extreme environments and may help researchers design more potent industrial strains.
Ubiquinone plays a role in several essential cell processes including respiration, gene regulation and oxidative stress management. The majority of strictly Gram-negative bacteria produce exclusively CoQ through a highly conserved CoQ biosynthetic pathway ( and Poole, 1999). The formation of 4-HBA from chorismate is the first committed step in this process and chorismate-pyruvate lyase encoded by ubiC or Rv2949c is the only known enzyme to catalyse the conversion (Siebert et al., 1994; and Poole, 1999; Meganathan, 2001; Stadthagen et al., 2005). The present study shows that XanB2 functions as UbiC to convert chorismate into 4-HBA for CoQ8 biosynthesis in Xcc (Fig. 5). To the best of our knowledge, this is the first report showing an alternative chorismate lyase-like enzyme involved in 4-HBA production and CoQ biosynthesis. This represents a new mechanism for ubiquinone biosynthesis. XanB2 homologues and similar CoQ biosynthetic pathways lacking UbiC are present in a wide range of bacterial species belonging to different genera (Fig. S8), suggesting that XanB2-dependent 4-HBA production and CoQ biosynthesis is a conserved mechanism in bacteria.
The present study also shows that XanB2 links the shikimate pathway to CoQ8 biosynthesis and xanthomonadin biosynthesis in Xcc (Fig. 8). XanB2 homologues, similar xanthomonadin biosynthetic gene clusters, the new ubiquinone biosynthetic pathways and the shikimate pathway are also found in other Xanthomonas spp., X. fastidiosa, P. spadix BD-a59 and P. suwonensis 11-1, Rhodanobacter sp. 2APBS1, F. aurantia DSM 6220, Nitrosococcus spp., S. lithotrophicus ES-1, T. sibiricaATCC700588, H. arsenicoxydans and V. paradoxus (Figs S8 and S9). All these strains produce membrane-bound yellow or purple pigments, suggesting that the pigment biosynthesis might be regulated in the same XanB2-dependent manner in these bacterial species. However, not all XanB2-containing bacterial species produce yellow pigment and encode xanthomonadin biosynthesis clusters (Table S1; Fig. S9). It will be interesting to investigate 3-HBA- or 4-HBA-associated biological functions in these non-pigmented bacterial species. Recently, Hyg5- or Bra8-derived 3-HBA has been proposed for involvement in the biosynthesis of BC325, a 3-HBA-containing analogue of the immunosuppressant rapamycin, or brasilicardin in Streptomyces hygroscopicus (Andexer et al., 2011). Considering that XanB2 homologues are widely conserved in the bacterial species with diverse ecological niches (Table S1), it is tempting to speculate that more novel 3-HBA- and 4-HBA-associated metabolisms could be unveiled in future studies.
Xanthomonas spp. and X. fastidiosa are plant pathogens and are normally restricted to the xylem tissues or intercellular spaces of infected plants. Increasing production of reactive oxygen species, including superoxide, H2O2 and OH, is associated with aerobic respiration and with active plant defence responses (Sutherland, 1991; Bestwick et al., 1997; Bindschedler et al., 2006). The ability of these pathogens to survive oxidative stress is therefore of critical importance for successful colonization in host plants. The membrane-bound yellow pigment xanthomonadins have been shown to play essential roles in protecting bacteria from photobiological and peroxidative damage, and are important for epiphytic survival and systemic infection in Xanthomonas species (Jenkins and Starr, 1982; Rajagopal et al., 1997; Poplawsky et al., 2000; He et al., 2011). Microbial ubiquinones also play roles in oxidative stress management. E. coli mutants defective in ubiquinone biosynthesis show sensitivity to chlorate, thiols, visible light and oxidative stress induced by H2O2, paraquat, γ-irradiation or methylmethanesulfonate ( and Poole, 1999). In the phytopathogens Xanthomonas and X. fastidiosa, both xanthomonadin and CoQ biosynthesis are linked together by a single metabolic enzyme, XanB2, which may confer a competitive advantage in survival and systemic infections inside the host plants. Consistent with this notion, the XanB2 deletion mutant, which is deficient for pigment and CoQ8 production, was more impaired in H2O2 resistance, systemic invasion and virulence in Chinese radish than the pigment-deficient mutant Xcc4015 (He et al., 2011).
Bacterial strains and growth conditions
Xcc strain XC1 was grown at 28°C in NYG medium (l-1) (5 g of peptone, 3 g of yeast extract, 20 g of glycerol, pH 7.0), or YEB medium (l-1) (5 g of yeast extract, 10 g of tryptone, 5 g of sodium chloride, 5 g of sucrose, 0.5 g of MgSO4) in the darkness. E. coli strains were grown at 37°C in LB medium. Antibiotics were added at the following concentrations when required: kanamycin (Km) 50 mg l−1, rifampicin (Rif) 25 mg l−1, ampicillin (Amp) 100 mg l−1, tetracycline (Tet) 5 mg l−1.
Gene deletion, functional complementation and XanB2 point mutation analysis
Xcc strain XC1 in-frame deletion mutants were generated using the methods described previously (He et al., 2006). Primers were listed in Table S2. For complementation analysis, the coding region of each gene was amplified by PCR (see primers in Table S2) and cloned into the expression vector pLAFR3 or pBBR1MCS (Kan) (Kovach et al., 1995). The resulting constructs were transferred to Xcc strains through triparental mating.
For point mutation analysis, xanB2 was first cloned to pGEM-T-Easy vector (Promega) and then point mutation was conducted using the primers listed in Table S2 following the protocol described in the QuikChange Site-Directed Mutagenesis Kit. After confirmation by sequencing, the mutated xanB2 was further cloned to pBBR1MCS (KanR) (Kovach et al., 1995). The resulting constructs were transferred to the ΔxanB2 strain through triparental mating. 3-HBA and 4-HBA production in the supernatants of the ΔxanB2 strain expressing mutated XanB2 was assayed by the method described in the following sections.
Quantification of pigment production
Pigments were extracted from Xcc using previously described procedures (He et al., 2011). Pigment was quantified using the absorbance (optical density at 445 nm) of the crude pigment extracts as described previously (Poplawsky and Chun, 1997).
Purification and structural analysis of 4-HBA
4-HBA extraction and purification were performed using the method described by He et al. (2011) with minor modifications. Briefly, the culture supernatants and reaction mixtures were adjusted to pH 4.0 by addition of 1 M HCl prior to extraction twice with an equal volume of ethyl acetate. The ethyl acetate fractions were collected and the solvent was removed by rotary evaporation to dryness at 40°C. The residues were then dissolved in methanol. The crude extract was subjected to flash column chromatography using a silica gel column (12 × 150 mm, Biotage Flash 12M cartridge), eluted with ethyl acetate–hexane (25:75, v/v, 0.05% acetic acid). The collected active component was then applied to HPLC on a C18 WAT094240 Symmetry column (4.6 × 150 mm, Waters, USA), eluted with methanol–water (30:70, v/v, 0.05% formic acid) at a flow rate of 1 ml min−1 in a Waters 2695 system with 996 PDA detectors. The elute containing the purified 3-HBA or 4-HBA was used for ESI-MS analysis on a Finnigan LCQ system. The purified 4-HBA was dissolved in methanol-d4 solution and 1H and 13C spectra were recorded on a Bruker DRX400 400 MHz spectrometer at room temperature.
Bioassay and quantification analysis of 3-HBA and 4-HBA
An agarose plate assay for 3-HBA and 4-HBA was conducted using a method described previously (He et al., 2011). 3-HBA and 4-HBA production was quantified using peak area in HPLC elute. The commercially available 3-HBA (Sigma) and 4-HBA (Sigma) were used as standards.
XanB2 purification and in vitro enzymatic activity assay
The coding region of XanB2 was amplified with the primers listed in Table S2 and fused in frame with the GST coding region in the expression vector pGEX-6p-1. The fusion construct was transformed into E. coli strain BL21. The resultant bacterial strain was inoculated separately in 200 ml LB medium supplemented with 100 μg ml−1 carbenicillin, and grown at 37°C with shaking at 200 r.p.m. until OD600 = 0.6. Expression of GST-fusion proteins was induced by addition of IPTG at a final concentration of 0.1 mM and the bacterial cultures were then incubated overnight at 18°C. The bacterial cells were collected by centrifugation, resuspended in 50 ml of PBS lysis buffer supplemented with one tablet of the CompleteTM Protease Inhibitor Cocktail (Roche). The bacterial cells were then lysed by sonication and the debris was removed by centrifugation at 14 000 r.p.m. for 30 min. The supernatants were loaded to an affinity column containing Glutathione Sepharose 4B beads for affinity binding, and the column was washed with five volumes of PBS buffer. The on-column GST-fusion protein cleavage with PreScission Protease (GE healthcare, 2 units/100 μg of bound proteins) was conducted at 4°C overnight. The cleaved fusion proteins were eluted and the purity was determined by SDS-PAGE analysis.
For in vitro analysis, 20 μg of purified XanB2 was dissolved in 0.2 ml of Tris buffer (Tris-HCl, 50 mM, pH 7.4) containing 100 μM chorismate in a plastic tube. The reaction mixture was incubated at 37°C for 60 min. The reaction was stopped by placing the plastic tube containing reaction mixture in boiling water for 5 min. 3-HBA and 4-HBA were extracted, purified and analysed by using HPLC described above.
H2O2 sensitivity testing
Xcc strains were grown in YEB medium with and without 4-HBA to an OD600 of 2.1. The cultures were then diluted with fresh YEB liquid medium to an OD600 of 0.1. Bacterial suspensions were treated with 880 μM H2O2 for 30 min. Bacterial cells were then spun down and resuspended in the same volume of YEB medium. Cells were serially diluted and spotted onto YEB agar plates, and then incubated at 28°C for 2 days before counting colony-forming units. The experiment was repeated three times.
Time-course of 4-HBA production during the growth
To determine the role of 4-HBA in Xcc, we first determined the dynamics of 4-HBA production during growth. Xcc strain XC1 was grown in YEB medium and 50 ml of culture was sampled at different points in time. The 4-HBA in cell-free supernatants was extracted and analysed by using HPLC analysis.
Ubiquinone extraction, purification and HPLC analysis
Ubiquinone was extracted as previously described with slight modifications (Cheng et al., 2011). Briefly, the Xcc cultures were grown in NYG medium until the stationary phase (OD600 = 2.0) and the cells from 50 ml of cultures were collected by centrifugation. Cell pellets were resuspended in 2 ml of 0.2 M acetate buffer (pH 5.6) after washing twice by 1× PBS. Cryo-impacted cell homogenate was subjected to sonication for a total of 90 s. Five millilitres of hexane–acetone (1:1, v/v) reaction mixture was added to the cell homogenate followed by sonication and vortexing. The hexane fractions were collected and the solvent was removed by rotary evaporation at 40°C to dryness. The residue was then dissolved in 1.5 ml of chloroform–methanol (1:1, v/v), followed by washing with 0.75 ml of 0.7% NaCl. The chloroform fraction was collected and the solvent was removed by rotary evaporation at 40°C to dryness. The residue was dissolved in 100 μl of methanol. The crude extract was further analysed by a liquid chromatographic system (Agilent 1260 Infinity, Agilent Technologies). The elute containing the purified ubiquinone was subjected to ESI-MS analysis on an Agilent 6230 TOF MS equipped with a Jet Stream ESI source (Agilent Technologies), using 100% methanol containing 1.5‰ formic acid as mobile phase, at a flow rate of 0.2 ml min−1. Data were acquired under the control of Mass Hunter workstation software.
We thank Mr Maolong Wei and Ms Wei Zhang for technical help with HPLC and MS analysis. This work was supported by a grant from National Natural Science Foundation of China (No. 31272005) and a research grant for Returned Scholars from Shanghai Jiao Tong University (No. WS3107208008). L.Z. was supported by a grant from China Postdoctoral Science Foundation (No. 2011M500770).
Y.W.H. and L.H.Z. designed the experiments and wrote the manuscript; L.Z. and J.Y.W. performed most of the experiment; A.R.P. participated in designing the experiment and writing the manuscript; S.L., B.Z., J.W., C.C. and T.Z. performed some of the experiments and analysed the data.
Conflict of interest
The authors declare that they have no conflict of interest.