INTRODUCTION

Top Abstract Introduction Materials & Methods Results Discussion References

Many prokaryotes produce extracellular polysaccharides that have important roles in pathogenesis (8). In the plant pathogen Ralstonia solanacearum, extracellular polysaccharides (EPS) are major virulence factors required to cause the agriculturally important disease bacterial wilt (15). Although the exact role of EPS has not been demonstrated, it may interfere with water transport in the plant by plugging the xylem vessels, leading to wilt (20).

Several genes involved in EPS production in R. solanacearum have been identified. Structural gene clusters include opsI (7, 22), opsII (29), rgnII (10), and epsI (10). The opsI and opsII gene clusters are important for both EPS and lipopolysaccharide syntheses since mutations in them affect the production of both macromolecules. The opsI cluster consists of at least seven genes, some of which are important for nucleotide sugar synthesis (7, 22). The rgnII cluster is largely uncharacterized since it is required for EPS production only in culture, not in plants (10). Mutational analyses of epsI suggest that it encodes proteins responsible for synthesis of the acidic component of EPS, which is absolutely required for R. solanacearum infection of plants (10, 23, 32).

Sequence analysis revealed that epsI encodes polypeptides transcribed from a single promoter (18). Regulation of epsI is complex, involving at least seven proteins, including the highly basic XpsR, which likely affects expression of epsI directly (17). The two-component regulatory systems VsrB-VsrC (19) and VsrA-VsrD (37) have been shown genetically to positively regulate epsI and xpsR expression, respectively. The LysR-like transcriptional regulator PhcA has been demonstrated to positively regulate xpsR expression (4, 5, 17). Finally, EpsR overexpressed in plasmids of four to six copies per cell can specifically reduce synthesis from the epsI promoter, decreasing EPS production from colonies (16, 21, 29). EpsR has sequence similarity to effector proteins of two-component regulatory systems (21). A thorough Tn5::lacZ mutagenesis of R. solanacearum selecting for EPS-defective strains resulted in the identification and lacZ tagging of 12 complementation groups of EPS genes (29). Overexpression of EpsR affected the expression of only epsI::lacZ genes, suggesting a specific interaction between EpsR and the epsI promoter. In this work, we provide evidence that the chromosomal copy of epsR encodes a positive regulator of epsI. An epsR insertional mutation also reduced the virulence of R. solanacearum. We show that a mutation at the putative phosphorylation site renders EpsR unable to regulate synthesis of epsI and prevents an EpsR-dependent gel mobility shift of the epsI promoter.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results Discussion References

Growth and maintenance of bacterial strains. The strains and plasmids used in the experiments described herein are listed in Table 1. R. solanacearum was routinely cultured in CPG medium (per liter, 10 g of tryptone, 5 g of glucose, 1 g of Casamino Acids, 1 g of yeast extract, and 15 g of agar as appropriate) at 30°C. Antibiotics, where used, were at the following concentrations: kanamycin, 50 µg/ml; tetracycline, 15 µg/ml; ampicillin, 100 µg/ml; streptomycin, 25 µg/ml.

Molecular techniques. Plasmids were isolated from Escherichia coli by using Qiagen (Chatsworth, Calif.) columns. Chromosomal DNA isolation and Southern hybridizations were done as previously described (21). To transform R. solanacearum, 2-ml cultures at an optical density at 600 nm (OD₆₀₀) of approximately 1.0 were washed three times with sterile water and finally resuspended in 100 µl of water. The cells were then electroporated with 0.5 µg of plasmid with a Gene Pulser (Bio-Rad, Hercules, Calif.) set at 25 µF with a field strength of 6,000 V/cm. After electroporation, the cells were incubated for 3 h in CPG broth before being plated onto selective medium.

Site-directed mutagenesis. The aspartate at residue 47 of EpsR was changed to alanine by amplifying epsR in two halves by using PCR with pGepsR as a template. The 5' half of epsR was generated with the following primers: PUC19PC (5' GCCTGCAGGTCGACTCTAG 3'), which hybridizes with sequence in the plasmid vector of pGepsR, and epsRD-A5' (5' CAGCGGCTGCGTGGGCGGCAAG 3'), which hybridizes to nucleotides (nt) 486 to 508 of epsR (21). The 3' half of epsR was amplified by using epsRD-A3', which hybridizes to nt 500 to 523 and contains an AlwNI site (5' CAGCCGCTGAACTGGCCGTGATC 3'), and epsR3' EcoRI, which hybridizes to nt 1125 to 1142 and contains an EcoRI site (5' GAATTCCCGCGACGCGACAGCGCG 3'). The PCR products encoded by the 5' and 3' halves of epsR were individually cloned into PCRII (Novagen, Milwaukee, Wis.), creating TAepsRD-A5' and TAepsRD-A3' respectively. Inserts from TAepsRD-A5' and TAepsRD-A3' were released with HindIII-AlwNI and EcoRI-AlwNI digestions, respectively. The fragments were ligated to pBSKS⁺ linearized with EcoRI and HindIII to reconstitute epsR encoding the amino acid change at residue 47, creating pBepsRD-A. Clones which contained epsRD-A mutations were verified by screening for the presence of the AlwNI site. Since ampicillin selection is ineffective in R. solanacearum K60, we cloned the  fragment (encoding streptomycin resistance) into the HindIII site in the polylinker region of pBepsRD-A, creating pepsRD-A. This plasmid was then transformed by electroporation into the epsI mutant S49, selecting for streptomycin resistance. To investigate the effect of overexpression of EpsRD-A in R. solanacearum, the EcoRI-HindIII fragment from pepsRD-A was cloned into the EcoRI and HindIII sites of pLAFR3, creating pepsRD-A.

In vivo labeling and Western blot analysis. Strains grown in 2 ml of CPG broth to an OD₆₀₀ of 0.5 were washed three times in low-phosphate medium (M9 medium containing 30 µM Na₂HPO₄ and KH₂PO₄) (36) and then resuspended in 2 ml of low-phosphate medium and incubated at 30°C for an additional 4 h. One-half millicurie of orthophosphate (200 mCi/mmol; Amersham, Arlington Heights, Ill.) was added to 1-ml aliquots of cells. Orthophosphate can freely diffuse into bacterial cells, be incorporated into nucleotides, and subsequently be used as a substrate for phosphorylation. After a 20-min incubation at 30°C, the samples were washed twice with sterile water and resuspended in 60 µl of 1× Laemmli loading dye (27). Ten microliters of each sample was loaded on a sodium dodecyl sulfate-10% polyacrylamide gel. After electrophoresis, the gel was wrapped in plastic wrap and exposed to X-ray film for 1 h at 80°C. After autoradiography, the gel was washed extensively with 1× Western transfer buffer (10% methanol, 200 mM glycine, 25 mM Tris [pH 8.3]). The gel was blotted onto a nitrocellulose membrane and probed with anti-EpsR antibodies as previously described (21).

Gel mobility shift. DNA probes used for gel mobility shift assays contained a 240-bp fragment (nt 140 to +100) of the epsI promoter. The 240-bp fragment was produced by digesting pTAepsI with EcoRI and then end labeling with Klenow polymerase, [-³²P]dATP (3,000 Ci/mmol), and dTTP. In competition experiments, a 320-bp fragment containing the opsG promoter was made by PCR as previously described (21). epsI promoter used in competition experiments was also synthesized by PCR by using K60 chromosomal DNA as the template and the following primers: epsI 5', which hybridizes to nt 1 to 20 (5' GAATTCTCTGTCGAATTGGG) and epsI 3', which hybridizes to nt 218 to 238 (5' GGATCCGCTTACGAACATGAATGCG 3') (18). Protein extracts were made from 50 ml of R. solanacearum culture grown to an OD₆₀₀ of approximately 1.0. The cells were harvested, washed once in extraction buffer (50 mM Tris-HCl [pH 7.9], 10 mM EDTA, 10% glycerol, 10 mM KCl, 1 mM dithiothreitol, 0.5 mM sodium pyrophosphate, 0.4 mg of phenylmethylsulfonyl fluoride per ml), and finally resuspended in 1 ml of extraction buffer. Lysates were prepared by sonication with three 10-s bursts with a model 50 Sonic Dismembrator (Fisher Scientific, Pittsburgh, Pa.) set to 30% output. Protein concentrations were determined by the Bradford assay with bovine serum albumin as the standard (3). DNA binding reaction components, including 50-µg protein extracts, competitor DNAs (when appropriate), and binding buffer [50 mM Tris-HCl (pH 7.9), 5 mM MgCl₂, 30 mM KCl, 12% glycerol, 0.5 µg of poly(dI-dC) per ml, 1 mM dithiothreitol] were incubated for 15 min at room temperature before the addition of 5 nM end-labeled epsI promoter. Reaction mixtures were incubated for a further 20 min at room temperature before electrophoresis in a 3.5% (wt/vol) polyacrylamide (Tris-glycine [pH 7.9]) native gel at 70 V for 12 h at 4°C.

	RESULTS

Top Abstract Introduction Materials & Methods Results Discussion References

EpsR can act as both a positive and negative regulator of epsI expression. We previously determined that overexpression of EpsR in plasmids of four to six copies per cell resulted in decreased expression from the epsI promoter (29). To examine the effect of inactivation of the chromosomal copy of epsR, we inserted an  cassette carrying streptomycin resistance into the unique SmaI site in epsR contained in PET11epsR, creating pepsR. This plasmid contains a ColEI replication origin which is not utilized in R. solanacearum. Exchange of epsR:: with the chromosomal copy of epsR was made in the wild type, K60, and the EPS mutants S49, S50, S70, and S112 by homologous recombination. The homologous integration of epsR:: was checked by Southern blotting by using labeled epsR DNA as a probe (data not shown). Mutations in epsR were obtained in several strains, S70 and S49, which have epsI::lacZ fusions, S50, which has an opsI::lacZ fusion, S112, which has a vsrB::lacZ or vsrC::lacZ fusion, and finally S90, which contains a rgnII::lacZ fusion (29).

View larger version (14K): [in this window] [in a new window]   FIG. 1.   (A) Effect of inactivation of EpsR on expression from different EPS genes. Cells used were grown to an OD600 of 1.5 and then assayed as described previously (29, 30). The bars represent an average of three independent trials. (B) Effect of the epsR mutation on expression from the epsI promoter at different cell densities. Cells used for the assay were collected at the indicated optical densities and then frozen at 70°C until use. -Galactosidase activities (micromoles of ONPG [o-nitrophenyl--D-galactopyranoside] hydrolyzed per minute per milligram of protein) were determined and plotted.

In planta analysis of EpsR. The effect of single and multiple copies of epsR on the ability of R. solanacearum to kill eggplant seedlings was assayed as previously described (7, 21, 22, 40). At least 10 plants were inoculated with either the wild type, K60, K60/pKL4 (overexpressing EpsR), or K60::epsR (epsR insertional mutant), and the results of two independently performed assays are presented (Table 2). The wild type, K60, caused death in a majority of the plants by 11 days postinoculation (Table 2). Overexpression of EpsR slowed the wilting process, with only approximately half of the plants killed at 11 days postinoculation (Table 2). However, at 14 days postinoculation, more than half of the plants inoculated with K60/pKL4 died, possibly due to the loss of pKL4, since there is no antibiotic selection in the plant. Strains with an epsR insertional mutation showed reduced wilting, with only approximately half of the inoculated plants killed at 11 days postinoculation. This result is consistent with reduction but not abolition of epsI expression due to the lack of EpsR. As expected for a stable genetic change, no significant increase in plant death occurred by 14 days postinoculation.

View this table: [in this window] [in a new window]   TABLE 2.   Effect of epsR mutation and overexpression on the ability of R. solanacearum to kill eggplant seedlings

Regulation of EPS genes. To facilitate a more convenient analysis of the effects of epsR, we fused the promoters of epsI, opsG, and epsR to pGL10::lacZ, which contains a promoterless lacZ gene. The epsI promoter contained nt 140 to +120 in plasmid pGepsI::lacZ, the opsG promoter contained nt 360 to +1 in pGopsG::lacZ, and the epsR promoter contained nt 280 to +40 in pGepsR::lacZ. Promoter activities when EpsR is overexpressed or absent were assayed by measuring -galactosidase activity. Transformation of strain K60 with pGepsI::lacZ and pKL4 resulted in a reduction of epsI expression to 15% of that of K60 transformed with pGepsI::lacZ and pLAFR3 (Table 3). Furthermore, -galactosidase activity resulting from strains transformed with pGopsG::lacZ was unchanged, as expected, whether EpsR was overexpressed or absent (Table 3). The epsI and opsG promoters on plasmids, thus, mimicked the phenotype of their chromosomal counterparts with regards to EpsR regulation and demonstrated the efficacy of the system (29). Strains transformed with pGepsR::lacZ had approximately 150 U of -galactosidase activity, demonstrating that the epsR promoter is expressed when contained on a plasmid (Table 3). However, the expression of the epsR promoter was not affected when strain K60 harbored both pGepsR::lacZ and pKL4 or when K60 contained pGepsR::lacZ and an epsR insertion mutation (Table 3). Therefore, EpsR does not regulate its own expression.

EpsR is phosphorylated in vivo. Most effector proteins of two-component regulatory systems are modified at a conserved aspartic acid by phosphorylation, which can modulate a variety of the protein's activities (33). EpsR contains an aspartate (D) at amino acid 47, which is the most likely site of phosphorylation (Fig. 2A). We changed D47 to an alanine by using the scheme diagrammed in Fig. 2A. The 5' and 3' halves of epsR were amplified by PCR by using primers which changed codon 47 to an alanine while adding a unique AlwNI site (Fig. 2A). The alanine codon created is one which is commonly used in R. solanacearum (6). DNA carrying the mutant version of epsR was cloned into pLAFR3, creating pepsRD-A.

View larger version (32K): [in this window] [in a new window]   FIG. 2.   (A) A schematic diagram of site-directed mutagenesis of the conserved aspartate (D47) in epsR is shown. Oligonucleotide primers are indicated with arrows. The internal pair of primers contains changes in the epsR sequence, indicated by ^ or . Changes made in the epsR sequence are shown at the bottom of the figure, and the encoded amino acids are in the standard one-letter code. The original aspartate in the EpsR sequence and the alanine in the EpsRD-A sequence are shown in bold type. (B) Western blot analysis shows that EpsRD-A is expressed in R. solanacearum (S70) when cells are transformed with pepsRD-A. The star designates nonspecific recognition of protein by the anti-EpsR antibodies. (C) EpsR (lane 1), but not EpsRD-A (lane 3), is phosphorylated. A longer exposure of this gel (lane 6) reveals the phosphorylation of chromosomally derived EpsR. (D) After autoradiography, the gel in Fig. 3C was subjected to Western blot analysis with anti-EpsR serum (21). The results show that EpsR and EpsRD-A are present in cell lysates in similar quantities.

EpsRD-A cannot modulate expression of epsI. To examine the effect of EpsRD-A on EPS production, K60 was transformed with pepsRD-A. The wild type, K60, became visibly mucoid approximately 48 h after plating (Fig. 3A). Overexpression of EpsRD-A did not grossly affect the abundance of EPS produced by K60, in contrast to overexpression of wild-type EpsR from plasmid pepsR, which visibly reduced the amount of EPS produced by K60 colonies (Fig. 3A). The plasmid pepsR contains a PCR-derived clone of epsR and demonstrates that EpsR is sufficient to reduce production of EPS. As previously reported, the effect of EpsR lessens as the culture ages, since colonies transformed for 72 h are noticeably more mucoid than those transformed for 48 h (21).

View larger version (35K): [in this window] [in a new window]   FIG. 3.   (A) Colony morphology of K60 transformed for 48 and 72 h with pepsR, pGepsRD-A, and pLAFR3. (B) Overexpression of EpsRD-A does not affect expression of the epsI promoter as measured in strain S49, containing the epsI::lacZ fusion. Cells containing the plasmids indicated at the bottom of the graph were grown to an OD600 of 1.2 and assayed for -galactosidase activity (see legend to Fig. 1 for clarification of values) as described in Materials and Methods.

View larger version (21K): [in this window] [in a new window]   FIG. 4.   (A) Schematic diagram for construction of S49 chromosomes with one copy of epsRD-A. The desired integration event and the resulting chromosome are shown. (B) Expression from the epsI promoter is reduced in S49 strains which contain epsRD-A.

EpsR participates in a complex with the epsI promoter. All of the results described above suggest that phosphorylation plays a critical role in the activity of EpsR. We used a gel mobility shift assay to examine the interaction of EpsR with the epsI promoter region and the role of phosphorylation in this interaction. Clarified extracts made from strains K60 and S70 overexpressing wild-type EpsR from either pKL4 or pepsR were incubated with a labeled 240-bp probe containing a functional epsI promoter. Three predominant bands were observed in a gel mobility shift assay, two of which were present in all extracts, including S70::epsR, which does not contain a functional copy of epsR, and the third band, with the slowest mobility, which was present only in extracts from cells overexpressing EpsR (Fig. 5A, compare lane 2 to lanes 3 and 5). This band likely contains EpsR, while the two lower bands are not specific to EpsR, since they appear in all reactions, including ones which used lysates lacking EpsR (Fig. 5A, lane 1). In addition, extracts made from K60 and from S70, which overexpresses EpsRD-A, did not give rise to the specific band (Fig. 5A, lanes 4 and 6). These results strongly suggest that EpsR expressed in either wild-type R. solanacearum or a mutant lacking the acidic exopolysaccharide can form a complex with the epsI promoter. Furthermore, phosphorylation of EpsR is required for complex formation with the epsI promoter.

View larger version (36K): [in this window] [in a new window]   FIG. 5.   (A) Gel mobility shift assay shows that EpsR mediates a gel shift of the epsI promoter. Strains used to make each extract are indicated above the lanes. The EpsR-dependent shifted band is designated with a star. S70::epsR containing pLAFR3 serves as a negative control since it lacks a functional copy of epsR. (B) Extracts overexpressing EpsR specifically shift the epsI promoter. A molar excess of nonradiolabeled DNA containing the opsG (nt 310 to +10) or epsI (nt 140 to +100) promoters was added as indicated above the autoradiogram. (C) The EpsR-mediated gel shift is detected in extracts made from strains S90 (xpsR), S112 (vsrB or vsrC), and S80 (phcA) overexpressing EpsR from pKL4 (29). Free probe is indicated to the right of each gel.

	DISCUSSION

Top Abstract Introduction Materials & Methods Results Discussion References

An acidic form of EPS produced by the epsI gene cluster is a major virulence factor of the bacterial wilt pathogen, R. solanacearum. Disruption of the epsI structural genes will reduce or eliminate the ability of R. solanacearum to kill plants. Three signal transduction systems, VsrA-VsrD (37), VsrB-VsrC (19), and PhcA (4, 5, 17), positively affect epsI expression. EpsR has been reported to be a negative regulator of EPS when present in multicopy plasmids, decreasing EPS production and expression from the epsI promoter (16, 21, 29). In this work, we have extended the characterization of EpsR and found that a single copy of epsR in the R. solanacearum chromosome is required for wild-type levels of epsI expression and virulence. Thus, the level of EpsR protein in the cell can result in different phenotypes. Furthermore, both the repressive and stimulatory effects of EpsR and the formation of an EpsR-specific complex with the epsI promoter require an aspartate residue, which is important for the phosphorylation of EpsR.

After visual inspection of colonies with an epsR mutation, it was previously reported that one copy of EpsR was not affecting EPS expression (21). However, when expression from the epsI promoter was examined with the more sensitive and quantitative lacZ promoter fusions, expression was reduced in strains with an epsR mutation. This observation, coupled with the fact that epsR mutant strains are reproducibly affected in virulence, suggests that the chromosomal copy of epsR is positively regulating epsI synthesis, and only when overexpressed from multicopy plasmids can EpsR act as a repressor of epsI. We note that epsI expression and virulence are not abolished in epsR mutant strains, suggesting that EpsR is contributing to, but not essential for, expression of epsI.

Together, our results suggest that EpsR directly regulates epsI expression. First, EpsR does not affect the expression of any of the other regulators of epsI. Second, overexpression of EpsR and the absence of the chromosomal copy of epsR affected expression from only epsI. Third, multicopy plasmids carrying the known positive regulators of epsI did not alleviate EpsR-mediated repression of epsI expression, suggesting that EpsR does not simply titrate one of these regulators (data not shown) (6). Our data does not eliminate the possibility that overexpression of EpsR interferes with the ability of another positive regulator to associate with the epsI promoter.

It is possible, although unlikely, that EpsR does not interact directly with the epsI promoter. Since our binding assays were done with crude cell extracts, we cannot rule out the possibility that overexpression of EpsR may in some way stimulate another protein to bind the epsI promoter, which in turn accounts for the observed epsI gel shift in extracts overexpressing EpsR. However, the EpsR-specific gel shift of the epsI promoter was observed with lysates putatively lacking PhcA, XpsR, and VsrB or VsrC, suggesting that these proteins are not required for the EpsR-induced gel shift of the epsI promoter. The C terminus of EpsR is homologous to the DNA-binding domains of other response regulators which have been shown to bind DNA (21). EpsR is highly basic (predicted pI of 8.5), which is consistent with the idea that EpsR has a natural affinity for DNA. All of these results are consistent with the working model that the activity of EpsR is mediated through formation of a complex with the epsI promoter.

EpsR-specific gel shift of the epsI promoter requires phosphorylation of EpsR. For R. solanacearum, this is the first direct demonstration of transcriptional regulation by protein modification. Most response regulators are thought to be phosphorylated at a conserved aspartate found in the N terminus. Several proteins have been shown experimentally to be phosphorylated, which in turn regulates many different aspects of the protein's activity, including dimerization, DNA binding, and transcriptional regulation (2, 11-13, 24, 33, 39). EpsR has extensive homology with response regulators in the OmpR class; however, its phosphorylation domain is unique. EpsR is missing two conserved aspartate residues at amino acid positions 13 and 14, which are present in nearly all response regulators. In NarL, these residues are thought to form an acid pocket which facilitates phosphorylation (1). This suggests that the mechanism of phosphorylation for EpsR may be different from that of proteins such as NarL and NarP. Although EpsR is missing these residues, D47 is clearly required for phosphorylation, likely serving as the phosphate acceptor. We were unable to detect an EpsR-mediated gel shift of the epsI promoter by using extracts made from E. coli overexpressing EpsR. Coincident with this, preliminary results indicate that EpsR is not phosphorylated in E. coli, providing further proof that phosphorylation is required for the EpsR-mediated gel shift observed when R. solanacearum extracts are used (data not shown) (6). For E. coli strains transformed with either pKL4 or pepsR, we were unable to identify a phosphorylated band corresponding to EpsR. The identity of the kinase is not known; however, a functional homolog appears to be one that is absent in E. coli.

Approximately 1% of the nonessential genes in R. solanacearum are directly or indirectly involved in EPS production (29). At present, all regulators of EPS production affect expression from the epsI promoter. In E. coli, the epsI promoter is not expressed when carried on the plasmid pGepsI::lacZ, although other promoters involved in EPS production (opsG) are expressed (6). Since expression of epsI in R. solanacearum is dependent on several regulatory proteins, it is not surprising that this gene is not expressed in E. coli. Determining the signals required for activation of epsI, how each of the regulators interacts with each of the other regulators, and what conditions in planta might result in overexpression of EpsR are potential topics of future study.

The paradigm for differential regulation of a promoter is the lambda phage protein, which regulates its own expression by binding to different operator sequences located in its promoter (35). Several other bacterial proteins have been shown to act as both activators and repressors of transcription. In E. coli, the response regulators NarL and NarP can both positively and negatively affect transcription by the location of their DNA binding sites with respect to the transcriptional start site (9). The flagellar genes of Caulobacter crescentus are also both positively and negatively regulated by the response regulator FlbD (31). Should EpsR levels be regulated in the cell, then EpsR may interact specifically with the epsI promoter at multiple sites and with differing affinities. While it is possible that the repressive effect of EpsR is due to artificial overexpression, an abundance of EpsR retains the ability to specifically interact with the epsI promoter. Perhaps, with normal cellular levels of EpsR, a high-affinity binding site will be recognized, leading to activation of epsI. However, in our in vitro experiments, we are unable to detect an EpsR-mediated gel shift of the epsI promoter by using extracts containing wild-type levels of EpsR.