GamR, the LysR-Type Galactose Metabolism Regulator, Regulates hrp Gene Expression via Transcriptional Activation of Two Key hrp Regulators, HrpG and HrpX, in Xanthomonas oryzae pv. oryzae

ABSTRACT

Xanthomonas oryzae pv. oryzae is the causal agent of bacterial leaf blight of rice. For the virulence of the bacterium, the hrp genes, encoding components of the type III secretion system, are indispensable. The expression of hrp genes is regulated by two key hrp regulators, HrpG and HrpX: HrpG regulates hrpX, and HrpX regulates other hrp genes. Several other regulators have been shown to be involved in the regulation of hrp genes. Here, we found that a LysR-type transcriptional regulator that we named GamR, encoded by XOO_2767 of X. oryzae pv. oryzae strain MAFF311018, positively regulated the transcription of both hrpG and hrpX, which are adjacent to each other but have opposite orientations, with an intergenic upstream region in common. In a gel electrophoresis mobility shift assay, GamR bound directly to the middle of the upstream region common to hrpG and hrpX. The loss of either GamR or its binding sites decreased hrpG and hrpX expression. Also, GamR bound to the upstream region of either a galactose metabolism-related gene (XOO_2768) or a galactose metabolism-related operon (XOO_2768 to XOO_2771) located next to gamR itself and positively regulated the genes. The deletion of the regulator gene resulted in less bacterial growth in a synthetic medium with galactose as a sole sugar source. Interestingly, induction of the galactose metabolism-related gene was dependent on galactose, while that of the hrp regulator genes was galactose independent. Our results indicate that the LysR-type transcriptional regulator that regulates the galactose metabolism-related gene(s) also acts in positive regulation of two key hrp regulators and the following hrp genes in X. oryzae pv. oryzae.

IMPORTANCE The expression of hrp genes encoding components of the type III secretion system is essential for the virulence of many plant-pathogenic bacteria, including Xanthomonas oryzae pv. oryzae. It is specifically induced during infection. Research has revealed that in this bacterium, hrp gene expression is controlled by two key hrp regulators, HrpG and HrpX, along with several other regulators in the complex regulatory network, but the details remain unclear. Here, we found that a novel LysR-type transcriptional activator, named GamR, functions as an hrp regulator by directly activating the transcription of both hrpG and hrpX. Interestingly, GamR also regulates a galactose metabolism-related gene (or operon) in a galactose-dependent manner, while the regulation of hrpG and hrpX is independent of the sugar. Our finding of a novel hrp regulator that directly and simultaneously regulates two key hrp regulators provides new insights into an important and complex regulation system of X. oryzae pv. oryzae hrp genes.

INTRODUCTION

In several plant-pathogenic bacteria, hrp genes are essential for bacterial pathogenicity (1). The hrp genes generally comprise more than 20 genes and are clustered in the genome of each bacterium. The genes mainly encode components of a type III secretion apparatus, through which the bacteria secrete numerous proteins, so-called effectors, directly into plant cells (2). The proteins play important roles in host plants in suppressing the plant immune systems and inducing the expression of genes that increase the susceptibility of plants to infection, resulting in vigorous bacterial growth in plants and, finally, the emergence of disease symptoms (3–5).

The expression of hrp genes is strictly regulated, and it is induced only in/on plants and not under typical culture conditions, except for certain nutrient-poor and low-pH media, so-called hrp-inducing media (1, 6–9). Numerous studies have revealed that many regulators and factors are involved in the regulation of hrp gene expression and that they form a complex, sophisticated regulatory network for the expression not only of hrp gene but also other virulence-related genes, such as those involved in the production of extracellular enzymes and extracellular polysaccharides, in each bacterium (10–14).

Xanthomonas oryzae pv. oryzae causes one of the most serious rice diseases, bacterial leaf blight, in rice-growing areas worldwide (15). The virulence of the xanthomonad bacteria generally relies on hrp genes, which are regulated by two key hrp regulators, HrpG and HrpX. Unlike in other bacteria, such as Pseudomonas syringae and Ralstonia solanacearum, these key hrp regulator genes in xanthomonads are located far from the hrp gene cluster, and they are neighbors to each other but have opposite orientations. One of the regulators, HrpG, is predicted to be an OmpR-type response regulator of a two-component signal transduction system (16, 17). A certain point mutation in HrpG, which makes the protein constitutively active without phosphorylation, results in the expression of hrp genes even under non-hrp-inducing conditions in Xanthomonas campestris pv. vesicatoria (17). Recently, Li et al. (18) discovered the cognate sensor kinase for HrpG, named HpaS, in X. campestris pv. campestris, though a homolog of HpaS is not encoded in the genome of X. oryzae pv. oryzae and the sensor kinase for HrpG remains unknown in the bacterium. The findings strongly suggest that activation of HrpG by phosphorylation via the unknown cognate sensor protein and unknown plant signal(s) is also essential for the induction of hrp gene expression in X. oryzae pv. oryzae. HrpG directly regulates the expression of hrcC on the hrpA operon, and another key hrp regulator, HrpX, then regulates the expression of other hrp genes, which are located separately in 5 operons (hrpB to hrpF operons). The regulator HrpX is a member of an AraC-type transcriptional activator family (19), and it upregulates the transcription of the hrpB to hrpF operons by directly binding to the consensus motif, called the plant-inducible promoter box, which is located upstream from each operon (20, 21).

Besides HrpG and HrpX, several regulators that are involved in hrp gene expression have been identified. Histone-like nucleoid-structuring (H-NS) proteins are small DNA-binding proteins that are widely conserved in Gram-negative bacteria and function as important global regulators, usually as repressors of transcription for a wide range of genes, including virulence-related genes and environment-responsive genes (22, 23). X. oryzae pv. oryzae possesses three H-NS proteins, and one of them, named XrvA, is involved in the upregulation of hrpG expression (24). However, we previously found that another H-NS protein, XrvB, negatively regulates hrpG expression (25). In addition, the GntR family transcriptional regulator Trh and the two-component signal transduction system PhoP/PhoQ have also been reported to be involved in the transcriptional regulation of hrpG (26, 27). Thus, not only the phosphorylation of the first key hrp regulator, HrpG, followed by the activation of the second key hrp regulator, HrpX, but also the transcriptional regulation of hrpG are likely to be important, although the transcriptional regulation of hrpG by the regulators that have been identified so far is probably all indirect.

The sugar composition is an important factor in hrp gene expression by X. oryzae pv. oryzae in culture. For an hrp-inducing medium, we reported and have used XOM2, a nutrient-poor synthetic medium containing xylose as the sole sugar source (7). When xylose is replaced with other sugar sources, such as glucose, fructose, galactose, and sucrose, the expression of hrp genes is dramatically reduced (7). Very recently, we found that high xylose-dependent accumulation of HrpX is associated with the high expression of hrp genes in XOM2 with xylose (28).

Thus, although several regulators have been shown to be involved in HrpG- and HrpX-mediated hrp gene expression, unknown regulators and unknown regulatory mechanisms are still predicted. In the work presented here, we found that a novel LysR-type transcriptional regulator regulated not only galactose metabolism-related genes but also the transcription of both key hrp regulators, HrpG and HrpX, by directly binding to the common upstream region of hrpG and hrpX. We named the novel regulator gene gamR (galactose metabolism-related-gene regulator).

MATERIALS AND METHODS

Bacterial strains, plasmids, and growth conditions.

The bacterial strains and plasmids used in this study are listed in Table 1. Escherichia coli strains were cultured at 37°C in LB medium (29). X. oryzae pv. oryzae strains were grown at 28°C using nutrient broth-yeast extract (NBY) liquid medium (30). For hrp induction, the nutrient-poor synthetic medium XOM2 (7) containing xylose as a sole sugar source was used. The medium was also used to examine the sugar utilization or hrp gene expression of bacterial strains by adding other sugars in place of xylose. All media were supplemented with the following antibiotics when required: ampicillin, 50 μg ml⁻¹ for E. coli; kanamycin, 25 μg ml⁻¹ for X. oryzae pv. oryzae and 50 μg ml⁻¹ for E. coli; and spectinomycin, 25 μg ml⁻¹ for X. oryzae pv. oryzae and E. coli.

TABLE 1.

Bacterial strains and plasmids used in this study

Strain or plasmid	Characteristics or purpose	Reference or source
Xanthomonas oryzae pv. oryzae strains
MAFF311018	Wild type, Japanese strain
74Hpa1::Lux	Bioluminescent strain with a lux operon controlled by the hpa1 promoter	26
Clone no. 34	74Hpa1::Lux-derived mutant with transposon EZ::TN <KAN2> inserted at position +38 of XOO_2767 (gamR)	This study
ΔGamR strain	gamR deletion mutant, lacking an internal 834-bp region (+25 to +858) in the 972-bp coding region	This study
Escherichia coli strains
DH5α MCR	Strain used for cloning	Stratagene
BL21(DE3)(pLysS)	Strain used for protein expression and purification	Merck
Plasmids
pBluescript II SK+	Plasmid vector, Ampr	Stratagene
pBSΔGamR::Kmr/sac	Plasmid used for the gamR deletion mutant, containing the upstream and downstream regions of gamR with the kanamycin resistance gene and levansucrase gene	This study
pET28b	Plasmid vector used in E. coli protein expression system	Merck
pETHis-GamR	Coding region of gamR cloned in pET28b	This study
pHM1	Broad-host-range vector, pSa ori, Spr	34
pHMGamR	Complementation plasmid for ΔGamR strain	This study
pHMHrpG::GUS-W	Promoterless gus gene preceded by −803 to +130 of hrpG cloned in pHM1	This study
pHMHrpG::GUS-M4	Plasmid similar to pHMHrpG::GUS-W but with base substitutions at the GamR-binding sequence	This study
pHMHrpX::GUS-W	Promoterless gus gene preceded by −891 to +42 of hrpX cloned in pHM1	This study
pHMHrpX::GUS-M4	Plasmid similar to pHMHrpX::GUS-W but with base substitutions at the GamR-binding sequence	This study
pHMXOO2768::GUS-W	Promoterless gus gene preceded by −439 to +92 of hrpG cloned in pHM1	This study
pHMXOO2768::GUS-M1	Plasmid similar to pHMXOO2768::GUS-W but with base substitutions at the GamR-binding sequence	This study
pHMGamR::GUS	Promoterless gus gene preceded by −504 to +22 of gamR cloned in pHM1	This study

Transposon mutagenesis.

An EZ::TN <KAN2> transposome, a mixture of the transposon EZ::TN <KAN2> and the EZ::TN transposase (Epicentre, Madison, WI, USA), was introduced by electroporation directly into X. oryzae pv. oryzae strain 74Hpa1::Lux, a strain that produces bioluminescence under the control of the HrpX-regulated hpa1 promoter (26), as described previously (31). Transposition clones were selected by plating on NBY agar containing kanamycin. The bioluminescence of kanamycin-resistant clones was examined as described below and compared with that of the parental strain, 74Hpa1::Lux.

Detection of bioluminescence.

Bioluminescent strains of X. oryzae pv. oryzae were preincubated on NBY agar overnight and then transferred to hrp-inducing XOM2 agar using toothpicks, followed by incubation at 28°C for 24 h. The bioluminescence signal from each bacterial colony was collected for 1 min, and the photons were counted using a video-intensified microscope (VIM) camera and the Argus-100 system (Hamamatsu Photonics, Hamamatsu, Japan). Alternatively, bacteria preincubated on NBY agar were suspended in distilled water (DW) and their concentration was adjusted to an A₆₀₀ of 1.0, and then 5 μl of each bacterial suspension was dropped onto XOM2 agar, followed by further incubation and measurement of bioluminescence as described above. For detection of bioluminescence from infected rice leaves, leaves of the susceptible rice cultivar IR24 were inoculated by the clipping method (32) with 74Hpa1::Lux or its derivative adjusted to an A₆₀₀ of 1.0, and 2 days after inoculation, bioluminescence was measured for 1 min from 1-cm-long leaf sections that included the inoculation sites. Simultaneously, the bacterial numbers in the same leaf sections were determined by the plating method to calculate bioluminescence (photons) per bacterial cell (count min⁻¹ 10⁸ CFU⁻¹).

Sequence analysis.

A dye terminator cycle sequencing reaction was performed with a BigDye Terminator version 3.1 cycle sequencing kit (Applied Biosystems, Foster, CA) according to the manufacturer's instructions, followed by electrophoresis and analysis with an autosequencer (model 373A; Applied Biosystems).

PCR.

For PCR, DNA polymerase Blend Taq (Toyobo, Osaka, Japan) or KOD plus (Toyobo) was used according to the manufacturer's instructions. The primers used in this study are listed in Table 2.

TABLE 2.

Primers used in this study

Purpose, primer	Sequencea
Generation of gamR deletion mutant
GamR_5_FW	5′-AACTGCAGGCGGGTGATCTTCATCCTTG-3′
GamR_5_RV	5′-TAGGATCCAGGGAATACCGGAAGTTG-3′
GamR_3_FW	5′-AAGGATCCGTCGGCGCGATTACAACCTG-3′
GamR_3_RV	5′-ATGAATTCCGCACTGACACGGATGATG-3′
Generation of gamR complementary plasmid
GamR_S	5′-TTGAATTCTGAAGTCATGCGGTTACTCGC-3′
GamR_AS	5′-ATGAATTCCTAAAGTGCAAACATCCGCGC-3′
Real-time qRT-PCR
hrpG_S	5′-GTTGCTCCGCGACGAAAATAC-3′
hrpG_AS	5′-CTTGCGCAGCTTGTAGATATG-3′
hrpX_S	5′-GACGATGAGGTCAGCCTGTT-3′
hrpX_AS	5′-GAAGCACCACTCTCCAGCTC-3′
hrpA_S	5′-GCGTTGGAGACGACGAACAAG-3′
hrpA_AS	5′-GCCATCCTCAATACGCACATC-3′
hrcU_S	5′-AGACACCGCTTGACATTTCC-3′
hrcU_AS	5′-CCTCGCTTTCCTTGTATTCG-3′
Generation of GamR expression plasmid
His_GamR_S	5′-CAACCGCACGCATATGGCAACTTC-3′
His_GamR_AS	5′-AAGAGCTCAGAGCGACGACTCCGGTGC-3′
Amplification of hrpG and hrpX promoter region
hrpG_Pro_AS	5′-CATCTAGACAGTTCATCGGAGAACACCGAG-3′
hrpX_Pro_AS	5′-ACTCTAGACAACGCAGAGATCGCTGCAAAG-3′
hrpG_Pro299_S	5′-CTTGGCGGCTGTGCGCAGTGCCGACGCCCA-3′
hrpG_Pro319_S	5′-CGCAATAGATTGGTCTATACCTTGGCGGCT-3′
hrpG_Pro339_S	5′-CGCGTGATTGCACTACGAATCGCAATAGAT-3′
hrpG_Pro359_S	5′-GACCGATGTATTCCAGAATGCGCGTGATTG-3′
Amplification of hrpG and hrpX promoter region with base substitutions
hrpG_ProM2_S	5′-GAATTGTTGACCGCCGTATTCCAGAATGCG-3′
hrpG_ProM3_S	5′-TTGTTGACCGATGCCTTCCAGAATGCGCGT-3′
hrpG_ProM4_S	5′-TTGTTGACCGATGTCCTCCAGAATGCGCGT-3′
hrpG_ProM4_AS	5′-ACGCGCATTCTGGAGGACATCGGTCAACAA-3′
hrpG_ProM5_S	5′-GTTGACCGATGTACCCCAGAATGCGCGTGA-3′
hrpG_ProM6_S	5′-GACCGATGTATTCCACCATGCGCGTGATTG-3′
hrpG_ProM7_S	5′-GACCGATGTATTCCAGCCTGCGCGTGATTG-3′
hrpG_ProM8_S	5′-GACCGATGTATTCCAGACCGCGCGTGATTG-3′
Amplification of XOO_2768 promoter region
XOO2768_Pro_S	5′-CATCTAGAACCAGGGAATACCGGAAGTTGC-3′
XOO2768_Pro_AS	5′-TAGGATCCGACATGGTCGAGCGTTTCTTCC-3′
XOO2768_Pro346_S	5′-ATATTTCCCCGGGCAAGGCTATCCAATG-3′
XOO2768_Pro369_S	5′-GCCGAAAATTTCATTTGCAGCA-3′
XOO2768_Pro395_S	5′-GTGTATTCTTCGGAGGAATAGAATAG-3′
Amplification of XOO_2768 promoter region with base substitutions
XOO2768_ProM1_S	5′-GTTGCGGCGGTGTGGTCTTCGGAGGAATAGAATAG-3′
XOO2768_ProM1_AS	5′-CTATTCTATTCCTCCGAAGACCACACCGCCGCAAC-3′
XOO2768_ProM2_S	5′-GTGTATTCTTCGGAGGAGGAGAATAGGCCG-3′
XOO2768_ProM3_S	5′-GTGTATTCTTCGGAGGAATAGAGGAGGCCG-3′
Amplification of gamR promoter region
GamR_Pro_AS	5′-GGTCTAGAGGGAATACCGGAAGTTGC-3′

^aUnderlining shows additive restriction sites, and boldface shows nucleotides that were base substituted.

Construction of the gamR deletion mutant and the complemented strain.

The 1,082-bp upstream region of gamR, including a 24-bp coding region, and the 1,133-bp 3′ downstream regions of the gene, including a 114-bp coding region, were amplified by PCR using X. oryzae pv. oryzae strain MAFF311018 genomic DNA as the template, and the fragments were serially cloned into the vector plasmid pBluescript II SK+. Into the resulting plasmid, a kanamycin resistance gene from the EZ::TN <KAN2> transposon and the levansucrase gene (sac) from pK18mobsacB (33) were also inserted to generate pBSΔGamR::Km^r/sac. The plasmid was introduced into MAFF311018 by electroporation, and the gamR mutant, called the ΔGamR strain, which lacks an internal 834-bp fragment in the 972-bp coding region of the gene, was generated through two consecutive homologous recombination events.

For construction of the complementation plasmid for gamR, the fragment that contained the gene with the 423-bp upstream and 62-bp downstream regions was amplified and cloned into a broad-host-range vector pHM1 (34). The resultant plasmid, named pHMGamR, was used to transform X. oryzae pv. oryzae by electroporation.

Real-time qRT-PCR.

Bacterial total RNA was extracted using an RNeasy minikit (Qiagen, Hilden, Germany). After the quality of the RNA sample was confirmed by agarose gel electrophoresis, 0.2 μg of each RNA sample was used for synthesis of cDNA using the reverse transcriptase ReverTra-Ace (Toyobo). Real-time quantitative reverse transcription-PCR (qRT-PCR) was performed on the Rotor-Gene Q system (Qiagen) using Thunderbird SYBR qPCR mix (Toyobo) according to the manufacturers' instructions. The comparative quantitation method (35) was used to calculate the mRNA levels of the mutant and complemented mutant strains relative to the corresponding transcripts in the wild type. As an internal standard, the expression level of 16S rRNA was used. The gene-specific primer sets used are listed in Table 2.

Preparation of GamR tagged with histidine residues.

An NdeI-SacI fragment containing the coding region of gamR was amplified by PCR with the primer set His_GamR_S and His_GamR_AS and cloned in the expression vector pET28b (Merck, Darmstadt, Germany) to generate pETHis-GamR. E. coli strain BL21(DE3)(pLysS) transformed with the plasmid was preincubated at 37°C in LB broth overnight, and then the culture (500 μl) was transferred into 50 ml fresh LB broth. After a 2-h incubation, isopropyl-β-d-thiogalactopyranoside (IPTG) was added at a final concentration of 1 mM, and the bacteria were further incubated at 30°C for 2 h. Bacterial cells were suspended in 10 ml of 50 mM phosphate buffer (pH 7.4) with 300 mM NaCl and 20 mM imidazole and sonicated. After 0.05% (final concentration) Tween 20 was added, the bacterial suspension was centrifuged to separate the soluble and insoluble fractions, and GamR tagged with 6 histidine residues at the N terminus (His-GamR) in the soluble fraction was purified using a His60 Ni gravity column purification kit (Clontech Laboratories, Mountain View, CA, USA) according to the manufacturer's instructions. The protein solution was buffer exchanged to 20 mM HEPES buffer (pH 7.9) with 100 mM KCl, 2 mM MgCl₂, 0.1 mM EDTA, and 20% glycerol using a PD10 column (GE Healthcare, Buckinghamshire, United Kingdom) and stored at −20°C.

Electrophoretic mobility shift assay.

Approximately 200 ng template DNA and 250 ng His-GamR protein were mixed into 20 μl of reaction solution containing 20 mM HEPES (pH 7.9), 100 mM KCl, 2 mM MgCl₂, and 0.1 mM EDTA, and the mixture was incubated at room temperature for 1 h, followed by electrophoresis using 1.2% agarose gel in 0.5× Tris-borate-EDTA buffer. The DNA bands were visualized by staining with ethidium bromide.

GUS reporter assay.

Bacterial strains with a gus reporter gene preincubated overnight on NBY agar were washed twice with DW and suspended 1.5 ml of either XOM2 or modified XOM2 in which xylose, the original sole sugar source, was replaced with one of various other sugars to an A₆₀₀ of 0.1. The concentration of each sugar in the medium was 0.18% (wt/vol). After incubation at 28°C for 18 h with shaking (120 rpm), bacterial cells were collected by centrifugation. Glucuronidase (GUS) activity was then measured as previously described (36, 37), with 1 unit of GUS activity defined as nanomoles of p-nitrophenol released per hour. Simultaneously, the concentration of bacterial cellular proteins per assay was also examined. The bacterial cells in 1 ml of culture were pelleted and resuspended with 100 μl of bacterial protein extraction reagent (B-PER; Thermo Fisher Scientific, IL, USA) to extract bacterial proteins. Protein concentrations were measured using a protein assay kit (Bio-Rad, Hercules, CA, USA), with bovine serum albumin as the reference. The GUS activity of each sample was calculated as units/μg bacterial cellular proteins.

Bacterial growth test in media with various sugars.

Bacterial strains grown on NBY agar were washed twice and adjusted to an A₆₀₀ of 1.0 with DW, and 200 μl of bacterial suspension was added to 20 ml of XOM2 with xylose or the modified XOM2 with galactose, glucose, or fructose. The concentration of each sugar was 0.18% (wt/vol). The bacteria were cultured at 28°C with shaking (120 rpm), and bacterial growth was examined by measuring the A₆₀₀ of each culture.

Bacterial growth test in rice leaves.

Bacterial strains grown on NBY agar were washed twice and adjusted to an A₆₀₀ of 0.03 with DW. The susceptible rice cultivar IR24 was inoculated with the bacterial suspension by leaf clipping (32). Every 24 h after inoculation, a 1-cm-long leaf section that included an inoculation site was homogenized with 1 ml DW and plated on NBY agar. After incubation at 28°C for 3 days, colonies were counted.

RESULTS

Identification of a novel hrp regulator gene, gamR.

To identify novel hrp regulators in X. oryzae pv. oryzae, we conducted transposon mutagenesis using the EZ::TN <KAN2> system. The transposon was randomly inserted into the genomic DNA of X. oryzae pv. oryzae strain 74Hpa1::Lux (26), which harbors a lux operon controlled by the promoter of hpa1, one of the genes in the hrp cluster and, therefore, produces bioluminescence under hrp-inducing conditions. Among 1,500 mutants with kanamycin resistance, we found a clone (clone no. 34) with weaker bioluminescence than the parental strain (approximately 15%) on the hrp-inducing agar plates (see Fig. S1 in the supplemental material), which suggested that, in this clone, a transposon was inserted and disrupted the gene involved in the expression of hrp genes, including hpa1. The regions flanking the transposon were sequenced, followed by a homology search of the genome database of X. oryzae pv. oryzae strain MAFF311018 (38), revealing that the transposon was inserted into position +38 of XOO_2767 (position +1 is A of the putative initiation codon). According to the domain search program of the National Center for Biotechnology Information (NCBI) Web server (http://blast.ncbi.nlm.nih.gov/Blast.cgi), the gene is predicted to encode a protein with the typical features of LysR-type transcriptional regulators, containing a substrate binding domain and a helix-turn-helix DNA-binding domain (see Fig. S2). We named the gene gamR.

To confirm that gamR functions as an hrp regulator, we generated an in-frame deletion mutant, the ΔGamR strain, in which the sequence comprising positions +25 to +858 of gamR (972 bp in length) was deleted (see Fig. S2 in the supplemental material) and examined the expression of the key hrp regulator gene hrpG, the HrpG-regulated genes hrpA and hrpX (another hrp regulator gene), and the HrpX-regulated gene hrcU in the mutant. Real-time qRT-PCR analysis after 16 h of incubation in the hrp-inducing medium XOM2 revealed that the expression levels of all hrp genes tested decreased in the mutant strain with the empty vector (pHM1) compared with their expression levels in the wild type with the vector and were restored by the introduction of the complementation plasmid pHMGamR (Fig. 1). The results suggest that the product of gamR is a novel hrp regulator that regulates the transcription of a key hrp regulator gene, hrpG, in X. oryzae pv. oryzae.

FIG 1.

Reduced accumulation of hrpG, hrpX, hrcU, and hrpA transcripts in the gamR mutant. The wild type (MAFF311018 with the empty vector; white bars), the gamR mutant (the ΔGamR strain with the empty vector; gray bars), and the complemented strain (the ΔGamR strain with pHMGamR; black bars) were incubated in XOM2 for 16 h, and total bacterial RNA was analyzed for accumulation of each transcript by real-time quantitative RT-PCR. As an internal standard, 16S rRNA was used, and values relative to those of the wild type were calculated. Mean results (±standard deviations [SD]) from 3 independent replicates are shown. Asterisks indicate significant differences as determined by t test (*, P < 0.05; **, P < 0.01).

GamR binds to the intergenic region between hrpG and hrpX and regulates the expression of both genes.

The finding that GamR, predicted to be a LysR-type transcriptional regulator with a DNA-binding domain, is involved in the expression of an hrp regulator gene, hrpG, led us to examine whether the protein binds to the promoter region of hrpG and directly regulates the transcription of the gene. For this purpose, an E. coli strain transformed with pETHis-GamR, which produces the full-length GamR protein tagged with 6 histidine residues at the N terminus (His-GamR), was incubated in Luria-Bertani (LB) broth, and protein production was induced by adding IPTG, followed by purifying the protein using a nickel column (see Fig. S3 in the supplemental material).

The hrpG and hrpX genes are serially located in opposite orientations from each other on the genome, and the promoter regions for both genes are between the two genes (38). When the 933-bp DNA fragment between positions +130 of hrpG and +42 of hrpX, amplified by PCR with the primer set hrpG_Pro_AS and hrpX_Pro_AS, was incubated with His-GamR and then subjected to electrophoresis, the DNA's mobility was retarded (data not shown), suggesting that GamR binds directly to the hrpG and/or hrpX promoter region(s). We treated the 933-bp DNA fragment with EcoRV, HincII, or SacI restriction enzyme and used the digestion products as the templates for mobility shift assays. As shown by the results in Fig. 2A and B, His-GamR bound to the DNA containing the 107-bp SacI-HincII fragment (positions −253 to −359 of hrpG, which corresponds to positions −403 to −509 of hrpX). Next, regions of various lengths upstream from hrpG (+130 to −299, −319, −339, and −359) were amplified and tested in the mobility shift assay. Among the DNA fragments that we tested, only the fragment containing positions +130 to −359 of hrpG could be the template for His-GamR binding (Fig. 2C). For a detailed analysis to identify the GamR-binding sequence(s), we amplified fragments with base substitutions at various positions between −339 and −359 upstream from hrpG and used them in the electrophoretic mobility shift assay. As shown by the results in Fig. 3, we found that binding of His-GamR to the fragment was prevented by substitutions at bases from −351 to −348 of hrpG, while the other base substitutions tested did not affect protein binding. These results suggest that the TATT sequence at positions −351 to −348 of hrpG, which corresponds to the region from −411 to −414 of hrpX, is indispensable for GamR binding.

FIG 2.

Determination of the region where His-GamR binds in the upstream region common to hrpG and hrpX. (A) Binding of His-GamR to restriction fragments from the upstream region common to hrpG and hrpX. The 933-bp PCR fragment from position +130 of hrpG to +42 of hrpX was digested with restriction enzyme EcoRV, HincII, or SacI to test for binding of His-GamR with each fragment in the electrophoretic mobility shift assay; + and – indicate the presence and absence of His-GamR, respectively. M, 100-bp DNA ladder marker (Toyobo). (B) Fragments produced by digestion with EcoRV, HincII, or SacI. Thick black lines and dotted gray lines represent His-GamR-binding and -nonbinding fragments, respectively. The numbers above the segments indicate the size of each fragment (bp). (Bottom) The black box with numbers above and below it shows the region that includes the His-GamR-binding site(s), and the numbers indicate the locations of the fragments relative to the starts of hrpG (top) and hrpX (bottom). (C) DNA binding assay using His-GamR and the different fragments of the hrpG upstream region. The DNA fragments comprising positions +130 to −299, −319, −339, and −359 of hrpG were amplified and used as the templates; + and – indicate the presence and absence of His-GamR, respectively.

FIG 3.

DNA-binding assay of His-GamR protein using fragments upstream from hrpG and hrpX with or without base substitutions. (A) Electrophoretic mobility shift assay using fragments with or without base substitutions as the templates. A mixture of each DNA fragment (approximately 200 ng) and His-GamR (approximately 250 ng) in 20 μl of reaction solution was subjected to 1.2% agarose gel–Tris-borate-EDTA (TBE) buffer electrophoresis after 1 h of incubation at room temperature. To prevent nonspecific binding of His-GamR to the template DNAs, approximately 500 ng of the 991-bp fragment that includes the gamR-coding region, amplified using primers His_GamR_S and His_GamR_AS, which cannot be a template of His-GamR, was also added to the reaction mixture (the band for the 991-bp fragment is not shown in this figure). The numbers above the gel correspond to those at the left in panel B, where the 5′-end sequence with base substitutions in each template fragment is shown; + and − indicate samples with and without the protein, respectively. Arrows indicate target DNA fragments with (top) or without (bottom) His-GamR. (B) The 5′-end sequences of the target DNAs used for the His-GamR-binding assay are shown. All fragments end at position +130 of hrpG. The numbers at the top show the distance from the putative start codon of hrpG or, in parentheses, from that of hrpX. Underlining indicates base substitutions.

To investigate the importance of GamR's binding to the common hrpG and hrpX upstream region in hrpG expression in X. oryzae pv. oryzae, we used a GUS reporter assay with plasmids that contained the promoterless gus gene preceded by the region upstream from hrpG and hrpX (positions +130 of hrpG to +42 of hrpX), with or without base substitutions in GamR-binding nucleotides. When the ΔGamR strain transformed with pHMHrpG::GUS-W (with no base substitution) was incubated in the hrp-inducing medium XOM2, the GUS activity was reduced compared with the level in the wild type carrying the same plasmid (Fig. 4), supporting the importance of GamR in hrpG expression. Next, we incubated the wild type carrying pHMHrpG::GUS-M4 with base substitutions at −350 and −349 of hrpG (AT was changed to CC), which correspond to the GamR-binding site, in XOM2. The GUS activity produced by the strain was also lower than that of the wild type carrying plasmid pHMHrpG::GUS-W, which had no base substitutions (Fig. 4). These results indicate that GamR activates hrpG expression by binding directly to the promoter region of the gene.

FIG 4.

Importance of GamR and the GamR binding site in the expression of hrpG and hrpX. The GUS activities of X. oryzae pv. oryzae strains carrying a plasmid harboring a promoterless gus gene preceded by the hrpG or hrpX promoter region were measured after 18 h of incubation in XOM2 medium with xylose or modified XOM2 in which the sugar source was replaced with glucose or galactose. The wild type (WT) and the gamR deletion mutant (ΔGamR) were used. Plasmids represented by (+) possess the wild-type hrpG or hrpX promoter region (pHMHrpG::GUS-W or pHMHrpX::GUS-W), and plasmids represented by (–) contain base substitutions at the GamR-binding sites (−350 and −349 of the hrpG start codon [mutation of AT to CC] or −413 and −412 of the hrpX start codon [mutation of TA to GG]; pHMHrpG::GUS-M4 or pHMHrpX::GUS-M4). The mean GUS activities (±SD) relative to that of the WT (+) strain incubated with xylose-containing XOM2 are shown (n = 4 replicates). The ΔGamR (+) strain was not tested because the growth of the mutant was restricted when galactose was the sole sugar. Significance was determined using Tukey's test; significant differences (P < 0.05) among strains in each medium are shown by different letters.

Interestingly, we found that GamR regulates not only hrpG but also hrpX expression. As shown by the results in Fig. 4, the GUS activity produced by the wild type-derived transformant carrying pHMHrpX::GUS-M4, which harbors the hrpX::gus fusion gene preceded by the hrpX promoter region (positions +42 of hrpX to +130 of hrpG) with base substitutions at −413 and −412 (TA was changed to GG), was significantly reduced compared with that of the wild-type strain transformed with plasmid pHMHrpX::GUS-W (without substitutions) after incubation in XOM2. Since hrpX-regulating, GamR-regulated HrpG is produced normally in the transformants, the results indicated the importance of GamR binding in hrpX expression, as well as in hrpG expression. The GUS activity of the gamR deletion mutant with pHMHrpX::GUS-W was lower than that of the wild type carrying the same plasmid. Here, both the lowered production of hrpX-regulating HrpG and the loss of direct activation of the hrpX promoter by GamR are likely to be responsible for the further reduction of HrpX::GUS production.

GamR also functions as a regulator for galactose metabolism-related genes.

The genomic database of X. oryzae pv. oryzae MAFF311018 and analysis using the Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway database (http://www.kegg.jp/kegg/pathway.html) revealed that gamR is flanked by galactose metabolism-related genes XOO_2768, XOO_2769, XOO_2770, and XOO_2771, which probably form an operon (XOO_2768–XOO_2771) (see Fig. S4 in the supplemental material). This fact led us to examine whether GamR is also involved in galactose metabolism, along with hrp gene expression, by regulating the expression of XOO_2768–XOO_2771. For this purpose, we first tested the growth of the GamR mutant in medium with galactose as the sole sugar source. As shown by the results in Fig. 5, the growth of the mutant in the galactose medium was significantly reduced compared with that of the wild type and was completely restored by complementation with plasmid pHMGamR. No differences were observed between the wild type and the mutant for growth in medium containing glucose, fructose, or xylose as the sole sugar (see Fig. S5).

FIG 5.

Reduced growth of the ΔGamR strain in medium with galactose as the sole sugar. X. oryzae pv. oryzae wild type with pHM1 (empty vector) (open squares), ΔGamR strain with pHM1 (open circles), and ΔGamR strain with pHMGamR (complementing plasmid) (closed triangles) were incubated in the modified XOM2 medium with 0.18% galactose, and the A₆₀₀ was measured every 12 h. Three independent replications yielded similar results; representative data are shown.

To examine whether GamR regulates neighboring galactose metabolism-related genes, we constructed a plasmid harboring the XOO_2768::gus fusion gene preceded by the XOO_2768 upstream region (positions −439 to +92; amplified by PCR using primer set XOO2768_Pro_S and XOO2768_Pro_AS). When the wild-type strain carrying the plasmid, named pHMXOO2768::GUS-W, was incubated in the synthetic medium containing glucose, xylose, or galactose as the sole sugar, GUS activity in the bacterium was higher in the presence of galactose than with glucose and xylose (Fig. 6). Moreover, adding galactose to the glucose- or the xylose-containing medium resulted in higher GUS activity of the bacterium than in the medium with glucose or xylose alone. These results indicated that the promoter of XOO_2768 is present in the region from −439 to +92 and that it is responsive to galactose. The GUS activity of the GamR mutant carrying pHMXOO2768::GUS-W was also measured after incubation in the synthetic medium containing glucose or xylose with and without galactose. Here, we did not use a medium with only galactose because the growth of the GamR mutant was considerably limited, as shown by the results in Fig. 5. The growth of the mutant in the medium with the various sugar additions was similar to that of the wild type (data not shown), while unlike the wild type, adding galactose to the glucose- or xylose-containing medium did not heighten GUS activity in the mutant (Fig. 6). The findings indicated that GamR regulates the neighboring galactose metabolism-related genes, or at least one of them, XOO_2768, in the presence of galactose.

FIG 6.

Expression of XOO_2768 is activated by GamR in the presence of galactose. The GUS activities in X. oryzae pv. oryzae wild type and the gamR deletion mutant (ΔGamR) harboring pHMXOO2768::GUS were measured after 18 h of incubation in modified XOM2 medium with various sugar sources. Glu, glucose; Gal, galactose; Xyl, xylose. Each sugar was added at a concentration of 0.18%, including when two sugars were used. Values are the mean results (±SD) from 4 replicates. Significance was determined using Tukey's test; significant differences (P < 0.05) among media are shown by different letters. Similar results were obtained in 2 independent experiments.

To clarify that GamR directly regulates the expression of XOO_2768, we performed an electrophoretic mobility shift assay using His-GamR and the PCR-amplified 531-bp fragment containing the XOO_2768 promoter region (positions −439 to +92). As expected, the protein bound to the DNA fragment, and the analysis using the restriction fragments from digestion with KpnI or SalI revealed that the region from −439 to −223 contains the target of the protein (Fig. 7A and B). Furthermore, when DNA fragments of various lengths were used as templates for the mobility shift assay, the fragment comprising positions −395 to +92 was found to be a possible target for the protein, while fragments comprising positions −369 to +92 and −346 to +92 were not, suggesting that the binding sequence of XOO_2768 is located between positions −395 and −369 of XOO_2768 (Fig. 7C). The fragments of the XOO_2768 upstream region with base substitutions between positions −395 and −369 were then PCR amplified and tested for His-GamR binding, revealing that base substitutions at −391 and −390 or at −378 and −377 interfere with the DNA-protein binding, while those at −373 and −372 do not (Fig. 8A and B). Plasmid pHMXOO2768::GUS-M1, which is the same as pHMXOO2768::GUS-W except for base substitutions at −391 and −390, was generated and used to clarify the importance of GamR and its binding sequence upstream from XOO_2768 in the expression of XOO_2768. As shown by the results in Fig. 8C, both the wild-type strain carrying pHMXOO2768::GUS-M1 and the GamR mutant carrying pHMXOO2768::GUS-W had lower GUS activities than did the wild type carrying pHMXOO2768::GUS-W after incubation in the XOM2-based medium supplemented with galactose and glucose (0.18% each) as sugar sources. These results indicate that GamR directly regulates the expression of galactose metabolism-related gene(s), in addition to the hrp regulatory genes.

FIG 7.

Determination of the binding region of His-GamR upstream from XOO_2768. (A) DNA-binding assay using restriction fragments from the upstream region of XOO_2768. The PCR fragment from position +92 to −439 of XOO_2768 was digested with restriction enzyme KpnI or SalI, and the binding activity of His-GamR with each fragment was assessed; + and – indicate the presence and absence of His-GamR, respectively. (B) Fragments produced by KpnI or SalI digestion. Thick black lines and dotted gray lines represent His-GamR-binding and -nonbinding fragments, respectively. The size of each fragment (bp) is shown above the segments. (Bottom) Numbers above the line representing the full-size fragment indicate locations relative to the start of XOO_2768. (C) DNA-binding assay using His-GamR and the different fragments of the XOO_2768 upstream region. DNA fragments comprising positions +92 to −346, −369, −395, and −439 of XOO_2768 were amplified and used as the templates; + and – indicate the presence and absence of His-GamR. M, 100-bp DNA ladder marker (Toyobo).

FIG 8.

Identification of GamR-binding site in the upstream region of XOO_2768. (A) The 5′-end sequences of the target DNAs used for His-GamR-binding assay are shown. All fragments end at position +92 of XOO_2768. Numbers show the distance from the putative start codon of XOO_2768. Underlining indicates base substitutions. (B) DNA-binding assay of His-GamR. Each DNA fragment whose sequence is shown in panel A was mixed with His-GamR, along with the fragment that includes the gamR coding region, used to prevent nonspecific binding of His-GamR with the target DNAs. The band of that fragment is not shown here. The mixtures were subjected to electrophoresis after 1 h of incubation; + and − show samples with and without the protein, respectively. Arrows indicate target DNA fragments with (top) or without (bottom) His-GamR. (C) Mean GUS activities (±SD) in X. oryzae pv. oryzae strains after incubation in the modified XOM2 medium with glucose and galactose (0.18% each) as sugar sources. WT (+), wild-type strain carrying the plasmid harboring XOO_2768::gus preceded by the wild-type XOO_2768 promoter region (−439 to + 92); ΔGamR (+), gamR deletion mutant carrying the same plasmid; WT (−), wild-type strain carrying the plasmid harboring XOO_2768::gus preceded by the XOO_2768 promoter region with the same base substitution as M1 in panel A. GUS activities are relative to that of the WT (+) strain from 4 replicates. Significance was determined using Tukey's test; significant differences (P < 0.05) among strains are shown by different letters. Similar results were obtained in 2 independent experiments.

Sugar-source-independent GamR regulation of hrpG and hrpX.

In our previous work (28), we showed that the expression of hrpG and hrpX is independent of the sugar source, although the expression of HrpX-regulated genes is specifically induced in the medium with xylose as a sole sugar source. On the other hand, the requirement of galactose for GamR-dependent expression of galactose metabolism-related gene(s) was found, as shown in the previous section. Here, we examined the dependence or lack of dependence on galactose for the GamR- and GamR-binding-site-dependent expression of hrpG and hrpX by using the GUS reporter assay. As shown by the results in Fig. 4, the expression of hrpG and hrpX decreased after gamR deletion or after the base substitutions at the GamR-binding site in any medium with any of the sugar sources that we tested, and the GUS activity of each strain did not differ significantly with any of the sugar sources. The results suggest that, unlike the regulation of the galactose metabolism-related gene(s), the regulation by GamR of hrpG and hrpX transcription is independent of the sugar source.

In planta hrp gene expression, bacterial growth, and virulence of the gamR mutant.

In planta hrp gene expression was examined using the transposon-inserted gamR mutant (clone no. 34) derived from 74Hpa1::Lux, which has bioluminescence production that is dependent on HrpG and HrpX. Two days after inoculation, bioluminescence was observed from leaves inoculated with either the parental strain or the mutant, but the intensity from the mutant was only approximately 25% of that of the parental strain and was recovered by the introduction of pHMGamR (Fig. 9). The results indicated that GamR also regulates hrp gene expression in planta, not just in the hrp-inducing medium.

FIG 9.

Bioluminescence produced by strain 74Hpa1::Lux and the gamR deletion derivative in rice leaves. Rice leaves were inoculated with X. oryzae pv. oryzae strain 74Hpa1::Lux (which has bioluminescence production that is dependent on HrpG and HrpX), its gamR deletion derivative clone no. 34, or the complemented strain of clone no. 34 harboring pHMGamR; bioluminescence from 1-cm-long leaf sections that included the inoculation sites was measured 2 days after inoculation. Bacterial numbers in the same leaf sections were determined by the dilution method. Nine leaves were used for each bacterial strain; the mean counts of bioluminescence per min per 10⁸ CFU (±SD) are shown. Significance was determined using Tukey's test; significant differences (P < 0.05) among strains in each medium are shown by different letters. Similar results were obtained in 2 independent experiments.

In spite of the low hrp gene expression, however, the population of the ΔGamR strain differed little from that of the wild-type MAFF311018 (Fig. 10), and the lengths of the lesions that each strain caused on leaves were similar when we measured them 2 weeks after inoculation (data not shown). Although the level of hrp gene expression in the mutant is low, it may be high enough for the mutant to be virulent, at least under our experimental conditions.

FIG 10.

Growth of Xanthomonas oryzae pv. oryzae wild type and the ΔGamR strain in rice leaves. Black line, wild-type MAFF311018; gray line, ΔGamR strain. Bacterial numbers in 1-cm-long leaf sections that included the inoculation sites were determined by the dilution method at 0, 24, 48, 72, and 96 h after inoculation. Five leaves were used for each bacterial strain; the mean results (±SD) are shown. Similar results were obtained in 2 independent experiments.

DISCUSSION

The expression of hrp genes in plant-pathogenic bacteria is highly influenced by the nutrient conditions and is usually repressed under nutrient-rich conditions (6–9). Among nutrients, sugar sources and their metabolism are likely to be closely related to hrp gene regulation. In X. oryzae pv. oryzae, we recently found that xylose functions as an inducer of hrp gene expression by promoting the accumulation of one of the hrp regulatory proteins, HrpX (28). For X. oryzae pv. oryzicola, Guo et al. (39) reported that fructose-bisphosphate aldolase, which is essential for glycolysis and gluconeogenesis to reversibly convert fructose-1,6-bisphosphate to dihydroxyacetone phosphate and glyceraldehyde 3-phosphate, is involved in positive regulation of two key hrp regulator genes, hrpG and hrpX. In the present study, we found that a LysR-type transcriptional regulator that regulates galactose metabolism also regulates hrpG and hrpX by binding directly to their common upstream region.

The LysR-type transcriptional regulators are well-characterized family members in bacteria and regulate various cellular functions, such as metabolism, quorum sensing, biofilm formation, and virulence (40). The regulators generally consist of two domains, the N-terminal DNA-binding domain with a helix-turn-helix motif and the C-terminal substrate (cofactor)-binding domain. The GamR protein of X. oryzae pv. oryzae possesses the typical features of this protein family and is well conserved in other xanthomonads. Moreover, the DNA-binding activity of GamR shown in this study supports that the protein is a member of this family.

One of the GamR targets was XOO_2768 or the putative operon that includes XOO_2768 to XOO_2771, which are annotated as genes involved in galactose metabolism (see Fig. S4 in the supplemental material). We predict that the interrupted palindrome with the sequence TATT-N₉-AATA corresponding to positions −392 to −376 of XOO_2768 is a binding motif of GamR (Fig. 8). The absence of GamR or the motif led to limited XOO_2768 expression, and at the same time, the growth of the GamR deletion mutant was limited in medium with galactose as a sole sugar source, indicating the importance of the regulators for galactose utilization by X. oryzae pv. oryzae.

The GamR-dependent induction of XOO_2768 was specific to galactose; no induction occurred without galactose (Fig. 6). The results suggest that galactose functions as a cofactor of GamR. In regulation by LysR-type transcriptional regulators, generally, two dimers bind to distinct DNA sites in the promoter region of the target gene, and they come into contact to form a tetramer, accompanied by a structural change (bending) in the DNA. The presence or absence of the coinducer determines the degree of DNA bending, which will turn on or off the transcription of the target gene (41). The sequence TATT-N₉-AATA at positions −392 to −376 of XOO_2768 is predicted to be one of the binding sites of the GamR dimer. Although we could not find another binding site for the dimer in this study, the tetrameric GamR molecules with galactose as the cofactor are likely to activate the transcription of XOO_2768, which is a typical manner of transcriptional activation by LysR-type regulators.

The region from position −392 to −376 of XOO_2768 corresponds to positions −37 to −21 of gamR. We therefore thought that the binding of GamR to the region might also affect the expression of gamR itself. When the GUS activities of the wild type and the GamR deletion mutant harboring pHMGamR::GUS, in which the promoterless gus gene is preceded by the fragment spanning positions −504 to +22 of gamR, were measured after growth in medium with or without xylose or galactose, the GUS activity was higher in the mutant derivative under all culture conditions that we tested, but the differences were slight (see Fig. S6 in the supplemental material). Also, there were no significant differences in the levels of expression of gamR under the different sugar conditions. It is likely that the binding of GamR to the region from −37 to −21 of gamR (−392 to −376 of XOO_2768) negatively affects the expression of the gene, but the effect is slight.

Besides its involvement in regulating galactose metabolism, interestingly, GamR is also involved in hrp gene expression by simultaneously regulating the expression of two key hrp regulators, hrpG and hrpX. These genes are located next to each other with reverse orientations. Our results indicated that GamR binds in between the genes, in their common promoter region, and then regulates them. In contrast to the requirement for galactose, probably as a cofactor, for the regulation of galactose metabolism-related gene(s), the sugar is not required for the regulation of the hrp regulatory genes, and the regulation was not dependent on any sugar (Fig. 4). Moreover, although a TATT sequence in the common promoter region of hrpG and hrpX is indispensable for GamR-dependent activation of the genes, no TATT-N₉-AATA sequence is found here, unlike in the XOO_2768 upstream region (Fig. 3). From the differences described above, the regulatory mechanism of hrp regulator genes by GamR is likely to differ from that of galactose metabolism-related gene(s) and to be cofactor independent. At present, how GamR regulates the two hrp regulators, which are located on each side of the GamR-binding site, remains unknown. Structural changes in the common hrpG and hrpX promoter region caused by the binding of GamR molecules may result in increased transcriptional activation of both genes. Alternatively, the binding of GamR might protect the promoter region from the binding of unknown negative regulator(s). One of the LysR-type transcriptional regulators in Salmonella enterica, LeuO, functions as an antagonist of H-NS and its paralogue StpA, which play roles in negatively regulating target genes by binding to and polymerizing on the promoter region of the target genes (22, 23, 42–45). Competition between H-NS (or StpA) and LeuO for DNA binding and inhibition of H-NS (StpA) polymerization by LeuO are proposed as reasons for the antagonism. In X. oryzae pv. oryzae, three H-NS proteins are encoded in the genome, and at least two of them, XrvA and XrvB, are reported to be involved in hrp gene regulation (24, 25). However, XrvA is involved in positive regulation of hrpG expression, which implies that XrvA must negatively regulate an unknown negative regulator of hrpG. Also, no XrvB binding to the hrpG upstream region has been reported (25). Therefore, GamR might function as an antagonist of the negative regulator(s) for hrpG and hrpX, but the targets could not be H-NS protein(s). Further analyses are required to clarify the details of the mechanism of regulation of hrpG and hrpX by GamR.

The results of the GUS reporter assay show that the effect of base substitutions at the GamR-binding site is more serious than that of gamR deletion on hrpG expression (Fig. 4). The base substitutions might also affect regulation by an unknown regulator(s) other than GamR. In addition, on the basis of the results of the reporter assays, the lack of GamR and its binding sequence reduces but does not eliminate the expression of hrpG and hrpX. These facts suggest that the regulation by GamR is one of multiple regulatory mechanisms for hrpG and hrpX expression or that GamR might assist other hrpG or hrpX regulators to function efficiently by changing the structure of the promoter region.

If GamR does not bind to the common hrpG and hrpX promoter region because of the absence of gamR or a GamR-binding site, hrpG can still be expressed but at a lower level (Fig. 4); the loss of GamR resulted in extremely decreased expression of hrpX and HrpX-regulated hrp genes, such as hrcU and hpa1 (Fig. 1; see also Fig. S1 in the supplemental material). According to the qRT-PCR analysis whose results are shown in Fig. 1, the hrpG transcript level in the GamR mutant reached as much as 70% of the level in the wild type. The accumulation of the HrpG-regulated hrpA transcript also decreased in the mutant but was 50% of the level in the wild type. On the other hand, the accumulation in the mutant of the hrpX transcript, which is regulated by HrpG similarly to hrpA, was reduced to less than 20% of the level in the wild type. Here, the lack of direct transcriptional activation by GamR and the lower transcriptional activation resulting from the reduced level of HrpG are likely to be responsible for the lower accumulation of the hrpX transcript, followed by lower expression of HrpX-regulated hrp genes.

Not only in culture but also in rice leaves, the gamR deletion led to reduced expression of hrp genes, at least hpa1 (Fig. 9). However, there was no significant difference either in bacterial growth in rice leaves or in virulence between the wild type and the gamR mutant (Fig. 10). This is interesting but not so surprising. In a previous paper (26), we reported that the X. oryzae pv. oryzae mutant lacking the GntR-type transcriptional regulator Trh, which is also involved in the activation of hrpG transcription, has lower hrp gene expression but population growth and virulence in rice leaves that are similar to those of the wild type, as in the case of the gamR mutant. It is likely that hrp gene expression is excessively abundant and that the reduction in expression from a lack of gamR or trh does not influence bacterial growth and virulence, at least under our experimental inoculation conditions.

In this study, the regulation of hrp gene expression and that of galactose metabolism-related gene expression utilized a common regulator, GamR. However, galactose and its metabolism are unlikely to be important for hrp gene expression, because the expression of hrpG and hrpX is not influenced by the presence/absence of galactose (Fig. 4) or by the deletion of galactose metabolism-related genes (XOO_2768 to XOO_2771) (data not shown). Why the regulation of these genes is controlled by a common regulator is an interesting question.

A homology search using the NCBI Web server (http://blast.ncbi.nlm.nih.gov/Blast.cgi) revealed that gamR is conserved among many Xanthomonas species with high homology (up to 95% identity at the amino acid level) and is located next to the galactose metabolism-related genes in each genome. We conducted ClustalW analyses, using the Bielefeld University Bioinformatics Server (http://bibiserv.techfak.uni-bielefeld.de/dialign/), on regions upstream from XOO_2768 and its homologs and on regions between hrpG and hrpX in X. oryzae pv. oryzae MAFF311018 (GenBank sequence accession number AP008229.1 [38]), X. oryzae pv. oryzicola strain BLS256 (CP003057.2 [46]), X. campestris pv. vesicatoria strain 85-10 (AM039952.1 [47]), X. axonopodis pv. citri strain 306 (AE008923.1 [48]), and X. campestris pv. campestris ATCC 33913 (AE008922.1 [48]). The sequence targeted by GamR, TATT-N₉-AATA, is well conserved in the upstream region of the XOO_2768 homolog in each bacterium (see Fig. S7 in the supplemental material), suggesting that GamR regulates galactose metabolism in other Xanthomonas species. Also, we found that the motif TATT exists at a similar location in all bacteria we examined, although the sequence similarity of the hrpG-hrpX intergenic region in X. oryzae pv. oryzae and other bacteria is notably low, except for X. oryzae pv. oryzicola (see Fig. S8 in the supplemental material). Whether GamR commonly regulates hrp gene expression in xanthomonads should be further investigated.

Funding Statement

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.