The bacterial strains and plasmids used in this study are listed in Table 2. Xoo strains were grown in rich media, e.g., nutrient agar (NA) and peptone sucrose agar (PSA), and in minimal media, M9 (Miller, 1972). Escherichia coli strains were cultured in Luria-Bertani (LB) medium. Antibiotics were used at the following concentrations: kanamycin (Km), 50 μg/ml; ampicillin (Amp), 100 μg/ml; cephalexin (Cep), 100 μg/ml; and gentamicin (Gm), 10 μg/ml.

For growth rate assay, Xoo WT and rpfE-and rpfC-mutant strains were incubated at 28ºC in M9 minimal medium containing 50 mM glucose, xylose, or sucrose, with shaking at 200 rpm. At indicated time points, 100-μL aliquots of each culture were spread on PSA medium, and the colony forming units (CFUs) were counted after incubation at 28ºC for 3 days.

All primers for the construction of the rpfE knockout mutant and rpfE complementation strains (Table 3) were designed on the basis of the Xoo KACC10331 genomic sequence (GenBank accession number AE013598) (Lee et al., 2005). The rpfE gene was amplified from Xoo genomic DNA and cloned into a TA-cloning vector (RBC Bioscience Corp., Taipei, Taiwan). Successful cloning was confirmed by sequencing and Xoo rpfE was subsequently cloned into the pUC19 vector using restriction enzymes (BamHI/HindIII) (Invitrogen Corporation, Carlsbad, CA, USA). The EZ-Tn5^TM<KAN-2> Transposome^™ insertional kit (Epicentre Biotechnology, Madison, WI, USA) was used to insert Tn5 into rpfE:: pUC19. A clone containing the Tn5 insertion in the middle of the rpfE gene was selected by restriction enzyme screening, and the location of insertion was confirmed by nucleotide sequencing analysis (Microgen Inc., Seoul, South Korea). The selected plasmid was introduced into Xoo KACC10859 by electroporation to generate the rpfE knockout strain. Transformants with homologous recombination were isolated on selection medium containing Km and Amp and confirmed using PCR-based analysis. To determine if phenotypic changes in the knockout mutant strain occurred through a polar effect, a complementation construct was generated using an expression vector, pML122 (Labes et al., 1990), as previously described (Cho et al., 2011; Labes et al., 1990). The rpfE gene containing no promoter region amplified by PCR using the primers rpfE-com-F/R (Table 3) was inserted into TA vectors. The cloned fragment was verified with sequencing analysis and inserted into pML122 using restriction enzymes (XhoI/BamHI), and the plasmid carrying Xoo rpfE was introduced into the knockout mutant strain by electroporation. The construct was also verified by restriction enzyme digestion and PCR-based analysis before Xoo transformation.

In this study, total DNA was extracted from Xoo cells using Exgene^™ cell SV (GeneAll Biotechnology Co. Ltd., Seoul, South Korea), plasmids were purified using the plasmid mini-prep Exprep^™ plasmid SV (GeneAll Biotechnology Co. Ltd., Seoul, South Korea), and PCR was performed with Ex Taq^™ (Takara Biotechnology Inc., Tokyo, Japan). The Gene Pulser Xcell^™ system (Bio-Rad Laboratories, Hercules, CA, USA) was used to transform Xoo with the following settings: voltage pulse, 3.0 kV; capacitance, 25 μF; and resistance, 200 Ω.

Cellulase activities were assayed as described (Chatterjee et al., 1995). The rpfE mutant was incubated in Nutrient Broth (NB) for 72 h, and the optical density of the cultures was then adjusted to 1.0 at 600 nm. The culture supernatant was added to cellulase assay medium and incubated for 24 h at 28°C. The cellulase detection plate was stained for 10 min with 0.1% Congo Red and washed 3 times with 1 M NaCl until it showed a clear zone. Xylanase activity and EPS dry weight were determined as previously described (Jeong, 2008). The rpfE mutant strain was cultured in NB for 72 h and the optical density of the cultures was then adjusted to 1.0 at 600 nm.

To determine xylanase activity, 1 ml of culture supernatant was incubated with 480 μl of xylanase assay solution for 30 min at 28°C. To this mixture, 960 μl of absolute ethanol was added and the reaction was incubated at room temperature for 40 min. The mixture was centrifuged at 10,000 × g for 10 min, and the optical density of the supernatant was measured at 590 nm. To measure EPS dry weight, 1.0% KCl (w/v) was added to the culture supernatant. Two volumes of absolute ethanol were then added to the tubes, and the mixture was incubated at −20°C overnight. The mixture was centrifuged at 10,000 ×g, the EPS pellets were dried for 12 h at 55°C and the dry weight of the pellet was measured.

The swarming test was performed on peptone sucrose soft agar containing 0.3% agar as previously described (Ryan et al., 2007; Shen et al., 2001). All virulence-related phenotype assays were independently repeated 3 times.

The total RNA was extracted from Xoo KACC10859, the rpfE mutant strain, and the complementation strain using RiboPure^™ Kits (Invitrogen Corporation) after culture for 2 days at 28°C. A QuantiTect^® Reverse Transcription Kit (QIAGEN, Düsseldorf, Germany) was used to synthesize cDNA, which was then subjected to quantitative real-time PCR (qRT-PCR) that was performed with SYBR^® Premix Ex Taq^™ (Takara Bio, Inc.) using a Smart Cycle^®II system (Takara Bio, Inc.) and SeeAMP^™ (Seegene, Inc.). Primers were designed on the basis of Xoo KACC10331 genome sequence in GenBank (Table 2). qRT-PCR was carried out as previously described (Jeong et al., 2008) with an initial denaturation step at 95°C for 30 sec, followed by 45 cycles of denaturation at 95°C for 5 sec, annealing at 56°C for 15 sec, and extension at 72°C for 20 sec. The quantity of the synthesized target mRNA was calculated using 16S rRNA as an internal control with Qubit^™ ver. 1.01 (Invitrogen Corp.), and the relative expression of each gene was calculated using the Smart-Cycle^® 3.0 software and Seegene viewer.

Each rpf-mutant strain and the Xoo WT strain were incubated in NB medium for 72 h at 28ºC. To remove bacterial cells, the cultures (500 ml) were centrifuged at 5,000 × g for 20 min at 4ºC and filtered through a 0.45-μm membrane. To extract DSF, an equal volume of ethyl acetate (EA, DaeJung Chemicals, Gyeonggi, Korea) was added to the supernatant and the mixture was incubated for more than 1 h with shaking at room temperature (He et al., 2010). The ethyl acetate fractions were separated and concentrated by rotary evaporator (Eyela, N-INW, Tokyo Rikakikai Co., Ltd., Japan). The residues were dissolved in 3 mL of CHCl₃ and concentrated to a small volume by vacuum centrifugation (SAVANT SpeedVac^® DNA120, Thermo Scientific, Vantaa, Finland). The samples were spotted on silica plates (TLC Silica gel 60 RP-18 F_254s, Merck Korea, Seoul, Korea) with DSF standard (cis-11-methyl-2-dodecenoic acid, Cayman Chemical Company, Ann Arbor, MI, USA), and then developed with 2-butanol: ethanol:water = 2:1:1 (BEW). Separated spots were detected under UV light at 254 nm and extracted in CHCl₃.

Xoo strains were inoculated to test virulence ability using the scissor-clip method (Tang et al., 1996). Xoo strains were grown in NB and PSB to optical densities of 0.1 at 600 nm for the inoculation, and the susceptible rice cultivar Milyang 23 or Dongjin 1 was used. Lesion lengths were measured at 14 days after inoculation for end point measurement or measured daily using digital vernier calipers (CD-15CP, Mitutoyo Corporation, Kanagawa, Japan).

For measurement of population change of the pathogen in the inoculated rice leaf, top 10 cm of the inoculated rice leaf were cut by 2 cm and numbered 1^st to 5^th from the top of the inoculation cut. Each leaflet was cut into small pieces with razor blade and macerated in sterilized water with microfuge tube and plastic pestle. The macerates were vortexed for 20 min and the suspension was plated on NA containing cephalexin, 10 μg/ml, after the suspension of the macerate was diluted.

The rpfE and flanking region was cloned and sequenced in X. campestris first (Dow et al., 2000; GeneBank AJ2455997.1). The rpfE and its flanking genes, and gene order and orientation were well conserved in X. oryzae pv. oryzae KACC10331 whose genome was sequenced completely (Lee et al., 2005; GeneBank AE013598.1). The 84% of nucleotide sequence and 87 % of amino acid sequence were identity between two genes. The 14 nucleotides were more in rpfE_xoo at just upstream of 1^st codon, which is a major gap between two genes showed by BLAST.

To generate rpfE knock-out mutant, Tn5 was inserted in Xoo rpfE cloned into the pUC19 with EZ-Tn5^TM<KAN-2> Transposome^™ insertional kit. The sequence of the Tn5 inserted DNA showed that Tn5 was inserted between 595^th and 596^th nucleotide of Xoo rpfE. By marker exchange, an rpfE-deficient strain (CBNUXO14) was generated. To test the virulence, 6-week-old rice plants (cv. Milyang 23) were inoculated with the mutant (CBNUXO14), a complementation strain (CBNUXO15) and wild-type (WT) KACC10859 strains, and the lesion lengths were monitored for 2 weeks. Lesion lengths were significantly reduced in rice leaves inoculated with the CBNUXO14 strain (8.5 ± 0.5 cm) compared to leaves inoculated with the WT KACC10859 (28.5 ± 1.5 cm) and the complementation strain CBNUXO15 (27.9 ± 1.1 cm) (Table 1). This result indicated that loss of the rpfE gene function substantially reduced the virulence of Xoo.

The disease progress in the inoculated rice leaf was analyzed by time-course monitoring the lesion length and bacterial population using Dongjin 1 cultivar. The lesion length was enlarged rapidly after 7 days post inoculation in wild-type inoculated leaves while the extent and enlargement of lesion was much less in the rpfE-mutant inoculated leaves (Fig. 1). The mutant population (CBNUXO14) was about 10-50 times lower than the wild type population in first 3 days in the inoculation leaflet (Fig. 2, 1^st). The rpfE mutant did not build up in the 3^rd-5^th leaflets, the lower parts of the inoculation leaf, at 7 or 10 days post inoculation (Fig. 2). This result suggests the rpfE mutant did not move to downward from the inoculation point efficiently.

To determine whether the production of virulence factors may be affected in the rpfE mutant, extracellular enzyme (cellulase and xylanase) activities and EPS levels were assayed in the CBNUXO14 and CBNUXO15 strains. In the CBNUXO14 strain, cellulase activity was significantly reduced to about 85% of that of the WT KACC10859 strain (Table 1). The EPS dry weight was also reduced to about 40% of that of the WT KACC10859 dry weight (Table 1). The reduction of cellulase activity and EPS production were completely restored by complementation of Xoo rpfE with pML122 (Table 1). However, xylanase activity of the CBNUXO14 strain did not differ from that of the WT KACC10859 strain (Table 1). These results demonstrate that inactivation of rpfE induced reduction of cellulase activity and EPS production in Xoo.

To test the hypothesis that reduction of cellulase activity in the CBNUXO14 strain is dependent on DSF signaling mediated by RpfC/RpfG TCS, cellulase activity in rpfC and rpfG mutant strains was also investigated (Fig. 3). Although reduction of cellulase activity of these mutant strains was observed, it remained more than 50% of that of the WT strain in both mutant strains (Fig. 3).

Another virulence factor, group motility was tested in the rpfE mutant strain. CBNUXO14 had significantly reduced motility in a swarming assay, whereas motility of the complementation strain CBNUXO15 was similar to that of the WT strain (Table 1).

We measured the expression of genes involved in Xoo pathogenicity using quantitative real time-polymerase chain reaction (qRT-PCR) to test if the phenotypes observed in the CBNUXO14 strain resulted from changes in gene expression. The expression of related genes in EPS production (gumG and gumM), swarming motility (flhF), and cellulase activity (engXCA and celS) was strongly reduced in the rpfE mutant strain, CBNUXO14, compared to the WT KACC10859. The expression of LPS-related genes (xanA and xanB) and the xylanase production-related gene (xynB) did not differ (Fig. 4). The complementation strain, CBNUXO15, did not differ from the WT strain KACC10859 in the expression of these genes (Fig. 4). These results show that the knock-out of Xoo rpfE decreased the expression of genes involved in EPS synthesis, motility, and cellulase production, which probably resulted in a decline in Xoo virulence.

To test the RpfE requirement for DSF synthesis, we compared DSF production in rpfC, rpfE, and rpfF mutant strains with that in the WT strain (Fig. 5). Using thin layer chromatography (TLC), we analyzed DSF (cis-11-methyl-2-dodecenoic acid) in extracts of suspension medium from each strain culture. The retention factor (Rf) value for the DSF standard was 0.683 (lane 1 in Fig. 5). WT, DSF-overexpressing mutant (rpfC) and rpfE mutant strains produced visible spots under UV illumination that migrated with the DSF standard. The calculated Rf values of the spots from each mutant strain closely matched that of the DSF standard (wild type, 0.675; rpfC, 0.678; rpfE, 0.675), suggesting that rpfC and rpfE mutant strains produced DSF (Fig. 5). The rpfF mutant strain, however, does not produce DSF because rpfF is required for DSF synthesis and the strain is DSF- deficient (He et al., 2010). The result of the TLC analysis indicated that rpfE deficiency does not affect DSF synthesis (Fig. 5).

Pathogenic bacteria in the host environment must assess local conditions, including nutrient availability, and adjust their virulence functions accordingly (Feng et al., 2009; Nakamura et al., 2006; Zhao et al., 2012). We tested the growth of the rpfE mutant strain under different carbon source to determine the response of rpfE expression to nutrient conditions. The rpfE mutant strain showed almost no growth on M9 medium containing sucrose and xylose, whereas the growth rate recovered on M9 medium containing glucose (Fig. 6). These results suggested that rpfE expression is sensitive to the carbon source and may regulate genes involved in carbon source utilization, especially for sucrose and xylose.