INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Although many studies have been devoted to controlling the disease, bacterial wilt of tomato plants caused by Ralstonia solanacearum still commonly occurs in tropical and warm temperate areas of the world (5). However, some soils are known to have a low incidence of the disease (7). In an area of Yamaguchi Prefecture, Japan, tomato plants have been cultivated with little occurrence of bacterial wilt despite more than a decade of continuous monoculture, so we began research to determine how the occurrence of bacterial wilt is suppressed in this area. In a previous report (14), we compared the population dynamics of R. solanacearum in nonrhizosphere soil, rhizoplane, roots, and stems of tomato plants in a wilt-conducive soil and in a suppressive soil that were both newly infested with the pathogen. Survival of the pathogen in nonrhizosphere soil was better in the suppressive soil than in the conducive soil, while multiplication of the pathogen at the tomato rhizoplane was better in the conducive soil than in the suppressive soil. Therefore, we deduced that the suppression of the multiplication of the pathogen at the rhizoplane was one of the reasons for the lower incidence of bacterial wilt of tomato in this suppressive soil.

In this study, to evaluate the role of biotic factors in the suppressiveness of tomato bacterial wilt, we bioassayed tomato plants in three treated soils. Then, the types of rhizoplane bacteria were characterized by 16S ribosomal DNA (rDNA)-restriction fragment length polymorphism (RFLP) analysis. In addition, the growth on pectin, which is thought to be a main root exudate of tomato (15), and the polygalacturonase activity (8, 16, 18) of rhizoplane microorganisms from the tested soils were investigated.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Soil samples. Two soil samples, Mutsumi and Yamadai (14), were used in this study. Mutsumi soil was obtained from a commercial greenhouse field in Yamaguchi Prefecture in May 1996. The sampled field has received annual dressings of cow manure, and tomato plants have been cultivated there for 8 consecutive years with little occurrence of bacterial wilt. Yamadai soil was taken from a field of ryegrass at Yamaguchi University. Both soil samples were sieved (4 mm) and stored at room temperature with a water potential of 25 kPa until further experiments. Soil pH, texture, and organic carbon and total nitrogen contents are shown in Table 1. Before each experiment, soil samples were preincubated for 2 weeks at 30°C with a water potential of 6.2 kPa.

Assessment of suppressiveness of tomato bacterial wilt in treated Mutsumi soil. Mutsumi soil was separated into three subsamples and treated as follows to modify its microbial characteristics. Two of the subsamples were autoclaved for 20 min at 121°C on 3 consecutive days to sterilize them completely and then mixed with either intact Mutsumi soil or intact Yamadai soil (ratio of 9:1 by dry weight). The other subsample was fumigated with chloroform vapor in a vacuum chamber for 24 h because this treatment is assumed to have little influence on the physicochemical properties of the soil in contrast to autoclaving, although it does not completely sterilize the microorganisms in soil.

Isolation of root surface bacteria. To isolate the rhizoplane bacteria, tomato plants were cultivated in Mutsumi and Yamadai soils as described above except that the pathogen was not inoculated into the soil. At about 2 weeks after the transplanting, tomato seedlings were carefully removed from two pots. Five seedlings grown in each pot were combined, and soil particles adhering to the roots were removed in sterile water by using sterilized forceps. Each set of roots was then weighed and washed by shaking four times in a 100-ml flask containing sterile water (20 times the weight of the fresh roots) to collect rhizoplane bacteria, as was reported previously (14). All of the washing solution from each set of root samples was collected in a 100-ml flask, serially diluted, and spread on the surface of YG agar plates (yeast extract, 1.0 g; K₂HPO₄, 0.3 g; KH₂PO₄, 0.2 g; MgSO₄ · 7H₂O, 0.2 g; agar, 15 g; water, 1 liter [pH 6.8]). Then, 250 colonies on the YG agar plate at a dilution level of 10⁵ were isolated from the rhizoplane of tomato plants grown in each soil type. The isolated bacteria were suspended in 50% glycerol and kept at 80°C until further experiments.

DNA extraction from rhizoplane isolates. Bacterial cells were lysed by different methods depending on whether they were gram negative or gram positive. Gram-negative bacteria were grown overnight in 10% solution of nutrient broth (Difco), after which 1.5-ml samples of culture were pelleted by centrifugation and washed twice in 1 ml of 1 M NaCl. Bacterial pellets were resuspended in 0.3 ml of lysis buffer (0.1 M Tris, pH 9.0; 0.1 M NaCl; 1% sodium dodecyl sulfate) and vortexed. After centrifugation, the supernatants were transferred to clean 1.5-ml tubes. Gram-positive bacteria were cultured overnight on YG agar plates. Each colony was suspended in 0.3 ml of lysis buffer containing proteinase K (1 mg/ml) (Boehringer Mannheim), incubated overnight at 37°C, and then incubated at 100°C for 5 min to inactivate the proteinase K. After centrifugation, the supernatants were transferred to clean 1.5-ml tubes and purified in a 1-ml mixture containing phenol, chloroform, and isoamyl alcohol (25:24:1). The DNA extracts were diluted 1:10 and used for PCR.

16S rDNA amplification. PCR was carried out on samples of individual extracts of DNA by using the following primers: 5'-AGA GTT TGA TCM TGG CTC AG-3' (positions 8 to 27 in Escherichia coli numbering) as the forward primer and 5'-TAC CTT GTT ACG ACT T-3' (positions 1507 to 1492 in E. coli numbering) as the reverse primer. Each 50-µl sample of PCR reaction mixture consisted of 5 µl of 10× PCR buffer (TaKaRa), 4 µl of 0.2 mM deoxynucleoside triphosphate mixture (TaKaRa), 1 µl of each primer (500 nM), 1 U of Ex-Taq DNA polymerase (TaKaRa), 1 µl of template DNA, and sterile distilled water to up to 50 µl of reaction volume. PCR amplifications were performed in an automated thermal cycler with an initial denaturing (94°C for 1 min), followed by 35 cycles of denaturation (94°C for 1 min), annealing (50°C for 1 min), and extension (72°C for 1.5 min) and concluded by a single final extension (65°C for 1 min). After PCR, the amplified DNA was purified by polyethylene glycol precipitation (6), vacuum dried, and dissolved in 10 µl of sterile distilled water.

Restriction endonuclease digestion and gel electrophoresis. To generate restriction fragments, these PCR products were digested for 10 h at 37°C with the restriction endonuclease HinfI (10 U) and MboI (10 U) in the reaction buffer supplied by the manufacturer. For each digestion reaction, 3 to 10 µl of amplified DNA was used, depending on the amount of amplified DNA. Digests were separated on 2% agarose gel in 1× TBE with a molecular marker, 100-bp DNA Ladder (GIBCO-BRL), and characterized by gel image analysis (Gel Doc 100; Bio-Rad).

Dendrogram and cluster analysis. To compare the types and diversity of bacteria present at the rhizoplanes of the tomato plants grown in Mutsumi and Yamadai soil samples, band patterns of 16S rDNA digested with HinfI and MboI were analyzed for genetic similarity (cluster analysis) by Ward methods (17). Dendrograms were constructed according to the genetic similarity.

Sequence analysis of 16S rDNA. After the results of the cluster analysis were considered, strains for partial 16S rDNA sequencing were randomly selected in the dominant clusters in Mutsumi and Yamadai rhizoplane bacteria to ascertain their phylogenetic positions. DNA sequences corresponding to positions 8 through 531 in E. coli 16S rRNA numbering were determined with an automated DNA sequencer (ABI 377; Perkin-Elmer). Homology searching of the sequence data was carried out in comparison with sequences in the nucleotide database at DDBJ/EMBL/GenBank by using BLAST (1).

Polygalacturonase activity of rhizoplane microorganisms. Tomato rhizoplane microorganisms were extracted by four washes of the roots 2 weeks after the transplantation as described above. The number of bacteria in the washing solution was measured by direct microscopic counting after ethidium bromide staining (19). Then, 10⁷ bacterial cells were inoculated into pectin medium (final volume, 10 ml; 10 g of pectin, 0.2 g of MgSO₄ · 7H₂O, 0.5 g of NH₄NO₃, 0.5 g of KH₂PO₄, 0.002 g of ZnSO₄ · 7H₂O, 0.001 g of CuSO₄, 0.001 g of MnCl₂, and 0.001 g of Na₂MoO₄ in either 1 liter of 0.05 M [pH 4.0] sodium acetate buffer or 1 liter of 0.05 M [pH 6.0] sodium citrate buffer; final concentration of bacteria, 10⁶ cells/ml). Some of the pectin media were supplemented with cycloheximide (final concentration, 100 mg/liter) to suppress the fungal activities. Inoculated samples were incubated at 30°C shaken at 150 rpm for 12 h; the number of bacterial cells in the samples was then determined by microscopic counting (19). Subsequently, to provide clear supernatants to use in measuring polygalacturonase activity (3), the samples were centrifuged at 10,000 rpm for 15 min. Reaction mixtures containing 0.5 ml of the supernatant, 0.5 ml of 0.5% polygalacturonic acid, and 0.005 M EDTA were incubated at 37°C for 60 min in either 0.1 M sodium acetate buffer (pH 4.0) or 0.1 M sodium citrate buffer (pH 6.0). The reaction was terminated by the addition of Somogy's reagent (20); the amount of reducing sugar was then measured by using the method of Nelson (13). One unit of polygalacturonase activity was defined as the amount of enzyme that produced 1 µmol of oligogalacturonate per min under the assay conditions.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Severity of tomato bacterial wilt in variously treated Mutsumi soils. To demonstrate that the suppression of tomato bacterial wilt in Mutsumi soil was due to the organisms residing in it, we used chloroform-fumigated soil since the fumigation is assumed to have little influence on the physicochemical properties of the soil compared to the autoclaving process. After 30 days of preincubation, soil microbial diversity, based on the appearance of the colonies in the fumigated soil, appeared to be reduced compared with that in the nonfumigated soil, as was reported previously (22), although the number of microorganisms was higher in the fumigated soil (data not shown).

View larger version (16K): [in this window] [in a new window]   FIG. 1.   Occurrence of tomato bacterial wilt grown in chloroform-fumigated () and nonfumigated () Mutsumi soil infested with R. solanacearum SL8 at an initial density of 103 CFU/g of soil. Standard errors are indicated by bars or are within each symbol.

View larger version (17K): [in this window] [in a new window]   FIG. 2.   Occurrence of tomato bacterial wilt grown in nine parts autoclaved Mutsumi soil mixed with either one part intact Mutsumi soil () or 1 part intact Yamadai soil () and infested with R. solanacearum SL8 at an initial density of 103 CFU/g of dry soil. Standard errors are indicated by bars or are within each symbol.

RFLP diversity of rhizoplane isolates. We carried out our initial assessment of bacterial diversity by using the plate count method. The number of culturable bacteria was slightly higher in samples from the rhizoplanes of tomato plants grown in Mutsumi soil than from those grown in Yamadai soil (Table 2). Mutsumi samples also had greater numbers of culturable actinomycetes and fluorescent pseudomonads than in the Yamadai samples. The Mutsumi samples had 18 times more culturable fluorescent pseudomonads than the Yamadai samples. The morphology of the colonies that formed on YG agar from Mutsumi samples showed more diversity than those from the Yamadai samples (data not shown).

View larger version (24K): [in this window] [in a new window]   FIG. 3.   Relationships among 22 groups composed of 316 isolates of culturable bacteria from the rhizoplanes of tomato plants. The square distance (genetic similarity) was determined from 16S rDNA-RFLP patterns obtained by digestion with HinfI and MboI by the Ward method. Isolates within 130 of the square distance were combined into a group.

View larger version (16K): [in this window] [in a new window]   FIG. 4.   Percentage of 16S rDNA-RFLP pattern groups in the rhizoplanes of tomato plants in Yamadai (gray bar) or Mutsumi (black bar) soils.

Ascertainment of the phylogenetic position of the dominant groups of culturable isolates at the tomato rhizoplane. The phylogenetic positions of several strains from the dominant bacterial groups of the Yamadai and Mutsumi tomato rhizoplane samples were estimated through an analysis of 16S rDNA sequences. Table 3 shows the results of homology searching with BLAST. Thirteen strains examined in group 19 were close to Burkholderia pickettii (homology, 97.6%), which belongs to the -subdivision of proteobacteria. The strains of group 17, predominant in both tomato rhizoplanes, showed high homology to Xanthomonas sacchari (-proteobacteria; homology, 93.5%), Burkholderia sp. (-proteobacteria; homology, 92.4%), and Hydrogenophaga palleronii (-proteobacteria; homology, 94.5%). All examined strains from group 20 matched well with Pseudomonas sp. strain P51 (-proteobacteria; homology, 97.6%). The results indicate that the major part of the bacterial isolates in the tomato rhizoplane from both soils are -proteobacteria.

Polygalacturonase activity and growth rate of microbial communities at the tomato rhizoplane. Rhizoplane microorganisms from Mutsumi soil showed a higher polygalacturonase activity both at pH 6.0 and 4.0 than those from Yamadai soil (Table 4). The number of bacteria after 12 h of incubation was also higher in samples from Mutsumi soil than in those from Yamadai soil. These results suggest that the rhizoplane microbial population in the Mutsumi soil samples had a greater ability to assimilate pectin than did the Yamadai microbial population.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

In our previous study, we found that under laboratory conditions tomato plants grown in Mutsumi soil were not as susceptible to bacterial wilt as those grown in Yamadai soil and that the growth of R. solanacearum was suppressed at the rhizoplane of tomato plants in Mutsumi soil (14). In order to evaluate the biotic factors that suppress tomato bacterial wilt in Mutsumi soil, we compared the incidence of bacterial wilt in soils that were treated in three different ways. In chloroform-fumigated soil, where the physicochemical properties of the soil are assumed to be quite similar to those of nonfumigated soil, the incidence of bacterial wilt was greater than in nonfumigated Mutsumi soil (Fig. 1), indicating that the organisms in natural Mutsumi soil are involved in the suppressiveness of bacterial wilt. In the Mutsumi soil that was autoclaved and mixed with intact Yamadai soil, the incidence of bacterial wilt was higher than when autoclaved Mutsumi soil was mixed with intact Mutsumi soil (Fig. 2). Owing to the low ratio (9:1) of the mixtures, the physicochemical properties of the soils were likely similar, so the native microorganisms in the intact Mutsumi soil must be responsible for the suppressive ability of Mutsumi soil. One hypothesis is that the suppression of colonization and multiplication of R. solanacearum at the rhizoplane in Mutsumi soil (14) is due to the competitive ability of the indigenous microbial community present at the rhizoplane.

To compare the make-up of the populations of bacteria present at the rhizoplane in wilt-suppressive and -conducive soils, we characterized culturable bacteria at the rhizoplane by 16S rDNA-RFLP analysis. At the Yamadai rhizoplane, a soil conducive to tomato bacterial wilt, a single type (group 19) predominates (73.1%). In contrast, the distribution of 16S rDNA-RFLP clusters showed a much more even distribution of diverse bacteria at the Mutsumi rhizoplane (Fig. 3).

In the rhizosphere, where the plant meets the soil, there can be steep gradients in water content, pH, and other environmental factors (2). A rhizobacterial community with greater diversity would present more candidates suitable for the plant-influenced environment and would provide more competition to the extreme dominance of a single species at the rhizoplane when roots alter environmental conditions. Thus, R. solanacearum might fail to colonize and multiply at the tomato rhizoplane in Mutsumi soil. In contrast, a single species, R. solanacearum, might easily dominate at the tomato rhizoplane in Yamadai soil owing to the simple composition of the indigenous bacterial population.

In vitro enzyme assay revealed that the rhizoplane of tomato plants in Mutsumi soil has a bacterial population that vigorously utilizes pectin (Table 4). This result, together with previous reports (9, 10), implies that, in Mutsumi soil, the presence of bacteria with a strong competitive ability to utilize pectin at rhizoplane retards the multiplication of R. solanacearum.

Further work is needed to clarify the mechanism of tomato bacterial wilt suppression in Mutsumi soil.