Induction of the Viable but Nonculturable State of Ralstonia solanacearum by Low Temperature in the Soil Microcosm and Its Resuscitation by Catalase

Ralstonia solanacearum is the causal agent of bacterial wilt on a wide variety of plants, and enters a viable but nonculturable (VBNC) state under stress conditions in soil and water. Here, we adopted an artificial soil microcosm (ASM) to investigate the VBNC state of R. solanacearum induced by low temperature. The culturability of R. solanacearum strains SL341 and GMI1000 rapidly decreased at 4°C in modified ASM (mASM), while it was stably maintained at 25°C in mASM. We hypothesized that bacterial cells at 4°C in mASM are viable but nonculturable. Total protein profiles of SL341 cells at 4°C in mASM did not differ from those of SL341 culturable cells at 25°C in mASM. Moreover, the VBNC cells maintained in the mASM retained respiration activity. Catalase treatment effectively restored the culturability of nonculturable cells in mASM, while temperature increase or other treatments used for resuscitation of other bacteria were not effective. The resuscitated R. solanacearum from VBNC state displayed normal level of bacterial virulence on tomato plants compared with its original culturable bacteria. Expression of omp, oxyR, rpoS, dps, and the 16S rRNA gene quantified by RT-qPCR did not differ significantly between the culturable and VBNC states of R. solanacearum. Our results suggested that the VBNC bacterial cells in mASM induced by low temperature exist in a physiologically unique state.


Introduction
Ralstonia solanacearum is a plant pathogenic bacterium that causes severe bacterial wilt on over 400 plant species across 50 families. Bacterial wilt by R. solanacearum causes significant yield losses of many economically important crops [1]. However, effective management practices are not available for bacterial wilt on a variety of plant species. Additionally, chemical control cannot be used to eliminate the bacterial pathogen from soils and waters. Cultivation of resistant cultivars is an option, but resistant stocks are available for only specific plant species [2]. This bacterium can survive in the soil or water for long periods of time and infects host plants through the roots [3,4,5]. It has been documented that R. solanacearum survives in water or soil for extended periods of time, and can cause disease outbreaks when environmental conditions turn favorable for infection [5,6]. Therefore, long-term survival of R. solanacearum could be a threat to crop cultivation, which complicates bacterial wilt management.
Investigation of the long-term survival of R. solanacearum is dependent on bacterial cultivation. However, cultivation-dependent bacterial detection has some limitations. For example, Gramnegative bacteria exposed to stresses cannot be cultivated, but stably maintain their viability. This phenomenon was first termed a viable but nonculturable (VBNC) state in Escherichia coli and Vibrio cholera that were not cultivated under stress conditions [7]. For R. solanacearum to survive in soil or water habitats before infection of host plants, bacteria must be able to tolerate biotic and abiotic stress conditions such as nutrient competition, antibiotic production from other bacteria, UV radiation, and temperature variations. Many bacteria enter VBNC states under various environmental stresses, and VBNC induction has been reported for many non-spore-forming bacteria [8,9,10,11,12]. It has been reported that R. solanacearum strains also enter a VBNC state upon low temperature, starvation or copper stress [6,13,14,15,16]. VBNC induction by copper exposure or low temperature has been also demonstrated in other bacteria [17] including a plant pathogenic Agrobacterium tumefaciens and a legume plant symbiotic Rhizobium leguminosarum [18], and Erwinia amylovora [19].
To characterize bacterial VBNC, it is important to examine the recovery of culturability (resuscitation) of the presumed VBNC bacteria. The VBNC state of Vibrio strains were induced using many stress factors, and their culturability was restored when stress factors were removed [20,21]. However, some bacteria do not return to their culturable state after elimination of the stress factors [22]. Recently, microbial autoinducers and antioxidants have been evaluated for the recovery of bacterial cell culturability [23,24]. Attempts to resuscitate R. solanacearum cells in the VBNC state have been made, and amendment of sodium pyruvates was shown to revive VBNC cells [24,35].
Gene expression in several VBNC bacteria has been investigated. The protein expression profile of Enterococcus faecalis in a culturable and VBNC state has been compared [26]. In V. cholera, mRNA expression levels in VBNC cells were monitored over time using reverse-transcription quantitative polymerase chain reaction (RT-qPCR), and the total transcriptional profile of the VBNC state was analyzed using a microarray method [27,28]. However, changes in bacterial physiology and gene expression in the VBNC state remain poorly characterized.
The VBNC state of soil bacteria is difficult to analyze because natural soils contain various organisms and unknown compounds. Furthermore, soils from different origins are heterogeneous, which make experimental results difficult to reproduce. Therefore, a specific and uniform soil microcosm system is necessary to investigate soil microbial ecology. Ellis [28] developed the artificial soil microcosm (ASM) to model the microbial ecology of soil, which comprises the essential components of soil. In this study, we adopted this ASM to investigate the VBNC state of R. solanacearum strain SL341 (phylotype I, race 1). The main aim of this study was to characterize the VBNC state of R. solanacearum in the modified ASM (mASM) in comparison with culturable cells. We found that low temperature can induce the VBNC state of R. solanacearum in the soil microcosm, while catalase treatment restores the culturability of VBNC cells.

Bacterial strains and growth conditions
Two Ralstonia solanacearum strains were used to investigate the correlation between low temperature and bacterial survival in artificial soil systems. R. solanacearum strain SL341 was isolated from diseased tomato plants and identified as race 1, biovar 3, and phylotype I [30]. Another strain, GMI1000, was originally isolated from tomato in French Guyana, and its genome has been completely analyzed [31]. These strains were routinely cultured in casamino acid peptone glucose (CPG) broth [32] or CPG supplemented with 0.005% (w/v) 2,3,5-triphenyl tetrazolium chloride (TZC) medium at 30uC [33]. When necessary, tryptic soy agar (TSA) and nutrient agar (NA) were used to investigate bacterial culturability.

Artificial soil microcosm and induction of a nonculturable state
Each of the components, including sand, kaolinite, bentonite, CaCO 3 and humic acid, of the ASM were purchased from Sigma-Aldrich and used in the following quantities: 35 g, 10 g, 5 g, 0.1 g and 1 g, respectively [29]. ASM was sterilized twice by autoclaving. Modified ASM (mASM) was ASM lacking humic acid.
R. solanacearum SL341 and GMI1000 cells were cultured in CPG broth at 30uC for 24 h with mild shaking at 150 rpm, and were harvested by centrifugation at 16,1006g for 3 min. The bacterial cells were washed once with distilled sterile water, harvested again by centrifugation at 16,1006g for 3 min, and finally resuspended in distilled water to a density of 10 9 CFU/mL. The mASM, which typically consisted of 50 g of the component mixture, was hydrated with 15 mL of adjusted bacterial cell suspension in 125-mL Erlenmeyer flasks and incubated at 25uC and 4uC. The bacterial cell densities were monitored over time using the dilution plating count method on CPG medium with 1 g of mASM suspended in sterile water. The experiments were conducted in triplicate.

Total proteins and total DNA extraction from mASM
Isolation of proteins in mASM containing R. solanacearum SL341 was performed by modifying a metaproteome extraction and purification method [34]. Briefly, 5 g of the mASM were suspended in 10 mL protein extraction buffer (50 mM Tris-Cl (pH 8.0), 10 mM EDTA (pH 8.0), 0.5% (w/v) SDS, 1% (w/v) bmercaptoethanol) for 30 min at room temperature, and cells in mASM were homogenized by ultrasonication (Sonic Dismembrator 500, Branson, USA). The ultrasonic wave amplitude was set at 10% of the maximum power with an irradiation pulse (10 sec on and 10 sec off) for 1 min. The suspension was mixed with 10 mL of phenol (pH 8.0) and 3 g of sucrose prior to centrifugation. The upper layer of about 10 mL of the phenol phases was transferred to a new centrifuge tube and proteins were precipitated using a fivefold volume of 0.1 M ammonium acetate in methanol at 220uC for 2 h. Protein pellets were then centrifuged at 16,1006g for 15 min at 4uC and washed in 80% acetone. Total proteins were dissolved in 20 mL deionized water, resolved by SDS-PAGE (14%), and visualized using Coomassie brilliant blue G-250 staining.
Total genomic DNA was isolated from culturable cells at 25uC and nonculturable cells at 4uC in mASM over time. mASM (1 g) from two temperature conditions was immediately used to isolate total DNA. DNA extraction from the soil microcosm was performed as described previously and analyzed by agarose gel electrophoresis [35]. Total DNA from mASM were quantified by Nanodrop 2000 (Thermo Fisher Scientific, MA, USA).

Bacterial respiration analysis
Bacterial respiration was investigated using 5-cyano-2, 3ditolyltetrazolium chloride (CTC) staining (Invitrogen, Camarillo, USA) following the manufacturer's instructions. Briefly, mASM (1 g) containing bacterial cells after incubation at different temperatures was suspended in 9 mL of sodium-phosphate buffer (pH 7.0). The culturability of bacterial cells in mASM was examined by spreading the cells on CPG agar media. When bacterial cells in mASM at 4uC were completely induced to the VBNC state by 30 days of incubation, 5 mL of a 50 mM CTC staining solution were added to 50 mL of mASM suspensions and incubated for 2 h at 37uC in the dark. The CTC-stained cell suspension was observed using an Olympus BX50 fluorescence microscope at 4006 magnification. Dead bacterial cells and aseptic mASM lacking bacteria were used as negative controls. To prepare dead cells, R. solanacearum SL341 cells actively grown in CPG broth were harvested by centrifugation at 16,1006g for 10 min at room temperature. The bacterial pellet was quickly resuspended in 70% isopropanol, incubated for 1 h, and finally harvested by centrifugation at 16,1006g for 10 min at room temperature [36]. The dead cells were washed and re-suspended in sodium-phosphate buffer (pH 7.0) for CTC staining.

Resuscitation of nonculturable bacteria
To evaluate the culturability of bacterial cells in mASM, mASM (1 g) containing nonculturable cells was suspended in 10 mL of distilled water, and 100 mL of the suspension were spread on different complex media for R. solanacearum, such as CPG, TZC and TSA agar media. To confirm the absence of culturable cells in the mASM with SL341 at 4uC (which causes re-growth of a small number of culturable cells present in the mASM), 10 mL of soil supernatant containing the bacterial cells were recovered from soil particles by pipetting. The bacterial cells in the suspension were subsequently harvested by centrifugation at 16,1006g for 5 min. The pellet was resuspended in 1 mL of sterile water and the suspension was spread on several CPG, TZC, and TSA agar plates to confirm that no R. solanacearum colonies formed.
Recovery of nonculturable bacterial cells after a temperature increase was performed as follows; 1 g of mASM containing nonculturable bacteria cells was suspended in 10 mL of sterile water and incubated at 30uC for 3 days. The sample suspension exposed to 30uC was investigated using the dilution plating count method described above. To examine their effects on nonculturable cells in mASM, we supplemented the suspension with the following compounds: thiamine hydrochloride, catalase, sodium pyruvate, CPG broth, V8 juice and acyl-homoserine lactone (AHL). Briefly, sodium pyruvate (0.2% (w/v)), thiamine hydrochloride (1.5 mg/mL) and 100 mM AHLs (3-oxo-C 8 -homoserine lactone (HSL), C 8 -HSL, and C 6 -HSL) were added directly to the sterile water. mASM (1 g) was resuspended in 10 mL of sterile water containing the compounds mentioned above. The negative control was mASM suspended in sterile water lacking chemical supplements. The soil microcosm suspensions were incubated at 30uC for 3 days, and bacteria were enumerated using the dilution plating count method on several culture media.

Plant growth and bacterial virulence assays
Tomato cultivar Moneymaker was used for bacterial virulence assays. The surface sterilized tomato seeds were germinated and tomato plants were grown in pots containing commercial horticulture nursery-media soil (Punong Co., Ltd, Korea) and maintained in the growth chamber for 3 weeks at 28uC. The virulence of the resuscitated cells and culturable SL341 strain were investigated in 3-week-old tomato plants 'Moneymaker' using soil soaking inoculation [32]. Bacterial inoculums were prepared by growing bacterial cells in CPG broth at 30uC for two days and resuspending bacterial cells in sterile water. The inoculum were adjusted the O.D.600 of 0.5 (4,5610 8 CFU/mL) with distilled water. For soil soaking method, each tomato was poured bacterial suspension with the final concentration ,10 7 CFU/g soil. A total of randomly selected 10 different resuscitated bacterial strains used for bacterial virulence assay with their wild type strain SL341. All inoculations included fifteen plants for each strain and the noninoculated controls with three replications. All inoculated plants were incubated in a growth chamber at 28uC in light for 14 h and in the dark for 10 h. Plants were monitored for disease progress over time after inoculation, and bacterial wilt was rated using the following scale: 0, no wilting; 1, 1-25% leaves wilted; 2, 25-50 leaves wilted; 3, 51-75% leaves wilted; 4, 76-100% leaves wilted.

Effect of catalase on bacterial resuscitation
The resuscitation of nonculturable cells by catalase was performed as follows. mASM (1 g) with nonculturable cells was inoculated into 10 mL of sterile water or yeast peptone (YP) medium (yeast extract, 0.25% (w / v); peptone, 0.5% (w / v)), and the suspension was divided into two tubes. One of the tubes was supplemented with catalase to a final concentration of 1,000 U/mL and incubated at 30uC for 3 days; bacteria were then enumerated using the plate count method. Catalase from bovine liver was purchased from Sigma Corp., suspended in double-distilled water, and sterilized by passing through 0.2-mm pore size membrane filters.
Bacterial resuscitation was investigated using catalase treatment in mASM containing R. solanacearum SL341 at both 25uC and 4uC. Sterile water with or without autoclaved catalase (inactivated catalase) were used as negative controls. mASM (1 g) with SL341 incubated at 25uC and 4uC was suspended in distilled sterile water. The suspension was divided into three tubes; two were supplemented with catalase or autoclaved catalase and incubated for 3 days at 30uC. Bacteria were then enumerated using the plate count method on CPG medium following incubation for 2 days. The experiments included three replicates.

RNA extraction from the soil microcosm and reversetranscription quantitative PCR (RT-qPCR)
The RNA expression levels of R. solanacearum SL341 in mASM were determined by RT-qPCR. Bacterial total RNA was isolated using the Hybrid-R RNA extraction kit (GeneAll Bio Inc., Seoul, Korea) from 1 g of mASM at various stages. The RNA was eluted in 30 mL of RNase-free water, and 5 mL were directly used as template for cDNA synthesis by the Prime Script RT Master Mix kit (TaKaRa Biotech. Co., Japan), following the manufac- turer's instructions. Potential contamination of DNA in the RNA was excluded by DNaseI treatment prior to cDNA synthesis. Quantitative PCR using the cDNA was conducted using a CFX384 Real Time System (Bio-Rad, Hercules, CA, USA). The qPCR reaction components contained SYBR Premix Ex Taq mix (TaKaRa Biotech. Co., Japan), 1 mL of diluted cDNA template, 10 mM both forward and reverse primers (Table S1), and RNase free water. Thermal cycling included two reaction steps; an initial preheat for 3 min at 95uC, followed by 40 cycles at 95uC for 5 sec, 60uC for 10 sec (50uC for 10 sec for the V3 primers), and 72uC for 35 sec. The qPCR data were displayed using the CFX manager software version 1.6. The oligonucleotide primers were designed based on the R. solanacearum GMI1000 sequence (Table S1).
Standard curves between the cDNA concentration and C (t) value of qPCR for individual genes were obtained using various DNA concentrations of target DNA for qPCR. First, target DNA was obtained by PCR using the indicated primers (Table S1), and the concentration of the purified PCR product was measured using a NanoDrop 2000 spectrophotometer (Thermo Scientific, Wilmington, USA). The template DNA was diluted serially 10-fold to various concentrations, and the corresponding C (t) value was obtained by qPCR. Finally, standard curves were generated by plotting DNA concentrations and their corresponding C (t) values for individual genes.
RT-qPCR results of individual genes were evaluated using the iCycler iQ Real-Time PCR Detection System. The C (t) values of qPCR products of each gene were used to determine the target cDNA concentration based on the standard curves. The molar ratio of cDNA copy numbers of each gene was determined using the following equation: The molar ratio of cDNA copies = the amount of cDNA (ng)/ the molecular weight of expected PCR product of dsDNA (ng) The amount of cDNA was determined based on the cDNA concentration after multiplying by the fold-dilution of each gene. The molecular weight of the expected PCR product was calculated using the Sequence Manipulation Suite software (http://www.bioinformatics.org/sms2/index.html). The cDNA copies for each gene of R. solanacearum in mASM were compared to the number of cDNA copies of the 0 day bacterial incubation in mASM at 25uC. The RT-qPCR experiments included three replicates.

Bacterial viability in mASM at two temperatures
We adopted the ASM system to investigate R. solanacearum survival in soil. ASM was first developed by Ellis [29] to contain the majority of essential soil components. However, the presence of humic acid, which is a nutrient component of ASM, from different manufacturers (or even the same company) resulted in inconsistent survival rates of R. solanacearum SL341 and GMI1000. Therefore, we modified the ASM by excluding humic acid, and referred to the modified ASM as mASM. Interestingly, culturable cells of strain SL341 were stably maintained at 25uC in mASM for 13 days. In the same system, culturable R. solanacearum GMI1000 strain also maintained its population at 25uC for 13 days, albeit with a slightly reduced cell density (Fig. 1). Since low temperature stress induces the nonculturable state of R. solanacearum in water [14], we inoculated R. solanacearum SL341 and GMI1000 in mASM, incubated at 4uC, and enumerated culturable bacterial cells. The culturability of SL341 and the GMI1000 decreased over time at 4uC. Finally, no culturable SL341 or GMI1000 cells were detected on nutrient agar, CPG, or minimal medium after 11 days and 7 days of incubation in mASM (Fig. 1). This suggested that mASM could be used to investigate the ecology of R. solanacearum strains and its survival at different temperatures.

Viability of nonculturable bacteria
Because low temperature (4uC) suppresses the culturability of R. solanacearum strains, cells at such low temperatures may be dead.
To determine whether the bacteria are alive or dead at low temperatures in mASM, we compared total protein patterns and total DNA of R. solanacearum SL341 in mASM incubated at two temperatures. SDS-PAGE of total protein showed that the majority of proteins remained intact in R. solanacearum SL341 incubated for 14 days in mASM at both 25uC and 4uC, although the levels of some proteins were decreased slightly at 4uC (Fig. 2A). The content and size of total DNA of R. solanacearum SL341 in mASM at 25uC or 4uC were not remarkably different after incubation for 3, 7 and 14 days (Fig. 2B and C). When total DNA and RNA from the microcosm were quantified, any remarkable difference was not apparent between culturable state and nonculturable state (Fig. 3). Bacterial respiration is an important indicator of viability. We adopted the CTC staining method to investigate bacterial respiration in mASM. CTC staining has been used widely to examine bacterial respiration in various microbial communities [37,38] CTC responds to the bacterial electrontransport chain and becomes reduced to form formazan precipitates (CTF), which generates red fluorescence. Our analysis revealed the presence of respiring bacterial cells in mASM suspension after incubation with CTC solution for 2 h. Respiration was also apparent in bacterial cells in a culturable (25uC) and VBNC state (4uC) in mASM for 30 days (Fig. 4). The number of respiring bacterial cells (red fluorescence) in the VBNC state (Fig. 4B) was similar to that of respiring culturable cells (Fig. 4A). We confirmed complete loss of culturability of VBNC cells maintained in mASM for 30 days at 4uC, while the number of culturable bacterial cells in mASM for 30 days at 25uC was 2610 8 CFU/g of mASM. However, we could not detect respiration of dead cells in suspension or aseptic mASM lacking bacterial cells (Fig. 4C and D).

Catalase treatment resuscitates nonculturable bacteria
Recovery of viable bacterial cells (resuscitation) is used to determine the viability of R. solanacearum SL341 [39] at low temperatures in mASM. Therefore, we used several methods to resuscitate nonculturable SL341 cells in mASM, including supplementation and incubation for up to 3 days in the presence of CPG broth, V8 juice, tomato extract, thiamine hydrochloride, autoinducers such as AHLs, and pyruvates. None of the supplements promoted resuscitation of VBNC cells in mASM. An increase in temperature to 30uC in mASM did not result in resuscitation of nonculturable R. solanacearum SL341 cells at 14 days after incubation in mASM at 4uC (Fig. 5). However, an increase in the temperature of mASM containing a small number of culturable cells resulted in bacterial re-growth (see day 13, Fig. 5). Interestingly, catalase treatment resuscitated nonculturable R. solanacearum SL341 incubated in mASM at 4uC. Briefly, mASM containing nonculturable bacteria was suspended in YP broth or distilled water supplemented with 1,000 U/mL catalase. Following incubation at 30uC for 1 to 3 days, nonculturable SL341 formed bacterial colonies on TZC or TSA medium, indicative of resuscitation (Fig. 6). However, bacterial cells from the same mASM suspension lacking catalase did not form colonies (Fig. 6).
The recovered cell density was similar to the initial inoculum in the presence of catalase, while the cell density in YP broth with catalase increased compared to the inoculum, (Fig. 6). The catalase treatment also showed the resuscitation of GMI1000 strain (data not shown). Moreover, all of the resuscitated cells   maintained full virulence by soil soaking inoculation compared with culturable SL341 wild type strain to 3-week-old tomato plants (cv. Moneymaker), which is a tomato cultivar susceptible to bacterial wilt (Fig. 7).

Three phases of the nonculturable state of R. solanacearum
Resuscitation of SL341 in mASM by catalase over time was investigated. Nonculturable bacteria SL341 in mASM were not resuscitated by incubation in sterile water lacking catalase or with autoclaved catalase, although some resuscitated cells were present until 4 days after incubation in mASM at 4uC (Fig. 8A and B). It is likely that the resuscitated bacterial cells 4 days after VBNC induction were not in a true nonculturable state. Bacterial cells at 5 days after induction of the nonculturable state were not resuscitated by water alone or water containing autoclaved catalase, but were resuscitated 16 days after the addition of catalase (Fig. 8C). Therefore, the true nonculturable state of SL341 in mASM occurs ,12 days after transition to the nonculturable state. Moreover, nonculturable cells cannot be recovered to the culturable state by incubation with catalase for 12 days (after 30 days at 4uC in mASM) (Fig. 8C).
Therefore, the VBNC state can be divided into three phases: Phase I, pre-nonculturable stage (which still contains small number of culturable bacteria); Phase II, nonculturable stage (resuscitated by catalase); and Phase III, nonculturable stage (not resuscitated by catalase).

Gene expression by culturable and nonculturable bacteria
Gene expression by culturable and nonculturable bacteria at 25uC and 4uC in mASM was investigated by RT-qPCR. The stages for comparison included the initial SL341 inoculum, prenonculturable stage, and nonculturable stage, during which SL341 can be resuscitated by catalase. The omp gene was used as a positive control because it was strongly expressed as determined by detection of the product using SDS-PAGE ( Fig. 2A, the arrow) and protein identification (data not shown). mRNA expression of omp, oxyR, rpoS, dps, and the V3 region of 16S rRNA gene differed slightly between the two temperatures (Fig. 9). However, no significant differences in gene expression at the transcriptional level between bacterial cells at 25uC and 4uC in mASM were observed.

Discussion
The nonculturable or VBNC state of R. solanacearum has been reported in watercourses, liquid microcosms and soil, and can be induced by low temperature, starvation or copper stress [5,13,14,15,16]. However, the VBNC state is not well understood. Here, we used an artificial soil microcosm (ASM) system [29] to investigate the nonculturable state of R. solanacearum. Previously, sterilized environmental soil has been used to investigate the microbial ecology of soil [40,41] because it mimics natural soil in terms of its physico-chemical composition. Nonetheless, sterilized natural soil lacks reproducibility due to the variable characteristics of soils from different origins. Ellis [29] first established ASM by mixing commercial components of soils and showed that Pseudomonas strains stably maintained their cell density. In this study, ASM was used to study the nonculturable state of R. solanacearum strains in soils.
The viability of R. solanacearum strains SL341 and GMI1000 in ASM at 25uC was maintained for over 1 month. However, the culturable cell population of SL341 and GMI1000 varied due to the presence of humic acid, which is included in ASM as a nutrient (data not shown). The variability of culturable cell density in ASM may be due to use of humic acid of different origins because it is an organic compound [42]. Interestingly, both R. solanacearum strains GMI1000 and SL341 stably and reproducibly maintained their cell density in mASM (ASM without humic acid) under unstressed conditions. R. solanacearum strains can survive for extended periods of time in soil lacking nutrients [6]. The reproducible survival of the R. solanacearum strains suggests that R. solanacearum ecology can be investigated using mASM, which mimics the bacterial ecology of soil.
When mASM containing R. solanacearum SL341 and GMI1000 was incubated at 25uC or 4uC, the culturability of both strains was similar. While the proportion of culturable bacteria was maintained at 25uC, that at 4uC decreased over time (Fig. 1). It has been reported that while R. solanacearum population decreased during the winter season in soil or water microcosm, the culturable cell density increased with increasing temperature [13,14]. This phenomenon is due to the presence of VBNC R. solanacearum cells. Therefore, the loss of culturability of R. solanacearum at 4uC could result from bacteria entering the VBNC state. Although the nonculturable bacterial cells were unable to form colonies on regular culture media, maintaining total DNA and proteins in these cells suggested that they retain fundamental metabolism. However, the intact DNA and proteins in mASM incubated at 4uC cannot be taken as direct evidence of R. solanacearum cell viability.
The resuscitation of VBNC cells in mASM indicates their viability. Vibrio strains in the VBNC state induced by low temperature were recovered by removing stress factors, such as increasing the temperature [20], which indicated their viability. We also attempted to resuscitate VBNC cells by increasing the temperature. Our results revealed that a temperature increase (before complete loss of culturability) recovered bacterial culturability (Fig. 5). However, VBNC cells in mASM did not recover culturability by a simple temperature increase. This result suggested that the culturability of nonculturable cells in mASM could not be restored by increasing the temperature. The addition of vitamin B 1 , sodium pyruvate, nutrients (CPG, V8 juice) and AHLs has been used to recover bacterial culturability since these are known to resuscitate VBNC cells of other bacterial taxa  [21,23,24,43,44]. However, none of the additives resuscitated SL341 cells in mASM.
Recently, it was reported that the oxidative stress response in bacteria is associated with cold stress [45], and VBNC cells of Vibrio, Ralstonia and Escherichia induced by low temperature were successfully resuscitated by treatment with antioxidant agents [21,24,25,46]. Our results also showed that culturability of VBNC cells of R. solanacearum SL341 was recovered in the presence of catalase (Fig. 6), which indicates that oxidative stress induced internally or externally may be important for induction of the VBNC state of R. solanacearum. However, sodium pyruvate, which protects against H 2 O 2 , did not resuscitate nonculturable bacterial cells, despite the fact that supplementation with sodium pyruvate in mASM or TZC medium delayed induction of the VBNC state in both GMI1000 and SL341 strains (data not shown). Therefore, it is likely that sodium pyruvate has a minor effect on the resuscitation of VBNC cells. The resuscitation of VBNC SL341 and GMI1000 cells by catalase treatment was reproducible. To exclude the possibility that catalase itself serves as a nutrient to support the re-growth of culturable SL341 cells, autoclaved catalase was used as a negative control (Fig. 8). The resuscitation of VBNC SL341 cells by catalase (but not by autoclaved catalase) suggested hydrogen peroxide accumulation in cells exposed to low temperature stress and suppression of colony formation on regular culture medium. It remains unclear whether the resuscitation of VBNC cells corresponds to re-growth or resuscitation [47,48]. The number of bacterial cells resuscitated by distilled water containing catalase was equal to the initial inoculum. Failure of resuscitation in the presence of autoclaved catalase indicated that catalase itself does not support the regrowth of culturable cells. Since mASM contains no nutrient that would support the growth of culturable cells, resuscitation is the result of recovery of the culturability of VBNC R. solanacearum SL341 cells. The recovered SL341 cells showed similar phenotypes to the original strain, including virulence towards tomato plants. Since GMI1000 VBNC cells showed the similar resuscitation by catalase treatment, our mASM microcosm system could be generally used for VBNC state and its resuscitation of phylotype I strains.
H 2 O 2 can permeate through bacterial cell membranes, although the permeability is not unlimited [49]. H 2 O 2 produced in SL341 cells in response to low temperature stress might have accumulated to a critical concentration extra-or intracellularly, which suppressed colony formation on culture media. In fact, copper-treated R. solanacearum SL341 lost culturability and accumulated H 2 O 2 compared to cells in the culturable state [16]. Therefore, an external supply of catalase, which removes H 2 O 2 in mASM and cells, may reduce the intracellular H 2 O 2 concentration, thus facilitating colony formation on regular culture media. The resuscitation of cells exposed to low temperature stress by catalase has been reported for other bacterial taxa [20,21]. Timecourse experiments of resuscitation of R. solanacearum SL341 VBNC cells suggested the following three phases of SL341 cell culturability at 4uC in mASM: Phase I, pre-nonculturable stage; Phase II, nonculturable stage can be resuscitated by catalase; Phase III, nonculturable stage but cannot be resuscitated by catalase. The failed resuscitation of VBNC cells by catalase in phase III may be due to entry to a death stage. Otherwise, other stress factors induced by the low temperature (other than H 2 O 2 ) might also have affected the cells. It remains unclear whether phase III represents a bacterial death stage. SL341 cells could be injured by H 2 O 2 in this phase [39], and thus may not be resuscitated by its removal. Interestingly, the nonculturable bacterial cells in phase III were respiration-positive based on CTC staining (Fig. 4).
To assess gene expression in VBNC cells, the V3 region of 16S rRNA and omp genes were used as a positive control for expression in both culturable and nonculturable cells. Omp was selected because an Omp protein was produced constitutively in both culturable and VBNC cells (see the major protein at ca. 32 kDa in Fig. 2A). The rpoS, oxyR and dps genes were selected because they are involved in regulating the expression of genes associated with bacterial survival and the stress response [50,51,52]. The rpoS and oxyR genes are involved in the production of antioxidant enzymes [51,53], and the dps gene encodes a starvation-or oxidative-stressinduced DNA-binding protein in Escherichia coli [54]. Although there is no difference in the expression of those genes between culturable cells and VBNC cells, their constitutive expression of those genes in VBNC SL341 cells may suggest continued transcription of essential transcripts and stable maintenance of their transcript level (Fig. 9).
In conclusion, we used the mASM system to demonstrate that low temperature induced the VBNC state of R. solanacearum. Moreover, VBNC R. solanacearum SL341 cells maintains their respiration, gene expression and protein production, and could be resuscitated by catalase treatment. VBNC is considered a survival strategy for pathogenic bacteria, such as Ralstonia, in the soil. Therefore, elimination of VBNC R. solanacearum cells could be a novel approach to controlling or preventing plant infection by harmful microorganisms.