Ralstonia solanacearum isolates were collected by Dr. Young Kee Lee (T-numbered strains) and Dr. Seungdon Lee (SL-numbered strains) from 1998 to 2003 in Korea. Among the twenty-five locations surveyed, potato bacterial wilt was observed in twelve cities in six Korean provinces (Fig. 1). We analyzed ninety-three isolates, which are listed in Table 1. All isolates were identified to R. solanacearum by 16S rRNA sequence analysis. These bacteria were cultured on tetrazolium chloride (TZC) agar medium (peptone 10 g, glucose 2.5 g, casamino acid 1 g, agar 18 g, TZC 50 mg in 1 L distilled water) at 28°C for 48 h and frozen in 40% glycerol stock at −70°C (Kelman, 1954).

To prepare genomic DNA, bacterial cells grown on TZC agar medium were subcultured in LB broth (peptone 10 g, yeast extract 5 g, sodium chloride 5 g in 1 l distilled water) in a 28°C shaking incubator for 16 h. Genomic DNA of all isolates was extracted using the Wizard Genomic DNA Purification Kit (Promega, Madison, WI, USA) according to the manufacturer’s instructions.

The ninety-three isolates were classified into phylotypes as previously described (Prior and Fegan, 2005; Sagar et al., 2014). We determined phylotype using the method of Sagar (2014) based on phylotype-specific multiplex PCR (Pmx-PCR) using the following primers: Nmult:21:1F, 5′-CGTTGATGAGGCGCGCAATTT-3′ (specific for phylotype I, amplicon size is 144 bp when paired with Nmult22:RR); Nmult:21:2F, 5′-AAGTTATGGACGGTGGAAGTC-3′ (phylotype II, 372 bp); Nmult:23:AF, 5′-ATTACGAGAGCAATCGAAAGATT-3′ (phylotype III, 91 bp); Nmult:22:InF, 5′-ATTGCCAAGACGAGAGAAGTA-3′ (phylotype IV, 213 bp), and Nmult:22:RR, 5′-TCGCTTGACCCTATAACGAGTA-3′ (reverse primer for all phylotypes). PCR reactions were carried out in a total volume of 20 μl with Profi-Premix (Bioneer) with primer sets and genomic DNA in an automated thermocycler (model PTC-200, MJ Research Inc., Waltham, MA, USA) as follows: initial denaturation at 96°C for 5 min, followed by 30 cycles of denaturation at 95°C for 15 sec, annealing at 59°C for 15 sec, and extension at 72°C for 30 sec, with a final extension at 72°C for 10 min. Seven microliters of each PCR product was examined by electrophoresis through 1% agarose gel, stained with ethidium bromide, and visualized on a UV-trans-illuminator.

The ability of each isolate to oxidize three disaccharides (maltose, lactose, and cellobiose) and three hexose alcohols (mannitol, sorbitol, and dulcitol) was evaluated by inoculating the isolates on biovar plates following a modified Hayward method (Hayward, 1964). Along with the isolates, 1 ml of the basal medium (NH₄H₂PO₄ 1 g, KCl 0.2 g, MgSO₄·7H₂O 0.2 g, peptone 1 g, bromothymol blue 8 mg, agar 1.5 g in distilled water 1 l, pH 7.1) containing 1% sterilized carbon sources (maltose, lactose, cellobiose, mannitol, sorbitol, and dulcitol) was dispensed into the wells of 24-well plates (SPL Life Sciences, Seoul, Korea). All isolates were inoculated into individual wells with 5 μl of the 10⁸ cfu/ml bacterial suspensions, with two replicates per isolate. The plates were incubated at 28°C for 14 days. The plates were observed every day and the color change of the medium was recorded two weeks after inoculation. The test was repeated twice.

We carried out PCR to validate the presence of rsa1, which has been reported to be a gene for avirulence of R. solanacearum on pepper hosts (Jeong et al., 2011). To eliminate errors, we prepared two rsa1 gene primer pairs. One pair produced a 720-bp fragment containing the full rsa1 gene including the promotor region (720-rsa1F; 5′-GCCGCTCGCCGCAATGCTGCC-3′, 720-rsa1R; 5′-TGGGCTGGGTGGGACTTAACC-3′), and the other produced a 315-bp fragment containing a partial rsa1 open reading frame (ORF) region (315-rsa1F; 5′-ATCACCAAGATTACCGGAAAG-3′, 315-rsa1R 5′-TGGGCTGGGTGGGACTTAACC-3′). The reactions were carried out as described previously for phylotype determination with a primer set of 720-rsa1F/720-rsa1R at an annealing temperature of 70°C, and with a primer set of 315-rsa1F/315-rsa1R at an annealing temperature of 59°C.

The 16S rRNA genes of the 93 isolates were amplified by PCR in 25 μl reaction volumes containing 1.25 U of Pfu Turbo DNA Polymerase (Stratagene), 2.5 μl of 10× Pfu polymerase buffer, 0.25 mM of each dNTP, 1 μl of 10 pmoles of each primer (9F, 5′-GAGTTTGATCCTGGCTCAG-3′; 1512R, 5′-ACGGCTACCTTGTTACGACTT-3′), and 50 ng of genomic DNA. The reaction was performed in an automated thermocycler (model PTC-200, MJ Research Inc.) with initial denaturation at 95°C for 5 min, followed by 30 cycles of denaturation at 95°C for 45 sec, annealing at 55°C for 45 sec, and extension at 72°C for 1 min, with a final extension at 72°C for 10 min. The 750-bp partial endoglucanase gene was amplified by PCR in the same reaction as the 16S rRNA gene with primer pairs of Endo-F (5′-ATGCATGCCGCTGGTCGCCGC-3′) and Endo-R (5′-GCGTTGCCCGGCACGAACACC-3′) (Poussier et al., 2000). After the PCR amplicons of 16S rRNA and endoglucanase genes were confirmed by electrophoresis, the sequences of the two genes were determined using an ABI BigDye Terminator v3.1 Cycle Sequencing Kit and an ABI3730XL automated DNA sequencer (Applied Biosystems Inc., Foster City, CA, USA) according to the manufacturer’s instructions. The 16S rRNA and endoglucanase gene sequences were confirmed and edited with BioEdit 7.2.5 software and trimmed with EditSeq software (DNASTAR Lasergene8, DNASTAR Inc., Madison, WI, USA). The analyzed 16S rRNA and partial endoglucanase gene sequences were deposited in the GenBank database, and the accession numbers are listed in Table 1. Reference sequences for the 16S rRNA gene and endoglucanase gene were obtained from the NCBI website with the following GenBank accession numbers: AL646052.1 (GMI1000), FP885891.2 (PSI07), FP885896.1 (CMR15), FP885897.1 (CFBP2957), and CP002820.1 (Po82). The ClustalW method was used for sequence alignment and the construction of phylogenetic trees using 1,000 bootstrap replicates in the MegAlign program (DNASTAR Lasergene8, DNASTAR Inc.).

For the pathogenicity test, tomato (Lycopersicon esculentum cv. Seokwang) (Lee and Kang, 2013), eggplant (Solanum melongena cv. Heukmajang) (Lee, 1999), and pepper (Capsicum annuum cv. Nokkwang) (Jeong et al., 2007) were grown in a greenhouse at 25-35°C under natural light conditions. For the positive control, potatoes (Solanum tuberosum cv. Sumi) (the original cultivar name: Superior) (Park et al., 2016) were grown in a greenhouse at 20-25°C. Two-week-old seedlings of eggplant and pepper, and 10-day-old seedlings of tomato were transplanted into plastic pots 7 cm in diameter containing commercial soil (Baroker, Seoul Agriculture Materials Co., Seoul, Korea) and grown in a greenhouse for 2-3 weeks. To prepare the inoculum, all isolates were grown on TZC plates for 48 h at 28°C. Bacterial cells were suspended in sterile distilled water and the concentration was adjusted to OD₆₀₀ 0.1. After wounding the root of each plant by stabbing with a 3-cm-wide scoop, 50 ml of bacterial suspension of the 93 isolates was poured into each pot. Three plants from each crop were inoculated with each isolate, and pathogenicity assays were repeated two or three times in a greenhouse under natural light conditions. Symptom development was observed every 3 days and recorded at 28 days post-inoculation using the following scale: −, no symptoms; +, one to three leaves wilted; ++, four to six leaves wilted; and +++, seven or more leaves or whole plant wilted.

The tomato plants were photographed at 12 days post-inoculation, and the eggplants and pepper plants were photographed at 19 days post-inoculation.

The Korean potato bacterial wilt isolates were analyzed for biovar, phylotype, and presence of the rsa1 gene. Of the 93 isolates analyzed, twenty-two were of biovar 2 (about 24%), eight were of biovar 3 (less than 8%), and sixty-three were of biovar 4 (68%) (Table 1). The isolates of biovar 3 and biovar 4 were classified as phylotype I, and biovar 2 was classified as phylotype IV.

It has been reported that the Rsa1 protein, which is an aspartic protease secreted through the type II secretion systems (T2SSs), is responsible for the loss of bacterial virulence to pepper (Jeong et al., 2011). To validate the presence of this gene, we carried out rsa1 detection PCR. For greater certainty, we used two sets of rsa1 gene primers. One pair was for the full rsa1 gene including the promotor (720-rsa1F/720-rsa1R), while the other was for a partial rsa1 ORF (315-rsa1F/315-rsa1R). From the ninety-three isolates of potato bacterial wilt, all phylotype IV-biovar 2 isolates produced a full 720-bp rsa1 gene fragment at an annealing temperature of 70°C. On the other hand, biovar 3 and biovar 4 isolates of phylotype I did not produce this fragment under these annealing conditions. Partial rsa1 gene PCR was carried out with a 315-rsa1F/315-rsa1R primer set at an annealing temperature of 59°C, and the results were the same as for the PCR of the full rsa1 gene, which was detected in all isolates of phylotype IV and not detected in any isolates of phylotype I.

Fig. 1 shows the geographical distribution of the potato bacterial wilt isolates in Korea from 1998 to 2003. Ralstonia solanacearum was first isolated in southern provinces (Gyeongsangnam-do, Jeollanam-do, and Jeju-do, Korea) with biovar 2 and biovar 4 isolates in 1998. The biovar 3 bacterium, which infects potato, was first identified in Gyeongsangnam-do in 1999, and was subsequently found in Jeollanam-do, Jeollabuk-do, and Jeju-do. Biovar 4 bacteria that infect potatoes have been isolated in all locations where potato bacterial wilt has been observed. R. solanacearum of biovar 2 has been isolated in seven locations, which are the same as the regions where biovar 3 bacteria have been isolated.

To evaluate the genetic relationship of Korean potato isolates, we sequenced the 16S rRNA gene (about 1,453 nucleotides) and partial endoglucanase gene (760 and 766 nucleotides) of 93 isolates (Table 1) and compared the sequences using the ClustalW program (Fig. 2).

The 16S rRNA sequences of biovar 2 isolates were identical with each other, and the sequences of biovar 3 and biovar 4 isolates were also identical. To downsize the phylogenetic tree, SL2729 was chosen to represent the biovar 4 strains (Fig. 2-A). In the 16S rRNA phylogenetic tree, Korean R. solanacearum isolates were separated into two groups following GMI1000 (phylotype I) and PSI07 (phylotype IV). Korean biovar 2 isolates, which were identified as phylotype IV, were clustered with PSI07 of a representative phylotype IV strain. The isolates of biovar 3 and 4, which were identified as phylotype I, were not only clustered with GMI1000, but also identical with GMI1000 in the sequenced 1,453 nucleotides.

Since the partial endoglucanase gene (egl) sequences were identical among biovar 4 isolates, we analyzed the egl sequence of SL2729 as the representative strain of biovar 4 (Fig. 2B). While the sequences of biovar 4 isolates were identical, the egl sequences of biovar 3 isolates showed subtle differences in nucleotides. In a phylogenetic tree based on the egl gene, biovar 3 and biovar 4 isolates were grouped to phylotype I without GMI1000, which was placed as the outgroup to Korean phylotype I.

To determine the host range of the bacteria that cause potato bacterial wilt, we assessed the pathogenicity of isolates on several solanaceous plants-potato, tomato, eggplant, and pepper. The patterns of pathogenicity were divided into four types: first, pathogenic only to potato, but nonpathogenic to tomato, eggplant, and pepper (P); second, pathogenic to potato and tomato, but nonpathogenic to eggplant and pepper (PT); third, pathogenic to potato, tomato, and eggplant, but nonpathogenic to pepper (PTE); and finally, pathogenic on all four tested crops (potato, tomato, eggplant, and pepper, PTEP) (Table 2).

The isolates of phylotype IV-biovar 2 showed various patterns of host pathogenicity, including P, PT, and PTE. Of the twenty-two biovar 2 isolates, eleven infected only potato (P), four infected only potato and tomato (PT), and seven infected potato, tomato, and eggplant (PTE). These results showed that none of the biovar 2 isolates could infect pepper. The isolates of phylotype I (including biovar 3 and 4) could be divided into two groups: pathogenic to potato, tomato, and eggplant, but nonpathogenic on pepper (PTE); and pathogenic to all test plants (PTEP). For phylotype I, one biovar 3 isolate and four biovar 4 isolates were classified as PTE (7% of phylotype I), and seven biovar 3 and fifty-nine biovar 4 isolates were classified as PTEP (93% of phylotype I). Fig. 3 shows the host range of representative pathotype strains.