We found that a WRKY-like gene was significantly activated in tobacco (Nicotiana tabacum) under R. solanacearum infection (unpublished data) based on preliminary transcriptomic data (unpublished data). We thus designed primers according to the predicted CDS sequence for qRT-PCR analysis. The ubiquitin gene is among the most stable genes in tobacco (Rotenberg et al., 2006) and is commonly used as internal reference gene in QRT-PCR. To roughly confirm the stability of ubiquitin gene UBI3 under various stress conditions, the expressions of WRKY gene were analyzed using three reference genes EF1α (Rotenberg et al., 2006), ACT9 (Rotenberg et al., 2006) and UBI3, or one reference gene UBI3. The results indicate the transcript levels of WRKY gene were quite similar between two analyses suggesting UBI expressions were quite stable under different conditions (Supplementary Figure S1). Hence, UBI3 was used as qRT-PCR reference gene.

To verify the transcriptional profiles of this WRKY-like gene under R. solanacearum infection, 8-week-old tobacco plants were inoculated with 10 ml of bacterial cells at a concentration of 10⁸ colony-forming units (CFU) ml^-1 via wounding of the roots. At 1 day post inoculation (dpi), no wilt symptoms were observed, and very few bacterial titers were detected in tobacco roots. However, the bacterial titers significantly increased, and the wilt symptoms became more serious at 3 and 5 dpi (Figures 1A,B). The whole roots and stem base of inoculated plant were harvested at 0, 1, 3, and 5 dpi, respectively. Total RNA was extracted, and cDNA was further generated by reverse transcription. The qRT-PCR analysis showed that the expression of WRKY-like gene was induced at 1 dpi and increased over time, until 5 dpi (Figure 1C). These results suggest a consistent and increasing activation of WRKY-like gene expression during R. solanacearum infection.

A pair of primers were designed, and polymerase chain reaction was performed to isolate this WRKY-like gene. The results showed that a sequence of 576 base pairs encoding 191 amino acids was amplified. Nucleotide BLAST demonstrated that the deduced WRKY gene shows 100% similarity to NtWRKY50. Then, this protein was termed NtWRKY50. The encoded NtWRKY50 protein contains the WRKY domain WRKYGKK, which contains a substitution of G to K in the typical WRKYGQK sequence, and a zinc-finger motif (C-X_4-5-C-X_22-23-H-X₁-H) (Figure 2).

We next performed a phylogenetic analysis for NtWRKY50 using MEGA 5.1 and the neighbor-joining method. The results showed that WRKYs were generally divided into three categories: WRKY I, WRKY II, and WRKY III. The WRKY II category was further divided into five subgroups, including IIa, IIb, IIc, IId, and IIe. NtWRKY50, along with Arabidopsis thaliana AtWRKY8 and AtWRKY28, was classified into the IIc subgroup (Figure 3). This subfamily was purported to contain a large amount of variation in the WRKYGQK heptapeptide sequence, and this variation was consistent with our results (Figure 2) (Xiang et al., 2016).

To investigate the localization of NtWRKY50, the open reading frame sequence of NtWRKY50 was fused in frame to the C-terminus of a green fluorescent protein (GFP) gene controlled by the cauliflower mosaic virus CaMV35S promoter in a pEGAD vector. Using Agrobacterium-induced transient expression in the leaves of tobacco, we observed that GFP-fused NtWRKY50 localized in the nucleus of leaf epidermal cells (Figure 4).

To investigate the role of NtWRKY50 in plant disease defense, we generated NtWRKY50-overexpression transgenic tobacco plants, in which the NtWRKY50 expression is controlled via the cauliflower mosaic virus CaMV35S promoter. Among seven transgenic tobacco lines, the OE2 line (NtWRKY50-OE2) showed a higher expression of NtWRKY50 and was used for the pathogenicity analysis (Figure 5A). When 8-week-old tobacco plants were inoculated with 10 ml of R. solanacearum cells at a concentration of 10⁸ CFU ml^-1 by root irrigation. NtWRKY50-OE2 plants were more resistant than WT and presented lower disease incidence (Figure 5B). To assess the bacterial growth, roots and base stems were collected from plants inoculated via the root-wounding method. The results indicate the bacterial growth was influenced and was significantly lower in NtWRKY50-OE2 than in WT (Figure 5C). In addition, DAB (3, 3′-diaminobenzidine) and NBT (nitro blue tetrazolium) staining showed stronger ROS in the NtWRKY50-OE2 plants (Figure 6). Thus, these results suggest that NtWRKY50 positively regulates tobacco resistance to R. solanacearum.

To further determine the role of NtWRKY50 in regulating plant disease resistance, we used the antisense silencing technique to knock down the expression of NtWRKY50 in tobacco. The transcript levels of NtWRKY50 in the silenced lines and WT tobacco plants were detected using qRT-PCR. The results showed that the S2 line had the lowest transcription levels of NtWRKY50 (Supplementary Figure S2A). Then, S2 plants were inoculated with R. solanacearum cells as were the NtWRKY50-overexpression plants. The result showed that there was no difference in bacterial titers and disease symptoms between NtWRKY50-silenced lines and WT (Supplementary Figures S2B,C). These results indicate that silencing NtWRKY50 does not alter tobacco disease resistance.

Salicylic acid, JA, and ET are important defense signaling molecules. We next detected the effects of NtWRKY50 overexpression on the expression of SA-, JA-, and ET-responsive genes. Total RNA was extracted from both 8-week-old NtWRKY50-OE2 and WT plants before and after R. solanacearum inoculation, and cDNA was generated. SA- and JA/ET-responsive PR genes, such as PR1A/C, PR1B, PR2, PR3; ET biosynthesis and signaling marker genes, including ACC oxidase, ACS1, and EFE26; HR-related genes HSR201 and H1N1; and genes encoding ROS-scavenging enzymes, including catalase (CAT), superoxide dismutase (SOD), and glutathione S-transferase (GST), were analyzed using qRT-PCR. Before pathogen infection, 3–6-fold up-regulation was observed for the gene expression of PR1B, PR2, ACS1, EFE26 in NtWRKY50-OE2 line compared with WT line, while PR3 and H1N1 increased 20–30 fold (Figure 7A). Other genes did not show significant changes. However, the transcript levels of the majority of genes in NtWRKY50-OE2 were 50–800-fold more abundant than in WT after pathogen infection (Figure 7B). ROS-scavenging enzymes exhibited different regulatory patterns: CAT decreased, while GST was up-regulated 10 fold in the NtWRKY50-OE2 line compared to WT line. These findings indicate NtWRKY50 may induce hormone-mediated defense pathways, especially after pathogen infection.

To investigate whether NtWRKY50 is altered by various plant hormones and abiotic stresses, 8-week-old WT plants were treated with the plant hormones SA, JA and ET and with abiotic stresses, such as hydrogen peroxide (H₂O₂), heat, cold, NaCl, and wounding. Materials were harvested at indicated time points. The results showed that NtWRKY50 expression increased markedly to the highest value (10 and 7 fold) at 24 and 6 h after spraying with SA and JA, respectively (Figure 8A). When treated with ET, NtWRKY50 increased by approximately 5 fold at 12 h, which was its highest peak. Other abiotic stress conditions also led to the up-regulation of NtWRKY50 (Figure 8B). Cold, heat, and NaCl treatments resulted in 4–6-fold more expression of NtWRKY50 at 12, 6, and 6 h, respectively. Wounding and H₂O₂ treatments showed the most inconspicuous induction; the highest value was the double of that of the control.

WRKY genes probably respond to multiple plant pathogens. Therefore, NtWRKY50 expression in response to different types of pathogens was tested. Phytophthora parasitica (P. parasitica), Rhizoctonia solani (R. solani) and Potato virus Y (PVY), which belong to oomycetes, fungi and viruses, respectively, were applied to leaves of tobacco. NtWRKY50 transcripts were also activated under various pathogen infections, although the expression patterns were distinct (Figure 8C), indicating that NtWRKY50 gene expression responds to environmental changes.

Regarding the strong morphological phenotypes observed in overexpression plants, one possibility to explain the mode of action of NtWRKY50 is that this TF might be involved in endogenous hormone biosynthesis. To test this possibility, we determined whether the endogenous levels of hormones implicated in plant defense were affected in NtWRKY50 transgenic plants. Therefore, the levels of SA and JA, both of which are important in defense responses, were determined in NtWRKY50-OE2 and WT plants before and after R. solanacearum inoculation. As shown in Figure 9, a small amount of SA was detected in WT, whereas NtWRKY50-OE2 possessed a relatively high amount of SA before pathogen inoculation (Figure 9). However, after infection by R. solanacearum, SA levels increased dramatically in WT but stayed unchanged in NtWRKY50-OE2. The amount of JA was similar between WT and NtWRKY50-OE2 before pathogen infection and JA increased markedly to a high level in WT but barely changed in NtWRKY50-OE2 (Figure 9). These findings suggest NtWRKY50 influences SA and JA amounts in the presence and absence of pathogen invasion.

Plants recognize pathogens through a series of signal transduction processes, leading to the final induction of plant defense genes. In these processes, WRKY TFs play an important role in activating or inhibiting the expression of defense genes through interaction with other related proteins. At the same time, the WRKY TF itself is also regulated by other factors. Changes in expression levels of specific WRKY TFs greatly alter stress resistance. In this study, we identified and characterized a tobacco WRKY member belonging to group IIc. These findings suggest that constitutive overexpression of NtWRKY50 promotes tobacco resistance to the bacterial pathogen R. solanacearum. Loss-of-function data indicate a redundancy feature of NtWRKY50 with other TFs. Notably, NtWRKY50 overexpression led to a marked increase in SA production but inhibited pathogen-induced JA accumulation; on the other hand, transcripts of NtWRKY50 were markedly induced by SA and JA, indicating an important role in the SA and JA regulatory networks. Promotion of NtWRKY50 expression by various abiotic stresses and pathogens might be a subsequent result from the effects of this TF on hormone production and suggests this protein is involved in basal responses to environmental stimuli.

Ralstonia solanacearum (R. solanasearum) isolate CQPS-1 (race 1, biovar 3, phylotype I, sequevar 17) was cultivated at 30°C. Plants of tobacco (Nicotiana tabacum) cultivar Yunyan87 [wild type (WT)] were grown in a growth chamber at 28°C and 70% relative humidity under a 14-h photoperiod.

The full-length coding region of NtWRKY50 was inserted into the XmaI and BamHI restriction sites of pEGAD, after which the resultant plasmid or empty control plasmid were transformed into Agrobacterium tumefaciens strain EHA105.

The coding sequence of NtWRKY50 was predicted from transcriptome sequencing results. The primers used in this study are listed in Supplementary Table S1. The coding sequence was inserted into the BglII and BstEII restriction sites of the plant expression vector pVCT2024 or Eco31I site of pBWA(V)HS to generate overexpression lines and RNAi lines, respectively, under the control of the CaMV35S promoter. Tobacco transformation was accomplished according to the procedures of Horsch et al. (1985). Transgenic plants were selected using kanamycin (100 mg L^-1) and confirmed using PCR. Positive plants were propagated asexually in Murashige Skoog (MS) medium and then transferred to soil; plants at the 8-leaf stage were ready for bacterial inoculation. WT Yunyan87 tobacco plants were also cultivated in MS medium, propagated and transplanted at the same time as the transgenic tobacco plants and served as controls for R. solanacearum inoculation and other analyses.

Ralstonia solanacearum was grown in B liquid medium (Vasse et al., 2000) for 16 h at 30°C during shaking. Bacteria were then collected by centrifugation and resuspended in distilled water.

For the disease incidence analysis, 10 transgenic and WT plants were root inoculated with 10 ml (1 × 10⁸ CFU ml^-1) of R. solanacearum cells. The experiment was repeated three times. Disease grades were as follows: 0, no wilting; 1, 1–25% wilted; 2, 26–50% wilted; 3, 51–75% wilted; and 4, 76–100% wilted or dead. The experiment was repeated three times.

For the bacterial growth assays, transgenic and WT plants were wounded on the roots and inoculated with 10 ml (1 × 10⁸ CFU ml^-1) of R. solanacearum cells. Roots (three plants were pooled into one sample) were sampled at each time point of bacterial inoculation, weighed and then rinsed with distilled water five times before being ground using 10 ml of distilled water. Dilutions of bacteria were spotted on B agar medium, and colonies were counted after 2 days, which allowed the bacterial density (CFU g^-1) in the roots to be calculated. The experiment was repeated three times. The Student’s t-test was then used to determine significant differences between the transgenic plants and WT plants.

The total RNA of tobacco plants was extracted from roots and base stem using TRIzol^® Reagent (Invitrogen, Carlsbad, CA, United States). The RNA concentration and integrity were then analyzed using an Agilent Bioanalyzer 2100 system (Agilent Technologies, Santa Clara, CA, United States).

qRT-PCR was performed using a C1000 Touch^TM Thermal Cycler (BIO-RAD, United States) and a CFX96^TM real-time system, with SsoFAST^TM Eva Green Supermix. Transcripts of genes were standardized using UBI3 (Schmidt and Delaney, 2010). Primers used in this assay were designed using Primer 5.0 software (Premier Biosoft International) and are listed in Supplementary Table S1. PCR amplification was performed according to the following conditions: 95°C for 3 min; 40 cycles of 95°C for 10 s, 55°C for 15 s, 72°C for 15 s; and a melt cycle from 65 to 95°C.

Before and 1 day after R. solanacearum inoculation, the leaves of WT and NtWRKY50-OE2 plants were stained with 3, 3′-diaminobenzidine (DAB, 1 mg ml^-1) or nitro blue tetrazolium (NBT, 0.1 mg ml^-1) for 24 and 6 h, respectively, at 25°C in the dark. The DAB-stained leaves were cleared by boiling in [lactic:glycerol:absolute ethanol (1:1:3, V:V:V)] and then destained overnight in absolute ethanol (Korasick et al., 2010). The NBT-stained leaves were destained overnight in absolute ethanol directly. After destaining, the leaves were soaked and preserved in fresh ethanol at room temperature and imaged.

Eight-week-old WT plants were sprayed with 2 mM SA, 0.1 mM JA, 7 mM ET, 10 mM hydrogen peroxidase (H₂O₂) or 200 mM NaCl. Plants were incubated at 38 or 15°C for heat and cold treatments, respectively. Wounding consisted of eight parallel cuts caused by a blade.

For biotic stress treatments, 8-week-old WT plants were mechanically inoculated with potato virus Y (PVY) or were wounded using a scalpel, after which the leaves were inoculated with Phytophthora parasitica or Rhizoctonia solani. Samples of PVY-infected plants were collected at 0, 6, 10, 11, and 12 dpi, whereas the plants infected with the other two pathogens were harvested at 0, 1, 2, 3, and 4 dpi. Three plants were pooled into one sample for qRT-PCR analysis, and the experiment was repeated three times.

Free SA and JA were quantitatively analyzed using high-performance liquid chromatography–mass spectrometry according to the protocol of Pan et al. (2010). SA (transition of 137/92.9) and JA (transition of 209.2/58.9) were detected in 50 mg of fresh -tobacco crude leaf extracts.