2.1 The L. regale subgroup-IIa WRKY3 gene was found to encode a nuclear protein
In our previous study, we identified a series of WRKY genes from a Fusarium wilt-resistant wild lily (L. regale) by transcriptome sequencing. According to the gene expression pattern data, LrWRKY3 was induced by F. oxysporum and responded to signaling molecule treatments [12]. In the present study, the function of LrWRKY3 was therefore further characterized to understand the transcriptional regulation of WRKY in L. regale in response to Fusarium wilt. The full-length cDNA of LrWRKY3, which was found to be 722 bp in length with a 537-bp open reading frame (ORF), was predicted to encode a subgroup-IIa WRKY protein with 178 amino acid residues. The calculated molecular mass of the deduced LrWRKY3 was 20.3 kDa. According to our sequence analysis, LrWRKY3 contained a highly conserved ‘WRKYGQK’ sequence and the zinc-finger motif CX4-5CX22-23HXH (C2H2), which suggests membership in the IIa WRKY subgroup (Fig. 1a). The deduced amino acid sequence of LrWRKY3 was highly similar to that of some WRKY proteins, including Phoenix dactylifera WRKY75 (GenBank no. XP_008809548.2), Cocos nucifera WRKY75 (KAG1370131.1), Elaeis guineensis WRKY75 (XP_010913024.1), and Durio zibethinus WRKY75 (XP_022740913.1) (Fig. 1b).
The result of PSORT prediction suggested that the protein encoded by LrWRKY3 is localized in the plant cell nucleus. To confirm the subcellular location of LrWRKY3, a fusion expression vector of LrWRKY3 and the green fluorescent protein (GFP) gene was constructed and expressed in onion epidermal cells through Agrobacterium tumefaciens-mediated transformation. The green fluorescence signal of the LrWRKY3-GFP fusion protein was exclusively distributed in the nuclei of onion epidermal cells and colocalized with the nuclear localization marker PI (Fig. 1c). In contrast, green fluorescence was detected throughout the entire onion epidermal cell in the control. This result indicates that LrWRKY3 is localized in the plant cell nucleus.
2.2 Overexpression of LrWRKY3 in tobacco increased resistance to F. oxysporum and enhanced transcription levels of some JA/SA signaling pathway, PR, and SOD genes
To further understand the biological function of LrWRKY3, the plant overexpression vector pCAMBIA2300s-LrWRKY3 was constructed and transformed into WT tobacco leaf disks. A total of 50 T0 transgenic tobacco plants were obtained according to PCR analysis with LrWRKW3-specific primers. qRT-PCR was used to analyze the expression levels of LrWRKY3 in leaves of 11 T2 generation transgenic lines (W3-1/2/3/6/19/20/27/34/35/36/44), with WT serving as a negative control. According to the results, LrWRKY3 transcripts accumulated and were differentially expressed in the 11 transgenic tobacco lines (Fig. 2a). LrWRKY3 transcription was highest in line W3-6, with a relative expression value of 10.25. LrWRKY3 relative expression levels in W3-20, W3-35 and W3-44 were also high: 7.53, 8.40, and 7.25, respectively. These data indicate that LrWRKY3 was stably expressed in T2 generation transgenic tobacco.
Four transgenic lines (W3-6/20/35/44) with high LrWRKY3 expression and WT plants were inoculated with a spore suspension of F. oxysporum to evaluate their resistance to the pathogen. The roots of WT tobacco turned black and rotted, and whole leaves had wilted 7 days after inoculation (Fig. 2b). In contrast, the roots of T2 generation transgenic tobacco lines showed no evident disease symptoms, and the leaves were healthy, bright green, and non-wilted. The results of leaf inoculation experiments were consistent with these observations. The leaves of the transgenic lines exhibited only slight deterioration and local yellowing near the inoculation wound, whereas leaves of the WT were obviously more yellowed and decayed, with a larger damaged area, 7 days after inoculation (Fig. 2c–d). These results indicate that the overexpression of the LrWRKY3 gene in tobacco improves disease resistance to F. oxysporum.
Transcription levels of some JA/SA signaling and defense-related genes in transgenic tobacco lines and WT tobacco were analyzed by qRT-PCR. As shown in Fig. 3, transcription levels of JA biosynthetic pathway-related genes, including NtLOX, NtAOC, NtOPR, NtAOS, NtJMT, NtKAT, and NtPACX, were obviously elevated in LrWRKY3 transgenic lines compared with the WT. The expression level of NtJMT was highest in line W3-44: 16.89-fold higher than in WT tobacco. Compared with their expressions in the WT, NtLOX and NtAOS were up-regulated in W3-44 by 6.49- and 2.98-fold, respectively. Relative expression levels of the SA signaling pathway-related genes NtPR1 and NtNPR1 were increased in the transgenic tobacco lines. The expression of NtPR1 in line W3-44 was 24-fold that of the WT, and the expression of NtNPR1 in line W3-35 was 7-fold that of the WT. In W3-44, the expression levels of pathogenesis-related protein (PR) genes NtGlu2 and NtCHI were respectively 24- and 8.5-fold higher than that of the WT. In addition, the expression level of PR gene Ntosmotin in line W3-20 was 3.4-fold higher than in the WT. Furthermore, antioxidant stress-related superoxide dismutase (SOD) genes, including NtSOD, NtCu-ZnSOD, and MnSOD, were all up-regulated in transgenic tobacco compared with the WT. In conclusion, the overexpression of the LrWRKY3 gene in tobacco up-regulated the expressions of JA/SA signaling pathway-related genes, PR and SOD genes.
2.3 Transient expression of the LrWRKY3 RNAi vector in L. regale scales increased susceptibility to F. oxysporum
To further analyze the function of LrWRKY3, a LrWRKY3 RNAi construct was transiently expressed in L. regale scales via Agrobacterium tumefaciens mediation, and the scales were then inoculated with F. oxysporum spore suspension. Obvious blackening and decay appeared on the infected scales expressing the LrWRKY3 RNAi construct, but disease symptoms were much less evident on L. regale scales transformed with Agrobacterium tumefaciens containing an empty RNAi vector (Fig. 4a). Calculation of the area of damage on F. oxysporum-infected L. regale scales confirmed that L. regale expressing the LrWRKY3 RNAi fragment was less resistant to F. oxysporum infection than the control (RNAi empty vector) (Fig. 4b). In addition, qRT-PCR analysis revealed that the expression level of LrWRKY3 in RNAi fragment-expressing L. regale was evidently lower than that in control scales, whether inoculated with F. oxysporum or not (Fig. 4c). These data clearly indicate that the decreased expression of LrWRKY3 in L. regale scales enhanced sensitivity to F. oxysporum.
2.4 The L. regale defensin gene Def1 was revealed to be a F. oxysporum resistance gene
To explore the regulation of PR gene expression by LrWRKY3, a L. regale defensin gene, LrDef1, was cloned on the basis of transcriptome sequencing data (unpublished). The full-length cDNA of LrDef1 was 501 bp, with a 225-bp coding region, and was predicted to encode a protein containing 74 amino acid residues. The predicted molecular weight of the deduced protein LrDef1 was approximately 8.06 kDa. The LrDef1 promoter sequence (850 bp) was obtained by genome walking. Using the PlantCARE prediction program, we identified many cis-acting elements in the LrDef1 promoter, including the TGACG motif (MeJA response element), W box (ET, SA, and MeJA response element), and ABRE (ABA response, hypersaline, and dark induction element), and a series of cis-elements involved in response to plant hormone signals and abiotic and biotic stresses (Table S3).
In the qRT-PCR analysis, LrDef1 transcripts were detected in L. regale roots, stems, leaves, flowers, and scales (Fig. 5a). In particular, LrDef1 was strongly expressed in scales. After F. oxysporum inoculation, the expression of LrDef1 in L. regale roots was rapidly induced, with a peak at 24 h (Fig. 5b) when the expression level was approximately 4.1-fold of the control. According to these results, LrDef1 is a F. oxysporum infection-induced gene and is dominantly expressed in L. regale scales. An N-terminal signal peptide was detected in the deduced protein LrDef1, which was predicted to localize in the plant cell wall, thus indicating that LrDef1 may be a secretory protein. Moreover, the subcellular localization of LrDef1 also examined by fusion expression with GFP in onion epidermal cells. The green fluorescence signal of the fusion gene of LrDef1 and GFP was distributed in the cell wall of onion epidermal cells (Fig. 5c). This result demonstrates that LrDef1 is located in the plant cell wall as an extracellular protein.
The recombinant plasmid pET-32(a)-LrDef1-His containing a His-tag was transformed into Escherichia coli strain BL21 (DE3) for heterologous expression. As revealed by SDS-PAGE, the molecular mass of the induced His-LrDef1 fusion protein was consistent with its predicted size, 25 kDa (Fig. 6a). The recombinant protein was purified by Ni-NTA-sepharose (Sangon Biotech, China) column affinity chromatography and imidazole elution buffer (Fig. 6b) and then used in antifungal experiments. As shown in Fig. 6c–e, the LrDef1 recombinant protein had different inhibitory effects on the mycelial growth of three fungi: F. oxysporum, F. solani, and Alternaria alternata. The recombinant protein had the strongest inhibitory effect on F. oxysporum, followed by Alternaria alternata and F. solani. Moreover, antifungal activity increased along with the mass of the LrDef1 recombinant protein (Fig. 6f).
PCR analysis with LrDef1-specific primers identified 35 T0 LrDef1 transgenic tobacco plants. In addition, qRT-PCR was used to analyze the expression levels of LrDef1 in 12 T2 generation transgenic lines (F1/2/4/5/8/9/11/16/17/19/20/21). The qRT-PCR analysis revealed that LrDef1 was expressed in all transgenic lines (Fig. 7a). We selected four of the T2 generation LrDef1 transgenic tobacco lines (F8/9/11/16) to study their resistance to F. oxysporum. Seven days after inoculation with F. oxysporum spore suspension (2 × 106 spores/mL), WT tobacco roots were black and rotten, and some leaves had started to shrink. In contrast, roots of the transgenic tobacco lines were only slightly darkened, and the leaves were still fully extended (Fig. 7b). Overexpression of LrDef1 in tobacco thus enhanced resistance to F. oxysporum infection.
2.5 The LrWRKY3 recombinant protein specifically bound to the LrDef1 promoter fragment with a W box
The recombinant vector pET-32(a)-LrWRKY3 was constructed and transformed into E. coli strain BL21 (DE3) for heterologous expression to obtain a LrWRKY3 recombinant protein. EMSA was used to analyze the binding between the LrWRKY3 recombinant protein and probes to reveal whether the specific binding site of LrWRKY3 is W box. As shown in Fig. 8a, lanes 1 and 2 on the EMSA gel both contained biotin-labeled probes designed from the LeDef1 promoter fragment with the W box. The band in lane 1, corresponding to biotin-labeled probe with no LrWRKY3 recombinant protein added, was not retarded on the gel. The presence of retarded bands in lanes 2 and 3 after the addition of LrWRKY3 indicates that LrWRKY3 was able to bind to the probes containing the W box, which slowed their migration rates on the gel. The band in lane 3 was less delayed than the band in lane 2 because of competition due to the large number of unlabeled probes, which were 50 times more abundant than the biotin-labeled ones. Moreover, the LrWRKY3 protein was unable to bind to the mutant probe in lane 4 and thus no mobility shift was observed. These data fully illustrate that LrWRKY3 specifically binds to the W box.
2.6 LrWRKY3 transcriptionally activated LrDef1 in yeast cells
To determine whether LrWRKY3 has transcriptional activation activity, we integrated the ORF of LrWRKY3 into the prey vector pGADT7 AD in a Y1H system followed by co-transformation with the recombinant bait vector pAbAi-pLrDef1 in Y1HGold yeast cells, with the pAbAi-p53 plasmid used as the positive control and the pAbAi empty vector as the negative control. Yeast cells co-transformed with pGADT7 AD-LrWRKY3 and pAbAi-pLrDef1 vectors were able to grow normally on auxotrophic SD/-Leu/AbA solid medium (Fig. 8b). In contrast, yeast cells co-transformed with pGADT7 AD-LrWRKY3 and bait empty vectors could not grow on the solid medium. These results indicate that pGADT7 AD-LrWRKY3 could be integrated into the yeast genome containing bait plasmid to produce a fusion protein, Gal4-LrWRKY3, capable of recognizing and activating the W box of pLrDef1 in yeast cells. The Gal4-LrWRKY3 protein activated the expression of the UR1-C gene in the recombinant bait plasmid pAbAi-pLrDef1, which enabled yeast cells to grow normally on the SD/-Leu/AbA medium. This outcome suggests that the LrWRKY3 protein can specifically bind to the LrDef1 promoter in yeast and has trans-activation activity.
2.7 LrWRKY3 positively regulated the expression of the F. oxysporum-resistance gene LrDef1 in tobacco
Next, LrWRKY3 and the LrDef1 promoter were co-expressed in tobacco to explore the effect of LrWRKY3 on the transcriptional activity of the LrDef1 promoter. According to a GUS activity analysis, the β-glucuronidase (GUS) gene had the highest activity in pBI121-GUS transgenic tobacco, approximately 88 pM 4-MU min−1 mg−1 (Fig. 8c). At the same time, GUS activity in LrWRKY3/pLrDef1 transgenic tobacco was significantly higher than that in pLrDef1 transgenic tobacco. The average GUS activity of the four pLrDef1 transgenic tobacco lines (PD-2/3/7/10) was approximately 24, 26, 33, and 31 pM 4-MU min−1 mg−1, respectively. The average GUS activity of the four LrWRKY3/pLrDef1 co-expressing transgenic tobacco lines (Co-1/7/9/12), however, was higher than 40 pM 4-MU min−1 mg−1, namely, approximately 52, 48, 46, and 42 pM 4-MU min−1 mg−1, respectively. Noteworthily, the average activity of Co-1 was approximately 1.7-fold higher than that of PD-7. These results indicate that the specific bindings of LrWRKY3 with the W box in the LrDef1 promoter activated the expression of the downstream reporter gene GUS driven by the LrDef1 promoter, thereby enhancing GUS activity.