Nicotiana benthamiana plants were cultivated in a phytotron at 24°C with a 12-h day/night cycle. Arabidopsis seeds were sown in a suitable controlled plant growth condition at 22°C with a 12-h day/night cycle at 75% relative humidity (Niu et al., 2011).

Pseudomonas syringae pv. tomato (Pst) DC3000 was cultured in liquid Kings’ B medium containing 50 mg of rifampicin per liter at 28°C overnight as previously described (Niu et al., 2011). Culture cells were harvested and the final concentration of cell suspensions was adjusted to 5 × 10⁸ CFU/mL using 10 mM MgSO₄ containing 0.01% (vol/vol) surfactant Silwet L-77 (Sigma, St. Louis, MO, United States).

Valsa mali strain 03-8 was grown on potato dextrose agar plates and incubated at 25°C in the dark for 3 days before inoculation onto apple leaves.

The sequence for BAR11 without the signal peptide was inserted into the pET28a vector (Novagen, United States) upstream and downstream of two 6×His tags. The final plasmid was then introduced into Escherichia coli BL21 (DE3) cells to express the BAR11-His recombinant protein. The E. coli BL21 cells were subsequently grown in LB-medium containing 50 mg of kanamycin per liter at 37°C to an optical density at 600 nm (OD₆₀₀) of 0.5–0.6, and then induced with 0.1 mM isopropyl b-D-1-thiogalactopyranoside (IPTG) (Sigma, United States) for 18–20 h at 16°C, from which the protein BAR11-His was extracted as follows. The induced culture cells were harvested by centrifugation (5,000 ×g, 20 min). The insoluble fraction was resuspended in phosphate-buffered saline (pH7.4) and disrupted using an ultrasonic disruptor (Scientz, China). After centrifugation (18,000 ×g, 4°C, 30 min), the supernatant containing recombinant protein was gathered into a new pellet and filtered through a 0.22-μm filter (Millipore, China). The filtrate was then purified via a HisTrapTM HP column (GE Healthcare, United States) following the manufacturer’s instructions. Further purification of the eluted protein was performed via dialysis desalination and freeze-drying. The protein purity and molecular weight were determined by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and staining with Coomassie Brilliant Blue G250. The protein was then used for biological activity assay.

The disease control ability of BAR11-treated plants was determined by supervising for disease symptoms after pathogen inoculation with protein elicitors at a concentration of 20 μM. This was determined by detached soaking ‘Fuji’ apple leaves for 10 min with the elicitor. For the negative control, sterile water was used in the same way. The apple leaves were inoculated with V. mali at 24 h after BAR11 treatment, and the inoculation assay was conducted as described below (Li et al., 2015, 2016).

Four-week-old A. thaliana leaves were treated with 5 mL of 20 μM recombinant BAR11, and sterile water was used as the control. Three days later, systemic leaves were sprayed with Pst DC3000 cell suspension in 10 mM MgSO₄ containing surfactant Silwet L-77 (OD₆₀₀ = 0.1). Inoculated plants were maintained in a growth chamber (Percival AR800, United States) at 22°C with 70% relative humidity and a 16-h day/8-h night cycle (Liu et al., 2016).

The production of H₂O₂ in the A. thaliana leaves was observed as previously described (Niu et al., 2011) without any change in our study. The A. thaliana leaves were treated with 20 μM BAR11, and sterile water was used the control. The leaf sections of 12 h post-treatment (hpt) were harvest for DAB-staining. The final leaves were then placed in a destaining solution bath (0.15% trichloroacetic acid w/v dissolved in 3:1 ethanol mixed with chloroform solution) to clear the chlorophyll, and then leaf segments were preserved in the saturated chloral hydrate to be transparent for 2–3 days. H₂O₂ production in the A. thaliana leaves was then examined under a light microscope (Olympus BX51, Japan). To determine callose deposition, A. thaliana leaves were sampled at 12 hpt and stained by the aniline blue method as described previously (Niu et al., 2011) without any modification in our study. These experiments were performed three times in the same way.

The relative expression levels of pathogenesis-related proteins (PR1, PR2, and PR5), radical-responsive glutathione transferase (GST1), lipoxygenase 1 (LOX1), lipoxygenase 2 (LOX2), plant defensin gene 1.2 (PDF1.2), and non-expressor of pathogenesis related 1 (NPR1) were analyzed in plants. Total RNA was extracted from frozen A. thaliana leaf samples using a quick RNA isolation kit (Huayueyang, Beijing, China). Using the oligo dT primers, the first-strand cDNA was synthesized in 20-μL reactions containing 1 μg total RNA using the RevertAid first strand cDNA synthesis kit (Thermo, United States). Primers specific to PR1, PR2, PR5, GST1, LOX1, LOX2, PDF1.2, NPR1, and EF-1a were designed using the Primer Quest Tool by IDT¹. Twenty-five cycles of independent PCR were performed using 1-μL cDNA samples and gene-specific primers. EF-1a, a constitutively expressed gene, was used as an endogenous control in the RT-PCR examination.

Agroinfiltration was carried out following Goto et al. (2007) with a slight modification. Briefly, a cell suspension of pre-cultured Agrobacterium tumefaciens strain GV3101 carrying an expression plasmid (pCAMBIA-1302:BAR11:GFP) was adjusted to OD₆₀₀ = 0.4 and infiltrated into the upper leaves of 4-week-old N. benthamiana plants using a 1-mL syringe. A. tumefaciens cells carrying the bar gene (pCAMBIA-1302:bar) were subsequently infiltrated into the same site. Control plants were infiltrated with A. tumefaciens carrying an empty pCAMBIA-1302 vector. The A. tumefaciens leaves were then shredded 2–3 days post-injection (dpi), and a laser confocal microscope (Olympus BX-51) with an excitation wavelength of 485 nm was then used to observe for the localization of green fluorescent protein (GFP). Furthermore, 0.8 M of sorbitol was used to treat the tobacco cells for 15 min for analyzing the fluorescence distribution.

For immunoprecipitation experiments, total protein was extracted from the inoculated N. benthamiana leaves at 2 dpi with the 1302:BAR11:GFP or 1302:AtCAT (1, 2, 3):HA constructs, and subjected to immunoprecipitation via GFP-Trap^®_A per the manufacturer’s protocol. GFP-trap beads (20 μL; Chromotek) were used following the standard operating protocols to immunoprecipitate GFP-bar::AtCAT (1, 2, 3)-HA from protein extracts (Lyzenga et al., 2013).

Filtrated leaves from transiently expressed leaves of N. benthamiana were harvested and frozen in liquid nitrogen. Samples frozen with liquid nitrogen were lysed in 200 μL Lysis Buffer (50 mM Tris-HCl, pH 7.4, 150 mM NaCl, 1 mM EDTA, 5 mM DTT, 0.1% Triton X-100, 1× protease inhibitor cocktail, and 1 mM PMSF) in an ice-bath for 30 min and vortexed every 10 min, and then 100 μL of loading buffer was added (Hong et al., 2017). The extracts were boiled for 10 min and then centrifuged at 12000 ×g for 10 min at 4°C. The supernatant (20 μL) was separated on a 10% SDS-PAGE gel, followed by a transfer onto PVDF membrane via wet electroblotting. For the co-immunoprecipitation assay, detection of the GFP or HA tags in transiently expressed N. benthamiana leaves was performed using a mouse monoclonal GFP or HA antibody (Sigma-Aldrich) for 12 h at 4°C, respectively, followed by alkaline phosphatase-conjugated goat anti-mouse IgG secondary antibody for 1 h at 25°C. The alkaline phosphatase chromogen kit (Thermo, United States) was used to detect the secondary antibody.

All data were analyzed separately using the SPSS 16.0 software (SPSS Inc., United States). Significant differences between mean values were determined using the student’s t-test (P < 0.05).

BAR11 is a hypothetical protein secreted by Saccharothrix yanglingensis strain Hhs.015. The results showed that the molecular weight of the BAR11 peptide is 27900 Da, the number of negatively charged residues is 21, and the number of positive residues is 25. In addition, the peptide also consists of 23.1% alanine, 13.6% leucine, and 10.1% glycine. A Secondary Structure Prediction Server analysis for the secondary structure of BAR11 indicated that the peptide chain has four α-helices with no β-folding and irregular curl. The predicted results for the transmembrane domain showed that there is a transmembrane domain in the N-terminal region, which has a distinct hydrophilic and hydrophobic region and a distinct hydrophilic region at position 120–125. Furthermore, the N-terminal, C-terminal, and central regions of the sequence have a hydrophobic region. A BLAST comparison showed that the BAR11 nucleotide sequence (Accession number: MG586232) shares the highest sequence similarity to proteins of several beneficial microorganisms and contains the conserved domain of DUF305 (Figure 1A). The three-dimensional structure of the protein DUF305 secreted by Streptomyces coelicolor (2qf9. 1.A) was used as a template to model the three-dimensional structure of the BAR11 sequence. The structure is mainly composed of an α-helix, and the same result was obtained with the Phyre² Prediction Server (reliability, 100%; coverage, 73%; Figure 1B).

Functional analysis of BAR11 was investigated via heterologous expression in E. coli to produce enough protein. The crude extracts were applied onto a HisTrapTM HP column (GE Healthcare), eluted with 50–200 mM imidazole, and then purified by dialysis desalination and freeze-drying (Bu et al., 2014). Finally, the recombinant protein showed a single band with the relative molecular mass of 27 kDa by SDS-PAGE analysis (Figure 2A). The protein concentration was 10 μM as determined using a BCA Protein Assay Kit (PIERCE, United States), followed by a transfer onto a PVDF membrane via wet electroblotting. His-BAR11 was then subjected to Western blot analysis using a mouse monoclonal His antibody (Figure 2B).

In order to investigate whether BAR11 can enhance plant disease resistance, apple leaves were inoculated with the pathogen V. mali at 24 hpt after treatment with BAR11 (1 mg/mL) or sterile water. Compared with sterile water-treated control leaves, BAR11-treated leaves showed noticeable weaker disease symptoms (Figures 2C,E). At 4 dpi, pathogen density in A. thaliana leaves which were pretreated with BAR11 was led to a significant reduction (P < 0.05) in pathogen density in A. thaliana leaves. A. thaliana bottom leaves were infiltrated with BAR11 or sterile water. Three days after the initial treatment, A. thaliana upper leaves (non-treated with BAR11) on the same plant were challenged with a virulent strain of DC3000. Plants with different pre-treatments and DC3000 infections were examined 4 dpi. We observed that BAR11 induced effective SAR against Pst DC3000 infection (Figures 2D,F). These results indicate that BAR11 significantly increased plant defense levels against the pathogen.

Systemic acquired resistance is usually associated with the occurrence of early cell defense reactions, such as rapid explosion of reactive oxygen and accumulation of callose. To investigate whether BAR11-induced disease resistance is associated with primed defense responses in leaves, we examined H₂O₂ accumulation and callose deposition in A. thaliana. H₂O₂ accumulation and callose deposition induced by BAR11-treatment and H₂O control were detected at 12 hpt in the A. thaliana leaves of plants (Figure 3). H₂O₂ was assayed using DAB and significant brown precipitates were observed in recombinant BAR11-treated A. thaliana leaves. Callose deposition in the A. thaliana leaves was visualized with ultraviolet excitation.

To further investigate the resistance mechanisms of recombinant BAR11-treated plants, we soaked apple leaves and sprayed 4-week-old A. thaliana leaves with recombinant BAR11 protein and then detected the expression levels of several plant defense-related genes. The pathogenesis-related proteins PR1, PR2, and PR5 are all markers of the SA-dependent defense pathway (Uknes et al., 1992). The qRT-PCR results for the apple leaves demonstrated that these pathogenesis-related proteins were significantly upregulated in response to recombinant BAR11 (Figure 4A). The PR1 expression reached a peak by 6.4-fold at 24 hpt after BAR11 treatment. Expression of PR2 continuously increased by 8.4-fold at 36 hpt. PR5 was significantly upregulated by approximately 10-fold at 48 hpt. The levels of these genes then decreased, but were still upregulated when compared to that in the control. Meanwhile, recombinant BAR11 treatment triggered the expression of PDF1.2 (Figure 4A), which is a crucial JA-responsive protein. The BAR11-induced PDF1.2 reached 2.4-fold at 24 hpt after the recombinant BAR11 treatment, and was subsequently upregulated by 4.8-fold at 48 hpt. Furthermore, NPR1 was upregulated by 4.9-fold (Figure 4A). The qRT-PCR results for the A. thaliana leaves demonstrated that the expression of pathogenesis-related proteins PR1, PR2, and PR5 was significantly upregulated in response to recombinant BAR11 (Figure 4B). The expression levels of the PR1, PR2, and PR5 gene were significantly upregulated at 1 and 2 dpt, and the maximum level of the gene increased by 4.9-fold at 1 dpt. Meanwhile, recombinant BAR11 triggered LOX2 expression and PDF1.2 upregulation (Figure 4B); PDF1.2 expression was increased by 310-fold at 4 dpt and subsequently upregulated by 338-fold at 5 dpt. All these are JA-responsive marker genes. NPR1 and GST1 were also upregulated (Figure 4B). The present results indicate that the BAR11-induced host plants’ resistance to pathogens may affect both SA and JA/ET signaling pathways. Furthermore, qRT-PCR demonstrated notable induction of these marker genes, suggesting an induction of SA and JA signaling in BAR11-treated plants.

The cellular localization of BAR11 was determined by generating a fusion with enhanced green fluorescent protein (eGFP). To this end, we fused bar11 without the signal peptide sequence to egfp, to produce the pCAMBIA-1302-bar11-egfp construct. Then the A. tumefaciens strain carrying the pCAMBIA-1302-bar11-egfp construct was infiltrated into N. benthamiana leaves, and GFP fluorescent signals were observed in the nucleus and membrane. After 15-min treatment with 0.8 M sorbitol, the cells were clearly separated from the cell wall and the fluorescence was localized in the cell membrane and nucleus (Figure 5A). Isolated proteins were used for Western blot analysis with GFP antibodies (Figure 5B).

To identify potential protein targets in N. benthamiana, in planta co-immunoprecipitation with GFP-Trap^®_A beads was conducted, followed by liquid chromatography–tandem mass spectrometry (Figure 5C). The results showed that there were additional candidate target proteins, and combined with the pre-active oxygen accumulation phenotype observed, which suggests that CAT may be the target of BAR11 interaction. Then the interaction between the candidate and BAR11 was examined via bimolecular fluorescence complementation. The A. tumefaciens strain carrying the BAR11-YC and NbCAT1-YN constructs were co-infiltrated into N. benthamiana leaves, and the results showed that BAR11 could interact with NbCAT1 (Figure 5D).

The A. thaliana genome contains a small group of homologous genes belonging to the catalase family, including CAT1, CAT2, and CAT3 (Frugoli et al., 1996). In order to further clarify whether BAR11 interacts with AtCATs (1, 2, 3), BAR11 and AtCATs (1, 2, 3) on the YC-1301 and YN-1301 were constructed, and the recombinant vectors BAR11-YC and AtCATs (1, 2, 3)-YN were then obtained. To determine the subcellular site(s) for interaction of the elicitor BAR11 interacts with plant catalases, BAR11-YC and AtCATs (1, 2, 3)-YN fusions were transiently co-expressed via agroinfiltration in N. benthamiana leaves, with the empty vectors YN-1301 and YC-1301 as negative controls. Yellow fluorescence signal was observed under laser confocal microscopy. Co-expression of BAR11-YC with AtCATs (1, 2, 3)-YN (Figure 6A) in tobacco leaf epidermal cells resulted in a fluorescent signal in the nucleus and cytoplasm. In contrast, transient expression of the empty vectors YN-1301 and YC-1301 did not trigger YFP. Taken together, the results suggest that there was an interaction between BAR11 and AtCATs (1, 2, 3).

In the A. thaliana vegetative tissues, CAT2 and CAT3 play key roles to represent major catalase activity (Frugoli et al., 1996; Mhamdi et al., 2010). As a result, here, possible interactions of BAR11 with CAT2 in plants were tested. The A. tumefaciens strain carrying the GFP:BAR11 construct was co-infiltrated with that containing the AtCAT2:HA construct into N. benthamiana leaves, and total proteins from the treated N. benthamiana leaves at 2 dpi were extracted for follow-up experiments. We used the GFP-trap beads (20 μL; Chromotek) to immunoprecipitate GFP-BAR11 from total protein extracts. Isolated proteins were then analyzed by Western blot with GFP or HA antibodies (Figure 6B). The results demonstrated that the elicitor BAR11 interacted with AtCAT2 and then interfered with the plant’s H₂O₂ levels, thus influencing the defense pathways and improving plants resistance against pathogens.