Tobacco (Nicotiana tabacum cv. Samsun NN) and tomato (Lycopersicon esculentum cv. Zhongza-9) plants were grown in a greenhouse at 25°C, with a day/night period of 16/8 h, and cotton (G. hirsutum cv. Guoxin) plants were maintained at 28°C, with a day/night period of 16/8 h. V. dahliae strain XH-8 was preserved at the China General Microbiological Culture Collection of China (CGMCC no. 7611). B. cinerea strain BC-98 was preserved at the CGMCC (CGMCC no. 7057). The fungi were grown on potato dextrose agar medium (PDA) at 25°C in the dark. Pseudomonas syringae pv. tabaci and Agrobacterium tumefaciens AGL-1 were stored in 20% glycerol at −80°C and grown on the KB (Rif⁵⁰) and LB (K⁵⁰ and Rif⁵⁰) medium plate at 28°C, respectively.

1 L of CDB liquid medium (NaNO₃, 3.0 g/L, MgSO₄·7H₂O, 0.5 g/L, KCl, 0.5 g/L, FeSO₄·7H₂O, 0.01 g/L, K₂HPO₄, 1.0 g/L, sucrose 30 g/L) with 5% roots fragments (w/v) from 6-week-old cotton plants was inoculated with 1 ml of a V. dahliae spore suspension (10⁵/ml) and subsequently cultured for 2 days in the dark at 25°C with shaking at 180 rpm. The culture filtrate was then centrifuged at 2,500 g for 30 min at 4°C, after which the supernatant was collected and passed through a 0.22 μm filter (Millipore, Suzhou, China). The proteins in the supernatant were precipitated with TCA as described in Wang et al. (2007). Protein concentrations were determined using a BCA™ Protein Assay Kit (TransGen Biotech, Beijing, China) with BSA as a standard. The mixture of proteins was analyzed via LC-MS/MS (Q Exactive, Bruker Daltonics K.K., Tokyo, Japan), combined with a BLAST search of the V. dahliae VdLs.17 genome database (http://www.broadinstitute.org/annotation/genome/verticillium_dahliae/MultiHome.html).

A 360 bp vdcp1 fragment (amplified with primers vdcp1F/vdcp1R, Table S1) without the predicted signal peptides and stop codons was inserted into the pPICZαA plasmid at the EcoRI and XbaI sites. The recombinant plasmid pPICZαA-vdcp1, pPICZαA-His-tag or empty plasmid pPICZαA were linearized with PmeI and transformed into Pichia pastoris KM71H for expression. The transformants were cultured in BMGY (1% yeast extract, 2% peptone, 100 mM potassium phosphate, pH 6.0, 1.34% yeast nitrogen base, 4 × 10⁻⁵ % biotin, 1% glycerol) to culture OD₆₀₀ = 6, after which the cell pellet was collected and resuspended in BMM (100 mM potassium phosphate, pH 6.0, 1.34% yeast nitrogen base, 4 × 10⁻⁵ % biotin, 0.5% methanol), followed by induction for 3 days with 0.5% methanol. The resulting culture supernatant was used for the purification of VdCP1 or His-tag (Zhang et al., 2017a,b). The purification was carried out with HisTrap HP pre-packed minicolumns (GE Healthcare Life Sciences, Uppsala, Sweden). The purified VdCP1 or His-tag was kept in protein buffer (20 mM Tris, pH 8.0). VdCP1 was analyzed via Tris-SDS-PAGE and confirmed using Western blotting for His-tag and with MS, and His-tag peptide was analyzed via Tris-Tricine-SDS-PAGE with the low molecular mass marker (RTD6110, Beijing Real-Times Biotechnology Co. Ltd., China) and confirmed by Western blotting. The concentration of the purified protein was measured using the BCA™ Protein Assay Kit, and the protein was then frozen at −80°C in small aliquots until use.

The necrosis-inducing activity, the production of ROS, and the measurement of ion electrolyte leakage, phenylalanine ammonia-lyase (PAL), polyphenol oxidase (PPO) and peroxidase (POD) activities and SA production were assayed in VdCP1-infiltrated plants. His-tag/BSA-infiltrated plants were used as controls. VdCP1, His-tag or BSA was infiltrated into one leaf of 4-week-old tobacco, cotton and tomato plants using a blunt 1-ml syringe, 24 h later the leaves were cut and treated with DAB (1 mg/ml, pH 3.8). Leaves were vacuum infiltrated and immersed with DAB solution for 3 h. To visualize DAB deposits, leaves were then incubated in absolute ethanol to eliminate chlorophyll and photographed (Frías et al., 2011). The DAB staining intensity on photographs was quantified using software ImageJ and is shown as the number of DAB pixels relative to total pixels according to a previous study (Chang et al., 2014). To assay electrolyte leakage, one leaf of 4-week-old tobacco plant was infiltrated with VdCP1, His-tag or BSA, three leaf discs of 1 cm diameter were cut at 6 hpi and submerged in 1 ml water at 4°C with shaking at 200 rpm, and the conductivity was measured at the indicated time points with a MP526 ion conductivity meter (LabSen, Shanghai, China) (Frías et al., 2011). Three tobacco plants were treated to detect ROS and electrolyte leakage in every time point, three biological replicates were carried out. One cotyledon of 4-week-old cotton plant was infiltrated with VdCP1, His-tag or BSA, and the leaves were collected to measure the activities of PAL, PPO and POD according to previously described methods (Hano et al., 2008; Zhang et al., 2010). 3 cotton plants were treated to measure the enzymes activities in every time point, three biological replicates were performed. At 72 h after infiltrating one of cotton cotyledons with VdCP1, His-tag or BSA, the roots were weighed and frozen in liquid nitrogen. For each sample, 0.1 g of the frozen tissue was extracted and quantitated for free SA as described previously (Lv et al., 2016). 3 cotton plants were treated and three biological replicates were performed.

VdCP1 (30 μM) was incubated with 10 mg chitin from crab shells (catalog no. C9752, Sigma-Aldrich) in 50 mM sodium acetate buffer (1 ml, pH 6.5) at 38°C with shaking at 320 rpm (Baccelli et al., 2013; Frischmann et al., 2013). BSA at the same concentration was included as a control. After the indicated times, the tubes were centrifuged at 12,000 × g for 2 min, the supernatants were then transferred to new tubes, and the pellets were washed with buffer at least 3 times to remove any remaining protein solution. Next, the pellet was resuspended in buffer containing 2% SDS and incubated at 99°C. After 10 min, the suspension was centrifuged at 12,000 × g for 2 min, and the supernatant was transferred to a new tube and precipitated using the methanol and chloroform method (Wang et al., 2007). Unbound proteins from the supernatant and bound proteins from the pellet were analyzed via sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE). In addition, VdCP1 (30 μM) was incubated with 1 cm-diameter Whatman no. 1 filter paper (Whatman, Milan, Italy) in 50 mM sodium acetate buffer (1 ml, pH 6.5) (Baccelli et al., 2013). Protein buffer only or buffer containing BSA at the highest concentration were used as negative controls, and protein CP expressed in P. pastoris KM71H was used as positive control. At the end of the incubation period (48 h, 38°C, 320 rpm), the presence of paper fragments in the suspension was detected via spectrophotometric measurements of absorbance at 500 nm and photographed.

The wells of a microtitre plate were filled with 200 μl of potato dextrose broth (PDB) at pH 6.85 containing 0.2 U of chitinase from Streptomyces griseus (catalog no. C6137, Sigma-Aldrich) that had been resuspended in the buffer (50 mM potassium phosphate, pH 6.0). The controls were carried out by adding the buffer instead of the chitinase solutions. A final concentration of 10⁶/ml spores of V. dahliae WT, Δvdcp1 and Δvdcp1-Res strains was added into PDB, and 68 μg/ml of resazurin dye (catalog no. R7017, Sigma-Aldrich) was added to each well to measure fungal growth. Resazurin is reduced from blue to light pink color in presence of actively growing cells. The plates were incubated at 25°C in the dark, and the absorbance was measured at 578 nm at the indicated times. V. dahliae growth inhibition. The percentage of V. dahliae growth inhibition was calculated as 100 minus the percentage of the ratio between net absorbance values of each sample and net absorbance of the medium with the resazurin dye but without V. dahliae (Sella et al., 2014; Quarantin et al., 2016).

V. dahliae with mutations in the vdcp1 (Δvdcp1) gene and vdcp1 complementary strain (Δvdcp1-Res) were constructed using the A. tumefaciens-mediated transformation method (Liu et al., 2013). Primers vdcp1aF/vdcp1aR, vdcp1bF/vdcp1bR and Hyg-F/Hyg-R were used to amplify the 1000-bp 5′ region of vdcp1, the 1000-bp 3′ region of vdcp1 and the 1800-bp hygromycin resistance cassette, respectively. A nested PCR reaction was carried out with primers Nest-F/Nest-R, which contain gateway BP reaction adaptors, these were then employed to amplify the fusion PCR sequence from the fusion of the three amplicons. Subsequently, a recombinant plasmid was constructed via homologous recombination between the final fusion amplicon and the pGKO2-Gateway vector, which contains the herpes simplex virus thymidine kinase (HSVtk) gene in T-DNA and was used as a lethal gene for counterselection against ectopic transformants with 5-fluoro-2′-deoxyuridine (F2dU). To generate the vdcp1 complementation vector, the primers vdcp1CF and vdcp1CR, which contain KpnI and XbaI enzyme sites, respectively, were employed to amplify the sequence containing vdcp1 and its 1,100-bp 5′ and 3′ regions from the V. dahliae genome. The vdcp1 cassette was ligated into the pCAM vector (Liu et al., 2013), which contains geneticin and hygromycin resistance selection markers. All primers are listed in Table S1. The Agrobacterium strain AGL-1 was used to transform the recombinant plasmid into V. dahliae spores.

To analyse the expression of vdcp1 gene, V. dahliae was cultured and collected from different culture media. The stem base of 2-week-old cotton plant was inoculated with a spore suspension of 10⁸/ml, and the inoculated tissues were collected at the indicated time points. 3 plants were inoculated for the assays in every time point and three biological replicates were performed. Total RNA from V. dahliae and V. dahliae-infected cotton plants was extracted with an E.Z.N.A.® Fungal RNA Kit. To analyse the expression of defense-related genes, one leaf of 4-week-old tobacco and cotton plants was treated with VdCP1 for every time point, and control plants were infiltrated with His-tag or BSA solution, 3 plants were treated for the assays and three biological replicates were performed. Total RNA was extracted from the indicated tissues at different time points using a plant RNA kit (TransGen Biotech). Amplification was performed in an ABI7500 Real-Time PCR system with TransStart Green qPCR SuperMix UDG (TransGen Biotech). All primers are listed in Table S1. The V. dahliae, tobacco and cotton actin genes, vdActin (XM_009660170), NtEF1α (XM_016658252) and GhActin (AY305737), respectively, were used to correct for sample-to-sample variation in the amount of RNA. The relative mRNA quantities were calculated from the threshold cycle using the ΔΔCt method (Schmittgen and Livak, 2008). Triplicate reactions were performed for all of the biological replicates to calculate the mean and deviation from the mean was calculated from the standard deviation.

Assays for the pathogenicity of the pathogens P. syringae pv. tabaci and B. cinerea were performed as previously described (Zhang et al., 2015). 4-week-old tobacco plant was infiltrated with VdCP1 on one side of the central vein of one leaf and cultured in the phytotron for 3 days, control plant was infiltrated with His-tag solution, 20 plants were treated for the assays and three biological replicates were carried out to calculate the average values. The treated leaves were then soaked in a P. syringae pv. tabaci suspension (OD₆₀₀ = 0.05 in 10 mM MgCl₂) with gentle swirling for 30 s. Three days after challenge with the bacteria, three punches were collected from each leaf and ground with a pestle in 1 ml of sterilized water. The sample suspensions were vortexed thoroughly and serially diluted to the appropriate dilution. The bacteria were subsequently spread on KB (Rif⁵⁰) plates, and the plates were incubated for 48 h at 28°C. The number of colonies on each plate was then counted, the plates were photographed, and the degree of infection was calculated. For B. cinerea inoculation, the treated tobacco leaves were inoculated with 10 μl of the spore suspension (5 × 10⁵/ml in 60 mM KH₂PO₄, 10 mM glycine, 0.01% Tween 20, 0.1 M glucose) and cultured at 22°C under conditions of high humidity for 3 days. Lesions were observed at the indicated time points, and lesion sizes were calculated. Pathogenicity assays for the vdcp1 WT, knockout and complementary strains were conducted using 2-week-old cotton seedlings. For each strain, conidial suspensions at 5 × 10⁶ /ml in sterile water were prepared. Cotton seedlings grown on soil in each pot and 15 pots were prepared for each strain, and this setup was replicated three times. 15 × 15 ml conidial suspension was pre-placed in 15 new trays respectively and each pot was fully immersed in the new tray until complete absorption. The inoculated seedlings were then grown for 2–3 weeks at 28°C, with a day/night period of 16/8 h. In addition, to test the ability of VdCP1 in inducing cotton disease resistance against V. dahliae, 2-week-old cotton seedlings were pre-infiltrated with VdCP1 on one cotyledon and cultured in the phytotron for 3 days, control plants were infiltrated with His-tag solution. The pre-treated seedlings were inoculated with WT conidial suspensions and grown as described above. The degree of wilt disease was divided into five grades as previously described: grade 0—healthy plant; grade 1—yellowing cotyledons; grade 2—wilting of the cotyledons and one true leaf; grade 3—wilting of all leaves and chlorosis; and grade 4—plant death (Liu et al., 2013). The control efficiency for V. dahliae was calculated using the following formulas: disease index value = [∑ (the number of seedling of every grade × relative grade)/ total seedlings × highest score (4)] × 100. Disease reduction (%) = [(disease index of control – disease index of treatment)/ disease index of control] × 100.

Many studies indicate that secretory proteomics offers an effective tool for exploring interesting fungus-plant interactions proteins (Medina et al., 2005; Kim et al., 2007; Paper et al., 2007; Shah et al., 2009; Espino et al., 2010). Although several secretory proteins from V. dahliae have been reported, few proteomic studies have been conducted on this fungus. Here, two integrated analyses of the V. dahliae early secretome were carried out individually. In total, 256 proteins, many of which were present due to cell lysis (Supplementary File 1), were identified in the supernatant. A number of putative pathogen-host interaction proteins in the secretome are highlighted and listed in Table 1, these proteins include cellulase, pectate lyase, chitinase, proteases, etc., as well as a potential effector (elicitor) protein that was analyzed from the Pathogen Host Interactions database (http://www.phi-base.org). It is meaningful to explore the roles of these interaction proteins in deciphering the pathogenic mechanism of V. dahliae.

Among these V. dahliae-cotton interaction proteins, one interesting SnodProt1 protein (VDAG_06199), designated VdCP1, was identified in the culture medium. VdCP1 contains 138 amino acids, has a predicted signal peptide at the cleavage site between amino acids 18 and 19 (SignalP 4.1 server), and is a typical CPP family homolog, exhibiting a low molecular weight, four conserved cysteine residues and high hydrophobicity. The mature VdCP1 protein contains 120 amino acid residues, has a predicted molecular mass of 12,607.1 Da. The amino acid sequence of VdCP1 indicated that it is highly homologous to Epl1, Sm1, BcSpl1 and CP (70.3, 68.1, 61.2, and 50.8%, respectively), which are among the best-characterized members of the CPP family (Figure S1). An amino acid sequence analysis of the post-translational modification sites using the NetOGlyc 4.0 and NetNGlyc 1.1 servers (http://www.cbs.dtu.dk) revealed the presence of potential O-glycosylation (residue 38, S) and N-glycosylation sites (residue 132, N), respectively, as indicated by the two arrows in Figure S1. In addition, hydropathicity plot (Kyte and Doolittle) analysis revealed that VdCP1 contains a high percentage of hydrophobic residues (Figure S1).

Like many previously characterized CPPs, VdCP1 was found in the culture medium. To obtain more details regarding the transcription of vdcp1, the vdcp1 mRNA level in different culture mediums and at various stages of infection was measured via qPCR. The vdcp1 mRNA levels increased with respect to ungerminated conidia in every medium studied at 48 h post-inoculation (hpi) (Figure 1A). In CDB culture medium supplemented with 5% cotton/tobacco roots, vdcp1 mRNA levels showed relatively high upregulations, and vdcp1 mRNA levels didn't show statistical differences between CDB and CDB with sucrose/nitrogen stress. During V. dahliae infection process, the amounts of vdcp1 mRNA continuously increased and reached a maximum 456-fold increase over the levels observed at time zero (ungerminated conidia) (Figure 1B), thus we can conclude that vdcp1 is expressed in the presence of all studied culture mediums and during all stages of the infection process, especially in planta during the late stages of infection.

To investigate the role of VdCP1 in the V. dahliae invasion process, the Δvdcp1 and Δvdcp1-Res strains were constructed using the strategy shown in Figure S2. Two random transformants that exhibited hygromycin resistance and were unaffected by F2dU were selected. Genomic DNA and cDNA from both mutant strains were analyzed by PCR amplification, and all showed the expected bands (Figure S2). Moreover, qPCR using cDNA obtained from the transformants revealed no expression of vdcp1 (data not shown), indicating that the vdcp1 gene had been successfully knocked out in the selected transformants (Δvdcp1.1 and Δvdcp1.2). Complemented strains (Δvdcp1.1-Res and Δvdcp1.2-Res) were generated and characterized in a similar manner (Figure S2). The phenotypes of the Δvdcp1 and Δvdcp1-Res strains with respect to their growth rate, sporulation capacity and mycelium/spore morphology were not discernibly different from those of the WT (Figure S3), whereas compared with the WT, both Δvdcp1 strains displayed a statistically significant decrease in their virulence to cotton, with nearly 20% infection, as calculated by the disease index, and the Δvdcp1-Res strains showed restored virulence (Figures 2A,B).

Previous studies have shown that many CPPs are elicitors of plant defense responses. To test whether VdCP1 exhibits similar elicitation activity, vdcp1 was inserted into the pPICZαA plasmid and expressed in P. pastoris KM71H. The purified VdCP1 showed a single band of 18 kDa upon SDS-PAGE and Western blotting and showed substantial matching with the VdCP1 sequence by mass spectrometry (MS) identification (Figure S4). Approximately 30 mg of VdCP1 was obtained from 100 ml of BMM culture supernatant.

In general, an elicitor can be recognized by PRRs, and this recognition results in a series of characteristic symptoms such as the production of ROS, expression of defense genes and leakage of ion electrolytes (Chinchilla et al., 2009). Here, the production of hydrogen peroxide was detected using 3,3′-diaminobenzidine (DAB) staining. The result showed that VdCP1 (10 μM) triggered a marked increase in brown DAB precipitate in tobacco, cotton and tomato leaves (Figure 3A), and the intensity of signals with VdCP1 differed significantly from that with His-tag or BSA (Figure 3B). Moreover, the expression of 6 defense-related genes in treated tobacco leaves was elevated after infiltration with 10 μM VdCP1 solutions (His-tag or BSA were used as controls, respectively) (Figure 3C), these genes included (1) PR-1a (XM_009800164) and PR5 (NM_001325216), which are pathogenesis-related genes (Spoel et al., 2009), (2) LOX (lipoxygenase, NM_001325784) (Veronesi et al., 1996) and EDS1 (enhanced disease susceptibility 1, XM_016636310) (Rietz et al., 2011), which encode a key enzyme involved in the biosynthesis of jasmonic acid (JA) and a lipase that modulates salicylic acid (SA)-dependent disease resistance, respectively, as well as (3) HSR203J (XM_016619229) and HIN1 (Y07563), which are two HR marker genes in tobacco (Pontier et al., 1994, 2001). Interestingly, two HR marker genes were induced (Figure 3C), however, no apparent necrosis appeared in the infiltration area at 24 hpi (data not shown). Additional assays using different concentrations of VdCP1 were conducted to further examine its necrosis-inducing activity. At concentrations of up to 200 μM, necrosis was observed in the infiltration area at 24 hpi (Figure S5). Cotton and tomato plants showed similar results (Figure S5), indicating that a high concentration of VdCP1 is required for HR-inducing activity. Additionally, the infiltration of tobacco leaves with VdCP1 (100 μM) also induced electrolyte leakage from the treated leaf tissue, as demonstrated by an increase in conductivity over time compared with that in control samples treated with His-tag or BSA at the same concentration (Figure 3D). When the assays were performed using different concentrations of VdCP1, the conductivity varied in a dose-dependent manner (Figure 3E). The minimum concentration of VdCP1 required to cause ion electrolyte leakage at 6 hpi was 50 μM.

Furthermore, the activities of three additional defense-related enzymes, PAL, POD and PPO, were assayed from 0 to 168 h after treating cotton leaves with 10 μM VdCP1, His-tag or BSA. The data showed that VdCP1 enhanced the activity of PAL, POD, and PPO, the activities of these enzymes peaked at 72, 96 and 120 h, respectively (Figures 4A–C).

To further explore the ability of VdCP1 to induce systemic acquired resistance (SAR) in plants, a qPCR analysis was performed to analyse the expression of GhNPR1 (a master gene for SAR, DQ325523) and two defense-related genes [PR genes, GLU (β-1,3-glucanase, CD486342) and CHT (chitinase, CD485880)] in cotton roots. The expression of GhNPR1, GLU, and CHT was significantly upregulated, respectively, at 72 h after inoculating cotyledons with VdCP1 (100 μM) compared with control plants treated with His-tag or BSA (Figure 5A). In the meantime, at 72 h after inoculating cotyledons with VdCP1 (100 μM), the SA levels in cotton roots increased prominently compared with those in control plants (Figure 5B).

To explore whether VdCP1 could confer plants disease resistance, tobacco leaves infiltrated with 100 μM VdCP1 or His-tag were assayed for resistance against B. cinerea and P. syringae pv. tabaci. Remarkably, tobacco leaves pre-treated with VdCP1 showed disease resistance compared to the control plants. VdCP1-treated tobacco plants showed a significant reduction in the lesion area caused by B. cinerea on leaves (Figure 6A), the B. cinerea lesion area decreased by 29% (Figure 6B). The number of P. syringae pv. tabaci colonies of VdCP1-treated tobacco leaves on the plate was significantly lower than that of control leaves (Figure 6A), and the incidence of P. syringae pv. tabaci infection was as low as 43% that of the control (Figure 6B). In addition, cotton roots were inoculated with V. dahliae after 72 hpi of cotyledons with 100 μM VdCP1 or His-tag. After 2 weeks, wilt symptoms on the leaves of the VdCP1-induced and control cotton seedlings are shown in Figure 6C. The results showed a 32% reduction in the disease index compared with that of the control plants (Figure 6D).

The above results suggest that VdCP1 contributes to the pathogenicity of V. dahliae and elicits plant defense responses. Several CPPs with dual functions have been shown to bind to chitin and they show an expansin-like activity on cellulosic materials (Baccelli et al., 2013; Frischmann et al., 2013). To determine the chitin-binding activity of VdCP1, we performed experiments using chitin from crab shells. The results showed that VdCP1 could bind to chitin rapidly (5 min after incubation), while there was no binding activity between BSA and chitin (Figure 7A). The presence of protein CP on filter paper discs resulted in the appearance of paper fragments in the suspension (Baccelli et al., 2013), we then tested the expansin-like activity of VdCP1 on filter paper. No filter paper-loosening activity was detected for VdCP1 (Figure 7B), even at high concentrations of 200 μM (data not shown), however, CP (10 μM), which was used as the positive control, was able to loosen the filter paper significantly (Figure 7C).

The direct toxic activity of VdCP1 (100 μM) was tested separately with the fungal pathogen B. cinerea and the bacterial pathogen P. syringae pv. tabaci. No observable effects on growth inhibition were found (Figure S6). This result could rule out the possibility that VdCP1, which we used to induce plant disease resistance, has a direct negative effect on the growth of B. cinerea and P. syringae pv. tabaci.

Although no single phenotypic difference was observed in Δvdcp1 strains compared with WT, VdCP1 was shown to bind to chitin. To determine whether the absence of VdCP1 makes the cell wall of the Δvdcp1 mycelium more susceptible to enzymatic degradation by chitinase, the sensitivity of WT, Δvdcp1 and Δvdcp1-Res strains to the presence of chitinase was studied. The result showed that chitinase could significantly affect the growth of all fungal strains, and the two mutant strains, Δvdcp1.1 and Δvdcp1.2, showed greater growth inhibition than the WT and Δvdcp1-Res strains (Figure 8 and Figure S7), suggesting that VdCP1 can protect the V. dahliae cell wall from degradation by chitinase.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.