The Fon race 1 strain ZJ1, isolated from diseased watermelon from Zhejiang province of China, was used as wild type (WT) strain for fungal transformation, gene targeted disruption and complementation experiments in this study. For growth and stress tolerance assays, the tested Fon strains were cultivated on potato dextrose agar medium (PDA) (200 g/l potato and 20 g/l dextrose, pH7.0) at 26°C under constant fluorescent light. For conidiation assays, the tested Fon strains were cultivated in mung bean liquid (MBL) broth (30 g mung beans boiled in water for 20 min, filtered through cheesecloth and brought to 1 l).

BLASTp was performed using human Not2 protein sequence (GenBank accession no. XP_011536702) as a query to search against downloaded local F. oxysporum genome database. Sequence alignment was carried out using ClustalW program in LaserGene software and phylogenetic tree was constructed using full-length Not2 proteins by neighbor-joining method in MEGA5 software with 1000 replications.

Construction of FonNot2-targeted disruption vector was performed using double-joint PCR method (Yu et al., 2004). A 4.3 kb fragment including FonNot2 coding sequence and putative 5′ and 3′ UTR sequences was used as target region for disruption. A 699 bp 5′ UTR fragment was amplified with primers FonNot2-5UTR-F and FonNot2-5UTR-R (containing a 28 bp overlapping HPH cassette sequence at 5′ end) (Supplementary Table S1); while a 646 bp 3′-UTR fragment was amplified with primers FonNot2-3UTR-F (containing a 28 bp overlapping HPH cassette sequence at 3′ end) and FonNot2-3UTR-R (Supplementary Table S1). A 1349 bp fragment containing the HPH cassette, which encodes hygromycin phosphotransferase under control of the Aspergillus nidulans TrpC promoter, was amplified from plasmid pBS-HPH with primers HPH-F and HPH-R (Supplementary Table S1). The amplified 5′ and 3′ UTR flanking sequences of the FonNot2 gene were fused to the HPH cassette via double-joint PCR using primers FonNot2-5UTR-F and FonNot2-3UTR-R (Supplementary Table S1), yielding a 2694 bp fragment of the targeted disruption vector FonNot2-TD. Fungal protoplasts were generated from WT strain and were directly transformed with the targeted disruption vector FonNot2-TD after purification according to previously described method (Di Pietro and Roncero, 1998). Transformants were selected on PDA supplemented with 100 mg/l hygromycin (Hyg) and the resulting Hyg^R transformants were initially identified by PCR using primers FonNot2-JD-F and FonNot2-JD-R with genomic DNA prepared by CATB method (Luo et al., 2005). To construct the complementation vector, a 3515 bp fragment containing the native promoter and coding sequence of FonNot2 was amplified from genomic DNA with primers FonNot2c-F and FonNot2c-R (Supplementary Table S1) and inserted into plasmid YF11-neo, yielding complementation vector pYF11-neo-FonNot2. Protoplasts were prepared from the ΔFonNot2 strain and were transformed with the complementation vector pYF11-neo-FonNot2. Transformants were selected on PDA with 50 mg/l neomycin and were confirmed by PCR using primers FonNot2C-JD-F and FonNot2C-JD-R (Supplementary Table S1) with genomic DNA prepared by CATB method (Luo et al., 2005). The obtained FonNot2-targeted disruption mutant ΔFonNot2 and complementation strain ΔFonNot2-C were purified via single spore isolation. For Southern blotting, genomic DNA was extracted from individual strain, digested completely with KpnI, separated by electrophoresis on a 0.8% agarose gel, and transferred by capillary action overnight onto a Hybond-N⁺ membrane (Amersham Biosciences, Little Chalfont, UK) using 20 × SSC solution. A 805 bp 5′ UTR flanking fragment of FonNot2 was amplified with primers FonNot2-TZ-F and FonNot2-TZ-R (Supplementary Table S1) and labeled with DIG by random priming method using a DIG High Prime DNA Labeling and Detection kit (Roche Diagnostics, Shanghai, China). Prehybridization, hybridization and detection were performed following the manufacturer’s recommendations.

Radial growth was measured on PDA and colony morphology was photographed after incubation at 26°C for 7 days. Conidia were harvested from 7-day-old cultures and counted with a haemacytometer. Microscopic observation of conidiophores was performed by placing hyphal blocks from 7-day-old cultures on glass cover slides for 24 h and then examined. Conidia harvested from 7-day-old cultures were added into YEPD (3 g yeast extract, 10 g peptone and 20 g dextrose in 1 l, pH 7.0) broth and incubated at 26°C on a rotary shaker at 180 rpm for 12 h. Percentage of conidial germination was determined by microscopic examination of at least 100 randomly selected conidia per field. For cellophane penetration assays, fungal colonies were grown for 3 days at 26°C on cellophane membranes placed on PDA and the cellophane membranes along with the fungal colonies were removed. The PDA plates were incubated for another 1 day to examine the presence of mycelial growth. Stress response assays were performed by inoculating 3-day-old mycelial plugs on PDA plates supplemented with 5 mM H₂O₂ or 3 mM paraquat for oxidative stress assays, or with 0.2% Calcofluor White (CFW) (Sigma–Aldrich, St. Louis, MO, USA) or 0.2% Congo Red (CR) (Sigma–Aldrich, St. Louis, MO, USA) for cell wall sensitivity assays. After incubation at 26°C for 7 days, radial colony growth was measured and percentage of inhibition was calculated by comparing the colony diameter under stressed condition with the diameter under normal condition.

To prepare inoculum, colonized agar plugs from a 7-day-old PDA culture were transferred into a liquid MBL and grown for 4 days on a rotary shaker at 170 rpm at room temperature. Spores were collected by filtering through four layers of sterile cheesecloth and the spore suspension was adjusted to approximately 1 × 10⁷ spores/ml. Pathogenicity was tested by root-dip method using watermelon (Citrullus lanatus L.) cultivar Zaojia (a susceptible cultivar to race 1). Seedlings were grown in a vermiculite: plant ash: perlite (6:2:1) mixture in a growth room at 24°C with a cycle of 16 h light/8 h dark. Two-week-old seedlings were carefully uprooted and washed in tap water to remove soil particles. Roots of the seedlings were dipped for 30 s in spore suspensions prepared from WT, targeted disruption ΔFonNot2 and complementation ΔFonNot2-C strains. The inoculated seedlings were carefully replanted in the same growth medium and allowed for disease development. Disease scores were assessed 3 weeks after inoculation using the following rating scales: 0 = no symptom, 1 = yellowing, 2 = wilting and 3 = death. Fungal biomass measurement was performed as described previously (Thatcher et al., 2009). Root and stem samples were collected at 3, 6, and 9 days post-inoculation and the transcript levels of Fon FonOpm12 and watermelon ClRps10 genes were determined by qRT-PCR using a pair of FonOpm12-specific primers FonOpm12-F and FonOpm12-R and a pair of watermelon ClRps10-specific primers ClRps10-F and ClRps10-R, respectively (Supplementary Table S1). Relative fungal biomass was calculated by normalizing FonOpm12 to watermelon ClRps10 (Lin et al., 2010).

Chitin content was measured according to a previously described method (Bulik et al., 2003). Briefly, ten Fon colonized agar plugs were inoculated into 200 ml YEPD liquid medium and incubated at 28°C on a rotary shaker at 180 rpm for 2 days. Mycelium was collected by filtering through three layers of Waterman filters and ground in liquid nitrogen. Five milligrams of the ground powder was resuspended in 1 ml 6% KOH and heated at 80°C for 90 min. After centrifugation at 12000 g for 10 min, the pellets were re-suspended in 1 ml 10 mM PBS (pH7.4) and the suspensions were re-centrifuged at 12000 g for 10 min. The resultant pellets were re-suspended in 100 μl Mcllvaine buffer (pH6.0) containing 5 μl chitinase from Streptomyces plicatus (Sigma–Aldrich, St. Louis, MO, USA) and incubated at 37°C for 16 h. One hundred microliters of the samples were combined with 100 μl 0.27 M sodium borate and heated for 10 min at 100°C. After cooling, 1 ml of freshly diluted (1:10) of DMAB reagent (10 g p-dimethylaminobenzaldehyde in 12.5 ml concentrated HCl and 87.5 ml glacial acetic acid, diluted 1:10 with glacial acetic acid) was added and incubated at 37°C for 20 min. The absorbance at 585 nm was recorded and the quantity of glucosamine in samples was calculated by referencing to a standard curve prepared with N-acetylglucosamine (Sigma–Aldrich, St. Louis, MO, USA). For quantification of FA content, fifteen Fon colonized agar plugs were inoculated into 200 ml Czapek Dox liquid medium and incubated at 28°C on a rotary shaker at 180 rpm for 15 days and the cultures were filtrated with Waterman filters to remove mycelium and conidia. After centrifugation at 4000 rpm for 10 min, the supernatant was adjusted to pH2.5 with 2 M HCl and extracted with ethyl acetate for three times. The combined ethyl acetate extractions were dried on a rotary evaporator and the resultant residues were dissolved in 5 ml methanol. A high performance liquid chromatography (HPLC) system (Model LC-20AD, Shimadzu Corporation, Kyoto, Japan) with an Agilent EclipseF XDB-C18 (4.6 × 250 mm, 5 μm) was employed to analyze FA content. Elution was carried out using a mobile phase comprising 30% acetonitrile and 70% H₃PO₄ (0.4%) for 20 min at flow rate of 1 ml/min with a UV detector at 280 nm. Before injection, the samples were filtrated through 0.45 μm filters. Quantification of FA in samples was done on the basis of the peak area compared to a standard curve prepared using the same procedure. For measurement of peroxidase (POD) activity, ten Fon colonized agar plugs were inoculated into 200 ml PDA liquid medium and incubated at 28°C on a rotary shaker at 180 rpm for 2 days. Culture filtrates were collected by passing through three layers of Waterman filters and 1 ml of the filtrates were centrifuged at 5000 g for 5 min. Five hundred microliters of the supernatants were added to 1.5 ml reaction mixture containing 50 mM acetate buffer (pH 5.0) and 10 mM ABTS and incubated at 25°C for 5 min with 3 mM H₂O₂. Absorbance at 420 nm was measured spectrophotometrically.

For detection of ROS in vegetative mycelia, hyphae were collected from 2-day-old culture and stained with 1.5 mM 2′, 7′-dichlorodihydrofluorescein diacetate (DCFH-DA) (Beyotime Biotechnology, Nantong, China) in 50 mM phosphate buffer (pH8.5) for 20 min under dark conditions. After washing twice with the same buffer, the samples were observed for fluorescence using a Nikon eclipse Ni-U microscope with a selective GFP filter and digital images were captured with a Nikon DS-Fi1c CCD camera using NIS-Elements D45000 software (Nikon, Tokyo, Japan). For detection of superoxide anion, spores were harvested from 7-day-old cultures and allowed to germinate in YEPD broth at 26°C on a rotary shaker at 180 rpm for 12 h. Hyphae from 2-day-old culture or spores germinated for 12 h were stained with 0.05% (wt/vol) nitroblue tetrazolium (NBT) solution for 20 min. After stopping by the addition of ethanol, the formazan staining was observed under a Nikon eclipse Ni-U microscope using NIS-Elements D44000 software and digital images were captured by a Nikon DS-Fi2 camera (Nikon, Tokyo, Japan). Mean pixel intensity within regions of interest in the digital images of DCFH-DA-stained mycelia or NBT-stained mycelia and spores was calculated using Image Pro Plus 6.0 software. At least 20 representative mycelia or germinating spores were examined in an independent experiment.

For SEM observation, 2-week-old seedlings were inoculated by dipping the roots into Fon spore suspension (1 × 10⁷ spores/ml) on a rotary shaker at 80 rpm and root samples were collected 24 h later. The samples were fixed overnight in phosphate buffer (pH7.0) with 2.5% glutaraldehyde, and then post-fixed with 1% osmic acid in phosphate buffer (pH7.0) for 1 h. After dehydration in an ethanol series, the samples were transferred to a mixture of alcohol and isoamyl acetate (v:v = 1:1) for 30 min and then transferred to pure isoamyl acetate for 1 h. The samples were dehydrated in Hitachi Model HCP-2 critical point dryer (Hitachi, Tokyo, Japan) with liquid CO₂. Finally, the dehydrated samples were coated with gold-palladium and observed in Hitachi Model TM-1000 scanning electron microscope (Hitachi, Tokyo, Japan) at 15 kV.

Fungal mycelial, conidial and germinating conidial samples or Fon- and mock-inoculated watermelon root samples were used for analysis of the FonNot2 expression, whereas mycelial samples collected from 7-day-old cultures were used for analysis of the expression of genes associated with pathogenicity, toxin and chitin biosynthesis and oxidative stress. Total RNA was extracted using Trizol reagent (Invitrogen, Shanghai, China) and treated with RNase-free DNase (TaKaRa, Dalian, China). First-strand cDNA was synthesized from 1 μg of total RNA using AMV reverse transcriptase (TaKaRa, Dalian, China) according to the manufacturer’s recommendation. qRT-PCR was performed with three technical replicates and run on a CFX96 real-time PCR system (BioRad, Hercules, CA, USA). SYBR Premix Ex Taq kit (TaKaRa, Dalian, China) was used to prepare qRT-PCR reaction, which contained 10 μl 2 × SYBR Premix Ex Taq buffer, 1 μg synthesized cDNA and 10 μmol of each of gene-specific primers in a final volume of 20 μl. FonActin was used as an internal control to normalize the data for comparing the relative transcript abundance of the target genes. Relative expression of genes of interest was calculated using the 2^-ΔΔCT method. Gene-specific primers used in qRT-PCR are listed in Supplementary Table S1.

All experiments were repeated independently for three times and each experiment was set with three replicates. Data were statistically analyzed using Student’s t-test at p = 0.05 level.

To identify FonNot2, putative F. oxysporum Not2 gene (FOMG_12536) was identified via BLASTp searching in the F. oxysporum genome database. A 2760 bp genomic fragment containing the FonNot2 gene and a 1545 bp cDNA fragment was amplified and sequenced. Alignment of the genomic and cDNA sequences revealed that FonNot2 contains one intron of 819 bp and that the entire coding sequence of FonNot2 is 1548 bp, encoding a 515 amino acid protein. The FonNot2 protein contains a conserved NOT2/NOT3/NOT5 domain at its C-terminal (Collart and Struhl, 1994) (Figure 1A). Phylogenetic tree analysis indicated that FonNot2 is closely related to F. graminearum FgNOT2, Colletotrichum graminicola CgNOT2, Verticillium dahliae VdNOT2, Magnaporthe oryzae MoNOT2 and Gaeumannomyces graminis var. tritici GgtNOT2 (Figure 1B) and shows 71, 52, 48, 49, and 52% of amino acid sequence identity to the above-mentioned Not2 proteins from different pathogenic fungal species, respectively.

We first analyzed the expression profile of FonNot2 in mycelia, conidia and germinating conidia and during root infection by qRT-PCR. In comparison to the expression in mycelia, the expression of FonNot2 was significantly increased in conidia and showed a further increase in germinating conidia, resulting in >200-fold of increases over that in mycelia (Figure 1C). In mock-inoculated watermelon roots, no transcript of FonNot2 was detected (Figure 1D). Over a 9-day period after Fon inoculation, in planta growth of Fon in roots of the inoculated plants increased gradually, as shown by the ratios of the transcript levels of FonOpm12/ClRps10 (Figure 1D). Similarly, the expression level of FonNot2 increased gradually with the progress of disease development in Fon-inoculated watermelon roots, as shown by the ratios of the transcript levels of FonNot2/FonActin (Figure 1D). The expression level of FonNot2 in inoculated watermelon roots was significantly higher than that in mycelia of axenic cultures (Figures 1C,D), indicating that FonNot2 could be induced during infection process. These data indicate that FonNot2 is highly expressed in conidia and germinating conidia and during plant infection process.

To explore the function of FonNot2, we generated a FonNot2-targeted disruption strain ΔFonNot2 by replacing a 4399 bp fragment containing the entire FonNot2 ORF and partial 5′ and 3′ UTR sequences with the 2694 bp targeted disruption vector containing HPH cassette (Figure 2A). After protoplast transformation with the targeted disruption vector, 21 Hyg-resistant (Hyg^R) transformants were obtained. Seven candidates of the targeted disruption mutants were initially screened from these Hyg^R transformants by PCR amplification of an expected 2015 bp fragment in disruption mutants instead of a 3720 bp fragment in WT strain. Southern blotting of genomic DNA probed with an 805 bp fragment in the 5′ UTR flanking region revealed that the 3.9 Kb KpnI fragment in WT strain was replaced by a 7.7 Kb KpnI fragment in the targeted disruption mutants (Figure 2B), demonstrating the targeted disruption of the FonNot2 gene in the mutants, which was named ΔFonNot2. Because the successful FonNot2-targeted disruption strains had the same phenotype, only one ΔFonNot2 strain was chosen for further study. Complementation of the ΔFonNot2 strain was performed by introducing a 3515 bp DNA fragment encompassing the promoter and entire coding sequence of the FonNot2 gene into the ΔFonNot2 strain. Several neomycin-resistant transformants were obtained and verified by PCR amplification of a fragment identical to that obtained from the WT strain but absent from the ΔFonNot2 strain. Southern blotting of genomic DNA probed with the 805 bp fragment in the 5′ UTR flanking region revealed that, in addition to a 7.7 Kb KpnI fragment identical to that in the ΔFonNot2 strain, the obtained complementation transformant contained an additional KpnI fragment, demonstrating that the complementation of the FonNot2 disruption in the ΔFonNot2 strain, which were named ΔFonNot2-C. qRT-PCR analysis revealed that the expression of FonNot2 in the ΔFonNot2 strain was undetectable while the expression level in the ΔFonNot2-C strain was similar to that in WT strain (Figure 2C), confirming that the obtained ΔFonNot2 strain is a null mutant of FonNot2.

To investigate the function of FonNot2 in fungal growth and development, phenotype of the ΔFonNot2 strain was compared with the WT and ΔFonNot2-C strains. When grown on PDA, the ΔFonNot2 strain grew slowly, by a reduction of 23% in growth rate, and produced less pink pigment, as compared with WT strain (Figures 2D,E). The mycelial growth defect in the ΔFonNot2 strain was restored by complementation with FonNot2 (Figures 2D,E), which showed a growth rate and colony pigmentation similar to that of the WT strain. When grown in MBL, the ΔFonNot2 strain produced significantly less macroconidia, resulting in a reduction of 98% in comparison to that in WT strain (Figure 3A). Microscopic examination revealed that only a few conidiophores were observed in the ΔFonNot2 strain whereas typical conidiophores were normally developed in WT strain (Figure 3B). qRT-PCR analysis indicated that expression levels of FonFGA1 and FonFGB1, encoding the heterotrimeric G protein subunits that are involved in conidiation in F. oxysporum (Jain et al., 2002, 2003), were significantly decreased in the ΔFonNot2 strain, as compared with those in WT strain (Figure 3C). Microscopic examination revealed that the macroconidia of the ΔFonNot2 strain lacked the typical morphology and were shorter, leading to a reduction of 39% in length, than those of the WT strain (Figures 3D,E). CFW staining assays showed that 80% of the macroconidia of the ΔFonNot2 strain had a single septum, and 8 and 12% of the macroconidia had 0 and 2 septa, respectively (Figures 3D,F). In contrast, 93% of the macroconidia from the WT strain had 3 septa (Figures 3D,F). The complementation strain ΔFonNot2-C showed indistinguishable changes in macroconidia morphology and conidiation ability as the WT strain (Figures 3A,B,D–F). The macroconidia produced by the ΔFonNot2 strain were able to germinate with similar germination rate to those of the WT and ΔFonNot2-C strains (Figure 3G). These results indicate that targeted disruption of FonNot2 not only affects the vegetable growth but also influences conidiation and conidial morphology.

To examine the role of FonNot2 in virulence, root infection assays were performed by dipping the roots of 2-week-old seedlings in conidial suspension of the WT, ΔFonNot2 and ΔFonNot2-C strains. In our repeated experiments, the ΔFonNot2-inoculated plants showed significantly reduced disease symptom development (Figure 4A), with 65 and 35% of the inoculated plants showing yellowing symptom on leaves or no obvious symptom at 3 weeks after inoculation (Figure 4B). By contrast, the WT- and ΔFonNot2-C-inoculated plants showed severe disease severity and all of the inoculated plants died at 3 weeks after inoculation (Figures 4A,B). To determine whether targeted disruption of FonNot2 affected in planta fungal growth, we quantified fungal biomass in root and stem tissues by analyzing the transcript levels of Fon FonOpm12 and watermelon ClRps10 genes as an indicator and an internal reference, respectively. The fungal biomass in root and stem tissues of the WT strain-inoculated plants increased during 3–9 days post inoculation (Figures 4C,D); however, the fungal biomass in root and stem tissues of the ΔFonNot2-inoculated plants were significantly reduced at 3, 6, and 9 days post inoculation (Figures 4C,D), showing reduction of 20-fold in roots and 10-fold in stem at 9 days post inoculation. These data suggest that targeted disruption of FonNot2 reduces the virulence of Fon on watermelon and weakens the ability of Fon to grow within watermelon tissues, demonstrating that FonNot2 is required for full virulence toward watermelon.

We examined whether targeted disruption of FonNot2 affected the penetration of Fon on watermelon roots. To test this hypothesis, watermelon roots inoculated with conidia of the WT and ΔFonNot2 strains were examined after 24 h by SEM. In the WT strain-inoculated roots, typical penetration events through openings at the junctions between epidermal cells were seen (Figure 5A, left) and 78% of the germinated conidia showed penetration event (Figure 5B). In the ΔFonNot2-inoculated roots, however, most of the germinated conidia grew on the root surface but failed to penetrate (Figure 5A, right) and only 8% of the germinated conidia showed penetration event (Figure 5B). We further compared the expression levels of some infection-related pathogenicity genes including FonFOW2, FonFMK1, FonMSB2, FonSHO1 and FonFVS1 (Di Pietro et al., 2001; Imazaki et al., 2007; Pérez-Nadales and Di Pietro, 2011, 2015; Iida et al., 2014) in axenic cultures of the WT and ΔFonNot2 strains. qRT-PCR analysis showed that the expression levels of these genes in the ΔFonNot2 strain were significantly reduced in comparison to the levels in the WT strain (Figure 5C). Furthermore, cellophane penetration assays were performed to confirm the defect in penetration ability of the ΔFonNot2 strain. As shown in Figure 5D, the ΔFonNot2 strain was unable to penetrate the cellophane membrane whereas the WT and complementation ΔFonNot2-C strains penetrated the cellophane membranes. These results indicate that targeted disruption of FonNot2 weakens the ability of Fon to penetrate on watermelon roots, demonstrating that FonNot2 contributes to penetration of Fon toward roots of watermelon plants.

FA, a polyketide-derived secondary metabolite, is thought to act as a mycotoxin contributing to severity of F. oxysporum-induced vascular wilt diseases (Brown et al., 2012, 2015). To examine whether targeted disruption of FonNot2 affected the biosynthesis of FA, we compared the FA production in Czapek medium by the WT and ΔFonNot2 strains using HPLC method. Under our experimental condition, FA was detected in filtrates of the WT and ΔFonNot2 strains (Figure 6A). However, the amount of FA produced by the ΔFonNot2 strain was significantly decreased, resulting in a reduction of 94%, compared with that of the WT strain (Figure 6B). To confirm this, we further compared the expression levels of FonFUB1, FonFUB4, FonFUB5, FonFUB6, FonFUB8 and FonFUB10, which are homologues of the FA-biosynthetic genes in F. oxysporum (Brown et al., 2015), in axenic cultures of the WT and ΔFonNot2 strains. As shown in Figure 6C, the expression level of all these FA biosynthetic genes in the ΔFonNot2 strain was markedly reduced, showing less than 10% of the levels in the WT strain. These results indicate that targeted disruption of FonNot2 affects the FA biosynthetic pathway in Fon, suggesting that FonNot2 plays a role in regulation of FA production.

To determine whether FonNot2 has a function in maintaining cell wall integrity, we compared the mycelial growth of the WT, ΔFonNot2 and ΔFonNot2-C strains on PDA supplemented with or without 0.2% CR or 0.2% CFW, which are well-known fungal cell wall perturbing agents. After 7 days incubation, the ΔFonNot2 strain showed an extremely small colony in comparison to that of the WT strain on CR or CFW plates (Figure 7A). The mycelial growth inhibition rates of the ΔFonNot2 strain were increased by 35 and 116% on CR and CFW plates, respectively, as compared with those in the WT strain (Figure 7B), indicating the ΔFonNot2 strain was hypersensitive to cell wall perturbing agents. Considering that chitin is one of the major components in F. oxysporum cell wall (Schoffelmeer et al., 1999) and both of CR and CFW can inhibit fungal cell wall assembly by binding chitin, the chitin contents in the WT, ΔFonNot2 and ΔFonNot2-C strains were also compared. As shown in Figure 7C, the ΔFonNot2 strain showed a significantly reduced chitin content, leading to a reduction of 25%, as compared with those in the WT strains. The sensitivity to cell wall perturbing agents and the chitin content in the ΔFonNot2-C strain were comparable to those of the WT strain (Figures 7A–C). Because chitin synthases (CHS) are the enzymes responsible for chitin synthesis in F. oxysporum (Madrid et al., 2003; Martín-Udíroz et al., 2004; Martín-Urdíroz et al., 2008), we further compared the expression levels of FonCHS1, FonCHS2, FonCHS4, FonCHS5 and FonCHS6 in the WT and ΔFonNot2 strains. As expected, the expression levels of these tested CHS genes in the ΔFonNot2 strain were significantly downregulated, showing less than 10% of the levels in the WT strain (Figure 7D). These results indicate that targeted disruption of FonNot2 affects chitin biosynthesis and that the reduced chitin content in the ΔFonNot2 strain is due to the downregulated expression of the CHS genes, suggesting that FonNot2 plays a role in regulation of chitin biosynthesis in Fon.

It was recently shown that the production of ROS and/or detoxification of ROS are required for virulence in some phytopathogenic fungi including F. oxysporum (Molina and Kahmann, 2007; Chi et al., 2009; Qi et al., 2013). To examine whether FonNot2 has a function in oxidative stress response, we compared the mycelial growth of the WT, ΔFonNot2 and ΔFonNot2-C strains on PDA supplemented with or without 5 mM H₂O₂ or 3 mM paraquat. As shown in Figure 8A, the ΔFonNot2 strain showed an extreme hypersensitivity to exogenous H₂O₂ and paraquat, as compared to the WT strain. The mycelial growth inhibition rates of the ΔFonNot2 strain were increased by 63 and 108% in the presence of H₂O₂ and paraquat, respectively, over those in the WT strain (Figure 8B). Accordingly, the extracellular POD activity in culture filtrates of the ΔFonNot2 strain was decreased, showing 7% of the activity in the WT strain (Figure 8C). The sensitivity to H₂O₂ and paraquat and the extracellular POD activity in the ΔFonNot2-C strain were comparable to those of the WT strain (Figures 8A–C). Considering that POD and peroxidase synthase (PODS) are effective enzymes in the ROS-detoxification systems, we thus compared the expression levels of three FonPODs (FonPOD3, FonPOD4 and FonPOD5) and three FonPODSs (FonPODS1, FonPODS2 and FonPODS3) in axenic cultures of the WT and ΔFonNot2 strains. As expected, the expression levels of the FonPODs and FonPODSs in the ΔFonNot2 strain were significantly downregulated, showing less than 20% of the levels in the WT strain (Figure 8D). These results indicate that targeted disruption of FonNot2 weakens the oxidative stress tolerance and that the hypersensitivity of the ΔFonNot2 strain to oxidative stress is due to the downregulated expression of the genes involved in the ROS-detoxification systems, suggesting a role for FonNot2 in regulation of the degradation of extracellular ROS.

We further investigated whether disruption of FonNot2 affects the production of endogenous ROS by comparing the levels of ROS and superoxide anion in vegetative mycelia and germinating spores of the WT and ΔFonNot2 strains using DCFH-DA and NBT staining, respectively. Results from fluorescence detection and surface plot analysis indicate that the hyphae of the ΔFonNot2 strain produced more ROS than WT (Figure 9A). Similarly, examination of blue formazan precipitate produced from the reduction of NBT by superoxide anion, and quantitative analysis of mean pixel intensity revealed a significant increase in production of superoxide anion in hyphae of the ΔFonNot2 strain than that in WT (Figures 9B,C). Surprisingly, microscopic examination and analyses of mean pixel intensity revealed that the blue formazan precipitate accumulated significantly in germinating spores of WT strain but only less blue formazan precipitate in germinating spores of the ΔFonNot2 strain (Figures 9B,C). This result indicates that, during spore germination, germinating spores and the germ tubes of the ΔFonNot2 strain produced much less superoxide anion than WT spores. Taken together, these data suggest that disruption of FonNot2 has differential effects on endogenous ROS production in vegetative mycelia and germinating spores in Fon.