Enhancement of trichothecene mycotoxins of Fusarium oxysporum by ferulic acid aggravates oxidative damage in Rehmannia glutinosa Libosch

Rehmannia glutinosa is an important medicinal herb that cannot be replanted in the same field due to the effects of autotoxic substances. The effects of these substances on R. glutinosa in continuous cropping systems are unknown. In the present study, bioassays revealed that R. glutinosa exhibited severe growth restriction and higher disease indices in the FO+FA (F.oxysporum pretreated with ferulic acid) treatment. The increases in the contents of MDA and H2O2 were greater in the FA+FO treatment than in the FA or FO only treatments, respectively. Consistent with this result, the enzyme activities in the seedlings increased with treatment time. To identify the main factor underlying the increased pathogenicity of FO, macroconidia and trichothecene mycotoxins coproduced by FO were separated and used to treat R. glutinosa seedlings. The MDA and H2O2 contents were similar in the seedlings treated with deoxynivalenol and in the FA+FO treatment. Quantification of the relative expression of certain genes involved in Ca2+ signal transduction pathways suggested that trichothecene mycotoxins play an important role in the increased pathogenicity of FO. In conclusion, FA not only directly enhances oxidative damage in R. glutinosa but also increases wilting symptom outbreaks by promoting the secretion of trichothecene mycotoxins by FO.

. The data are presented as the mean of three independent replicates, with n = 6 plants per replicate. The error bars represent the standard error of the mean. Different letters indicate significant differences between groups according to one-way ANOVA followed by a post-hoc LSD test (P < 0.05).
Scientific RepoRts | 6:33962 | DOI: 10.1038/srep33962 To cope with oxidative damage, enzyme activities in the seedlings increased progressively when treated with FA and FO (with or without FA pretreatment) during the treatment period (days 1-3) compared to untreated seedlings. The SOD activity was higher in the FA-or FO-treated seedlings compared to the control (Fig. 3C), and the SOD activity of the treated seedlings increased with time. The enzyme activities of the seedlings increased with treatment time, and the increases in enzyme activities became significant after 3 days of treatment. The SOD activity of the treated seedlings increased with FO treatment, and the trends of the activities of POD and CAT were similar to those of SOD, which increased with prolonged (days 1-3) exposure (Fig. 3D,E). Notably, the enzyme activities were higher in the seedlings treated with FA+ FO than in the other treatments at 48 h after inoculation, which may indicate a corresponding increase in severe oxidative damage.
FO mycelial growth and macroconidia numbers as well as production of trichothecene mycotoxins by FO are significantly increased after FA treatment. To further determine the link between FA and FO, agar plugs (0.7 cm diameter) of a 7-day-old colony of FO were placed in the center of PDA agar plates containing different concentrations (0 μ mol·L −1 control, 50 μ mol·L −1 , 75 μ mol·L −1 , 100 μ mol·L −1 , 125 μ mol·L −1 , 200 μ mol·L −1 , and 300 μ mol·L −1 ) of FA. After 72 h of incubation at 26 °C in the dark, the colony diameter was estimated. The results showed that FA increased FO mycelial growth at concentrations from 50 μ mol·L −1 to 100 μ mol·L −1 , but a drastic decrease in growth was observed after 3 days when the concentration of FA was greater than Figure 2. FA facilitates the pathogenicity of F. oxysporum on R. glutinosa. F. oxysporum cultured in nutrient solution at 1 × 10 4 macroconidia per ml was treated withincreasing concentrations of FA for 2 days and was then used to inoculate 6-leaf-stage R. glutinosa plants. After 9 days, the plants with root rot and vascular bundle browning were counted and assessed a scale of 0-5 as follows: 0, healthy without any browning; 1, white shoot with scarce browning; 2, light shoot rot and browning; 3, mild shoot rot and browning; 4, severe shoot rot and browning; and 5, death of the whole plant. The statistical enrichment (A) for each treatment were demonstrated through the R Programming Language software. The mean disease index (B) for each treatment was tested. The data are presented as the mean of three independent replicates, with n = 6 plants per replicate. The error bars represent the standard error of the mean. Different letters indicate significant differences between groups according to one-way ANOVA followed by a post-hoc LSD test (P < 0.05). Photographs (C) of infected samples were the root rot symptoms and the vascular bundle browning symptoms which assessed at each scale. 100 μ mol·L −1 (Fig. 4A). To display the effect of FA on the macroconidia production, 100 μ l of spores of FO (10 4 macroconidia per ml) were cultured in liquid potato dextrose medium mixed with different concentrations of FA solution described above. The numbers of sporulation of FO macroconidia in each treatment was measured after incubation using a haemocytometer. Compared to the control (untreated), macroconidia numbers were significantly higher from 50 μ mol·L −1 to 100 μ mol·L −1 FA, with the highest FO macroconidia number in the presence of 100 μ mol·L −1 FA after 3 days (Fig. 4B).
In addition, the production of trichothecene mycotoxin by FO was also evaluated with FA treatment (Fig. 5A). Deoxynivalenol (DON) and 3acetyldeoxynivalenol (3ADON) accumulation were higher after treatment with FA (at concentrations of 50 μ mol·L −1 , 100 μ mol·L −1 , and 200 μ mol·L −1 ) compared to the control (untreated), with the highest production at 100 μ mol·L −1 FA, these results consistent with the fact that concentrations of FA above 100 μ mol·L −1 not only inhibit R. glutinosa growth (Fig. 1), but also FO mycelial growth and sporulation of FO macroconidia (Fig. 4). However, regardless of the concentration of exogenous FA, the quantity of 15-ADON (15-acetyldeoxynivalenol) did not increase.  The sporulation of FO macroconidia in the six treatments was measured by a haemocytometer after incubation at 26 °C in the dark with shaking. Each treatment was performed in triplicate. The error bars represent the standard error of the mean. Different letters indicate significant differences between groups according to a oneway ANOVA followed by a post-hoc LSD test (P < 0.05).
Scientific RepoRts | 6:33962 | DOI: 10.1038/srep33962 We next focused on the dynamics of trichothecene mycotoxin production by FO. Trichothecene mycotoxin were collected and detected 120 h after treatment. DON and 3-ADON production were higher after treatment with FA compared to the control (untreated) and increased at an exponential rate in 48 h. However, DON and 3-ADON production was relatively slow in untreated FO (Fig. 5B,C). DON production by FO pretreated with FA was increased by more than 5.53-fold compared to the control. Similarly, 3-ADON was increased by more than 5.26-fold, up to 528 mg·L −1 . Not surprisingly, production of 15-ADON was not obviously different from that of the control (Fig. 5D).

Trichothecene mycotoxin plays a key role in the pathogenicity of FO.
To determine if trichothecene mycotoxins are the main factor in the pathogenicity of FO macroconidia, we separated the trichothecene mycotoxins from the macroconidia. Spores produced by FO were diluted in a nutrient solution at 10 4 macroconidia per ml after filtering the culture supernatant of FO; the filtrate without spores was also collected. The MDA and H 2 O 2 contents of R. glutinosa seedlings treated with the macroconidia suspension or filtrate were then measured (Fig. 6). Treatment with the filtered liquor led to increases in the MDA and H 2 O 2 contents, consistent with the trend in the contents of R. glutinosa seedlings treated with DON (500 mg·L −1 ). By contrast, there were no obvious increases in seedlings treated with the macroconidia suspension ( Fig. 6), suggesting that the induced oxidative damage in R. glutinosa seedlings was due to increased trichothecene mycotoxins, not the increased mycelial growth or macroconidia production.
We further studied the expression of Ca 2+ sensors in plant signalling, including calmodulin (CaM), calcium-dependent protein kinase (CDPK), calcineurin B-like protein (CaBL), and calmodulin-like protein (CaML), in R. glutinosa seedlings treated with FA and FO. Quantitative real-time PCR revealed similar patterns of mRNA expression in all treatments with differential upregulation of these genes (Fig. 7). Compared to the controls, the transcript levels of CaM were significantly increased after FA or FO treatment by more than 8-to 10-fold. The transcript levels of CDPK exhibited a similar pattern after treatment, with an increase of more than 4-to 8-fold at 24 h after treatment. The expression levels of CaML and CaBL also increased significant and were more than 3-fold higher than in the control plants. Among these treatments, the FA+ FO treatment, the filtered liquor treatment and the DON treatment exhibited similar expression levels of Ca 2+ sensors. These results indicate that DON is the main factor underlying the pathogenicity of FO.

Discussion
Continuous cropping obstacles are common phenomena for various crops, including greenhouse crops 21,24 , trees 25,26 , and, in particular, medicinal plants 27,28 . More than 70% of root-harvested medicinal plants, such as Panax ginseng and American Ginseng, can only be replanted once every 7-20 years due to these obstacles 26,29 . The reasons for these obstacles remain unclear but include autotoxicity of root exudates and microbial community shifts. Greenhouse crops and leguminous crops tend to suffer from devastating fungal diseases or pests 21,24 , and medicinal plants and fruit trees are heavily impacted by autotoxicity issues 28,29 .
Both are considered key factors in the inability to continuously crop R. glutinosa. In the present study, R. glutinosa growth was affected not only by autotoxic substances but also by fungal diseases. Our field observations indicated that seedling growth was always inhibited by autotoxic substances at the early growth stage, whereas serious fungal diseases occurred in the late stage of R. glutinosa growth. Therefore, R. glutinosa provides a model for the study of the problem and its typical characteristics.
In the present study, FA had an inhibitory effect on the growth of R. glutinosa (Fig. 1). MDA and H 2 O 2 , indicators of oxidative damage, were enhanced by FA in the cells of R. glutinosa plants (Fig. 3). Similar increases in MDA and H 2 O 2 have been reported in response to treatment with other phenolic compounds 30,31 . Moreover, FO inoculation, particularly its trichothecene mycotoxin, resulted in slight increases in MDA and H 2 O 2 contents. We thus propose that both FA and FO trigger ROS generation and induce oxidative stress, disrupting cellular membrane structure and restraining seedling growth.
Furthermore, the activity and expression of most antioxidant enzymes are stimulated by ROS accumulation 10,32 . The accumulation of H 2 O 2 in R. glutinosa in response to these treatments enhanced lipid peroxidation and caused  , and calmodulin-like protein (CaML) was measured in the untreated (control) group and groups treated with FA, F. oxysporum, F. oxysporum pretreated with FA, filtered and resuspended macroconidia, filtered liquor or DON solution. Gene expression was profiled by quantitative real-time RT-PCR using the comparative C t method. Each treatment was performed in triplicate. The error bars represent the standard error of the mean. Different letters indicate significant differences between groups according to a one-way ANOVA followed by a post-hoc LSD test (P < 0.05). The statistical analyses were only conducted for a given target gene and that statistical differences between target genes were not considered.
severe oxidative stress, resulting in the disruption of metabolic activity in the cells. Antioxidant enzymes, such as CAT, SOD, and POD, remove accumulated H 2 O 2 . In the present study, increases in the activities of these antioxidant enzymes paralleled the accumulation of MDA and H 2 O 2 in R. glutinosa plants after treatment. Increased POD activity in response to phenolic compounds has also been confirmed in cucumber roots 19,29 .
Autotoxin compounds shape rhizosphere microbiology by deterring or attracting certain microbial species 17,20 . In the present study, application of exogenous FA affected F. oxysporum mycelial growth, conidiophore production and trichothecene mycotoxin accumulation. Thus, FA produced by R. glutinosa not only significantly inhibits the growth of R. glutinosa plants but also significantly increases the incidence of root rot and vascular wilt. Indeed, previous investigations 13 have reported that vascular wilt, a major soil-borne disease in many crops, is also promoted by exposure to autotoxins such as cinnamic acid, FA and 3,4-dihydroxybenzoic acid, suggesting an association of this disease with soil toxicity 8,30,33 .
In particular, our results demonstrate that treatment of R. glutinosa with FO leads to oxidative stress due to the overproduction of ROS. The effect observed is primarily attributable to the trichothecene mycotoxins, since that the possibility that that effect was due to increased mycelial growth was ruled out (Figs 6 and 7).
These results suggest that FA directly and indirectly exerts detrimental effects by triggering oxidative stress and disturbing seedling metabolism. In addition, FA promoted the production of DON by FO and thus led to oxidative damage in R. glutinosa plants via ROS generation, resulting in decreased plant growth and wilting symptoms. In conclusion, our results indicate that FA produced by fibrous roots plays an important role in soil-related disease outbreaks in monoculture systems.
Ca 2+ ions play a vital role as second messengers in plant cells during various developmental processes and in response to environmental stimuli, acting as important sensors of Ca 2+ flux in plants2 3,24 . The oxidative burst is another important component of pathogen defence. The present study, together with previously published data, reveals that the expression of certain Ca 2+ sensor genes was altered in response to all the treatments. Hence, it is reasonable to suggest that Ca 2+ signalling systems may be involved in oxidative damage processes.
Continuous cropping of R. glutinosa is a complicated issue, and further studies are needed to characterize additional signalling compounds involved in the R. glutinosa-F. oxysporum interaction.

Materials and Methods
Plant materials. The plant materials in this paper were micropropagated R. glutinosa (supplement Fig. S1), and the seedlings reached the 6-leaf stage and exhibited healthy roots without any physiological or morphological disorders 34 . Ferulic acid solution preparation. FA dissolved in ethanol was added to the nutrient solution at concentrations of 0 μ mol·L −1 , 50 μ mol·L −1 , 75 μ moL·l −1 , 100 μ mol·L −1 , 125 μ mol·L −1 , 200 μ mol·L −1 , and 300 μ mol·L −1 . The final concentration of ethanol in each solution, including the control, was 0.1% (v/v), a concentration that has a negligible effect on R. glutinosa plants 8 and has no effect on F. oxysporum 35 (Supplement Table 1).
F. oxysporum culture. A single strain of F. oxysporum (coded No. CCS043) was used for the fungal experiments and was isolated from soil under continuous cropping of R. glutinosa in Henan Province in China. The fungus was stored as a macroconidia suspension in 30% glycerol at − 80 °C and was regularly transferred to growth plate cultures on potato dextrose agar medium (PDA) at 26 °C in the dark.
Growth inhibition experiments of R. glutinosa seedlings. Using untreated healthy R. glutinosa micropropagated plants as a control, 3 micropropagated plants were treated with FA (100 μ mol·L −1 ) for 2 days or F. oxysporum for 9 days. Each treatment was replicated 6 times in a completely randomized design.
In addition, the healthy R. glutinosa seedlings were inoculated with F. oxysporum that had been pretreated with FA (100 μ mol·L −1 ). The untreated or treated seedlings were sampled at 24 h, 48 h, and 72 h; the tissue was frozen in liquid nitrogen and stored at − 70 °C until analysis. At the end of the experiment (72 h after treatment), the seedling length and fresh weight were measured 32 .
Pathogenicity test of F. oxysporum pretreated with ferulic acid. F. oxysporum cultured (PDA medium at 26 °C in the dark) in a nutrient solution at 1 × 10 4 macroconidia per ml 36 was treated with FA at concentrations of 0 μ mol·L −1 , 50 μ mol·L −1 , 75 μ moL·l −1 , 100 μ mol·L −1 , 125 μ mol·L −1 , 200 μ mol·L −1 , and 300 μ mol·L −1 for 2 days and was then used to inoculate R. glutinosa micropropagated plants. Each treatment contained 3 plants and was performed in 6 times. The experiment was terminated at 9 days after inoculation, when the FO-inoculated plants exhibited wilting symptoms with yellowing leaves. At the end of the experiment, each plant was harvested to assess root rot and vascular bundle browning on a scale of 0-5, as follows: 0, healthy without any browning; 1, white shoot with scarce browning; 2, light shoot rot and browning; 3, mild shoot rot and browning; 4, severe shoot rot and browning 8 ; and 5, death of the whole plant.

Measurement of MDA and H 2 O 2 . Lipid peroxidation was followed by measuring MDA accumulation
using the method of Zhao et al. with some modifications 37 . Seedling samples (0.2 g) were homogenized in 0.1% trichloroacetic acid in phosphate buffer (5 ml, pH 7.8) and centrifuged at 12,000 rpm for 15 min. The supernatant (1 ml) was added to 0.5% thiobarbituric acid in 20% trichloroacetic acid (4 ml). The mixture was placed in a water bath at 100 °C for 10 min and then quickly cooled in an ice bath for 15 min. The samples were centrifuged at 12,000 rpm for 5 min, and the absorbance of the supernatant was measured at 450, 532, and 600 nm 37 .
H 2 O 2 in the supernatant was measured according to Kang et al. 38 . Seedlings (0.5 g) were homogenized in 5 ml of pre-cooled acetone and centrifuged for 10 min at 1500 × g. Titanium chloride (0.1%, w/v) and concentrated ammonia (0.2 ml) were added to the supernatant (1 ml), and the mixture was allowed to react (10 min at 25 °C).
Scientific RepoRts | 6:33962 | DOI: 10.1038/srep33962 The reaction mixture was then centrifuged at 1500 rpm for 10 min. The absorbance at 410 nm was measured, and the H 2 O 2 concentration was calculated according to a standard curve.
Extraction and assay of enzyme activities. Antioxidant enzymes (SOD, POD, and CAT) were extracted according to the method of Liu et al. 39 and Sofo et al. 40 , with some modifications. Seedling samples (0.5 g) were homogenized in phosphate buffer (8 ml, 0.1 mol·L −1 , pH 7.5) containing 2% (w/v) polyvinylpyrrolidone. The homogenate was centrifuged (12,000 rpm for 20 min), and the supernatant was used for enzyme analysis. All assays were performed at 2-4 °C.
Assessment of the effects of ferulic acid on F. oxysporum growth. The effects of FA on growth were evaluated on PDA medium in Petri dishes (diameter = 9 cm). After the medium was autoclaved, an ethanol solution of FA (filtered through a 0.22-μ m filter membrane) was added, and the medium was thoroughly mixed and poured into Petri dishes 35 ; the final concentrations of FA were 0 μ mol·L −1 , 50 μ mol·L −1 , 75 μ mol·L −1 , 100 μ mol·L −1 , 125 μ mol·L −1 , 200 μ mol·L −1 , and 300 μ mol·L −1 19 .
Individual wells (0.7 cm diameter) were formed in the plates by cutting a square in the centre of the plates. Then, an agar plug (0.7 cm diameter) of a 7-day-old colony of F. oxysporum was placed in the well. After 72 h of incubation at 26 °C in the dark, the colony diameter was estimated (cm). The experiments for each condition were performed in triplicate and were independently replicated 6 times 9,41 .
Assessment of the effects of ferulic acid on macroconidia production. As described by Ling et al. 36 , the sporulation of F. oxysporum macroconidia was induced in liquid potato dextrose culture at 26 °C in the dark. After 10 days, the spores were carefully filtered through two layers of sterile lens paper to eliminate mycelial fragments; Counts were determined in liquid potato dextrose cultures using a haemocytometer. Then, 100 μ l of spores of F. oxysporum (diluted into 1 × 10 4 macroconidia per ml) was cultured in liquid potato dextrose medium mixed with different concentrations of FA solution 42,43 . The final concentrations of FA are described above. The germination assay was performed at 26 °C in the dark with shaking at 140 rpm for 72 h 43 ; There were three replicates for each treatment. The sporulation of F. oxysporum macroconidia in the 6 treatments (5 concentrations of FA and one control at 0 μ mol·L −1 FA) was measured after incubation using a haemocytometer.

Assessment of the effects of ferulic acid on trichothecene mycotoxins. For the hyphal growth
and spore-bearing experiments, 4 treatments were designated as follows: control (0 μ mol·L −1 FA); 50 μ mol·l −1 FA; 100 μ mol·L −1 FA; and 200 μ mol·L −1 FA. For the mycotoxin analyses, F. oxysporum macroconidia were selected. Macroconidia suspensions (10 4 macroconidia per ml) were generated in the liquid potato dextrose medium and were mixed with FA solution in Erlenmeyer flasks 44 . The cultures were incubated at 26 °C in the dark with shaking at 200 rpm. The toxin levels in the treatment are expressed in micrograms of toxin per litre of liquid medium after 5 days of control culture. Each treatment was replicated 6 times in a completely randomized design.
Quantification of trichothecene mycotoxins produced by F. oxysporum. The macroconidia cultures were stopped at the point time and filtered through two layers of sterile lens paper. The DON, 3-ADON, and 15-ADON mycotoxins were extracted from the filtrates using 3-fold volumes of ethyl acetate according to the protocol described by Ponts et al. 18 . The samples were shaken vigorously, sonicated on ice, and allowed to stand for 30 min for phase separation. Thereafter, the organic phase was evaporated to dryness at 70 °C under a nitrogen stream. The dried samples were re-dissolved in 200 μ l of methanol/water (50%, vol/vol) before analysis by HPLC 44 .
To determine the levels of trichothecene mycotoxins produced by F. oxysporum, DON, 3-ADON, and 15-ADON mycotoxins purchased from Sigma (HPLC grade) were used as standards. Quantification was performed with external calibration using standard solutions of DON and 15-ADON prepared from commercial pure powders (Sigma-Aldrich. Saint Louis, Missouri 63103 USA).
According to the protocol described by Boutigny et al. 4 and Ponts et al. 18,44 , the mycotoxins extracted from the different treatments were analysed using a Waters 2695 HPLC system with a 250 × 4.6 mm Luna 5 μ m C18 100 Å column (Phenomenex, Torrance, CA). A photodiode array (PDA) detector was used with an isocratic solvent system [methanol: water-methanol containing 5% (v/v) (90:10) ratio]. The PDA detector measured the UV spectrum (190-500 nm). The samples were dissolved in acetonitrile, and 10 μ l was loaded onto the column using an automatic injector. The mycotoxins were eluted with solvent or the mobile phase at a rate of 0.75 ml·min −1 for 25 min. Standard curves for the respective mycotoxins were generated based on 5 different concentrations of pure toxins, and the absorbances obtained from the HPLC analyses. The mycotoxins produced by the same population in each treatment were sampled over time and replicated 3 times in a completely randomized design.
Autotoxin and trichothecene mycotoxin treatments. When the R. glutinosa seedlings reached the 6-leaf stage with a healthy root, batches of uniform seedlings were transferred to the seedling medium; 3 micropropagated plants were transplanted into glass growth vessels, and each treatment was performed six times. The seedlings were allowed to acclimate to the hydroponic conditions for 7 days.
The medium of F. oxysporum pretreated with 100 μ mol·L −1 FA was extracted to evaluate the effects of FA on trichothecene mycotoxins. Based on previous experiments investigating the effects of FA on trichothecene mycotoxins 29 , we designated 4 treatments: control (0 μ mol·L −1 FA solution); FA (100 μ mol·L −1 FA solution); FO (extract of the F. oxysporum medium without pretreatment with FA); and FO+FA (extract of the F. oxysporum medium pretreated with 100 μ mol·L −1 FA). Five millilitres of each of the solutions and the extracts, which were filtered through a 0.45-mm filter, was added to the robust seedling medium for the experiment, and each treatment was replicated 6 times in a completely randomized design. The seedlings were added with an equivalent volume of distilled water and ethanol as the control. All glass growth vessels were maintained in a tissue culture room at 26 °C with fluorescent lights for 11 h (8:00-20:00), and the fluorescent light intensity was 4.17 ± 0.18 × 10 3 lux 10 .
qRT-PCR analysis was performed using a One-step Quanti-Tect SYBR Green RT-PCR Kit (Qiagen, Shanghai, China). The 18 S rRNA gene was used as the internal control. The primer pairs are listed in Table 1 [45][46][47][48] .
The PCR reactions were performed according to the manufacturer's instructions. The RT-PCR conditions were 1 cycle of 95 °C for 5 min; 40 cycles of 95 °C for 10 sec, 59 °C for 30 sec, and 72 °C for 30 sec; and 1 cycle of 72 °C for 7 min. The data were analysed using the comparative C t method.
Statistical Analysis. All data were subjected to analysis of variance using the Statistical Analysis System Program (SPSS 9.0). Each value was expressed as the mean of 3 replicates ± the standard error (SE). Statistical significance was analysed using Student's t-test and one-way analysis of variance (ANOVA). The precision of the method was confirmed by least-significant difference (LSD, %). The values were considered significant when the P value was < 0.05.