The MfCUT1 gene encoding the major cutinase of Monilinia fructicola, a causal agent of blossom blight and fruit rot of stone fruits, is a virulence factor of the pathogen. The pathogen remains quiescent on stage II fruit, which contain high levels of chlorogenic acid, a quinate ester of caffeic acid (CA). A medium shift system was established to show that MfCUT1 expression is down-regulated by CA, consistent with previous findings in continuous culture, and by the antioxidants glutathione (GSH), N-acetyl-l-cysteine and ascorbic acid. However, MfCUT1 expression is up-regulated by the superoxide-generating oxidant menadione and by the GSH synthesis inhibitor buthionine sulphoximine (BSO). The changes in MfCUT1 transcript levels induced by CA or BSO are related to the levels of intracellular GSH and H2O2. In addition, GSH reductase activity was increased by CA. Monilinia fructicola genes encoding enzymes that may contribute to redox homeostasis (GSH reductase and GSH peroxidase) were cloned and their expression was found to be slightly affected following exposure of cultures to CA. The hybrid tea rose cv. Flaming Peace was identified as an alternative host to complement pathogenicity assays of M. fructicola on stone fruits, which are seasonally restricted. Pathogenicity assays on detached peach fruit and rose petals revealed that CA suppresses and H2O2 enhances formation of brown rot lesions. The results indicate that changes in cellular redox status impact MfCUT1 expression and virulence in M. fructicola, and suggest that redox cycling of GSH is related to this regulation.
The fungal pathogen Monilinia fructicola causes brown rot blossom blight and fruit rot in many species of Prunus (stone fruits) and other members of the Rosaceae family (Batra, 1991). The disease results in severe yield losses and can be a major limiting factor in stone fruit production worldwide (Ogawa et al., 1995). Peach and nectarine fruit are prone to M. fructicola infection during the early stage of rapid expansion of the pericarp (stage I) and then again at harvest maturity when the endocarp and pericarp are fully developed (stage III) (Biggs & Northover, 1988; Luo & Michailides, 2001). Developmentally intermediate stage II fruit, the so-called ‘pit hardening’ stage due to lignification of the endocarp, are considerably less susceptible to infection. If infection occurs during stage II, the pathogen usually remains quiescent until the fruit are mature, when invasion and colonization resumes.
The exocarp of some resistant peach (Prunus persica) cultivars have higher concentrations of chlorogenic acid (5-O-caffeoylquinic acid; CGA) and caffeic acid (3,4-dihydroxycinnamic acid; CA) than susceptible cultivars (Lee & Bostock, 2007). Biochemical analysis of tissue from peach fruit at different stages of development revealed that stage II fruit are enriched in CGA, CA and other phenols such as catechin and epicatechin relative to stage I or stage III fruit, implicating a potential contributing role in arresting quiescent infections during fruit development (Senter & Callahan, 1990; Kubota et al., 2000; Lee & Bostock, 2007). Neither CGA nor CA at concentrations up to 1 mm inhibit M. fructicola growth or conidial germination. However, both compounds markedly suppress the production of cutinase (EC 18.104.22.168) and polygalacturonase (EC 22.214.171.124), enzymes that are probably involved in invasion and colonization of host tissue by the pathogen.
MfCUT1 encodes the major cutinase in M. fructicola (Wang et al., 2002). MfCUT1 over-expressing transformants produce larger lesions on Prunus flower petals than the wildtype strain, providing evidence that MfCUT1 contributes to virulence (Lee et al., 2010). MfCUT1 expression is strongly inhibited by CA and antioxidants, and is enhanced by H2O2 in axenic culture (Lee et al., 2010). CA inhibition of cutinase gene expression and enzyme activity is associated with changes in the electrochemical redox potentials in axenic culture (Wang et al., 2002; Lee & Bostock, 2007). Monilinia fructicola causes smaller lesions on peach petals when inoculated in the presence of CGA, CA or redGSH (Lee & Bostock, 2007). These observations suggest redox-mediated regulation of infection by M. fructicola in stone fruit.
The antioxidant properties of plant phenolic compounds such as CGA and CA have long been recognized. CA and its phenethyl ester have antitumour and anti-inflammatory activity in mammalian cells, and protect human skin cells and monocytes from oxidative damage (Nardini et al., 1998; Neradil et al., 2003). CA and its various esters also contribute to the antioxidant properties and presumed health benefits associated with consumption of wine and other foods containing them. However, CA has also been shown to generate quinone and phenoxy radicals, as well as superoxide (Galati et al., 2002). Thus CA, CGA and other phenolic compounds, particularly under certain conditions (e.g. presence of redox-active transition metals, high pH), can function as pro-oxidants and be cytotoxic (Galati et al., 2002).
CA and CGA influence redox status in systems where they have been studied in part through regulation of glutathione metabolism. CA increases cellular glutathione (GSH; l-γ-glutamyl-l-cysteinylglycine) in animal cells, and both CA and CGA have been shown to increase the activity of glutathione-S-transferase (GST; EC 126.96.36.199) (Nardini et al., 1998). Cellular redox status is indexed by the ratio of reduced GSH to oxidized GSH (GSSG; glutathione disulphide), and cells can undergo oxidative stress as this ratio decreases (Han et al., 2006). Conversion and cycling between GSH and GSSG involve NADPH and the integrated activities of GSH reductase (GR; EC 188.8.131.52), GSH peroxidase (GPx; EC 184.108.40.206), and glucose-6-phosphate dehydrogenase (G6PD; EC 220.127.116.11) (Penninckx, 2000). GPx oxidizes GSH to GSSG, with concomitant detoxification of H2O2. GR converts GSSG back to GSH using G6PD-generated NADPH as electron donor to complete the cycle.
Redox metabolism and reactive oxygen species (ROS) signalling are important for transitions occurring during differentiation, development and parasitism in fungi (Heller & Tudzynski, 2011). However, relatively little is known about how changes in plant redox chemistry, such as occurs during fruit development, influences parasitic activity of plant pathogenic fungi. Previous experiments by the authors were performed in prolonged culture (up to 10 days), and intracellular redox potentials in the cultures could change over time even in the absence of CA. Thus, the mechanisms underlying the inhibitory effect of CA on MfCUT1 expression and other virulence or pathogenicity factors remain unresolved.
To better understand how host phenolic compounds regulate expression of the virulence gene, MfCUT1, medium shift experiments were conducted in which the fungus was initially grown in basal medium and then transferred to medium containing redox-active compounds. The effects of various antioxidants and pro-oxidants on electrochemical redox potentials, GSH metabolism, ROS in the form of H2O2, and MfCUT1 expression were determined in these cultures. Genes encoding enzymes involved in redox cycling of GSH, including GSH reductase and GSH peroxidase from M. fructicola were cloned, and their expression analysed following treatment of cultures with CA. In addition, an alternative host was identified, the hybrid tea rose cv. Flaming Peace, for pathogenicity assays of M. fructicola, which is not as restricted by season as stone fruits. Pathogenicity tests of M. fructicola on peach fruit and rose petals were conducted with and without exogenous redox active chemicals. The results indicate that extracellular redox changes that impact MfCUT1 expression and virulence in M. fructicola are related to redox cycling of glutathione.
Materials and methods
Fungal strain, growth and medium-shift protocol
The MUK-1 strain of M. fructicola used in this study is a single-spore isolate cultured from an infected peach fruit in California (Bostock et al., 1999). The fungus was maintained on V8 juice agar plates at room temperature for sporulation and genomic DNA isolation. Freshly prepared conidial suspensions (2 × 105 mL−1) were transferred into a 9 cm Petri dish with 20 mL starter medium containing 0·1% yeast extract, glucose and Czapek-Dox salts (1 g K2HPO4, 0·5 g MgSO4, 0·5 g KCl, 0·01 g FeSO4 per litre, pH 7·0). After incubating at 23°C for 3 days, fungal mycelium was washed twice with sterile water (2 × 20 mL), mixed with 20 mL Czapek-Dox salt supplemented with 0·1% 16-hydroxyhexadecanoic acid (16HAA) and redox test compounds, and incubated for an additional 24–96 h. Cultures were filtered through a preweighed Whatman filter paper and washed twice with sterile distilled water. The filters were dried at 80°C for 48 h and weighed again to obtain mycelial dry weights, with three replicates for each treatment. For RNA purification and biochemical analysis, mycelia were collected without drying.
Determination of redox potentials of culture filtrates
Intracellular GSH was determined as described previously (Lee & Bostock, 2007). Absolute electrochemical (redox) potentials of culture filtrates were measured with a Mettler InLab redox electrode using a Mettler SevenEasy meter (Mettler) with Zobell's solution (potassium ferro–ferricyanide solution) as a calibration reference.
To assay for total cellular GSH, GSH reductase activity and soluble protein, freshly harvested fungal mycelia were ground into a fine powder in liquid nitrogen and dissolved into 20 mm sodium phosphate buffer (pH 7·4). Supernatants were collected by centrifugation at 22 896 g at 4°C for 15 min. Soluble protein content was determined using the Bio-Rad protein assay kit according to the manufacturer's instructions (Bio-Rad Laboratories). Total cellular GSH was determined as described previously (Lee & Bostock, 2007). GSH reductase activity was measured as described by Emri et al. (1997). The reaction mixture contained 100 μL of 0·2 m potassium phosphate buffer (pH 7·5), 50 μL of 3 mm 5,5′-dithio-bis (2-nitrobenzoic acid) (DTNB), 10 μL of 2 mm NADPH, 15 μL H2O and 15 μL sample and was incubated at 25°C for 3 min. The change in absorbance at 405 nm after addition of 10 μL of 20 mm oxidized glutathione (GSSG) was recorded at 15 s intervals for 5 min using a Labsystems Multiskan RC microplate reader (ThermoFisher Scientific Inc.).
Effect of redox compounds on fungal growth
Redox compounds used in this study include buthionine sulphoximine (BSO), caffeic acid (CA), reduced GSH, N-acetyl-l-cysteine (NAC), ascorbic acid (AA), H2O2, menadione and 1-chloro-2,4-dinitrobenzene (CDNB). CA, CDNB and menadione were dissolved in methanol and added to cultures such that the final concentration of methanol did not exceed 0·4%. Other compounds used in this study were dissolved in water. Fungal conidia (1 × 104 mL−1) were incubated in starter medium containing 0·1% glucose in a 96-well microplate at 23°C for 24 h and then challenged with redox compounds and incubated for an additional 72 h. Each treatment contained at least four replicates. The mock controls were treated with 0·4% methanol or water, as appropriate. Fungal growth was estimated based on the optical density of each well at 550 nm using a microplate reader.
Detection of H2O2
Starter medium containing conidia (1 × 104 mL−1) and 0·5, 1 or 2 mm CA or BSO were placed in a 96-well black plate (Nunclon) with 180 μL per well and four replicates for each treatment, and incubated at 23°C for 24 h. To detect and quantify H2O2, 20 μL of 0·1 mm 2,7-dichlorofluorescin diacetate (DCFH-DA) was added to each well and incubated at 37°C for 30 min. The fluorescence was measured with a multimode reader (Infinite M200; Tecan Group) with excitation wavelength at 485 nm and emission wavelength at 535 nm.
Gene cloning and analysis
The genes encoding GSH reductase (MfGR1) and GSH peroxidase (MfGPx1) were obtained from genomic DNA of M. fructicola by PCR with degenerate primers (Table 1). Primers were designed based on highly conserved regions of the genes found in Sclerotinia sclerotiorum and Botryotinia fuckeliana. The PCR products were cloned into a TOPO TA cloning vector (Invitrogen) and sequenced from both directions by an ABI 3730 DNA sequencer (Applied Biosystems). The similarity of the amplified fragment was determined by searching against the databases at the National Center for Biotechnology Information (NCBI) using the blastx program. The deduced amino acids of each gene were aligned with published sequences using clustalW. Functional domains were predicted with the ScanProsite program available at http://prosite.expasy.org/scanprosite. Sequence data from this article can be found in the EMBL/GenBank data libraries (MfGR1 accession no. JQ837262; MfGPx1 accession no. JQ837263).
Table 1. Primers used in this study
RNA extraction, real-time quantitative reverse transcriptase PCR (qRT-PCR) and RNA blot analysis
Mycelia from three replicates were mixed for RNA extraction. RNA was purified and precipitated with LiCl as previously described (Lee et al., 2010), treated with DNase and used to synthesize the first strand of cDNA with the SuperScript II reverse transcriptase (Invitrogen) and oligo-dT (5′-(T)25VN-3′). Primers (Table 1) used for quantitative PCR (qPCR) were designed using the Roche online design tool (http://www.roche-applied-science.com/sis/rtpcr/upl/index.jsp?id=UP030000). The qPCR was conducted with a Rotor-Gene Q PCR machine (QIAGEN) using HOT FIREPol EvaGreen qPCR Mix Plus kit (Solis BioDyne). The cycling profile for amplification was as follows: 95°C for 15 min, immediately followed by 40 cycles of 95°C for 15 s, 58°C for 20 s and 72°C for 20 s. Amplification of specific transcripts was confirmed by melting curve and agarose gel electrophoreses with ethidium bromide staining. The β-tubulin gene was used as internal control. The expression levels of each MfPG1 transcript were quantified relative to the expression level of β-tubulin using the comparative Ct method (Schmittgen & Livak, 2008). The qRT-PCR data presented here are from one of three independent experiments with similar results. Each experiment contains three replicates for each treatment, with three PCR replicates for each gene in each qPCR reaction.
For RNA blot analysis, 20 μg RNA was separated and denatured in a formaldehyde-containing agarose gel as described (Lee et al., 2010). RNA was blotted onto a nylon membrane and hybridized with a digoxigenin (Dig)-labelled DNA probe (500 bp) at 55°C. DNA probes were amplified and labelled by PCR and detected with the DIG high prime DNA labelling and detection starter kit II (Roche) following the manufacturer's instructions. Post-hybridization washes also were performed at 55°C. The density of each band was analysed with id image analysis software (Eastman Kodak Co.).
Fungal pathogenicity was assessed on detached rose (Rosa cv. Flaming Peace) petals and peach (Prunus persica) fruits, collected from local farms (Taichung, Taiwan). Due to seasonal unavailability of peach fruit, flower petals of Rosa cv. Flaming Peace, which are also highly susceptible to M. fructicola, were used for disease assays. Fungal conidia were collected from V8 juice agar plates by flooding with distilled water and low speed centrifugation. Five microlitres of a suspension containing 2 × 105 conidia mL−1 were placed on the surface of rose petals at each inoculation site. To study the effect of redox compounds on pathogenicity of M. fructicola, peach fruits were surface sterilized before inoculation by dipping in 0·1% Tween 20 for 10 s, 10% bleach for 25 s, and rinsing twice with reverse osmosis purified water. After air-drying on a lab bench for 8–10 h, fruit were wounded with a sterile 1000 μL pipette tip to generate a 1 cm depth wound before inoculation. The fruit were inoculated with 5 μL conidial suspension (1 or 2 × 105 mL−1) with and without a test compound in the droplet on each side of a fruit. The mock control was treated with 0·4% methanol or water only. The inoculated samples were incubated in a moist chamber at room temperature (c. 25°C) for lesion development.
Statistical analysis and replication
All experiments were performed at least twice, with at least three replicates for each treatment. All data presented and discussed in this study represent at least two independent experiments with similar results. The significance of differences was determined by paired t-test, regression or one-way anova using spss software, v. 10 (SPSS Inc.).
The effect of antioxidants on MfCUT1 gene expression
MfCUT1 gene expression was assessed by RNA gel blot and semiquantitative (sq) RT-PCR following a shift of M. fructicola hyphae from a modified Czapek-Dox medium after 3 days of growth to a medium containing the cutinase inducer 16-hydroxyhexadecanoic acid (16HAA) and CA or other redox-active compounds. The cultures were incubated for an additional 24 h after the shift and then extracted for RNA. Controls included cultures shifted to media with the corresponding solvent diluted to the same concentration used to deliver each test compound. CA, reduced GSH, N-acetyl-l-cysteine (NAC) and l-ascorbic acid had little or no effect on fungal growth at all tested concentrations (data not shown). MfCUT1 expression relative to the β-tubulin gene, TUB2, by sqRT-PCR decreased after a shift to 1 mm CA (data not shown). Time course studies by RNA gel blot also confirmed the inhibitory effect of CA on MfCUT1 expression (Fig. 1). Exposure of M. fructicola to GSH or NAC (at 1 or 0·5 mm, respectively) slightly increased expression of MfCUT1 (Fig. 2a). However, MfCUT1 transcript accumulation declined in medium containing 4 mm GSH, 2 mm NAC or ascorbic acid at 5 and 10 mm. Although all antioxidants reduced the extracellular redox potential of the cultures, the changes in the level of MfCUT1 transcripts did not directly correspond to the relative reduction in redox potential (Fig. 2a).
The effect of pro-oxidants on MfCUT1 gene expression
MfCUT1 expression increased when M. fructicola was shifted to 16HAA-medium containing the superoxide generator menadione (2-methyl-1,4-naphthoquinone; Jamieson, 1992) at 10–50 μm, with three replicates for each treatment (Fig. 2b). When M. fructicola was shifted to 16HAA-medium containing 1-chloro-2,4-dinitrobenzene (CDNB), a GST substrate that can deplete cellular GSH levels, the accumulation of MfCUT1 transcripts was slightly repressed, as evidenced in both RNA gel blots (Fig. 2b). Fungal growth was not affected by menadione at concentrations up to 50 μm. However, M. fructicola is highly sensitive to CDNB. Mycelial degradation was observed when cultures were incubated with 100 μm CDNB.
The effect of caffeic acid on intracellular redox status in M. fructicola
CA decreased extracellular redox potentials in medium-shift experiments (Fig. 3a). In the mock control treated with 0·4% methanol (diluent for CA), electrochemical redox potentials of the culture filtrates increased with fungal growth. Upon exposure to CA, the redox potential of the culture was reduced by 11% (285–254 mV) at 24 h, and then recovered to a level similar to the control by 48 h. In contrast, CA increased the total intracellular GSH levels relative to the control by 28% at 24 h (P =0·084, Tukey's test) and 268% at 96 h (P =0·004, Tukey's test; Fig. 3b). The GSSG levels varied among samples, excluding precise quantitation of the ratios between oxidized (GSSG) and reduced forms of GSH. However, to further evaluate cellular redox state, GSH reductase activity was measured (Fig. 3c). Activity of GSH reductase, which converts GSSG to reduced GSH, increased when cultures were shifted to 16HAA-medium containing CA. The level of H2O2 within hyphae, evaluated with the redox-sensitive fluorescent probe DCFH-DA, was reduced following addition of CA to the cultures relative to the control. The fluorescence intensity of DCFH-DA changed in a CA concentration-dependent manner, with CA reducing the DCFH-DA fluorescence intensity by 27% at 1 mm and by 35% at 2 mm (slope = −2265/mm, P <0·0001, R2 = 0·80; Fig. 3d).
The effect of a GSH biosynthesis inhibitor on MfCUT1 expression
Buthionine sulphoximine (BSO) irreversibly inhibits γ-glutamylcysteine synthase (γGCS), an enzyme required for the biosynthesis of GSH (Mehdi & Penninckx, 1997). When M. fructicola was shifted to 16HAA-medium containing BSO (at 0·5, 1 and 2 mm), the level of total GSH declined (Fig. 4a). BSO treatment did not significantly increase intracellular H2O2 accumulation in fungal hyphae (Fig. 4b), but increased MfCUT1 gene expression (Fig. 4c). However, MfCUT1 transcript levels decreased following a shift to medium containing both CA and BSO. Application of CA and BSO, either separately or in combination, reduced electrochemical redox potentials of the cultures (Fig. 4d). BSO did not inhibit fungal growth at the concentrations tested.
Expression of M. fructicola genes involved in GSH redox cycling
To assess further how CA influences redox status in M. fructicola, the expression of two genes involved in redox cycling of GSH was evaluated. Partial sequences corresponding to GSH reductase (MfGR1) and GSH peroxidase (MfGPx1) were obtained by PCR with degenerate primers. A 1420 bp MfGR1 DNA fragment and a 480 bp MfGPx1 DNA fragment were independently amplified. Sequence analysis and similarity searches against fungal sequences deposited in NCBI verified their identities (also see experimental procedures). The deduced partial amino acid sequence of MfGR1 contains a pyridine nucleotide–disulphide oxidoreductase class-I site at its N-terminus predicted with ScanProsite, which is highly similar to BCGR1 (XM_001546364.1) of B. fuckeliana and SSGR1 (XM_001588227.1) of S. sclerotiorum (Fig. S1). A GSH peroxidase active site and a GSH peroxidase signature 2 were identified in this partial sequence of MfGPx1 with ScanProsite (Fig. S2). The expression of MfGR1 and MfGPx1 was analysed by qRT-PCR after exposing mycelium to CA for 24 h. The β-tubulin gene (Tub2) was used as the internal control for the comparative Ct method assay because the growth of M. fructicola was not significantly affected by CA, suggesting β-tubulin expression was not affected by CA (this study; Lee & Bostock, 2007). As shown in Figure 5, the expression of MfCUT1 was inhibited by CA, which confirms the results observed by RNA gel blot (Fig. 1). However, MfGR1 and MfGPx1 gene expression was slightly up-regulated by CA (Fig. 5).
The effect of CA and H2O2 on virulence of M. fructicola
Pathogenicity assays on detached peach fruit revealed that brown rot lesions were significantly greater when H2O2 at 1 mm was applied with M. fructicola inoculum onto the fruit surface of both mature (P =0·009; one-way anova) and immature (P <0·0001; paired t-test) peaches (Table 2). In contrast, inoculation of M. fructicola with 1 mm CA resulted in smaller brown rot lesions on mature peach fruit, compared to those inoculated with the fungus with 0·4% methanol. When tested on immature peach fruit, the addition of CA in the infection droplet had no effect on brown rot lesion formation. H2O2 at 1 mm in the infection droplet enhanced lesion expansion on peach fruit. However, H2O2 at 0·1 mm (data not shown) or 0·5 mm did not have a significant effect on lesion expansion on peach fruit. H2O2 (0·1, 0·5 or 1 mm) or CA at 1 mm did not cause any adverse effects on conidial germination or appressorium formation in wounded peach fruit (data not shown).
Table 2. Effect of caffeic acid (CA) or H2O2 on brown rot lesion development caused by Monilinia fructicola in peach fruit and rose petals
Peach flower petals were available seasonally, thus limiting their use for disease assays. It was recently found that after screening 34 rose cultivars, the flower petals of the hybrid tea rose Rosa cv. Flaming Peace are highly susceptible to M. fructicola. Monilinia fructicola inoculated onto fully opened petals of this cultivar induced brown rot lesions within 24–48 h post-inoculation (hpi; Fig. 6a), resembling those induced on peach petals. Brown rot lesions expanded until the entire petal became necrotic at 96 hpi. Monilinia fructicola was apparently able to produce appressoria, penetrate host tissue and eventually produce sporodochia within the lesions (Fig. 6b). Monilinia fructicola produced significantly smaller brown rot lesions on the rose petals when inoculated in the presence of 1 mm CA or H2O2, compared to those produced without addition of CA or H2O2 (Table 2; Fig. 6a). However, H2O2 at 0·1 mm in the infection droplet promoted lesion expansion on rose petals, similar to the effect observed on peach fruit treated with 1 mm H2O2. Microscopic examination of the rose petals revealed that 1 mm H2O2 attenuated conidial germination and hyphal elongation, whereas H2O2 tested at 0·1 mm had no significant effects on the growth of M. fructicola in rose petals (Fig. 7). Therefore, inhibition of lesion expansion on rose petals in the presence of 1 mm H2O2 is probably due to the inhibition of fungal germination and growth. CA (1 mm) or H2O2 (0·1, 0·5 or 1 mm) alone did not cause any microscopically visible lesions on either rose petals or peach fruit (data not shown).
Preformed chemical factors and cuticle architecture contribute to differences in disease resistance during fruit development and maturation in various host species (Prusky, 1996). Similarly, CA and CGA have been proposed as resistance factors to brown rot disease in developing peach fruit (Bostock et al., 1999). However, in contrast to classical preformed factors that may directly inhibit the pathogen, this study reveals that CA, CGA and related phenols may have a more subtle, indirect action in the Prunus–Monilinia interaction by modulating cellular redox to suppress expression of virulence and pathogenicity factors. The study provides further evidence that CA regulates fungal virulence through regulating cellular redox. The activity of glutathione reductase, which can help maintain a high ratio of GSH to GSSG in the cell, was increased by CA, while endogenous H2O2 levels were decreased by CA. MfCUT1 expression was also affected by the addition of antioxidants and pro-oxidants to cultures. Two redox homeostasis enzymes were cloned and their expression was shown to be regulated by CA.
The limited seasonal availability of host materials is a major impediment for researchers working on the interaction of M. fructicola with Prunus and other rosaceous host species. Here, 34 rose cultivars were screened and it was found that flower petals from several cultivars are susceptible to M. fructicola. Although rose petals are not as susceptible to M. fructicola as petals from natural hosts such as almond and peach (Lee & Bostock, 2006), perhaps because of the thicker cuticle of rose petals, Rosa cv. Flaming Peace presents highly reproducible brown rot symptoms for quantitative disease assays. In addition, immature rose petals are more resistant to M. fructicola infection than mature petals (data not shown) and could provide an additional system for studies of age-related changes in host resistance and susceptibility for this interaction. Monilinia fructicola also formed appressoria on rose petals, similar to that observed previously on peach petals (Lee & Bostock, 2006). The formation of small sporodochia was also observed on rose petals (Fig. 6b), similar to those that form on twig cankers and other tissues of infected Prunus species (Batra, 1991).
GSH functions as an important, and often the most abundant, redox buffer in many organisms, and can serve as a sulphur reservoir and a modulator of redox-regulated signal transduction (Meister & Anderson, 1983; Cnubben et al., 2001). GSH-associated metabolism is an important mechanism for protecting cells against ROS. Cellular GSH levels are tightly regulated, but often increase rapidly under oxidative stress or attack by microorganisms (Cnubben et al., 2001; Foyer & Noctor, 2011; Heller & Tudzynski, 2011). Reduced GSH is synthesized from glutamate, cysteine and glycine by γ-glutamyl cysteine synthase (γGCS; EC 18.104.22.168) and glutathione synthase (GS; EC 22.214.171.124). NAC and ascorbic acid are also well known cellular antioxidants. Both reduced GSH, and NAC can be converted to cysteine, a precursor for GSH biosynthesis (Whillier et al., 2009). Similar to CA, GSH at 4 mm, NAC at 2 mm and ascorbic acid at 5–10 mm suppressed MfCUT1 transcript accumulation (Fig. 2a). GSH and ascorbic acid may function in an integrated manner to detoxify H2O2, and as total GSH content increases there could be a concomitant decline in H2O2 levels (Foyer & Noctor, 2011).
There are five putative binding sites for the redox-sensitive transcription factor activator protein-1 (AP-1), and these are located at positions −691 to −243 in the MfCUT1 promoter (Lee et al., 2010). It is tempting to speculate that these putative regulatory sites contribute to MfCUT1 regulation by changes in cellular redox status (see below). In contrast, the superoxide-generating agent menadione, the GSH synthesis inhibitor BSO and H2O2, all inducers of oxidative stress, enhanced MfCUT1 transcript accumulation (this study; Lee et al., 2010). However, CDNB, which acts very differently to menadione and H2O2 to induce the oxidative stress response in yeast (Mutoh et al., 2005), had little or no effect on MfCUT1 expression. The data also indicate that changes in extracellular redox potentials following addition of antioxidants do not directly correspond to changes in the intracellular redox status and MfCUT1 transcript levels.
The present study suggests that redox-cycling of GSH occurs following treatment of cultures with antioxidants and pro-oxidants that alter MfCUT1 expression and virulence in M. fructicola. The addition of CA to cultures initially reduced electrochemical redox potentials and resulted in an elevation of total intracellular GSH, a strong induction of GSH reductase activity, a decline in H2O2 levels in hyphae, and a modest increase in transcript levels of MfGR1 and MfGPx1. The combined activities of GSH reductase and GSH peroxidase provide a mechanism to cycle GSH between reduced and oxidized states. CA decreases redox potentials, probably via increasing GSH reductase activity to support balanced GSH levels. Gene expression reveals that MfGR1 transcripts were slightly up-regulated by CA, which is consistent with the enzymatic activity. blastx search using the compiled sequence of MfGR1 as a query identified two different GSH reductase-coding genes in the sequenced genomes of B. fuckeliana and S. sclerotiorum. In contrast, only one copy of each of the genes encoding GSH peroxidase was identified. Redox cycling with corresponding changes in GSH and GSSG pools in M. fructicola can probably be inferred from the collective results above and from the inverse effects of antioxidants and pro-oxidants on the responses measured in this and previous studies (Lee & Bostock, 2007). However, the determinative role of MfGR1 and MfGPx1 in the redox regulation of virulence in M. fructicola is unresolved and requires further study.
Oxidants such as H2O2 readily interact with cellular proteins and lipids. In the M. fructicola–rose petal interaction, low concentrations of H2O2 (0·1 mm) inhibited lesion expansion but did not affect fungal growth and development, suggesting that the expression of virulence factors, such as MfCUT1, is up-regulated in the presence of ROS during infection. However, in wounded peach fruit, the addition of low concentrations of H2O2 (0·1 or 0·5 mm) at the infection site did not affect M. fructicola growth and virulence. The absence of a discernible effect on the pathogen and disease phenotype might be explained by the instability and rapid dissipation of H2O2 in the presence of competing host reactions at the wound site (e.g. peroxidase, catalase). Using a similar line of reasoning, the lack of an effect of CA in the infection droplet on lesion formation in immature (stage II) peach fruit might be explained by the presence of already high levels of phenols in the pericarp of these fruit that mask any further incremental effects of added CA (Lee & Bostock, 2007).
Redox control in fungal development and pathogenicity is an emerging area of research in plant–microbe interactions (Heller & Tudzynski, 2011), with implications for disease management in pre- and post-harvest settings. NADPH oxidase, which catalyses superoxide production, is required for pathogenicity in Magnaporthe oryzae (Egan et al., 2007). Sclerotium formation of S. sclerotiorum, Sclerotium rolfsii and Rhizoctonia solani is induced by oxidative stress (Patsoukis & Georgiou, 2008). In addition, γGT catalysing the transfer of a γ-glutamyl moiety of GSH to amino acids and other compounds (Meister & Anderson, 1983) was recently reported to be required for maintaining total GSH levels and for development of sclerotia and compound appressoria in S. sclerotiorum (Li et al., 2011). Sclerotinia sclerotiorum also produces oxalic acid during infection, in part, to manipulate the host redox environment (Cessna et al., 2000; Williams et al., 2011). The yeast AP-1-like transcription factor (YAP1) is activated under oxidative stress and has a crucial role in fungal development and pathogenicity in many fungal pathogens (Molina & Kahmann, 2007; Guo et al., 2011). Studies of citrus pathotypes of Alternaria alternata have revealed that detoxification of host-generated ROS by the pathogen is mediated by redox activation of the transcription regulator YAP1, which is required for fungal pathogenicity and lesion formation (Lin et al., 2009, 2011; Yang et al., 2009). A YAP1 gene (MfAP1) from M. fructicola has been cloned and its expression was found to be regulated by CA and H2O2 (P.-L. Yu, C.-M. Chiu, M.-H. Lee, National Chung–Hsing University, Taiwan, unpublished data). Functional analysis of MfAP1 in pathogenicity and fungal development is currently underway.
The authors thank Drs W.-H. Ko, F.-J. Jan, W.-L. Deng (NCHU), and S.-J. Lee (National Health Research Institutes, Taiwan) for their helpful comments and suggestions. Research was supported by grants (96-2311-B-005-003, 97-2313-B-005-034-MY3 and NSC 101-2911-I-005-301) from the National Science Council of Taiwan to MHL.