The effect of silicon on antioxidant metabolism of wheat leaves infected by Pyricularia oryzae

Authors


Abstract

This study investigated whether the increase in wheat resistance to blast, caused by Pyricularia oryzae, potentiated by silicon (Si) is linked to changes in the activity of antioxidative enzymes. Wheat plants (cv. BR 18) were grown in hydroponic culture with either 0 (–Si) or 2 mm (+Si) Si and half of the plants in each group were inoculated with P. oryzae. Blast severity in the +Si plants was 70% lower compared to the −Si plants at 96 h after inoculation (hai). Superoxide dismutase (SOD), catalase (CAT), peroxidase (POX), ascorbate peroxidase (APX) and glutathione-S-transferase (GST) activities were higher in the leaves of the −Si plants compared with the +Si plants at 96 hai. This indicates that other mechanisms may have limited P. oryzae infection in the +Si plants restricting the generation of reactive oxygen species, obviating the need for increased antioxidative enzyme activity. In contrast, glutathione reductase (GR) activity at 96 hai was higher in the +Si plants than in the −Si plants. Although the inoculated plants showed significantly higher concentration of malondialdehyde (MDA) than the non-inoculated plants, lower MDA concentrations were observed in the +Si plants compared with the −Si plants. The lower MDA concentration associated with decreased activities of SOD, CAT, POX, APX and GST, suggest that the amount of reactive oxygen species was lower in the +Si plants. However, GR appears to play a pivotal role in limiting oxidative stress caused by P. oryzae infection in +Si plants.

Introduction

Leaf blast, caused by Pyricularia oryzae (teleomorph: Magnaporthe grisea), emerged in 1986 as a new wheat disease in Brazil that caused considerable losses to crop yields (Igarashi et al., 1986). Today, P. oryzae has spread into all regions of Brazil where wheat is cultivated, and blast is the most prevalent wheat disease in the savannas (Debona et al., 2012). Blast symptoms on leaves begin as grey-green water-soaked lesions that develop dark green borders after they have completely expanded (Reis & Casa, 2005). Elliptical lesions with a grey centre and brown borders can be observed on glumes (Reis & Casa, 2005). On spikes, the symptoms appear as a bright black point on the rachis (Reis & Casa, 2005). Colonization of the rachis by Poryzae decreases nutrient translocation to the grains, which become shrivelled in appearance, small, misshapen and underweight (Goulart & Paiva, 1992).

It has been reported that silicon (Si) can alleviate biotic (e.g. pathogen or insect attack) and abiotic (e.g. excess salt, heavy metals or drought) stresses (Epstein, 2009). With respect to biotic stress, there is a growing body of literature demonstrating the effectiveness of Si for controlling diseases of economically important crops such as barley, corn, cucumber, grape, rice, ryegrass, sorghum and strawberry (Datnoff et al., 2007; Resende et al., 2009). In wheat, it has been shown that resistance to powdery mildew (Bélanger et al., 2003; Rémus-Borel et al., 2005, 2009; Guével et al., 2007), leaf streak (Silva et al., 2010), spot blotch (Domiciano et al., 2010) and leaf blast (Xavier Filha et al., 2011) can be increased by supplying Si to the plants. Although researchers have recognized the importance of Si for increasing host resistance to disease, the mechanisms involved in this phenomenon are unknown. Two hypotheses were proposed to explain increases in plant resistance potentiated by Si. Initially, it was proposed that Si deposition in the form of amorphous silica on plant leaves prevented fungal penetration of epidermal cells (Fauteux et al., 2005). Although this mechanism may partially explain the effect of Si, monomeric Si is also considered biologically active and may trigger faster and more extensive defence responses than polymerized silica (Fauteux et al., 2005). This hypothesis was initially proposed to explain cucumber resistance to Podosphaera xanthii but was later shown to explain disease resistance in grasses. Histological and ultrastructural analyses revealed that the epidermal cells of wheat plants supplied with Si reacted to Blumeria graminis f. sp. tritici infection with specific defence responses, including papilla formation, the production of callose and the release of electrodense osmiophilic material that was identified by cytochemical labelling as glycosylated phenolics (Bélanger et al., 2003). In another study of the same pathosystem, Rémus-Borel et al. (2005) observed the presence of fungitoxic aglycones in Si-treated plants but not in control plants. High performance liquid chromatography analysis revealed that at least three compounds were produced in higher amounts following Si treatment of infected plants. Only one study has shown that Si increases wheat resistance to blast. In that study, the positive effect of Si on wheat resistance was attributed to an increase in the activities of chitinases and peroxidases (Xavier Filha et al., 2011).

Plant cells produce reactive oxygen species (ROS) when infected by several different pathogens (Mandal et al., 2008). Among the most important ROS are hydrogen peroxide (H2O2), the superoxide anion (math formula) and the hydroxyl radical (OH) (Mandal et al., 2008). The enhanced production of ROS causes oxidative damage, leads to lipid peroxidation and damages macromolecules such as pigments, proteins, nucleic acids and lipids (Apel & Hirt, 2004). Considering the drastic damage that ROS can cause to plant cells, enzymatic and non-enzymatic systems are necessary for protection against ROS and the products of secondary reactions (Scandalios, 1993). Enzymes that are known to be involved in protection against ROS and are produced in wheat plants during infection by P. oryzae include superoxide dismutase (SOD), catalase (CAT), peroxidase (POX), ascorbate peroxidase (APX) and glutathione-S-transferase (GST). SOD is involved in the removal of math formula, whereas the other enzymes are associated with the removal of H2O2, phospholipid hydroperoxides and glutathione reductase (GR). GR produces reduced glutathione (GSH), an important compound associated with the reduction of oxidative stress (Debona et al., 2012).

In a previous study, the infection of wheat plants by P. oryzae resulted in oxidative stress, and an increase in the activity of antioxidative enzymes was shown to be important for wheat resistance to leaf blast (Debona et al., 2012). In addition, the positive effect of Si on the antioxidative capacity of plants occurs when the plants are subjected to diverse types of stress (Zhu et al., 2004; Liang et al., 2006; Shi et al., 2010; Mohaghegh et al., 2011; Song et al., 2011; Li et al., 2012; Resende et al., 2012). Accordingly, this study was carried out to investigate whether Si-related wheat resistance to leaf blast was associated with a higher antioxidative capacity in plants.

Materials and methods

Plant material

Wheat seeds of cv. BR 18 were surface sterilized in 10% (v/v) NaOCl for 1·5 min, rinsed in sterilized water for 3 min and germinated on distilled water-soaked germitest paper (Fisher Scientific Co.) in a germination chamber at 25°C for 6 days. Germinated seedlings were transferred to plastic pots with one-half strength nutrient solution without the presence of Si for 1 week. After this period, the plants were transferred to new plastic pots with 5 L of nutrient solution prepared with or without Si. The nutrient solution was prepared based on Clark (1975) with some modifications, and included the following macronutrients and micronutrients: 1·04 mm Ca(NO3)2.4H2O, 1 mm NH4NO3, 0·8 mm KNO3, 0·069 mm KH2PO4, 0·931 mm KCl, 0·6 mm MgSO4.7H2O, 19 μm H3BO3, 2 μm ZnSO4.7H2O, 0·5 μm CuSO4.5H2O, 7 μm MnCl2.4H2O, 0·6 μm Na2MoO4.4H2O, 90 μm FeSO4.7H2O and 90 μm disodium ethylenediaminetetraacetic acid (EDTA). Silicon (2 mm) was supplied to plants in the form of monosilicic acid, which was prepared by passing potassium silicate through a cation exchange resin (Amberlite IR-120B, H+ form; Sigma-Aldrich). The pH of the nutrient solution was 5·6 and was not affected by the addition of monosilicic acid. Four holes were made in the cover of each pot. A styrofoam block (4 cm height, 5 cm diameter) was placed in each cover containing a central hole in which one plant was anchored. The nutrient solution was aerated and was changed every 4 days. Electrical conductivity and the pH of the nutrient solution were checked daily. The pH was maintained at ≈5·5 using NaOH or HCl (1 m) when needed.

Plant inoculation

All the leaves of 60-day-old wheat plants (growth stage 65; Zadoks et al., 1974) were inoculated. A pathogenic isolate of P. oryzae (UFV/DFP-01), which was obtained from the leaves of wheat plants (cv. BR 18), was used to inoculate wheat leaves in this experiment. The isolate was preserved on strips of filter paper placed into glass tubes containing silica gel at 4°C. Pieces of filter paper with fungal mycelia were transferred to Petri dishes containing potato dextrose agar (PDA). After 3 days, PDA plugs containing fungal mycelia were transferred to new Petri dishes containing oat media. These Petri dishes were maintained in a growth chamber at 25°C with a 12 h photoperiod for 10 days. After this period, mycelia producing conidia were carefully removed from the Petri dishes to obtain a suspension of conidia. A total of 25 mL of P. oryzae conidial suspension (105 conidia mL−1) was applied as a fine mist to the adaxial leaf blades of each plant until run-off using a VL Airbrush atomizer (Paasche Airbrush Co.). Gelatin (1%, w/v) was added to the suspension to aid conidial adhesion to the leaf blades. Immediately after inoculation, the plants were transferred to a growth chamber with a temperature of 25 ± 2°C and a relative humidity of 90 ± 5% and were subjected to an initial 24-h dark period. After this period, the plants were transferred to a plastic mist growth chamber (MGC) inside a greenhouse for the duration of the experiment. The MGC was constructed using wood (2 m wide, 1·5 m high and 5 m long) and covered with transparent plastic (100 μm thickness). The temperature inside the MGC ranged from 25 ± 2°C (day) to 20 ± 2°C (night). The relative humidity was maintained at 92 ± 3% using a misting system in which nozzles (model NEB-100; KGF Company) sprayed mist every 30 min above the plant canopy. The relative humidity and temperature were measured with a thermohygrograph (TH-508; Impac). The maximum natural photon flux density at plant canopy height was c. 900 μmol m−2 s−1.

Experimental design

A 2 × 2 factorial experiment, consisting of two Si concentrations (0 or 2 mm, referred to as −Si and +Si plants, respectively) and inoculated and non-inoculated plants, was arranged in a completely randomized design with four replications. Each experimental unit consisted of one plastic pot containing four plants.

Three separate groups of plants were used: one to evaluate blast severity, another for biochemical analysis and a third for analysis of Si concentration. For the evaluation of blast severity, the flag leaves from each main shoot per replication of each treatment were marked with wool yarn and used to evaluate the blast severity at 96 h after inoculation (hai). The diseased leaves were harvested and scanned at 600 dpi resolution. The images were then processed using the software quant (Resende et al., 2009) to determine the severity by the QUANT (SDQ).

Enzyme assays

The enzyme assay methodology was based on that described by Debona et al. (2012). Samples from the flag leaves of each plant (a total of 12 leaves per replication of each treatment) were collected at 0, 48, 72 and 96 hai. The leaf samples were kept in liquid nitrogen during sampling and subsequently stored at −80°C until further analysis. To determine the activities of superoxide dismutase (SOD), catalase (CAT), peroxidase (POX), ascorbate peroxidase (APX) and glutathione-S-transferase (GST), a total of 300 mg of leaf tissue (mix of the 12 leaves collected per replication of each treatment) was ground into a fine powder in a mortar and pestle with liquid nitrogen. The fine powder was homogenized in an ice bath in 2 mL of a solution containing 50 mm potassium phosphate buffer (pH 6·8), 0·1 mm EDTA, 1 mm phenylmethylsulphonyl fluoride (PMSF) and 2% (w/v) polyvinylpolypyrrolidone (PVPP). The homogenate was centrifuged at 12 000 g for 15 min at 4°C, and the supernatant was used as the crude enzyme extract. To determine the glutathione reductase (GR) activity, a total of 300 mg of leaf tissue was ground as described above. The fine powder was homogenized in an ice bath in 2 mL of a solution containing 100 mm potassium phosphate buffer (pH 7·5), 0·1 mm EDTA, 1 mm dl-dithiothreitol, 1 mm PMSF and 2% (w/v) PVPP. The homogenate was centrifuged as described previously.

The SOD activity was determined following the methodology of Del Longo et al. (1993) by measuring the ability of the enzyme to photochemically reduce nitroblue tetrazolium (NBT). The reaction mixture consisted of 50 mm potassium phosphate buffer (pH 7·8), 13 mm methionine, 75 μm NBT, 0·1 mm EDTA and 2 μm riboflavin. The reaction was started after the addition of 60 μL of the crude enzyme extract to 1·94 mL of the reaction mixture. The reaction was allowed to proceed at 25°C under a 15 W lamp. After 10 min of light exposure, the light was blocked and the production of formazan blue, which resulted from the photoreduction of NBT, was monitored using a spectrophotometer (Evolution 60, Thermo Fisher Scientific Inc.) to determine the increase in absorbance at 560 nm according to Giannopolitis & Ries (1977). The control reaction mixtures were kept in darkness for 10 min, and the absorbance was measured at 560 nm. The control values obtained were subtracted from the experimental values obtained. One unit of SOD was defined as the amount of enzyme necessary to inhibit NBT photoreduction by 50% (Beauchamp & Fridovich, 1971).

The CAT activity was determined based on the rate of H2O2 decomposition at 240 nm for 1 min at 25°C (Havir & McHale, 1987). The reaction was initiated after the addition of 50 μL of the crude enzyme extract to 1·95 mL of the reaction mixture containing 50 mm potassium phosphate buffer (pH 6·8) and 20 mm H2O2. An extinction coefficient of 36 m−1 cm−1 (Anderson et al., 1995) was used to calculate the CAT activity.

The POX activity was assayed following the colorimetric determination of pyrogallol oxidation (Kar & Mishra, 1976) after the addition of 15 μL of the crude enzyme extract to 1·98 mL of the substrate mixture containing 25 mm potassium phosphate (pH 6·8), 20 mm pyrogallol and 20 mm H2O2. The POX activity was determined based on the absorbance of coloured purpurogallin recorded at 420 nm for 1 min at 25°C. An extinction coefficient of 2·47 mm−1 cm−1 (Chance & Maehley, 1955) was used to calculate the POX activity.

The methodology proposed by Nakano & Asada (1981) was used to determine APX activity. The reaction was started after the addition of 50 μL of the crude enzyme extract to 1·95 mL of the reaction mixture containing 50 mm potassium phosphate buffer (pH 6·8), 1 mm H2O2 and 0·8 mm ascorbate. The APX activity was measured based on the rate of ascorbate oxidation at 290 nm for 1 min at 25°C. An extinction coefficient of 2·8 mm−1 cm−1 (Nakano & Asada, 1981) was used to calculate the APX activity.

For GST, 150 μL of the crude enzyme extract was added to 1·35 mL of a mixture containing 50 mm potassium phosphate buffer (pH 6·5) and 50 mm reduced glutathione (GSH; Habig et al., 1974). The reaction was initiated after the addition of 500 μL of 30 mm 1-chloro-2,4-dinitrobenzene at 25°C. The absorbance was measured at 340 nm over 3 min. An extinction coefficient of 9·6 mm−1 cm−1 (Habig et al., 1974) was used to determine the GST activity.

The reaction mixture used to determine GR activity contained 100 mm potassium phosphate (pH 7·5), 1 mm EDTA, 1 mm oxidized glutathione (GSSG) and 0·1 mm NADPH prepared in 0·5 mm Tris-HCl buffer (pH 7·5) (Carlberg & Mannervik, 1985). The reaction was started after the addition of 100 μL of the crude enzyme extract to 1·9 mL of the substrate mixture. The decrease in absorbance at 340 nm was determined for 1 min at 30°C. An extinction coefficient of 6·22 mm−1 cm−1 (Foyer & Halliwell, 1976) was used to calculate GR activity.

To determine the activity of each enzyme, four separate extractions were performed on samples from each treatment. Each extraction was read three times. The soluble protein concentrations of the extracts were measured by the method of Bradford (1976) using bovine serum albumin as the standard protein.

Oxidative damage and silicon concentration

Oxidative damage in the leaf cells was estimated as the concentration of total 2-thiobarbituric acid (TBA) reactive substances and expressed as equivalents of MDA according to Cakmak & Horst (1991). Leaf tissue (100 mg) was ground into a fine powder with liquid nitrogen using a mortar and pestle. The fine powder was homogenized in 2 mL of 0·1% (w/v) trichloroacetic acid (TCA) solution in an ice bath. The homogenate was centrifuged at 12 000 g for 15 min at 4°C. After centrifugation, 0·5 mL of the supernatant was reacted with 1·5 mL of TBA solution (0·5% in 20% TCA) for 30 min in a boiling water bath at 95°C. After this period, the reaction was terminated in an ice bath. The samples were centrifuged at 9000 g for 10 min, and the specific absorbance was determined at 532 nm. The nonspecific absorbance was estimated to be 600 nm and subtracted from the specific absorbance value. An extinction coefficient of 155 mm−1 cm−1 (Heath & Packer, 1968) was used to calculate the MDA concentration.

After the termination of the experiment, a total of 50 leaves per replication of each treatment were collected, washed in deionized water, dried for 72 h at 65°C, ground and passed through a 40-mesh screen with a Thomas-Wiley mill (Thomas Scientific). The Si concentration in leaf tissue was determined by colorimetric analysis of 0·1 g of dried and alkali-digested tissue (Korndörfer et al., 2004).

Data analysis

The experiment was repeated once, and the data from all variables were subjected to analysis of variance (anova). Within each sampling time, the means from the −Si and +Si treatments (for non-inoculated or inoculated plants) or the means from the inoculated and non-inoculated plants (for the −Si and +Si treatments), were compared by the t-test ( 0·05) using sas (v. 6.12; SAS Institute, Inc.). For anova, the design was considered to be a 2 × 2 × 4 factorial experiment consisting of two Si concentrations, non-inoculated or inoculated plants and four sampling times (24, 48, 72 and 96 hai). The Pearson linear correlation technique was used to determine the association among the following variables: enzyme activities, concentration of MDA, leaf Si concentration and blast severity. Only data from inoculated plants at 96 hai were used to determine these correlations because this was the only time at which the Si concentration and blast severity were determined.

Results

The leaf Si concentration was significantly increased by Si treatment; the concentration in +Si plants (4·5%) was approximately tenfold greater than that in −Si plants (0·4%). Blast severity was also affected by Si concentration and was significantly lower in the +Si plants (8·8%) compared with the −Si plants (29·5%).

Si concentration, plant inoculation and sampling time were significant for the activities of SOD, CAT, POX, APX and GST (Table 1). However, only plant inoculation was significant for GR activity. Most of the double interactions were significant for the enzyme activities, with the exceptions of Si concentration × plant inoculation (GST and GR), Si concentration × sampling time (CAT and GST) and plant inoculation × sampling time (GR). The Si concentration × plant inoculation × sampling time interaction was significant for the activities of all enzymes except GR.

Table 1. Analysis of variance of the effects of silicon concentration, plant inoculation and sampling time on the activities of superoxide dismutase (SOD), catalase (CAT), peroxidase (POX), ascorbate peroxidase (APX), glutathione-S-transferase (GST) and glutathione reductase (GR), and malondialdehyde concentration (MDA)
Sources of variationd.f.F valuesa
SODCATPOXAPXGSTGRMDA
  1. aLevels of probability: nsnot significant, * = 0·05.

Silicon concentrations (Si)162·76*40·67*104·63*107·08*9·25*0·01ns20·18*
Plant inoculation (PI)1223·03*109·81*1712·09*323·13*34·06*4·93*141·03*
Sampling times (ST)336·12*26·69*135·49*53·85*4·67*1·59ns15·24*
Si × PI128·12*4·89*17·73*35·79*1·23ns0·73ns53·37*
Si × ST37·15*1·57ns10·52*11·76*2·12ns5·21*0·18ns
PI × ST322·12*14·65*162·79*33·75*4·29*1·28ns18·51*
Si × PI × ST38·26*3·34*19·15*19·60*6·57*2·59ns3·97*

The Si concentration, plant inoculation and sampling time were significant for MDA concentration, as were the following interactions: Si concentration × plant inoculation, plant inoculation × sampling time and Si concentration × plant inoculation × sampling time (Table 1).

For the non-inoculated plants, a significant difference in SOD activity was observed between the −Si and +Si plants at 72 hai, with the lower values occurring in the +Si plants (Fig. 1a). For the inoculated plants, SOD activity was significantly lower in the +Si plants compared with the −Si plants at 48, 72 and 96 hai with differences of 31, 37 and 53%, respectively (Fig. 1b). Inoculation with P. oryzae resulted in a significant increase in SOD activity at 48, 72 and 96 hai for the −Si plants and at 72 hai for the +Si plants.

Figure 1.

Activities of: (a,b) superoxide dismutase (SOD), (c,d) catalase (CAT), (e,f) peroxidase (POX), (g,h) ascorbate peroxidase (APX), (i,j) glutathione-S-transferase (GST), and (l,m) glutathione reductase (GR) in the leaves of wheat plants (cv. BR 18) grown in hydroponic culture containing 0 or 2 mm silicon (−Si or +Si, respectively), either inoculated (b, d, f, h, j and m) with Pyricularia oryzae or non-inoculated (a, c, e, g, i and l). The means from the −Si and +Si treatments within each sampling time that are followed by an asterisk (*) are significantly different ( 0·05) by t-test. The means from the non-inoculated and inoculated treatments within each sampling time that are followed by ▼ for the −Si treatment or ▲ for the +Si treatment are significantly different ( 0·05) by t-test. Bars represent the standard errors of the means. Two experiments were conducted with consistent results; the data presented represent the best experiment.

For the non-inoculated plants, there was significant difference in CAT activity between the −Si and +Si plants at 72 hai, with the lower values occurring in the +Si plants (Fig. 1c). The CAT activity in +Si inoculated plants was significantly lower than that in the −Si plants by 26, 15 and 41% at 48, 72 and 96 hai, respectively. CAT activity was significantly higher in inoculated plants than in non-inoculated plants at 48, 72 and 96 hai in the absence of Si and at 48 and 72 hai in the presence of Si.

POX activity in non-inoculated plants was significantly lower in the +Si plants than in the −Si plants at 24 and 72 hai (Fig. 1e). For the −Si inoculated plants, the POX activity was significantly lower than in the +Si plants by 28, 9, 13 and 39% at 24, 48, 72 and 96 hai, respectively (Fig. 1f). POX activity was significantly increased at 24, 48, 72 and 96 hai for both –Si and +Si inoculated plants.

For the non-inoculated plants, APX activity was significantly lower in the +Si plants than in the −Si plants at 24 and 72 hai (Fig. 1g). The APX activity for the +Si inoculated plants was significantly lower by 58, 32 and 33% at 48, 72 and 96 hai, respectively, compared with the −Si plants (Fig. 1h). In the absence of Si, APX activity was significantly higher in the inoculated plants at 48, 72 and 96 hai compared to the non-inoculated plants. In the presence of Si, APX activity increased significantly at 24, 72 and 96 hai in the inoculated plants compared with the non-inoculated plants.

GST activity in the non-inoculated plants was not affected by Si (Fig. 1i). Among the inoculated plants, the GST activity in the +Si plants was significantly (58%) lower than that in the −Si plants at 96 hai (Fig. 1j). Significant differences in GST activity between the inoculated and non-inoculated plants occurred at 72 hai for both −Si and +Si plants and at 96 hai for −Si plants. The higher values occurred in the inoculated plants.

GR activity in non-inoculated plants was not affected by Si treatment (Fig. 1l). Among the inoculated plants, there was a significant difference in GR activity between the −Si and +Si plants at 96 hai, with GR activity being 52% higher in the +Si plants (Fig. 1m). There were no significant differences in GR activity between the non-inoculated and inoculated plants, regardless of Si supply or sampling time.

Regarding the MDA concentration in non-inoculated plants, there was a significant difference between the −Si and +Si plants at 72 hai with the higher MDA values occurring in the +Si plants (Fig. 2a). For the inoculated plants, the MDA concentration was significantly lower in the +Si plants compared with the −Si plants. The MDA concentrations in the +Si plants were 28, 19, 33 and 30% lower than those in the −Si plants at 24, 48, 72 and 96 hai, respectively (Fig. 2b). Significant increases in the MDA concentration occurred at 24, 48, 72 and 96 hai for the −Si plants but only at 96 hai for the +Si inoculated plants.

Figure 2.

Concentration of malondialdehyde (MDA) in the leaves of wheat plants (cv. BR 18) grown in hydroponic culture containing 0 or 2 mm silicon (−Si or +Si, respectively) and non-inoculated (a) or inoculated (b) with Pyricularia oryzae. Means from −Si and +Si treatments within each sampling time that are followed by an asterisk (*) are significantly different ( 0·05) by t-test. The means from the non-inoculated and inoculated treatments within each sampling time that are followed by ▼ for the −Si treatment and ▲ for the +Si treatment are significantly different (≤ 0·05) by t-test. Bars represent the standard errors of the means. Two experiments were conducted with consistent results, and the data presented represent the best experiment.

The activities of SOD, CAT, POX, APX and GST were positively correlated with each other and with MDA concentration and blast severity (Table 2). The activities of SOD, CAT, POX, APX and GST were negatively correlated with GR activity and leaf Si concentration. The correlation between GR activity and MDA concentration and the correlation between GR activity and blast severity were both negative. GR activity was positively correlated with leaf Si concentration. Leaf Si concentration was negatively correlated with both MDA concentration and blast severity. The correlation between MDA concentration and blast severity was positive (Table 2).

Table 2. Pearson correlation coefficients for the activity of superoxide dismutase (SOD), catalase (CAT), peroxidase (POX), ascorbate peroxidase (APX), glutathione-S-transferase (GST) and glutathione reductase (GR); malondialdehyde concentration (MDA); silicon concentration in leaf tissue (Si); and blast severity (Sev) in the leaves of wheat plants (cv. BR 18) 96 h after inoculation with Pyricularia oryzae
VariableaSODCATPOXAPXGSTGRMDASiSev
  1. aLevel of probability: *0·05.

SOD 0·89*0·95*0·92*0·88*−0·81*0·94*−0·94*0·84*
CAT  0·95*0·83*0·94*−0·77*0·85*−0·90*0·84*
POX   0·93*0·98*−0·69*0·91*−0·94*0·84*
APX    0·84*−0·67*0·97*−0·87*0·69*
GST     −0·62*0·82*−0·89*0·83*
GR      −0·75*0·69*−0·59*
MDA       −0·92*0·76*
Si        −0·96*

Discussion

This study provides novel evidence regarding the effect of Si on the antioxidative system in the leaves of wheat plants during P. oryzae infection. In contrast to previously reported studies mainly related to the abiotic stress system (Zhu et al., 2004; Liang et al., 2006; Shi et al., 2010; Song et al., 2011; Li et al., 2012), in the present study, lower levels of antioxidative enzyme activity were observed in the +Si plants compared with the −Si plants upon infection with P. oryzae. This result may be as a result of the activation of other mechanisms that limit fungal infection, thereby reducing oxidative stress.

Debona et al. (2012) found that wheat plants showed a significant increase in the concentrations of math formula and H2O2, two important ROS, when infected by P. oryzae. To prevent oxidative stress, various enzymes act cooperatively to remove ROS (Debona et al., 2012). Of these enzymes, SOD represents the first line of defence, catalysing the dismutation of math formula to H2O2 and O2 (Giannopolitis & Ries, 1977). Similar to what was reported by Debona et al. (2012), the results of this study showed that the SOD activity in wheat leaves was increased in response to P. oryzae infection. However, this increase was more dramatic in the −Si plants compared with the +Si plants. The dramatic increase in SOD activity in the −Si plants may reflect higher rates of math formula generation as the result of massive leaf tissue colonization by P. oryzae.

Similarly, the activities of CAT, POX, APX and GST increased in response to P. oryzae infection; however, these activities were decreased in the leaves of plants supplied with Si. This finding was unexpected because of the significant body of literature showing that increases in host resistance to abiotic and biotic stress potentiated by Si are associated with the stimulation of the host antioxidative system (Zhu et al., 2004; Liang et al., 2006; Shi et al., 2010; Mohaghegh et al., 2011; Song et al., 2011; Li et al., 2012; Resende et al., 2012). For example, Mohaghegh et al. (2011) showed that the activities of CAT and APX increased in the roots of cucumber plants infected by Phytophthora melonis and supplied with Si. Similarly, sorghum plants supplied with Si had higher SOD, CAT, APX and GR activities than plants that were not supplied with Si. In this case, the enzyme activity limited the production of math formula and H2O2 during the process of infection by Colletotrichum sublineolum (Resende et al., 2012).

The limited increase in the activities of SOD, CAT, POX, APX and GST in the leaves of infected +Si plants relative to −Si plants suggests that other mechanisms are acting to enhance plant resistance to leaf blast. Kim et al. (2002) reported that fortification of the cell wall in rice leaf cells was potentiated by Si and associated with resistance to blast. In another study, high Si concentrations in the leaves of wheat plants resulted in a longer incubation period, a smaller number of lesions and a smaller area under the blast progress curve (Xavier Filha et al., 2011). In this case, however, Si also appeared to play an active role in resistance to blast through an increase in the activity of chitinases (Xavier Filha et al., 2011).

In contrast to the reduced SOD, CAT, POX, APX and GST activities in +Si plants relative to their −Si counterparts, the presence of Si was associated with higher GR enzyme activity, particularly during the late stages of P. oryzae infection. GR is a NAD(P)H-dependent enzyme that catalyses the reduction of oxidized glutathione (GSSG) and produces two molecules of reduced glutathione (GSH), a key compound involved in the elimination of oxidative stress (Schaedle & Bassham, 1977; Carlberg & Mannervik, 1985; Kataya & Reumann, 2010; Noctor et al., 2012). It is known that an elevated GSH:GSSG ratio is needed for optimal protein synthesis and that an imbalance in the ratio inhibits protein synthesis by converting a transcription factor to its inactive form (Nagalakshmi & Prasad, 2001). Recently, it was demonstrated that Si alters the leaf proteome of rice plants under stress caused by cadmium (Cd) (Nwuego & Huerta, 2011). In rice plants supplied with Cd, reductions in the production of antioxidative proteins and reductions in physiological variables such as maximum net CO2 assimilation rate, carboxylation efficiency of RuBisCO and quantum efficiency of open PS2 centres in a dark-adapted state were observed (Nwuego & Huerta, 2011). However, in the presence of Si, the oxidative stress preventing the reductions in those variables was limited (Nwuego & Huerta, 2011). In wheat, drought-induced oxidative stress was limited by Si via an increase in the activity of certain antioxidative enzymes, including GR (Gong et al., 2005). As a consequence, Si supply improved the leaf relative water content and water potential under drought (Gong & Chen, 2012). Taken together, these data suggest that the higher GR activity sustained by plants supplied with Si and subjected to stress may help to maintain protein synthesis.

Typically, the activity of GSH is reduced during pathogen infection (Gonnen & Schlösser, 1993; Kuzniak & Sklodowska, 1999) and plants with greater GR activity have elevated levels of GSH in their cells (Foyer et al., 1994). Hernández et al. (2001) observed higher GR activity in a disease-resistant apricot cultivar compared with a susceptible cultivar after inoculation with Plum pox virus, indicating that the resistant plants had a higher capacity for GSH generation. Cultivars of barley and oat that were resistant to B. graminis f. sp. hordei and B. graminis f. sp. avenae, respectively, showed increases in GSH activity in the apoplast, suggesting a possible relationship between GSH enzyme levels and host resistance to powdery mildew (Vanacker et al., 1998a,b). Debona et al. (2012) studied the activities of antioxidative enzymes in wheat cultivars with different degrees of resistance to leaf blast and demonstrated that GR activity increased only in the resistant cultivar. Similarly, in this study, GR activity appeared to be important in wheat resistance to blast in the presence of Si. The importance of GR activity was corroborated by the positive and negative correlations of GR activity with foliar Si concentration and leaf blast severity, respectively. Increases in GR activity may have contributed to maintaining adequate levels of GSH in the leaves of +Si plants. Li et al. (2012) found that muskmelon fruits showed an increase in GR activity in the presence of Si during inoculation with Trichothecium roseum. Consequently, the fruits showed an increase in the concentration of GSH, which reduced oxidative stress during fungal infection.

Non-selective toxins produced by P. oryzae, which are highly cytotoxic to rice (Ou, 1985) and potentially cytotoxic to wheat, may result in lipid peroxidation and greater levels of electrolyte leakage as a consequence (Debona et al., 2012). The data from this study showed that cellular damage, evidenced by high MDA concentrations, increased in response to infection by P. oryzae. The positive correlation between MDA concentration and blast severity supports this result. Cellular damage was sharply reduced in the +Si plants, as evidenced by the negative correlation between MDA concentration and leaf Si concentration. This finding is consistent with the study of Xavier Filha et al. (2011), who observed that cellular damage caused by P. oryzae was reduced in the leaves of wheat plants supplied with Si because of reduced diffusion of non-selective toxins and lytic enzymes.

In conclusion, the lower level of cellular damage in plants infected by P. oryzae and the decreased SOD, CAT, POX, APX and GST activities in wheat plants supplied with Si indicate that ROS generation was limited by Si. However, a high level of GR activity, particularly in the late stages of fungal infection, may play a pivotal role in reducing oxidative stress, potentially by maintaining high levels of GSH.

Acknowledgements

Professor F.A. Rodrigues thanks the National Council for Scientific and Technological Development (CNPq) for his fellowship. D. Debona, J.A. Rios and K.J.T. Nascimento were supported by CNPq. The authors thank Dr Douglas Lau and Dr Márcio Só e Silva (EMBRAPA-Centro Nacional de Pesquisa de Trigo) for providing the wheat seeds. This study was supported by grants from CAPES, CNPq and FAPEMIG to F.A. Rodrigues.

Ancillary