Introduction

Oxidative stress emerges as a result of unfavorable environmental conditions leading to generation of reactive oxygen species (ROS), such as superoxide radicals (O2 −), hydroxyl radicals (HO), and hydrogen peroxide (H2O2), which are highly detrimental to the cells. Similarly, oxidative stress occurs in response to pathogen attack and during the natural processes of senescence. It is a widespread phenomenon and leads to undesirable events in plants such as reductions in growth, yield and crop quality.

ROS are highly reactive towards membrane lipids, proteins and nucleic acids. They are key factors causing rapid cell damage under stress (Hariyadi and Parkin 1993; O’Kane et al. 1996; Prasad 1996), but also play a signaling role interrelating different stimuli and respective responses (Desikan et al. 2004). Among all ROS, hydrogen peroxide is the most appropriate molecule for this function because of its longer half-life and higher stability.

It is known that the accumulation of hydrogen peroxide is an early, common response to various stress factors, such as pathogen attack, wounding, drought, extreme temperatures, physical and chemical shock, UV-irradiation (Doke et al. 1994; Foyer and Noctor 2000). A stress environment leads to rapid synthesis of H2O2 in chloroplasts and other cell organelles and in the apoplast (Foyer et al. 1997).

The generation of ROS and especially H2O2 is acknowledged as a signal for activation of plant defense mechanisms under biotic and abiotic stress (Doke et al. 1994; Prasad et al. 1994; Foyer et al. 1997). It is assumed that the different forms of ROS interact with molecular targets, which are sensitive to changes in the ROS concentration within and outside the cells and thus transform this information into alterations in gene expression (Laloi et al. 2004).

It was found that application of H2O2 at low concentrations could induce stress tolerance in plants. The preliminary H2O2 treatment of Arabidopsis or tobacco protects the plants from oxidative damages due to high light intensity (Karpinski et al. 1999; Gechev et al. 2002). Tolerance to low positive temperatures was demonstrated after treatment with low concentrations of H2O2 in maize seedlings, Phalaenopsis and Vigna radiata (Prasad et al. 1994; Yu et al. 2002, 2003), and similarly treated potato nodal explants were found to be resistant to high temperature (Foyer et al. 1997; Lopez-Delgado et al. 1998). In these cases it was established that the pre-treatment with H2O2 causes alterations in the activity of several antioxidant enzymes and/or the levels of some antioxidants such as glutathione.

Paraquat is a bipyridilium contact herbicide, which by diverting electrons from Photosystem I forms bipyridilium radicals. The bipyridilium radicals, by reaction with oxygen produce superoxide radicals, and this is followed by chain reactions leading to the generation of other reactive forms of oxygen. Paraquat inhibits CO2 fixation in plants, but its rapid herbicidal action is attributed mainly to generation of ROS (Harvey and Fraser 1980), and thus the biomembranes are subjected to free radical attack (Farrington et al. 1973). The double-bonds of biomembrane unsaturated fatty acids are the primary targets for ROS, which initiates chain reactions of lipid peroxidation followed by destruction and disorder in the normal membrane permeability (Szigetti et al. 2001; Cobb 1992). Numerous hypotheses have been evolved on possible reasons for PQ resistance. Inhibited translocation, binding of PQ to the cell wall, its sequestration to vacuoles, or enhanced activity of oxygen radical detoxifying enzymes (Norman et al. 1993; Hart and DiTomaso 1994; Yu et al. 2004) were proposed as possible mechanisms of resistance.

In our previous investigation (Moskova et al. 2007), we established that pre-treatment of young pea plants with low concentrations of hydrogen peroxide decreased the damage caused by paraquat.

Here we report the changes in some parameters related to paraquat toxicity and the alterations of the antioxidant system in H2O2 pre-treated plants.

Materials and methods

Seedlings of Pisum sativum L. were grown on a half-strength Hoagland-Arnon nutrition medium in a growth chamber under the following conditions-12/12 h photoperiod; light intensity 70 μmol m2 s−1; temperature 25 ± 1°C.

Ten-day-old seedlings were sprayed with an aqueous solution of H2O2 (2.5 mM) containing 0.1% Tween 80, and 24 h later, part of the plants were sprayed with 0.2 mM aqueous solution of paraquat containing 0.1% Tween 80. Hydrogen peroxide- or PQ-non-treated plants (as well as the controls), were sprayed with water containing 0.1% Tween 80 at each chemical application stage. Paraquat treatment was carried out after the end of the light phase in the photoperiod cycle, thus giving a 12 h of incubation period during which the herbicide effects are usually weak (Chang and Kao 1997).

Samples for measurement were taken at the end of the dark phase (0 h), 2 h after the beginning of illumination when there were still no visible herbicide effects, and 5 h after beginning of illumination when the first visible symptoms of injury were observed.

Biochemical analyses were carried out on fresh plant material which was immediately extracted and assayed according to the appropriate methods listed below. Lipid peroxidation was estimated by the amount of malondialdehyde (MDA), a product of unsaturated fatty acid peroxidation. MDA concentration was measured by the method of Kramer et al. (1991). Membrane permeability of leaves was measured by electrolyte leakage (Palta et al. 1977). Leaves were cut and placed in tubes containing distilled water and after 24 h at room temperature the conductivity of the solution was measured. After twofold refreezing the conductivity was measured again. The results are presented as a percentage of the total conductivity of solutions after double freezing and unfreezing. Hydrogen peroxide was measured spectrophotometrically according to Alexieva et al. (2001). The free thiol-groups were determined according to Ellman (1959) with modifications by Edreva and Hadjiiska (1984). Glutathione content was measured according to Gronwald et al. (1987), and ascorbate content was measured as described by Yabuta et al. (2007). For determination of enzymatic activities, leaves were homogenized at 4°C in 100 mM K-phosphate buffer (pH 7.5), containing 2 mM DTT and 2 mM EDTA. The homogenate was centrifuged at 19,000g for 30 min to yield a crude enzyme extract. Catalase activity was determined by the method of Aebi (1984); peroxidase activity according to Dias and Costa (1983), using guaiacol as an external donor of electrons; superoxide dismutase (SOD) activity according to Beauchamp and Fridovich (1971); glutathione-S-transferase (GST) activity by the method of Gronwald et al. (1987); ascorbate peroxidase (APX) activity according to Guo et al. (2007). Soluble protein was determined by the dye-binding technique (Bradford 1976) using bovine serum albumin as a protein standard.

The experiments were performed three times in three replicates. Data presented are mean values with standard errors (±SE). Different letters indicate significant differences between treatments for the respective measurement point according to Duncan’s multiple range test at P < 0.05.

Results

Stress marker content

It was found that at the end of the dark period (0 h) as well as 2 h after the beginning of illumination there were no significant difference between the levels of malondialdehyde in treated and control plants (Fig. 1a). After 5 h of illumination the amount of malondialdehyde in paraquat treated plants increased by up to 42% in relation to the control, but remained near the control level in H2O2 pretreated plants.

Fig. 1
figure 1

Malondialdehyde content (a), and leakage of electrolytes (b), in pea plants treated with H2O2 and paraquat. Data are mean values ± SE. Different letters indicate significant differences between treatments at P < 0.05

Full size image

Treatment with hydrogen peroxide alone did not alter the membrane permeability during the whole period (0, 2, and 5 h), but the plants treated with paraquat showed considerably increased leakage of electrolytes (Fig. 1b): 14% (0 h), 44% 2 h after illumination, and 60% after 5 h. Pretreatment with H2O2 prevented that induced by paraquat leakage of electrolytes to some extent.

Endogenous hydrogen peroxide and enzyme activities

Application of hydrogen peroxide did not markedly influence endogenous H2O2 concentration. On the other hand, paraquat treatment led to a significant decrease in endogenous hydrogen peroxide in both variants—PQ-only treated and pretreated with hydrogen peroxide plants (Fig. 2a). This occurred during the dark phase (0 h) and after illumination hydrogen content was not increased as expected. The activity of superoxide dismutase (Fig. 2b) was not influenced by H2O2 treatment but in the variants with paraquat application a decrease in the activity of SOD during the light period (2, 5 h) was detected.

Fig. 2
figure 2

Hydrogen peroxide content (a), activity of superoxide dismutase (b), catalase (c) and guaiacol peroxidase (d) in pea plants treated with H2O2 and paraquat. Data are mean values ± SE. Different letters indicate significant differences between treatments at P < 0.05

Full size image

At the end of the dark phase (0 h) catalase activity (Fig. 2c) increased slightly after H2O2 application. Treatment with PQ alone caused an inhibition of catalase activity, and the pretreatment with H2O2 lessened the negative effect of the herbicide to some extent. Marked alterations in catalase activity were not found during the illumination phase, with the exception of the variant of combined treatment with H2O2 and PQ where a significant increase was observed after 5 h of illumination.

Treatment with hydrogen peroxide provoked an increase in guaiacol peroxidase activity (Fig. 2d) during the whole period studied as compared to the control. The herbicide applied alone caused a considerable inhibition of peroxidase activity, especially during the dark phase. However, under light conditions an increase in enzymatic activity was observed. Pretreatment with hydrogen peroxide mitigated the effects of PQ to some extent and at the end of the experimental period peroxidase activity reached the control level.

Hydrogen peroxide treatment caused an inhibition of ascorbate peroxidase as compared to the control. Paraquat also inhibited APX activity but to a much higher degree than H2O2, and at the end of the experiment it was 2.4% of the control. This trend was initiated during the dark phase and the effect of the herbicide was enhanced by illumination. On the other hand pretreatment with H2O2 preserved APX activity in PQ-treated plants to some extent.

Hydrogen peroxide applied alone did not alter the activity of glutathione-S-transferase (Fig. 3a). Paraquat caused an increase in GST activity after 5 h of illumination. However, the combined treatment caused a significant increase in the activity of GST 2 h after the beginning of illumination, and this tendency was maintained till the end of the experiment.

Fig. 3
figure 3

Glutathione-S-transferase activity (a), content of free thiol groups (b), reduced glutathione (c) and oxidized glutathione (d) in pea plants treated with H2O2 and paraquat. Data are mean values ± SE. Different letters indicate significant differences between treatments at P < 0.05

Full size image

Content of non-enzymatic compounds

Treatment of plants with H2O2 increased the amount of free thiol-group containing compounds (Fig. 3b). On the other hand, the herbicide caused a decrease in the content of thiol-group containing compounds, which occurred during the dark phase and continued to diminish considerably till the end of the experiment. Pretreatment with H2O2 followed by application of PQ alleviated the herbicide effect slightly, and after 5 h in light, thiol group content was double that observed in plants treated with herbicide alone.

Hydrogen peroxide by itself caused a slight increase in reduced glutathione (Fig. 3c) after 5 h of illumination. Paraquat treatment markedly decreased GSH content and this low level of GSH was maintained during the whole period. Combined application of both compounds resulted in higher GSH content compared to that measured after herbicide application alone, especially after 5 h of illumination. In contrast to GSH, a significant increase in oxidized glutathione was detected after PQ application (Fig. 3d). Treatment with H2O2 had no effect on GSSG content. There was a decline in the paraquat-induced augmentation of GSSG only after 5 h of illumination.

A twofold decrease in ascorbate content after treatment with hydrogen peroxide (Table 1) was observed, and paraquat applied alone caused a drastic fall in the ascorbate pool. Similarly to AsA, application of H2O2 diminished dehydroascorbate content, while paraquat provoked an increase in the dehydroascorbate level. In both cases this was observed during the dark phase and the effects persisted till the end of the experiment. On the other hand pretreatment with H2O2 prevented ascorbate decline and dehydroascorbate rise caused by paraquat to some extent.

Table 1 Ascorbate peroxidase activity, ascorbate, and dehydroascorbate content in pea plants treated with 2.5 mM H2O2 and 0.2 mM Paraquat (PQ)
Full size table

Discussion

Malondialdehyde production, as well as the enhanced leakage of electrolytes from plant tissues are the result of oxidative damage and imbalance in biomembrane permeability, and thus they are considered to be sensitive stress markers reflecting numerous unfavorable environmental factors including herbicides (Kunert and Dodge 1989; Foyer et al. 1994). We observed that there were no substantial changes in the content of malondialdehyde during the dark phase and after 2 h of illumination, but 3 h later there was an increase only in paraquat treated plants. In contrast, the leakage of electrolytes was already initiated before the light period, it was sharply enhanced by the second hour and continued to rise due to paraquat action. Since MDA is one of the bioproducts of membrane oxidation and does not represent the overall membrane imbalance the observed disparity between changes in MDA content and electrolyte leakage is not unexpected. Similar membrane disorders provoked by paraquat were reported in rice plants by Guo et al. (2007). The restoration to the control level of malondialdehyde and electrolyte leakage decrease observed in pea pretreated with 2.5 mM H2O2 followed by paraquat is an indication that hydrogen peroxide lessened herbicide damage. This is in conformity with our observation that plants pretreated with H2O2 and sprayed with paraquat had higher survival rates compared to those treated with PQ only (Moskova et al. 2007).

Since in the present experimental model, treatment with a naturally-occurring constituent (H2O2) has been investigated, the issue we were interested in was whether exogenously applied hydrogen peroxide will affect endogenous H2O2 concentration. It should be noted that no changes in endogenous H2O2 were detected immediately after the treatment with hydrogen peroxide and this tendency lasted till the end of the experiment (data not shown). This is in accordance with the observation by Yu et al. (2003) of lack of change in the endogenous H2O2 content in Vigna radiata after treatment with hydrogen peroxide. In contrast, paraquat application resulted in a decrease in hydrogen peroxide content especially after 5 h of exposure to light. It suggests that PQ treatment provokes a chemical environment facilitating reactions which produce hydroxyl radicals and this results in a concomitant decrease of hydrogen peroxide content. For example Shen et al. (1999) showed that in cucumber cultivars, treated with chilling temperature, H2O2 content decreased significantly accompanied by a marked increase in highly reactive hydroxyl radical generation.

Similarly to other ROS, the endogenous concentration per se of H2O2 depends on the balance between the rates of its generation (as a metabolite of the normal physiological processes) versus its enzymatic and non-enzymatic utilization or decomposition. In this relation the activities of SOD, catalase, ascorbate peroxidase and guaiacol peroxidase were also evaluated. A minor increase in the activity of SOD was detected in the H2O2-treated plants, while the application of paraquat led to a slight inhibition of the total superoxide dismutase activity under light conditions. Most probably different isoforms of SOD were influenced by the herbicide—Iturbe-Ormaetxe et al. (1998) established that paraquat inhibited Fe-SOD isoforms in pea plants, but did not affect Mn-SOD and CuZn-SOD isoforms. Catalase, guaiacol peroxidase and ascorbate peroxidase are key enzyme systems utilizing hydrogen peroxide in plants, and play important roles in the protection of plants against oxidative damage (Asada and Badger 1984; Puntarulo et al. 1988, 1991). We observed that paraquat slightly inhibited catalase and more significantly, guaiacol peroxidase activity, but ascorbate peroxidase was most substantially affected by the herbicide. On the other hand, treatment with H2O2 did not influence catalase, but caused some inhibition of ascorbate peroxidase, and increased the activity of peroxidase. In all these enzyme systems pretreatment with hydrogen peroxide opposed to some extent the effects of subsequently applied herbicide, especially ascorbate peroxidase. Similar effects of paraquat and pretreatment with 2-aminoethanol were reported by Masher et al. (2005) in barley.

Glutathione-S-transferases are multifunctional enzymes involved in detoxification of a number of xenobiotics and endogenous reactive metabolites. We observed that H2O2 did not influence the activity of GST, but paraquat caused a considerable increase after 5 h of illumination. On the other hand, the combined treatment of H2O2 and paraquat showed a positive effect on GST activity after 2 h in light. Since there is no direct evidence that paraquat is detoxified by conjugation with glutathione via GST, the observed increase in the activity of glutathione-S-transferase might be attributed to its peroxidase activity. Cummins et al. (1999) reported that glutathione-S-transferases can function as glutathione peroxidases in plants treated with herbicides causing oxidative stress, and such activity is of importance for the survival of the resistant species.

Compounds bearing a free thiol group, such as low molecular cell metabolites like glutathione, as well as a number of enzymes which are active only in a reduced state play a key role in important cellular functions, and the massive oxidation of—SH groups could be regarded as an aspect of oxidative toxicity (Haugaard 2000). The relationship between the—SH state in plants and their resistance to various stress factors is well known. Glutathione is a major fraction of the—SH pool in the cells (Alsher et al. 1997). Additionally, glutathione along with ascorbate are key components of the AsA-GSH cycle and play a pivotal role in the defense systems of plants as antioxidants and GSH particularly in detoxification of xenobiotics via GST (Marrs 1996). The changes in free thiol-group containing compounds and reduced glutathione followed a comparable pattern the application of H2O2 caused an increase, and paraquat sharply decreased GSH content. In contrast, GSSG was significantly augmented by the herbicide. The pretreatment with H2O2 prevented to some extent that caused by paraquat transition of GSH into GSSG. It is well documented that the glutathione pool in plants can be influenced by a number of stress factors such as drought, herbicides, heavy metals etc., and some authors considered the accumulation of GSH to be a prerequisite for plant stress resistance (Foyer et al. 1997; Loggini et al. 1999; Kocsy et al. 2000; Tari et al. 2002). On the other hand, Yu et al. (2002) showed that GSH accumulation was substantial but not the only factor for H2O2-induced cold tolerance of Vigna radiata L. plants.

Similarly to glutathione, we observed a drastic depletion of the ascorbate pool after paraquat application. According to the data presented here, the disturbance of glutathione and ascorbate pools takes place during the dark phase, indicating that the herbicide causes considerable damage to the components of the AsA-GSH cycle prior to exposure in light. This cannot be attributed to the primary mode of action of paraquat, but evidence exists that the herbicide also causes injuries in darkness (Marino et al. 2006).

Several authors reported that hydrogen peroxide pretreatment gives protection in plants subjected to various stressors (Yu et al. 2003; Gechev et al. 2002). However, the exact mechanism of the protective action of low H2O2 doses against different unfavorable factors (including herbicides) is not yet elucidated. In a previous study we established that H2O2 pretreatment prevented the inhibition of photosynthesis and chlorophyll loss provoked by PQ, and the plants showed higher survival rates (Moskova et al. 2007). The results presented here demonstrate that the negative effect of PQ, which was considerable in relation to AsA and GSH content, was eliminated to a great extent by H2O2 pretreatment. The application of H2O2 prior to PQ treatment influenced the activities of defense enzymes and especially of APX. One of the possible mechanisms explaining the protective action of H2O2 against PQ toxicity could be related to processes of oxidative coupling of ferulate and tyrosine residues of cell wall proteins and cell wall polysaccharides which form intra- and interchain bridges. These covalent cross-links between wall polymers are a physiologically significant strategy contributing to the termination of wall extensibility, wall strengthening, and the process catalyzed by peroxidase and H2O2 is intensified by various types of stress (Fry 1986; Liyama et al. 1994; Edreva 1996). Another explanation of an H2O2 protective mechanism could be the influence of AsA-GSH cycle components on the cellular redox state. GSH has been considered as an inducer of defense genes (Wingate et al. 1988), and changes in cell thiol-disulphide status provide redox signals leading to gene expression alterations (Foyer et al. 1997). In the present study with the model system described above, this assumption could also be considered. However additional investigations to confirm such a possibility are required. Most probably, hydrogen peroxide used in the present model system acts as a mild stressor hardening treated pea plants.