- Top of page
- Materials and Methods
One of the earliest of many diverse defence reactions activated in plant tissues in response to pathogen attack is the accumulation of reactive (or active) oxygen species (ROS) (Bolwell et al., 1995; Bolwell, 1999; Apel & Hirt, 2004), which include the superoxide radical (O2−), hydrogen peroxide (H2O2), and the hydroxyl radical (OH•). The phenomenon of ROS accumulation is often termed the oxidative burst. Chemically, H2O2 is the most stable of the ROS species, and hence the most easy to study (Costet et al., 2002). H2O2 has a direct antimicrobial effect and is involved in cross-linking in cell walls, induction of gene expression, hypersensitive cell death, phytoalexin production and induced systemic resistance (Peng & Kuc, 1992; Wu et al., 1995; Thordal-Christensen et al., 1997; Bolwell, 1999; Costet et al., 2002; Apel & Hirt, 2004; Gechev & Hille, 2005). Several studies have attempted to elucidate the role of H2O2 in host–pathogen systems. For example, it has been reported that biotrophic pathogens are inhibited by H2O2 accumulation (Thordal-Christensen et al., 1997; Vanacker et al., 2000; Mellersh et al., 2002), whereas necrotrophic pathogens are reported to be favoured by H2O2 production or even to stimulate its production (Gönner & Schlösser, 1993; Govrin & Levine, 2000; Kumar et al., 2001; Able, 2003). The question is, then, as follows: how do pathogens exhibiting characteristics of both lifestyles, that is, the so-called hemibiotrophic pathogens, react to H2O2 during their different phases of growth? In order to understand the role of ROS in relation to pathogen biology and to avoid oversimplifications about this role, it is necessary to study several different host–pathogen interactions, involving monocots and dicots as well as different types of pathogens. Subsequently, this information will provide the basis for advanced molecular and genetic studies to verify the mechanisms of formation and role of individual ROS beyond any doubt. The interaction between wheat (Triticum aestivum) and Septoria tritici, the causal agent of speckled leaf blotch or septoria tritici blotch, is currently being established as a model system for studies involving hemibiotrophic pathogens. Defence response studies (Ray et al., 2003; Shetty et al., 2003; N. P. Shetty, unpublished) and full genome sequencing of the pathogen have been initiated (http://www.jgi.doe.gov/sequencing/why/CSP2005/mycosphaerella.html).
Recently, Shetty et al. (2003) found that the infection of wheat by S. tritici was associated with a large and early accumulation of H2O2 in incompatible interactions, coinciding with pathogen arrest and thus indicating a role for H2O2 in the active defence of wheat. In a compatible interaction, very little H2O2 accumulated during the initial biotrophic phase of the interaction, where hyphae grew in the host apoplast after initial stomatal penetrations. This phase lasted until c. 11 d after inoculation (dai) and was followed by a massive accumulation of H2O2, tissue collapse and necrosis, coinciding with pathogen sporulation (c. 5 dai). This pattern suggests that H2O2 production is a defence response during the early biotrophic phase of the interaction, whereas the pathogen thrives with the late, massive accumulation of H2O2 during the necrotrophic phase.
In this study, we manipulate H2O2 concentrations in wheat to determine the causal relationship between H2O2 and the amount of fungal infection in the interaction with S. tritici. This was achieved by scavenging H2O2 with catalase or supplementing with additional H2O2. Furthermore, we attempted to elucidate the origin of the late accumulation of H2O2 by measuring the amounts of chlorophyll and soluble sugars in the leaves to determine the effect of the pathogen on photosynthesis. Based on these data, we propose a hypothesis to explain the adaptation of lifestyle adopted by S. tritici to survive and reproduce in wheat.
- Top of page
- Materials and Methods
We investigated the role of H2O2 in wheat infected by the hemibiotrophic pathogen S. tritici and show that infiltration with catalase in wheat leaves effectively scavenged early produced H2O2, thus resulting in decreased latent period (Fig. 1), increased penetration and mesophyll colonization (Tables 1, 2) and fungal biomass (Fig. 2) in both a susceptible and a resistant cultivar. By contrast, infiltration with H2O2 had the reverse effect, generally resulting in reduced fungal colonization and biomass and increased latent period. The initial changes in fungal biomass were, however, so small that significant differences could only be observed from 12 dai. Such clear changes in susceptibility by manipulation of a single defence compound such as H2O2 (both ways) using a simple and delicate approach has, to our knowledge, not been demonstrated before. The only known role for catalase is to scavenge H2O2, and, furthermore, the control (infiltration of heat-inactivated catalase) in cv. Sevin showed that it is the catalase enzyme activity and not perception of the catalase protein molecule per se that is responsible for the effect observed.
Scavenging of H2O2 in the resistant cv. Stakado also resulted in enhanced pathogen growth in the substomatal cavities, but did not result in pycnidium formation, fungal growth most often stopping when reaching the mesophyll. Presumably, the reason for this is that this cultivar has several additional defence responses, in addition to H2O2, which contribute to pathogen arrest. Observed responses, which may contribute to resistance, include accumulation of peroxidase, β-1,3-glucanase, chitinase and callose (Shetty et al., 2003; N. P. Shetty, unpublished). However, another possible reason for the lack of a clear effect of catalase infiltration is that H2O2 concentrations in Stakado are so high, they cannot be depleted by catalase infiltration. A similar observation has been made in the barley–Blumeria graminis f. sp. hordei pathosystem, where it was not possible to scavenge all the H2O2 produced, despite external application of antioxidants (Zhou, 1998). In addition, fungal growth in our study was probably not affected by the infiltrated substances some days after the last infiltration because of their degradation, making it likely that the plants reverted to their ‘natural’ mode of interaction with the pathogen.
Septoria tritici is generally classified as a hemibiotrophic pathogen with a long symptomless phase that is generally considered to be biotrophic. This is followed by a necrotrophic phase, which starts with tissue collapse and necrosis, later leading to pycnidium formation (Parbery, 1996; Shetty et al., 2003). Accumulation of H2O2 as a defence response that inhibits S. tritici during its initial biotrophic phase is in accordance with studies of other biotrophic pathogens that have been found to be inhibited by this ROS (Thordal-Christensen et al., 1997; Mellersh et al., 2002). Since S. tritici shifts from a biotrophic to a necrotrophic lifestyle, H2O2 could be anticipated to have a different role during the latter phase. Thus, necrotrophic pathogens have been reported to benefit from elevated H2O2 concentrations. For example, Govrin & Levine (2000) concluded that Botrytis cinerea infecting Arabidopsis thaliana induced accumulation of H2O2 in order to kill host cells to facilitate invasion. Likewise, it has been concluded that ROS were important for pathogenicity of other necrotrophic and hemibiotrophic pathogens such as Drechslera spp., Rhynchosporium secalis and Bipolaris sorokiniana (Gönner & Schlösser, 1993; Kumar et al., 2001; Able, 2003). A massive accumulation of H2O2 occurred in susceptible wheat infected by S. tritici during the late stages of the interaction (Shetty et al., 2003). If S. tritici induces the same pattern of H2O2 accumulation as necrotrophic pathogens, scavenging of H2O2 during this stage should make the plants more resistant and hinder pathogen colonization and symptom expression. However, surprisingly, scavenging of H2O2 with catalase increased pathogen growth and reduced the latent period, whereas infiltration of H2O2 had the reverse effect. This demonstrates that H2O2 is also harmful to S. tritici during the necrotrophic phase, thus contradicting previous reports on the role of H2O2 for necrotrophic pathogens.
Among the several potential roles for H2O2 in defence (Lamb & Dixon, 1997; Neill et al., 2002), a direct antimicrobial effect could, among others, be envisaged in the wheat–S. tritici pathosystem. In vitro experiments demonstrated a direct inhibitory effect of 10 mm H2O2 on 4-d-old inoculum. Similarly, millimolar concentrations of H2O2 were found to inhibit pathogen growth in other host–pathogen systems (Peng & Kuc, 1992; Wu et al., 1995). The concentration of H2O2 applied by infiltration gave a strong and clear effect, though less than that which had an effect on fungal growth in vitro. The probable reason for this is that the H2O2 is highly localized in planta and the infiltration procedure is not able to mimic the timely and localized release of H2O2 occurring naturally in the plant in response to infection and spread of the pathogen. We found that the in vitro growth of the S. tritici harvested from 16-d-old cultures was inhibited less by H2O2 concentration than inoculum harvested from 4-d-old cultures. Obviously, it is difficult to relate changes observed in vitro with the in vivo changes, since the latter occur in a close interaction with the host. Nevertheless, it is interesting that this difference in sensitivity parallels the change in fungal physiology occurring during the transition from biotrophic to necrotrophic growth. An interesting observation from these experiments is that, whereas the initial stages of penetration by S. tritici were generally enhanced by the early infiltrations of catalase, H2O2 infiltrations often reduced them. These observations indicate that H2O2 has a very strong effect on penetration events. This is supported by the fact that H2O2 accumulation is particularly strong in the stomatal complexes during infection (Shetty et al., 2003), and that the pathogen is highly sensitive to H2O2 just after inoculation (cf. Fig. 3).
Collectively, our data demonstrate that H2O2 is very important in the defence of wheat against S. tritici during the biotrophic phase of the interaction and that the pathogen does not benefit from the H2O2 present during the necrotrophic phase, even though it can be tolerated (Shetty et al., 2003). This raises the question as to the origin of the massive, late accumulation of H2O2 and how the pathogen copes with it. We do not anticipate fungal origin of the H2O2, and the mitochondria are normally relatively unimportant ROS generators in photosynthesizing tissue (Foyer & Noctor, 2000; Apel & Hirt, 2004; Kużniak & Skłodowska, 2005). A likely source is therefore the chloroplasts. Thus, if damage occurs to the chloroplasts, chlorophyll will be released and this chlorophyll will in turn release H2O2 and other ROS if it is not readily degraded by the plant (Kariola et al., 2005). Therefore, we examined whether photosynthesis was affected by inoculation with S. tritici. Photosynthesis in cv. Sevin inoculated with S. tritici was only down-regulated from 11 dai, corresponding with the first appearance of symptoms. This was, however, not related to a decrease in chlorophyll content in this cultivar even though the tissue became necrotic and H2O2 accumulated (Shetty et al., 2003). The reason for the increase in chlorophyll content per mg FW seen in cv. Sevin at 7 dai is not known, whereas the increase at 15 dai is the result of the reduced FW of the leaves caused by their wilting and necrosis. In conclusion, it appears less likely that the late H2O2 accumulation is the result of stress on photosynthesis. The other major source of ROS in photosynthesizing tissue is the peroxisomes, and it is likely that stress on these bodies is important. Thus, peroxisomes have been shown to be implicated in natural leaf senescence (Pastori & del Río, 1997; Distefano et al., 1999). Furthermore, Kużniak & Skłodowska (2005) showed that B. cinerea infecting tomato induced stress on the peroxisomes, leading to collapse of their antioxidant system, thus partly explaining the pathogen-related leaf death. Whether the same explanation also applies for S. tritici remains to be elucidated. We are currently investigating the source of H2O2 in the wheat–S. tritici system.
It has been shown that carbohydrates are involved as signalling compounds in down-regulation of photosynthesis (Wright et al., 1995; Ehness et al., 1997; Sinha et al., 2002), and increased sugar concentrations have been shown previously to increase host defence response gene expression following pathogen attack (Ehness et al., 1997). In the wheat–S. tritici interaction, we found that photosynthesis was down-regulated from 11 dai and this was followed by release of H2O2 by 13 dai (Shetty et al., 2003). Therefore, we quantified fructose, glucose and sucrose to study if this could explain the effect on photosynthesis and ROS release. We found a very large accumulation of all three sugars in cv. Sevin inoculated with S. tritici starting before initiation of the necrotrophic phase of the interaction (indicated by cell collapse, necrosis and increase in fungal biomass), with a small increase also occurring in cv. Stakado. The sugar accumulation pattern we observed is contrary to that reported for the necrotrophic pathogen B. cinerea (Berger et al., 2004), but is in accordance with observations made for the biotrophic pathogen Blumeria graminis f. sp. tritici (Wright et al., 1995). S. tritici will benefit from this late, massive sugar release as it feeds on apoplastic carbohydrates (Rohel et al., 2001). Nevertheless, Rohel et al. (2001) observed that during sporulation, there were indications of carbohydrate starvation in the fungus, suggesting a need to increase availability of nutrients. Therefore, when apoplastic carbohydrates are depleted, the pathogen probably stimulates sugar release to initiate pycnidium formation, thereby eliciting host tissue necrosis. Sugar release could also be the consequence of toxin production by the pathogen (Shetty et al., 2003).
Sugar release could result in host tissue necrosis and the reduction of photosynthesis, or, alternatively, stress on the peroxisomes could trigger subsequent ROS release. Tissue necrosis will result in activation of defence reactions, releasing H2O2, which could participate in cell death and reduction of photosynthesis. In support of this suggested chain of events is the observation that, in cv. Sevin, S. tritici lives practically unnoticed until sporulation, at which time defence responses are finally expressed (Shetty et al., 2003; N.P. Shetty, unpublished). Alternatively, as a consequence of the repression of photosynthesis, the host may react, at the onset of symptom expression, by changing to heterotrophic metabolism (Ehness et al., 1997), or assimilates from the chloroplasts may be translocated to other parts of the plant as a response to stress (Krupinska & Humbeck, 2004). Irrespective of the source, the late H2O2 accumulation probably stimulates the pathogen to enter its necrotrophic phase as a requirement to survive. The pathogen thus copes with the toxic substances accumulating at this stage and survives the death of host tissues. During the initial, biotrophic stage, the pathogen does not penetrate any cell (Cohen & Eyal, 1993; Kema et al., 1996; Shetty et al., 2003) and it survives, presumably by attempting to avoid recognition until it has produced a sufficient amount of biomass to initiate reproduction. In this respect, we can speculate whether the symptomless phase is really biotrophic sensu stricto, or whether the pathogen is, in fact, an endophyte, that is, living saprohytically and not pathogenically. In further support of this proposition is the fact that S. tritici lives strictly in the apoplast with no special intracellular feeding structures (Palmer & Skinner, 2002).