The influence of the light environment and photosynthesis on oxidative signalling responses in plant–biotrophic pathogen interactions

Authors


Philip M. Mullineaux. Fax: +44 1206872592; e-mail: mullin@essex.ac.uk

ABSTRACT

Plants grow in a constantly fluctuating environment, which has driven the evolution of a highly flexible metabolism and development necessary for their sessile lifestyle. In contrast to the situation in the natural world, the detailed dissection of the regulatory networks that govern plants’ responses to abiotic insults and their interaction with pathogens have been studied almost exclusively in controlled environments where a single challenge has been applied. However, the question arises of how such pathways operate when the plant is subjected to multiple stresses, especially where the expression of overlapping gene sets and common signalling molecules, such as reactive oxygen species (ROS), are implicated. This review will focus on the responsiveness of leaves to their light environment and how this might influence both basal and induced resistance to infection by biotrophic pathogens. While several signalling pathways operate in a complex network of defence responses, the functioning of the salicylic acid (SA) signalling pathway will receive specific consideration. This is because information is becoming available of its role in abiotic stress responses and it dependency on light. This article covers several topics, some of which formerly have received scant attention. These include the effects of infection on photosynthetic performance and carbohydrate metabolism, the parallels between the induction of acclimation to high light and immunity to pathogens, the role of light in the functioning of the SA signalling pathway and the light sensitivity of lesion formation and the use of lesion mimic mutants and transgenic plants. Finally, a model is proposed that attempts to extrapolate these controlled environment-based studies to the functioning of defences against pathogens in a field-grown crop.

INTRODUCTION

Much of the work concerning detailed analysis of signalling pathways and gene regulation in response to both infection by pathogens and to abiotic stress have been conducted using the model plant Arabidopsis thaliana L. Heynh (Arabidopsis). Although a plant of open habitats, in the laboratory Arabidopsis can grow at photosynthetically active photon flux densities (PPFDs) of at least between 60 and 1800 µmol m−2 s−1 (Weston et al. 2000; Müller-Moulé, Golan & Niyogi 2004). Arabidopsis is also a good subject for studying the effects of a sudden increase in light intensity, mimicking the effects of a loss of shade (Karpinski et al. 1997, 1999; Müller-Moulé, Havaux & Niyogi 2003). Such sudden increases in light intensity, to which the plant is not acclimated, result in an increase in excitation energy in excess of that required for photosynthetic metabolism (Asada 1999; Ort & Baker 2002). Excess excitation energy (EEE) is potentially highly damaging to the plant leading to the destruction of photosynthetic reaction centres and oxidative damage often manifested as bleaching, chlorosis or necrosis of leaves (Karpinski et al. 1999, 2003; Karpinska, Wingsle & Karpinski 2000; Kulheim, Agren & Jansson 2002; Mateo et al. 2004). However, EEE is not always associated with a rise in light intensity. Any environmental stresses that limit photosynthetic metabolism, either through effects on gas exchange or upon primary metabolism can have similar consequences under otherwise benign light conditions. Typical abiotic stress conditions that promote an increase in EEE are ozone, drought, salinity and extremes of temperature (Havaux & Kloppstech 2001; Guidi, Degl’Innocenti & Soldatini 2002; Tsonev et al. 2003; Tezara et al. 2003).

EEE is dissipated by several mechanisms that act in concert; broadly classified as non-photochemical and photochemical quenching processes. Such dissipation mechanisms have been extensively reviewed in recent years and the reader is referred to these (Asada 1999; Niyogi 2000; Ort & Baker 2002). Of relevance to considerations here is that reactive oxygen species (ROS) can be produced by several processes linked to an increase in, or the dissipation of excitation energy. These include singlet oxygen (1O2), as a bi-product of over-excitation of chlorophylls in the reaction centre of photosystem II or in the antenna system, superoxide anion (O2·−) and hydrogen peroxide (H2O2) as bi-products of the photo-reduction of molecular triplet state oxygen (O2) at photosystem I (the Mehler reaction) and, indirectly, the production of H2O2 in the peroxisome from the glycollate oxidase-catalysed reaction of the photorespiratory cycle (Fufezan et al. 2002; Asada 1999; Niyogi 1999). This range of ROS and associated radicals closely parallels those resulting from ozone uptake into the mesophyll apoplast (Fiscus, Booker & Burkey 2005). All of these reactions that produce ROS in response to an increase in excitation energy are capable of promoting cellular oxidative damage, but have also been implicated in signalling (Willekens et al. 1997; Karpinski et al. 1999; Hilpert et al. 2001; Op den Camp et al. 2003; Foyer & Noctor 2005). In relation to high light stress responses, ROS-mediated signalling has been argued to be at least one component for the establishment of local and systemic acclimation to high light (Karpinski et al. 1999; Mullineaux & Karpinski 2002; Fryer et al. 2003; Mateo et al. 2004). The signalling pathways implicated in response to high light and other abiotic stresses that promote EEE and ROS production have not been elucidated but are likely to be complicated, implicating at least photosynthetic electron transport, leaf water status, heat stress, abscisic acid and glutathione redox state (Karpinski et al. 1997, 1999; Panchuk, Volkov & Schöffl 2002; Fryer et al. 2003; Ball et al. 2004; Chang et al. 2004). Not surprisingly perhaps, this may indicate that multiple signalling pathways are integrated in establishing acclimation to changed light conditions.

Resistance to pathogens and the role of ROS

In this article the prime consideration from a plant–pathogen perspective, is the defence mechanism(s) elicited upon infection by a range of biotrophic pathogens. Although plants often continually express a range of genes associated with defence against pathogens at low levels, expression of these and additional genes is strongly induced upon contact with specific avirulent races of a pathogen. Such races are avirulent because they produce specific recognition molecules, which activate one or more resistance (R) gene-driven signalling pathways leading to the successful induction of plant defences at the site of infection. The best known of these is the hypersensitive response (HR; Feys & Parker 2000), which also leads to the establishment of whole plant immunity (often called systemic acquired resistance; SAR; Levine et al. 1994; Lamb & Dixon 1997) to subsequent infections by virulent forms of the same or different biotrophic pathogens (Salt, Pan & Kuc 1988; Pan, Ye & Kuc 1993). After the initial recognition, very often ROS and reactive nitrogen species (RNS) such as nitric oxide (NO) are quickly produced (Lamb & Dixon 1997; Wendehenne et al. 2001). This closely parallels the HR response triggered by acute ozone exposure (see Kangasjärvi et al. 2005). Ozone, in causing ROS production, may elicit or signal the responses that have evolved for defence against pathogens.

There are several origins of ROS produced in plant cells during the elicitation of R gene-mediated resistance, the best studied being that from the plasmamembrane, although cytoplasmic and chloroplast sources have also been noted (Bestwick et al. 1997; Allan & Fluhr 1997). In the plasma membrane, it is most likely that membrane-associated NADPH oxidases, and possibly cell wall-bound peroxidases and amine oxidases catalyse the production of ROS (Torres, Dangl & Jones 2002; Bolwell et al. 2002; Yoda, Yamaguchi & Sano 2003). This burst of ROS and RNS then leads to the activation of a complex network of signal transduction pathways involving small molecules such as salicylic acid (SA), jasmonic acid (JA) and ethylene, which via cascades of protein phosphorylation and activated transcription factors leads to changes in the expression of a large number of genes. Many of these genes overlap with induction in response to a range of abiotic stresses (Sharma et al. 1996; Borsani, Valpuesta & Botella 2001; Kunkel & Brooks 2002; Clarke et al. 2004; Kangasjärvi et al. 2005). Further evidence for overlapping signalling responses to abiotic and biotic stress has been shown with an Arabidopsis mutant, adr1, which displays enhanced ectopic expression of an R gene and displays enhanced tolerance to drought, but not salinity or heat stress (Chini et al. 2004). The drought tolerance displayed by adr1 required functioning abscisic acid (ABA) and SA signalling pathways.

PHOTOSYNTHESIS IN DISEASE-SUSCEPTIBLE AND RESISTANT LEAVES

In both resistant and susceptible plant–pathogen interactions, contained or spreading tissue chlorosis or necrosis on leaves is very often observed, which indicates that perturbations in light harvesting and photosynthesis may play a major role in the development of such symptoms. Perturbations in phytohormones levels (such as cytokinin and auxins) in infected leaves may also accelerate or inhibit tissue senescence, inevitably impacting on photosynthetic performance (Pennazio & Roggero 1998; Jameson 2002). Accordingly, the impacts on photosynthesis and allied processes such as gas exchange in leaves have been investigated in a range of plant–pathogen interactions. In infected leaves or isolated mesophyll cells challenged with elicitor preparations from pathogens, depression of photosynthetic electron transport and simultaneous increases in both non-photochemical- and photochemical quenching have been observed. These data indicate that leaves suffer a decline in light use efficiency to drive photosynthetic metabolism and an increase in EEE. In some studies it has been noted that the effects on photosynthesis are not confined to the leaf zones that develop necrotic lesions (Luque et al. 1999; Allen et al. 1999; El-Omari et al. 2001; Arias, Lenardon &Taleisnik 2003; Roloff, Scherm & van Iersel 2004). However, it should be noted that some papers report increases in photosynthesis and photosynthetic electron transport in tissue surrounding lesions (Tecsiet al. 1994; Berger et al. 2004) and in some cases negative effects on photosynthesis were only observed on plants doubly challenged with environmental stresses, such as potatoes simultaneously infected with a root nematode and fungal pathogens (Saeed et al. 1999; Rotenberg et al. 2004). Some of the increases in EEE have been associated with reduced CO2 assimilation caused by decreased stomatal conductance, which creates conditions that would lead to increased ROS production (Mateo et al. 2004).

An additional feature of lesion formation, especially in plant virus-infected leaves, is the accumulation of starch in the lesion itself, in surrounding tissue and also in leaves remote from the development of symptoms, irrespective of the effects on CO2 assimilation rates (Tecsi et al. 1992, 1994; Shalitin & Wolf 2000; Arias et al. 2003). Much of this accumulation is most likely due to disruption of sugar loading into the phloem, accompanied by perturbation of metabolic activity in sink leaves. Thus for this aspect alone, the prevailing light environment will affect rates of photosynthesis and consequently will have an impact on the progression of the disease in virus-infected plants. The higher starch and sugar content observed in such plant–virus interactions would cause a down-regulation of the Calvin cycle, possibly further increasing EEE.

DISRUPTED PHOTOSYNTHESIS AND INDUCTION OF DEFENCE GENE EXPRESSION

Another intriguing possibility is that disrupted photosynthesis and carbohydrate metabolism may lead to induction of plant defence gene expression. This view initially came from transgenic tobacco plants expressing vacuolar and apoplastic yeast invertase which show HR-like lesions in fully expanded (source) leaves (von Schaewen et al. 1990; Sonnewald et al. 1991) and perturbed ROS and antioxidant metabolism (Polle 1996). Such plants accumulate photo-assimilate in their leaves and are disrupted in phloem loading of sucrose (Herbers et al. 1996a). This was accompanied by an accumulation of foliar SA, induction of transcripts of defence genes (such as PR1) and increased resistance to potato virus Y infection. The minimum leaf sugar content associated with induction of defence gene expression was also the same at which repression of photosynthesis-associated gene expression occurred. The observation that a cytosol-located invertase did not produce these effects pointed to disrupted phloem loading of sucrose as being the more likely initiator of the observed alterations in gene expression. However in discs of wild-type tobacco source leaves fed sugars, induction of the same defence genes was shown to be independent of SA (Herbers et al. 1996b). This suggests that the picture left to us from the study of the transgenic invertase plants (Herbers et al. 1996a) is more complicated than it first appeared, with perhaps multiple pathways of plant defence signalling being activated. Again this parallels inhibition of phloem loading that has been reported for some ozone-treated leaves (see Fiscus et al. 2005). This impression of complexity is reinforced by a recent microarray study that shows altered expression of a number of defence and stress-related genes in Arabidopsis seedling treated with glucose (Price et al. 2004). Furthermore increased rates of respiration under conditions of higher sugar content in leaves may lead to increased ROS production. Mitochondrial-sourced ROS have been implicated in a range of abiotic and biotic stress responses and may signal the induction of defence genes, such as those coding for ROS metabolism and protein repair mechanisms (Bowler et al. 1989; Møller 2001; Bechtold, Murphy & Mullineaux 2004). These observations may suggest that prevailing light conditions experienced by source leaves, which would strongly influence their carbohydrate metabolism, may affect the type of defence signalling pathway activated by a pathogen. This theme is developed further below.

COMPARISON BETWEEN THE ESTABLISHMENT OF WHOLE PLANT ACCLIMATION TO HIGH LIGHT AND IMMUNITY TO INFECTION

The above considerations lead to the notion that establishment of acclimation to increased light exposure and immunity to infection might have a number of common features, at least in Arabidopsis: Both processes involve systemic signalling that leads to increased tolerance to high light exposure or pathogen resistance in naïve leaves (Alvarez et al. 1998; Karpinski et al. 1999). In both cases, initiation of signalling involves ROS (Levine et al. 1994; Alvarez et al. 1998; Karpinski et al. 1999; Fryer et al. 2003; Mateo et al. 2004). Both high light stress and infection can induce the expression of common genes, for example GPX7, PR2 and GST6 (Levine et al. 1994; Alvarez et al. 1998; Mullineaux et al. 2000; Ball et al. 2004). High light exposure produces micro-HR like lesions (discrete zones of cell death and accumulation of ROS), similar to those observed on systemically resistant leaves (Alvarez et al. 1998; Zeier et al. 2004). Finally both treatments cause an initial decline in the redox status of the antioxidant glutathione, followed by a doubling of total glutathione content, albeit with different timescales (Karpinski et al. 1997; Mou, Fan & Dong 2003; Müller-Mouléet al. 2003). Other, perhaps less well-described features may also include increased photoinhibition of photosynthetic electron transport (see above and Karpinski et al. 1997, 1999; Fryer et al. 2003) and a critical dependency on leaf water status for the development of HR or induction of sinks for dissipation of EEE (May, Hannond-Kosack & Jones 1996; Yoshioka et al. 2001; Fryer et al. 2003; Mateo et al. 2004). Some of these comparisons are expanded upon below.

LIGHT AND SA

The SA signalling pathway is one of the best-studied signal transduction routes in this complex network of defences. Many studies have shown that SA plays an important role in signalling during local and systemic defence responses upon infection with an avirulent pathogen (Gaffney et al. 1993; Delaney et al. 1994; Vernooij et al. 1994). Local responses may include ion fluxes, strengthening of the cell wall, accumulation of SA and other phenolic compounds, activation of defence genes and induction of the HR (Hammond-Kosack & Jones 1996; Dempsey, Shah & Klessig 1999; Wildermuth et al. 2001). Systemic parts of the plants frequently exhibit increased defence gene expression and elevated levels of SA indicating they have developed SAR. Extensive genetic analysis of the SA signalling pathway has revealed a number of key steps that are shown in Fig. 1. Extensive reviews on this pathway and the role of key SA pathway regulatory genes such as NPR1 have been published recently (Feys & Parker 2000; Dong 2004).

Figure 1.

The salicylic acid (SA) signalling pathway and its relationship to initiation of signalling triggered by detection of a pathogen avirulence (Avr) factor. The nomenclature of the proteins shown is as follows: NON RACE SPECIFIC DISEASE RESISTANCE1, NDR1; PHYTOALEXIN DEFICIENT4, PAD4; ENHANCED DISEASE SENSITIVITY1, EDS1; SALICYLIC ACID INDUCTION DEFICIENT2, SID2; LESION SIMULATING DISEASE1, LSD1; NON-EXPRESSOR OF PR-1 1, NPR1; TGA class basic leucine zipper (bZIP) transcription factor family, TGA; PATHOGENESIS RELATED PROTEIN 1, PR-1. Treatment of leaves with salicylic acid or an incompatible (avirulent) pathogen causes a change in cellular redox status as a consequence of a change in cellular reduced glutathione content (Δ redox; Mou et al. 2003; Dong 2004). The interaction between a resistance gene product (R) and Avr is not direct. Both EDS1-dependent and -independent signalling routes are shown, whose activity depends on the precise Avr and R genes involved. This pathway is modified and combined from several publications (Feys & Parker 2000; Jirage et al. 2001; Rustérucci et al. 2001; Kunkel & Brooks 2002; Mou et al. 2003; Dong 2004). The short parallel lines (//) denote inhibition of the signalling involving ROS by LSD1.

More recently, the accumulation of SA has been shown in several laboratories to be light-dependent in Arabidopsis plants that were either untreated, mock-inoculated or challenged with an avirulent strain of Pseudomonas syringae (Genoud et al. 2002; Karpinski et al. 2003; Zeier et al. 2004). This agrees with observations that in the SA-mediated induction of the HR, light and/or a chloroplastic factor is required, which has been suggested to be a ROS signal (Genoud et al. 2002; Mateo et al. 2004). Thus conditions that promote an increase in excitation energy may be part of a signal that controls SA accumulation, and may provide a common explanation for the observed accumulation of, or requirement for SA in plants subjected to several types of abiotic stresses (Borsani et al. 2001; Munné-Bosch & Peñuelas 2003; Chini et al. 2004; Scott et al. 2004; Clarke et al. 2004). One of these factors could be the impact that increases in EEE may have on the cellular glutathione pool and consequently its redox state; this in turn might impact on the functioning of NPR1 (Fig. 1), which has been shown to be redox sensitive in vivo and in vitro (Mou et al. 2003). It should be noted that although there was a light–dependency of SA accumulation this did not necessarily correlate with SA dependency of induced whole plant immunity at high light levels (Zeier et al. 2004) suggesting that non-SA mechanisms of resistance may operate in high light environments. This may have important implications for ozone studies. It has been assumed that dependence of ozone damage on light is due to access of ozone to the mesophyll via the open stomata (e.g. Fiscus et al. 2005). The evidence from pathogen studies suggests that light may also have an essential role in the development of damage.

In addition to the requirement of light via effects on the chloroplast, there has also been shown to be a dependency on a phytochrome-mediated system for the SA-mediated induction of defence gene expression and the development of the HR. The photoreceptors implicated in this signal system were PhyA and PhyB and the response was shown to be independent of any potential sugar signal (see above; Genoud et al. 2002). Employing a variegated mutant of Arabidopsis, the authors were able to make two important observations. First, whereas light was an absolute requirement for induction of defence genes, functional chloroplasts were not. Second and in contrast, functioning chloroplasts were required for the propagation of HR lesions. This requirement for functional chloroplasts was consistent with observations made for lesion mimic mutants (see below) and was in addition to an absolute requirement for phytochrome (Genoud et al. 2002; Mateo et al. 2004). This indicates at least two levels of control over the functioning of the SA pathway operate that involve the light environment of the leaf. At the time of writing it is not clear if this work is relevant to the same or different genes responsive to abiotic stresses, but earlier work indicated that a wide panel of stress-responsive genes was not controlled by a phytochrome-mediated pathway (Genoud et al. 1998).

The comparisons made above raise the suggestion that any environmental condition, which would bring about an increase in EEE leading to the induction of an acclimation response, could perhaps elicit immunity to a virulent pathogen. In agreement with this hypothesis, recent experiments in our laboratory have shown that Arabidopsis plants that acclimated to a combination of high light and chilling, show some increased resistance to a virulent strain of Pseudomonas syringae pv. tomato when placed back at growth temperature (Fig. 2). Interestingly, chilling has been shown to cause an increase in SA levels in Arabidopsis (Scott et al. 2004), which might be linked to the increase in resistance observed in our experiments. There was a correlation between the level of SA and the growth of plants under such conditions, emphasizing the metabolic cost that induction of defences incurs. This notion is supported by the observed negative correlation between growth rate and elicited resistance to pathogens in spring wheat (Heil et al. 2000). Chemically induced SAR without a pathogen challenge has severe effects on plant growth and yield. It suggests a metabolic competition between limiting resources, which is directly influenced by light-driven processes. More recently it has been shown that limitations in reduced nitrogen supply compromises constitutive and induced resistance to pathogens (Dietrich, Ploß & Heil 2004). Since photosynthesis provides the carbon skeletons and energy for nitrogen assimilation, any disturbance that causes an imbalance in photosynthetic processes would likely affect pathogen resistance by this process alone.

Figure 2.

Pseudomonas syringae pv tomato DC3000 infections of plants grown at low light and 22 °C (▪) or at high light and 10 °C (□). Plants were acclimated to high light (850 µmol m−2 s−1) and chilling (10 °C) 6 d prior to bacterial infection. Bacterial growth tests were performed by infiltration of a suspension of Pseudomonas syringae pv tomato DC3000 (1 × 105 colony-forming units mL−1) in 10 m m MgCl2 and 0.01% Silwet L-77 into the leaves of 3–4 weeks old-plants (n = 18). After infiltration plants were transferred to normal growth conditions (150 µmol m−2 s−1 and 22 °C) and colony forming units were counted at 0, 2 and 4 d post infection (dpi; n = 6 per timepoint). The data are from two independent experiments.

In the context of responses to the environment the observation that a chloroplast stroma localized carbonic anhydrase (CA) is a potent SA binding protein in tobacco may be relevant (Slaymaker et al. 2002). Silencing of the CA gene led to a direct suppression of the HR in disease resistance, perhaps suggesting that an interaction between CA and SA in the chloroplast is part of the establishment of immunity to infection in plants. The authors have explained this observation by suggesting that SA may influence a possible antioxidant activity of CA, although the lack of a suitable assay prevented the testing of this hypothesis. The classical activity of CA in catalysing the interconversion of CO2 and HCO3 (Badger 2003) was not inhibited when the enzyme was bound to SA (Slaymaker et al. 2002). In C3 plants the role of CA remains obscure, especially since a loss of up to 95% of CA in antisense plants has no discernible impact on CO2 fixation rates (Badger 2003). Thus the higher plant enzyme may not play the same role as in algae and cyanobacteria in the delivery of CO2 to RuBisCo (Slaymaker et al. 2002; Badger 2003). Therefore at present much of these interesting observations remain unexplained, but again it does suggest an interaction between processes associated with photosynthesis and SA, and it may not be a coincidence that this strong SA binding activity is located in the chloroplast where SA synthesis partly occurs (Wildermuth et al. 2001).

THE IMPACT OF LIGHT ON THE PHENOTYPES OF LESION MIMIC MUTANTS

The yeast invertase transgenic tobacco plants described above can be regarded as one class of lesion mimic transgenic or mutant plants that have been described for several species (Lorrain et al. 2003). Lesion mimics can be divided into two classes, those that show discrete lesions and less frequently, a spreading propagative lesion phenotype (Lorrain et al. 2003, 2004). Among the latter class, the initiation of spreading lesions occurs under defined environmental conditions or at a particular stage in the plant's life cycle. In maize, rice and Arabidopsis, some of the lethal spreading lesion phenotypes are a consequence of changes in the light environment and its impact on leaves as they develop (Dietrich et al. 1994; Gray et al. 1997; Hu et al. 1998; Takahashi et al. 1999; Mateo et al. 2004; Lorrain et al. 2004).

One of the best-studied lesion mimic mutants is lesion simulating disease 1 (lsd1) of Arabidopsis (Dietrich et al. 1994, 1997; Jabs, Dietrich & Dangl 1996; Kliebenstein et al. 1999; Mateo et al. 2004). This mutant forms O2·−-dependent spreading of chlorotic/necrotic lesions when plants are in long day (> 16 h photoperiod) conditions or under continuous illumination, or when short day (8 h photoperiod) low-light-grown plants are exposed to short (45 min) 20-fold higher irradiance. In the latter example, short-term responses to this photo-inhibitory light stress were no different from the parental wild-type accession, but after 72 h, spreading lesions did develop on the leaves of lsd1. Under permissive conditions (i.e. short day photoperiods), the spreading lesion phenotype could be initiated by inoculation with an avirulent pathogen, treatment with SA or artificially blocking stomata. From such studies it is clear that LSD1 acts as a negative regulator of cell death by controlling ROS homeostasis. Consequently, LSD1 acts upon major plant processes dependent on these molecules, such as resistance to pathogens and acclimation to changing light conditions (Fig. 1). In part, LSD1 may achieve this by regulating the timely expression of at least two types of gene encoding for ROS-scavenging enzymes; superoxide dismutases (SODs) and peroxisomal catalase (CAT1). LSD1 may prevent the spread of a HR-induced lesion by controlling SOD and CAT activity such that accumulation of ROS is halted once a critical threshold has been attained (Kliebenstein et al. 1999; Mateo et al. 2004). Similarly an increase in light intensity or prolonged closure of stomata may require increased dissipation of excitation energy by photorespiration, thus requiring increased rates of removal of H2O2 generated by the photorespiratory cycle. This would be achieved by LSD1-regulated CAT1 gene expression and activity (Mateo et al. 2004). Interestingly under such conditions, microscopic zones of cell death form on high light-exposed leaves, which are reminiscent of HR-like lesions (Zeier et al. 2004; Mateo et al. 2004). In support of this contention, transgenic tobacco plants with suppressed levels of peroxisomal catalase also show a spreading lesion phenotype characterized by its light dependency, accumulation of H2O2, induction of pathogen-associated defence genes and hyper-responsiveness to pathogens (Willekens et al. 1997; Mittler et al. 1999).

Taken together, these data imply a co-ordination of the processes that regulate light acclimation and establishment of immunity to some pathogens. This may be achieved in part by light-driven processes leading to increased EEE, which may contribute to the establishment of the HR and subsequent immunity. In wild-type Arabidopsis, while initiation of the HR, in terms of both cell death and accumulation of ROS was independent of light, subsequent propagation of lesions was light-dependent (Mateo et al. 2004). Therefore, this suggests that factors such as localized closure of stomata and other conditions that may promote EEE and subsequent propagation of ROS have a strong influence on the development of the HR (Mateo et al. 2004). Conversely, factors that slow the closure of stomata may be expected to inhibit the progression of HR lesions, which indeed is the case in both tomato and Arabidopsis leaves kept in high humidity environments (May et al. 1996; Jambunathan, Siani & McNellis 2001; Yoshioka et al. 2001).

One noticeable feature of the leaf's reaction to high irradiance, inoculation with an avirulent pathogen or wounding in the light is the accumulation of H2O2 in veinal tissue and to a lesser extent in neighbouring mesophyll cells (Orozco-Cárdenas, Narváez-Vásquez & Ryan 2001; Fryer et al. 2003; Chang et al. 2004; Mateo et al. 2004). In the case of responses to local high light and wounding in Arabidopsis leaves, it has been argued that the production of ROS in such tissue, possibly derived from the Mehler reaction at photosystem I is required for systemic signalling (Karpinski et al. 1999; Fryer et al. 2003; Chang et al. 2004). Furthermore, the same tissues have been suggested to have a lower threshold for initiating programmed cell death (Alvarez et al. 1998) thus implying an increased sensitivity to environmental cues. In this respect the recent description of a mutant, vascular associated death 1 (vad1), displaying a vein-associated spreading lesion phenotype, is of considerable interest (Lorrain et al. 2004). This lesion mimic mutant shows a light-dependent spreading necrosis along veins starting with petioles and extending into main and secondary veins as the leaf ages. Coincident with the development of lesions, vad1 shows accumulation of H2O2 in the veinal regions and precocious cell death in mesophyll cells adjacent to veins. High levels of SA, expression of defence genes associated with the SA, ethylene and jasmonic acid (JA) signalling pathways and increased resistance to both virulent and avirulent Pseudomonas syringae infection are also observed in the vad1 mutant. Under permissive low light conditions, the mutant behaved similar to wild-type in all of these parameters. The resistance phenotype was found to be dependent on SA, but not NPR1 (see Fig. 1), while the key defence regulators, NDR1 and EDS1 (Fig. 1), were required for all aspects of the vad1 phenotype. Both EDS1 and NDR1 gene products may be required for the integration of ROS into defence signalling networks (Rustérucci et al. 2001). VAD1 encodes a novel putative membrane-associated protein, containing a so-called GRAM domain. This domain is important in intracellular protein binding and lipid-mediated signalling. The GRAM domain is present in a wide range of organisms as part of widely occurring proteins such as the Rab-like GTPase activators. From these considerations, the authors speculated that VAD1 plays a role in defence and cell death signalling associated with the cell membrane. No further investigation of the role of light in initiating the phenotype was reported, which was surprising given the enhanced capacity of veinal tissue to accumulate H2O2 when exposed to light (Fryer et al. 2003). It is tempting to speculate that ROS derived from the action of photosynthetic processes may have initiated the vad1 phenotype and in wild-type plants may be implicated in the amplification or initiation of systemic signalling for the establishment of immunity. In this respect, it may be relevant to the link between disease resistance and response of vascular-associated tissue, that cell death occurs more readily in those cells adjacent to veinal tissue in the catalase-suppressed tobacco plants described above (Dat et al. 2003).

There is also some relevance to the above considerations on lesion mimic mutants in model plants to the situation in crop plants. The economically important mlo-based resistance to powdery mildew (Uncinula necator) in spring barley varieties induces faint discrete lesions on leaves of the growing plant (Lyngkjaer et al. 2000). This durable broad spectrum resistance trait is a consequence of mutant alleles (mlo) of the wild-type gene (Mlo). Homozygous mutations in Mlo have a resistance phenotype because the wild-type gene product (MLO) is a negative regulator of one or more defence responses. Barley MLO is one of a family of so-called seven-transmembrane proteins found in higher plants and bryophytes, but not in prokaryotes, yeast or animals. MLO has some structural features similar to G-protein coupled receptors (Lyngkjaer et al. 2000; Piffanelli et al. 2002). More precise information on MLO function is not available, but it is clear that it can exert control over oxidative processes associated with cell wall cross-linking, inhibition of hyphal growth, formation of HR lesions and possibly intracellular ROS-mediated signalling. However, while mlo has promoted durable resistance against this pathogen, there is a yield penalty in unchallenged crops and compromised responses to some abiotic stresses in field conditions (Baker et al. 1998; Lyngkjaer et al. 2000; Brown 2002). This may suggest that factors such as interaction of homozygous mlo barley with their light environment and other abiotic factors are altered in comparison with wild-type Mlo genotypes. In support of this, laboratory-based studies have demonstrated the importance of MLO for tolerance to a range of abiotic stresses as well as biotic stresses (Piffanelli et al. 2002). The conclusion from these studies on mlo-mediated resistance is that there is a limit to how far one can push a plant into constitutive expression of HR-mediated disease resistance before problems of interaction with the abiotic environment appear.

CONCLUSIONS

Light and its interactions with both photoreceptors and photosynthesis-related processes impact strongly on the susceptibility of plants to infection. Despite the sometimes conflicting information available, several features of the role of light in the development of immunity to biotrophic pathogens become apparent. Firstly, light at low fluence is required for the expression of at least the SA signalling pathway in the plant defence network and conversely in the night, other signalling pathways for the induction of defences operate (Zeier et al. 2004). Secondly, light intensity may also affect which pathways operate, since it may be that the SA pathway does not operate at very high light intensities and yet a similar panel of genes is expressed and HR-like lesions can form. The high rates of photosynthesis encountered in high light conditions may drive the formation of both ROS- and sugar-mediated signals, which may overlap with ABA-mediated signalling (Fryer et al. 2003; Brocard-Gifford et al. 2004; Porteau et al. 2004). This might explain some of the effects that leaf water status has upon the development of the HR.

It is interesting to note that SA promotes increased sensitivity of low-light-grown leaves to conditions that promote EEE, manifested as increased photo-inhibition of photosynthetic electron transport (Mateo et al. 2004). Thus the SA signalling pathway may be an adaptation to a low-light-shaded environment, maintaining the involvement of chloroplast-derived activities (photo-inhibition, altered carbon metabolism, increased ROS production). This would parallel the important physiological responses in the establishment of immunity in photosynthetic tissues under high light conditions.

Most of the above considerations have concerned the response of Arabidopsis rosettes to changing light conditions. Can the observations made be extrapolated in any way to an erect canopy-forming crop species? It can be hypothesized that in sun leaves of a dense crop canopy, the parts of the signalling network (e.g. sugar- or ABA-mediated) deployed would be different from that in shade leaves (e.g. SA-mediated), while the outcome in terms of defence gene expression and successful resistance to a pathogen may be similar or identical. Such considerations might then explain why a phytochrome-mediated control of the SA signalling operates, since this could be important for determining when this transition in the choice of defence signalling pathways occurs. One might further surmise that the choice of signalling pathway may be dictated by the allocation of resources required under different conditions. In this respect the diminished growth of plants constitutively expressing resistance pathways has already been noted. Whether or not such speculation is correct, it is clear that the impact of the light environment on disease resistance pathways is of fundamental importance in understanding how plant–pathogen interactions work in the external environment.

Ancillary