Singlet oxygen-mediated and EXECUTER-dependent signalling and acclimation of Arabidopsis thaliana exposed to light stress

Abstract

Plants respond to environmental changes by acclimation that activates defence mechanisms and enhances the plant's resistance against a subsequent more severe stress. Chloroplasts play an important role as a sensor of environmental stress factors that interfere with the photosynthetic electron transport and enhance the production of reactive oxygen species (ROS). One of these ROS, singlet oxygen (¹O₂), activates a signalling pathway within chloroplasts that depends on the two plastid-localized proteins EXECUTER 1 and 2. Moderate light stress induces acclimation protecting photosynthetic membranes against a subsequent more severe high light stress and at the same time activates ¹O₂-mediated and EXECUTER-dependent signalling. Pre-treatment of Arabidopsis seedlings with moderate light stress confers cross-protection against a virulent Pseudomonas syringae strain. While non-pre-acclimated seedlings are highly susceptible to the pathogen regardless of whether ¹O₂- and EXECUTER-dependent signalling is active or not, pre-stressed acclimated seedlings without this signalling pathway lose part of their pathogen resistance. These results implicate ¹O₂- and EXECUTER-dependent signalling in inducing acclimation but suggest also a contribution by other yet unknown signalling pathways during this response of plants to light stress.

1. Introduction

Plants are sessile organisms exposed to a highly variable environment that may adversely affect their growth and development [1]. One strategy of plants to cope with environmental stress and mitigate its negative impact is acclimation. Stress acclimation is often triggered by minor environmental changes and results in an enhanced protection against a subsequent more severe stress [2–4]. Such an enhanced stress resistance may be confined to the stress that triggers acclimation or it may provide cross-protection also against other forms of stress [5–7]. In order to acclimate, plants need to monitor and sense continuously their environment. Environmental cues are transformed into signals that are relayed to the nucleus and modulate the plant's capacity to withstand future stress. As plants are continuously challenged not only by one but several environmental factors whose relative impact may change, acclimation is difficult to analyse. Signals activated by environmental changes seem to operate via a signalling network that integrates various cues at the same time. The stability and robustness of such a network has been ascribed to the integration of numerous and redundant pathways with partially overlapping specificities [8]. How these signalling pathways interact and determine stress acclimation in plants is largely unknown.

Chloroplasts act as an important sensor of environmental factors such as high light, low or high temperature and drought that interfere with the photosynthetic electron transport [9,10]. Under these environmental conditions, light absorbed by photosynthetic membranes exceeds the plant's capacity to assimilate CO₂ and leads to the over-reduction of the electron transport chain [11,12]. In order to maintain the photosynthetic electron transport chain in an at least partially oxidized state, photosynthetic organisms have evolved several ways to dissipate excess light energy that may result in increased production of reactive oxygen species (ROS) [13–15]. The term ‘ROS’ is used to describe the effects of several chemically distinct ROS. While oxygen in its ground state is chemically inert, it may be converted to ROS by either electron or energy transfer reactions. The former leads by stepwise reduction to the formation of superoxide, hydrogen peroxide (H₂O₂) and hydroxyl radical [16], whereas the latter results in the formation of singlet oxygen (¹O₂) [17]. Depending on their concentration, each of these ROS may cause oxidative damage or is believed to initiate signalling. In plants under light stress, concentrations of these ROS increase almost simultaneously, thus making it difficult to link a given ROS to a particular cellular response. To overcome these obstacles and dissect the complexity of signalling in plants under light stress, the conditional fluorescent (flu) mutant of Arabidopsis thaliana has been established as an experimental system to describe in molecular terms the biological role of one particular ROS, ¹O₂ [18,19], and determine whether or not it takes part in triggering stress acclimation in plants under light stress [20].

2. ¹O₂-mediated and EXECUTER-dependent signalling in the flu mutant

Mutations of the FLU gene interfere with the negative feedback control of tetrapyrrole biosynthesis in flu mutants and allow the accumulation of protochlorophyllide (Pchlide), the immediate precursor of chlorophyllide (Chlide), in the dark [18,21,22]. FLU is a nuclear-encoded protein that, after import and processing, becomes tightly associated with plastid membranes and forms part of a complex that comprises also the four enzymes catalyzing the final steps of chlorophyll synthesis [23]. It links chlorophyll synthesis and the target of feedback control, Glu tRNA reductase, the first enzyme committed to tetrapyrrole biosynthesis, thereby allowing the Mg²⁺ branch to control the initial steps of this pathway [24]. In contrast to dark-grown wild-type seedlings, flu seedlings are no longer able to restrict the accumulation of Pchlide [18]. When these seedlings are transferred from the dark to the light, they rapidly bleach and die because of the release of ¹O₂ [19] that is formed due to the photosensitizing activity of excess amounts of free Pchlide. The flu mutant remains viable though, when it is kept from the very beginning under continuous light. Under these conditions, Pchlide in the flu mutant is immediately photoreduced to Chlide and in this way does not reach critical levels that might lead to an enhanced production of ¹O₂. As long as flu plants are kept under continuous light, no obvious difference between mutant and wild-type can be observed [18,19]. These properties of the flu mutant have been exploited to study the physiological role of ¹O₂ by growing flu plants initially under continuous light and then transferring them to the dark and re-exposing them to light. The response of the flu mutant to ¹O₂ has been analysed in seedlings and in mature plants that are ready to bolt. Seedlings of flu bleach and die, when they are kept under repeated 16 h L : 8 h D cycles, whereas mature plants grown initially under continuous light develop necrotic lesions on their leaves and stop their growth immediately after being shifted to daily dark/light cycles [19]. By varying the length of the dark period, one can modulate non-invasively the level of the photosensitizer Pchlide, and define conditions that reduce the toxicity of ¹O₂ and reveal a genetic basis of its signalling role [25–27]. Under these optimized conditions, the nucleus-encoded and chloroplast-localized EXECUTER1 (EX1) and EXECUTER2 (EX2) proteins are essential for initiating ¹O₂-mediated responses. Inactivation of these proteins in an ex1/ex2/flu triple mutant abolishes ¹O₂-mediated cell death of seedlings and growth inhibition of mature plants [25,28]. As the triple mutant in the dark over-accumulates Pchlide and upon re-illumination generates ¹O₂ similar to the parental flu line but yet behaves like wild-type, ¹O₂-mediated responses in the flu mutant are not due to photo-oxidative damage but are the result of ¹O₂-mediated and EX-dependent signalling [28].

3. Control of ¹O₂-responsive nuclear gene expression in the flu mutant

After transferring flu seedlings from the dark to light, the release of ¹O₂ triggers an immediate induction of nuclear gene expression changes, with the number of affected genes rapidly increasing over time [19]. As these nuclear gene expression changes are suppressed in ex1/ex2/flu triple seedlings, they were initially attributed to a single retrograde plastid-to-nucleus signalling pathway that is activated within chloroplasts by the release of ¹O₂ and depends on the plastid proteins EX1 and EX2 [19,27]. Subsequent studies of the flu mutant revealed, however, that the regulation of ¹O₂-responsive genes in the flu mutant seems to be more complex than originally anticipated. One of the first cellular changes that occur within less than 30 min following the dark-to-light shift is a loss of chloroplast integrity and the rupture of the central vacuole [28]. These two events mark the beginning of a ¹O₂-mediated and EX-dependent programmed cell death response of the flu mutant and both precede the upregulation of most of the ¹O₂-responsive nuclear genes reported earlier [28,29]. The loss of cellular integrity is expected to impact nuclear gene expression. Hence, many of the ¹O₂-responsive genes of the flu mutant reported earlier are probably only indirectly affected by ¹O₂-mediated and EX-dependent signalling and their upregulation seems to mark activation of other signalling pathways closely associated with cellular damage [28,30]. This interpretation is supported by a comparison of nuclear gene expression changes in flu following a dark-to-light shift and in wild-type seedlings transferred from moderate light (90 µmol photons s⁻¹ m⁻², 20°C) to a combined low temperature/higher light stress (270 µmol photons s⁻¹ m⁻², 12°C) that previously had been shown to activate ¹O₂-mediated and EX-dependent signalling without triggering an immediate cell death response as seen in flu seedlings [28]. Most of the genes upregulated in flu following the release of ¹O₂ are not affected in wild-type exposed to the combined lower temperature/higher light stress, and vice versa only a minor fraction of the stress-induced and EX-dependent nuclear genes of wild-type is also affected in the flu mutant [29].

Oxidative damage in flu seedlings kept under optimized light/dark conditions is a consequence of ¹O₂-mediated and EX-dependent signalling that occurs after ¹O₂-mediated loss of chloroplast integrity had been initiated. Polyunsaturated fatty acids are a preferred target of ROS, and non-enzymatic peroxidation of these molecules has been used as a marker for oxidative damage [31,32]. At the beginning of re-illumination of pre-darkened flu seedlings, peroxidation of linolenic acid, the most prominent polyunsaturated fatty acid of chloroplast membranes [33], happens almost exclusively enzymatically by lipoxygenases and not through direct attack by ¹O₂ [19,26,28]. However, during the following ¹O₂-mediated programmed cell death response, non-enzymatic peroxidation of polyunsaturated fatty acids starts to prevail [28]. The signature of peroxidation products indicates that in flu seedlings during the progression of cell death, non-enzymatic peroxidation is caused primarily by ¹O₂ and not by other ROS [28,32]. Thus, in flu seedlings kept under conditions that favour ¹O₂-mediated and EX-dependent signalling, two different modes of ¹O₂ activity can be distinguished that occur sequentially during re-illumination of pre-darkened flu seedlings. First, ¹O₂ activates retrograde plastid-to-nucleus signalling without causing detectable cellular damage, and then, as a consequence of the ¹O₂-mediated and EX-dependent cell death response, ¹O₂ accelerates the loss of cellular integrity and lesion formation in flu seedlings by directly reacting with various cellular components, e.g. polyunsaturated fatty acids in chloroplast membranes. Initially, ¹O₂ seems to be exclusively generated by excitation of the photosensitizer Pchlide, but during the following loss of chloroplast integrity also chlorophyll seems to act as a photosensitizer that in a positive feedback loop enhances ¹O₂ production. As one would expect, this second phase of ¹O₂ formation is still part of the genetically controlled programmed cell death response initiated by ¹O₂- and EX-dependent signalling and is suppressed in flu/ex1/ex2 seedlings [28].

4. ¹O₂-mediated and EX-dependent signalling in wild-type

¹O₂-mediated and EX-dependent activation of a suicidal programme in flu seedlings or the arrest of bolting and flower formation during the transition from vegetative to generative growth in mature flu plants have not been found in wild-type plants, when these are exposed to environmental stress conditions that are known to promote ¹O₂ formation in chloroplasts. ¹O₂-mediated and EX-dependent signalling as seen in the flu mutant operates under conditions that are different from those to which the wild-type has adapted, and thus its physiological significance in wild-type may not be directly deduced from ¹O₂-mediated responses of the flu mutant. In the light-grown wild-type transferred to the dark, Pchlide does not over-accumulate as in flu but is barely detectable due to strict negative feedback control of tetrapyrrole biosynthesis [34]. Whereas in flu ¹O₂ signalling occurs without concomitant signalling induced by other ROS [19], in the wild-type ¹O₂ is released under stress conditions that also enhance the production of superoxide, H₂O₂ and other stress signalling molecules [28,30,35] that modulate ¹O₂-mediated responses shown by the flu mutant [20,36]. In wild-type seedlings transferred to the combined low temperature/higher light stress mentioned above, ¹O₂-mediated and EX-dependent signalling is activated as indicated by the rapid upregulation of the ¹O₂-responsive marker gene AAA-ATPase that is suppressed in ex1/ex2 seedlings [28]. However, unlike ¹O₂-mediated and EX-dependent signalling in flu seedlings, in wild-type seedlings this signalling pathway does not correlate with an immediate programmed cell death response and the subsequent collapse of seedlings. Only when seedlings have been kept for several days continuously under the combined low temperature/higher light stress, a ¹O₂-mediated and EX-dependent cell death response is initiated, which leads to the formation of microlesions in the wild-type without causing detectable oxidative damage or reducing the viability of seedlings [28].

5. Formation of ¹O₂ in wild-type under light stress

In the wild-type under light stress, chlorophyll acts as the photosensitizer that generates ¹O₂ [37]. Formation of ¹O₂ in chloroplasts may occur within either light-harvesting antennae complexes [38] or the reaction centre of photosystem II (PSII) [11,12]. Chlorophyll in light-harvesting antennae complexes is in close contact with carotenoids that efficiently quench excess light energy and suppress the release of ¹O₂ by scavenging directly this ROS whenever it is formed within their immediate vicinity [39] (figure 1). Only when the excitation energy absorbed by chlorophyll remains trapped inside the antennae complexes and exceeds the light-quenching capacity of carotenoids does the release of ¹O₂ occur, leading to non-enzymatic oxidation of lipids and carotenoids and causing oxidative damage [33,40] (figure 1). Some of the non-enzymatic oxidation products may act as signalling molecules [41,42]. Signalling under these conditions can be regarded as a consequence of oxidative damage and induces responses that are not dependent on EX1 and EX2 [28,42]. ¹O₂ production within the PSII reaction centre occurs whenever the electron acceptor site of PSII is reduced and unable to oxidize the excited reaction centre chlorophyll P680 [11,12]. This may happen even under low light intensities [43]. ¹O₂ generated by the reaction centre chlorophyll of PSII has been suggested to interact with the D1 protein of the PSII reaction centre that may act as a scavenger of ¹O₂ and, following its oxidation, needs to be replaced by a newly synthesized D1 polypeptide [12] (figure 1). ¹O₂ formed under mild light stress conditions does not seem to cause oxidative damage and its signalling activity depends on EX1 and EX2 as indicated by the absence of lesion formation in ex1/ex2 seedlings exposed to the combined low temperature/higher light stress [28].

Figure 1.

Production of ¹O₂ at two different sites in chloroplasts of Arabidopsis exposed to light stress. Under severe stress, excitation energy remains trapped inside the light-harvesting antenna pigment complexes shown on the left side and gives rise to excess amounts of ¹O₂ that exceed the light-quenching capacity of carotenoids closely connected to chlorophyll. ¹O₂ production under these conditions leads to non-enzymatic oxidation of lipids and carotenoids and photo-oxidative damage. ¹O₂ production within the PSII reaction centre shown on the right side occurs whenever the primary electron acceptor at the acceptor side of PSII remains reduced and is unable to oxidize the P680 chlorophyll of the PSII reaction centre. Carotenoids are not in contact with the reaction centre chlorophyll P680, and ¹O₂ formation may happen even under low light intensities. The PSII reaction centre protein D1 (D1) has been suggested to act as a ¹O₂ scavenger [12] and upon its oxidation to be degraded.

6. Light stress acclimation

In previous studies with Chlamydomonas reinhardtii [2] and a chlorophyll b-less light-sensitive mutant of Arabidopsis [44], it has been shown that ¹O₂ formed under light stress conditions may affect the susceptibility of cells to abiotic stress [2]. A transient or moderate elevation of ¹O₂ resulted in the activation of a subset of ¹O₂-responsive genes and protection against a subsequent more severe light stress. To assess the possible presence of a ¹O₂-mediated and EX-dependent stress acclimation in Arabidopsis thaliana wild-type plants, Arabidopsis seedlings grown under moderate light (90 µmol photons s⁻¹ m⁻², 20°C) were first exposed for 24 h to a brief combined low temperature/high light stress (270 µmol photons s⁻¹ m⁻², 12°C) [28]. These plants were then returned to the previous moderate light condition for 6 h to allow recovery from the pre-stress before they were re-exposed for various lengths of time to the combined low temperature/higher light stress (figure 2a). The induction of acclimation under these conditions was monitored by the effect of pre-stress on PSII activity. The functional state of PSII was determined by the maximum quantum efficiency of PSII expressed as the ratio of variable to maximum chlorophyll fluorescence (F_v/F_m) (figure 2b). During 24 h of combined low temperature/higher light treatment, wild-type seedlings suffered from photoinhibition of PSII as indicated by a drop of the F_v/F_m ratio from 0.86 to approximately 0.7. ex1/ex2 seedlings exposed to the same 24 h-pre-stress treatment showed a similar photoinhibition of PSII. However, when this stress treatment was continued for up to 96 h, PSII activity in ex1/ex2 seedlings was less inhibited than in the wild-type (figure 2b). After pre-stressed seedlings had been returned to the original moderate light conditions, they recovered from photoinhibition of PSII within the next 6 h and restored the original maximum quantum efficiency of PSII (figure 2b). The 24 h stress pre-treatment was sufficient to induce stress acclimation. When pre-stressed seedlings following the recovery phase were re-exposed to the same stress as used during the pre-stress treatment, they showed a strongly enhanced stress resistance as indicated by the reduced photoinhibition of pre-stressed seedlings relative to non-pre-stressed control seedlings (figure 2b). Upregulation of the ¹O₂-responsive marker gene AAA-ATPase in the wild-type and its suppression in ex1/ex2 reveals that ¹O₂-mediated and EX-dependent signalling was activated during the pre-stress treatment [28]. However, the protection against stress-induced photoinhibition of PSII in pre-stressed wild-type and ex1/ex2 seedlings was almost identical, suggesting that this acclimatory response was not dependent on EX1 and/or EX2 (figure 2b).

Figure 2.

Stress acclimation of Arabidopsis seedlings. (a) Five-day-old seedlings initially grown at room temperature under moderate light are transferred for 24 h to a combined low temperature/higher light stress and moved back to the original moderate light conditions for 6 h, before they are exposed again for various lengths of time to the combined low temperature/higher light stress regime. (b) Following the 24 h pre-stress and 6 h of recovery, plants are acclimated as shown by the enhanced protection of PSII in pre-stressed seedlings relative to the non-pre-stressed controls. Results represent mean values of three independent experiments ±s.d.

7. Cross-protection of seedlings acclimated to light stress against other forms of stress

The possible contribution of ¹O₂-mediated and EX-dependent signalling to cross-protection against other forms of stress has been analysed. Several lines of evidence suggest that ¹O₂- and EX-dependent nuclear gene expression changes may enhance pathogen resistance. Among the 36 ¹O₂-responsive transcription factor genes upregulated in the flu mutant shortly after the beginning of enhanced ¹O₂ production, 22 are also activated in response to pathogens (table 1; http://ausubellab.mgh.harvard.edu/nsf2010/). Microlesions similar to the ones formed in wild-type seedlings exposed to moderate light stress are closely associated with an enhanced pathogen resistance (figure 3a) [46] and the concomitant upregulation of the PATHOGEN-RELATED (PR) PROTEIN1 gene in light-stressed seedlings (figure 3a) resembles PR1 expression changes in response to various other abiotic and biotic stresses [47]. The light-stress-induced microlesion formation and PR1 expression were suppressed in ex1/ex2 seedlings (figure 3a). Collectively, these data suggest that ¹O₂-mediated and EX-dependent stress acclimation may reduce the susceptibility of seedlings to pathogens. To verify this suggestion, plants were spray-inoculated with virulent Pseudomonas syringae DC3000 that constitutively expresses a LUCIFERASE reporter gene [45]. Bacterial growth was determined after different lengths of incubation by measuring the luciferase activity in extracts of inoculated plants. In a first experiment, non-pre-stressed wild-type and ex1/ex2 seedlings grown under moderate light conditions were exposed to the pathogen and the number of bacteria in plants was determined after 2 days of incubation. Inactivation of the ¹O₂- and EX-dependent signalling pathway in ex1/ex2 seedlings did not seem to affect the bacterial growth (figure 3b). Both lines were equally susceptible to the pathogen. When non-pre-stressed wild-type seedlings and wild-type seedlings exposed for 24 h to the combined low temperature/higher light pre-stress and a subsequent 6 h recovery phase were compared, bacterial growth was strongly suppressed in pre-stressed seedlings, demonstrating that the light-stress-induced acclimation of the wild-type triggers cross-protection against the virulent Pseudomonas strain (figure 3c). Inactivation of ¹O₂-mediated and EX-dependent signalling diminished significantly the pathogen resistance of pre-acclimated seedlings but did not fully restore the pathogen susceptibility of non-pre-stressed wild-type seedlings that one would expect if cross-protection against pathogens in the pre-acclimated seedlings would be exclusively controlled by ¹O₂-mediated and EX-dependent signalling (figure 3d). These results confirm that ¹O₂₋ and EX-dependent signalling contributes to the light-stress-induced cross-protection against pathogens, but at the same time indicate that other signalling events must be considered to fully account for the enhanced pathogen resistance in light-stress-acclimated Arabidopsis seedlings. One of the features of the plant's reaction to a shift to high light is a rapid accumulation of H₂O₂ [5,28]. An increase in H₂O₂ is closely associated with stress acclimation and has been suggested to contribute also to an enhanced resistance of pre-acclimated Arabidopsis plants to pathogens [5,48]. Enhanced levels of ¹O₂ and H₂O₂ stimulate the production of hormones such as jasmonic acid, salicylic acid and ethylene that activate different sets of genes encoding antimicrobial proteins and enhance the resistance against distinct microbial pathogens [5,19,44,49]. Even though ¹O₂ and H₂O₂ operate via separate signalling pathways, they may interact with each other [36]. During light-stress-induced acclimation, both appear to undergo also extensive crosstalk with other stress-related signalling pathways [44,48,49]. One of the big challenges of future research will be to dissect this complex signalling network and to identify other signals that during light stress interact with ¹O₂ and H₂O₂ and control stress acclimation.

Table 1.

Identification of ¹O₂-responsive transcription factor genes of the flu mutant that are also activated in response to pathogens. Thirty-six transcription factor genes have been listed that are upregulated shortly after the dark-to-light shift of the flu mutant [19]. Genes in bold have also been reported to be activated in response to pathogens (http://ausubellab.mgh.harvard.edu/nsf2010/).

	common name	AGI number
ERF/AP2	ERF/AP2	At1g21910
	ERF11	At1g28370
	ERF13	At2g44840
	ERF4	At3g15210
	ERF6	At4g17490
	ERF2	At5g47220
	ERF5	At5g47230
	ERF/AP2	At5g51190
	ERF/AP2	At5g61590
	ERF/AP2	At5g61600
	DREB2A	At5g05410
		At1g25560
	RAV1	At1g13260
	RAV2	At1g68840
MYB	MYB51	At1g18570
	MYB15	At3g23250
	MYB77	At3g50060
WRKY	WRKY40	At4g31550
	WRKY33	At2g38470
	WRKY46	At2g46400
	WRKY22	At4g01250
	WRKY53	At4g23810
	WRKY11	At4g31550
	WRKY18	At4g31800
Zinc finger	ZAT10/STZ	At1g27730
	ZAT11	At2g37430
		At5g04340
	ZAT12	At5g59820
	AZF1	At5g67450
	CZF1/ZFAR1	At2g40140
		At3g55980
NAC	ANAC036	At2g17040
	ANAC062	At3g49530
	ANAC102	At5g63790
others	HSFA4A	At4g18880
	BT5	At4g37610

Figure 3.

¹O₂-mediated and EX-dependent acclimation of Arabidopsis seedlings exposed to light stress confers cross-protection against the biotrophic pathogen P. syringae DC3000. (a) ¹O₂-mediated and EX-dependent signalling in wild-type seedlings exposed to the combined low temperature/higher light stress for 96 h as shown in figure 2a induces programmed cell death as revealed by trypan blue (TB) staining and upregulation of PR1 gene expression, both of which are suppressed in ex1/ex2 seedlings. (b–d) Plants were spray-inoculated with virulent P. syringae DC3000 that constitutively expresses a LUCIFERASE reporter gene [45]. Bacterial growth was determined after different lengths of incubation by measuring the luciferase activity in extracts of inoculated plants. (b) Non-pre-stressed wild-type and ex1/ex2 seedlings show a similar high susceptibility to the virulent pathogen. (c) Wild-type seedlings acclimated to light stress show an enhanced resistance against the pathogen as indicated by the strongly reduced number of bacteria grown on seedlings 2 days after infection (DPI). (d) Inactivation of the ¹O₂- and EX-dependent signalling pathway in ex1/ex2 seedlings only partially restores the susceptibility of non-acclimated control wild-type seedlings. Results in (b–d) represent mean values of three independent experiments ±s.d.

Funding statement

Work done in our laboratory has been supported by the Boyce Thompson Institute for Plant Research and the National Institutes of Health (grant no. R01-GM085036 to K.A.).