Diversity and Evolution of Ascorbate Peroxidase Functions in Chloroplasts: More Than Just a Classical Antioxidant Enzyme?

Abstract

Reactive oxygen species (ROS) have dual functions in plant cells as cytotoxic molecules and emergency signals. The balance between the production and scavenging of these molecules in chloroplasts, major sites for the production of ROS, is one of the key determinants for plant acclimation to stress conditions. The water–water cycle is a crucial regulator of ROS levels in chloroplasts. In this cycle, the stromal and thylakoid membrane-attached isoforms of ascorbate peroxidase (sAPX and tAPX, respectively) are involved in the metabolism of H₂O₂. Current genome and phylogenetic analyses suggest that the first monofunctional APX was generated as sAPX in unicellular green algae, and that tAPX occurred in multicellular charophytes during plant evolution. Chloroplastic APXs, especially tAPX, have been considered to be the source of a bottleneck in the water–water cycle, at least in higher plants, because of their high susceptibility to H₂O₂. A number of studies have succeeded in improving plant stress resistance by reinforcing the fragile characteristics of the enzymes. However, researchers have unexpectedly failed to find a ‘stress-sensitive phenotype’ among loss-of-function mutants, at least in laboratory conditions. Interestingly, the susceptibility of enzymes to H₂O₂ may have been acquired during plant evolution, thereby allowing for the flexible use of H₂O₂ as a signaling molecule in plants, and this is supported by growing lines of evidence for the physiological significance of chloroplastic H₂O₂ as a retrograde signal in plant stress responses. By overviewing historical, biochemical, physiological and genetic studies, we herein discuss the diverse functions of chloroplastic APXs as antioxidant enzymes and signaling modulators.

Introduction

Chloroplasts are plant organelles that are responsible for photosynthesis, which generates molecular oxygen and reducing power from the splitting of water under illumination, thereby making them significant sources of reactive oxygen species (ROS), such as hydrogen peroxide (H₂O₂), the superoxide radical (O₂^–), singlet oxygen (¹O₂) and the hydroxyl radical (·OH) (Asada 1999). The rate of ROS production is greatly affected by environmental conditions. The over-reduction of the photosynthetic electron transport (PET) chain occurs under a high light intensity because of the exhaustion of NADP⁺, a terminal electron acceptor, which facilitates ROS production (Asada 1999). This also occurs under conditions with a normal (or even low) light intensity, in which the Calvin cycle is inhibited, e.g. drought, salinity and low temperatures. Since ROS are cytotoxic molecules, the development of rigorous antioxidant system(s) during evolution must have been essential for plant survival under changeable environments. The discovery of a set of enzymes related to ascorbate metabolism, including ascorbate peroxidase (APX), led to the water–water cycle being identified as a plant-specific system involving the production and scavenging of ROS in chloroplasts (Asada 1999, Shigeoka et al. 2002, Foyer and Shigeoka 2011).

In the water–water cycle, electrons excised from water at PSII are transferred to oxygen by PSI, resulting in the formation of O₂^– (Asada 1999), which is subsequently converted into H₂O₂ by membrane-attached copper/zinc superoxide dismutase (Cu/Zn-SOD). Thylakoid membrane-bound ascorbate peroxidase (tAPX) reduces H₂O₂ back into water using ascorbate as an electron donor. The reactions of these enzymes act as the first layer of ROS scavenging, followed by their removal by iron SOD (Fe-SOD) and stromal APX (sAPX) as the second layer in stroma. The oxidized form of ascorbate, which is generated by the APX reaction, is reduced by ferredoxin-, glutathione- and NAD(P)H-dependent pathways. Since electrons are consumed at many steps, including the photoreduction of oxygen and recycling of reductants, the water–water cycle acts not only as an antioxidant system, but also as a system for dissipating excess electrons from PET, i.e. an electron sink (Asada 1999). A number of studies based on biochemistry and physiology have indicated the physiological significance of this cycle, as described in the sections below. Furthermore, the extremely fragile nature of chloroplastic APXs has been identified as the source of a bottleneck in the water–water cycle and, thus, affects the capacity of plants to tolerate stress (Shigeoka et al. 2002, Kitajima 2008). Consistent with this finding, a large number of stress-resistant plants have been genetically engineered by enhancing the H₂O₂-scavenging activity of chloroplasts (e.g. Shikanai et al. 1998, Yabuta et al. 2002). These findings suggest that loss-of-function mutants of chloroplastic APXs show severe growth defects, light sensitivity or even lethality in plants (Yabuta et al. 2002, Rizhsky et al. 2003). In spite of this plausible hypothesis, researchers have unexpectedly failed to identify such a ‘severe’ phenotype in the loss-of-function mutants of chloroplastic APXs, at least under laboratory conditions (Giacomelli et al. 2007, Kangasjärvi et al. 2008, Maruta et al. 2010, Caverzan et al. 2014).

Evidence to support the function of ROS as signaling molecules and their physiological significance in plant stress responses is accumulating (Mittler et al. 2011, Shigeoka and Maruta 2014). It is now accepted that there are production site- and kind-specific pathways for Reactive Oxygen Species (ROS) signaling that allow plants to respond rigorously and flexibly to stress conditions (Gadjev et al. 2006, Vaahtera et al. 2014). A previous study demonstrated that the conditional knockdown of chloroplastic APXs affects the expression of a large set of nuclear genes (Maruta et al. 2012), which may improve plant tolerance, at least to specific stress conditions such as heat and osmotic stress (Miller et al. 2007). These have provided a new insight into the role of antioxidant enzymes as signaling modulators through the spatio-temporal regulation of ROS levels.

Based on the authoritative review by Asada (1999), which clearly describes the mode of action of the water–water cycle, we herein focus on chloroplastic APXs in an attempt to elucidate the roles of these enzymes in plant stress responses. We start with a historical view of the occurrence of chloroplastic isoforms during plant evolution, and, from the point of view of the classical oxidative stress theory, then discuss their antioxidative functions by providing an overview of the findings of biochemical, physiological and genetic studies. In the last section, we focus on another aspect of chloroplastic APXs as signaling regulators by reviewing recent advances in chloroplastic H₂O₂-mediated stress responses.

Appearance of Chloroplastic APX Isoforms During Plant Evolution

APXs are heme peroxidases and members of Class I non-animal peroxidases, which also include cytochrome c peroxidases (CCPs) and bacterial catalase peroxidases (CPs) (Welinder 1992, Passardi et al. 2007). CCPs are present in photosynthetic and non-photosynthetic eukaryotes, whereas APXs are only found in plastid-containing organisms, with some exceptions (Teixeira et al. 2004, Passardi et al. 2007, Nedelcu et al. 2008). While cyanobacteria have no APX gene, which is supported by genome-sequencing studies (Passardi et al. 2007), most eukaryotic algae possess more than one APX gene (see below). A recent analysis using a large-scale phylogenetic tree of the whole peroxidase–catalase (non-animal peroxidase) superfamily suggested that monofunctional APXs are evolutionary descendants of atypical APX–CCP hybrid A1 peroxidases, which are found in non-photosynthetic kinetoplastids, such as Trypanosoma and Leishmania (Zámocký et al. 2014). Other hybrid A2 peroxidases found in photosynthetic euglenids, such as Euglena gracilis (Ishikawa et al. 2010), evolved in parallel with monofunctional APXs (Zámocký et al. 2014).

In order to obtain further insights into the evolutionary history of APXs, particularly chloroplastic isoforms, we herein examined the sequenced genomes of Archaeplastida, Stramenopiles, Alveolates, and Rhizaria (SAR) and Cryptophyta, Centrohelida, Telonemia, and Haptophyta (CCTH) (i.e. they include photosynthetic organisms) for the presence of APX gene(s). The methods used for the genome search and subsequent alignment analysis are described in the footnote of Supplementary Table S1; it is important to note that a large number of APX-like proteins, which have deletion(s) or substitution(s) in amino acid(s) essential for APX activity, were removed (e.g. Arabidopsis APX4/TL29 and APX6). As summarized in Supplementary Table S1, we failed to find an APX gene in the porphyridiophyceae, glaucophytes, oomycetes, ciliates, chlorarachniophytes, foraminiferae or haptophytes, which was consistent with recent findings (Wheeler et al. 2015). One exception was that the chlorarachniophyte Bigelowiella natans was previously reported to possess an APX gene (Wheeler et al. 2015); however, the protein (BnaAPx01/2512) was removed because of a large N-terminal deletion. The other organisms analyzed here had more than one APX gene.

We constructed a phylogenetic tree using robust ‘monofunctional’ APX sequences, estimated their subcellular localization and analyzed them for the presence of a transmembrane domain (Fig. 1; Supplementary Table S2). As shown in Fig. 1, these sequences, except for that of the red alga Cyanidioschyzon merolae APX (Cme-APX), were roughly divided into four groups: chloroplastic, cytosolic, peroxisomal and red algal groups. Consistent with the findings of Teixeira et al. (2004), there was a divergence between the chloroplastic group and other groups. One of the most identifiable characteristics of typical chloroplastic APXs in higher plants is that they have two extra insertions (chloroplastic domains 1 and 2) compared with cytosolic and peroxisomal APXs (Teixeira et al. 2004). Both domains were detected in all APXs in the chloroplastic group, but not in those in the cytosolic, peroxisomal and red algal groups. Among the APXs in the chloroplastic group, rice Os-APX5 and Os-APX6 are targeted solely to mitochondria, and Os-APX7 and Os-APX8 to chloroplasts (Xu et al. 2013). In contrast, Arabidopsis sAPX (At-sAPX) is dual targeted to both organelles (Xu et al. 2013). Thus, this group includes not only chloroplastic APXs, but also mitochondrial and dual-targeted enzymes; however, in this review, we refer to them as ‘chloroplastic’ in order to prevent complexity. Most APXs in this group were successfully estimated to be located in chloroplasts and/or mitochondria (see Supplementary Table S2); however, in the case of some algal APXs, it is currently difficult to estimate their localization because these sequences do not start with methionine.

Fig. 1

Phylogenetic tree of APXs in plants and algae. APX sequences from Arabidopsis thaliana (At), Populus trichocarpa (Pt), Oryza sativa (Os), Zea mays (Zm), Nicotiana tabacum (Nt), Spinacia oleracea (So), Pisum sativum (Ps), Physcomitrella patens (Pp), Selaginella moellendorffii (Sm), Klebsormidium flaccidum (Kf), Chlamydomonas reinhardtii (Cre), Volvox carteri (Vca), Chlorella variabilis (Cva), Coccomyxa subellipsoidea C-169 (Csu), Ostreococcus tauri (Ota), Ostreococcus lucimarinus (Olu), Micromonas pusilla CCMP1545 (Mpu), Chondrus crispus (Ccr), Porphyridium purpureum (Ppu), Cyanidioschyzon merolae (Cme), Galdieria sulphuraria (Gsu), Galdieria partita (Gpa), Thalassiosira pseudonana (Tps), Phaeodactylum tricornutum (Ptr), Fragilariopsis cylindrus (Fcy), Ectocarpus siliculosus (Esi) and Guillardia theta (Gth) were used. The phylogenetic tree was generated by the Neighbor–Joining method with 1,000 bootstraps using MEGA6 software. A red asterisk indicates tAPX.

Very importantly, only APXs classified into the chloroplastic group were found in all unicellular green algae, diatoms and brown algae. In most cases, these organisms contained one APX, which is consistent with previous biochemical analyses showing that Chlamydomonas reinhardtii and Coccomyxa subellipsoidea C-169 (Chlorella vulgaris) have only one APX in their chloroplasts (Takeda et al. 1997, Takeda et al. 1998). As exceptions, the cryptophyte Guillardia theta and diatom Phaeodactylum tricornutum had two APXs, one of which was estimated to be in the cytosol, while the other was in chloroplasts (or possibly the endoplasmic reticulum), although both APXs were classified into the chloroplastic group (Supplementary Table S2). Thus, APXs in the chloroplastic group were widely distributed among all organisms having APX, except for red algae, which possessed a specific group of APXs having features of both cytosolic and chloroplastic APXs (Kitajima et al. 2002). In contrast, the cytosolic and peroxisomal groups of APXs were only found in the streptophytes, i.e. land plants and charophyta. These findings indicate that the first monofunctional APX occurred as the chloroplastic type in unicellular green algae. tAPX has a transmembrane domain in its C-terminus in order to attach to the membrane. Consistent with previous findings (Pitsch et al. 2010), we also failed to find the transmembrane domain of APXs in the unicellular algae examined, suggesting that APXs in the chloroplastic group act as stromal sAPXs in unicellular algae (if they are actually targeted to chloroplasts). Interestingly, we found that the charophyte Klebsormidium flaccidum, a filamentous terrestrial alga, clearly had not only sAPX (Kf-APX2), but also tAPX (Kf-APX3) (see Supplementary Table S2). Furthermore, using the PeroxiBase database (Fawal et al. 2013), we also found putative tAPXs in the charophytes Chara braunii and Nitella hyaline (CbraAPx01/12952, NhyAPx02/7740 and NhyAPx05/11147) (data not shown). These findings demonstrate that the earliest tAPX occurred in charophytes, which had an attached specific transmembrane region at the C-terminus of sAPX. Considering the general acceptance that the ancestor(s) of terrestrial plants are closely related to charophytes (Timme et al. 2012), the appearance of tAPX might be important for the transition of a charophyte(s) to the first land plants in order to cope with the harsh features of terrestrial environments, such as drought and high light intensities.

The Rapid Inactivation of Chloroplastic APXs Under Photooxidative Stress: Desertion Under Enemy Fire?

Despite their enzymological functions, chloroplastic APXs are extremely sensitive to H₂O₂ under low ascorbate levels (Chen and Asada 1989, Miyake and Asada 1996). The half-inactivation time of chloroplastic APXs is 15 s when the concentration of ascorbate is <10 µM, while that of the cytosolic enzyme is >40 min (Kitajima 2008). The molecular mechanism underlying this inactivation has been extensively examined, as reviewed by Kitajima (2008). Structural and biochemical analyses have revealed that the irreversible cross-linking of heme to the distal Trp41 and radical formation in Cys31 and Cys125 are crucial for this process. Importantly, these amino acids are generally conserved in the stable cytosolic isoform (Supplementary Table S2). A chloroplastic isoform-specific insertion (chloroplastic domain 2) moves a loop structure, which is in the vicinity of the propionate side chains of heme, away from the propionate side chains, which may facilitate the cross-linking process (Kitajima 2008). The H₂O₂ sensitivity of tobacco sAPX has been improved by triple mutations in such amino acids or by the absence of chloroplastic domain 2 (Kitajima et al. 2008, Kitajima et al. 2010).

The photooxidative stress-induced inactivation of chloroplastic APXs has been reported in higher plants in vivo (Miyake et al. 2006, Yoshimura et al. 2000). Together with a number of in vitro and in vivo studies, the finding that Arabidopsis vitamin C defective mutants contain chloroplastic APXs that exhibit low activities provides further evidence that low ascorbate levels trigger their inactivation in vivo (Conklin et al. 1997). Nevertheless, there have been difficulties with elucidating the exact relationship between the inactivation of these enzymes and ascorbate levels in vivo. For example, in intact tobacco chloroplasts, a moderate light intensity (250 µmol photons m^–2 s^–1) almost completely inactivates APXs within 30 min without any change in ascorbate levels or redox states (Miyake et al. 2006). Similar findings have been reported in spinach leaves (Yoshimura et al. 2000). Furthermore, the ascorbate pool size is actually enhanced by application of a high (or moderate) light intensity to leaf cells (Dowdle et al. 2007), particularly in chloroplasts (Zechmann et al. 2011). One possible explanation for this may be that if ascorbate levels are sufficient, the strong viscosity in stroma inhibits the diffusion of ascorbate for APX, making it easier for the smaller molecule H₂O₂ to attack the enzyme (Miyake et al. 2006). Alternatively, another unknown mechanism(s) may be involved in the inactivation process. tAPX activity in vivo was recently shown to be inactivated through direct phosphorylation by a specific kinase in wheat during pathogen infections (Gou et al. 2015). Although the site of phosphorylation in wheat tAPX is currently unknown, phosphoproteomic studies have successfully identified Arabidopsis tAPX and sAPX as phosphorylated proteins (e.g. Roitinger et al. 2015). A heme-containing APX-related (APX-R) protein is located in chloroplasts and mitochondria, in which it physically interacts with APX, possibly to modulate its activity (Lazzarotto et al. 2011).

Since the first monofunctional APX occurred in unicellular alga(e) as the chloroplastic type (see the previous section), this finding suggests that the enzyme is also highly susceptible to H₂O₂. As described above, C. reinhardtii and C. subellipsoidea C-169 have only one APX (Cre-APX and Csu-APX, respectively) in chloroplasts (Takeda et al. 1997, Takeda et al. 1998), both of which have been classified into the chloroplastic group. All three amino acids and chloroplastic domain 2 involved in enzyme inactivation are conserved in both algal APXs (Supplementary Table S2). One exception is that Cys125 is substituted with valine in Csu-APX; however, a single mutation in this amino acid has negligible effects on the H₂O₂ sensitivity of tobacco sAPX (Kitajima et al. 2008). Nevertheless, the half-inactivation times of Cre-APX and Csu-APX are 10 and 15 min, respectively (Takeda et al. 1997, Takeda et al. 1998); thus, these enzymes are at least 40-fold more stable than the chloroplastic enzyme in higher plants (with a half-inactivation time of 15 s) and slightly less stable than the cytosolic enzyme (40–60 min). This may be very important for their function in the scavenging of H₂O₂ because the concentration of ascorbate is very low in such green algae (for example, its intracellular concentration in C. reinhardtii is approximately 67 µM; Takeda et al. 1997), in contrast to that in higher plants (millimolar range).

In summary, although the exact mechanism responsible for the inactivation process in vivo remains controversial, the activities of chloroplastic APXs are rapidly inhibited under photooxidative stress, at least in higher plants. Since algal APXs are considerably stable, the fragile nature of the enzyme was acquired during plant evolution. This may represent an evolutionary trade-off between stability and the turnover rate for catalysis (see Kitajima 2008); however, according to the antioxidant theory, chloroplastic APX activity should be required under photooxidative stress. Chloroplastic 2-Cys peroxiredoxins (2-Cys Prxs), which have the ability to form an alternative H₂O₂-scavenging system in the water–water cycle, are also susceptible to H₂O₂ (Wood et al. 2003, Kitajima 2008). The sensitivity of bacterial 2-Cys Prxs to oxidative inactivation is markedly less than that of eukaryotic enzymes, which is very similar to that of APXs, thereby providing the ‘floodgate’ hypothesis that fragile characteristics maintain cellular H₂O₂ at very low levels under favorable conditions, but permit higher levels under stressful conditions that require H₂O₂ to function as a signaling molecule (Wood et al. 2003). The potential role of chloroplastic APXs as signaling regulators is discussed in the last section.

Reinforcement of the Fragile APX System in Chloroplasts: A Successful Way to Enhance Plant Stress Tolerance

Based on the fragile nature of chloroplastic APXs and the photooxidative stress theory, it is plausible that APX activity is a bottleneck in the water–water cycle and a good target for improving plant stress tolerance. The first approach was to overexpress an Escherichia coli catalase (encoded by KatE), which has relatively high affinity for H₂O₂ and does not require ascorbate for its reaction, in tobacco chloroplasts. The ectopic expression of the enzyme does not protect chloroplastic APXs from oxidative inactivation under a combination of drought and a high light intensity (Shikanai et al. 1998) or with methyl viologen treatment (Miyagawa et al. 2000), but protects thiol-modulated enzymes in the Calvin cycle, thereby mitigating the inhibition of photosynthesis. These findings indicate that the sensitivity of chloroplastic APXs to oxidative damage is markedly higher than that of thiol-modulated enzymes. The inhibition of D1 protein translation, which is one of the major causes of photoinhibition (Nishiyama et al. 2001), is also alleviated in transgenic tobacco under a combination of salt and a high light intensity (Al-Taweel et al. 2007). As a consequence, a strong resistant phenotype against photooxidative stress has been detected in transgenic tobacco plants (Shikanai et al. 1998, Miyagawa et al. 2000). The overexpression of tAPX itself has almost the same effects on photooxidative stress tolerance in tobacco (Yabuta et al. 2002). One exception is that unlike KatE, the overexpression of tAPX clearly suppresses its inactivation under photooxidative stress. A plausible explanation for this gap may be the difference in affinity for H₂O₂ between the overexpressed enzymes; the K_m values of spinach tAPX and KatE are 40 µM and 4 mM, respectively. Thus, the fragile nature of chloroplastic APXs may simply be compensated for by their overaccumulation, possibly by maintaining H₂O₂ in chloroplasts at very low concentrations. A comparison between wild-type and transplastomic tobacco chloroplasts, which overexpress H₂O₂-resistant Galdieria partita APX in stroma, showed that APX activity is required for water–water cycle-dependent electron transport (Miyake et al. 2006). These lines of evidence have strongly indicated that the inactivation of APXs is a significant cause of photooxidative damage from abiotic stress in plant cells. Consistent with this, genetic engineering of the chloroplastic antioxidant system has succeeded in enhancing tolerance against a wide range of abiotic stresses in various plant species (Foyer and Shigeoka 2011).

A Significant Gap Between the Clear-Cut Antioxidant Theory and Phenotype of the Loss-of-Function Mutant Under a High Light Intensity

Considering the findings as summarized above and the fact that APXs are plant-specific enzymes, it had been concluded that APX activity in chloroplasts is a key determinant for plant stress tolerance. Hence, the loss-of-function mutants of chloroplastic APXs had been expected to cause severe growth defects or even lethality in plants under illumination (Yabuta et al. 2002, Rizhsky et al. 2003) as observed in ascorbate-deficient mutants (Müller-Moulé et al. 2004). The first loss-of-function analysis of chloroplastic APXs was performed on a hexaploid wheat mutant S-SV8, which spontaneously originated from an inbred R-SV8 line (Danna et al. 2003). The S-SV8 line lacks tAPX-6B, one of the three tAPX genes, and possesses 60% of the tAPX activity of the R-SV8 line. Although the S-SV8 line shows no visible symptoms under normal light conditions, the biomass of the mutant is smaller than that of the wild type when grown under conditions with more than moderate light intensity (400 µmol photons m^–3 s^–1), suggesting the importance of tAPX for plant growth even under mild light stress conditions, as expected. However, light intensity affects the mutant biomass in a dose-independent manner. In addition, no significant difference has been observed in photosynthetic parameters between genotypes under the moderate light conditions. It is also important to note that the wheat mutant has a deletion on the distal region of chromosome 6BL (Danna et al. 2003). Thus, it currently remains unclear whether tAPX-6B is the only gene absent in the mutant and responsible for its phenotype.

Concerted efforts have been made to perform loss-of-function analyses using Arabidopsis knockout and knockdown lines. After the short-term application of a high light intensity (>1,000 µmol photons m^–2 s^–1 for up to 6 h), the accumulation of H₂O₂ and decrease in PET activity were found to be more pronounced in single and/or double sapx and tapx mutants than in the wild-type plants (Kangasjärvi et al. 2008, Maruta et al. 2010). Nevertheless, no obvious phenotypic difference was observed between these mutants and the wild type under the short- and even long-term application of a high light intensity (Giacomelli et al. 2007, Kangasjärvi et al. 2008, Maruta et al. 2010). The slight inhibition of growth was only noted in these Arabidopsis mutants when artificial oxidative stress (a methyl viologen treatment) was applied (Kangasjärvi et al. 2008). These findings indicate that chloroplastic APXs contribute, to some extent, to photooxidative stress tolerance, but are not essential in Arabidopsis. A similar conclusion was reached in a rice double mutant with the reduced expression of Os-APX7 (sAPX) and Os-APX8 (tAPX) (Caverzan et al. 2014). Thus, no robust evidence currently exists to support the genetic requirement of chloroplastic APXs for antioxidant defense against photooxidative stress tolerance.

Why can loss-of-function mutants survive under light stress conditions without (or with only minor) visible symptoms? Previous studies have suggested that the water–water cycle does not act as a major electron sink for PET (Driever and Baker 2011), implying that the negligible (or weak) phenotype of the mutants may simply be due to the low rate of production of O₂^–/H₂O₂ in chloroplasts. However, the significant accumulation of O₂^–/H₂O₂ has also been observed in chloroplasts under light stress in a dose-dependent manner (e.g. Fryer et al. 2003, Driever and Baker 2011). In line with this finding, chloroplast development is defective in Arabidopsis knockout mutants lacking Fe-SODs (FSD2 and FSD3), providing genetic evidence for a non-negligible amount of O₂^–/H₂O₂ being produced in chloroplasts (Myouga et al. 2008). It should be noted that only laboratory conditions have been tested for the role of chloroplastic APXs. Arabidopsis non-photochemical quenching 1 (npq1) and npq4 mutants, which cause critical substitution in the essential proteins for the xanthophyll cycle involved in the dissipation of excess light energy as heat, grow as do the wild type in the laboratory even under photooxidative stress conditions, while they show severe growth defects in the field (Külheim et al. 2002). In contrast, an Arabidopsis lesion stimulating disease 1 (lsd1) mutant that exhibits a runaway cell death phenotype is highly sensitive to high light in the laboratory, but not at all sensitive to variable light and temperature in the field (Wituszynska et al. 2013). Thus, the application of field conditions is of utmost importance to evaluate the physiological role of chloroplastic APXs in future.

Compensation by other antioxidant enzyme(s) represents a plausible explanation for the negligible phenotype of chloroplastic APX mutants in laboratory conditions. Chloroplasts contain other peroxidase, including typical and atypical peroxidase as well as glutathione peroxidases (GPXs) (Foyer and Shigeoka 2011). In contrast to animal GPXs, plant enzymes efficiently utilize thioredoxin as an electron donor (i.e. Prx activity), but exhibit no or very weak activity in the presence of glutathione (Foyer and Shigeoka 2011). A reverse genetic approach has revealed that Arabidopsis GPX1 and GPX7, chloroplastic isoforms, actually constitute an alternative route for the scavenging of H₂O₂ in the water–water cycle (Chang et al. 2009). A recent study reported growth retardation in a complete double mutant lacking 2CPA and 2CPB under a light intensity of 160 µmol photons m^–2 s^–1 and clearly visible stress symptoms (leaf bleaching) following exposure to a high light intensity, while the single tapx mutant showed the wild-type phenotype (Awad et al. 2015). The phenotype of 2cpa 2cpb was further facilitated by an additional defect in tAPX (i.e. a triple mutant). Based on these findings, it was concluded that 2-Cys Prxs are the predominant components of the water–water cycle, and that tAPX is only functional in the absence of 2-Cys Prxs (Awad et al. 2015). However, it is important to note that 2-Cys Prxs are multifunctional proteins and, thus, have the ability to act as peroxidases, molecular chaperones and possibly signal transducers through protein–protein interactions and/or the redox modification of proteins (Dietz 2011). Difficulties are currently associated with determining which type of enzyme has predominance in the water–water cycle through phenotype comparisons between mutants. It has been also proposed that cytosolic APX compensates for the lack of chloroplastic APXs by cross-compartment protection (Davletova et al. 2005). In contrast, an unexpected finding was obtained from an Arabidopsis apx1 tapx double mutant, which was more resistant to oxidative stress than the wild type (Miller et al. 2007). This is similar to the case of another double mutant lacking APX1 and catalase 2 (CAT2), a predominant H₂O₂-scavenging enzyme in peroxisomes, in which the DNA damage response is constitutively activated (Vanderauwera et al. 2011). The balanced expression of cytosolic APXs appears to be important for the acclimation of chloroplastic APX mutants to photooxidative stress.

Other mechanisms are also involved in regulating the production of ROS from photosynthesis. In addition to linear electron transport, cyclic electron transport (CET) around PSI via the proton gradient regulation 5 (PGR5)- and chloroplast NADH dehydrogenase-like (NDH) complex-dependent pathways greatly contributes to the formation of a proton gradient across the thylakoid membrane (i.e. low pH in the lumen), which activates the xanthophyll cycle (Shikanai 2014). It has not yet been established whether these pathways contribute to the alleviation of oxidative damage in chloroplastic APX mutants.

Alternative Role of Chloroplastic APXs in Fine-Tuning Chloroplast-Derived H₂O₂ Signaling

We have discussed the role of chloroplastic APXs in antioxidant defense based on the concept that ROS are cytotoxic molecules. In contrast, it has been widely accepted that ROS also function as signaling molecules essential for modulating plant stress responses and growth and development (Foyer and Shigeoka 2011, Mittler et al. 2011). Production site- and kind-specific pathways have been described for ROS signaling (Gadjev et al. 2006, Vaahtera et al. 2014, Shigeoka and Maruta 2014). Although the mode of action of each pathway remains largely unclear, the integration and cross-talk of multiple pathways have been considered for plants to fine-tune stress responses. Therefore, the antioxidant networks located in various cellular compartments may play a key role in integration and cross-talk processes through the spatio-temporal regulation of ROS signal availability in cells. Under these conditions, chloroplastic APXs regulate the levels of a retrograde H₂O₂ signal from the organelles to the nucleus in order to modulate plant stress responses, which may provide an explanation for why the loss-of-function mutants of chloroplastic APXs show a negligible phenotype under photooxidative stress. In the last section, we discuss the potential role of chloroplastic APXs as signaling regulators in plant stress responses.

Transcriptional reprogramming is one of the significant outputs of ROS signaling (Gadjev et al. 2006). Microarray data using an Arabidopsis sapx tapx double mutant showed that when the threshold for responses was set to a 1.5-fold change in expression from that in the wild type, 286 genes were significantly (P < 0.05) up- or down-regulated in the sapx tapx mutant under normal and/or high light intensity conditions (Kangasjärvi et al. 2008). However, when the threshold was set to a 2-fold change, only 46 genes were differentially expressed in the mutant. Therefore, the constitutive lack of both APXs appears to have only a minor influence on nuclear gene expression. It is important to note that stress responses are generally transient processes. If this is applied to the case of the sapx tapx mutant, the transcriptome in the double mutant may only reflect the end of the acclimation process. In an attempt to overcome this issue and investigate the earlier stages of stress responses, a conditional system for tAPX silencing in Arabidopsis has been developed using an estrogen-inducible RNA interference (RNAi) method (Maruta et al. 2012). The expression of tAPX was almost zero 48 h after an estrogen treatment, resulting in the enhanced oxidation of chloroplastic proteins under normal light conditions, which indicates that chloroplastic H₂O₂ levels increased in response to tAPX silencing without the application of any stress. tAPX silencing itself has no effect on photosynthesis, the levels and redox states of antioxidants (ascorbate and glutathione), or growth and development, suggesting that enhancements in H₂O₂ levels in tAPX-silenced plants were not high (i.e. not cytotoxic). Nevertheless, 365 and 409 genes were at least 2-fold (P < 0.05) up- and down-regulated in tAPX-silenced plants, respectively, and are referred to as Responsive to tAPX Silencing (RTS) genes (Maruta et al. 2012). These findings indicate that even small amounts of H₂O₂, which are efficiently scavenged by tAPX in the wild type, markedly influence the expression of a number of genes.

One of the interesting features of RTS genes is that they rarely include typical marker genes for oxidative stress, which have been identified by comparing the transcriptomic data of several ROS-related mutants and plants treated with ROS-producing agents (Maruta et al. 2012). For example, RTS genes contained no antioxidant defense gene, such as cytosolic APX2, one of the representative marker genes. Rather, the activation of APX2 expression under a high light intensity is inhibited in the tapx mutant (Maruta et al. 2010, Maruta et al. 2012). In addition, RTS genes only slightly overlapped with genes whose expression was affected by peroxisomal H₂O₂ (i.e. in the cat2 mutant) (Queval and Foyer 2012). These findings suggest that the signaling function of chloroplastic H₂O₂ differs from that of the ROS produced from other cellular compartments as well as that of other kinds of ROS. A similar conclusion was recently provided by an alternative approach, in which a photorespiratory glycolate oxidase (GO) was ectopically overexpressed in Arabidopsis chloroplasts (GO plants) to produce H₂O₂ in a conditional manner (Sewelam et al. 2014). It remains an open question how chloroplastic H₂O₂ interacts with other ROS signaling pathway(s). An antagonistic cross-talk between chloroplastic H₂O₂ and ¹O₂ was proposed on the basis of the finding that ¹O₂-evoked gene expression and programmed cell death are significantly affected by tAPX overexpression, which reduces H₂O₂ availability (Laloi et al. 2007).

Classification of RTS genes and the identification of oxidative stress-sensitive and insensitive mutants from the T-DNA insertion lines of these genes have indicated a regulatory role for tAPX in metabolic pathways related to abiotic stress acclimation in plants. One of these mechanisms is the biosynthesis of anthocyanins, which protect cells from photoinhibition and photodamage through the absorption of excessive solar radiation (Chalker-Scott 1999). tAPX silencing transiently activates the expression of genes encoding the transcription factors PRODUCTION OF ANTHOCYANIN PIGMENT1 (PAP1) and PAP2, which are required for the biosynthesis of anthocyanins, in parallel with the induction of FAH1, resulting in the accumulation of anthocyanins without the application of any stress (Maruta et al. 2014). Consistent with this finding, the accumulation of anthocyanins under photooxidative stress is also pronounced in Arabidopsis mutants lacking tAPX (Miller et al. 2007). This may protect chloroplastic APX mutants from photooxidative stress. A role for tAPX in the regulation of γ-aminobutyric acid (GABA) metabolism, which may be linked to the oxidative stress response, has been also proposed (Maruta et al. 2013).

It has been demonstrated that one of the significant roles of chloroplastic H₂O₂ is associated with plant immunity. This may be linked to cross-talk between light acclimation and immunity (Mühlenbock et al. 2008), in which salicylic acid (SA), a phytohormone crucial for immune responses (Fu and Dong 2013), plays a central role. There are growing lines of experimental evidence for the involvement of tAPX in plant immunity. For example, tAPX silencing enhanced the accumulation of SA and subsequent transcriptional activation of defense-related genes without the application of any stress (Maruta et al. 2012). In addition, the overexpression of tAPX suppresses pathogen-induced programmed cell death (Yao and Greenberg 2006). More direct evidence was recently provided by the chloroplastic protein kinase WHEAT KINASE START1.1 (WKS1.1) enhancing cellular H₂O₂ levels upon pathogen attack in order to activate defense responses through the direct phosphorylation and inactivation of tAPX (Gou et al. 2015) (Fig. 2). These findings clearly indicate that chloroplastic APXs are involved in immune responses by regulating H₂O₂ levels within the organelle.

Fig. 2

Chloroplastic APXs function as antioxidant enzymes and signaling modulators. Under normal conditions, chloroplastic APX (chlAPX) maintains H₂O₂ at low levels to prevent oxidative damage in chloroplasts and, possibly, to suppress unwanted stress response. Abiotic and biotic stresses inactivate chlAPX, allowing for the flexible use of H₂O₂ for stress responses.

Perspectives

We herein overviewed evolutionary, biochemical, physiological and genetic studies on chloroplastic APXs, and concluded that these enzymes function not only as classical antioxidant enzymes, which prevent oxidative damage in plant cells, but also as modulators of a retrograde H₂O₂ signal in order to fine-tune abiotic and biotic stress responses (Fig. 2). Based on the widely accepted concept of ROS as signals, our conclusion may no longer be surprising; it is applicable to all enzymes/proteins involved in the metabolism of ROS. However, chloroplastic APXs, particularly tAPX, are still attractive targets for future ROS studies for the following reasons: (i) tAPX has been detected in charophytes, suggesting that its occurrence might be important for the transition of multicellular alga(e) to the first land plants; and (ii) the susceptibility of these enzymes to H₂O₂ acquired during plant evolution, which may have allowed for the flexible use of H₂O₂ as a signaling molecule in plants. These may be linked to the current concept for the evolution of ROS signaling functions in which plants first acquired ROS-scavenging mechanisms, developed the ability to control intracellular levels of ROS and then started to use these molecules for signaling purposes (Mittler et al. 2011). Along with this evolutionary history, chloroplastic APXs as well as other enzymes may have acquired another function as signaling regulators along with the diversity of ROS actions. In this regard, the following key questions remain to be addressed. Was the occurrence of tAPX essential for controlling intracellular levels of ROS? Was the occurrence involved in the terrestrialization of plants? When did tAPX (and sAPX) acquire susceptibility to H₂O₂ during evolution? Since chloroplastic H₂O₂ plays a key role in abiotic and biotic stress responses, are transgenic plants with enhanced H₂O₂-scavenging activity at a disadvantage? Furthermore, the evaluation of chloroplastic APXs as signaling modulators has just begun; thus, the current situation can serve as a starting point for future detailed analysis. Although more than two decades have passed since the discovery of chloroplastic APXs, there are now more questions than answers. One of the great motivators is the chloroplast redox system being strongly linked to a diverse physiological process (Foyer and Shigeoka 2011).

Funding

This work was supported by the Ministry of Education, Culture, Sports, Science, and Technology of Japan [Grant-in-Aid for Young Scientists (B) (to T.M: 23880018) and for Scientific Research (B) (to T.I: 24380186)].

Abbreviations

Abbreviations
 
• APX

  ascorbate peroxidase

 
• CCP

  cytochrome c peroxidase

 
• GPX

  glutathione peroxidase

 
• PET

  photosynthetic electron transport

 
• Prx

  peroxiredoxin

 
• ROS

  reactive oxygen species

 
• sAPX

  stromal APX

 
• SOD

  superoxide dismutase

 
• tAPX

  thylakoid membrane-attached APX

Disclosure

The authors have no conflicts of interest to declare.

Acknowledgments

We are grateful to Dr. Hironori Takasaki (The University of Tokyo/Ghent University) for his help in the phylogenetic analysis. We apologize to the authors and our colleagues whose work could not be cited owing to space limitations.