ROS and Cell Redox Control and Interaction with the Nuclear Gene Expression
The discovery of multisensory systems is important for understanding the basic question how perception and transmission of stress signals operate in large cell cultures and tissue. As is shown in the example of the multisensory protein Hik33 in Synechocystis one protein is able to react to a variety of different stress signals (chap. 4.2). Hik33 can sense Redox stress but additionally salt and cold stress and all lead to the induction of similar responsive patterns. Multisensory systems become more complicated in plants than in cyanobacteria. Apparently, changes in membrane fluidity, regardless of the nature of the stress effects, are a signal that is perceived by sensory histidine kinases or ion channels localized in the membranes (Los et al., 2013). The ROS signals are transduced by rather macroscopic structures like membranes that regulate single molecules like Hik signaling directly to the molecular fundament of each cell, the genome. In such way, involving histidine kinases as one example ROS signaling covers a large hierarchy from top down and bottom up mechanisms involving the genome, the proteome, cell tissues and the whole organism. Bottom up ROS can oxidize EF-G (see chapter 4.1) and translation of genes is stopped. Top down the ROS level determines membrane rigidity and therefore the activity of histidine kinases.
Structural modifications give rise to detachment of weakly bound enzymes from the cell wall, as shown for the isoforms of peroxidases (Minibayeva and Gordon, 2003). Concomitantly, ROS modulate the activity of antioxidant enzymes, e.g. in case of catalase and ascorbate peroxidases (Shao et al., 2008). Likewise, also the redox state of cell components acting as antioxidants or being involved in signaling (glutathione system, ascorbate system, plastoquinone pool, thioredoxin, etc.) is prone to changes by ROS. Furthermore, the activity of ion channels can be affected, going along with variations in the concentration of relevant ions like Ca2+ in the cytosol.
The by far most important regulatory control in acclimation of organisms to different stress factors is the modulation of gene expression. ROS in general and H2O2 in particular play an important role in cell signaling pathways and are involved in the regulation of gene expression (Apel and Hirt, 2004; Laloi et al., 2007). Studies using DNA microarrays (Gadjev et al., 2006; Scarpeci et al., 2008) revealed that an increase of the ROS concentration affects the expression of a rather large number of genes. This response can sometimes comprise up to one third of the entire genome. Experiments performed with the unicellular green alga Chlamy- domonas (C.) reinhardtii showed that H2O2 and :AO2 interact with different targets leading to activation of specific promoters (Shao et al., 2007).
With respect to regulatory and signaling effects of ROS including the function of second messengers, the general control of cellular processes by the ambient redox conditions should be pointed out clearly as a generalized working framework of the cell chemistry rather than ROS taking the role of “isolated” signal molecules (Foyer and Noctor, 2005; Pfannschmidt et al., 2009; Shao et al., 2008; Buchanan and Luan, 2005). Several models have been proposed for this redox control which includes oxidation/reduction of thiol groups, iron-sulfur centers, hemes, and flavin (Foyer and Noctor, 2005; Vranova et al., 2002). The redox homeostasis in cells is mainly controlled by the presence of large pools of the thiol buffer glutathione and of NADPH/NADP+ and also by high concentrations of ascorbic acid (Foyer and Noctor, 2005). The fraction of reduced glutathione is normally higher than 90% (Noctor et al., 2002). Concomitant with these hydrophilic compounds, tocopherols can func?tion as a lipophilic redox buffer system (for the antioxidant efficiency of different tocopherol species, see (Krumova et al., 2012). These systems protect lipids and other membrane components of chloroplasts by physical scavenging and chemical interaction with ROS (Krumova et al., 2012).
The main sources of ROS in plant cells are still the chloroplasts, where ROS are produced as the so called “waste product” during photosynthesis. Within the chloroplasts primarily the components of the photosynthetic ETC produce ROS (Shao et al., 2007). Several possible sources are known for chloroplast signals, which can affect gene expression in the nucleus of the plant cells (Galvez-Valdivieso and Mullineaux, 2010; Buchanan and Luan, 2005). These sources include the biosynthetic pathway of tetrapyrrole compounds, changes in the redox state of photosynthetic ETC components (e.g. the PQ pool) and ROS generation. All these pathways induced by different signals are interconnected and therefore often considered as tightly coupled. At present, only rather limited information is available on the exact mechanisms of the transduction of signals from the chloroplast to the nucleus due to the accumulation of redox-sensitive compounds and ROS in the chloroplasts (Fey et al., 2005a,b; Pogson et al., 2008; Shao et al., 2008, Pfannschmidt et al., 2009; Kreslavski et al., 2011; Schmitt et al., 2014a).
One of the key sensors in adaptation of the ETC to light is the redox state of the PQ pool which regulates the phosphorylation of lightharvesting complexes II (LHC II) (Vener et al., 1998) and also acts as a signal for regulation of the expression of a set of plastid and nuclear genes (Pfannschmidt et al., 2003), such as Lhcb, petE, APX2, and ELIP2 encoding light-harvesting complex proteins, plastocyanin, ascorbate peroxidase 2 and early light inducible protein, respectively. Likewise, the expression of SOD is affected (Shaikhali et al., 2008). It should be noted that the expression of Lhcb genes is only partially controlled by the redox state of PQ, because additional factors are involved like ATP synthesis and the electric potential difference across the thylakoid membrane (Yang et al., 2001).
The redox state of PQ is proposed to induce two signaling pathways, which are initiated under the influence of high and low light and subsequently activate the expression of plastid and nuclear genes (Fey et al., 2005a,b; Pfannschmidt et al., 2009). ROS arising from reactions at the acceptor side of PSII could be one type of signals which trigger the regulation of these pathways (Ivanov et al., 2007). Under high light stress of A. thaliana, the PQ pool was shown to be oxidized by both :AgO2 and less electron input from PS II due to the effect of NPQ with implications on redox signaling (Kruk and Szymanska, 2012). Furthermore, a plastid terminal oxidase (PTOX) leading to slow PQH2 oxidation is probably involved in a ROS-triggered signal transduction cascade (Trouillard et al., 2012). The underlying mechanism of the specific role of PTOX in acclimation of plants to high light remains to be clarified.
The expression of genes encoding PSI proteins (psaD and psaF) is also affected (Pogson et al., 2008). Changes in the redox state of components on the PSI acceptor side contribute to the regulation of nuclear and chloroplast genes (Shaikhali et al., 2008). This effect is primarily related to the redox state of thioredoxin, which depends on the rate of electron transport from ferredoxin (Scheibe et al., 2005). The redox states of thioredoxin, glutathione and glutaredoxin act as signals for regulation of stress-responsive genes (Mullineaux and Rausch, 2005; Schurmann and Buchanan, 2008).
Several examples of regulation by thiols, in particular chloroplast gene translation and transcription have been described in recent reviews (Oelze et al., 2008). In addition to signaling modes where ROS generated in response to different types of stress act either directly as signal molecules, or, via generation of second messengers (e.g. oxidation products of Cars and lipids, vide supra), also pathways operating in the opposite direction are established in plants. This does not only account for producing ROS as defensive molecules but it also accounts for the active production of ROS as messenger molecules. In both these cases, ROS are produced as second messengers or as reactive species in response to stress, e.g. in the defense to biotic infection.
Special proteins are involved in the development of cell response to changes of the redox state. These proteins are encoded by so-called reporter genes. Investigations on redox signaling between chloroplasts and nucleus have been focused on the induction of genes of cytosolic ascorbate peroxidases APX1 and APX2, genes ZAT10 and ZAT12 encoding zinc-finger transcription factors, and also gene ELIP2 encoding the early light-induced chlorophyll-binding protein ELIP2. Both transcription factors, ZAT10 and ZAT12, favor the induction of gene clusters related to activation of the photosynthetic ETC under high light by switching on the expression of genes APX1 and APX2 (Davletova et al., 2005; Pogson et al., 2008). The expression of genes APX2, ZAT10, and ZAT12 is stimulated by treatment with H2O2 (Karpinski et al., 1999; Davletova et al., 2005; Pogson et al., 2008) It seems reasonable to assume that H2O2 regulates the expression of these genes, thus acting via direct or indirect effects of the redox state signaling from chloroplast through cytosol to the nucleus. Kinetic experiments revealed that the redox signal arising from the response to a change in light quality (spectral composition) is transmitted within about 30 min from the chloroplast to the nucleus (Zhang et al., 2000).
Studies on high-light effects in A. thaliana plants showed that the nuclear genes encoding cytosolic peroxidases APX1 and APX2 were active- ted in 15-20 min. It was also found that the activation of these genes is part of the systemic response to superfluous light exposure (Karpinski et al., 1999). The induction of chloroplast gene expression occurred also in the range of 15-20 min in response to changes of the redox state in the organelles induced by changes in light quality (Pfannschmidt et al., 2009). These kinetic results on signal transduction suggest that some components of the signal cascade triggered by stress-induced changes of ROS concentration are present in the cells already under optimum conditions and do not need to be synthesized in response to stress. This idea explains the correspondence of the kinetics of signal transduction in response to ROS appearance and of the transduction of other intracellular stimuli (Pfannschmidt et al., 2009).
A genetic screen aimed at identifying :AgO2-responsive genes (Baruah et al., 2009) led to the proposal of the gene named “pleiotropic response locus 1” (PRL1) acting as a point of convergence of several different signaling pathways, thus integrating various intra- and extracellular signals. Under pathogen-induced stress, the gene “enhanced disease susceptibility 1” (EDS1) plays a role in development of the hypersensitive reaction and in mediating EXE1/EXE2-regulated cell death induced by :AgO2 (Ochsenbein et al., 2006). The EDS1 protein has been shown to be required for the resistance to biotrophic pathogens and the accumulation of SA. SA likely enhances the plant defenses by inducing the synthesis of pathogen-related proteins (Mullineaux and Baker, 2010). EDS1 seems to play a pivotal role in a mutually antagonistic system, integrating ROS signals from chloroplasts in cells suffering from photooxidative stress (Straus et al., 2010).
A general problem in identifying different 1AgO2-induced signal pathways and their (synergistic) interplay has to be mentioned. The effect on the gene expression pattern is expected to depend on the nature of the nearest neighborhood of :AgO2 formation, if one accepts that signaling directly by :AgO2 can take place only at a site very close to its generation. This would also imply that the signal pathway comprises the participation of oxidation products of Cars, lipids and other molecules acting as second messengers which can induce different genetic responses. The :AgO2 site differs in WT plants and in mutants like flul and, also, if :AgO2 is generated by using exogenous sensitizers (Hideg et al., 1994; Krieger-Liszkay, 2005). Therefore, different types of second messenger species are likely to be formed in mutant studies in contrast to wild type studies. Thus, it is difficult to gather straightforward conclusions on the mechanism of 1AgO2 signaling from studies performed under different assay conditions and using different sample material including single gene mutants. This important problem needs to be further addressed in forthcoming studies.
ROS can be involved in several signaling pathways by modulating the activity of different components like MAPKs and phosphatases, transcription factors, and calcium channels (Pei et al., 2000; Mori and Schroeder, 2004; Pfannschmidt et al., 2009; Kreslavski et al., 2011; Pogson et al., 2008; Kovtun et al., 2000; Gupta and Luan, 2003). Heterotrimeric G-proteins may also participate in the signaling pathways initiated by ROS (Joo et al., 2005). An effect of ROS on the activity of Ca2+ channels was shown to arise for both abiotic stress and plant-pathogene interaction (Demidchik et al., 2003).
Very few details have been resolved so far on the nature of steps which link various pathways in coordinating ROS signaling. One piece of the puzzle is the finding that MAPKs are involved in transducing signals derived from ROS generated by sources in chloroplasts (Liu et al., 2007).
One mode of ROS-induced signaling is given by the activation of transcription factors containing SH groups like OxyR in eubacteria and/or iron-sulfur clusters (Zheng et al., 1998). Formation of S-S bridge(s) by H2O2 is expected to change the structure of OxyR, thereby inducing the transition from the inactive into the active form, as is schematically illustrated in Figure 69.
But another possibility of ROS signaling changes of the subcellular distribution of these factors is seen in the case of yeast. Yeast cells express the protein Yap1, which is functionally homologous to the transcription factor OxyR in eubacteria. Yap1 can regulate the transcription of specific genes in response to changes of the redox state of the cell (Liu et al., 2005). The inactive form of Yap1 is localized in the cytoplasm. H2O2 oxidizes Yap1 via the peroxidase Gpx3 (Delaunay et al., 2000) under formation of disulfide bonds between neighboring cysteines, thus leading to conformational changes, which enables the transport of this Yap1 form to the nucleus, where it induces the expression of genes encoding for components of antioxidant defense system(s). Mutants lacking Yap1 were shown not to be able to induce antioxidant defense upon treatment with H2O2 (Liu et al., 2005).
In analogy to the thiol-based sensor OxyR of bacteria, Yap1 is part of a relatively simple regulatory loop, where ROS induce the expression of certain antioxidative enzymes. Although a gene homologous to OxyR
is absent in higher plants, attempts to complement mutations of the gene OxyR in E. coli by using the expression gene library of A.thaliana identified the AnnAtl gene encoding annexin as to be capable of restoring a functional defect in the OxyR bacterial mutant (Gidrol et al., 1996). Recently identified signaling proteins undergoing thiol modulation (modification) in plants include a protein tyrosine phosphatase (Dixon et al., 2005) and a histidine kinase ETR1, which is involved in ethylene signaling (Desikan et al., 2005).
Figure 69. Hypothetical scheme of regulation of bacterial transcription factor OxyR activity. The inactive form contains thiol groups (SH). Under the influence of H2O2, the thiol group is oxidized with the formation of an SOH group and then rapid formation of a disulfide bond occurs and OxyR transits into its active form.
Based on these findings, it seems reasonable to assume that analogous mechanisms also exist in plant cells, but they are likely to be more complex. Figure 70 presents a proposed simplified scheme for redox-sensitive sensors (RsS) acting as primary sensors of H2O2 signal transduction.
The signal can be transmitted directly from H2O2 or via RsSs to the MAPK cascade and/or to transcription factors (Neill et al., 2002; Pogson et al., 2008; Pfannschmidt et al., 2009). The conformation and activity caused by reversible oxidation of cysteine residues of regulatory proteins, which are involved in gene expression at different developmental stages, offer a simple and elegant mechanism for regulation of transcription and translation systems under oxidative stress. As a result, transcription of nuclear genes required for ROS scavenging is activated. By oxidation of the translation elongation factor G (EF-G) in chloroplasts and by blocking translation of new proteins, H2O2 can also regulate gene expression on the level of translation, in particular (see chap. 4.1 for more details) (Nishiyama et al., 2011; Murata et al., 2012).
Figure 70. Effects of H2O2 on processes of transcription and translation. RsS, MAPK and TF are redox-sensitive sensor(s), MAP-kinase and transcription factor(s), respectively.
Figure 71 presents a hypothetical scheme of pathways of photosynthetic redox signal transduction in plants. It summarizes selected mechanisms described in the chap. 4.1, 4.2 and 4.3 in a general picture that aims to denote the complex networking of different species to establish ROS-iniated and ROS-mediated signaling pathways between cell organelles.