**2.3.4 Tetrapyrrole and metabolite signaling**

Pioneering experiments designed to understand plastid signaling were based on the application of norflurazon. This bleaching herbicide was found to be a potent experimental tool by its ability to profoundly disrupt chloroplast function, due predominantly by its strong inhibition of carotenoid biosynthesis. It was also shown to trigger the release of ROS upon illumination and prevent the light-dependent induction of nuclear photosynthesisrelated genes (Oelmuller & Mohr, 1986). The isolation and characterization of the so called *gun* mutants (for *genomes uncoupled*), most of which coded for proteins that are involved in tetrapyrrole biosynthesis, led some workers to suggest that tetrapyrrole intermediates serve as a plastid signal to regulate the expression of nuclear genes for photosynthetic proteins (Mochizuki et al., 2001; Larkin et al., 2003; Strand, 2003; Koussevitzky et al., 2007). However, the conclusive results derived from subsequent studies strongly suggested that tetrapyrrole pathway intermediaries are not directly linked to plastid signaling (Mochizuki et al., 2008; Moulin et al., 2008).

Conversely, the concept of metabolic signaling arose from the unlikeness that ROS or redox compounds themselves act as signaling molecules that traverse the cytosol (see above). In this context, messengers that are metabolically more inert and less readily inactivated during diffusion through the cell represent much more promising signaling candidates (Baier & Dietz, 2005). This alternative signaling pathway is justified on the informative nature regarding the metabolic state of the chloroplast that is contained in the relatively large exchange of primary metabolites, such as carbohydrates, or of the xanthophyll derivative abscisic acid (ABA) phytohormone, with the rest of the cell. In this regard, sugarsignaling pathways have been considered as likely candidates for the regulation of photosynthetic acclimation under various stress conditions. Moreover, the expression of nuclear encoded photosynthetic genes (e.g. *CAB2* and *rbcS*) was found to be inversely correlated with intercellular soluble sugar levels (Oswald et al., 2001).

Thus, one proposed scenario envisions that metabolite concentration changes could be sensed by cytosolic or nuclear receptors to regulate nuclear gene expression. One of these sensors could be the cytosolic hexokinase, known to be crucial for sensing and responding to glucose signals intra-cellularly (Figure 1). Alternatively, the transport of carbohydrates across the chloroplast membrane might directly communicate information on the redox state of the chloroplast by means of so called 'redox valves'. Two well-studied examples of carbohydrate shuttles that export reducing power from the chloroplast are the malate/oxaloacetate and triose-phosphate shuttles (Heineke et al., 1991). The 'malate valve' comprises the malate/oxaloacetate translocator (Taniguchi et al., 2002) together with the chloroplast and cytosolic isoforms of NAD(P)H malate dehydrogenase, and is thought to constitute the central mechanism for the export of excess reducing power from the chloroplast. The dihydroxyacetonephosphate/3-phosphoglycerate shuttle involves the triosephosphate translocator (TPT) (Flügge et al., 1989) that functions primarily as a dihydroxyl acetonephosphate/phosphate exchanger to maintain sucrose synthesis in the

*redox-imbalanced* (*rimb*) mutants, which were detected using an Arabidopsis reporter gene line expressing luciferase under control of the redox-sensitive 2-cysteine peroxiredoxin A (2CPA) promoter (Heiber et al., 2007). Valuable information shedding light on the nature of redox signaling should be expected when the identity of the RIMB genes and their

Pioneering experiments designed to understand plastid signaling were based on the application of norflurazon. This bleaching herbicide was found to be a potent experimental tool by its ability to profoundly disrupt chloroplast function, due predominantly by its strong inhibition of carotenoid biosynthesis. It was also shown to trigger the release of ROS upon illumination and prevent the light-dependent induction of nuclear photosynthesisrelated genes (Oelmuller & Mohr, 1986). The isolation and characterization of the so called *gun* mutants (for *genomes uncoupled*), most of which coded for proteins that are involved in tetrapyrrole biosynthesis, led some workers to suggest that tetrapyrrole intermediates serve as a plastid signal to regulate the expression of nuclear genes for photosynthetic proteins (Mochizuki et al., 2001; Larkin et al., 2003; Strand, 2003; Koussevitzky et al., 2007). However, the conclusive results derived from subsequent studies strongly suggested that tetrapyrrole pathway intermediaries are not directly linked to plastid signaling (Mochizuki et al., 2008;

Conversely, the concept of metabolic signaling arose from the unlikeness that ROS or redox compounds themselves act as signaling molecules that traverse the cytosol (see above). In this context, messengers that are metabolically more inert and less readily inactivated during diffusion through the cell represent much more promising signaling candidates (Baier & Dietz, 2005). This alternative signaling pathway is justified on the informative nature regarding the metabolic state of the chloroplast that is contained in the relatively large exchange of primary metabolites, such as carbohydrates, or of the xanthophyll derivative abscisic acid (ABA) phytohormone, with the rest of the cell. In this regard, sugarsignaling pathways have been considered as likely candidates for the regulation of photosynthetic acclimation under various stress conditions. Moreover, the expression of nuclear encoded photosynthetic genes (e.g. *CAB2* and *rbcS*) was found to be inversely

Thus, one proposed scenario envisions that metabolite concentration changes could be sensed by cytosolic or nuclear receptors to regulate nuclear gene expression. One of these sensors could be the cytosolic hexokinase, known to be crucial for sensing and responding to glucose signals intra-cellularly (Figure 1). Alternatively, the transport of carbohydrates across the chloroplast membrane might directly communicate information on the redox state of the chloroplast by means of so called 'redox valves'. Two well-studied examples of carbohydrate shuttles that export reducing power from the chloroplast are the malate/oxaloacetate and triose-phosphate shuttles (Heineke et al., 1991). The 'malate valve' comprises the malate/oxaloacetate translocator (Taniguchi et al., 2002) together with the chloroplast and cytosolic isoforms of NAD(P)H malate dehydrogenase, and is thought to constitute the central mechanism for the export of excess reducing power from the chloroplast. The dihydroxyacetonephosphate/3-phosphoglycerate shuttle involves the triosephosphate translocator (TPT) (Flügge et al., 1989) that functions primarily as a dihydroxyl acetonephosphate/phosphate exchanger to maintain sucrose synthesis in the

correlated with intercellular soluble sugar levels (Oswald et al., 2001).

biochemical function is determined.

Moulin et al., 2008).

**2.3.4 Tetrapyrrole and metabolite signaling** 

cytosol, but that could also export both ATP and NADPH from chloroplasts into the cytosol (Figure 1). Interestingly, the altered nuclear gene expression detected in the *tpt* (Biehl et al., 2005) and *cue1* mutants, the latter affected in the phosphoenolpyruvate/phosphate translocator (PPT) in the inner chloroplast envelope (Streatfield et al., 1999) supports the proposed role played by metabolite exchange in redox signaling .

Finally, the role of ABA as a signal relying information on the chloroplast status to the nucleus can be explained by evoking the multiple effects that photosynthetic activity rates have on the biosynthetic ABA pathway that is partially localized in the chloroplast (Baier & Dietz, 2005). For instance, oxidative stress conditions leading to increased ABA through inductive effects on the synthesis of its xanthophyll precursor in the chloroplast, could provide a link between the redox state and ROS levels in the plastid and gene expression in the nucleus. In this respect, it has been speculated that the repressed photosynthetic gene expression produced by norflurazon treatment might be associated with reduced levels of ABA resulting from a depressed carotenoid biosynthesis (Kleine et al., 2009).

Fig. 1. Proposed plastid signal transduction pathways (redox, *metabolic* and ABAdependent) involved in the regulation of acclimation responses to stress, including state transitions, long-term responses (LTR) and/or the activation/ repression of specific target genes in the chloroplast and nucleus. Redox signals are generated within the electron transport chain (purple) or by generation of reactive oxygen species (ROS) (black). The plastoquinone pool (PQ/PQH2, in purple letters) is the origin for at least two redox

Friend or Foe? Exploring the Factors that Determine the Difference

Between Positive and Negative Effects on Photosynthesis in Response to Insect Herbivory 169

processes, in addition to abiotic and biotic stress tolerance responses (Kleine et al., 2009; Tognetti et al., 2011). For instance, the local and systemic acclimation in Arabidopsis leaves in response to excess excitation energy associated with cell death and regulated by specific redox changes of the PQ pool, also caused a rapid decrease of stomatal conductance, global induction of genes involved in ROS scavenging and pathogen resistance, increased ROS production and enhanced ethylene signaling. In addition, evidence was provided which showed that multiple hormonal/ROS signaling pathways not only regulate the plant's response to excess excitation energy, but also control induced systemic acquired resistance and basal defenses to virulent bacterial pathogens. The balanced activity of the disease resistance and signaling-related proteins coded by the *LSD1*, *EDS1*, *PAD4*, and *EIN2* genes was found to be necessary to regulate the steps leading to programmed cell death, light acclimation, and defense responses that are initiated, at least in part, by redox changes of the PQ pool (Mühlenbock et al., 2008). Further evidence coupling chloroplast-controlled disease resistance with ROS accumulation was obtained with the Arabidopsis mutant *rph1* (for *resistance to Phytophthora 1*), which was found to be susceptible to the pathogen *Phytophthora brassicae* as a consequence of a reduced oxidative burst, a runaway cell-death response, and failure to properly activate the expression of defense-related genes. The finding that the *RPH1* gene encodes an evolutionary highly conserved chloroplast protein was in accordance with a prominent chloroplast-dependent role in the activation of immune responses to

*Phytophthora*, not only in Arabidopsis but in potato, as well (Belhaj et al., 2009).

signals (Dietz et al., 2010).

In addition, signal integration at the level of transcription factor (TF) activation appears to be majorly controlled by the family of APETALA 2/ ethylene response element binding protein (AP2/EREBP) TFs, which are abundantly represented in Arabidopsis, poplar and rice (Dietz et al., 2010). By dint of their activation of different innervating pathways, or their ability to bind to multiple target elements, AP2/EREBP TFs are known to integrate several signaling inputs. A couple of examples are the ERF1 TF, which is controlled by ethylene and JA (Lorenzo et al., 2004), and the dehydration-responsive element binding TF TINY, that connects abiotic stress signaling via DRE-dependent regulation to biotic stress signaling via ethylene response elements (Sun et al., 2008). It is considered that combinatorial target gene regulation by different signals may involve different mechanisms, including: (i) cross-talk in the signaling pathways; (ii) stimuli-dependent TF activation, e.g. by homo- or heterotypic dimer and oligomer formation, respectively; (iii) competition for the same or binding to different cis elements; and (iv) amplification cascades that can be modulated by interfering

**3.2 The negative effect of abiotic and biotic stress on photosynthetic gene expression**  The down-regulation of photosynthetic gene transcription is frequently observed in plants subjected to stress. Thereby, environmental stresses, including drought, salinity and low temperatures can negatively affect photosynthetic gene expression in addition to an induction of compensating physiological and biochemical alterations (Saibo et al., 2009; Chaves et al., 2009). Similarly, a whole gamut of biotic insults caused by arthropods, fungi, bacteria and viral pathogens triggers a uniform and apparently regulated reduction in transcription of nuclear genes coding for the major components of photosynthesis, regardless of the plant host. The widespread negative effect on photosynthesis caused by biotic stressors was recently evidenced by a meta-genomic analysis in which the transcriptome data from microarray experiments representing twenty two different forms of

signaling pathways that are active under low or high light. These can lead to rapid state transitions, involving reversible association of the mobile pool of LHCII to PSI or PSII, or LTR. Qo (red oval) represents the docking site where plastoquinol binds to the cytochrome *b6f* complex. Both processes are dependent on the redox-regulated activity of the STN7 kinase (possibly in conjunction with STN8; purple letters). The TAP38 phosphatase (in red letters) regulates state transitions by specifically dephosphorylating LHCII. On the other hand, LTR involves chloroplast and nuclear gene expression (purple arrows). LTRdependent plastid gene expression is believed to be regulated by the *Chloroplast Sensor Kinase* (CSK, in purple letters), while nuclear gene expression might require a putative LTR signal (in purple box). LTR-dependent changes in gene expression and protein accumulation can also lead to the establishment of two distinct metabolic states (purple letters, in *cytoplasm*) derived from the exposure to either PSI or PSII light (yellow letters and arrows). ROS are generated by transfer of electrons from PSI or reduced ferredoxin (Fd) to oxygengenerating superoxide (O2-, in black letters). This is detoxified by superoxide dismutase (SOD, in black letters) to hydrogen peroxide (H2O2; in black letters). Un-scavenged H2O2 might able to diffuse freely across the chloroplast envelope or through water channels (or aquaporins, A; blue letter in blue circle) and is thought to start MAP kinase cascades in the cytosol. Singlet oxygen (1O2; in black letters), is generated at PSII. Its high reactivity and short half-life require additional signaling components, such as Executer 1 and 2 (EX1, EX2, in black letters), although evidence in green algae suggests that 1O2 might be able to diffuse out of the chloroplast*. Metabolic* plastid signaling has been proposed to require the activity of the malate/oxaloacetate (MOT; in green letters) and triosephosphate (TPT; in green letters) translocators needed for the export of excess reducing power and ATP from the chloroplast. Alternatively, metabolite concentration changes (e.g., glucose originated from the Calvin cycle; green letters) could be sensed by cytosolic or nuclear receptors (i.e. hexokinase; HK, green letters in yellow circle) to modify nuclear gene expression. Abscisic acid (ABA, in orange letters) whose synthesis is derived from ROS-sensitive xanthophyll precursors (XP, in orange letters) in the chloroplast, has also been proposed to act as a signal relying information on the chloroplast status to the nucleus. Dotted arrows represent pathways mediated by unknown components that have not been entirely elucidated. Tetrapyrrole and ROS-scavenging-associated signaling pathways are not shown (Adapted from Kleine et al., 2009, Pfannschmidt et al., 2009, and Pesaresi et al., 2010).
