Peroxisomal plant metabolism – an update on nitric oxide, Ca²⁺ and the NADPH recycling network

Peroxisomes were first described at the cellular level in animal and plant organs in the 1950s and 1960s (Rhodin, 1954; Frederick et al., 1968), when their biochemical composition also began to be analyzed. They have been shown to have a basic oxidative metabolism, characterized by high hydrogen peroxide (H₂O₂) content and the presence of catalase (De Duve et al., 1960; Tolbert et al., 1968). This enzyme, which is exclusively found in peroxisomes, is considered to be one of the main antioxidant enzymes and part of the first line of defense against oxidative stress. Subsequent advances in the study of these organelles have revealed that they are extraordinary subcellular compartments characterized by great endogenous metabolic adaptability and highly active interactions with other subcellular compartments (van der Zand et al., 2006; Mast et al., 2015; Wanders et al., 2016). This adaptability is even more remarkable for plant peroxisomes, because their enzymatic composition reflects the functionality of the respective tissue and can change depending on the cell type in each plant organ (root, cotyledon or leaf), the stage of development or external conditions (Goto-Yamada et al., 2015; Reumann and Bartel, 2016). An excellent example of plant peroxisome adaptability and flexibility is the dynamic and reversible interactions between leaf peroxisomes and chloroplasts, which varies depending on illumination conditions. Here, the external appearance of peroxisomes changes from a spherical (dark conditions) to an elongated elliptical (light conditions) shape in order to increase the contact surface with chloroplast, thus facilitating effective metabolite interchange (Oikawa et al., 2015; Corpas, 2015).  The use of new biochemical, proteomic and cellular approaches has increased our knowledge of the components and functions of plant peroxisomes (Palma et al., 2009; Hu et al., 2012; Kataya et al., 2015; Reumann and Bartel, 2016; Cui et al., 2016; Corpas et al., 2017). Furthermore, H₂O₂ and nitric oxide (NO) are not only found in chloroplasts and mitochondria, but are also present in peroxisomes. This raises the question as to whether they function as signals between the different cellular compartments. This review provides an update on the metabolic interconnections between some of the newly identified components in peroxisomes, including NO, Ca²⁺, NADPH and nitro fatty acid (NO₂-FA). Thus, the involvement of enzymes, such as NADP-dehydrogenases (NADP-DHs), NAD kinases (NADKs) and Nudix hydrolase 19 (NUDX19, also known as NUDT19), contributes to the regulation of peroxisomal redox status mediating a change in the NADPH:NADP and GSH:GSSG ratio. These molecules are also connected to the enzymatic generation of NO, which involves NADPH, calmodulin (CaM) and Ca²⁺. In addition, the presence of lipid-derived signals, such as jasmonic acid (JA) and NO₂-FA, opens up new avenues of research into the importance of peroxisomes and their interaction with other cellular compartments.

NO is a versatile molecule involved in numerous physiological processes and in the mechanism of defense against both biotic and abiotic stresses (Domingos et al., 2015; Simontacchi et al., 2015). Given the prominence of peroxisomal oxidative metabolism over the last 15 years, research has especially focused on the potential presence of NO and related molecules called reactive nitrogen species (RNS) in peroxisomes. We have analyzed the generation of this molecule and its modulation of proteins whose function in peroxisomes is modified upon post-translational modification, such as nitration and S-nitrosation (Corpas and Barroso, 2014a; Chaki et al., 2015; Begara-Morales et al., 2016; Corpas et al., 2017), as well as its potential role as a signaling molecule promoting intracellular communication. NO has been detected in peroxisomes in different plant species using techniques such as spin trapping electron paramagnetic resonance (EPR) spectroscopy and fluorescence probes (Corpas et al., 2004, 2009). In addition, there is biochemical evidence to show that peroxisomal NO is generated by an L-arginine-dependent NO synthase (NOS)-like activity in plant cells (Barroso et al., 1999; Corpas and Barroso, 2016). Interestingly, NOS proteins have also been detected in animal peroxisomes (Stolz et al., 2002), suggesting that this enzyme is a constitutive component of peroxisomes, although the identity of the plant NOS protein is still unknown. However, it is known that the NOS protein contains a type 2 peroxisomal-targeting signal (PTS2), which is imported into the peroxisomal matrix; this domain is present in enzymes responsible for NO generation in plant and animal systems (Loughran et al., 2013; Corpas and Barroso, 2014b). Plant peroxisomal NOS-like activity has been found to require additional molecules – especially NADPH, Ca²⁺ and CaM, similar to what is found during the regulation of NOS activity in animal cells (Barroso et al., 1999; Corpas et al., 2004; Alderton et al., 2001). Peroxins (PEXs), which are peroxisomal biogenesis factors located in the cytosol and the peroxisomal membrane itself, are involved in matrix protein transport (Cross et al., 2016). In addition, the import of peroxisomal NOS proteins requires PEX12, as well as both of the PTS2 receptors PEX5 and PEX7. Through the use of Ca²⁺ channel blockers and CaM antagonists, we have also shown that this import is strictly dependent on both Ca²⁺ and CaM (Corpas and Barroso, 2014b). Fig. 1 shows a working model of the known elements involved in the peroxisomal protein import mechanism responsible for NO generation.

It is well established that Ca²⁺ and CaM are key regulatory elements in many cellular processes in animals and plants (Hepler, 2005; Clapham, 2007; Perochon et al., 2011). The presence of these factors in plant peroxisomes, as well as their potential metabolic functions, has recently begun to be studied. Ca²⁺ can move through membranes using voltage-dependent anion channels, among others (Gincel et al., 2001; Jammes et al., 2011). These channels are a class of porin ion channel, and their presence has been described in peroxisomal membranes (Fig. 1) (Reumann et al., 1997, 1998; Corpas et al., 2000). However, the presence of Ca²⁺ and its effect on catalase have only recently been assessed in plant peroxisomes (Costa et al., 2010; Yang and Poovaiah, 2012). The presence of Ca²⁺ also closely correlates with that of Ca²⁺-dependent protein kinase (CDPK) CPK1 in Arabidopsis peroxisomes. CPK1 appears to be not only involved in lipid metabolism and oxidative stress (Dammann et al., 2003), but also in mediating pathogen resistance (Coca and San Segundo, 2010). Similarly, CDPK2, which is involved in pollen tube growth, has been found in Petunia inflata peroxisomes (Yoon et al., 2006; Guo et al., 2013), which is indicative of a direct involvement of Ca²⁺ and peroxisomal NO in this process (Prado et al., 2004, 2008).

Additionally, Arabidopsis peroxisomes have been reported to contain the calmodulin-like (CML) protein CML3 (Chigri et al., 2012). CML3 can modulate Ca²⁺ signals, as it contains amino acid sequences commonly found in canonical CaM-binding motifs (Perochon et al., 2011). It mediates the dimerization of the Arabidopsis thaliana peroxisomal ATP-independent serine endopeptidase (DEG15), which belongs to the family of degradation of periplasmic protein (DEG) proteases; these are also orthologs of glyoxysomal-processing protease (GPP) in the watermelon (Helm et al., 2007). In its monomeric form, GPP exerts general protease activity; in its dimeric form, it is capable of specifically cleaving PTS2 domains (Fig. 1) (Schuhmann et al., 2008; Dolze et al., 2013). The PTS2 domain is present in plant peroxisomal enzymes, such as malate dehydrogenase, citrate synthase, acyl-CoA oxidase and thiolase; after cleavage by this protease, the proteins become functional (Mullen, 2002; Johnson and Olsen, 2003; Pracharoenwattana et al., 2007; Meng et al., 2014).

Recently, we reported that different CaM antagonists cause a reduction in peroxisomal Ca²⁺ levels, which adversely affects peroxisomal protein import (Corpas and Barroso, 2017b). Several peroxisomal enzymes, including catalase, hydroxypyruvate reductase (HPR) and glycolate oxidase (GOX), involved in photorespiration, were also found to be downregulated upon such treatment (Corpas and Barroso, 2017b). GOX is an important source of H₂O₂, suggesting that there is a close correlation between Ca²⁺ levels and H₂O₂. Therefore, it is plausible to assume that peroxisomal Ca²⁺ is crucial for normal function of several organelles. Table 1 summarizes the effects of Ca²⁺ on the peroxisomal protein function.

Thus, the newly identified peroxisomal molecules NO, Ca²⁺ and CaM show metabolic interactions, which evokes new questions regarding the physiological role of peroxisomes in both optimal and adverse environmental conditions.

The reducing power of the NADPH cofactor is well known to be essential for normal cell growth and proliferation. As mentioned previously, NADPH is also necessary as an electron donor for the generation of NO in peroxisomes. In plants, the main source of NADPH in photosynthetic cells is the ferrodoxin-NADP reductase (FNR) enzyme. However, in non-photosynthetic cells, such as roots, or in photosynthetic cells during the dark phase, a family of NADP-DHs regenerates NADPH. These NADP-DHs include two enzymes of the oxidative step of the pentose phosphate pathway, the glucose-6-phosphate dehydrogenase (G6PDH) and 6-phosphogluconate dehydrogenase (6PGDH) proteins, as well as the NADP-malic enzyme (NADP-ME), also known as NADP-malate dehydrogenase, and NADP-isocitrate dehydrogenase (NADP-ICDH) (Corpas and Barroso, 2014c). All these NADP-DHs have different isozymes that are present in the respective subcellular compartments, including the cytosol, the chloroplast, mitochondria and peroxisomes.

A set of NADPH-recycling enzymes have been detected in plant peroxisomes, including the above mentioned NADP-ICDH (Corpas et al., 1999; Reumann et al., 2007; Leterrier et al., 2016), as well as G6PDH and 6PGDH proteins (Corpas et al., 1998; Fernández-Fernández and Corpas, 2016; Hölscher et al., 2016) (see also Table 2). These NADP-DHs operate in coordination with certain import mechanisms in order to supply the substrate corresponding to each enzyme (Fig. 2). With regard to isocitrate, which is necessary for NADP-ICDH activity, the most plausible import mechanism is through the isocitrate–2-oxoglutarate shuttle (Reumann, 2000; Rottensteiner and Theodoulou, 2006; Linka and Theodoulou, 2013). However, the import system for glucose 6-phosphate (G6P), which is used by the G6PDH enzyme, has – to our knowledge – not been described in the peroxisomal membrane. One suspects the presence of a G6P translocator, similar to what is known to occur in plastids (Kunz et al., 2010). Nevertheless, an alternative import route for G6P could be through a porin in the peroxisomal membrane (Fig. 2) (Reumann et al., 1998).

Another way to maintain NADPH regeneration is to provide an appropriate NADP⁺ pool, which can be affected through the NAD⁺ metabolism. Recently, the membrane-located peroxisomal NAD⁺ carrier (PXN), which enables NAD⁺ to be imported in exchange for adenosine monophosphate (AMP), has been described in Arabidopsis (Bernhardt et al., 2012; van Roermund et al., 2016). In addition, NAD kinase (NADK) activity is known to catalyze the direct phosphorylation of NAD⁺ to NADP⁺ (Pollak et al., 2007) (Fig. 2). Of the three NADKs identified in Arabidopsis (Turner et al., 2005), NADK3 has been reported to be localized in the peroxisomal matrix (Waller et al., 2010), where it contributes to the generation of the peroxisomal NADP⁺ pool (Chai et al., 2006; Li et al., 2014) and maintains peroxisomal NADP-DH activity (Fig. 2).

In addition to the capacity to regenerate NADPH, the peroxisomal NADP-DHs are involved in other potential biological processes, such as leaf senescence, stomata movements and pollen-tube–ovule interactions. For example, leaf senescence is a highly regulated process characterized by antioxidative capacity loss, in which peroxisomes are involved (Pastori and del Río, 1997; Jiménez et al., 1998; Zimmermann et al., 2006). With regard to peroxisomal NADP-ICDH activity, a comparative analysis of its kinetic parameters in young and senescent leaves of pea plants has revealed that the NADP-ICDH affinity (K_m) for their isocitrate substrate decreased drastically during leaf senescence. This leads to over a 20-fold increase in isocitrate affinity, while the maximum rate (V_max) remains unchanged (Corpas et al., 1999). There are two possible explanations for this finding: first, that this allows the peroxisomal intracellular pool of isocitrate to outcompete the glyoxylate cycle enzyme isocitrate lyase, which appears in peroxisomes during leaf senescence (Pastori and del Río, 1997); or second, that the increase provides a larger supply of NADPH to eliminate excess H₂O₂ produced during senescence, when catalase activity decreases sharply (Corpas et al., 1999; Zimmermann et al., 2006). Stomatal movement is regulated by a complex network of signals, including Ca²⁺, NO, H₂O₂ and phytohormones (abscisic acid, JA and salicylic acid) (Murata et al., 2015). Recently Arabidopsis null mutants for the peroxisomal NADP-ICDH were shown to have stomatal movement malfunction (Leterrier et al., 2016). In this regard, it is worth pointing out that senescence and stomatal movement are also regulated by NO levels in peroxisomes (Corpas et al., 2004; Leterrier et al., 2016). Another example is peroxisomal 6PGDH in Arabidopsis, whose activity is necessary during growth and developmental processes, such as male gametophyte growth and pollen-tube–ovule interaction (Hölscher et al., 2016). Interestingly, Ca²⁺ and peroxisomal NO are also directly involved in the latter processes (Prado et al., 2004, 2008).

Apart from the involvement of NADPH in generating NO, as mentioned above, this reducing agent is also required for several other peroxisomal enzymatic activities that are involved in various pathways (see Table 2). For example, 2,4-dienoyl-CoA reductase (DECR) activity catalyzes the conversion of 2-trans,4-trans-dienoyl-CoA+NADPH into 3-trans-enoyl-CoA+NADP⁺ (Behrends et al., 1988; Goepfert and Poirier, 2007). This peroxisomal DECR is part of the alternative β-oxidation pathway present in plant peroxisomes, which is necessary to degrade unsaturated fatty enoyl-CoA esters with double bonds in even- and odd-numbered carbons (Fig. 3) (Rottensteiner and Theodoulou, 2006). Furthermore, 12-oxophytodienoate reductase (OPR) enzymes catalyze the conversion of 12-oxo-phytodienoic acid (OPDA)+NADPH into 3-oxo-2[2′(Z)-pentenyl]-cyclopentane-1-octanoate (OPC-8:0) and NADP⁺. In Arabidopsis, OPR3 is located in peroxisomes and is involved in JA biosynthesis, which has a role in signaling during development and in the response to adverse stresses, such as mechanical wounding and pathogens (Schaller et al., 2000; Strassner et al., 2002; Baker et al., 2006; Tani et al., 2008; Gfeller et al., 2010, see also below).

Another peroxisomal enzyme that requires NADPH is glutathione reductase (GR; At3g24170). This enzyme is part of the ascorbate–glutathione cycle, which facilitates the regeneration of the antioxidant glutathione (GSH) in the following reaction: GSSG+NADPH→GSH+NADP⁺. GR, whose activity has been observed to increase during natural leaf senescence and in response to abiotic stresses, has been detected in different plant species (Jiménez et al., 1998; Romero-Puertas et al., 2006; Kataya and Reumann, 2010). Similarly, its product, peroxisomal GSH, is increased under stress conditions and can reach a concentration of ∼5 mM in Arabidopsis leaf peroxisomes (Koffler et al., 2013; Zechmann, 2014). GSH can also react with NO to form S-nitrosoglutathione (GSNO), which may function as an intracellular NO reservoir; it mediates post-translational modifications, such as S-nitrosation. This affects the function of the protein, as previously described for catalase, ascorbate peroxidase and monodehydroascorbate peroxidase (Begara-Morales et al., 2016). This highlights connections between the peroxisomal metabolism of NADPH, antioxidant molecules and some the NO-related RNS molecules (Barroso et al., 2013; Corpas et al., 2017).

The last group of enzymes involved in the use of NADPH are pyrophosphatases that contain a highly conserved amino acid sequence or Nudix box, Gx₅Ex₇REUxEExGU, in which U is a hydrophobic amino acid such as Val, Ile or Leu (Ogawa et al., 2008). NUDX19 hydrolyzes NADPH to reduced nicotinamide mononucleotide (NMNH) plus 2′,5′-ADP. Amino acid sequence analysis suggests that Arabidopsis NUDX19 is located in both peroxisomes and chloroplasts (Ogawa et al., 2008; Lingner et al., 2011). Specifically, it has a C-terminal PTS1 tripeptide (SSL), which is regarded as a weak, but effective, type 1 PTS (PTS1). Indeed, the peroxisomal presence of Arabidopsis NUDX19 has been experimentally confirmed in Arabidopsis (Lingner et al., 2011). Owing to its NADPH pyrophosphohydrolase activity, NUDX19 can modulate the cellular levels and redox states of this pyridine nucleotide, which could be involved in NADP-DH regulation and hormonal signaling. Recent studies demonstrate that the loss of NUDX19 function affects NADPH content and mediates the increased resistance to arsenic and photooxidative stresses in Arabidopsis (Corpas et al., 2016; Maruta et al., 2016). The presence of NUDX19 in peroxisomes and chloroplasts points to metabolic and physical interactions during photorespiration between these organelles (Oikawa et al., 2015). Interestingly, in the Arabidopsis NUDX19 knockout mutant, a significant increase in all NADP-dehydrogenase activities (NADP-ICDH, G6PDH and 6PGDH) is triggered in roots and leaves, especially under induced arsenic stress (Corpas et al., 2016). This highlights the importance of NUDX19 in NADPH homeostasis in plants.

The cofactor NADPH could therefore be a necessary link in the redox metabolism of peroxisomes and must be regenerated to function normally; NADPH also represents a connection between several pathways, such as NO generation, the antioxidant system and β-oxidation.

Plant peroxisomes house all the components necessary for the β-oxidation of fatty acids (FAs); this oxidation is essential for the mobilization of the lipid reservoir required for seedling development (Graham, 2008). Additionally, peroxisomes are also important for the metabolism of several lipid-derived signals that have a crucial physiological role in the entire plant. The import of FAs into peroxisomes is mediated by ATP-binding cassette (ABC) transporter 1 [ABCC1, also known as COMATOSE (CTS)], which is involved in the import of other molecules such as 12-oxo-phytodienoic acid (OPDA), a precursor of jasmonic acid (JA), as well as 4-dichlorophenoxybutyric acid and indole butyric acid (IBA); the latter of which are precursors of the phytohormone auxin, which is involved in plant growth and development (Baker et al., 2015).

As mentioned above, JA is a linolenic acid-derived cyclopentanone mainly involved in mediating wounding responses (Theodoulou et al., 2005; Delker et al., 2007; Gfeller et al., 2010; León, 2013). JA and its derivatives also modulate a wide range of activities, such as plant development, as well as abiotic (including salinity, drought and UV radiation) and biotic stress tolerance; these activities also provide a defense against pathogen infection and insect attacks (Wasternack and Song, 2017). In this context, the interplay between NO, phytohormones (indoleacetic acid, abscisic acid and ethylene) and other plant growth regulators (polyamines) has been well established (Freschi, 2013; Asgher et al., 2017). This interplay could be of physiological significance in the peroxisomal metabolism, during which these molecules are generated. Indeed, in Arabidopsis, there is evidence that NO increases JA production by increasing the expression of the OPR3 gene in peroxisomes (Fig. 3) (Huang et al., 2004; Mur et al., 2013; Hussain et al., 2016). A similar effect of NO on JA has been observed in tomato roots in response to nematode infection (Zhou et al., 2015). OPR3 activity also affects the metabolism of other molecules, including those of glucosinolates and tryptophan (Cao et al., 2016). Tryptophan is a precursor of serotonin and indoleacetic acid (IAA), which are involved in plant growth and stress responses (Zolman et al., 2008; Schlicht et al., 2013; Spiess and Zolman, 2013). Therefore, apart from the important role played by peroxisomal β-oxidation in lipid mobilization during germination, there are other derived lipids, such as JA, whose synthesis specifically requires NADPH for OPR3 activity. However, expression of the OPR3 gene is also stimulated by NO, which establishes another connection point between NADPH, JA and NO in peroxisomes (Fig. 3).

NO₂-FAs are derived from NO and produced by the addition of a nitro group (–NO₂) to unsaturated fatty acids. These molecules were first shown to have important physiological functions in animals, such as cardiac ischemic injury protection, regulation of inflammatory processes and reduction of atherosclerosis (Freeman et al., 2008; Trostchansky et al., 2013). In higher plants, the presence of NO₂-FAs was hypothesized a few years ago (Sánchez-Calvo et al., 2013). Recently, we have reported on the endogenous presence and function of NO₂-FAs in Arabidopsis, with the most prominent NO₂-FA being nitro-linolenic acid (NO₂-Ln) (Mata-Pérez et al., 2016a). The highest NO₂-Ln content, which diminished during plant development, was found in Arabidopsis seeds. NO₂-Ln activates a molecular chaperone network in response to diverse abiotic stresses and has been shown to have a powerful signaling function. NO₂-Ln has the capacity to either increase or decrease the gene expression of hundreds of genes with a remarkable induction capacity, and ∼25% of the total upregulated genes encode heat-shock proteins (Mata-Pérez et al., 2016a). Recently, mass spectrometry analysis of the peroxisomal fraction isolated from the leaves of pea plants has revealed that peroxisomes also contain NO₂-Ln (Mata-Perez et al., 2017). This demonstrates that there is a highly active NO metabolism in plant peroxisomes, as these molecules may function as a NO reservoir with the capacity to release NO (Mata-Pérez et al., 2016b). Alternatively, the electrophilic property of NO₂-Ln could allow it to contribute to nitroalkylation of a protein that contains nucleophilic residues (Cys, His or Lys), which could contribute to the signaling action of NO₂-FAs. Taken together, the physiological function of these molecules in plant peroxisomes raises issues that need to be addressed and could provide new evidence to show that these organelles have a potential signaling function mediated by GSNO and NO₂-FAs.

The metabolism of plant peroxisomes is characterized by adaptability and flexibility, and depends on the organ, stage of development and response to adverse environmental conditions. Consequently, peroxisomes should be regarded as nodes at the crossroads of different metabolic pathways. Examples of these pathways are photorespiration, lipid mobilization and the nitro-oxidative metabolism, which endow cells with signaling molecules, including H₂O₂, NO, GSNO, NO₂-FA and JA. Fig. 3 summarizes a model for the interplay between the various molecules described in this review of plant peroxisomes. Their NADPH content is affected by NADP-DHs, NADKs and NUDX19, which help to regulate the peroxisomal redox status. Moreover, NADPH, synthesis of NO, ascorbate–glutathion cycle and β-oxidation are important for several essential metabolic pathways; NADPH also mediates the generation of lipid-derived signals (JA and NO₂-FA).

However, several points regarding these new peroxisomal molecules need to be addressed in future research: the mechanisms that control their transport through the peroxisomal membrane of Ca²⁺ and the substrates specific to NADP-DHs; the identification of the protein responsible for NO synthesis; the peroxisomal proteins susceptible to post-translational modifications mediated by NO (nitration, S-nitrosation or nitroalkylation); the specific function of peroxisomal NO₂-FAs under physiological and stress conditions; and the cellular function of peroxisomal molecules as compared to molecules from other subcellular compartments. The answers to all these points could provide a better understanding of the cellular role of peroxisomes as a source of intra- and inter-cellular signaling molecules (H₂O₂, NO, NO₂-FA and JA). Such an understanding may also be useful for researchers working on peroxisomes in other organisms, including yeast and mammals, in order to provide a more complete picture of the signals involved in biochemical communication between organelles.

We would also like to thank Michael O'Shea for proofreading the text.