Lipid Peroxide-Derived Short-Chain Carbonyls Mediate Hydrogen Peroxide-Induced and Salt-Induced Programmed Cell Death in Plants  

Abstract

Lipid peroxide-derived toxic carbonyl compounds (oxylipin carbonyls), produced downstream of reactive oxygen species (ROS), were recently revealed to mediate abiotic stress-induced damage of plants. Here, we investigated how oxylipin carbonyls cause cell death. When tobacco (Nicotiana tabacum) Bright Yellow-2 (BY-2) cells were exposed to hydrogen peroxide, several species of short-chain oxylipin carbonyls [i.e. 4-hydroxy-(E)-2-nonenal and acrolein] accumulated and the cells underwent programmed cell death (PCD), as judged based on DNA fragmentation, an increase in terminal deoxynucleotidyl transferase dUTP nick end labeling-positive nuclei, and cytoplasm retraction. These oxylipin carbonyls caused PCD in BY-2 cells and roots of tobacco and Arabidopsis (Arabidopsis thaliana). To test the possibility that oxylipin carbonyls mediate an oxidative signal to cause PCD, we performed pharmacological and genetic experiments. Carnosine and hydralazine, having distinct chemistry for scavenging carbonyls, significantly suppressed the increase in oxylipin carbonyls and blocked PCD in BY-2 cells and Arabidopsis roots, but they did not affect the levels of ROS and lipid peroxides. A transgenic tobacco line that overproduces 2-alkenal reductase, an Arabidopsis enzyme to detoxify α,β-unsaturated carbonyls, suffered less PCD in root epidermis after hydrogen peroxide or salt treatment than did the wild type, whereas the ROS level increases due to the stress treatments were not different between the lines. From these results, we conclude that oxylipin carbonyls are involved in the PCD process in oxidatively stressed cells. Our comparison of the ability of distinct carbonyls to induce PCD in BY-2 cells revealed that acrolein and 4-hydroxy-(E)-2-nonenal are the most potent carbonyls. The physiological relevance and possible mechanisms of the carbonyl-induced PCD are discussed.

In plants, environmental stressors such as extreme temperatures, drought, intense UV-B radiation, and soil salinity can cause tissue damage, growth inhibition, and even death. These detrimental effects are often ascribed to the action of reactive oxygen species (ROS) produced in the stressed plants for the following reasons: (1) various environmental stressors commonly cause the oxidation of biomolecules in plants; and (2) transgenic plants with enhanced antioxidant capacities show improved tolerance to environmental stressors (Suzuki et al., 2014). The production of ROS such as superoxide anion radical and hydrogen peroxide (H₂O₂) is intrinsically associated with photosynthesis and respiration (Foyer and Noctor, 2003; Asada, 2006).

Plant cells are equipped with abundant antioxidant molecules such as α-tocopherol, β-carotene, and ascorbic acid and an array of ROS-scavenging enzymes such as superoxide dismutase and ascorbate peroxidase to maintain low intracellular ROS levels. When plants are exposed to severe and prolonged environmental stress, the balance between the production and scavenging of ROS is disrupted and the cellular metabolism reaches a new state of higher ROS production and lower antioxidant capacity. Then, the oxidation of vital biomolecules such as proteins and DNA proceeds, and as a consequence, cells undergo oxidative injury (Mano, 2002). The cause-effect relationship between ROS and tissue injury in plants is thus widely accepted, but the biochemical processes between the generation of ROS and cell death are poorly understood.

Increasing evidence shows that oxylipin carbonyls mediate the oxidative injury of plants (Yamauchi et al., 2012; for review, see Mano, 2012; Farmer and Mueller, 2013). Oxylipin carbonyls are a group of carbonyl compounds derived from oxygenated lipids and fatty acids. The production of oxylipin carbonyls in living cells is explained as follows. Lipids in the membranes are constitutively oxidized by ROS to form lipid peroxides (LOOHs; Mène-Saffrané et al., 2007) because they are the most immediate and abundant targets near the ROS production sites. There are two types of LOOH formation reaction from ROS (Halliwell and Gutteridge, 2007). One is the radical-dependent reaction. Highly oxidizing radicals, such as hydroxyl radical (standard reduction potential of the HO^•/H₂O pair, +2.31 V) and the protonated form of superoxide radical (HO₂/H₂O₂, +1.06 V), can abstract a hydrogen atom from a lipid molecule, especially at the central carbon of a pentadiene structure in a polyunsaturated fatty acid, to form a radical. This organic radical rapidly reacts with molecular oxygen, forming a lipid hydroperoxyl radical, which then abstracts a hydrogen atom from a neighboring molecule and becomes a LOOH. The other reaction is the addition of singlet oxygen to a double bond of an unsaturated fatty acid to form an endoperoxide or a hydroperoxide (both are LOOHs). A variety of LOOH species are formed, depending on the source fatty acid and also by the oxygenation mechanism (Montillet et al., 2004). LOOH molecules are unstable, and in the presence of redox catalysts such as transition metal ions or free radicals, they decompose to form various aldehydes and ketones (i.e. oxylipin carbonyls; Farmer and Mueller, 2013). The chemical species of oxylipin carbonyl formed in the cells differ according to the fatty acids and the type of ROS involved (Grosch, 1987; Mano et al., 2014a).

More than a dozen species of oxylipin carbonyls are formed in plants (for review, see Mano et al., 2009). Oxylipin carbonyls are constitutively formed in plants under normal physiological conditions, and the levels of certain types of oxylipin carbonyls rise severalfold under stress conditions, detected as increases in the free carbonyl content (Mano et al., 2010; Yin et al., 2010; Kai et al., 2012) and by the extent of the carbonyl modification of target proteins (Winger et al., 2007; Mano et al., 2014b). Among the oxylipin carbonyls, the α,β-unsaturated carbonyls, such as acrolein and 4-hydroxy-(E)-2-nonenal (HNE), have high reactivity and cytotoxicity (Esterbauer et al., 1991; Alméras et al., 2003). They strongly inactivate lipoate enzymes in mitochondria (Taylor et al., 2002) and thiol-regulated enzymes in chloroplasts (Mano et al., 2009) in vitro and cause tissue injury in leaves when they are fumigated (Matsui et al., 2012).

The physiological relevance of oxylipin carbonyls has been shown by the observation that the overexpression of different carbonyl-scavenging enzymes commonly confers stress tolerance to transgenic plants (for review, see Mano, 2012). For example, 2-alkenal reductase (AER)-overproducing tobacco (Nicotiana tabacum) showed tolerance to aluminum (Yin et al., 2010), aldehyde dehydrogenase-overproducing Arabidopsis (Arabidopsis thaliana) showed tolerance to osmotic and oxidative stress (Sunkar et al., 2003), and aldehyde reductase-overproducing tobacco showed tolerance to chemical and drought stress (Oberschall et al., 2000). In addition, the genetic suppression of a carbonyl-scavenging enzyme made plants susceptible to stressors (Kotchoni et al., 2006; Shin et al., 2009; Yamauchi et al., 2012; Tang et al., 2014). Under stress conditions, there are positive correlations between the levels of certain carbonyls and the extent of tissue injury (Mano et al., 2010; Yin et al., 2010; Yamauchi et al., 2012). Thus, it is evident that oxylipin carbonyls, downstream products of ROS, are causes of oxidative damage in plant cells.

To investigate how oxylipin carbonyls damage cells in oxidatively stressed plants, we here examined the mode of cell death that is induced by oxylipin carbonyls and identified the carbonyl species responsible for the cell death. We observed that oxylipin carbonyls cause programmed cell death (PCD), and our results demonstrated that the oxylipin carbonyls mediate the oxidative stress-induced PCD in tobacco Bright Yellow-2 (BY-2) cultured cells and in roots of tobacco and Arabidopsis plants. We then estimated the relative strengths of distinct carbonyl species to initiate the PCD program. Our findings demonstrate a critical role of the lipid metabolites in ROS signaling.

RESULTS

Oxylipin Carbonyls Formed in H₂O₂-Stressed Cells Can Cause PCD

To investigate the process of oxidative injury in plant cells, we first used tobacco BY-2 cells and gave them an oxidative stimulus with H₂O₂. Cells that underwent a 4-d culture (0.7 g fresh weight per flask) propagated to double weight in approximately 20 h under the normal culture conditions (Fig. 1A, untreated). When H₂O₂ was added to 1 mm, the cells stopped propagation and their fresh weight started to decrease. At 20 h, the fresh weight was reduced to 0.4 g (Fig. 1A) and the cells were apparently dead, as detected by Trypan blue staining (Fig. 1, B and C).

Figure 1.

H₂O₂-induced growth inhibition and cell death of tobacco BY-2 cells and the suppression of them by carnosine and hydralazine. Fifty milligrams of cells from 7-d culture was subcultured in 50 mL of fresh culture medium, and after 4 d, the culture medium was supplemented with either 1 mm  H₂O₂ or a carbonyl scavenger (1 mm carnosine or 0.2 mm hydralazine) or both. A, Changes in the fresh weight (per flask) of cells. Cells were collected at the indicated time points and weighed immediately. Carnosine (A-i) and hydralazine (A-ii) were included as indicated. Means ± se of three independent experiments are shown. B, Detection of cell death with Trypan blue staining. BY-2 cells treated as in A were collected at 20 h and stained as described in “Materials and Methods.” Cells forming a single layer under microscopy were chosen for the evaluation. Typical images of the Trypan blue staining are shown: untreated cells as a blank control (i), 1 mm  H₂O₂ (ii), 1 mm  H₂O₂ + 1 mm carnosine (iii), 1 mm  H₂O₂ + 0.2 mm hydralazine (iv), 1 mm carnosine (v), and 0.2 mm hydralazine (vi). White arrows indicate dead cells. Bar = 50 µm. C, The fraction of dead cells (Trypan blue-stained cells) at 20 h of incubation. A total of 200 cells were counted in each treatment. Means ± se of three independent experiments are shown. Differences among treatments were analyzed by Tukey’s test: P < 0.05.

H₂O₂ induces PCD in BY-2 cells, appearing as morphological changes and nuclear DNA fragmentation (Houot et al., 2001). We also detected PCD-associated events in the BY-2 cells in the 20-h H₂O₂ treatment group, as follows. (1) The fragmentation of genomic DNA into 0.18 kb and its multiples was detected as DNA laddering (Fig. 2A). (2) More than 80% of the cells had positive nuclei in the terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) assay, which represents the fragmentation of DNA, whereas less than 2% of the untreated control cells were positive (Fig. 2B). (3) Retraction of cytosol from the cell wall (Reape and McCabe, 2008) was observed (Fig. 2D). Thus, under our experimental conditions, too, H₂O₂ induced PCD in BY-2 cells.

Figure 2.

PCD-associated events in BY-2 cells induced by H₂O₂ were suppressed by carbonyl scavengers. Four-day-cultured cells were treated with 1 mm  H₂O₂ with or without a carbonyl scavenger (1 mm carnosine or 0.2 mm hydralazine). After a 20-h incubation, the cells were used for genomic DNA extraction, a TUNEL assay, and cytoplasm retraction observation as described in “Materials and Methods.” A, Agarose gel electrophoresis of genomic DNA. Cells were treated as indicated at the top of each lane. The left-most lane is for molecular weight markers. White arrows indicate DNA fragments of 0.18, 0.36, and 0.54 kb. B, Fraction of the cells with TUNEL-positive nuclei. Cells forming a single layer under microscopy were chosen for evaluation. The total cell number was counted under phase contrast observation, and the TUNEL-positive cells were counted under fluorescence observation. All values are means ± se, and the data represent three independent experiments. Differences among treatments were analyzed by Tukey’s test: P < 0.05. C, Typical fluorescence microscopy images of the TUNEL assay results: untreated cells as a blank control (i), 1 mm  H₂O₂ (ii), 1 mm  H₂O₂ + 1 mm carnosine (iii), 1 mm  H₂O₂ + 0.2 mm hydralazine (iv), 1 mm carnosine (v), 0.2 mm hydralazine (vi), positive control (vii), and negative control (viii). Bar = 50 µm. D, Typical phase-contrast microscopy images of cell morphology for cytoplasm retraction: untreated control cells (i), 1 mm  H₂O₂ (ii), 1 mm  H₂O₂ + 1 mm carnosine (iii), and 1 mm  H₂O₂ + 0.2 mm hydralazine (iv). The white arrow indicates cytosolic retraction. Bar = 50 µm.

According to our previous observations of tobacco leaves and roots (Mano et al., 2010; Yin et al., 2010), oxidative stress treatment will increase the levels of oxylipin carbonyls before apparent cell death is observed. At 2 h after treatment, when the H₂O₂-treated cells had just stopped growth (Fig. 1A), we extracted carbonyls from the cells, derivatized them with 2,4-dinitrophenylhydrazine, and analyzed them by HPLC. Eight species of carbonyls of C1 to C9 were detected in the untreated cells (Fig. 3A). In the H₂O₂-treated cells, the levels of HNE, n-hexanal, n-heptanal, malondialdehyde, acetaldehyde, and propionaldehyde were significantly higher (Fig. 3B). Acrolein and 4-hydroxy-(E)-2-hexenal (HHE) also tended to be higher in the H₂O₂-treated cells, although the difference was not significant.

Figure 3.

Effects of H₂O₂ and carbonyl scavengers on the carbonyl contents in BY-2 cells. Four-day-cultured cells were treated with water as a control, 1 mm  H₂O₂, and 1 mm  H₂O₂ plus 2 mm carnosine or 0.2 mm hydralazine for 2 h. Carbonyls were extracted from them, derivatized with 2,4-dinitrophenylhydrazine, and separated by HPLC as described in “Materials and Methods.” A, Typical chromatograms showing the carbonyls in the control (top) and H₂O₂-treated (bottom) BY-2 cells. The identified aldehydes are labeled at the top of each peak. B, Intracellular contents of malondialdehyde, acetaldehyde, HHE, acrolein, propionaldehyde, HNE, n-hexanal, and n-heptanal. Means ± se of three independent experiments are shown. Differences among treatments were analyzed by Tukey’s test: P < 0.05. FW, Fresh weight.

We also determined carbonyls at 5 h (Supplemental Fig. S1), when the H₂O₂-treated cells showed a notable loss of fresh weight. Highly electrophilic and reactive carbonyls such as HNE and HHE were significantly increased, and other less reactive saturated carbonyls such as n-hexanal and n-heptanal were increased by at least 10-fold, compared with the 2-h treated cells (Supplemental Fig. S2). It should be noted that the α,β-unsaturated carbonyls HNE and acrolein were increased by the H₂O₂ treatment at an early stage and continued to increase afterward. These compounds can induce cell death in animals (Kruman et al., 1997; Liu-Snyder et al., 2006) and cause tissue injury in plants (Alméras et al., 2003; Mano et al., 2005; Yin et al., 2010; Kai et al., 2012).

We tested the ability of acrolein, HNE, n-hexanal, n-heptanal, (Z)-3-hexenal, and propionaldehyde to cause PCD in BY-2 cells. The addition of acrolein (Fig. 4) and HNE (Supplemental Fig. S3) at 0.2 mm and n-hexanal (Supplemental Fig. S4) at 3 mm to the cells resulted in DNA laddering, an increase in the percentage of TUNEL-positive nuclei, and cytoplasm retraction. DNA laddering was also found with 3 mm  n-heptanal, 3 mm (Z)-3-hexenal, and 50 mm propionaldehyde (Supplemental Fig. S5). These carbonyl concentrations required for inducing PCD were clearly higher than their endogenous levels. This is probably because exogenously added carbonyls were first scavenged by the cell with glutathione and specific enzymes, and the residual unscavenged carbonyls exerted cytotoxicity or signaling actions.

Figure 4.

Induction of PCD in BY-2 cells by acrolein. Four-day-cultured cells were treated with 0.2 mm acrolein or a carbonyl scavenger (1 mm carnosine or 0.2 mm hydralazine) or both as indicated. After 20 h of incubation, the cells were used for genomic DNA extraction, the TUNEL assay, and cytoplasm retraction observation as described in “Materials and Methods.” A, Agarose gel electrophoresis of genomic DNA. White arrows indicate the DNA fragments of 0.18, 0.36, and 0.54 kb. B, Fraction of the cells with TUNEL-positive nuclei. The total cell number and the TUNEL-positive cells were counted as in Figure 2B. Means ± se of three independent experiments are shown. Differences among treatments were analyzed by Tukey’s test: P < 0.05. C, Typical fluorescence microscopy images of the TUNEL assay results: untreated cells as a blank control (i), 0.2 mm acrolein (ii), 0.2 mm acrolein + 1 mm carnosine (iii), and 0.2 mm acrolein + 0.2 mm hydralazine (iv). Bar = 50 µm. D, Typical phase-contrast microscopy images of cell morphology for cytoplasm retraction: untreated control cells (i), 0.2 mm acrolein (ii), 0.2 mm acrolein + 1 mm carnosine (iii), and 0.2 mm acrolein + 0.2 mm hydralazine (iv). The white arrow indicates cytosolic retraction. Bar = 50 µm.

Carbonyl Scavengers Suppressed the Intracellular Carbonyls and the H₂O₂-Induced PCD

We hypothesized that if oxylipin carbonyls formed after H₂O₂ treatment are responsible for the PCD, the scavenging of them would stop the cell death. To test this hypothesis, we used two carbonyl-scavenging agents, carnosine and hydralazine, and examined their effects on the death of the H₂O₂-treated cells. Carnosine is a dipeptide (β-alanyl-l-His) found in skeletal muscles and brain in vertebrates at millimolar levels. It can covalently trap carbonyls at the amino end of its β-alanyl moiety (Aldini et al., 2002; Burcham et al., 2002). Hydralazine (1-hydrazinylphthalazine) can trap carbonyls at its hydrazine moiety (Burcham et al., 2002).

Our findings confirmed that carnosine and hydralazine suppressed the carbonyl-induced PCD. When carnosine (1–5 mm) or hydralazine (0.2 mm) was added with acrolein, HNE, n-hexanal, n-heptanal, (Z)-3-hexenal, or propionaldehyde in the culture medium, all of the PCD-associated events caused by these aldehydes were suppressed: the DNA laddering (Fig. 4A; Supplemental Figs. S3A, S4A, and S5), the increase in TUNEL-positive nuclei (Fig. 4B; Supplemental Figs. S3B and S4B), and the cytosol retraction (Fig. 4C; Supplemental Figs. S3C and S4C).

In the H₂O₂-treated cells, the addition of carnosine and hydralazine greatly suppressed the carbonyl levels (Fig. 3B). Carnosine effectively prevented the H₂O₂-induced growth inhibition and the death of BY-2 cells (Fig. 1A-i). Hydralazine also suppressed cell death, although the compound itself showed weak cytotoxicity (Fig. 1A-ii). As expected, all of the PCD-associated events caused by H₂O₂ were strongly suppressed by carnosine and hydralazine: the DNA laddering (Fig. 2A), the increase in TUNEL-positive nuclei (Fig. 2B), and the cytosol retraction (Fig. 2D). It appears that hydralazine had a weaker effect on PCD markers, but this was probably due to the low dose of hydralazine used here. Because it showed a toxicity on the growth of the cells (Fig. 1A-ii), we did not add it at a higher concentration. At the concentration we used, hydralazine might have been depleted below optimal scavenger levels by the carbonyls that were continuously produced following the addition of H₂O₂.

These pharmacological results suggested that the H₂O₂-induced PCD was prevented by the scavenging of oxylipin carbonyls. There is, however, a possibility that these scavengers also suppressed ROS, because carnosine and hydralazine have been reported to have antioxidant capacities (Aruoma et al., 1989; Daiber et al., 2005). We investigated the possibility that these compounds suppressed ROS and LOOH in the following four ways. First, to examine the direct scavenging of H₂O₂ by these compounds, we incubated H₂O₂ at 5 mm with either carnosine at 5 mm or hydralazine at 1 mm in 50 mm phosphate buffer (pH 7.4) and 0.5 mm diethylenetriaminepentaacetic acid at 25°C. After a 30-min incubation, we found no significant decrease in the H₂O₂ amount as determined by the catalase-mediated oxygen evolution (data not shown). Thus, these carbonyl scavengers do not significantly scavenge H₂O₂ at the concentrations we used.

Second, we examined the effects of these compounds on the intracellular H₂O₂ level by using BES-H₂O₂-Ac, a highly specific H₂O₂ indicator (Maeda, 2008; Supplemental Fig. S6). At 2 h after the addition of H₂O₂ to BY-2 cells, the BES-H₂O₂ fluorescence level was lower. The decrease in H₂O₂ was probably due to the induction of H₂O₂-scavenging enzymes such as peroxidases (Tsukagoshi et al., 2010; Xu et al., 2011). In both H₂O₂-stimulated and untreated cells, carnosine and hydralazine, at the concentrations we used, did not affect the intracellular H₂O₂ level.

In the third experiment, we determined the intracellular level of a broader range of ROS by monitoring the oxidation of 2′,7′-dihydrodichlorofluoresein (H₂DCF; Fig. 5). H₂DCF, formed from the exogenously added H₂DCF diacetate via enzymatic hydrolysis in the cell, is oxidized by hydroxyl radical, organic peroxyl radicals, and the reactive nitrogen species NO^• and ONOO^– to form the fluorescent dye dichlorofluorescein (DCF). This dye is used for determinations of the formation of general reactive species rather than specific ROS (Halliwell and Gutteridge, 2007).

Figure 5.

Carnosine and hydralazine did not affect the increases in ROS levels in BY-2 cells after H₂O₂ treatment. A, Four-day-cultured cells were incubated with either 1 mm  H₂O₂ or a carbonyl scavenger (1 mm carnosine or 0.2 mm hydralazine) or both for 2 h. DCF fluorescence was recorded using a fluorescence microscope as in “Materials and Methods.” Typical photographs are shown: untreated cells as a control (i), 1 mm  H₂O₂ (ii), 1 mm  H₂O₂ + 1 mm carnosine (iii), 1 mm  H₂O₂ + 0.2 mm hydralazine (iv), 1 mm carnosine (v), and 0.2 mm hydralazine (vi). Bar = 50 µm. B, DCF fluorescence intensity of cells. The fluorescence intensity was integrated per cell with ImageJ software. A total of 200 cells were counted in each treatment. Means of three runs ± se are shown. Differences among treatments were analyzed by Tukey’s test: P < 0.05. a.u., Absorbance units.

Untreated control cells showed weak fluorescence, representing the basal level of constitutively formed ROS. When H₂O₂ was added to the cells, the fluorescence level was increased slightly at 30 min, and at 2 h, it became 4-fold stronger than that in the untreated control cells (Fig. 5B). The fluorescence had become weaker at 5 h (data not shown). This indicated that the intracellular ROS level was transiently increased by the H₂O₂ stimulus. It should be noted that the changes in DCF fluorescence and BES-H₂O₂ fluorescence after H₂O₂ treatment were apparently different. While the H₂O₂ level was suppressed in stressed cells, the general ROS level was increased. Thus, in the H₂O₂-treated BY-2 cells, DCF fluorescence represented not only the level of H₂O₂ but also the levels of a broader range of ROS. Carnosine and hydralazine did not affect the intracellular ROS levels before or after the H₂O₂ treatment (Fig. 5B), providing evidence that these reagents did not scavenge ROS efficiently in the BY-2 cells.

In the fourth experiment, we examined the effects of carnosine and hydralazine on the LOOH level. LOOHs are products of ROS and the immediate precursors of oxylipin carbonyls. If the LOOH level is not affected by carnosine or hydralazine, then the observed suppression of carbonyls by these compounds can be explained primarily as a direct scavenging of carbonyls. We determined the LOOH level with the fluorescent probe Spy-LHP (Soh et al., 2007). Spy-LHP, a derivative of diphenyl-1-pyrenylphosphine, has a bulky hydrophobic tail and reacts very rapidly with LOOH to form its oxidized product, Spy-LHPOx, which fluoresces intensely. The reactions of Spy-LHP with O₂  ^–•, alkyl hydroperoxyl radical, nitric oxide, and peroxinitrite are insignificant (Soh et al., 2007). Its reaction with H₂O₂ is very slow and can be distinguished from that with LOOH by the measurement of fluorescence increase kinetics (Khorobrykh et al., 2011).

We found that the addition of H₂O₂ to BY-2 cells increased the LOOH level 4-fold in 30 min, and it remained high up to 2 h. Neither carnosine nor hydralazine significantly affected the LOOH level (Fig. 6). Thus, under our experimental conditions, carnosine and hydralazine primarily acted as carbonyl scavengers rather than ROS scavengers. The suppression of H₂O₂-induced PCD by these scavengers suggests that the carbonyl species formed in stressed cells participated in the initiation of PCD.

Figure 6.

The carbonyl scavengers carnosine and hydralazine did not suppress LOOH levels in BY-2 cells. Four-day-cultured cells were incubated with either 1 mm  H₂O₂ or a carbonyl scavenger (1 mm carnosine or 0.2 mm hydralazine) or both for the indicated times. The LOOH levels were detected as in “Materials and Methods.” Means ± se of the data represent three independent experiments. Differences among treatments at the same time point were analyzed by Tukey’s test: P < 0.05. FW, Fresh weight.

Involvement of Oxylipin Carbonyls in Root PCD

We then examined whether oxylipin carbonyls are involved in PCD in planta by combining the pharmacological approach described above with a genetic approach. The addition of H₂O₂ or salt (NaCl) to roots causes PCD in the epidermal cells (Demidchik et al., 2010). Arabidopsis plants grown on an agar plate for 6 d were stressed with 0.2 mm  H₂O₂ (Supplemental Fig. S7, A-ii and B) or 150 mm  NaCl (Supplemental Fig. S8, A-ii and B) for 20 h. Both the H₂O₂ and NaCl treatments resulted in approximately 60 TUNEL-positive cells per half millimeter of root apex. The protoplast retraction in root hairs and the concomitant loss of the fluorescein diacetate (FDA)-fluorescing ability in the roots treated with H₂O₂ (Supplemental Fig. S7C-ii) or NaCl (Supplemental Fig. S8C-ii) also indicated the occurrence of PCD (Reape and McCabe, 2008; Hogg et al., 2011).

When carnosine (3 mm) or hydralazine (0.2 mm) was supplemented in the medium, all of the PCD-associated events caused by H₂O₂ and NaCl were largely suppressed. Specifically, the number of TUNEL-positive nuclei was decreased (Supplemental Figs. S7 and S8A, iii and iv) and protoplast retraction in root hairs was prevented (Supplemental Figs. S7 and S8C, iii and iv).

Carbonyl scavengers suppressed oxylipin carbonyls but did not affect the intracellular levels of H₂O₂ in the root tissues. In Arabidopsis plants exposed to 150 mm  NaCl for 20 h, an accumulation of carbonyls along the vascular cylinder was observed as a pink color development of Schiff’s reagent (Supplemental Fig. S9-ii). When carnosine or hydralazine was added to the plants concomitantly with NaCl, the color developed only weakly (Supplemental Fig. S9, iii and iv). These carbonyl scavengers also did not affect the BES-H₂O₂ fluorescence level in H₂O₂-treated root apex (Supplemental Fig. S10). Thus, oxylipin carbonyls appeared to be involved in in planta PCD as well.

To obtain genetic evidence for the involvement of oxylipin carbonyls in PCD, we used transgenic tobacco plants overexpressing Arabidopsis AER (Mano et al., 2005). AER catalyzes the NADPH-dependent reduction of the α,β-unsaturated bond in 2-alkenals such as HNE (Mano et al., 2002). The AER-overexpressing tobacco plants accumulate smaller amounts of carbonyls upon oxidative stress than the wild-type SR1 (Mano et al., 2010; Yin et al., 2010).

In this study, SR1 and the AER-overexpressing line P1#18 plants grown on agar plates were exposed to H₂O₂ or NaCl. Treatment with 2 mm  H₂O₂ and 175 mm  NaCl for 20 h respectively resulted in approximately 50 and 40 TUNEL-positive nuclei per 1 mm of root apex in SR1 plants, whereas in the P1#18 plants, both treatments produced only approximately 10 TUNEL-positive nuclei (Fig. 7B). Most of the root hair cells in the SR1 plants exposed to H₂O₂ and NaCl showed protoplast retraction and lost the ability to exhibit FDA fluorescence, whereas the root hair cells in the P1#18 plants showed much fewer PCD symptoms (Fig. 8).

Figure 7.

PCD in roots of the wild-type tobacco line SR1 and the AER-overexpressing line P1#18. Six-day-old plants were transferred to fresh Murashige and Skoog (MS) medium supplemented with either 2 mm  H₂O₂ or 175 mm  NaCl. After a 20-h incubation, root tips of about 5-mm length were excised from the plant and used for TUNEL assay as described in “Materials and Methods.” A, Typical fluorescence images for TUNEL assay. Bar = 250 µm. B, Fraction of the cells with TUNEL-positive nuclei. TUNEL-positive cells were counted in 1 mm from the tip of the primary root. A total of nine roots were counted in each treatment. Differences between the lines were examined by Student’s t test. n.s., Not significant.

Figure 8.

Root hair PCD in the wild-type tobacco line SR1 and the AER-overexpressing line P1#18. Six-day-old plants were treated with H₂O₂ or NaCl as in Figure 7. After a 20-h incubation, root tips of 5-mm length were excised and used for FDA staining as described in “Materials and Methods.” Typical phase-contrast images for cytoplasm retraction (top row for each line) and fluorescence images of the same field (bottom row) are shown. White arrows indicate cytoplasm retraction. Bars = 100 µm.

The overexpression of AER did not affect the intracellular ROS level (Fig. 9). The DCF fluorescence level was low under nonstressed conditions, and it was increased by treatment with H₂O₂ and NaCl, especially at the elongation zone in the root apex. The SR1 and P1#18 plants showed the same fluorescence level before the stress treatments and after the treatments. Specifically, the lowered PCD rate in the AER-overexpressing plants was ascribed to the scavenging of carbonyls by AER. These pharmacological and genetic results indicate the direct contribution of oxylipin carbonyls to PCD in the H₂O₂- and NaCl-stressed roots.

Figure 9.

Distribution of ROS in roots of the wild-type tobacco line SR1 and the AER-overexpressing line P1#18. Six-day-old plants were treated with H₂O₂ or NaCl as in Figure 7. After a 20-h incubation, root tips of 5-mm length were excised from the primary root. DCF fluorescence was recorded as in “Materials and Methods.” Typical images of DCF fluorescence are shown. Bar = 250 µm.

Strength of Each Carbonyl to Cause PCD

The α,β-unsaturated aldehydes such as HNE and acrolein are highly electrophilic and generally highly reactive, whereas saturated aldehydes such as n-hexanal and propionaldehyde are less reactive. To obtain insights into the roles of individual carbonyls that are formed in stressed plant cells, we evaluated the PCD-inducing strength of these carbonyls. We defined the strength of a carbonyl species as the lowest concentration that would cause DNA laddering in tobacco BY-2 cells (Table I; Supplemental Figs. S4 and S5). As expected, the strongest carbonyls were acrolein and HNE. These caused the DNA laddering, TUNEL-positive nuclei, and cytoplasm retraction (Fig. 4; Supplemental Fig. S3) at concentrations as low as 0.2 mm. The next strongest carbonyls were n-hexanal, n-heptanal, and (Z)-3-hexenal (3 mm required for PCD), and the weakest was propionaldehyde (50 mm).

Comparison of the strength of carbonyls to induce PCD in BY-2 tobacco cells as determined by DNA fragmentation

Table I.

Comparison of the strength of carbonyls to induce PCD in BY-2 tobacco cells as determined by DNA fragmentation

Cells were incubated with a carbonyl at the indicated concentrations for 20 h. Fragmentation of the chromosomal DNA was judged based on the generation of 180 bp and its multiples by agarose gel electrophoresis. + and –, Detectable and undetectable levels of the DNA fragments, respectively. nt, Not tested.

Carbonyl	Concentration
	0.1 mm	0.2 mm	0.5 mm	3 mm	10 mm	50 mm
Acrolein	–	+	+	+	nt	nt
HNE	–	+	+	nt	nt	nt
(Z)-3-Hexenal	–	–	–	+	+	nt
n-Hexanal	–	–	–	+	+	nt
n-Heptanal	–	–	–	+	+	nt
Propionaldehyde	–	–	–	–	–	+

We also tested the PCD-inducing ability of carbonyls in planta and found that acrolein and HNE at 1 mm and propionaldehyde at 10 mm effectively developed PCD markers in root hairs of Arabidopsis, such as protoplast retraction and a concomitant loss of the FDA-fluorescing ability (Supplemental Fig. S11). Thus, the PCD-inducing strength of carbonyls differs by species, but most species of oxylipin carbonyls that were generated in H₂O₂-stressed cells (Fig. 3) can cause PCD.

DISCUSSION

Oxylipin Carbonyls Can Mediate H₂O₂- and NaCl-Induced PCD

We investigated the roles of oxylipin carbonyls in plant injury under oxidative stress. As experimental models, we used H₂O₂-induced damage to tobacco BY-2 cells and H₂O₂- and NaCl-induced damage to roots of tobacco and Arabidopsis. In these experiments, we found that the stress treatment commonly induced PCD and that the scavenging of endogenous oxylipin carbonyls with chemical scavengers or a specific enzyme blocked the PCD. We identified the oxylipin carbonyl species that were increased in H₂O₂-treated BY-2 cells. These species when added exogenously could induce PCD in BY-2 cells and roots. From these results, we concluded that oxylipin carbonyls, downstream products of ROS, significantly contributed to the initiation of PCD in H₂O₂- or NaCl-stressed cells.

The involvement of oxylipin carbonyls in plant PCD shown in this study may not be surprising. For animal cells, it is established that reactive oxylipin carbonyls such as HNE have the potential to cause apoptosis, and their involvement in oxidative signal-induced PCD has been studied extensively (for review, see Dalleau et al., 2013). For plant cells, the toxicity of oxylipin carbonyls has been recognized recently, but their ability to produce cellular damage has not been investigated. An achievement in this study was that we identified the carbonyl species that are generated and can cause PCD in stressed cells. In addition, our findings demonstrated that not only ROS but also oxylipin carbonyls were necessary for the development of PCD under H₂O₂ and NaCl stress. Specifically, these oxylipin carbonyls were the chemical entities that initiated plant PCD. Although such critical involvement of oxylipin carbonyls may be valid only for certain types of stress, our results provide a new clue to understanding the toxicity of ROS in plant cells.

Multiple Species of Oxylipin Carbonyls Appear to Contribute to PCD

The oxylipin carbonyls identified in the H₂O₂-stressed cells included the α,β-unsaturated carbonyls acrolein and HNE, the saturated carbonyls n-hexenal, n-heptanal, and propionaldehyde, and (Z)-3-hexenal. All carbonyls can modify proteins via Schiff-base formation with an amino group. The α,β-unsaturated carbonyls are known to have higher cytotoxicity than other types of carbonyls because they are strong electrophiles and can form Michael adducts with the Cys, His, and Lys residues in addition to the Schiff base. Indeed, HNE and acrolein induced PCD in BY-2 cells at 0.3 mm. What we did not expect was that saturated carbonyls such as n-hexanal and n-heptanal also have the ability to initiate PCD. The strength order of these carbonyls to induce PCD corresponded with the order of chemical reactivity (Table I).

Considering that n-hexenal and n-heptanal were more greatly increased than acrolein and HNE in the H₂O₂-stimulated cells, we suggest the possibility that these saturated carbonyls also substantially contributed to the induction of PCD in the stressed cells. The inhibition of PCD by the overexpression of AER does not exclude the involvement of saturated carbonyls, because in the AER-overproducing plants, the scavenging of the α,β-unsaturated carbonyls resulted in the suppression of the stress-induced increase in saturated carbonyls, probably as a secondary effect (Yin et al., 2010). These two types of carbonyls might mediate PCD by different mechanisms (i.e. via different receptors/sensors [discussed below]). It is necessary to investigate the effect of distinct carbonyls on different pathways.

In addition, the finding that various carbonyls have the capacity to induce PCD may explain the diversity and number of carbonyl-scavenging enzymes. There are four classes of carbonyl-scavenging enzymes in plants: aldehyde dehydrogenases (Kirch et al., 2001), aldo-keto reductases (Oberschall et al., 2000), 2-alkenal reductases (Mano et al., 2005), and glutathione S-transferases (Gronwald and Plaisance, 1998). In the Arabidopsis genome, a dozen isozymes or more are encoded for each class. These enzymes in different classes have different substrate specificities, and even in the same class, isozymes show different specificities. There must have been selective pressure on plants to obtain a large set of scavenging enzymes to harness oxylipin carbonyls so that they would not trigger PCD unless it is to be commanded.

Oxylipin Carbonyls as Mediators of Oxidative Signals in Environmental Stress Responses

ROS-induced cell death in plants has been demonstrated for various environmental stressors (Mittler et al., 2011; Petrov et al., 2015). In this study, we investigated the cell death induced by H₂O₂ and NaCl. NaCl treatment activates the plasma membrane NADPH oxidases to facilitate superoxide production, from which H₂O₂ is formed via catalysis with superoxide dismutase. NaCl treatment also activates polyamine oxidase, which catalyzes the production of H₂O₂ in the oxidation of spermine and spermidine. Thus, NaCl treatment also imposes H₂O₂ stress on plant cells (Petrov et al., 2015). When they were stressed, both cultured cells and root cells showed TUNEL-positive nuclei, cytosolic retraction, and DNA laddering, which are typical of apoptosis-like PCD (Reape and McCabe, 2008).

Two oxidative signal pathways for PCD are reported: one is the mitochondria-dependent activation of caspase-like proteases, as in animal cell apoptosis (Reape and McCabe, 2008), and the other is the mitogen-activated protein kinase (MAPK) cascade (Nakagami et al., 2006). In either case, the receptors or sensors for ROS have been largely elusive (Gadjev et al., 2008). Our finding that the oxylipin carbonyls caused PCD suggests that these compounds participate in the signal pathway, at least in part, as mediators just downstream of ROS.

The PCD-initiating role of oxylipin carbonyls in whole plants may not be restricted to NaCl stress. The overexpression of carbonyl-scavenging enzymes improves a plant’s tolerance not only to salinity but also to heavy metals, drought (Sunkar et al., 2003), UV-B light (Hideg et al., 2003), strong light (Mano et al., 2010), and aluminum (Yin et al., 2010). Among these stressors, UV-B light and aluminum are known to cause PCD. Analysis of the mode of cell death and oxylipin carbonyls will clarify the roles of these compounds in PCD induced by the different conditions.

From a broader perspective, ROS-triggered PCD has been observed in a hypersensitive response after pathogen infection (Torres et al., 2005) and certain developmental events such as lateral root cap shedding (Pennell and Lamb, 1997), aleurone layer death during the germination of cereal grains (Bethke and Jones, 2001), the development of tracheary elements in the xylem of vascular plants (Kuriyama and Fukuda, 2002), leaf senescence (Zapata et al., 2005), and the nucellar cell elimination during endosperm formation in Sechium edule (Lombardi et al., 2010). Because the production of oxylipin carbonyls is closely associated with the occurrence of ROS, we can expect their involvement in these physiological responses. The use of carbonyl scavengers such as carnosine or enzymes in combination with the chemical analysis of the involved carbonyls will be a good test to judge the broader physiological aspects of the operation of the carbonyl signal in plant PCD.

How Do Oxylipin Carbonyls Initiate PCD?

Because of their high reactivity, the α,β-unsaturated carbonyls such as HNE and acrolein may exert indiscriminate damage to proteins, but they target only limited types of proteins in vivo. Winger et al. (2007) analyzed the HNE modification of mitochondrial proteins in Arabidopsis cells. They identified 16 distinct proteins that were sensitively modified under oxidative stress. In salt-stressed Arabidopsis leaves, even when the tissue started to die, only 17 protein species showed more than a 2-fold increase in HNE modification (Mano et al., 2014b). Thus, oxylipin carbonyls have chemical specificity required for ROS signaling (Møller and Sweetlove, 2010). In other words, specific receptors/sensors of oxylipin carbonyls may exist. Indeed, the redox-regulated Nonexpresser of PR gene proteins and the TGA transcription factors are candidates for the carbonyl sensors (Farmer and Mueller, 2013) but not for PCD induction.

As judged by the morphological change and DNA fragmentation, the H₂O₂- and NaCl-induced PCD in this study was of the apoptosis-like mode, in which a mitochondria-dependent activation of proteases (caspase-like proteases) is involved (Reape and McCabe, 2008). In the mitochondria-dependent apoptosis in animal cells, several proteins are suggested as HNE targets, such as the Fas protein and the tumor suppressor/cell cycle regulator protein p53 (Dalleau et al., 2013), but plant cells do not have proteins apparently homologous to these HNE targets. The primary effect of HNE on plant mitochondria might be the inactivation of sensitive target enzymes such as pyruvate dehydrogenase complex and Gly decarboxylase complex (Taylor et al., 2002). A rapid consumption of glutathione via Michael adduct formation with the α,β-unsaturated carbonyls (Esterbauer et al., 1991) would also affect the redox status of mitochondria. These changes would facilitate the deterioration of the mitochondrial membranes and the formation of pores on the outer membrane to allow the release of mitochondrial PCD-associated proteins.

On the other hand, in the H₂O₂-inducible PCD in Arabidopsis, a signal transduction mechanism via the regulation of proteasomal degradation is known to be involved in the MAPK cascade. Specifically, MEKK1, the MAPK kinase kinase, which is the most upstream factor of the H₂O₂-induced Arabidopsis MAPK3 (AtMPK3)- and AtMPK6-dependent PCD pathway, is regulated in a proteasome-dependent manner (Nakagami et al., 2006). It was reported in a study of mammalian cells that oxylipin carbonyls are involved in proteasome-dependent signal regulation via the carbonyl sensor Kelch-like ECH (for erythroid cell-derived protein with Cap‘n’Collar [CNC] homology)-associated protein1 (Keap1; Higdon et al., 2012). In the relaxed state of the cells, Keap1 binds the transcription factor Nuclear factor E-2-related factor2 (Nrf2), and this binding facilitates the ubiquitination and the subsequent proteasome-dependent degradation of Nrf2, so as to repress the Nrf2-dependent gene expression. HNE formed under oxidative stress modifies Keap1 on the Cys-151 residue. The modified Keap1 no longer binds Nrf2, which then escapes from ubiquitination. Thus, Keap1 receives and transduces the carbonyl signal via the attenuation of the proteasome-dependent degradation of Nrf2 to induce defense responses (Higdon et al., 2012). It is unclear how H₂O₂ suppresses the proteasome-dependent MEKK1 degradation, but it is likely that MEKK1 degradation is regulated by a protein that is functionally similar to Keap1. Oxylipin carbonyls may act on such a regulator.

We recently identified 17 species of proteins that are sensitively modified with HNE in the leaves of Arabidopsis under salt stress (Mano et al., 2014b). Among them, several proteins are candidates to determine cell fate, as follows. One is the chloroplastic peptidyl-prolyl cis-trans-isomerase (also named cyclophilin20-3), which plays a key role in Cys biosynthesis. The modification of its redox-active Cys residues with 12-oxo-phytodienoic acid, a long-chain oxylipin carbonyl, triggers the formation of Cys synthase complex, thereby activating sulfur assimilation and building up cellular redox potential (Park et al., 2013). Modification of these carbonyl-prone Cys residues with short-chain oxylipin carbonyls might have different effects on the protein’s function.

Another candidate is Nitrilase1 (Nit1), a highly conserved protein that functions as a tumor suppressor (Sun et al., 2009). In Arabidopsis, changes in the polymer structure and the intracellular localization of Nit1 are closely associated with the transition from proliferation to differentiation (Doskočilová et al., 2013). It is unclear, however, whether the HNE modification of Nit1 causes such changes. We are now investigating the effects of the carbonyl modification of each target protein on cell fate.

MATERIALS AND METHODS

Culture of Cells and Plants

The suspension of tobacco (Nicotiana tabacum) BY-2 cells was cultured in MS medium supplemented with Suc (30 g L⁻¹), myoinositol (100 mg L⁻¹), KH₂PO₄ (200 mg L⁻¹), thiamine HCl (0.5 mg L⁻¹), and 2,4-dichlorophenoxyacetic acid (0.2 mg L⁻¹), pH 5.6. The cells were cultured in darkness at 25°C with continuous rotation at 120 rpm. Every 7 d, 50 mg of cells was transferred to a flask of 50 mL of fresh medium. Four-day culture cells were used for PCD treatment. In this exponential growth phase, cells remain in the maximum mitotic stage (Nagata et al., 2004). Cells were collected by filtration and washed once with distilled water for analysis.

Wild-type tobacco ‘Petit Havana’ SR1, the transgenic line P1#18 that overexpresses Arabidopsis (Arabidopsis thaliana) AER (Mano et al., 2005), and Arabidopsis ecotype Columbia-0 were grown vertically on 1% (w/v) agar plates (one-half-strength MS medium), in sterile conditions, at 23°C in a 14-h-light/10-h-dark cycle (light intensity, 100 μmol m^–2 s^–1 with white fluorescence lamps).

Trypan Blue Staining for Cell Death

A 0.2-mL aliquot of cell suspension was added to a mixture of 0.3 mL of 3% (w/v) Suc and 0.5 mL of 4% (w/v) Trypan blue solution. After 10 min, cells were washed once with phosphate-buffered saline (PBS), and then 10 μL of the cell suspension was transferred to a microscopic slide, and the dead (Trypan blue-stained) and live (unstained) cells were counted. Cell death is expressed as the percentage of dead cells.

Isolation of Genomic DNA and Gel Electrophoresis

Approximately 200 mg of BY-2 cells was ground with a mortar and pestle that were chilled with liquid nitrogen. The resulting powder was transferred to a 1.5-mL microtube, and 0.7 mL of extraction buffer containing 2% (w/v) cetyltrimethylammonium bromide, 1.4 m  NaCl, 20 mm EDTA, 100 mm Tris-HCl, pH 8, and 0.2% (v/v) β-mercaptoethanol was added immediately. The mixture was incubated at 65°C for 45 min. After centrifugation at 12,000g for 15 min at room temperature, the supernatant was treated with RNase A (10 mg mL⁻¹) at 37°C for 1 h and then mixed well with an equal volume of the chloroform:isoamyl alcohol mixture (24:1). After centrifugation at 10,000g for 12 min, the upper aqueous phase was collected. The DNA was precipitated with a two-thirds volume of isopropanol, washed in 70% (v/v) ethanol, dried, and resuspended in sterile distilled water. The DNA was then electrophoresed on a 1% (w/v) agarose gel followed by visualization with ethidium bromide.

Detection of Nuclear DNA Fragmentation Using a TUNEL Assay

We used the In Situ Cell Death Detection Kit: Fluorescence (Roche Diagnostics) to determine cell death, according to the manufacturer’s instructions with minor modification. Four-day culture cells were treated for 20 h with H₂O₂ or acrolein plus scavengers and then harvested for staining. The cells were fixed with 4% (w/v) fresh paraformaldehyde in PBS and labeled by the TUNEL reaction mixture: the terminal deoxynuleotidyl transferase solution and the label solution described in the kit’s manual. The transferase was omitted for the negative control. For the positive control, we incubated fixed and permeabilized cells with DNase I to induce DNA strand breaks prior to the TUNEL reaction. The technique was adjusted for the study of the whole root and observation of the epidermal cells. Roots were permeabilized for 20 min in a solution containing 0.1% (w/v) Triton-X, 0.1% (w/v) sodium citrate, 0.1 mm KCl, and 0.1 mm CaCl₂. The stained cells and roots were analyzed with a fluorescence microscope (Leica AF6000). The excitation and emission wavelengths were 488 and 515 nm, respectively.

Detection and Quantification of Carbonyls Using HPLC

Freshly harvested cells (0.45 g) were added to 3 mL of acetonitrile containing 60 nmol of 2-ethylhexanal (as an internal standard) and 0.005% (w/v) butylhydroxytoluene and then incubated in a screw-capped glass tube at 60°C for 30 min. Carbonyls in the extract were derivatized with 2,4-dinitrophenylhydrazine, and the dinitrophenylhydrazone derivatives were extracted and analyzed by HPLC as described (Yin et al., 2010). We identified dinitrophenylhydrazone derivatives of carbonyls by their retention times and determined their contents by a comparison with authentic compounds (Matsui et al., 2009).

ROS Detection with H₂DCF Diacetate and H₂O₂ Detection with the BES-H₂O₂-Ac Probe

BY-2 cells, collected with a brief centrifugation followed by a washing with distilled water, were incubated in 20 µm  H₂DCF diacetate (Molecular Probes) in PBS at 37°C for 30 min or 50 µm BES-H₂O₂-Ac (Wako Pure Chemical) for 30 min in darkness. The cells were washed twice with PBS. For tobacco and Arabidopsis roots, 5 mm of root apex was excised and incubated in 50 µm  H₂DCF diacetate in 20 mm potassium phosphate, pH 6, or in 50 µm BES-H₂O₂-Ac in darkness for 30 min and washed two times with PBS. The fluorescence was monitored using a microscope with excitation at 488 nm and emission at 530 nm for both DCF and BES-H₂O₂.

Detection of LOOHs with Spy-LHP

BY-2 cells were collected via filtration and washed once with distilled water. Then, 0.4 g of cells was taken in a tube, and 100 µL of suspension medium (50 mm MES-NaOH, pH 6.5, and 35 mm  NaCl) and 2.7 µm Spy-LHP (Dojindo Laboratories) in ethanol (900 µL) were added. After a 30-min incubation at 37°C, the samples were centrifuged at 12,000g for 2 min. The supernatant was collected, and its fluorescence was measured with a spectrofluorometer (FP 8300; JASCO) with excitation at 524 nm and emission at 538 nm. To determine the amount of peroxides, a standard curve was prepared with the model LOOH  m-chloroperbenzoic acid (Khorobrykh et al., 2011).

Visualization of the Accumulation of Carbonyls in Roots

Root tips (approximately 2 cm from the tip) were excised and stained with Schiff’s reagent (Wako) for 20 min at room temperature, rinsed with a freshly prepare sulfite solution (0.5% [w/v] K₂S₂O₂ in 0.05 m HCl), and kept in the sulfite solution and observed using a light microscope (Leica LED3000).

Root Hair PCD Assay

Roots were stained with FDA (Wako) for the detection of PCD. Only viable root hairs are able to cleave FDA to form fluorescein and fluoresce green using a microscope with excitation at 485 nm. Dead cells do not fluoresce, and their protoplasts retract from the cell wall. Roots were stained in a 10 µg mL^–1 solution of FDA on microscope slides, and root hairs were immediately observed using a fluorescence microscope (Leica AF6000).

Supplemental Data

The following supplemental materials are available.

• Supplemental Figure S1. Chromatograms of carbonyls in BY-2 cells (5 h experiment).

• Supplemental Figure S2. Scavengers' effects on carbonyl levels in BY-2 cells (5 h experiment).

• Supplemental Figure S3. Induction of PCD in BY-2 cells by HNE.

• Supplemental Figure S4. Induction of PCD in BY-2 cells by n-hexanal.

• Supplemental Figure S5. DNA fragmentation by carbonyl treatments on BY-2 cells.

• Supplemental Figure S6. Scavengers' effects on H₂O₂ levels in BY-2 cells.

• Supplemental Figure S7. H₂O₂-induced PCD in Arabidopsis roots.

• Supplemental Figure S8. NaCl-induced PCD in Arabidopsis roots.

• Supplemental Figure S9. Accumulation of carbonyls in Arabidopsis roots.

• Supplemental Figure S10. Scavengers' effects on H₂O₂ in Arabidopsis roots.

• Supplemental Figure S11. Root hair PCD made by carbonyls in Arabidopsis.

ACKNOWLEDGMENTS

We thank Atsushi Sakamoto (Hiroshima University) for providing the tobacco BY-2 cells and Kenji Matsui (Yamaguchi University) for critical reading of the article.

Glossary

 
• ROS

  reactive oxygen species

 
• H₂O₂

  hydrogen peroxide

 
• LOOH

  lipid peroxide

 
• HNE

  4-hydroxy-(E)-2-nonenal

 
• AER

  2-alkenal reductase

 
• PCD

  programmed cell death

 
• BY-2

  Bright Yellow-2

 
• TUNEL

  terminal deoxynucleotidyl transferase dUTP nick end labeling

 
• H₂DCF

  2′,7′-dihydrodichlorofluoresein

 
• DCF

  dichlorofluorescein

 
• FDA

  fluorescein diacetate

 
• MS

  Murashige and Skoog

 
• PBS

  phosphate-buffered saline

 
• NaCl

  salt

Author notes

1

This work was supported by the Japan Society for the Promotion of Science (KAKENHI grant no. 26440149).

2

Present address: Science Research Center, Yamaguchi University, Yoshida 1677–1, Yamaguchi 753–8515, Japan.

*

Address correspondence to [email protected].

The author responsible for distribution of materials integral to the findings presented in this article in accordance with the policy described in the Instructions for Authors (www.plantphysiol.org) is: Jun’ichi Mano ([email protected]).

M.S.B. performed most of the experiments; J.M. supervised the experiments.

[OPEN]

Articles can be viewed without a subscription.

www.plantphysiol.org/cgi/doi/10.1104/pp.115.256834