Lipoxygenase in singlet oxygen generation as a response to wounding: in vivo imaging in Arabidopsis thaliana

Wounding, one of the most intensive stresses influencing plants ontogeny and lifespan, can be induced by herbivory as well as by physical factors. Reactive oxygen species play indispensable role both in the local and systemic defense reactions which enable “reprogramming” of metabolic pathways to set new boundaries and physiological equilibrium suitable for survival. In our current study, we provide experimental evidence on the formation of singlet oxygen (1O2) after wounding of Arabidopsis leaves. It is shown that 1O2 is formed by triplet-triplet energy transfer from triplet carbonyls to molecular oxygen. Using lipoxygenase inhibitor catechol, it is demonstrated that lipid peroxidation is initiated by lipoxygenase. Suppression of 1O2 formation in lox2 mutant which lacks chloroplast lipoxygenase indicates that lipoxygenase localized in chloroplast is predominantly responsible for 1O2 formation. Interestingly, 1O2 formation is solely restricted to chloroplasts localized at the wounding site. Data presented in this study might provide novel insight into wound-induced signaling in the local defense reaction.

Under reducing conditions, LOOH is reduced by transition metals to lipid alkoxyl radical (LO • ) which might further abstract hydrogen from another polyunsaturated fatty acid forming another lipid radical (L • ) and hydroxy polyunsaturated fatty acids (lipid hydroxide, LOH) 25 . Under oxidizing condition, LOOH is oxidized to lipid peroxyl radical (LOO • ) by oxidized transition metals, ferric heme iron of cytochrome c, peroxynitrite, chloroperoxide, and hypochlorous acid. The cyclization of LOO • is known to form a cyclic endoperoxide (dioxetane), whereas recombination of the two LOO • forms a linear tetroxide 26,27 . These high energy intermediates decompose to triplet carbonyls ( 3 L=O * ) which might transfer triplet energy either to pigments forming excited pigments or molecular oxygen forming singlet oxygen ( 1 O 2 ). In addition, tetroxide might decompose directly to 1 O 2 by Russell mechanism [28][29][30] (Fig. 1).
Singlet oxygen imaging with SOSG showed that 1 O 2 is formed in the wounded Arabidopsis leaves 31 . The authors proposed that 1 O 2 formation is accompanied by recombination of two LOO • via Russell mechanisms formed during enzymatic lipid peroxidation. Nevertheless, no experimental data supporting this proposal have been published yet. In this study, we provide experimental evidence on the role of lipoxygenase in 1 O 2 formation after wounding of Arabidopsis plants. Formation of 3 L=O * monitored by ultra-weak photon emission and 1 O 2 detected by the Singlet Oxygen Sensor Green (SOSG) fluorescence assessed by a laser confocal scanning microscopy was studied in lox2 mutant, which lacks chloroplast lipoxygenase. Our data revealed that in wounded Arabidopsis plants, the lipoxygenase plays a key role in the formation of 3 L=O * and 1 O 2 .

Material and Methods
Chemical Reagents. All chemicals were purchased of analytical grade from Sigma Aldrich GmbH (Germany) and Molecular Probes Inc. (Eugene, OR, USA). Arabidopsis   soaking in distilled water. The plants were grown at a photoperiod of 16 h light/8 h dark (photon flux density 100 µmol photons m −1 s −1 ) for 6 weeks at a temperature of 25 °C and at 60% relative humidity. The wounding of plant leaves was performed under diffused green light using sharp blade while any external mechanical pressure was avoided at other parts on the surface of plant leaves. Measurement was performed 20 min after wounding.

Arabidopsis plants.
Charge coupled device imaging. Highly sensitive CCD camera VersArray 1300B (Princeton instruments, Trenton, NJ, USA) was used for two-dimensional photon imaging. Dark current of the CCD camera was achieved by cooling it down to −110 °C using a liquid-nitrogen cooling system. Spectral sensitivity of the CCD camera was within the range of 350-1000 nm. The quantum efficiency was almost 90% in the visible range of the spectrum. The measurement was done in the image format of 1340 × 1300 pixels and the data correction was done by subtracting the background noise from every measurement. The following CCD camera parameters were used: scan rate, 100 kHz; gain, 2 and accumulation time of 20 min, based on the parameters described in Prasad et al. 33 . The CCD camera was situated in a black box located in an experimental dark room with an approximate dimension of 3 m × 1.5 m × 5 m. To avoid any possible interference by external light, the data recording computer was installed in the outer dark room. Prior to the two-dimensional ultra-weak photon emission measurements, the Arabidopsis plant was dark-incubated for approximately 2 h duration to prevent any intervention of delayed luminescence. The data accumulation from Arabidopsis plants and leaves were started 20 min after wounding.
Confocal laser scanning microscopy. Imaging of 1 O 2 was achieved using Singlet oxygen sensor green (SOSG). Wounding of Arabidopsis leaves was exerted by a sharp razor blade used to cut ca 5 × 5 mm pieces from marginal leaf lamina while avoiding other mechanical injury. These cuts were incubated in the presence of 50 µM SOSG for 30 min, washed in 40 mM HEPES buffer (pH 7.5) and forthwith visualized by Fluorview 1000 confocal laser scanning microscope (Olympus Czech Group, Prague, Czech Republic). The excitation was achieved by a 488 nm line of an argon laser and the emission recorded using a 505-525 nm band-pass filter. Negative controls were treated with 5 mM catechol during the staining procedure. Integral distribution of fluorescence signal intensity corresponding to singlet oxygen localization within figures was visualized in FV10-ASW 4.0 Viewer software (Olympus).

Triplet carbonyl formation in WT Arabidopsis monitored by ultra-weak photon emission.
Formation of 3 L=O * in WT Arabidopsis plants subjected to wounding was monitored by two-dimensional imaging of ultra-weak photon emission. It is well established that singlet chlorophylls ( 1 Chl * ) contribute predominatly to photon emission in plant tissue 27 . As 1 Chl * is formed solely by excitation energy transfer from 3 L=O * to chlorophylls, ultra-weak photon emission might serve as an indirect indicator of 3 L=O * formation. Figure 2 shows a photograph (A) and two-dimensional image of ultra-weak photon emission (B) measured in nonwounded and wounded leaves. Spontaneous ultra-weak photon emission observed from the non-wounded area of Arabidopsis plants is caused by oxidative metabolic processes. The wounding of Arabidopsis plant ( Fig. 2A, red circles; Supplementary data 1) resulted in the enhancement of ultra-weak photon emission from the cut edges of the leaves which are caused by wound-induced oxidative processes (Fig. 2B, red circles). The spatial profile of photon emission shows a higher photon emission at the injured part (Fig. 2C). This observation indicates that wounding of Arabidopsis plants results in 3 L=O * formation restricted solely to the wounded areas of plant. To verify the contribution of the propagative reaction and associated ultra-weak photon emission, we have measured photon emission up to 2 h after wounding (Supplementary data 2). It can be observed that the ultra-weak photon emission as a response to wounding last in the scale of hours. Thus, it can be hypothesized that the ultra-weak photon emission originates at the initial stage due to burst of oxidative reactions after injury and propagative reaction continuing up to few hours.

Effect of catechol on triplet carbonyl formation in WT Arabidopsis.
To study the involvement of lipoxygenase in 3 L=O * formation, the effect of lipoxygenase inhibitor catechol on ultra-weak photon emission was studied in WT Arabidopsis leaves. It is well established that binding of catechol to the ferric non-heme iron of lipoxygenase leads to inactivation of the enzyme active site 34 . Figure 2 shows a photograph (D) and two-dimensional imaging of ultra-weak photon emission (E) measured in wounded Arabidopsis leaves in the absence (left) and presence (right) of catechol. The topical application of catechol on the wounded areas of the Arabidopsis leaves suppressed significantly ultra-weak photon emission ( Fig. 2E and F). The spatial profile of photon emission measured in the presence of catechol (red dotted rectangle) shows that photon emission was decreased to the level of photon emission observed in the non-wounded site (Fig. 2F). As a control, the exogenous application of catechol (n = 3) was also tested in the non-wounded leaves which showed no changes in ultra-weak photon emission (Supplementary data 3). These results reveal that lipoxygenase is involved in 3 L=O * formation in Arabidopsis leaves exposed to wounding.

Formation of triplet carbonyls in lox2 mutant. To further clarify the involvement of lipoxygenase in
SCientiFiC REPORtS | 7: 9831 | DOI:10.1038/s41598-017-09758-1 Singlet oxygen formation in WT Arabidopsis detected by confocal laser scanning microscopy. To visualize the formation of 1 O 2 in wounded Arabidopsis leaves, we used SOSG which is highly sensitive and specific fluorescent probe for 1 O 2 . SOSG fluorescence was detected by confocal laser scanning microscopy. In this method, the formation of SOSG endoperoxide by cycloaddition of 1 O 2 to SOSG results in the enhancement of SOSG fluorescence. Figure 3 represents the Differential interference contrast; SOSG fluorescence and DIC + fluorescence channel for comparison of tissue/cell details measured at different magnifications (for objectives 10x, 20x and 40x) where the margins indicate the wounding site which reflects higher SOSG fluorescence signal. SOSG fluorescence was pronouncedly higher in the wounded area compared to none or very low SOSG fluorescence in the intact area of Arabidopsis leaf lamina. Highest SOSG fluorescence signal originated mainly from the first layer of cells on the cutting edge ( Fig. 3 and Fig. 4I,A-C). However, the number of cell The ultra-weak photon emission was measured utilizing highly sensitive CCD camera. The photographs (A,D and G) and the corresponding two-dimensional images of ultra-weak photon emission (B,E and H). In B, twodimensional images of spontaneous ultra-weak photon emission from non-injured and mechanically injured part of the leaves of Arabidopsis was measured. In E, ultra-weak photon emission imaging was measured in the absence (left leaf) and presence of 5 mM catechol (right leaf) in WT Arabidopsis leaves and in H; ultra-weak photon emission imaging was measured in WT (left leaf) and lox2 mutants (right leaf) of Arabidopsis leaves. Prior to the measurements, the Arabidopsis plants and leaves were kept in the complete darkness for a period of 2 hrs. The circles in red indicate the mechanically injured part of the leaves. Ultra-weak photon emission imaging was measured after 20 min of wounding with an integration time of 20 min. In C,F and I, the spatial profile of photon emission in a particular strip of the image [Y = 433; Y = 657 and Y = 594, respectively] is presented. Y-axis denotes the number of photon counts accumulated after 20 min, whereas the X-axis denotes pixel of the image. In C, the dotted rectangle represents the position within the Arabidopsis plants at the point of mechanical injury. In F, the dotted rectangle represents the position at the point of mechanical injury (black dotted rectangle) and mechanical injury + catechol (red dotted rectangle) in WT Arabidopsis leaves and in I, the dotted rectangle represents the position within the Arabidopsis leaves at the point of mechanical injury in WT (black dotted rectangle) and lox2 mutant (red dotted rectangle). The evaluation was performed by transporting the photon intensity at pixel points for the total image frame (1300 × 1340 pixels).
layers with signal can be influenced by the sharpness of razor blade, i.e. mechanical injury intensity within tissues (Fig. 3, 10X). These observations revealed that wounding in leaves results in 1 O 2 formation at the site of injury restricted predominantly to the first, i.e. most impacted, layer of cells and a limited signal from adjoining cells.

Effect of catechol on singlet oxygen formation in WT Arabidopsis.
To confirm the involvement of lipoxygenase in 1 O 2 formation caused by wounding, the effect of catechol on SOSG fluorescence was measured in WT Arabidopsis leaves. Figure 4I shows the Nomarski DIC (D), SOSG fluorescence (E) and SOSG intensity (F) measured in wounded Arabidopsis leaves in the presence of catechol. Topical application of catechol on the wounded area of Arabidopsis leaves caused significant suppression of SOSG fluorescence (Fig. 4I,E) as compared to non-catechol treated Arabidopsis leaves (Fig. 4I,B). The distribution of SOSG fluorescence intensity reveals that SOSG fluorescence in wounded Arabidopsis leaf (Fig. 4I,C) was lower as compared to wounded Arabidopsis leaf treated with catechol (Fig. 4I,F). Based on these results, it is concluded that lipoxygenase is involved in the formation of 1 O 2 in Arabidopsis leaves under wounding.

Singlet oxygen formation in lox2 mutant.
To identify the involvement of lipoxygenase in the formation of 1 O 2 during wounding in Arabidopsis leaves, SOSG fluorescence was measured in lox2 mutant. Negligible SOSG fluorescence was observed in lox2 mutant at the cut edges of the wounded Arabidopsis leaves (Fig. 4I-H). The distribution of SOSG fluorescence intensity shows that SOSG fluorescence in wounded leaf of lox2 mutant (Fig. 4I-I) is pronouncedly lower as compared to wounded WT Arabidopsis leaf (Fig. 4I-C). These observations reveal that chloroplast lipoxygenase LOX2 play a key role in 1 O 2 formation during wounding in plants. The negligible SOSG fluorescence in few numbers of cells on the cut edge is believed to be contributed by the lipoxygenase located within the cell other than the chloroplasts. Supplementary data 5 and Fig. 4II shows the intensity of SOSG  fluorescence channel of confocal images from non-injured edge and the injured edge of WT Arabidopsis leaves, respectively. The results indicate an enhancement in intensity of SOSG fluorescence by 3 times in wounded edge of Arabidopsis leaf as compared to non-wounded areas of the Arabidopsis leaves. The effect to catechol was observed to suppress the SOSG intensity close to the value comparable to non-injured area of the Arabidopsis leaf.

Effect of desferal and trolox on singlet oxygen formation in WT and lox2 mutant. The effect
of Desferal (deferoxamine mesylate), which is an iron chelator that forms nontoxic ferrioxamine; can attenuate iron-induced oxidative stress and also known to interact with LO • was measured in WT and lox2 mutant of Arabidopsis leaves. Figure 5 (Fig. 5F) and lox2 mutant (Fig. 5H) as compared to non-desferal treated Arabidopsis leaves (Fig. 5B and D). As a termination agent for lipid peroxidation, effect of trolox (2-carboxy-2,5,7,8-tetramethyl-6-chromanol) which is a water soluble analogue of vitamin E was measured in WT and lox2 mutant of Arabidopsis leaves. Topical application of trolox on the wounded area of Arabidopsis leaves caused complete suppression of SOSG fluorescence in both WT (Fig. 5J) and lox2 mutant (Fig. 5L) as compared to non-trolox treated Arabidopsis leaves (Fig. 5B and D).
In addition, ultra-weak photon emission imaging was measured in non-wounded (right leaves of the panel) and wounded (left leaves of the panel) of WT and lox2 Arabidopsis leaves, respectively in the presence of desferal and trolox showing pronounced suppression in ultra-weak photon emission (Supplementary data 6). Based on these results, it is concluded that inhibition of the propagation and termination step of lipid peroxidation can lead finally to negligible 1 O 2 generation.

Discussion
In plants, local response to wounding comprises of oxidative damage of lipids and proteins at the wounding site, whereas systemic response mediated by hormones such as jasmonic acid, ethylene, salicylic acid, and abscisic acid is widespread over the plant tissue and organs 35 . In this study, we provided evidence that 3 L = O * formed during lipid peroxidation results in 1 O 2 formation in the local response to wounding.

Triplet carbonyl formation.
It is known that wounding of plant tissue is accompanied by oxidative damage of lipids and proteins 11 . Two-dimensional imaging of ultra-weak photon emission which is known to be a non-invasive indicator of oxidative stress 36 was found to be pronouncedly enhanced at the site of wounded plant tissue ( Fig. 2; Supplementary data 1). In agreement with our data, Flor-Henry et al. 37 proposed that lipid peroxidation occurs under wounding in detached Arabidopsis leaves using ultra-weak photon emission. More recently, suppression of LOH formation in lox2 mutant revealed that enzymatic lipid peroxidation is initiated by lipoxygenase 38 . Oxidative burst characterized by ROS production is known to be generated in plant tissues in response to wounding in plants 17,39 . Recent reports on studies involving Pisum sativum and other plant models have claimed activation of NADPH oxidase in response to wounding 17,[39][40][41] . Evidences have been provided on direct detection of superoxide anion radical (O 2 •− ) and hydrogen peroxide (H 2 O 2 ) measured during wounding in Arabidopsis leaves as monitored by NBT and DAB staining 42 . Due to the fact that light enhanced NBT and DAB signals, the authors proposed that O 2 •− and H 2 O 2 formation is related to electron transport. In addition, the treatment of Arabidopsis leaves with calcium blockers and calcium chelators after wounding of leaves abolished ROS signal indicating the involvement of calcium in the pathways that couples perception of wounding with the generation of ROS 43 . It has also been reported that LOX2 can be activated by calcium ion; however, its direct interaction is not sufficiently understood 44-47 . Singlet oxygen formation. In vivo imaging of 1 O 2 using SOSG fluorescence measured by confocal laser scanning microscopy revealed that wounding of Arabidopsis leaves caused 1 O 2 formation. The observation that lipoxygenase inhibitor catechol completely suppressed 1 O 2 formation indicates that lipid peroxidation is initiated by lipoxygenase. Suppression of 1 O 2 formation in lox2 mutant reveals that lipoxygenase localized in chloroplast is predominantly responsible for 1 O 2 formation. The observation that 1 O 2 formation is localized solely at the site of the wounded plant tissue indicates that 1 O 2 unlikely diffuse to surrounding plant tissue. Under dark conditions, the chloroplasts are known to be situated near the periphery attached to the cell membrane of the cells. In the mechanically injured Arabidopsis leaves, the SOSG fluorescence was observed in the periphery close to the cell membrane indicating the generation of 1 O 2 localized predominantly in the chloroplasts (Fig. 6). The SOSG fluorescence was observed in layers adjoining the site of mechanical injury indicating that the oxidative radical reaction occurs predominantly close to the site of mechanical injury and that the chain reaction is limited to a close proximity. The termination of chain reaction is likely to occur due to limitation of presence of initiators of the oxidative radical reaction which cannot diffuse to longer distance due to its shorter half-life period. The less probable reason which cannot be neglected completely can be the limited diffusion of the SOSG probe.
Physiological relevance. Based on the results obtained and understanding from our current study, the response of wounding and generation of 1 O 2 can lead to hypothesis on existence of wound induced signaling pathway mediated by 1 O 2 . The signal observed is predominantly contributed by the chloroplasts which were found suppressed almost entirely in the lox2 mutant leads to the conclusion that lipoxygenase plays a major role in wound-induced 1 O 2 production which is in agreement with experiments performed in model system 48 . However, a lower signal as observed using confocal microscopy may indicate the diffusion of 1 O 2 to the neighboring cells. The direct contribution of 1 O 2 or oxidized biomolecules can thus be hypothesized to play a role in cellular signaling and opens a new perspective in the signaling pathway 49 . It is proposed here that 1 O 2 formed during wounding in plants can be involved in oxidation of either lipids or proteins which can act as signaling molecule.