The antagonistic effect of hydroxyl radical on the development of a hypersensitive response in tobacco Sheng Deng 1,2, *, Mina Yu 1,2, *, Ying Wang 1 , Qin Jia 1 , Ling Lin 2 and Hansong Dong 1 1 Key Laboratory of Monitoring and Management of Crop Diseases and Pest Insects, Ministry of Agricuture of R. P. China, Department of Plant Pathology, College of Plant Protection, Nanjing Agricultural University, China 2 Institute of Plant Protection, Jiangsu Academy of Agricultural Sciences, Nanjing, China Introduction In the course of their development, plants are often confronted with various potential pathogens. To pre- vent infection by these pathogens, plants have devel- oped a set of inducible defence response systems. The activation of plant defence responses is initiated through the recognition of pathogen-associated molecular patterns by plant receptors or the recogni- tion of pathogen effectors (avirulence proteins) by plant resistance gene products (R proteins) [1,2]. The defence responses include the production of reactive oxygen species (ROS) and phytoalexins; the reinforce- ment of cell walls; the deposition of callose; the Keywords elicitin; hydroxyl radicals; hypersensitive response; reactive oxygen species; riboflavin Correspondence H. Dong, Key Laboratory of Monitoring and Management of Crop Diseases and Pest Insects, Ministry of Agricuture of R. P. China, Department of Plant Pathology, College of Plant Protection, Nanjing Agricultural University, Nanjing 210095, China Fax: +86 25 8439 5325 Tel: +86 25 8439 9006 E-mail: hsdong@njau.edu.cn *These authors contributed equally to this work and are regarded as joint first authors (Received 19 June 2010, revised 24 August 2010, accepted 12 October 2010) doi:10.1111/j.1742-4658.2010.07914.x Reactive oxygen species (ROS) are important signalling molecules in living cells. It is believed that ROS molecules are the main triggers of the hyper- sensitive response (HR) in plants. In the present study of the effect of ribo- flavin, which is excited to generate ROS in light, on the development of the HR induced by the elicitin protein ParA1 in tobacco (Nicotiana tabacum), we found that the extent of the ParA1-induced HR was diminished by hydroxyl radical (OH • ), a type of ROS. As compared with the zones trea- ted with ParA1 only, the HR symptom in the zones that were infiltrated with ParA1 plus riboflavin was significantly diminished when the treated plants were placed in the light. However, this did not occur when the plants were maintained in the dark. Trypan blue staining and the ion leak- age measurements confirmed HR suppression in the light. Further experi- ments proved that HR suppression is attributed to the involvement of the photoexcited riboflavin, and that the suppression can be eliminated with the addition of hydrogen peroxide scavengers or OH • scavengers. Fenton reagent treatment and EPR measurements demonstrated that it is OH • rather than hydrogen peroxide that contributes to HR suppression. Accom- panying the endogenous OH • formation, suppression of the ParA1-induced HR occurred in the tobacco leaves that had been treated with high-level abscisic acid, and that suppression was also removed by OH • scavengers. These results offer evidence that OH • , an understudied and little appreci- ated ROS, participates in and modulates biologically relevant signalling in plant cells. Abbreviations ABA, abscisic acid; DAB, 3,3¢-diaminobenzidine; G + GOX, glucose plus glucose oxidase; H 2 O 2 , hydrogen peroxide; HR, hypersensitive response; O 2 •) , superoxide radical; OH • , hydroxyl radical; PA, ParA1 plus adenine; PC, ParA1 plus catalase; PR, ParA1 plus riboflavin; PT, ParA1 plus thiourea; PV, ParA1 plus ascorbic acid; ROS, reactive oxygen species; SOD, superoxide dismutase; WIPK, wounding-induced protein kinase. FEBS Journal 277 (2010) 5097–5111 ª 2010 Nanjing Agricultural University. Journal compilation ª 2010 FEBS 5097 expression of pathogenesis-related proteins; and, most drastically, the hypersensitive response (HR), which is a form of programmed cell death. The HR leads to cell death at the infection sites, thus limiting pathogen growth and establishing systemic acquired resistance in the whole plant [3,4]. Although numerous studies have been carried out regarding the development of the HR induced by plant pathogens or pathogen-derived elicitors, the underlying molecular networks have not been well elucidated. Undoubtedly, the early stage of HR development involves some or all of the key events, including Ca 2+ influx, ROS burst, mitogen-activated protein kinase cascades, nitric oxide production, cytochrome c release, lipid peroxidation and phytohormone imbalance [2,5– 7]. In these events, ROS burst is most often considered to be the protagonist to the occurrence of the HR [8,9]. ROS, especially superoxide radical (O 2 •) ), hydrogen peroxide (H 2 O 2 ) and hydroxyl radicals (OH • ), are very important signalling molecules [6]. They play pivotal roles in plant processes as diverse as development, growth, the response to biotic and abiotic stimuli, and programmed cell death [10–12]. In plant cells, O 2 •) can be converted into H 2 O 2 by superoxide dismutase (SOD), and H 2 O 2 can be converted into highly toxic OH • by the Fenton reaction or by peroxidase under certain conditions [11,13]. Plant cells can sense the var- iation of ROS production in location, amount, type, rate and duration; these variations direct the subse- quent responses of cells [12,14–17]. To date, the signalling roles of H 2 O 2 and O 2 •) have been intensively analysed, whereas other types of ROS, such as OH • and singlet oxygen ( 1 O 2 ), have been lar- gely ignored [11]. O 2 •) ,H 2 O 2 and 1 O 2 can be detoxi- fied by the internal antioxidants or by some specific enzymes. However, no specific scavenger for OH • has been identified in plants. As a result of its high reac- tiveness, OH • is under strict control in plants [14,18]. However, this does not mean that OH • has no other roles than that of a destroyer in cells. Foreman et al. [19] reported that ROS are involved in plant cell growth and root hair elongation. Specifically, OH • can loosen cell walls, thus helping plant organs to elongate and seeds to germinate [20–22]. The natural killer cell, which is an essential constituent of host defence sys- tems in humans, is activated by OH • exclusively, and OH • scavengers can inhibit the activity of this type of cell [23]. The present study showed that OH • can nega- tively regulate the HR development induced by elicitin and bacterium (Xanthomonas spp.) in tobacco ( Nicotiana tabacum L. cv NC89), thereby suggesting that under certain circumstances, OH • may differ significantly from other species of ROS with regard to biological activity. Purified elicitors are usually applied to given plants for the elucidation of plant defence mechanisms. ParA1, an elicitor derived from Phytophthora parasiti- ca, belongs to the elicitin family, which comprises a group of 10 kDa small proteins secreted by oomycetes of the genera Phytophthora and some Pythium [24–26]. In most Nicotiana species, elicitins such as cryptogein, ParA1 and INF can trigger various defence responses, including the HR [25,27–29]. Riboflavin, which is a water-soluble vitamin essen- tial to living cells, can reversibly accept or lose a pair of hydrogen atoms. Therefore, its two derivatives, flavin adenine dinucleotide and flavin mononucleotide, can perform key metabolic functions as coenzymes in electron transfer processes. It is generally believed that photoexcited riboflavin can generate various ROS under visible light [30–33]. Specifically, O 2 •) ,H 2 O 2 and OH • are generated through a type I reaction, whereas 1 O 2 is produced through a type II reac- tion [34]. In the present study, by exploiting the character of riboflavin, we found that OH • can inhibit the development of the HR triggered by ParA1 in tobacco leaves. Results The antagonistic effect of photoexcited riboflavin on the development of the HR induced by ParA1 It is known that ParA1 can induce the HR in tobacco leaves [25]. However, in the current experiment, HR development was suppressed by photoilluminated ribo- flavin. The purified ParA1 protein (obtained from the Pichia pastoris expression system; see Materials and methods section) and ParA1 plus riboflavin (PR) were infiltrated into different parts of a leaf (simplified as the same leaf or leaves in the following sections). After 48 h exposure to light, the HR spread over the entire ParA1-infiltrated zones, whereas just some small HR lesion spots were scattered in the PR-treated zones. By contrast, in the dark treatment, the infiltration of the two materials yielded similar effects on HR develop- ment (Fig. 1A). Additionally, in the zones infiltrated with riboflavin and the control (ParA1 purification buffer), no HR-like lesion spots could be found, either in the light or in the dark (Fig. 1A). To identify the most effective concentration of ribo- flavin for the suppression of the HR, ParA1 protein was mixed with riboflavin at concentrations ranging from 5 to 100 mgÆL )1 (data not shown). The most Suppression effect of hydroxyl radical on HR S. Deng et al. 5098 FEBS Journal 277 (2010) 5097–5111 ª 2010 Nanjing Agricultural University. Journal compilation ª 2010 FEBS effective riboflavin concentration was found to be 50 mgÆL )1 , which was used to produce the results shown in Fig. 1A. The trypan blue staining also supported the pheno- types. The HR extent in the PR zone was diminished as compared with that in the ParA1 zone, and the riboflavin zone was not stained as expected (Fig. 1B,C). Furthermore, ion leakage measurement exhibited the cell death progression in different infil- trated zones at the indicated time points (Fig. 1D). The extent of the HR in the PR zone was significantly suppressed as compared with that in the ParA1 zone. In the ParA1 zone, the ion leakage increased dramati- cally within 12–24 h, whereas in the PR zone the ion leakage rose to a peak of  25% at 12 h and then started to decline. The further development of HR symptom in these zones was monitored from 48 to 96 h, but no changes were noted, except that the dead zones in the leaves became dry and crisp. The results indicate that the photoexcited riboflavin can suppress the HR rather than simply delay its development. Photoexcited riboflavin suppresses the HR development by disturbing the HR signalling Riboflavin did not affect the HR development of plants that were kept in the dark (Fig. 1A). These find- ings suggest that riboflavin did not disturb the HR sig- nalling and the initial contact recognition between ParA1 and its receptor under dark condition. PR and ParA1 were infiltrated into the same leaf and the treated plants were placed in the dark. After 2, 4, 8 and 12 h, at least six plants at each time were transferred from the dark to the continuous light (termed predark treatment). Forty-eight hours after infiltration, ion leakage in each treated zone was mea- sured (Fig. 2B). The ParA1 zones were set as controls, with similar ion leakage levels found for all groups. In the PR zones, the ion leakage ratios of the first three predark groups climbed from 12% in the 2 h group to  25% in the 8 h group, and then the ratio rose dras- tically in the 12 h group, approximating to the level of ParA1 zones. It is known that the fluorescence inten- sity of riboflavin provides a means of assessing its level of concentration [35]. In the PR zones, the increase in ion leakage from 4 to 12 h could not be attributed to the decrease in riboflavin content, because no signifi- cant changes in fluorescence intensity had been found in these zones during predark treatment (Fig. 2B). In addition, four other groups were treated in a reverse manner (Fig. 2A; prelight treatment). In the PR-trea- ted zones of these plants, only the 2 h prelight group had a high ion leakage ratio ( 50%); in other prelight groups the ratio reduced to  10% or even lower. The mRNA levels of several elicitin response genes were investigated by semiquantitative RT-PCR A B D C Fig. 1. The antagonistic effect of riboflavin on HR development occurred under light. (A) ParA1, Rf (50 mgÆL )1 riboflavin), PR (ParA1 + 50 mgÆL )1 riboflavin) and Control (ParA1 purification buffer) were infiltrated into different interveinal segments in the same tobacco leaves. The treated leaves were photographed 48 h after they were placed under continuous light (lower row) or in the dark (upper row). The experiment was repeated five times (four plants per repeat, with two kept in the dark and two kept in the light) with simi- lar results (scale bars = 1 cm). (B) Dead cells were stained in situ with trypan blue. The leaf was stained 12 h after treatment, and infil- trated areas were encircled by a black line. (C) The details of each zone were observed and recorded under a universal microscope (scale bars = 500 lm). (D) The time course of cell death was moni- tored by ion leakage measurements in the different treatment zones. The ion leakages from leaf discs obtained from the corresponding zones were measured at the indicated time points after infiltration. Con represents control treatment. Mean values ± standard error of at least three replicates of ion leakage measurements are presented. S. Deng et al. Suppression effect of hydroxyl radical on HR FEBS Journal 277 (2010) 5097–5111 ª 2010 Nanjing Agricultural University. Journal compilation ª 2010 FEBS 5099 (Fig. 2D). With regard to the HR hallmark genes, including hypersensitive-related (hsr) Hsr515, Hsr203J and sensitivity-related (str) Str319 [36], the accumula- tion of their transcripts was inhibited and delayed sig- nificantly during the first 9 h in the PR-treated zones as compared with the ParA1-treated zones. Similar results were found for alternative oxidase, which is a marker gene for mitochondrial dysfunction [37]. Lipox- ygenase-1, a key HR-dependent gene [38], was substan- tially suppressed within 6–12 h. The transcript levels of the wounding- induced protein kinase (WIPK) gene and the 3-hydroxy-3-methylglutaryl CoA reductase gene, which represent the activation of WIPK and salicylic acid-induced protein kinase signalling pathways, respectively [5,39], were lower in the PR zones than in the ParA1 zones from 6 to 12 h. However, the tran- script levels of pathogenesis-related proteins 1a were not significantly different in ParA1 and PR treatment. Cytosolic ascorbate peroxidase was upregulated in all treatments, whereas catalase was only induced by ribo- flavin treatment. These results demonstrate that the photoexcited riboflavin can suppress HR development by disturbing the HR signalling rather than by blocking the contact recognition between ParA1 and its receptor. The ROS generated from photoexcited riboflavin are involved in HR suppression The photolytic products of riboflavin obtained by plac- ing riboflavin under light for 24 h failed to inhibit the development of the HR (data not shown). This result indicated that some active species for HR suppression had disappeared after riboflavin photolysis. Attention was then focused on the ROS that were generated dur- ing the riboflavin photolysis process, such as O 2 •) , H 2 O 2 ,OH • and 1 O 2 . In Fig. 3A,B, the riboflavin-infiltrated zone, which was covered by aluminium foil, could not be stained by the H 2 O 2 probe 3,3¢-diaminobenzidine (DAB). However, the uncovered zone could be stained by the probe, as was the zone that was infiltrated with glucose plus glucose oxidase (G + GOX; taken as the positive control) [15]. Further proof was obtained by H 2 O 2 measurement. Compared with the zone infiltrated by H 2 O, the relative content of H 2 O 2 in the riboflavin- infiltrated zone was increased by  40% (Fig. 3C). The findings indicate that riboflavin can generate ROS in tobacco leaves in the presence of light irradiation. To investigate the role of different types of ROS in HR suppression, we introduced various free radical scavengers, including catalase (scavenger of H 2 O 2 ) [40], SOD (scavenger of O 2 •) ) [11], ascorbic acid (redox reg- A B C Fig. 2. Photoexcited riboflavin suppresses HR development by dis- turbing HR signalling. (A) In the predark and prelight groups, the ion leakage in the ParA1- and PR-treated zones was measured 48 h after infiltration. After infiltration, at least six plants at each time were first placed in the dark for the indicated times (2, 4, 8 and 12 h) and then transferred to the continuous light (labelled as pre- dark treatment) and others were treated reversely (labelled as pre- light treatment). The treated plants kept in the dark for 48 h after infiltration were used as the control and labelled as ‘dark’. Mean values ± standard error of three replicates are presented. (B) In PR-infiltrated zones, the fluorescence emitted from riboflavin was observed at the indicated times (0, 4, 8 and 12 h) after predark treatment. (C) The levels of different transcripts in the zones, which were treated with ParA1 (P), riboflavin (R) or PR, were assayed by semiquantitative RT-PCR. After infiltration, the plants were placed under continuous light and the corresponding zones were collected at the indicated times (3, 6, 9 and 12 h) for RNA isolation. Con rep- resents the leaves that did not receive any treatments, and were taken as the control. The constitutively expressed gene elongation factor 1-alpha was used as an internal reference transcript. Three independent experiments were completed with similar results. Suppression effect of hydroxyl radical on HR S. Deng et al. 5100 FEBS Journal 277 (2010) 5097–5111 ª 2010 Nanjing Agricultural University. Journal compilation ª 2010 FEBS ulator of cells) [11], l -histidine (scavenger of 1 O 2 and OH • ) [41], adenine and thiourea (scavengers of OH • ) [20,23]. Leaves treated with these scavengers or with the mixture of each scavenger plus riboflavin mani- fested no symptoms 72 h after infiltration (data not shown). However, in the zones infiltrated with ParA1 plus each scavenger, HR cell death occurred as it did in the ParA1-treated zones (Fig. 3D). These results reflect that HR development induced by elicitins is independent of ROS, which is consistent with previ- ously reported results [42–45]. The ROS scavengers, when coinfiltrated with PR, were able to restore the HR symptom, with the excep- tion of SOD (Fig. 3D). l-histidine, a scavenger of 1 O 2 and OH • , had little effect on HR restoration even at 100 mm. For comparison, the specific scavengers of OH • , adenine and thiourea, can restore the HR sub- stantially at 1 and 50 mm, respectively (Fig. 3D). These findings suggest that H 2 O 2 or OH • (but not O 2 •) or 1 O 2 ) is involved in the suppression of the HR. It is noteworthy that the HR symptom was further suppressed in the PRS (PR + SOD) zone, which is probably an outcome of much more H 2 O 2 and OH • derived from SOD catalysing O 2 •) (Fig. S2). Because H 2 O 2 is the precursor of OH • [11,46], and because the OH • scavengers as well as catalase and ascorbic acid could restore the HR symptom (Fig. 3D), it was hypothesized that OH • was the HR suppressor. The suppressor role of OH • in HR development is verified by the Fenton reaction and EPR measurements OH • is introduced and generated by a Fenton-type reaction that takes place in the presence of Fe 2+ and H 2 O 2 [13,47]. To obtain stable and moderate H 2 O 2 generation, we used the G + GOX system again. In the presence or absence of Fe 2+ , ParA1 and G + GOX at increasing concentrations were coinfil- trated into the same leaves. Because Fe 2+ ⁄ Fe 3+ may interfere with the results of ion leakage measurement, five grades were adopted to assess the severity of the HR induced by ParA1 (Fig. 4A). In the Fe 2+ -present groups, the HR symptoms were significantly sup- pressed, with grade 0 (i.e. no visible HR lesion) accounting for  60% of the treated leaves or for even higher proportions in some treatments (Fig. 4A). Con- versely, in the Fe 2+ -absent groups, grade IV accounted for more than 60%. Moreover, the compromised HR was observed when ParA1 was coinfiltrated with Fe 2+ (Fig. 4A). All the suppression effect on HR develop- ment could be eliminated with the addition of the scavengers of H 2 O 2 or OH • (Fig. S3A). A C B D Fig. 3. The ROS generated from photoilluminated riboflavin influ- ence HR development. (A) In situ detection of H 2 O 2 was per- formed by DAB staining 1.5 h after treatment. Riboflavin (Rf; 50 mgÆL )1 ) and the positive control (G + GOX; 14 mM glucose and 2.5 unitsÆmL )1 glucose oxidase) were infiltrated into tobacco leaves. The upper part of the leaf was covered with aluminium foil (shaded area in the left-hand picture), and the lower part of the leaf was exposed to light. The experiment was repeated three times (two leaves per repeat) with similar results. (B) The details of DAB staining were observed under a universal microscope. The control was taken from the untreated zone (scale bars = 100 lm). (C) As compared with the H 2 O treatment, the relative content of H 2 O 2 in the riboflavin-infiltrated zone was measured 1.5 h after treatment. Mean values ± standard error of at least three repli- cates are presented. (D) Different ROS scavengers were used to investigate the role of corresponding type of ROS in HR suppres- sion. After infiltration, the treated plants were placed in the light, and the HR extent was evaluated by the ion leakage ratio at 12 and 48 h after infiltration. Mean values ± standard error of three replicates are presented. The treatment by protein purification buf- fer was regarded as the control. [P, ParA1; R, riboflavin (50 mgÆL )1 ); C, catalase (2000 unitsÆmL )1 ); S, SOD (100 uni- tsÆmL )1 ); V, ascorbic acid (50 mM); T, thiourea (50 mM); A, adenine (1 m M); H, L-histidine (100 mM); RC, riboflavin + catalase; PRC, ParA1 + riboflavin + catalase; other abbreviations follow the same pattern.] S. Deng et al. Suppression effect of hydroxyl radical on HR FEBS Journal 277 (2010) 5097–5111 ª 2010 Nanjing Agricultural University. Journal compilation ª 2010 FEBS 5101 The EPR method, a reliable way to analyse the for- mation of OH • [48,49], was applied to the evaluation of the OH • level during riboflavin photolysis in vitro and in vivo. In Fig. 4B, a strong signal was detected, which means that OH • is generated by riboflavin in vitro under light irradiation. The addition of the OH • scavenger adenine, and the dark maintenance of the sample could substantially reduce the signal. As expected, riboflavin also had similar effects in vivo (Fig. 4C). Against the H 2 O infiltration, the relative content of OH • in the riboflavin-treated zone was increased by  50% (Fig. 4D). These results supported A B C D a b c d e a b c d e Suppression effect of hydroxyl radical on HR S. Deng et al. 5102 FEBS Journal 277 (2010) 5097–5111 ª 2010 Nanjing Agricultural University. Journal compilation ª 2010 FEBS the previous findings that OH • is involved in HR sup- pression. It should also be noted that light irradiation can affect the background signal of spin trapping reagent [Fig. 4B (c) and (e)]. However, the signal is not produced by OH • generation, because the OH • scaven- ger cannot reduce the signal [Fig. 4B (d) and (e)]. As a result of endogenous OH • generation, HR development is suppressed by exogenous application of high-level abscisic acid Previous studies have proved that the exogenous appli- cation of high-level abscisic acid (ABA) can induce the accumulation of catalytic Fe, which is critical for the generation of OH • through the Fenton reaction [50]. ParA1, PR, PA (ParA1 plus adenine), PT (ParA1 plus thiourea), PV (ParA1 plus ascorbic acid) and PC (ParA1 plus catalase) were infiltrated into the same leaves of tobacco plants, which had been sprayed with 100 lm ABA, 400 lm ABA or H 2 O and kept in the light for 24 h. Twenty-four hours after ParA1 infiltration, the development of the HR in the leaves that had been pre- treated with ABA was significantly suppressed as com- pared with the leaves that had been pretreated with H 2 O (Fig. 5A,D). After 48 h, the HR further developed in the leaves pretreated with 100 lm ABA, but less in the leaves pretreated with H 2 O and 400 lm ABA. The EPR assays provided the evidence that the 400 lm ABA-pre- treated leaves had a higher OH • level after ParA1 treat- ment, and the level increased by  20% against H 2 O treatment (Fig. 5B,C). These findings suggest that the HR induced by ParA1 can be suppressed by the exoge- nous application of high-level ABA, probably as a result of the endogenous OH • formation. Additionally, in the PR-infiltrated zone, because of the ABA pretreatment, the HR suppression effect of photoexcited riboflavin was synergistic (Fig. 5A). In the zones infiltrated with OH • scavengers, both adenine and thiourea could restore HR development in ABA-pretreated leaves, although the formation of HR lesions in the PT zone lagged behind that seen in the PA zone, where the HR lesion could be observed 16 h after infiltration (Fig. 5A,D). Similarly, in PV and PC zones, the HR in ABA-pretreated leaves could be res- cued by the application of ascorbic acid and catalase (Fig. 5A,D). The phytopathogenic bacterium Xanthomonas oryzae pv. oryzicola (stain RS105), can cause disease in their host rice; in nonhost plants, such as tobacco, they can trigger the HR [51]. ABA- and H 2 O-pretreated leaves were infiltrated with stain RS105 and RS105 plus ade- nine. As expected, against H 2 O-pretreated leaves, the HR induced by RS105 was suppressed in ABA-pre- treated leaves, and this suppression could be restored by the addition of adenine (Fig. 5D). In addition, it was affirmed that the suppressed HR cannot be explained by the preferential decrease in the viable bacterial population, which was caused by the unfa- vourable environment in plant cells (data not shown). These findings indicate that the exogenous application of high-level ABA suppressed the HR development triggered by ParA1 and RS105 in tobacco leaves through the involvement of OH • , which was generated by the accumulation of ROS and catalytic Fe. Discussion In the present study, the HR induced by ParA1 was sig- nificantly suppressed when ParA1 and riboflavin were coinfiltrated, and the suppression only took place in the light and not in the dark. Further analysis suggested that the major suppressor was OH • derived from photo- excited riboflavin. In addition, the endogenous OH • could also suppress the development of the HR induced by ParA1 and the bacteria strain RS105 in tobacco. Accordingly, it is proposed that OH • , an understudied and little appreciated ROS, has an antagonistic effect in HR development by disturbing the HR signalling. Fig. 4. OH • was involved in the suppression of the HR. (A) The HR induced by ParA1 could also be suppressed by the Fenton reaction reagent. Five grades of HR severity were depicted and labelled as 0, I, II, III and IV. PG 0.5 O 0.5 (ParA1 mixed with 0.5 mM glucose and 0.5 unitsÆmL )1 glucose oxidase), PG 1 O 1 ,PG 5 O 2.5 were infiltrated into the same tobacco leaves in the presence or absence of Fe 2+ (0.5 mM, FeSO 4 ). ParA1 and ParA1 plus Fe 2+ were infiltrated as controls. The data were recorded and compiled 48 h after infiltration. Leaves that showed the HR extent of grade IV or III in ParA1-treated zones were selected for calculation (in total 21 leaves from 10 plants). (B) The EPR method was used to detect OH • generation during riboflavin photolysis in vitro 1.5 h after exposure to light. (a) Riboflavin under light; (b) riboflavin plus adenine (1 m M) under light; (c) riboflavin in the dark; (d) adenine (1 mM) with spin trapping assay reagent under light; and (e) spin trapping assay reagent under light. (C) OH • generated from riboflavin photolysis in vivo was measured by EPR 1.5 h after treatment. (a) Leaf tissues treated with H 2 O plus spin trapping assay reagent as a negative control; (b) leaf tissues treated with riboflavin; (c) leaf tissues cotreated with adenine (1 m M) and riboflavin; (d) leaf tissues treated with the Fenton reaction reagent (5 mM glucose, 2.5 unitsÆmL )1 glucose oxidase and 0.5 m M Fe 2+ ) as a positive control; and (e) leaf tissues obtained from the untreated zone as a background for the spectra. All spectra were representative of at least three measurements under indicated conditions. (D) Against the H 2 O treatment, the relative content of OH • in the riboflavin zone and the riboflavin plus adenine zone were calculated from the above spectra. Mean values ± standard error of three replicates are presented. S. Deng et al. Suppression effect of hydroxyl radical on HR FEBS Journal 277 (2010) 5097–5111 ª 2010 Nanjing Agricultural University. Journal compilation ª 2010 FEBS 5103 Riboflavin is one of the pivotal vitamins for living organisms. It is also an excellent photosensitizer that can generate ROS under light irradiation. Over the years, due to its photosensitization, riboflavin has been extensively studied in medicine, pharmaceutical chemis- try, foodstuffs, nutrition and other fields [33,34,52–54], but less so in plants. As plants are the major source of riboflavin taken up by animals and also that, to date, little evidence in plants has been reported about the involvement of photoexcited riboflavin (including its two derivatives) in light stress and light injury, the field in question is worthy of research effort. To identify the real HR suppressor, we used various ROS scavengers, because the photoilluminated ribofla- vin can generate varied species of ROS, including O 2 •) ,H 2 O 2 ,OH • and 1 O 2 . It seems a little complicated on the surface. However, in fact, the application of SOD failed to restore HR symptoms and l-histidine had little effect on HR restoration (Fig. 3D), which suggests that O 2 •) and 1 O 2 do not participate in HR suppression. Further experiments provided more evi- dence: (a) only the Fenton reaction can significantly affect HR development (Fig. 4A); (b) the signal strength of EPR is correlated with HR suppression (Figs 4C,D and 5B,C); (c) the presence of Fe 2+ can also suppress the HR induced by ParA1 to a large extent (Fig. 4A). This evidence indicates that the chem- ical species involved in HR suppression is OH • . It is believed that plants pretreated with sublethal stress (e.g. ozone exposure, ultraviolet irradiation and methyl viologen treatment) can build up a resistance to the subsequent lethal stress and pathogen infections with less cell death. These phenomena are termed cross-tolerance or acclimation [55,56]. Initially, we A D B C Fig. 5. The exogenous application of ABA counteracted HR development. (A) The HR extent was affected by ABA pretreatment. Twenty- four hours before the infiltration with ParA1, PR, PA, PT, PV and PC, tobacco plants were sprayed with H 2 O and ABA (100 lM, A100; 400 l M, A400). After infiltration, the HR extent in leaves was recorded at the indicated times in terms of the grade depicted in Fig. 4A (upper panel). Each treatment was repeated in 15 leaves from eight plants, and all recorded data were compiled. Furthermore, the HR extents were also assessed by ion leakage measurement (lower panel). Mean values ± standard error of at least three replicates are pre- sented. (B) The EPR measurements indicated that the leaves pretreated with ABA had a higher OH • level after ParA1 treatment. H 2 O, H 2 O pretreatment; H + ParA1, ParA1 infiltration after H 2 O treatment; A400, 400 lM ABA pretreatment; A + ParA1, ParA1 infiltration after ABA treatment. All the spectra were representative of at least three measurements under the indicated conditions. (C) The relative content of OH • in the H + ParA1, A400 and A + ParA1 zones against the H 2 O zone. Mean values ± standard error of three replicates are presented. (D) With ABA pretreatments, the development of the HR was compromised or suppressed. 1, ParA1; 2, PV; 3, PA; 4, PC; 5, PT; 6, RS105 (colony-forming unit = 10 8 ); 7, RS105 + A; 8, PR. Pictures were taken for each treated leaf at the indicated times, and they represent the general results from 15 infiltrated leaves for each treatment. Suppression effect of hydroxyl radical on HR S. Deng et al. 5104 FEBS Journal 277 (2010) 5097–5111 ª 2010 Nanjing Agricultural University. Journal compilation ª 2010 FEBS assumed that the suppression of the HR in this study was due to acclimation, but the evidence refuted the assumption. First, it takes hours for the preparation of more severe stress or pathogen infection after pretreat- ment. However, in the PR-infiltrated zone, the ribofla- vin and ParA1 functioned almost simultaneously. Second, the ABA-pretreated leaves only showed small HR lesions after infiltration with ParA1, which could be regarded as probably acclimation. However, the application of ROS scavengers can restore the HR, and the addition of riboflavin, which can generate ROS in light, suppresses the HR completely. These outcomes are not feasible in accordance with the classical concept of acclimation. Liu et al. [57] found that the autophagy process negatively regulates the HR triggered by tobacco mosaic virus in tobacco plants with N protein. The reason offered by the authors was that autophagy can restrict prodeath signal(s) from diffusing. Xiong et al. [58] confirmed that autophagy can be induced by oxidative stress in Arabidopsis. These results are strongly reminiscent of the HR suppression that was found in the present study. However, when ParA1 was infiltrated into the leaf zones that had been trea- ted with 10 or 20 mm H 2 O 2 5 h previously, no HR suppression occurred (Fig. S3B). Therefore, the HR suppression presented here has nothing to do with autophagy. Nitric oxide may be involved in HR suppression in the present work, because the production balance between nitric oxide and H 2 O 2 is crucial to trigger the HR [59]. However, increasing the H 2 O 2 level by adding the G + GOX system or by eliminating H 2 O 2 with specific scavengers had no effect on the final HR symptoms induced by ParA1 (Figs 3D and 4A), which probably means that the imbalance of nitric oxide and H 2 O 2 is not the key reason for the HR suppression. In the present study, the ROS seemed to be a sup- pressor of the HR. In Fig. 3D, 12 h after infiltration, the ratios of iron leakage in the PC, PV and PA zones were increased against ParA1 treatment, which sug- gests that the HR symptoms were promoted in these zones. Similar results were obtained by other research groups. Ruste ´ rucci et al. [60] reported that the devel- opment of cell death was strongly delayed and dimin- ished when leaves were exposed to continuous light (7000 lux of white light) 24 h before elicitin treatment. A similar phenomenon was also observed by Tronchet et al. [61]. Moreover, correlating with the HR inhibi- tion, the activity of 9-lipoxygenase and galactolipases as well as their mRNA levels w as significantly suppressed by high light exposure (350 lmol quantaÆm )2 Æs )1 ) after cryptogein treatment [62]. These findings imply that at least one species of ROS may negatively regulate the development of the HR. The short-lived OH • is the most reactive species of ROS. It is very unstable, and it can rapidly attack bio- molecules that are located at the site of its generation [13,63]. So far, in plant cells, research regarding the functions of OH • has mainly focused on cell wall loos- ening and its actions related to oxidative stress, whereas other fields involving OH • have rarely been addressed [18,22,64]. For example, is it involved in any signalling pathway, and how does it perform its func- tion as a signalling molecule? Conversely, the movable and relatively stable H 2 O 2 has been extensively studied. It is prevailingly believed that H 2 O 2 can diffuse in cells and between organelles, and that it can conduct its functions by modifying thiol groups in certain pro- teins, such as transcription factors and protein kinases [11,14,46]. In addition to ParA1, HrpN Ea protein (UniProt accession number: Q01099), a well-known HR trigger in tobacco [65], was investigated in the present study. The experimental results prove that photoexcited ribo- flavin fails to suppress the HrpN Ea -induced HR, which is ROS dependent and mitochondrial dysfunction dependent [66,67] (Fig. S4). To some extent, the photo- excited riboflavin promotes HR development (data not shown). Because the HR induced by elicitin is chloro- plast dependent and fatty acid peroxidation dependent [38], the effect of OH • produced by photoexcited ribo- flavin on HR suppression should be signalling pathway dependent. In plant cells, the signature of fatty acid peroxida- tion can be changed after OH • treatment [68]. Fatty acid peroxidation can be performed either by free radi- cals (not including H 2 O 2 and O 2 •) ) or by lipoxygenase pathways. The former process yields 9(R,S), 12(R,S), 13(R,S) and 16(R,S) fatty acid hydroperoxides, whereas enzymatic oxygenation yields 9(S) and 13(S) fatty acid hydroperoxides exclusively [69]. It has been reported that 9(S)-hydroperoxides are the crucial sig- nalling molecules for the execution of the HR, and that the enantiomer composition of different fatty acid peroxidation products determines cell fates (i.e. death or survival) after cryptogein treatment [38,70]. In the present study, it was hypothesized that the enantiomer composition and signature of fatty acid hydroperox- ides were changed, probably as a result of modification by OH • , thus leading to HR suppression. As plants in nature are subject to biotic and abiotic stress simultaneously, it should not be ignored that OH • formed under abiotic stress will probably exert disturbing effects on biotic stress response pathways, especially on the fatty acid-dependent response pathways. S. Deng et al. Suppression effect of hydroxyl radical on HR FEBS Journal 277 (2010) 5097–5111 ª 2010 Nanjing Agricultural University. Journal compilation ª 2010 FEBS 5105 However, whether the interference is positive or negative has to b e d etermined on t he basis of g iven conditi ons and mechanisms. Materials and methods Chemicals Riboflavin was obtained from Calbiochem (Merck KGaA, Darmstadt, Germany), reduced glutathione from Roche (Basel, Switzerland). Catalase, SOD, ascorbic acid and ABA were purchased from Sigma (St Louis, MO, USA). All other mentioned reagents were of analytical grade. Plant culture condition Tobacco plants (N. tabacum L. cv NC89) were grown in a growth chamber at 25 °C with 16 h of light (50 lmol quantaÆm )2 Æs )1 ). All treatments were performed on plants 14–16 weeks old. Treatments of tobacco leaves with ParA1, riboflavin and other reagents Before treatment, all operations involving riboflavin were performed in subdued light to protect riboflavin from deg- radation. ParA1 (UniProt accession number: P41801), ribo- flavin and different reagents were infiltrated by needleless syringes in the second, third and fourth bottom leaves, which were intact but not etiolated. The infiltrated zones were marked with a black marking pen. Unless there were special requirements, all the experimental treatments were as follows: the intensity of light was 50 lmol quantaÆm )2 Æs )1 (cool white light) and the concentrations of ParA1 and riboflavin were 3 nm and 50 mgÆL )1 , respectively. H 2 O 2 detection in situ In situ H 2 O 2 generated by photoexcited riboflavin was detected with the use of the DAB staining method (Sigma) described by Thordal-Christensen et al. [71], with the following modifications. Treated leaves were vacuum infiltrated in 1 mgÆmL )1 DAB solution, pH 3.8, for 15 min. After incubation at 25 °C for 12 h in the dark, samples were transferred in 95% ethanol at 95 °C for the removal of chlorophyll. Samples were observed under a universal microscope and stored in 70% ethanol. The observation of the fluorescence emitted from riboflavin After infiltration, the treated plants were placed in the dark. Then at indicated times (0, 4, 8 and 12 h), the fluorescence emitted from riboflavin in PR-infiltrated zones was observed in  480 nm light under a stereomicroscope SZX12 (Olympus, Tokyo, Japan) equipped with a mercury lamp fluorescence system and an excitation filter (460– 490 nm). Measurement of H 2 O 2 in treated tobacco leaves In accordance with a ferrous ammonium sulphate ⁄ xylenol orange method [72] with some modification, the content of H 2 O 2 in the treated leaves was measured. The relative increase in H 2 O 2 in riboflavin-treated zones was calculated in comparison with the H 2 O-treated zones in the same leaves. An area of 3–5 cm 2 in the infiltrated zone was cut and crushed into a coarse powder in liquid nitrogen. Approximately 7.5 mg powder was loaded into a tube that contained 1.5 mL precooled 5% trichloroacetic acid. After gentle vibration to mix, the tube was placed in the dark at room temperature for 2 min. The homogenate was then centrifuged at 10 000 g,4°C for 2 min. Next, 500 lL supernatant was taken and mixed with 500 lL assay solu- tion (containing 500 lm ferrous ammonium sulphate, 200 lm sorbitol, 200 lm xylenol orange, 50 mm H 2 SO 4 and 2% ethanol). The mixtures and the blank sample (500 lL 5% trichloroacetic acid and 500 lL assay solution) were placed at room temperature in the dark for 30 min. The absorbance of the sample was then measured at 560 nm using a spectrophotometer (U-2800; Hitachi, Tokyo, Japan). On the basis of the absorbance value, the standard curve (obtained by adding the concentration gradient of H 2 O 2 ) and the loaded weight of plant tissue, the content of H 2 O 2 was calculated. Protein expression, purification and concentration evaluation The full-length cDNA, which encodes ParA1 protein (Uni- Prot accession number: P41801), had been cloned from Phytophthora parasitica in the previous study. The synthetic ParA1 gene (together with seven histidine codons attached at the 3 ¢ end), which was flanked by EcoRI and KpnI restriction sites at the 5¢ and 3¢ ends, respectively, was obtained by PCR with the following oligonucleotide prim- ers: 5¢-TGAATTC AATAATGTCTAACTTCCGCGCTCT- GTTC-3¢ and 5¢-AGGTACCTCAATGATGATGATGAT GATGATGCAGTGACGCGCACGTAGA-3¢. For the suc- cessful protein expression, a yeast expression consensus sequence (including ATG) must be added to the 5¢ end pri- mer (underlined). The correct ParA1 gene sequence was cloned into pPICZ-A, the expression plasmid, through EcoRI and KpnI restriction sites. In accordance with the manual in the EasySelect TM Pichia Expression Kit bought from Invitrogen (Carlsbad, CA, USA), the yeast strain KM71H was used for the expression of ParA1 protein, and the Escherichia coli strain Suppression effect of hydroxyl radical on HR S. Deng et al. 5106 FEBS Journal 277 (2010) 5097–5111 ª 2010 Nanjing Agricultural University. Journal compilation ª 2010 FEBS [...]... the basis of the gradient concentration of bovine serum albumin, the ParA1 concentration was evaluated by quantity one, a software contained in Bio-Rad Gel Documentation System 2000 Finally, we obtained  75 ngÆlL)1 ( 6500 nm) ParA1 protein (Fig S 1A, B), which can induce the HR at 6000· dilution on tobacco leaves in 24 h Measurement of ion leakage from leaf discs Cell death can be assayed by measuring... respectively The spectra shown were obtained from the accumulation of five single scans RNA isolation and RT-PCR ParA1, riboflavin and PR were in ltrated into different interveinal segments in the same tobacco leaves, and the samples were obtained from the corresponding in ltrated regions at the indicated times (four leaves from two plants per time point) The sample from the untreated plant leaves was taken as the. .. supplementary material is available: Table S1 Information about the genes investigated in this study for RT-PCR Fig S1 The plasmid constructed for the expression of ParA1 in Pichia pastoris and the tricine–SDS-PAGE analysis of purified ParA1 Fig S2 The synergized effect of HR suppression in the PRS (ParA1, riboflavin and SOD) zone Suppression effect of hydroxyl radical on HR Fig S3 The in uence of the Fenton... water for 3 h in a growth chamber The conductivity of the bathing solution was measured with a conductivity meter (SY-2, Institute of Soil Science, Chinese Academy of Sciences, China): value A The leaf discs were then returned to the bathing solution and incubated at 95 °C for 25 min After cooling to room temperature the conductivity of the bathing solution was measured again: value B The result of. .. mixture was placed into a glass capillary with an inside diameter of 1.1 mm for the EPR assay The parallel control without riboflavin was carried out in a similar way For the in vivo assay of photoexcited riboflavin, the mixture was in ltrated into leaves with needleless syringes, and the treated plants were placed in the light After 1.5 h, two pieces of leaf tissue (6 mm in width, 15 mm in length) obtained... peroxide J Invest Dermatol 105, 608–612 31 Frati E, Khatib AM, Front P, Panasyuk A, Aprile F & Mitrovic DR (1997) Degradation of hyaluronic acid by photosensitized riboflavin in vitro Modulation of the effect by transition metals, radical quenchers, and metal chelators Free Radic Biol Med 22, 1139–1144 32 Hasan N, Ali I & Naseem I (2006) Photodynamic inactivation of trypsin by the aminophylline-riboflavin system:... precipitated in the presence of phosphate For the plants that were pretreated with ABA or H2O, ParA1 and the spin trapping assay reagent were coinfiltrated into the leaves, and the plants were placed under light After 6 h, the levels of OH• in the leaf tissues obtained from in ltrated zones were measured by EPR in a similar way to that described above EPR measurements The measurements were performed with an EMX... component of the cryptogein signaling pathway leading to defense responses and hypersensitive cell death in tobacco Plant Cell 14, 1937–1951 Takahashi Y, Berberich T, Miyazaki A, Seo S, Ohashi Y & Kusano T (2003) Spermine signalling in tobacco: activation of mitogen-activated protein kinases by spermine is mediated through mitochondrial dysfunction Plant J 36, 820–829 ´ Rusterucci C, Montillet JL, Agnel... al TOP10 F¢ was used for the construction of the expression plasmid pPICZA::ParA1 Ni2+ affinity columns (HisTrapÔ HP; Amersham Bioscience, Buckinghamshire, UK) were applied for ParA1 protein purification In accordance with the instructions, the optimal concentrations of imidazole for binding (100 mm) and elution (350 mm) were chosen to purify the protein, and most of the purified ParA1 was collected in. .. in the second tube and stored at )80 °C The purified recombinant ParA1 protein was determined by tricine–SDS ⁄ PAGE with the use of 4% stacking gels and 10% separation gels in accordance with the method introduced by Schagger [73] The protein electrophoresis ¨ was performed by constant current (10 mA in stacking gels, 15 mA in separation gels) with PowerPAC Universal (Bio-Rad, Hercules, CA, USA) On the . basis of the gradi- ent concentration of bovine serum albumin, the ParA1 con- centration was evaluated by quantity one, a software contained in Bio-Rad Gel. ParA1 zone, the ion leakage increased dramati- cally within 12–24 h, whereas in the PR zone the ion leakage rose to a peak of  25% at 12 h and then started