Báo cáo khoa học: The antagonistic effect of hydroxyl radical on the development of a hypersensitive response in tobacco pot

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 tabac

The antagonistic effect of hydroxyl radical on the

development of a hypersensitive response in tobacco

Sheng Deng1,2,*, Mina Yu1,2,*, Ying Wang1, Qin Jia1, Ling Lin2and Hansong Dong1

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 infiltratrea-ted 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 O2, hydrogen peroxide; HR, hypersensitive response; O2•), 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.

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 Ca2+

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 (O2 )), hydrogen

peroxide (H2O2) 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, O2)can

be converted into H2O2 by superoxide dismutase

(SOD), and H2O2 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 H2O2 and O2•)have

been intensively analysed, whereas other types of ROS,

such as OH• and singlet oxygen (1O2), have been

lar-gely ignored [11] O2 ), H2O2 and 1O2 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, O2 ), H2O2 and OH• are generated through a type I reaction, whereas 1O2 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

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)1riboflavin) 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.

(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 O2 ),

H2O2, OH•and1O2

In Fig 3A,B, the riboflavin-infiltrated zone, which was covered by aluminium foil, could not be stained

by the H2O2 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 H2O2 measurement Compared with the zone infiltrated by

H2O, the relative content of H2O2 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 H2O2) [40], SOD (scavenger of O2)) [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

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.

ulator of cells) [11], l-histidine (scavenger of 1O2 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 1O2 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 H2O2 or OH• (but not

O2 )or 1O2) 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 H2O2 and OH• derived from SOD catalysing O2 ) (Fig S2) Because

H2O2 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 Fe2+ and

H2O2 [13,47] To obtain stable and moderate H2O2 generation, we used the G + GOX system again

In the presence or absence of Fe2+, ParA1 and

G + GOX at increasing concentrations were coinfil-trated into the same leaves Because Fe2+⁄ Fe3+ 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 Fe2+-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 Fe2+-absent groups, grade IV accounted for more than 60% Moreover, the compromised HR was observed when ParA1 was coinfiltrated with Fe2+ (Fig 4A) All the suppression effect on HR develop-ment could be eliminated with the addition of the scavengers of H2O2 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 m M 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

H2O2 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 m M ); T, thiourea (50 m M ); A, adenine

(1 m M ); H, L -histidine (100 m M ); RC, riboflavin + catalase; PRC,

ParA1 + riboflavin + catalase; other abbreviations follow the same

pattern.]

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 H2O 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

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 H2O 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 H2O

(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 H2O 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 H2O

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 H2O-pretreated leaves were infiltrated with stain RS105 and RS105 plus ade-nine As expected, against H2O-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 PG0.5O0.5(ParA1 mixed with 0.5 m M glucose and 0.5 unitsÆmL)1glucose 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 m M , FeSO4) 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 m M ) 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 m M glucose, 2.5 unitsÆmL)1glucose 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 H2O 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.

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

O2), H2O2, OH•and1O2 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 O2 ) and 1O2 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 Fe2+ 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

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 l M , 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 H2O treatment; A400, 400 l M 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.

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 H2O2 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 H2O2 is crucial to trigger the

HR [59] However, increasing the H2O2level by adding

the G + GOX system or by eliminating H2O2 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

H2O2is 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 was significantly suppressed

by high light exposure (350 lmol quantaÆm)2Æs)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 H2O2has been extensively studied

It is prevailingly believed that H2O2can 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, HrpNEa 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 HrpNEa-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 H2O2 and O2 )) 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

However, whether the interference is positive or negative

has to be determined on the basis of given conditions

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 25C with 16 h of light (50 lmol

quantaÆm)2Æs)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)2Æs)1

(cool white light) and the concentrations of ParA1 and

riboflavin were 3 nm and 50 mgÆL)1, respectively

H2O2detection in situ

In situ H2O2 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)1DAB solution, pH 3.8, for 15 min

After incubation at 25C for 12 h in the dark, samples

were transferred in 95% ethanol at 95C 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 H2O2in treated tobacco leaves

In accordance with a ferrous ammonium sulphate⁄ xylenol orange method [72] with some modification, the content of

H2O2 in the treated leaves was measured The relative increase in H2O2in riboflavin-treated zones was calculated

in comparison with the H2O-treated zones in the same leaves An area of 3–5 cm2 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, 4C 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 H2SO4and 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

H2O2) and the loaded weight of plant tissue, the content of

H2O2was 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 parasiticain 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¢-TGAATTCAATAATGTCTAACTTCCGCGCTCT-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 EasySelectTM

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