Plant Physiol. (1984) 76, 125-1300032-0889/84/76/01 25/06/$0 1.00/0

Benzyl Viologen-Mediated Counteraction of Diquat and ParaquatPhytotoxicities'

Received for publication January 20, 1984 and in revised form May 1, 1984

EFRAIM LEWINSOHN AND JONATHAN GRESSEL*2Department ofPlant Genetics, The Weizmann Institute ofScience, Rehovot 76100 Israel

ABSTRACT

There was reason from bacterial and algl systems to expect thatpretreatments with a paraquat analog might confer tolerance against asubsequent paraquat treatment. Thus, a series of compounds were testedfor protective activity agait bipyridinium herbicides. These includedother bipyridinium compounds and derivatives, as well as compoundshaving similar or more positive redox potentials than paraquat andcompounds known to increase or maintain high superoxide dismutaseactivity levels in plants.

Only treatments with benzyl viologen, a benzyl analog of paraquat,protected Spirodela oligorrhiza (Kurz) Hegelm. colonies from othenrise

gi levels of diquat.NADP photoreduction by isolated thylakoids was inhibited by the

same concentrations of paraquat, diquat, and benzyl viologen given sep-arately. Thus, the benzyl viologen-mediated tolerance against the bipyr-idiaium herbicides is probably not due to a direct interction at thethylakoid level.

Superoxide dismutase activity was about 50% higher in broken plastidsof benzyl viologen-treated plants compared to controls, which may partlyexplain the observed tolerance.

The bipyridinium herbicides, paraquat (1,1 '-dimethyl-4,4'-bipyridinium ion, also known as methyl viologen) and diquat(6,7-dihydrodipyrido [1,2-a: 2',1 '-c] pyrazinediium ion) are cat-ions formed by two pyridine rings, each having a quaternaryamine and thus charged 2+. They have strongly negative redoxpotentials: (E' = -0.446 v for paraquat; E.' = -0.35 v for diquat)and are thus capable of acting as electron acceptors of PSI of thephotosynthetic electron transport chain (6). The cations are one-electron photoreduced to form stable bipyridinium radicalswhich are reoxidized by molecular oxygen (9, 13). Superoxideradicals, H202 and hydroxyl radicals may be liberated withcatalytic amounts of herbicide (9). Diquat free radical is pre-sumed to react in the same manner (cf. 13).

Paraquat rapidly inhibits '4CO2 fixation (8). The primarydamage is probably caused by the peroxidation of unsaturatedfatty acids in the membranes (9), with disintegration of theplasmalemma and disorganization of the whole cell. Finally, thecell contents escape, necrosis and ultimate desiccation of theleaves ensue (6).

Chloroplasts detoxify moderate levels of oxygen radicals withSOD,3 which transform superoxide radicals and hydroxyl radicals

' Supported in part by the Fund for Basic Research of the IsraelAcademy of Sciences and Humanities.

2J. G. holds the Gilbert de Botton Chair of Plant Sciences.3Abbreviations: SOD, superoxide dismutase; DMF, N-N'-dimethyl

formamide; FW, fresh weight; Iso, 50% inhibitory dose.

to H202 (29). H202 is much less reactive than the oxygen radicalsand can be further detoxified by the plastids, possibly due to thepresence of the following enzyme system (12):FH 02 > < Asc orbate \,,GIutathione (ox) - , NADPH

Ascorbate Dehydroascorbate Glutathioneperoxidase reductase red uctase

H20 \Dehydroascorbate Glutathione ( red)"J NADP

Paraquat-resistant biotypes have appeared in usually suscepti-ble species, in cases where multiple annual field treatments ofparaquat were given (13). A 300% increase in chloroplast SODactivity was found in resistant Conyza linifolia (= C. bonariensis)and a 56% increase in tolerant Lolium perenne when comparedto susceptible biotypes. This increase was accompanied by anincrease in catalase and peroxidase activities in the leaves, thoughnot in the chloroplasts, of resistant L. perenne (13).There are indications that SOD activity may be induced. SOD

was induced after exposure of Phaseolus vulgaris to ethylenedi-urea (an antioxidant) (16), and after exposure ofEscherichia colito sublethal doses of paraquat (14). A slight increase in paraquattolerance in the green alga Chlorella sorokiniana was correlatedto increased SOD activity obtained after exposure to low levelsof paraquat (22).

It was recently proposed that the induction of SOD mightconfer plant tolerance to paraquat, if high levels could be main-tained (10). Conversely, there might be ways to achieve toleranceto the bipyridinium herbicides without increasing SOD activitylevels. We thus designed a series of experiments to try to findcompounds which could protect against bipyridinium herbicidesbased on many possible criteria. The compound that conferredprotection, benzyl viologen, did not do so at the level of thethylakoid, as indicated below.

MATERIALS AND METHODS

Plant Material. Spirodela oligorrhiza (Kurz) Hegelm. colonieswere axenically cultured on the mineral medium used previously(18). Fronds were cultured in 1-L Fernbach flasks containing250 ml mineral medium with the addition of 1 g/l sucrose, andfortnightly were transferred to fresh medium. The flasks werekept in a growth chamber at 25°C under continuous cool-whitefluorescent light (25-30 uE/m2s-, PAR). Fronds were transferredto mineral medium without sucrose at least 4 d before eachexperiment.

Herbicide and Putative-Protectant Treatments. One or two S.oligorrhiza colonies (5 to 10 mg FW) were placed in a well of a"Costar" cluster dish containing 1.0 ml mineral medium in thepresence or absence of paraquat, diquat, or other chemicals.Light fluence rates, preincubation and exposure times, and con-centrations are specified for each experiment.

In each experiment, up to six 24-well cluster dishes wereemployed. Thus, 36 different treatments consisting of four rep-

125 www.plantphysiol.orgon March 13, 2020 - Published by Downloaded from Copyright © 1984 American Society of Plant Biologists. All rights reserved.

LEWINSOHN AND GRESSEL

licates each were performed within each experiment.

observations, FW, Chl, and photosynthetic capacity ("4C02ation) measurements were then made.

'4CO2 Fixation. Six cluster dishes were placed in a closed

1 acrylic box. '4C02 was released by acidifying 7 MCi NaH'4CO3

(Nuclear Research Center, Beer Sheva, Israel, 20.6

with tartaric acid as outlined in (18). A small fan

rapidly dispersed the '4CO2. The dishes were illuminated

to 85 ME/M2 s, PAR (cool-white fluorescent) for 15 to min.

The air was then flushed from the acrylic box through a

Ba(OH)2 solution in order to trap any unfixed '4CO2.

Irradiance Determination. The action of the bipyridinium

herbicides is positively correlated with light fluence rate

flux measurements were obtained using a Licor LI-

grating quantum radiometer/photometer in the uE/M2.S,

mode.Fresh Weight Determination. S. oligorrhiza colonies

placed onto filter paper until excess water was absorbed,

then weighed.Visual Estimation of Damage. Colonies were evaluated

ing to the following scale: 5, no visual signs of

separation of the fronds from each other, but no signs

rosis; 3, signs of chlorosis; 2, pronounced chlorosis;

almost complete (only traces of green color); 0, full

(white colonies).Chl Determinations and Scintillation Counting. Both

14C contents were estimated from the same samples as described

by Lewinsohn and Gressel (18).Thylakoid Preparations for NADP Photoreduction.

(type 'C') chloroplasts were prepared according to Avron

steps were carried out at0°C. The grinding buffer contained

sucrose, 0.1 NaCI, 50 mM Tricine (pH 7.8),

sodium ascorbate (Sigma). Grinding buffer(15 ml)

approximately 10 g FW S. oligorrhiza colonies. Grinding

performed in a Sorvall Omni-mixer, for 15 s at full

then at setting 7.5 for another 15 s. Preparations

through eight layers of cheesecloth and one of Miracloth,

then centrifuged for 1.5 min at 270g. The supernatant

centrifuged for 10 min at 1500g. This pellet was resuspended

15 ml grinding buffer without ascorbate and was further

fuged for 10min at 2500g. The pellet was then resuspended

grinding buffer without ascorbate and homogenized

glass homogenizer. Chl was determined according to

2).

NADP Photoreduction. Tricine (30 mm, pH 7.8), 30 mm

20Mm fefredoxin (Sigma), 0.5 mm NADP (Sigma), Mg

Chl/ml S. oligorrhiza thylakoids were placed in a 3-ml

metric cuvette. An actinic beam of 3800 ME/in2s,applied. The photomultiplier was protected from the

by 15 mm of a saturated CUS04 solution (24). NADP

was monitored continuously for 60 s at a full scale

0.2 absorbance units simultaneously at 340 to 400 nm.

ments were performed in an Aminco DW-2-dual

spectrophotometer at25°C. A millimolar absorbance

of 6.2 was assumed for NADPH (24).

Superoxide Dismutase Activity. Unwashed Thylakoid

ration. Approximately 0.5 g FW S. oligorrhiza fronds

transferred to a Petri dish containing 30 ml mineral

the presence or absence of mM benzyl viologen. After 24-h

incubation at 40 ME/in2. s, PAR, and at25°C, colonies

washed and weighed. All subsequent steps were performed

ice bath. The tissue was homogenized in 20 ml of

containing 0.4 m sucrose, 10 mm Tricine (pH 7.8), mm

NaCl in an Omni-mixer for 30 s at full speed, then

setting 7.5. Homogenates were filtered through eight

cheesecloth and one layer of Miracloth (11). Then

centrifuged for 5 min at 7700g at 4'C. The supernatant was

discarded and the pellet resuspended in 10 ml 0.1 M phosphatebuffer, pH 7.8.

Superoxide Dismutase Assay. SOD activity was determinedby a method similar to that described by Vaughan et al. (27),with some modifications (19). Superoxide radicals were enzym-

ically generated by the hypoxanthine/xanthine oxidase system.Hypoxanthine (Sigma) was used as it is twice as efficient as

xanthine in liberating superoxide radicals (19). Acetylation ofCyt c (Horse-heart type III, Sigma) was performed according toAzzi et al. (cf. 19). The following were added to a cuvette in thisorder: 250 Mg acetylated Cyt c, 25 Mg hypoxanthine, 10 to 500 ulS. oligorrhiza thylakoids, and 0.1 M phosphate buffer, pH 7.8, toa final volume of 2.0 ml. Then, 0.5 ml of 1 mg/ml xanthineoxidase from buttermilk (Sigma) was added. The xanthine oxi-dase solution was prepared immediately before use. The rate ofreduction of acetylated Cyt c at room temperature (followed at550 nm) was constant for the first 45 s after the addition ofxanthine oxidase. The control rate of acetylated Cyt c reductionwas from 0.03 to 0.04 A550/min. A SOD unit is defined as a

suitable concentration of enzyme causing a 25% decrease in therate of the acetylated Cyt c reduction. The limit of assay was50% inhibition (5).Cytochrome Oxidase Activity. The lack of interference by Cyt

oxidase in the unwashed thylakoid preparation was verified (27).Acetylated Cyt c was reduced with sodium dithionite (cf. 27) andadded to a cuvette at the same concentration as in the determi-nation of SOD. Hypoxanthine and xanthine oxidase solutionswere omitted and equal volumes of 0.1 M phosphate buffer pH7.8, were added instead. Finally, the S. oligorrhiza thylakoidpreparation was added and a decrease in A550 was monitored. Cytoxidase activity was negligible in all SOD extracts tested withacetylated Cyt.

Protein Determination. Protein was estimated by the methodof Bradford (4), with Coomassie brilliant blue. BSA solutionswere used as standards.

RESULTS

Many compounds were tested to find an agent which wouldprotect the plants from paraquat injury. The first group ofinactive compounds included 0.1 mM 2,2'-dipyridyl (Sigma),which structurally resembles diquat; 0.1 to10 gM rose bengal,which has similar physiological effects as paraquat (21);0.1 to1

Mm piperonyl-butoxide (4,5-methylenedioxy-2-propyl- benzyldi-ethyleneglycol-butylether, Fluka), which is known to preventozone injury in Nicotiana tabacum (15); 0.1 to 10 mm sulfiteions (used as an alternative toSO2), which increased SOD levelsin Populus euramericana leaves (cf.10);I to100gM ethylene-diurea, which prevents ozone injury in P. vulgaris and increasesSOD activity levels (16); and pretreatments with a sublethal levelof paraquat, which increased SOD levels in E. coli (14) andC.sorokiniana (22). We used 0.01Mm paraquat; higher doses were

already lethal.Next we tested a series of compounds using the following

criteria.(a) Compounds similar to paraquat in structure, or substituted

analogues:1,1 '-di(p.benzoic acid)-2,2'-bipyridinium dichloride;I ,methyl,1 '-(5-pentanoic acid)-4,4'-bipyridinium dichloride;1,1'-di(2-propene)-4,4'-bipyridinium dibromide;1-1'-di(2-hy-droxyethane)-4,4'-bipyridinium dichloride.

(b) Compounds having similar structural or physical charac-teristics to diquat, as size or hydrophobicity: triquat(1,1''-1,3-propylene]-2,2'-bipyridinium dibromide), tetraquat (1,1'-[1,4-butylene]-2,2'-bipyridinium dibromide).

(c) Compounds having a redox potential close to paraquat, or

which can act as one-electron acceptors: neutral red (3-amino7-dimethylamino-2-methylphenazine-hydrochloride), E.' =

-0.325 v; benzyl viologen (1,1I'-dibenzyl-4,4'dipyridinium di-

126 Plant Physiol. Vol. 76, 1984

www.plantphysiol.orgon March 13, 2020 - Published by Downloaded from Copyright © 1984 American Society of Plant Biologists. All rights reserved.

PROTECTION AGAINST BIPYRIDINIUM HERBICIDES

chloride) E.' = -0.36 v.(d) Compounds having a more positive redox potential than

the bipyridinium herbicides and thus less toxic: N42-propenyl)acridinium bromide, E.' = +0.24 v; 2,5-dimethylphenaziniumfluorosulfonate, E.' = +0.1 v; 1-methoxy, 5-methylphenaziniumfluorosulfonate, E.' = +0.11 v; 4-methyl-i ,4-benzodiaziniumfluorosulfonate, E.' = -0.16 v.

(e) Other compounds which contain either tertiary or quater-nary pyridine rings which might resemble paraquat: 4,4'-dithio-dipyridyl, 2,2'-dithiodipyridyl, acridine, 5-aminoacridine, N-n.butylacridinium bromide.

(f) A compound which reduces ozone injury: piperonal (3,4-(methylenedioxy)-benzaldehyde) (15).The Chl content and 4CO2 fixation capacity of the colonies

were measured after a 24-h preincubation with the putativeprotectant and a 24-h exposure to paraquat at 80 to 85 ME/M2.s, PAR.Most treatments did not significantly alter the paraquat effect

on Chl levels: 1 and 10 AM tetraquat; 0.32 MM 1,methyl,1'-5-pentanoic acid)-4,4'-bipyridinium dichloride; 1 and 10 Mm 1, 1'-di(p.benzoic acid)-2,2'-bipyridinium dichloride; 1 and 10 yM N-(2-propenyl) acridinium bromide; 1 and 10 Mm N-n .butylacridi-nium bromide; 1 and 10 gM 2,5-dimethylphenazinium fluoro-sulfonate; 1 and 10 MM 1-methoxy, 5-methylphenazinium fluo-rosulfonate; 1 and 10 Mm piperonal; 1 and 10 Mm neutral red.Other compounds were too toxic to the plants under these

conditions even in the absence of paraquat: 1 and 10 MM 1, l'-di(2-propene)-4,4'-bipyridinium dibromide; 1 and 5 MM 1-1'-di(2-hydroxyethane)-4,4'-bipyridinium dichloride. Other treat-ments protected more than 50% of the Chi from being bleachedby paraquat: I gM acridine, 1 FM 5-aminoacridine, IlOM neutralred. The best protection against Chl breakdown was obtained

0.1 0.3 T 0o C

I=M paraquat or diquat

40

0

0

0

cJ

0

IC-

FIG. 1. Effects of benzyl viologen in continuous co-treatment withparaquat or diquat on Chl bleaching and photosynthesis. Colonies wereincubated for 24 h with (A, B) paraquat only (@ *); paraquat togetherwith I FMbenzyl viologen (A---A); paraquat with 1OMm benzyl viologen(-... 0). C, D depict diquat-treated colonies preincubated in the pres-ence of 10m benzyl viologen (0---0) or in its absence (@ ). Lightintensity was 80 to 85 JE/m2 *s, PAR. Colonies were exposed for 30 minto "CO2. Control values correspond to (A) 0.87 mg Chl/g FW, (B) 1373cpm/mg FW, (C) 0.98 mg Chl/g FW, and (D) 182 cpm/mg FW.Radioactivity is expressed as DMF-insoluble cpm fixed per mg FW. Barsrepresent SE of 4 replicates.

with benzyl viologen. Essentially all the Chl was still presentfollowing a benzyl viologen pretreatment at 10 Mm and a 24-hexposure to I AM paraquat. The results of these experiments arediscussed in detail in Lewinsohn (17).None of the chemicals including benzyl viologen, effected

more than a 20% protection ofthe "CO2 fixation capacity oftheplants. Even when Chl bleaching was prevented, "CO2 fixationwas markedly suppressed. This indicated that the plants wereindeed affected by paraquat, although Chl breakdown was slower.1-1l'-Di(2-hydroxyethane)-4,4'-bipyridinium dichloride and1,1 '-di(2-propene)-4,4'-bipyridinium dibromide, were thus toxicby the criterion of 14CO2 fixation. Some treatments (i.e. I gMacridine, 1 AM 5-aminoacridine, 1 AM 4,4'-dithiodipyridyl) main-tained about 15% of the "CO2 fixation at 0.32 Mm paraquat,which normally totally suppresses "C02 fixation. When theparaquat level was increased to 1 ,M, the level of "CO2 fixationdropped in all but one treatment (i.e. 10 Mm benzyl viologen).The 14C02 fixation capacity remained at 17% ofthe control valuewith benzyl viologen.The ideal protectant situation of neither affecting Chl content

nor 14CO2 fixation capacity, even after exposure to relatively highparaquat levels was not obtained. However, pretreatment with10 AM benzyl viologen appeared to provide better protectionthan all other compounds by both criteria.Benzyl Viologen Toxicity. Continuously supplied 10 M benzyl

viologen did not significantly affect "C02 fixation in S. oligor-rhiza colonies for 24 h at 80 ME/M2. S, PAR. After a 48-hincubation, "CO2 fixation decreased by 80% and chlorotic re-gions appeared in the colonies. Benzyl viologen caused fullchlorosis in restricted zones within a frond while other parts ofthe same frond seemed to be unaffected. In contrast, paraquat-induced chlorosis affected the entire frond (data not shown).

Interaction between Benzyl Viologen, Paraquat, and Diquat.Continuous Co-Treatments. As benzyl viologen showed somepromise as a protective agent, we further assessed its interactionswith paraquat and diquat (Fig. 1). Benzyl viologen (10 MM)protected Chl when the plants were treated simultaneously withparaquat (Fig. 1A) but no significant protection of "CO2 fixationwas detected (Fig. 1B). A better protection was obtained againstdiquat damage (Fig. 1, C and D). "CO2 fixation was significantlyhigher in benzyl viologen treated versus untreated colonies ex-posed to 0.1 Mm diquat (Fig. 1D). Additionally, diquat alone ismore toxic than paraquat to most monocots, including S. oligor-rhiza (3).

Plants exposed continuously to these herbicide levels even-tually died even in the presence of benzyl viologen. Benzylviologen only delayed Chl breakdown and ultimate death, butdid not eliminate the toxic effects of paraquat or diquat. Fur-thermore, long exposures to the fluence rates used (80-85 ME/m s, PAR) were toxic, even to untreated colonies. Chloroticlesions and decreases in "C02 fixation were occasionally detectedin colonies which had been exposed at these fluence rates for 24h to 10 M benzyl viologen without the herbicides.

Pulse Treatments with Herbicide andPutative Protectants. Therate of disappearance of the bipyridinium herbicides is muchhigher in the field or in water than under our experimentalconditions. The presence of high sunlight, which promotes UV-mediated degradation, water movement, and conjugation to soilor other colloids cause the herbicide levels to rapidly decline (6).These factors are absent in a 10-mm diameter well containing 1ml medium in quasi-sterile conditions illuminated by relativelylow fluence cool-white fluorescent light. Thus, the next approachwas to mimic more 'natural' conditions by removing the herbi-cide from the medium after a fixed period. Fluence-rates werelowered to a level of 40 ME/mn2- s, PAR, which is closer to thenormal environmental levels of our S. oligorrhiza strain.

After a 24-h preincubation with or without 10 Mm benzyl

c

00

0

--

3

Un0

0'

EVI

CL

0

0

U

0C

127

www.plantphysiol.orgon March 13, 2020 - Published by Downloaded from Copyright © 1984 American Society of Plant Biologists. All rights reserved.

LEWINSOHN AND GRESSEL128

I _i .

o_

10

_ c.

5

_-,c

0-E._ .4

Plant Physiol. Vol. 76, 1984

11 11 1- I II

y volokgen

1000 C.)

10 30

it

1500

K)O u

1 000A"""1]~~~0

D i q u at (1LM)

FIG. 2. Benzyl viologen protectioti against a short treatment with paraquat (A, B, C) or diquat (D, E, F). Colonies were preincubated for 24 h inmineral medium in the presence or absence of 10 Mm benzyl viologen. Paraquat or diquat treatment was for 2 h and then colonies were transferredto fresh medium. The transfer was repeated daily for 3 d. Chi content and '4CO2 fixation were determined on the 4th d. Light intensity was 40 to45 uE/m2.s, PAR. Each value represents a mean of4 replicates + SE.

CP303 benzyl vialogn+diqwot-

E

r 200-

z100- diquat abne

0.0 Ql Q3 I 3 rDiquot (MM)

FIG. 3. Long-term protection by benzyl viologen against diquat tox-icity. Colonies were subjected to the benzyl viologen and diquat treatmentdescribed in Figure 2, and transferred to 30 ml mineral medium. FWincrease was determined after 10 d of culture under 40 to 45 M&E/m2-s,PAR, cool-white fluorescent light.

viologen, colonies were exposed for 2 h to paraquat or diquat,then transferred to fresh medium and re-transferred to freshmedium daily for 3 d. Chl content and "CO2 fixation were thendetermined. The results ofthese experiments are shown in Figure2.

Benzyl viologen treatments clearly protected against paraquat

(Fig. 2, A, B, C). Protection was apparent by visual evaluation(Fig. 2A) and Chl content (Fig. 2B) at 3.2 to 32 #M paraquat.Protection was also apparent at 1.0 Mm paraquat when photosyn-thesis was measured (Fig. 2C). The Iso value for Chl bleachingand for photosynthesis increased about 4-fold after a 10 Mmbenzyl viologen pretreatment (Fig. 2, B and C). Injury was notdetected at 1 ,gM paraquat by the criterion of Chl breakdown inunprotected plants, as Chl breakdown occurs after the abolitionof photosynthesis (8).The effect ofa 24-h preincubation with 10 Mm benzyl viologen

on diquat toxicity was even more pronounced than protectionagainst paraquat (Fig. 2, D, E, F). Benzyl viologen-treated plantswere relatively unaffected by 1.0 to 3.2 Mm diquat as assayed byvisual evaluation (Fig. 2D), Chl bleaching (Fig. 2E), and photo-synthesis (Fig. 2F). Diquat (3.2 gM) caused almost completebleaching and death of colonies which were not treated withbenzyl viologen within 3 d of exposure. There was almost nogrowth inhibition, Chl a or b loss (17), nor a significant drop inphotosynthesis in benzyl viologen-treated plants. The I5o for Chlbleaching increased more than 8-fold (Fig. 2E) and the Iso forphotosynthesis inced about 4-fold after benzyl viologenpreincubation (Fig. 2F).The benzyl viologen and diquat treated plants were further

cultured to ascertain if there was any late exprssion of damage(Fig. 3). Apparently normal colonies were obtained, as judgedboth by visual appearance, Chl content, and photosyntheticcapacity, after benzyl viologen-treated plants were exposed tootherwise lethal levels of diquat. Indeed, the I50 for diquat in-creased about 14-fold after the benzyl viologen pretreatment.

Ptraquat (JAM)

www.plantphysiol.orgon March 13, 2020 - Published by Downloaded from Copyright © 1984 American Society of Plant Biologists. All rights reserved.

PROTECTION AGAINST BIPYRIDINIUM HERBICIDES

IIg-EI I v *

benzyl-viologenalone

I iquat + benzyI - viologen

IFi_

*..... diquat aloneni I I I

I I1 4- mineral medium alone

'iparoquat benzyl-viologen

O4- paraquat alone

I I I I

100

80 XO _0._

60 -0

01

O

40 2L.-

20 E0

CU

FIG. 4. The effect of duration of benzyl viologenpretreatment on diquat and paraquat toxicities. Colo-nies were incubated in mineral medium containing 10pM benzyl viologen at 25C under 40 to 45 ,uE/m2 -s,PAR for the duration specified. Then, 3 #M paraquator diquat was added. '4CM2 fixation of the samples wasmeasured 2 h later. Bars represent SE of four replicates.The control value was 1010+ 77 cpm fixed per jg Chl.

V 0. . .

0 5 10 15 20 0 5 10 15; 20 25Benzyl viologen (I0LM) preicubation time (hours)

0.0

28-

EI= 6-a

4-4z0

0 ~~~~~~0PARAQUJAT-

0 2 DIQUAT

E 0BENZYL- VIOLOGENt 0.3 MLM BENZYL-VIOLOGEN

+ I3pM DIQUATO .L.1 1ll * 11 ' " '1

0 0.1 1 10Viologen concentration (CLM)

FIG. 5. The lack of benzyl viologen protection at the level of isolatedthylakoids. NADP reduction rates were measured in illuminated S.oligorrhiza thylakoids. The arrow indicates the value obtained when 0.3piM diquat and 0.3 FM benzyl viologen were added simultaneously, andconsidered additively.Table I. SOD Activity in Crude Thylakoid Pellets ofBenzyl Viologen-

Treated S. oligorrhiza ColoniesReduction of acetylated Cyt c by enzymically generated superoxide

radicals was monitored as described under "Materials and Methods". SEare in parentheses. The differences between treatments are significant atthe 0.05 level by the t test.

SOD Unitsper mg protein per mgFW

Control 129 (±3.0) 57.8 (±1.4)+10 uM benzyl viologen 210 (±17.2) 87.8 (±5.6)Increase 65% 52%

Plants were subcultured for three more lO-d periods withoutshowing any sign of residual diquat or benzyl viologen damage.We assayed for the minimal time needed to confer the plants

with tolerance toward diquat and paraquat. Benzyl viologen (10Mm) conferred protection both when applied simultaneously withthe herbicides (at 3Mm) or when it was administered up to 24 hbefore the toxic treatment, based on 14CO2 fixation capacity (Fig.4).NADP Reduction. It was important to ascertain whether the

benzyl viologen-induced protection could also be detected inisolated thylakoids. Benzyl viologen might serve as an artificialintermediate in the photosynthetic electron transport chain (25).If so, benzyl viologen would not inhibit photosynthetic NADPreduction. However, paraquat, diquat, and benzyl viologen allinhibited the rate of light-induced NADP reduction to approxi-mately the same extent (Fig. 5). Benzyl viologen appeared to be

as toxic as diquat and paraquat, and its toxicity was additive tothat of diquat. This is in contrast to the whole plant system,where benzyl viologen was much less toxic than the bipyridiniumherbicides.

Superoxide Dismutase Activity after Benzyl Viologen Treat-ment. We found a significant increase in thylakoid SOD activityafter the colonies were exposed for 24 h to IO0Mm benzyl viologenat 40 to 45 1E/m2 s, PAR, (Table I). The increase was approxi-mately 60% when expressed on a protein basis and about 50%on a FW basis.These results indicate that SOD activity is enhanced by benzyl

viologen pretreatment of S. oligorrhiza colonies, which maycontribute to the observed tolerance.

DISCUSSION

SOD is thought to play an important role in detoxifying theeffects of bipyridinium herbicides. Nevertheless, there are otherways to decrease paraquat and diquat toxicities which are clearlynot involved with enhanced SOD activity. PSII inhibitors, likemonuron, inhibit or delay the toxicity of paraquat (8) as theblockage precludes availability of electrons for free-radical for-mation by PSI. This is quite different from the effects of benzylviologen, as monuron-treated plants cannot fix CO2. In contrast,benzyl viologen-protected plants did not totally lose their pho-tosynthetic capacity (Figs. 1, 2, 4).

Amphiphilic viologens could bind to the thylakoid membranesby hydrophobic interactions and serve as artificial electron car-riers (7). Benzyl viologen may act as an intermediate electroncarrier, bypassing part of the normal photosynthetic electrontransport chain (25). Benzyl viologen can form relatively stablebipyridyl radicals by a one-electron reduction of the cation andthus should be herbicidal. Benzyl viologen was about 100-timesless inhibitory to '4C02 fixation than paraquat on S. oligorrhizacolonies in our conditions, somewhat less by visual evaluations(23). Plants seemed normal and healthy after a 24-h exposure to10 gM benzyl viologen at 40 to 45 pE/m2 _s, PAR, if they weresubsequently transferred to fresh medium. The plants even tol-erated doses of the herbicides that would have been lethal, if notfor the benzyl viologen treatment (Fig. 2). The benzyl viologenradical is more easily reduced than the respective paraquat ordiquat radicals (28). A second electron from the photosyntheticelectron transport chain might further reduce the radical to forma stable neutral molecule. Benzyl viologen could thus lose itstoxicity, as reduction of02 to form 02- would be very unlikely.

After benzyl viologen was applied to broken chloroplasts keptin aerobic conditions, superoxide radical was formed and therewas a breakdown of lipid constituents (1, 26). There was nomarked difference between the levels of 02- liberated by similarconcentrations of paraquat, diquat, triquat, or benzyl viologen,as measured by SOD-dependent reduction of Cyt c (1). Further-more, benzyl viologen itselfenhanced the rates of02- productionto a similar extent. Thus, at the level of the thylakoid, benzyl

125

m 100

z 0--a 75bO_0 0

:L %o 50a O-

E 250.

129

lr

I

www.plantphysiol.orgon March 13, 2020 - Published by Downloaded from Copyright © 1984 American Society of Plant Biologists. All rights reserved.

LEWINSOHN AND GRESSEL

viologen was rather toxic (Fig. 5), whereas at the whole plantlevel, benzyl viologen was rather inoffensive.

It is necessary to attempt to explain the discrepancy betweenthe benzyl viologen-mediated protection at the whole plant leveland its toxicity at the thylakoid level. In addition, it may bedifficult to conceive that a 50% increase in SOD activity issufficient to confer protection, especially as the protection wasimmediate (Fig. 4), and it takes time to increase enzyme activitylevels. The following explanations seem feasible:

(a) Benzyl viologen is rapidly metabolized to a protectivecomponent. Benzyl viologen could thus be toxic to brokenplastids if they did not contain the enzymes involved.

(b) Benzyl viologen binds to a site on the plasmalemma or theplastid envelope for which it has high affinity and preventspenetration of paraquat, diquat, and benzyl viologen itself. Am-phiphilic viologens bind strongly, affecting thylakoid structureand properties (7), supporting this possibility.

(c) Benzyl viologen interacts with a stromal or thylakoidalcomponent and together they interact to accept electrons fromparaquat and recycle them back into the electron transport chain.This hypothetical compound was lost during thylakoid prepara-tion. An Azotobacter vinelandii enzyme (20), catalyzed thisNADP reduction, with benzyl viologen but not paraquat as theelectron donor. However, such activity has not been detected inchloroplasts.

(d) It is, thus, possible that the herbicide-tolerance observedwas indeed caused by an increase in SOD activity, especially ifother factors involved in detoxificationof 02 (i.e. peroxidases,dehydroascorbate reductases, glutathione reductases, reducedglutathione, and ascorbate) were present at sufficient levels andjust SOD activity was limiting. Similar increases in SOD activitywere reported in paraquat-tolerant L. perenne biotypes comparedto susceptible ones (13).

Thus, as predicted (10), treatment with a herbicidally inactivebipyridinium derivative conferred tolerance toward the bipyri-dinium herbicides, possibly by inducing higher levels of SOD.However, other mechanisms which are not necessarily involvedwith SOD induction may be responsible for the benzyl viologen-induced tolerance in S. oligorrhiza.

Acknowledgmenis-We would like to express our sincere thanks to Dr. YosephaShahak, The Weizmann Institute of Science, Rehovot, Israel, for the interestingand useful discussions we had and for her kind instruction in measuring NADPreduction in broken chloroplasts. The viologens (except for benzyl viologen) werekindly provided by Dr. Stuart Ridley, ICI, Jealotts-Hill, UK.Ethylenediurea was akind gift of E.L. Jenner, E. I. Du Pont de Nemours.

LTERATURE CITED

1. ASADA K, KKLso, K YOSHIKAWA 1974 Univalent reduction of molecularoxygen by spinach chloroplasts on illumination. J Biol Chem 249: 2175-2181

2. AVRON M 1960 Photophosphorylation by swiss-chard chloroplasts. BiochimBiophys Acta 40: 257-272

3. BLACKBURN RD, LW WELDON 1965 The sensitivity ofduckweeds(Lemnaceae)and Azolla to diquat andparaquat. Weeds 13: 147-149

4. BRADFORD MM 1976 Arapid and sensitive method for the quantitaton ofmicrogram quantities of protein utilizing the principle of protein-dye bind-ing.Anal Biochem 72: 248-254

5. BUCHANAN AG, H LEES 1976 The effects of pH and temperature on the assayof superoxide dismutase. Can J Microbiol 22: 1643-1646

6. CALDERBANK A, P SLADE 1976 Diquat and paraquat. In PC Kearney, DDKaufman, eds, Herbicides Chemistry, Degradation and Mode of Actiion, Ed2. Marcel Dekker Inc, New York, pp 501-540

7. CUENDET P, M GRATZEL 1983 Interactions of hydrophilic and amphiphilicviologens with the thylakoid surface. Photobiochem Photobiophys 5: 335-345

8. DODGE AD 1982 The role of light and oxygen in the action of photosyntheticinhibitor herbicides. In E Moreland, JB St John, FD Hess, eds, BiochemicalResponses Induced by Herbicides. ACS Symposium series No. 181, Ameri-can Chemical Society, Washington, DC, pp 57-77

9. FARRINGTON JA, M EBERT, EJ LAND, K FLETCHER 1973 Bipyridylium quater-nary salts and related compounds. V. Pulse radiolysis studies of the reactionof paraquat with oxygen. Implications for the mode of action of bipyridylherbicides. Biochim Biophys Acta 314: 372-381

10. GRESSEL J, G EZRA, SM JAIN 1982 Genetic and chemical manipulation tocrops to confer tolerance to chemicals. InJS McLaren, ed, Chemical Manip-ulation of Crop Growth and Development. Butterworths, London, pp 79-91

11. GRESSEL J, Y REGEV, S MALKIN, Y KLEIFELD 1983 Characterization of an s-triazine-resistant biotype of Brachypodium distachyon. Weed Sci 31: 450-456

12. HALLIWELL B, CH FOYER, SA CHARLES 1981 The fate of hydrogen peroxidein illuminated chloroplasts. In G Akoyunoglou, ed, Photosynthesis II Elec-tron Transport and Photophosphorylation. Balaban Intl Sci Ser, Philadel-phia, PA, pp 279-283

13. HARVEY BMR, DB HARPER 1982 Tolerance to bipyridilium herbicides. In HMLe Baron, J Gressel, eds, Herbicide Resistance in Plants. J Wiley, New York,pp 215-233

14. HASSAN HM, I FRIDOVICH 1977 Regulation of the synthesis of superoxidedismutase in E. coli. Induction by methyl-viologen. J Biol Chem 252: 7667-7672

15. KoiwAl A, H KITANO, M FUKUDA, T KISAKI 1974 Methylenedioxyphenyl andits related compounds as protectants against ozone injury to plants. AgricBiol Chem 38: 301-307

16. LEE EH, JH BENNET 1982 Superoxide dismutase: A possible protective enzymeagainst ozone injury in snap beans (Phaseolus vulgaris). Plant Physiol 69:1444-1449

17. LEWINSOHN E 1983 A decrease in bipyridinium toxicity in benzyl viologentreated Spirodela oligorrhiza colonies. MSc Thesis. The Weizmann Instituteof Science, Rehovot, Israel (Available on request from the author)

18. LEWINSOHN E, J GRESSEL 1983 The determination of chlorophylls a and btogether with'4C02 fixation in the same plant tissue samples. Anal Biochem135: 438-442

19. MCCORD JM, JD CRAPO, I FRIDOVICH 1977 SOD assays: A review of meth-odology. In AM Michelson, JM McCord, I Fridovich, eds, Superoxides andSuperoxide Dismutases. Academic Press, London, pp1 1-17

20. NAGAI Y, RF ELLEWAY,DJD NICHOLAS 1968 Some properties of an NADH-benzyl viologen reductase from Azotobacter vinelandii. Biochim BiophysActa 153: 766-776

21. PERCIVAL MP, ADDODGE 1983 Photodynamic effects of rose-bengal onsenescent flax cotyledons. J Exp Bot 34: 47-54

22. RABINOWITCH HD, DA CLARE, JD CRAPO, I FRIDOVICH 1982 Positive corre-lation between superoxide dismutase and resistance to paraquat toxocity inthe greenalga Chlorella sorokiniana. Arch Biochem Biophys 225: 640-648

23. Ross JH, LOLIM, RI KRIEGER 1979 Herbicidal potency of 1,1'-alkyl-4,4'-bipyridylium salts as a function of their physicochemical constants in duck-weed. Drug Chem Toxicol 2:193-205

24. SHEFFER M, MAVRON 1982 An unusual sensitivity to salt of ferredoxin-dependent photoreactions in Dunaliella. Plant Sci Lett 25: 241-246

25. SHIN M, DI ARNON 1965 Enzymic mechanisms of pyridine nucleotide reduc-tion inchloroplasts. J Biol Chem 240:1405-1411

26. TAKAHAMA U, M NISHIMURA 1976 Effects of electron donors and acceptors,electron transfer mediators, and superoxide dismutase on lipid peroxidationin illuminated chloroplasts. Plant Cell Physiol 17: 111-118

27. VAUGHAN D, PC DEKOK, BG ORD 1982 The nature and localization ofsuperoxide dismutase in fronds of Lemna gibba L. and the effect of copperand zinc deficiency on its activity. Physiol Plant 54: 253-257

28. VOLKE J 1968 Relation between herbicidal activity and electrochemical prop-erties of quatemary bipyridinium salts. Collect Czech Chem Commun 33:3044-3048 (Chem Abstr 69: 85624d)

29. WINTERBOURN CC 1981 Production of hydroxyl radicals from paraquat radi-cals and H202. FEBSLett 128: 339-342

130 Plant Physiol. Vol. 76, 1984

www.plantphysiol.orgon March 13, 2020 - Published by Downloaded from Copyright © 1984 American Society of Plant Biologists. All rights reserved.