Cl - Plant Physiology · Since photoreduction differs from photosynthesis in that hydrogen gas replaces water as the ultimate electron donor, it seems reasonable to suppose that DCMUinterferes

INHIBITORS OF THE HILL REACTION , 2

NORMAN E. GOODPESTICIDE RESEARCH INSTITIIIE, UNIVERSITY SUB POST OFFICE, LONDON, CAN .\DA

Many commercially important herbicides seemto kill plants by inhibiting photosynthesis. Wesselsand van der Veen have shown that leaves or parts ofleaves treated with phenylureas irreversibly lose allability to assimilate CO.,. They also found thatchloroplasts treated with phenylureas are unable toreduce oxidants such as 2,6-dichlorophenolindophenolin the light (20). A wide variety of phenylcarba-mates similarly inhibit photochemical reductions bychloroplast (the Hill reaction) although somewhathigher concentrations of these carbamates are required(20, 11). Recently a number of acylanilides havebeen intro(luced as herbicides. These, too, interferewith the Hill reaction (12). The phenylcarbamates,phenylureas, and acylanilides are, of course, closelyrelate(d chemically: all are anilides of carboxyacids.However, the very (lissimilar herbicide Simazine[2-chloro-4,6-bis(ethylamino)-s-triazine] also inhibitsthe Hill reaction and does so at low concentrations(13).

It has been possible to identify more specificallythe process wvhich is inhibited in chloroplasts byphenylureas. Bishop discovered that 3-(3,4-dichloro-phenyl)-1,1-dimethylurea (DCMU) inhibited photo-synthesis but not photoreduction in the alga Scene-desmus (3). Since photoreduction differs fromphotosynthesis in that hydrogen gas replaces water asthe ultimate electron donor, it seems reasonable tosuppose that DCMU interferes with the mechanism ofwater oxidation-that is. with the process leading tothe prodluction of molecular oxygen. This hypothesishas been supported by experiments with chloroplastsfrom higher plants. Thus, Jagendorf showed thatATP formation (an(d therefore probably electronflux) was inhibited in chloroplasts by 3-(4-chloro-phenyl) -1,1-dimethylurea (CMNIU) when flavin mono-nucleotide (FMN) was the electron carrier but notwhen N-methylphenazonium ion (PM S) was thecarrier (8). Later Krall, Good, and Mayne demon-strated that the FMN catalysed process involved theproduction and reutilization of molecular oxygenwvhereas the PMS catalysedl process did not (10).Apparently reduced PMS can replace water as elec-tron donor in a reaction analogous to photoreduction.In accordance with Bishop's hypothesis, the process

' Received April 10, 1961.2 Contribution No. 194, Pesticide Research Institute,

Research Branch, Canada Agriculture, University SubPost Office, London, Ontario.

which is independent of the production of molecularoxygen is also resistant to CMAU inhibition.

In the work described below the chloroplast-inhibiting herbicides and a large number of relatedsubstances were investigated. This report deals withfour aspects of the problem:

I. More than 200 acylanilides, thioacylanilides,acylamides, ureas, and thioureas were prepared andidentified. About half of these were not describedin the readily accessible literature and probably manyare new.

II. The work of Wessels and van der Veen (20)and of Mforeland et al. (11, 12) on the relationship ofinhibitor structure to inhibitor potency has been ex-tended. Inhibition was routinely measured as a de-crease in the ability of isolated chloroplasts to re(luceferricyanide in the light.

lII. The Jagendorf criterion was use(d to (leter-mine the site of inhibition. Inhibition of the photo-chemical formation of ATP with FMN as catalystwas compared with the inhibition of ATP formationwith PMS as catalyst. This test was applied torepresentative acylanilides, alkylureas, and triazinesas well as to the previously examined phenylureas.

lV. An attempt was made to correlate the bond-ing potentials of the imino hydrogens of the variousinhibitors with inhibitor structure and biological ac-tivitv. Wessels and van der Veen (20) have sug-gestecl that the carbonyl oxygen of the cyclopentanonering of the chlorophyll molecule may form hydlrogenbonds witlh the imino hydrogen of phenylcarbamiiatesand phenylureas, thus disrupting the photochenmicalapparatus of the chloroplast. For reasons to be (lis-cussed, we do not consider that this particular bondcould account for the observe(d inhibitions. Never-theless some of the most striking properties of anilidlesand other amides (high melting points, low vaporpressures. prefercnce for polar solvents, etc.) are(lirectly related to intermolecular hydlrogen bonding.Therefore it seems quite probable that these inhibitorsbecome attached by hydrogen bonds to the as yetuni(lentified active site. Since the formation of ahydrogen bond results in a replacement of one ofthe characteristic N-H infra-red absorption bands byan N-H ... 0 band at a longer wave length, thebonding of the imino hydrogen of representative anil-ides and triazines to various carbonyl groups was ob-served by infra-red spectroscopy.

788

Copyright (c) 2020 American Society of Plant Biologists. All rights reserved.

GOOD-HILL REACTION INHIBITORS

MATERIALS & METHODSI. SYNTHESES OF INHIBITORS. Acylanilides were

prepared by treating the appropriate free anilineeither with acid anhydrides (acetic, propionic, butyric,& isobutyric) or with acid chlorides. The anhydridesor chlorides of the acids were added slowly to a ben-zene solution of the aniline, and the resulting suspen-sion wvas boile(d for a few minutes. Those anilideswlhiclh were insoluble in benzene were filtered off.dlissolved in alcohol, treated with Norite, and re-crystallized from alcohol or, more often, from alco-hol and water. Anilides which were soluble in ben-zene were precipitated by adding n-hexane. To bringout the anilidles of the higher fatty acids large amountsof hexane and cooling to about -15 C were necessary.The acid chlorides were prepared by heating the acidsunder reflux with an excess of thionyl chloride andwere used without purification after the bulk of theunreacte(d thionyl chloride had been removed by dis-tillation. Isobutyrylamides were prepared in a simi-lar manner except that no heating of the alkylamine-acid anhydridle mixture was required. Fornmanili(leswvere prepare(d by heating the anilides undler refluxin a large excess of formic acid.

4-Chloro- and 3,4-dichloroanilides of thiopropionicacid were prepared by treating the correspondingoxygen anilides with P.S5 (9). The solubility of thethioanilides in dilute KOH was used to facilitate theirisolation. They were recrystallized from a solutionof benzene and n-hexane.

Asymmetrical ureas were prepared either by theaction of dimethylcarbamyl chloride on the appropriateamine (3,5-dichloroaniline, 3-chloroaniline, octyl-amine, benzylamine, cyclohexylamine) at room tem-perature or by the action of alkylamines on the iso-cyanates derived from 3,4-dichloroaniline and 2-chloroaniline. Note that unless heat is avoided in thereaction involving dimethylcarbamyl chlori(le. a re-arrangement takes place which results in the forma-tion of symmetrical ureas. The reaction of dimethyl-carbamyl chloride with anilines is very slow at roomtemperature, requiring one to several weeks: with 2-chloroaniline it is prohibitively slow and even at roomtemnerature yields predominantly the symmetrical1 .3-bis (2-chlorophenyl) urea.

Asymmetrical thioureas were prepared by theaction of alkvlamines on 3.4-dichlorophenyl isothio-cyanate.

Symmetrical dliphenylureas and diphenylthioureas(most of which were too insoluble to test) were pre-pared by prolonged heating of the appropriate anilinewitlh dimethylcarbamyl chloride (see above) or di-methylthiocarbamyl chloride.

Alkylamines, anilines, and acids used in this in-vestigation were obtainedl from commercial soutrcesexcept 3,4-dichlorophenylacetic acid which was pre-pared by the Arndt-Eistert synthesis (1) from 3,4-dichlorobenzoyl chloride. The phenyl isocyanatesand phenyl isothiocyanates were prepared by slowlyadding the corresponding aniline to solvents contain-ing an excess of phosgene or thiophosgene (2).

The melting point of each of the potential inhibi-tors was determined on a melting point block. Thisblock had been calibrated with standard pure sub-stances having well-recognized melting points. Melt-ing points so obtained were identified with those inthe literature whenever possible. The only majordifference involved 1-benzyl-3,3-dimethylurea (found76770, literature 166°); the substance described inthe literature (19) is probably 1,3-dibenzylurea, mp1680. The melting point of another of our products,N-isobutyryl-2,6-dimethylaniline, could not be estab-lished. This substance softened and sublimed over arange of more than 1000. Some of the substitutedureas also had abnormal (double?) melting points.In these instances we have followed the literature inreporting only the higher temperature. (table I).

When no description of the substance could befound (or if there was reason to question the litera-ture) the identity of the substance was confirmed byelemental analysis. (table I).

II. DETERMINATION OF RELATIVE POTENCIES OFINHIBITORS. Chloroplasts from young pea plants(Pisum sativum L.) were isolated in the followingmanner: Freshly picked leaves were ground in achilled mortar with an ice-cold buffer consisting of135 g sucrose, 3.0 g tris (hydroxymethyl) aminometh-ane, and 1.0 g sodium chloride per liter, adjusted topH 7.95 with 1 N sulfuric acid. The homogenate wasfiltered through glass wool into ice-encased centrifugetubes. After centrifugation for 3 minutes at about4,000 times gravity, the supernatant was discardedand the chloroplasts were resuspended in more coldbuffer. The suspension was transferred to other ice-encased centrifuge tubes and was again centrifugedfor 2 minutes at 4,000 times gravity. The chloro-plasts were then suspended in a minimum volume ofbuffer and the suspension was filtered through glasswvool again to remove clumps. The final dense sus-pension was stored briefly in a flask which was deep-lv immersed in chopped ice and water.

The reaction mixture (final volume 2.0 ml) con-sisted of the above mentioned buffer, methylaminehydrochloride (24 ,umoles), potassium ferricvaiide(1.0 ,umole), and an amount of the chloroplast sus-pension containing about 30 ,ug chlorophyll. Ferri-cyanide reduction was followed spectroscopically inan apparatus described elsewhere (6). Reactiontemperatures were held at about 15 C by cooling thecuvette holder with rapidly circulating tap water.An intense beam of light from a 300 w projectionlamp was passed through a red glass filter and di-rected against the frosted side of the reaction vessel.a 1 cm cuvette. The inhibitors to be tested were dis-solved in alcohol. Different concentrations were pre-pared, usually by a process of serial dilution. Afterit had been ascertained that 5 % alcohol alone hadvery little effect on the rate of ferricyanide reduction.these alcoholic inhibitor solutions were added to thereaction mixture in amounts not exceeding 0.1 ml.WVith each inhibitor the concentration which halvedthe reaction rate was determined. For convenience

789


TABLE IAMIDES AS INHIBITORS OF HILL REACTION

GENERAL FORMULA: R-NH-CX-R1

ELEMENTARYMELTING POINTS ANALYSES

No. R X R' PI5 *50 FOUND LITERATURE FOUND CALCULATED

Part I-Anilides from unsubstituted aniline

12345

6789

10

11121314

Phenyl,,

,,

,,

00000

00000

00S0

Part II-4-Chloroanilides15 4-Chlorophenyl1617 "1819 "20

21

2223 "24 "25

26 "27 "28 "29 "30

000000

S

0000

0000S

MethylChloromethylDichloromethylTrichloromethylEthyl

1-Chloroethyl1,1-DichloroethylPropyl2-PropylPhenyl

2-Methyl-1-propenylDimethylaminoDimethylamino2-Propoxy

3.75.04.45.04.7

4.14.44.04.24.0

5.25.22.04.1

HMethylChloromethylDichloromethylTrichloromethylEthyl

Ethyl

1-Chloroethyl1,1-DichloroethylPropyl2-Propyl

Phenyl2-Methyl-1-propenyl2-Methyl-2-propylDimethylamino4-Chloroaniline

2.54.34.74.74.64.5

3.2

4.75.14.75.2

4.55.04.56.34.5s

Part III-3-Chloroanilides313233343536

373839404142

3-Chlorophenyl,,

000000

000000

HMethylChloromethylTrichloromethylEthylPropyl

2-Propyl2-Methyl-1-propenyl3-ChloroanilineDimethylamino2-Propoxy4-Chloro-but-3-yn-1-oxy

3.24.55.05.05.55.3

4.85.05.06.34.45.1

113-114133-13411894105

89-9099-10094-95105-106161-162

127-129129-130132-13387-88

101-102178-179169-170137-138127-128137

77-78

11294-95102-103148-150

192121-122148-149169178-179

55-567699-100101-10287-8845-46

112112-113246139-141

75-76

11413411894103

9210195105160

127-12912790

102179169

141

104153

N 8.04 N 8.00

Cl 44.5 Cl 44.6Cl 52.2 Cl 52.0

Cl 18.3S 16.0

Cl 32.1Cl 42.0

192-193Cl 17.0Cl 16.7

171Cl 23.8S 10.8

79Cl 34.9Cl 52.4

88-89Cl 17.6

Cl 17.8Cl 16.9

245 Cl 23.2Cl 17.9

Cl 17.8S 16.1Cl 32.5Cl 42.1

Cl 17.0Cl 16.8

Cl 23.9S 10.7

Cl 34.8Cl 52.0

Cl 17.9

Cl 17.9Cl 16.9Cl 23.0Cl 17.9

* pI50 is the loglo of the reciprocal of the molar concentration of inhibitor giving 50 % inhibition of ferricyanidereduction.

(Continued on next page)

790 PLANT PHYSIOLOGY



(Continued from preceding page)

TABLE IaAMIDES AS INHIBITORS OF HILL REACTION

GENERAL FORMULA: R-NH-CX R1


No. R X Ri pI50*Np FOUND LITERATURE FOUND CALCULATED

Part IV-2-Chloroanilides2-Chlorophenyl

,,0000

0000

TrichloromethylEthyl1-ChloroethylPropyl

2-Propyl2-Methyl-2-propyl2-Methyl-1-propenylDimethylamino

3.72.62.32.0

2.22.53.43.3

6292-935980

93-947688-8992-93

Cl 51.891

Cl 32.3Cl 17.7

Cl 17.7Cl 16.6Cl 16.9

95

Part V-3,5-Dichloroanilides51 3,5-Dichlorophenyl52 19

53 "5455

56575859

Part VI-2,4-Dichloroanilides60 2,4-Dichlorophenyl61 "

Part VII-3,4-Dichloroanilides62 3,4-Dichlorophenyl63 "646566

67686970

71

7273747576

7778798081

00000

0000

HMethylTrichloromethylEthyl2-Chloroethyl

Propyl2-Propyl2-Methyl-1-propenylDimethylamino

3.24.Os5.75.74.5

4.64.84.76.0

O MethylO 2-Propyl

00000

000S

127186121-122118-12097-98

85-86134-135106-107163-165

2.5s 143-1442.8s 124-125

HMethylChloromethylTrichloromethylBromomethyl

Ethyl1-Chloroethyl1,1-DichloroethylEthyl

0 2-Chloroethyl

00000

00000

2-Propenyl2-Methyl-2-propylPropyl2-Propyl3-Pentyl

2-ButylButyl2-Methyl-l-propyl2-PentylPentyl

3.55.56.56.56.6

6.86.26.04.3

108-109122-123106-107124-12699-101

91-92127-129110-11271-72

Cl 37.5186-187

Cl 58.1Cl 32.8Cl 41.9

Cl 30.9Cl 30.5Cl 29.0Cl 30.2

Cl 37.3

Cl 57.8Cl 32.7Cl 42.1

Cl 30.6Cl 30.6Cl 29.0Cl 30.4

144124

110-112121

91-92133-134

5.2 112-113

6.76.25.76.25.5

6.36.55.07.06.5

120-122145-14679-80

132-134125

112-11373.585-86105-10675-76

77-78135125

71.5-7382-87108-109

Cl 44.7Cl 57.8N 4.98

Cl 49.3Cl 29.5S 13.4

Cl 42.0

Cl 44.7Cl 57.7N 4.96

Cl 49.5Cl 30.4S 13.7Cl 42.1

Cl 28.9 Cl 28.8

Cl 28.8 Cl 28.9

Cl 27.3 Cl 27.3

50 % inhibition of ferricyanide

(Continued on next page)

43444546

47484950

Cl 52.0

Cl 32.5Cl 17.9

Cl 17.9Cl 16.7Cl 16.9

*pI50 is the log,, of the reciprocal of the molar concentration of inhibitor givingreduction.

791


(Continued from Ia)

TABLE IbAMIDES AS INHIBITORS OF HILL REACTION

GENERAL FORMULA: R-NH-CX R1


No. R X R' PI5*FOUND LITERATURE FOUND CALCULATED

Part VII-3,4-Dichloroanilides (Continued)8283848586

8788899091

9293949596

979899100101

102103104105106

107108109110111

112113114115116

117118119120121122

00000

00000

00000

0000C

00000

00000

00000

SSSS0S

Part VIII-Miscellaneous anilides

123 2,3-Dichlorophenyl124 2,5-Dichlorophenyl125 2,4,5-Trichlorophenyl126 2,4,6-Trichlorophenyl127 4-Bromophenyl

00000

HexylHeptylOctylNonylPhenyl

2-Chlorophenyl4-Chlorophenyl2,4-Dichlorophenyl3,4-DichlorophenylCyclohexyl

Benzyl3,4-DichlorobenzylPhenoxymethyl2,4-Dichlorophenoxymethyl2-Methyl-1-propenyl

3-PhenylpropylTrans-2-phenylethenyl2,4,5-Trichlorophenoxymethyl2-Naphthylmethyl1-Naphthylmethyl

MethylaminoEthylaminoPropylaminoButylaminoHexylamino

CyclohexylaminoBenzylamino2-HydroxylethylaminoDimethylaminoDiethylamino

DipropylaminoDi- (2-propyl) aminoPiperidinoMorpholinoDi- (2-hydroxyethyl) amino

DimethylaminoMethylaminoDiethylaminoEthylaminoAminoAmino

of isobutyric acid2-Propyl

5.26.05.45.45.2

4.85.75.56.06.0

6.85.65.05.55.9

5.15.25.25.45.6

7.06.25.86.75.5

5.55.54.87.56.8

4.75.36.86.44.5

4.53.74.33.84.73.3

3.Os3.Os4.55.Os5.0

61.54269-7070-71

145-146

152-153172-173156-157227-228137-138

132186-187141-142160-161103

74-75179-181141-142157-158170-172

156-157175-176128-129121-122104-105

183-184171-172137-138156-157111-112

96-97130-131172-173153-154153-154

161-162148-15095-96114-115152-154209-210

108-109137-139145-146151-152151-152

60-61.5Cl 24.4Cl 23.4Cl 22.5Cl 26.7

Cl 34.8Cl 35.5Cl 42.1Cl 42.3Cl 26.0

Cl 25.3Cl 40.3Cl 24.0Cl 39.0Cl 29.0

Cl 23.1Cl 24.7Cl 43.4Cl 21.6Cl 21.7

155.5179.5

Cl 28.2Cl 26.8Cl 24.9

188 Cl 24.6Cl 24.0Cl 28.6

158-159Cl 26.3

157.5157.0

166.0147.5

155.5203-204

150-151

Cl 24.2Cl 24.4Cl 26.1Cl 25.5Cl 24.3

Cl 25.6 Cl 25.6Cl 28.5 Cl 28.5

Cl 30.4Cl 30.4Cl 39.9Cl 39.4

*pI50 is the log10 of the reciprocal of the molar concentration of

reduction.inhibitor giving 50 % inhibition of ferricyanide

(CoIntinuedi onI next lpage)

Cl 24.6Cl 23.5Cl 22.4Cl 26.7

Cl 34.5Cl 36.0Cl 42.4Cl 42.4Cl 26.1

Cl 25.3Cl 40.7Cl 24.0Cl 39.0Cl 29.0

Cl 23.0Cl 24.3Cl 44.5Cl 21.6Cl 21.6

Cl 28.7Cl 27.1Cl 24.5Cl 24.6Cl 24.2Cl 28.5

Cl 26.2

Cl 24.5Cl 24.5C1 26.1Cl 25.8Cl 24.2

Cl 30.6Cl 30.6Cl 40.0Cl 40.0

792 PLANT PHYSIOLOGY



(Conitinued from Ib)

TABLE ICAMIDES AS INHIBITORS OF HILL REACTION

GENERAL FORMULA: R-NH-CX-R'

No. R xMELTING POINTS

FOUND LITERATURE

ELEMENTARYANALYSES

FOUND CALCULATED%/ %/

Part VIII-Miscellaneous anilides of isobutyric acid (Continued)2-Methylphenyl4-Methylphenyl3-Methylphenyl2-Methoxyphenyl4-Methoxyphenol

4-Nitrophenyl3-Nitrophenyl3-Chloro-4-methylphenyl2-Methyl-3-chlorophenyl2-Methyl-4-chlorophenyl

3-Nitro-4-methylphenyl1-Naphthyl5,6,7,8-Tetrahydro-2-naphthyl4-Dimethylaminophenyl2,6-Dimethylphenyl

00000

00000

00000

2-PropylPt

,,

,,9

,,p

2.34.54.82.05.Os

4.03.85.82.5s2.5s

4.82.Os5.03.02.7

Part IX-Alkylamides & alkylureasCyclohexyl 0 2-PropylBenzyl 0 "

; Octyl 0 Dimethylaminor.-1-11-11 n pp.ycionexyiBenzyl

3.32.35.64.83.60

115-116108-10982-8344109-111

167-16993146-147142-143163-164

106-107147-149102157-158

115-11610885

N 7.19 N 7.25N 7.26 N 7.26

N 13.4N 13.5Cl 16.8Cl 16.8Cl 16.8

NNNNN

116-11791-9227-28156-15776-77

12.56.576.57

13.87.26

N 8.32N 8.05N 13.7N 16.5N 15.7

N 13.45N 13.45Cl 16.8Cl 16.8Cl 16.8

NNNNN

NNNNN

12.66.586.45

13.67.35

8.307.92

14.016.415.7

*pI50 is the log1,0 of the reciprocal of the molar concentrationreduction.

of inhibitor giving 50 % inhibition of ferricyanide

these concentrations have been expressed on a loga-rithmic scale. Thus the log10 of the reciprocal ofthe molar concentration giving 50 % inhibition isreferred to as the pI50. The inhibitors with thelarger p150's are the more potent and a difference inpT50 of one unit represents a tenfold difference inpotency.

Several points on the concentration-inhibitioncurve were established for each compound in orderto avoid spurious conclusions which might arise froman inadvertent saturating of the medium with in-hibitor. (With saturated or transiently supersatur-ated solutions, adding more inhibitor does not neces-

sarily increase the effective concentration and mayeven decrease it by initiating crystallization.) Solu-bility problems did occur with a number of substances.In fact, many of the compounds prepared and testedare not included in this report because it was im-possible to achieve concentrations high enough tocause 50 % inhibition. Other substances were doubt-ful, but by utilizing the fleeting phenomenon of super-saturation often it was possible to obtain the desired

level of inhibition. However, to indicate our reserva-

tions we have added the letter "s" after the pI50value listed in table I whenever the true inhibitorconcentration was in doubt. Fortunately most of thehighly active inhibitors presented no problem, eitherbecause their water solubilities, although very small,were adequate to supply the extremely low inhibitorrequirement or because, at the very great dilutioninvolved, crystallization was so slow as to render thesupersaturated state quasi-stable.

III. INHIBITION OF ATP FORMATION. Chloro-plasts (60-90 Ag chlorophyll) were illuminated in a

reaction medium consisting of the pH 7.95 buffer,ATP (10-3 M), K2HPO4 labeled with p32 (10-2 M),magnesium sulfate (3 X 10-4 M), hexokinase (3.0mg), glucose (2 X 10-2 M), either FMN or PMS(10-4M), a trace of horse liver catalase to hastenthe decomposition of the hydrogen peroxide formedon oxidation of the reduced flavin, and variousamounts of inhibitor in 0.2 ml of alcohol. The totalvolume was 5.0 ml. The reactions were carried out

128129130131132

133134135136137

138139140141142

143144145146147

793


PLANT PHYSIOLOGY

in test tubes (15 mm I.D.) supported in a wire rackimmersed in a large rectangular glass water bath.Under the bath was a mirror. Light from a 500 wincandescent lamp was directed down on the bath andlight from twvo 300 w reflector lamps was directedagainst eaclh side of the batlh. The temperature ofthe bath was held constant + 1 C during the 10minutes of illumination either by adding ice or bycirculating tap water. The reaction temperature indifferent experiments varied from 3 C to 20 C, al-though most of the experiments were con(lucted atabouit ten Ceiltigrade. After illuminatioln, the reactionmixture wvas poured into 10 p(Ierchloric aci(d alnd theunreacte(l ortlhoplhosphate xvas extracted as phospho-molvbdate 1v the method of Nielsen an(d Leninger(15). The organic phosphate (glucose-6-phosphatesynthesise(d by hexokinase from glucose & labeledATP) w\Nasnmeasured as the residual radioactivity ofthe aci(lified medium.

IV. HYDROGEN BOND FOR-ATIONT. Variousacylanilides, acylamides, alkylureas, phenylureas, andtriazines were dissolved in chloroform (approximate-ly 0.067 or 0.5 M), in chloroform + 5 '; acetone(0.06/7 Ar), and in carbontetrachloride (0.067 WI).There is a characteristic absorption bancl of the N-Hbond at 2.9 ,u (wave number = 3,440 cm-1) and acorresponding absorption band of the N-H .. . 0 bondnear 3.0 ,u. Consequently tlle absorption spectrum be-tween 2.6 and 3.2 ju of each substance in each solventsystem was (letermined with a Perkin-Elmer Model21 infra-red spectrophotometer.

RESULTS

The substances with which the report deals maybe classified on the basis of their chemical structuresinto three unequal groups:

I. Amides having the general formula R-NH-CX-Rl wThere X is oxygen of sulfur, R and R1 arealkyl or aryl, substituted or unsubstituted; R1 mayalso be alkylamino, dialkvlamlino, or arvlamino. (Seetables I & II).

II. Miscellaneous compounds resembling groupI) in that they are also amiides but not fitting thegeneral formula, e.g., N-methylanilides, sulfonamides.

III. A small number of symmetrical triazineshaving the general formula:

x

vCN N

11N

where X is chlorine or miethoxy. Y and Z are ethyl-amino, or diethylamino. (See table III).

A. INHIBITION OF FERRICYANIDE REDUCTION.There are several considlerations xvhich may throw

doubt on comparisons of inhibitor potencies. It isconventional to express the activity of the inhibitorin terms which involve its molar concentration. Nowa) if the enzyme-inhibitor reaction is essentially ir-reversible or b) if the inhibitor's solubilities are suchthat it partitions betweeni water ani(l the biologicalmaterial in a manner whliclh strongly favors the latter,then the significant parameter becomes the amountof inhibitor in each chloroplast an(d not the originalconcentration of inhibitor in the suspension ime(liumn.Unfortunately, the real condition is frequently inter-me(liate andI neither the overall inhibitor concentrationnor total amoulnt of inhibitor addced is uniquely -sig-nificant. This may make kinetic analy-sis of the situa-tionialmost impossible. \Ve have followved the populartren(l and(I expressed inllibitor potencies as pI,0,,'s,that is, log,1 of the reciprocal of the molar concentra-tioIn cauisinig 50 C ilnlibition. However, all the in-hibitor stu(dies repor-tedl in tables I. IT, anld III werema(le wvithl the same or nearly the same plroportionof chloroplasts to medium. Therefore, the concen-tration (lata expressed as pI50's may be converted intodlata expressedl as amounts of inhibitor per chloroplastor per ug chlorophyll without chainging in any wANaythe or(ler- of potencies. Of course, this still leav-esun(lecided the questioni of the proportion of the applie(dinhibitor actually in the chloroplasts (which propor-tion may be quite (lifferent with (lifferelnt inlhibitors)and it must be recognizedl that these aln(d all simiiilarvaluies for inhibitor potency are the outcomlle of at leasttwo unrelated properties of the inhibitor: its inherenteffectiveniess as an enzyme poison aln(d the coefficientof its partitioning betweeln ater anid chloroplasts.

In or(ler to obtain the maximumll Hill reactionrates in our test system, we were very careful to utSesaturating light intensities an(d to uncouple the phos-phorylating electron transport system wvith methy-l-amine (6). Under these conditions the uninllibite(drates wN-ere of the or(ler of 1 millimlole ferricyanidere(luce(l per mg chlorophyll per hour.

The reproductibility of the p1,5,'s in our test variedlwitlh (lifferent substances. In compariing the poten-cies of the various substances, it is probably safe toconsider a difference of 0.2 as of borderline signifi-cance and a difference of 0.3 as quite significant.

I. ACYLANILIDES, THIOACYLANILIDES, ACYL-AMIDES, PHENYLUREAS, PHENYLTHIOUREAS. ALKYL-UREAS, & PHENYLCARBAMATES HAVING GENERALFORMIULA R-NH-CX-R1 (table 1).

Part I. Derivatives of uttsubstihtted aniilinie.The unsubstituted anilides are not nearly as effectiveas many of the substituted anilides (lescribed below.Among the unsubstituted anilides. the acetvl deriva-tive (No. 1) is a poor inhlibitor while the chloroacet-l(No. 2), the trichloracetyl (No. 4) and the ,'. /8-dimethylacrylyl (No. 11) are all fairly good and aboutequally potent. The well known herbicide, phenvldi-methylurea (No. 12), has the same activity as thebetter acyl derivatives but its sulfur analog (No. 13)is almost inactive. Note that the only carbamate(No. 14) is not verv effective as a Hill reaction in-

794



hibitor although it is a commercial herbicide. Theprobable mode of action of the phenylcarbamates willbe compared with the mode of action of the otheranilides in the discussion.

Part II. Derivatives of 4-chloroaniline. Theseexhibit a wider range of activities than do the un-substituted anilides; the potencies of the various de-rivatives do not follow an entirely parallel pattern.The formyl (No. 15) and the thiopropionyl (No. 21)derivatives are particularly poor. The a, a-dichloro-propionyl (No. 23), the isobutyryl (No. 25), andthe ,l, /3-dimethylacrylyl (No. 27) derivatives arereasonably good. The dimethylcarbamyl derivative(No. 29), (4-chlorophenyldimethylurea, CMU or

monuron), is at least 10 times more potent than itsnearest competitor.

Part III. Derivatives of 3-chloroaniline. Againthe sequence of relative potencies is different in some

details. The formyl derivative (No. 31) has verylow activity. The acetyl derivative (No. 32) and theisopropylcarbamate (No. 41) are rather poor. The3-chloro-analog of CMU (No. 40) is, like CMU, out-standingly better than the other members of thisgroup.

Part IV. Derivatives of 2-chloroaniline. With-out exception these anilides are feeble inhibitors.

Part V. Derivatives of 3,5-dichloroaniline. Al-though the average potency is about the same as in

the para- and meta-substituted series, these disubsti-tuted anilides do not differ as widely. Again thedimethylurea (No. 59) has the highest activity, butonly by a narrow margin.

Part VI. Derivatives of 2,4-dichloroaniline. Thetwo substances in this class resemble 2-chloroanilides.Apparently a chloro group in the ortho position inter-feres seriously with the mechanism of inhibition.

Part VII. Derivatives of 3,4-dichloroaniline.The 3,4-dichloroanilides are the most active inhibitorswe have investigated. Indeed, there is not a singleinstance of another anilide or alkylamide exceedingthe corresponding 3,4-dichloroanilide in potency.Consequently, it seemed appropriate to use 3,4-dichloroaniline as the common moiety in a large num-

ber of substances to determine how the structure ofthe rest of the molecule affected the inhibitory func-tions. Admittedly, this investigation was not verysystematic but, unfortunately, no principles on whichto base a systematic study were apparent. Even now,

widely applicable generalizations relating activity tothe structure of the acyl moiety of the anilides eludeus. However, it is possible to make certain generaliza-tions of limited applicability:

Among the anilides of aliphatic acids, inhibitoractivity is a periodic function of the chain length ofthe acid (See fig 1). There is a maximum at thepropionyl derivative (No. 67), another maximum

TABLE IIANILIDES OF ISOBUTYRIC ACID AS HILL REACTION INHIBITORS

POSITION OCCUPIED ON BENZENE RINGNo. p

2 3 4 5 6 50

9 4.225 Cl 5.237 Cl 4.847 Cl 2.261 Cl Cl 2.857 Cl Cl 4.875 Cl Cl 6.2123 Cl Cl 3.0124 Cl Cl 3.0125 Cl C1 Cl 4.5126 Cl Cl Cl 5.Os127 Br 5.0128 CH3 2.3130 CH3 4.8129 CH3 4.5142 CH3 CH3 2.7s132 CH30 5.0s131 CH30 2.0133 NO2 4.0134 NO2 3.8135 Cl CH3 5.8136 CH3 Cl 2.0137 CH3 Cl 2.5s138 NO2 CH3 4.8141 (CH3) 2N 3.0

$ pI is the loglo of the reciprocal of the molar concentration of inhibitor giving 50 % inhibition of ferricyanidereduction. "s" following pI50 indicates doubts concerning the actual amount of inhibitor in solution.

795


PLANT PHYSIOLOGY

N-ACY L - 3,4- DI CH LO R O AN ILI NhE S

2 3 4 5 6 7 8 9 10

No. Of Carbons In Acyl Group

FIG. 1. The effectiveness of 3,4-dichloroanilides as in-lhibitors of the reduction of ferricyanide by illuminatedchloroplasts. pI50 is the log, 0 of the reciprocal of themolar concentration giving 50 % inhibition. The abscissanumbers refer to the carbons in the normal acyl moiety.

which includes the valeryl (No. 78) andl the caproyl(No. 81) derivatives, and a third lower maximum atthe octanoyl derivative (No. 83). Branched or sub-stituted aliphatic acids somewhat resemble the parentstraight chain acid. Thus the methylpropionic (iso-butyric) anilide (No. 75), like the propionyl deriva-tive, is considerably more active than the butyricanilide (No. 74), while an-methylbutyric acid yieldsan anilide (No. 77) wlhich is intermediate (perhapsbecause it is also a-ethylpropionic acid?). In thisconnection, an interesting substance is the 3,4-di-chloroanilide of a-methylvaleric acid, the herbici(leKarsil (No. 80). This is one of the most effectiveinlhibitors of the Hill reaction (lescribe(l. It is con-

ceivable that Karsil owes its unlusual potency to thefact that a-methylvaleric acid is also a-butylpropionicaci(l: in a sense Kar-sil sits oni both the first and

secon(lmaxima shown in figuire 1.Replacing a methyl group) in the alipliatic aci(d by

a chlorine atom. (lespite the inevitable major rear-

ranlgement of electron delnsities, results in surprisinglylittle change in the activity of the anilide. Thechloroacetanilide (No. 64) resembles the highly activepronionyl derivative (No. 67). the anilide of 8-chloropropionic acid (No. 71) is similar to the anilideof u-butyric acid (No. 74), and(l tile a-chloropropionyl(lerivative (No. 68) lhas the samiie activity as the iso-l)utyryl derivative (No. 75). On the average, how-ever, the chloro- aci(ls formii 3,4-dichloroanilides ofslightly lower activity than (lo their methyl counter-parts. The effect of replacing a methyl group bya bromine atomii seems to be simiiilar since the brom-

acetanilide (No. 66) is also comparable to the pro-piollic anilide in activity.

Polar groups sharply reduce the activity of the in-hibitor. Thus. 3- (3,4-dichlorophenyl) -1- (2-hydroxy-ethyl) - and 3- (3,4-dichlorophenyl) -1, 1-bis (2-hy-droxyethvl)- ulreas (No. 109 & No. 116) are poor

inhibitors. On the other hand, the (lehy(lrated ver-sion of the latter compound, in which the two hy-droxyl groups are replacedl by a relatively non-polarether bridge (No. 115), is 100 times nmore activethan its polar counterpart. It is uncertain whethlerthis inactivity of polar substances should be explainedin terms of enzyme affinities or on the basis of par-titioning and penetration; inactivity couldl result fromthe fact that suclh inhibitors fail to enter the lipid-rich chloroplasts in significant amounts.

The thiopropionyl-3.4-(dichloraniline (No. 70) andthe 3,4-dichlorophenyl-thioureas (No. 117 to No. 120)are from 300 to 2,000 timles less effective than theoxygen analogs.

The important herbicide 3- (3,4-dichlorophenyl ) -l,j-dinmethylurea or DCAIU (No. 110) is the mostinhibitory substance we have encountere(l. In thisrespect, the 3,4-dichloroanilides resemble the otheranilides anid alkylamids examined: the dlimethylcarl)-amyl (lerivative is always the most active of the (le-rivatives tested. However, other 3,4-dichloroanilidesare also very good inhibitors; some of their activitiesapproaclh that of DCMU. Among these are the cor-respon(ling monomethylurea (No. 102), the diethyl-urea (No. 111), the anilide of a-methylvaleric aci(d(Karsil, No. 80), the anilides of propionic (No. 67)and phenylacetic (No. 92) acids, and the piperidinocompoun(l (No. 114).

Part VIII. Anilidcs of isobuftyric acid (table II).Isobutyric acid was chosen to contribute the commlonacyl nioiety when a study of the influence of theamine miioiety on inhibitor potency was undertaken.The reasons for this choice were the availability ofisobutyric anhydridle and the simplicity of its use informingi amides, the desirable crystal cliaracteri sti csof most of the amides. and the moderately high ac-tivity of Imlost of the isobutyranilides alrea(ly teste(l.

The (lata, most of wvhiclh are brought together intable TIT may be summarized as follows: \Vhenchloriine atoms, bromine atoms, methoxv groups. or-methyl groups replace the para- and meta-hydrogensof the aniline, there is a marked increase in activitv.The effect is greatest with chlorine, slightly less wN-itlbromine and methoxy, and again slightly less -withmethyl. M\Ieta-para dlisubstituted compoun(ds areloublv activated wvlhile meta-meta compounld(s are lnot.On the other lhanid, anv ortho substitution causes astriking reduction in activity. The adverse effect ofan ortho-clhloro caan be overcome in part by addinga meta-para pair as in the 2,4,5-trichloranilide (No.125). Surprisingly, two ortho-chloro atoms reducethe activity muclh less thaln one does. This may heseen by compalinig the 2,4-dichloroanilide (No. 61)with the 2,4,6-trichloroanilide (No. 126). Morelandan(l Hill lhave obserived a similar relationship betweenderivatives of 2-chloro- and 2,6-dichloroaniline (11).The same is not true of two ortho methyl groups; the2-methylanilide (No. 128) and the 2,6-dimethylanilide(No. 142) are both exceedingly poor inhibitors.Nitro groups, either meta or para, reduce the activityby a small amount while the dimethylamino group inthe para positioln almost abolislhes the inhibitory ef-

796

ul0


797GOOD-HILL REACTION INHIBITORS

TABLE IIITRIAZINES AS HILL REACTION INHIBITORS

y

EthylaminoIsopropylaminoEthylamilnoIsopropylaminoIsopropylaminoEthylaminoIsopropylamino

GENERAL FORMULA

x

/C\N NI DI

Y-C Xc-ZN

z

EthylaminoIsopropylaminoDiethylaminoDiethylaminoEthylaminoEthylaminoIsopropylamino

p150*6.46.34.74.96.65.75.8

TRADE NAME

SimazinePropazineTrietazineIpazineAtrazineSimetoneMethoxypropazine

*pIPI is the log1, of the reciprocal of the molar concentration giving 50 % inhibition of ferricyanide reduction.

fect (No. 141). The 3,4-cycloalkyl-substituted anil-ide, N-isobutyryl-5,6,7,8-tetrahydro-2-naphthylamine(No. 140) is a good inhibitor as might have been pre-

dicted, and the 2,3-disubstituted anilide, N-isobutyryl-1-naphthylamine (No. 139) has the low activity whichis almost always associated with ortho-substitutedsubstances.

The steric and other effects of ortho substitutionwill be discussed below under the topic of hydrogenbond formation.

Part IX. Alkylamnides & alkylureas. The greaterpart of this paper deals with acylanilides and phenyl-ureas, partly for historical reasons and partly becausethey seem to be more active than any alkylamide or

alkylurea yet tested. Indeed the amides formed fromthe smaller alkylamines, such as propylamine, ethyl-amine, and methylamine, are completely inactive.However, n-octylamine (No. 145) and cyclohexyl-amine (Nos. 143, 146) yield ureas and amides which

are fair inhibitors. No derivatives of higher amineshave been prepared to date, nor has the possibility ofusing branched or substituted aliphatic amines beeninvestigated.

II. N-METHYLANILIDES & SULFONAMIDES. N-methyl-N-chloracetylaniline was the only secondaryamide tested during this investigation. (Meltingpoints; found 68 C, literature 70 C). It was a verypoor inhibitor with a PI50 of 2.3. This agrees withthe observation of other workers who tested other N-methylaniline derivatives (20, 11). Apparently animino hydrogen is required for inhibition.

Several sulfonamides were prepared and tested.Two of these, the p-toluene-sulfonyl derivatives of n-

octylamine (mp 55 C) and 3,4-dichloroaniline (mp144 C), were fair inhibitors having pI50's of 4.5.Since no further tests were made on these compounds,it is not known if sulfonamides and the amides of car-

boxyacids inhibit the same biological processes.

TABLE IVINHIBITIONS OF ATP FORMATION IN FMN & PMS CATALYSED SYSTEMS

No. NAME PI50 WITH FMN PI50 WITH PMS(SPECIFIC INHIBITION) (UNSPECIFIC INHIBITION)

2 N-Chloracetylaniline 4.3 < 2.029 3-(4-Chlorophenyl)-1,1-dimethylurea (Monuron, CMU) 6.2 < 4.063 N-Acetyl-3,4-dichloroaniline 4.8 3.569 N-Propionyl-3,4-dichloroaniline 6.2 4.064 N-Chloracetyl-3,4-dichloroaniline 5.8 4.5

65 N-Trichloracetyl-3,4-dichloroaniline 6.3 4.880 N- (2-Methylvaleryl) -3,4-dichloroaniline (Karsil) 7.0 4.9110 3-(3,4-Dichlorophenyl)-1,1-dimethylurea (Diuron, DCMU) 7.2 <4.0145 3-Octyl-1,1-dimethylurea 4.6 3.2

2-Chloro-4,6-bis (ethylamino) -s-triazine (Simazine) 5.6 < 3.52-Chloro-4- (2-propylamino) -6-ethylamino-s-triazine (Atrazine) 6.2 < 3.52-Chloro-4,5-bis (2-propylamino) -s--triazine (Propazine) 5.7 < 3.5

x

C1ClClClClCH30-CH30-


PLANT PHYSIOLOGY

III. HERBICIDAL TRIAZINES (table III). Themost active triazine inhibitors are about as effectiveas the average 3,4-dichloroanilide an(d about 10 timesless active than the best of the anilides. Not until amore diverse group of triazines has been examinedwill it be possible to establish reliable rules for pre-dicting the potency of these inhibitors from theirmolecular structures. On the basis of the limited in-formation provided here, two tentative generalizationsare presented: Chlorine atoms are appreciably betterthan metlhoxy groups, and two imino hydrogens arevery muclh better than one. It is interesting to spec-

Fi g. 2

100

E

a.L4-

6.0 5 5 5.0

p I

Fig. 3Propazi n e

p I

FIG. 2. The inhibition of photophosphorylation by the3,4-dichloroanilide of a-methylvaleric acid (Karsil) andthe 4-chloroanilide of dimethylcarbamic acid (CMU).Electron transport was mediated either by riboflavin-5-phosphate (FMN) or by N-methylphenazonium ion(PMS). pI is the log1O of the reciprocal of the inhibitorconcentration.

FIG. 3. The inhibition of photophosphorylation by 2-chloro - 4,6 - bis (isopropylamino) - s -triazine (Propazine) .

Concentrations to the right of the dotted line representsupersaturated solutions. FMN, PMS, and pI have thesame significance as in figure 2.

ulate as to the significance of the common require-ment for imino hydrogens in these apparently verydifferent types of molecules-the triazines and theanilides.

B. Localigation, of Site of Inhibition. (Figs 2& 3 & table IV). It has been known for several vearsthat CMIU inhibits the formation of ATP by illumin-ated chloroplasts when either ferricyanide or FMNis the electron acceptor but not wvhen N-methylphena-zonium ions (PMS) are the acceptors. It was notsuirprising, therefore, to find that the chemicallv re-lated acVlanilides such as Karsil (No. 80) exhibiteda similar behavior. Actually Karsil has two inhibi-tory functions, one requiring at least 100 times moreinhibitor thlaln the other. (See the binmodal Karsilcurve in fig 2). Apparently the chloroplast prepara-tion employed for figure 2, unlike milost preparations.was unal)le to lw-pass completely the oxv-gen-procluc-ing systenm in the presence of PMvS. Conseqluientlyat verv low Karsil concentrations all of the FMN-catalysed ATP formationi an(d a part of the PM\S-catalvse(l ATP formlationi were inhibited. \Vith in-creasing inhibitor concentration no further increasein the inhibitioni of the P1MS-catalv-sed system occur-red utntil quite higlh concentrationis ha(l been reaclhed.Probably the curve for CMU inhibition also wvouildhave risen again hadl it been possible to achieve therequisite high concentrationi. W;Te slhall refer to thesetwo types of inhibitioln as specific (at lowv concentra-tions) an(l unspecific (at high concentrations).

Tn table IV is sho-wn the specific an(l unspecificpotency of several inhlibitors. [The reader should bewarned that the clata on unspecific inhibition (pT5,,'srith PATS as catalyst) care quantitatively, unreliablebecause of the limited solubilities of the inhibitors:concentrations (luoted are calculate(d on the basis ofthe amount of inhibitor addedl to the mle(liumi andignlore the possibility of saturation of the medium.This problem was particularly severe with the veryinsoluble DCMU & triazines.1 Tf the Jagendorf cri-terion is a reliable guide to the nature of the processinhibited, w\e nmust conclude that each of the twelveinhibitors liste(d in table IV-and presumablv mostof the large number of inhibitors with wlhich thispaper is concernedl-are primarily inhlibitors of themeclhanism of water oxidation. As already indicated.the chemical simiilarity of the phenvlureas. phenvl-carbamates. an(l acylanilides is such that a commonmeclhanismii of inhibition was not unexpecte(l. Howr-ever, the similar pattern of inhibition witlh the tria-zines was a surprise. Errors arising froml the solu-bility problems mlentioned above cannot have misledus since Simazine, Propazine (fig 3), an(d Atrazineinhibite(d ATP formation in the FAMN system com-pletely at concentrations which caused at most a fewpercent inhibition of the PiNIS system.

C. HA?drogen Bond Form1ation (Figs 4 & 5).The inhibiting substances (lescribed in this paper arevery stable compounds. They react only with power-ful chemical reagents and many of them resist degra-dation by the enzynmes of plants and( bacteria. Pre-

798

z

E

1-



sumably, they are absorbed and held on a catalyticsurface such as the active site of an enzyme by phys-ical forces rather than by chemical reactions. [Notethat phenylurethane, the classical narcotic or surface-blocking inhibitor of Warburg (18), belongs to theclass of anilides with which we have been dealing.Indeed, it was one of the chloroplast inhibitors in-vestigated by Wessels and van der Veen (20)]. Alikely binding force, one able to hold substances suchas these reasonably firmly, is the electrostatic attrac-tion between the positive imino hydrogen atom of theinhibitor and the negative carbonyl oxygen of a poly-peptide chain. With the amide and anilide inhibitors,this force could be augmented by the reciprocal attrac-tion between the inhibitor's carbonyl oxygen and oneof the protein's imino hydrogens.

In non-aqueous solutions of anilides, dimers andhigher aggregates are produced, the imino hydrogenforming an electrostatic bond with the carbonyl ofanother like molecule. There is, therefore, an amountof bonding which depends on the concentration ofthe anilide, on the electron densities throughout themolecule which determine the electro-positivity of thehydrogen and the electro-negativity of the oxygen,and on steric considerations which may limit the ac-cessibility of either the hydrogen or the oxygen. Todetermine the relative bonding potentials of the iminohydrogens of the various inhibitors without the com-plication of variations of the carbonyl oxygen, weadded an excess of acetone to some of the solutions.With an excess of a common carbonyl compound pres-ent, differences in the amount of bonding should de-

Fi g.

I

01c l '''

C o

64

CI

Oi- N C- CH 2 C H3

72

5

Cl

KI H CH3

112 ( C MU)

Cl

C'

CH3s ,CHsoC-N-C ,CN- C

CH3 H H C H 3

(Propazsin )

FIGS. 4 & 5. The absorption spectra of various anilides and Propazine between 2.6 A and 3.2 A. The solid linesshow the absorption of 0.67 M solutions in chloroform and the dotted lines show the absorption by the same substanceswhen the chloroform solvent contained 5 % acetone. For clarity of illustration the dotted curves have been displacedto the right; actually the narrower peaks on the left of both the solid and dotted curves represent absorption by theN-H bond and are at the same wave length. The broader peaks to the right, which are invariably higher in thedotted acetone curves, represent absorption by the N-H . .. 0 bond.

799

Fi g. 4

° CH3-N-C- C

H "'CH3

9

11 ,..7

47

A;

ci

Ci

77

CM3

Ci

37

Cl

128


PLANT PHYSIOLOGY

pend exclusively on differences in the nature andposition of the imino hydrogen. To investigate thecharacteristics of the imino hydrogens of triazines,which have no carbonyl oxygen of their own, it wasobviously also necessary to add an outside source ofelectronegative groups such as acetone.

Hydrogen bonding was measured as a decrease inthe N-H absorption in the region about 3,440 cm-'and the appearance of N-H ... 0 absorption about3,340 cm-'. Twenty-five acylanilides, thioacylanil-ides, phenylureas, phenylthioureas, and triazines werestudied. The findings are summarized in the follow-ing generalizations:

I. When the acyl moiety is acetyl, propionyl, orisobutyryl, anilides form hydrogen bonds readily.This tendency is increased in both the 3- and 4-chloro-anilides and is even greater in the 3,4- and 3,5-di-chloroanilides. The increase in hydrogen bond forma-tion is primarily due to changes in the hydrogen sincethe same increase occurs in the acetone-containingsolvent. Para- and meta-methyl substitution of theaniline has a similar but much smaller effect. Allthese observations parallel the inhibition data andconsequently support the hypothesis that hydrogenbonding plays an important role in the mechanism ofinhibition.

II. Also in support of the hypothesis is the factthat the inactive 2-chloroanilides form hydrogenbonds neither between like molecules nor with ace-tone. This also applies to the 2,4-dichloroanilide(No. 61) and the 2,4,5-trichloroanilide (No. 125).Moreover, the 2,4,6-trichloroanilide (No. 126), whichas we have noted, is an unexpectedly effective inhibi-tor, forms hydrogen bonds to about the same extentas the unsubstituted anilides. The inactive 2-methyl-anilide (No. 128) and the equally inactive 2,6-di-methvlanilide (No. 142) do form hydrogen bonds,but rather reluctantly and the N-H ... 0 absorptionband is wi(lened and shifted toward the N-H band.

III. Both 4-nitro (No. 133) and 4-dimethylamino(No. 141) anilides form hydrogen bonds but not asfreely as the unsubstituted anilides; these substitu-tions also reduce inhibition, very markedly in thelatter case.

IV. On the other hand, halogen substitution ofthe acyl moiety reduces the tendency to form N-H... 0 bonds without causing a corresponding diminu-tion of inhibition potency. In chloroform solutions thehighly active trichloracetvl-3,4-dichloroaniline (No.65) gives only a very small peak at 3,340 cm-1 andthe highly active chloracetyl-3,4-dichloroaniline (No.64) gives none. Even in the acetone solvent thehydrogen bonding tendency is very much reduced.

V. The thioanalogs of acylanilides and of phenyl-ureas have an absorption peak at about 3,340 cm-.In fact the spectra of the thio compounds and the cor-responding oxygen compounds are indistinguishablein the region between 2.6 and 3.2. I was surprised,since sulfur atoms (as in H,S) do not ordinarily

form this type of bond with hydrogens. However, itis well known that the sulfur atoms in compoundssuch as thioureas and thioamides are exceptionallyelectronegative. The tendency of the imino hydro-gen of thiopropionyl-3,4-dichloroaniline (No. 70) toform bonds with the carbonyl oxygen of acetone isapparently the same as in the oxygen anilides. (Seefig 5).

VI. The N-H ... 0 peak in the most powerfulinhibitors of all, the phenyldimethylureas, is lowerthan in the acylanilides; both the N-H and N-H ... 0peaks are moved through about 30 wave numbers to-wards shorter wave lengths.

VII. The imino hydrogens in the triazines donot form bonds with the carbonyl oxygen of acetoneto an appreciable extent.

DISCUSSION

STRUCTURE & ACTIVITY. A variety of substances,including compounds with widely different molecularconfigurations, are powerful inhibitors of the Hillreaction. Apparently these inhibitors interfere withthe same overall process since they prevent the reduc-tion of ferricyanide and FMN by illuminated chloro-plasts without seriously hindering PMS-catalysedphotophosphorylation. The steric requirements forhigh activity follow a consistent pattern among thedifferent classes of anilides and alkylamides and,therefore, there can le little (loubt that these act ata common site. There is, however, no convincingreason for believing that the triazines mlust inhibitat the same site. Possibly several catalyse(d reactionsnot required in PMS-mediate(d phosphorylations areinvolved in the reduction of ferricvanide and FMN.Consequently, it is possible that the poisoning of anyone of several catalysts (enzymes?) woul(d explainmy experimental results. On the other hand, if thegeneral formula is broadened for the anilide andamide inhibitors, R-NH-CX-R', to include not onlyX equals 0 but also X equals N-, the triazines maybe accommodated; moreover the results of experi-ments employing combinations of phenylureas andtriazine are consistent with a single site of inhibition(unpublished). In view of this uncertainty regard-ing the functional relationship of the triazines to theother inhibitors, the triazines will be disregarded inthe following discussion of structure and activity.

Even among the anili(les, compounds with quite(lifferent structures (e.g. DCMU & Karsil) havesimilar high activities. However, the apparent lowlevel of specificity on the part of the inhibited catalystcan scarcely be a manifestation of a general low af-finity; so few molecules of DCM\U are present in aninhibited preparation that a considerable proportionof them must reach, and remain at, the sensitive sites.Although qjuite different classes of amide inhibitorsmay share the property of inhibiting at low concen-trations, within any one class closely similar com-pounds may differ sharply. For instance. N-valeryl-3,4-dichloroaniline (No. 78) is 30 tinmes more active

800



than N-isovaleryl-3,4-dichloroaniline (No. 79).Steric considerations seem to be more important indetermining the effectiveness of these inhibitors thanare factors affecting the distribution of electron den-sities. Thus chloroacyl-anilines closely resemble thecorresponding methyl-substituted acyl compounds.Moreover, para-and meta-substitution of the anilinemoiety are equivalent and are opposite in effect toortho-substitutions; electronically the para and orthosubstituted anilines are related and contrast withthe meta. This preponderance of steric influences isconsistent with a mechanism of inhibition in whichthe active site is physically obstructed; the importantthing is that the inhibitor fit the enzyme well. Thegreat chemical stability of these anilides and thereversibility of the inhibition caused by them (20)argue the same type of inhibition. Electron dis-tributions are more pertinent to chemical reactivityand, therefore, may be expected to play a bigger partin determining the effectiveness of those inhibitorsactually reacting with the catalytically active site,e.g., organophosphate inhibitors.

Almost the only feature common to all the Hill re-action inhibitors investigated is the presence of animino hydrogen. Moreover, there is good reason tobelieve that this imino hydrogen must not onlv bepresent but also accessible if a substance is to havesignificant inhibitory action. In ortho-chloroanilides,the bulky acyl groups almost certainly lie in the planeof the benzene ring and remote from the also bulkychlorine atom. Such an orientation would force theimino hydrogen into close proximity with the chlorineatom where it would be shielded from interactionswith adjacent molecules. Consequently ortho-chloro-anilides may be expected to, and do, resemble N-methylanilides in having low melting points, lowbiological activities, and no hydrogen bonding. Incontrast, the imino hydrogen of 2,6-dichloroanilineswould be forced into a position away from the planeof the benzene ring and midway between the twochlorines. In this position the imino hydrogenshould be, and apparently is, available for hydrogenbond formation and for whatever other functions maybe involved in producing the inhibition.

If one were permitted to select his data, it wouldbe possible to obtain a striking correlation betweenthe activity of inhibitors and the tendencies of theirimino hydrogen to form bonds with carbonyl oxygen

atoms. Substitutions on the benzene ring of theanilides have completely parallel effects on these twoproperties. For instance, the series 3,4 > 3- and4->2,4,6-> unsubstituted > 2- applies equally to thehydrogen bonding and the inhibitory action of thechloroanilides. Unfortunately, this beautiful correla-tion does not apply at all when the acyl moietv isvaried. For example, the biologically active chloro-acylanilides form hydrogen bonds most reluctantly,and the extremely potent phenyldimethylureas(DCMU, CMU, etc.) form hydrogen bonds much lessfreely than do the moderately active isobutyrylanilidesand the weakly active acetylanilides. Since the au-

thor is inclined to follow Wessels in believing that

hydrogen bonds should play an important role inanilide inhibitions, he finds these exceptions puzzling.Of course, in this study we may have presented theinhibitor molecules with inappropriate carbonyl ox-ygens; it may be that hydrogen bonding with anoxygen (or nitrogen?) in a very unusual situation isrequired if the catalytic site is to be blocked, and,therefore, the general bonding potential of the iminohydrogen may be of limited interest. The conversefact that many inactive or almost inactive substancesreadily form hydrogen bonds is irrelevant since ob-viously many requirements other than the capacity toform hydrogen bonds must be satisfied if a substanceis to be a strong inhibitor.

SITE OF ACTION & PHOTOSYNTHETIC UNIT. TheBishop-Jagendorf-Krall interpretation of phenylureainhibition seems plausible in the light of our presentknowledge of photosynthesis. Nevertheless, wewould do well to remember that our information con-cerning the mechanisms of photophosphorylation andoxygen production is practically non-existent. Con-sequently the phosphorylation measurement describedabove cannot establish either with precision or withcertainty the loci of the inhibitions. All that we cansay with confidence is this: The anilides (which in-clude phenylureas), the alkylamides, and the triazineshave similar modes of action and nothing that we nowknow is inconsistent with the hypothesis that themechanism for the oxidation of water to molecularoxygen is the process which is primarily affected.

Wessel's suggestion that the imino hydrogens ofphenylureas form hydrogen bonds with the carbonyloxygen of the cyclopentanone ring of chlorophyll(20), encounters two serious objections. In the firstplace such a mechanism of inhibition should equallyaffect photosynthesis, photoreduction, the Hill reac-tion with ferricyanide or FMN, and the Hill reactionwith PMS. As we have seen, this is not the case.In the second place, the better inhibitors are muchtoo effective. The number of moles of DCMU orKarsil required for nearly complete inhibition issmaller than the number of moles of chlorophyllpresent, by a factor of at least 100. Moreover, thisfactor of 100 does not allow for the fact that not allof the inhibitor can have reached the site of inhibition.Actually we have shown that at least two-thirds ofthe DCMU does not even leave the medium under theconditions of our experiments. (This was done bothby varying the amount of chloroplast material & bypreincubating the medium with chloroplasts whichwere subsequently removed & discarded). Since thechloroplasts do not remove all of the inhibitor fromthe medium it follows that the inhibitor-sensitive sitesdo not remove all of the inhibitor from the other partsof the chloroplasts. In fact, one would expect ahigher concentration of inhibitor in the body of thechloroplast than in the medium in view of the solubil-ity characteristics of DCMU and the lipoidal natureof the structural components of the plastids. Con-sequently, it must be concluded that an unknown butpossibly quite small proportion of the inhibitor reaches

801


PLANT PHYSIOLOGY

its target. This target cannot be chlorophyll unlesswe postulate one uniquely situated chlorophyll mole-cule among hundreds or perhaps among thousands.Since excitation energy certainly can be transferredfrom chlorophyll molecule to chlorophyll moleculemany times, there is no reason why there might notbe a fewzr special chlorophyll molecules associated withenzvnmes or substrates to act as a trap for this mobileexcitation energy, but it surely stretches credulity alsoto attribute unique bonding powers to the cyclopent-anone carbonyl of these particular molecules.

Nevertheless, the above considerations suggest a

possible relationship of the inhibited site to the photo-synthetic unit. On the basis of kinetic data [flashsaturationi of photosynthesis (5), etc.1 it has been

postulated that a number of chlorophyll moleculescollaborate in some way, funnelling their excitationenerg y or an interilie(liate chemical product of theexcitation energy throughI a common channel. Thisphotosynthetic unit may have a morphological basisor it mayihave a purelv statistical basis in nature;accordling to Rabinowitch (16) "the now most plauisi-ble interpretationi of these phenomena is in terms ofan enizymatic component ordinarily present in coni-

centrations 1/300 to 1/2400 that of chlorophyll".Thle iliteresting point is that the unit may consist ofabout 300 molecules of chlorophyll, and the numberof units is, therefore, of the same order as the numberof miiolecules of DCAIU whiclh we can reasonably ex-

pect to be at the inhlibite(d sites when inhibition ispractically complete. Unfortunately, the number ofphotosynthetic units is uilcertain and estimates ofthe liumber of effective inhibitor molecules are neces-

sarily very crude. Therefore, there is a high prob-ability that the correspondence of the numbers isaltogether fortuitous.

HERBICIDAL ACTIVITY & MECHANISM OF KILLING.\We are inow experimenting to determine the effective-ness of these compounds as inhibitors of the in vivoprocess of photosynthesis. We are employing excisedbean leaves according to the method of Minshall (14).Our preliminary results indicate that the effectivenessof the chemicals in the intact organ is only roughlvcorrelated with in vitro potency. Thus, the thioureasare much too active in vivo and the better acylanilidesare not nearly active enough. On the basis of anal-ogy to the well known conversion of P = S to P = 0

in organophosphate metabolism, we suggest that thethioureas may be fairly rapidly transformed into theiroxvgen analogs in the plant. The unexpected lowtoxicity of the acylanilides could be explained in atleast two ways: These are generally less polar sub-stances anld they may be preferentially absorbed bylipids, never reaching the sensitive site; alternativelythe acylanilides may be hydrolysed enzymatically.If the activations of the thioureas and the inactivationsof the acylanilides are indeed enzymic reactions, andif the enzymes involved are not universally dis-tributed, there would seem to be some hope of dis-covering among the thioureas and acylanilides, herbi-cides with appreciable species specificity.

The herbicidal activity of the lihenylcarbaiiiatesis not completely understood. It has long been knownthat these substances interfere with cell division,especially in grasses (4), and it was taken for grantedthat the plants were thereby killed. However, More-land and Hill (11) have shown recently that nmost ofthe herbicidal phenylcarbamates also inhibit the Hillreaction and therefore they may inhibit photosyn-thesis. In fact both mechanisms may be operativeand both may contribute to the death of the plantbut the author is inclined to favor the original hy-pothesis for several reasons. Of these, the mostconmpelling is the fact that isopropyl(2-chlorophenyl)-carbamate and isopropyl (2-methoxyphenol) carbamateare typical carbamate herbicides (17). From ourwork and the work of Moreland and( Hill, it seemsmost improbable that suclh ortho-suibstituted anilidescotld(l be effective inhibitor.s of photosynthesis.

The manner in which the phenyluireas. acylanlilidle,an(l triazine herbicides kill plants is niot as olbinlis asone mighlt think. Naturally. inter-ference w\ith thephotosynthetic apparatus mlust ultimately lead to (leatlhof the plant tlhrough starvation. However, there isevidence that a mlor-e active pr-ocess is involved sincethe phenylureas are toxic to suigar fed algae in thelight but not in the (lark (7). Since in this case therecan be I10 question of stairation, it is presunle(l thatdeleterious oxidation processes accompany the ac-cumulation of anl oxidize(d photo-product, the photo-prod(uct which vould be recdtuce(l, in the absence ofinhibition, by water.

SUM\ MNARY

I. A large number of substanices having the gen-eral formula R-NH-CX-R' were obtainied comnmercial-ly or prepared and identified. These included acyl-anili(les, acylamides, thioacvlanilidles, phenylureas,alkvlureas, and phenylthioureas.

II. A study was made of the effectiveness withwhich these and miscellaneous other related com-pounds inhibit the reduction of ferricvanide by il-luminated chloroplasts. A similar study was madeof the inhibitory activity of several lherbicidal tria-zines. The observations made possible certaini gen-eralizations:

A. Substitution of the hydrogen on the 3-, 4- or5- positions of the benzene ring of the aniline deriva-tives by Cl, Br, CHO, or CH3 increases the inhibitoryeffect. Substitution of the ortho position practicallyabolishes inhibition. Other parts of the moleculebeing equal, the inhibitory activity of the chloroanil-ides follows this sequence: 3,4-> 3,5- and 3- and4- > 2,4,6- > unsubstituted and 2,4,5- > 2,5- and 2,3> 2-. Para- and meta-nitro groups reduce activityslightly and the p-dimethylamino group reduces ac-tivity sharply.

B. The effects of modifying the acyl moiety ofthe anilides are too complex to classify. Steric con-siderations seem to be paramount since the chloroacids and their methyl analogs yield ainilides of com-parable potency. Polar groups reduice activity.

802



C. Chloracetyl-N-methylaniline, an anilide whichlacks an imino hydrogen atom, is scarcely inhibitory.

D. Among the triazines, the substances with twoimino hydrogens are more inhibitory than those withone.

III. The acylanilides. phenylureas, and triazines,like CMU, inhibit chloroplast reactions when ferri-cyanide or FMN is the electron acceptor but not whenPMS is the acceptor. This probably means that themechanism inhibited by all of these chemicals is themechanism normally responsible for the oxidation ofwater to molecular oxygen.

IV. Attempts to relate inhibitor potency to thebonding tendency of the imino hydrogen were onlypartially successful; the role of hydrogen bonding inthe mechanism of inhibition remains uncertain. Sub-stitutions on the aniline moiety which favor hydrogenbond formation also increase the effectiveness of theinhibitor. However, modification of the acyl moietydid not reveal a similar correlation; the imino hvdro-gens of chloracetyl- and trichloracetyl-3,4-dichloro-aniline form bonds with carbonyl oxygens to a verylimited extent although both substances are excellentinhibitors. The same is true of the triazines.

ACKNOWLEDGMENTS

This investigation would have been impossiblewithout the able technical assistance of Mr. S. T.Bajura; I am much indebted to Mrs. Dorle Bongartfor the numerous analyses. I also wish to expressmy gratitude to Dr. G. D. Tlhorn for his advice.

The 3,4-dichloranilide of a-methvlvaleric acid(Karsil) was kindly supplied by the Niagara Chem-ical Division of the Food Machinery and ChemicalCorp. The triazines were contributed by the GeigyChemical Corp.

IITERATURE CITED

1. BACHMANN, W. E. & W. S. STRUVE. 1947. TheArndt-Eistert Synthesis. In: Organic Reactions,Vol. I, pp 38-62. John Wiley & Sons, Inc., NewYork.

2. BEAVER, D. J., D. P. ROMAN, & P. J. STOFFEL. 1957.The preparation & bacteriostatic activity of sub-stituted ureas. J. Am. Chem. Soc. 79: 1236-1245.

3. BISHop, N. I. 1958. The influence of the herbicide,DCMU, on the oxygen-evolving system of photo-synthesis. Biochim. Biophys. Acta 27: 205-206.

4. ENNIS, W. B., JR. 1949. Histological & cytologicalresponses of certain plants to some aryl carbamicesters. Am. J. Botany 36: 823.

5. EMERSON, R. & W. ARNOLD. 1932. The photo-chemical reactions in photosynthesis. J. Gen.Physiol. 16: 191-205.

6. GOOD, N. E. 1960. Activation of the Hill reactionby amines. Biochim. Biophys. Acta 40: 502-517.

7. HOFFMAN, C. E., R. T. HERSH, P. B. SWEETSER, &C. W. TODD. 1960. The effect of urea herbicideson photosynthesis. Proceedings 40th Ann. MeetingNE Weed Control Conf. Pp. 16-18.

8. JAGENDORF, A. T. 1958. The relationship betweenelectron transport & phosphorylation in spinachchloroplasts. In: The Photochemical ApparatusIts Structure & Function, R. C. Fuller, J. A. Berg-eron, L. G. Augenstine, M. E. Koshland, & H. J.Curtis, eds.

9. KLINGSBERG, E. & D. PAPA. 1951. Thiation withphosphorus pentasulfide in pyridine solution. J.Am. Chem. Soc. 73- 4988-4989.

10. KRALL, A. R., N. E. GOOD, & B. C. MAYNE. 1961.Cyclic & non-cyclic p)hotophosphorylation in chloro-plasts distinguished bv use of labeled oxygen.Plant Physiol. 36: 44X47.

11. MORELAND, D. E. & K. L. HILL. 1959. The actionof alkyl N-phenvlcarbamates on the photolytic ac-tivity of isolated chloroplasts. J. Agr. Food Chem.7: 832-837.

12. 'MORELAND, D. E. & K. L. HiLL. 1960. Inlhibitionof the photochemical activity of isolated chloroplastsby ring-chlorinated-N-ohenyl-2-methylpentanamides.Abstr. Meeting Weed Soc. America. Pp. 41-42.

13. MORELAND, D. E, W. A. GENTNER, J. L. HILTON,& K. L. HILL. 1959. Studies on the mechanism ofherbicidal action of 2-chloro-4,6-bis (ethylamino) -s-triazine. Plant Physiol. 34: 432-435.

14. MINSHALL, W. H. 1960. Effect of 3-(4-chloro-phenyl)-1,1-dimethylurea on dry matter production,transpiration, & root extension. Can. J. Botany38: 201-216.

15. NIELSEN, S. 0. & A. L. LEHNINGER. 1955. Phos-phorylation coupled to the oxidation of ferrocyto-chrome c. J. Biol. Chem. 215: 555-570.

16. RABINOWITCH, E. I. 1956. In: Photosynthesis &Related Processes. Vol. II, p. 1298. IntersciencePublishers Inc., New York.

17. SHAW, W. S. & C. R. SWANSON. 1953. The rela-tion of structural configuration to the herbicidalproperties & phytotoxicity of several carbamates& other chemicals. Weeds 2: 43-65.

18. WARBURG, 0. 1919. On the rate of photochemicaldecomposition of carbon dioxide in living cells.Biochem. Z. 100: 230-270.

19. WERNER, E. A. 1920. The constitution of carba-mides. XI. The mechanism of synthesis of ureafrom ammonium carbamate. The preparation ofcertain mixed trisubstituted carbamates & dithio-carbamates. J. Chem. Soc. 117: 1046-1053.

20. WESSELS, J. S. C. & R. VAN DER VEEN. 1956. Theaction of some derivatives of phenylurethan & of3-phenyl-1,1-dimethylurea on the Hill reaction.Biochim. Biophys. Acta 19: 548-549.

803


Documents

Cl - Plant Physiology · Since photoreduction differs from photosynthesis in that hydrogen gas replaces water as the ultimate electron donor, it seems reasonable to suppose that DCMUinterferes