Pyrazolidine-3,5-dione-based inhibitors ofphosphoenolpyruvate carboxylase as a new class ofpotential C4 plant herbicidesMarkus Dick1, German Erlenkamp1, Giang T.T. Nguyen2, Kerstin F€orster2, Georg Groth2 andHolger Gohlke1

1 Institute of Pharmaceutical and Medicinal Chemistry and Bioeconomy Science Center (BioSC), Heinrich-Heine-Universit€at D€usseldorf,

Germany

2 Institute of Biochemical Plant Physiology and Bioeconomy Science Center (BioSC), Heinrich-Heine-Universit€at D€usseldorf, Germany

Correspondence

H. Gohlke, Institute of Pharmaceutical and

Medicinal Chemistry and Bioeconomy

Science Center (BioSC), Heinrich-Heine-

Universit€at D€usseldorf, 40225 D€usseldorf,

Germany

Fax: +49(0)221-81-13847

Tel: +49(0)221-81-13662

E-mail: gohlke@uni-duesseldorf.de

(Received 28 July 2017, revised 1

September 2017, accepted 5 September

2017, available online 24 September 2017)

doi:10.1002/1873-3468.12842

Edited by Richard Cogdell

Phosphoenolpyruvate carboxylase (PEPC) is a key enzyme in the C4 photosyn-

thetic pathway of many of the world’s worst weeds and a valuable target to

develop C4 plant-selective herbicides. By virtual screening, analog synthesis,

and in vitro validation, we identified pyrazolidine-3,5-diones as a new class of

small molecules with inhibitory potential down to the submicromolar range

against C4 PEPC and a selectivity factor of up to 16 over C3 PEPC. No other

biological activity has yet been reported for the best compound, (3-bromophe-

nyl)-4-(3-hydroxybenzylidene)-pyrazolidine-3,5-dione. A systematic variation

in the substituents allowed the derivation of a qualitative structure–activity rela-

tionship. These findings make this compound class highly interesting for further

investigations toward generating potent, C4 plant-selective herbicides with a

low potential for unwanted effects.

Keywords: C4 photosynthesis; phosphoenolpyruvate; pyrazolidine-3, 5-diones;

selective C4 herbicides; virtual screening

The development of herbicides against weeds has

become a global challenge in agriculture as weeds com-

pete with crops for water, sunlight, and soil nutrients.

Additionally, the rapid development of resistances

against established herbicides demands to develop

bioactive compounds for new targets. For this, one

can exploit that many of the worst weeds are C4

plants, whereas the majority of crops use C3 photosyn-

thesis [1,2]. Hence, one should be able to identify C4

plant-selective herbicides by targeting enzymes of such

plants. In C4 photosynthesis, initially, inorganic car-

bon is fixed by the enzyme phosphoenolpyruvate

(PEP) carboxylase (PEPC). PEPC catalyzes the car-

boxylation of PEP to the four-carbon molecule

oxaloacetate [3], which is reduced to aspartate or

malate, or transaminated to asparagine [4]. Structural

and biochemical studies have revealed that residue

R884 of PEPC of the C3 plant Flaveria pringlei assists

in binding malate; malate then acts as a feedback inhi-

bitor. R884 is conserved in all typical C3 crop plants

(Fig. S1). In contrast, in C4 weeds, the homolog resi-

due often corresponds to glycine, serine, glutamate, or

glutamine, besides arginine (Fig. S2). The smaller resi-

dues have been linked with a weaker feedback inhibi-

tor-binding site [5], for example, G884 in PEPC of the

C4 plant Flaveria trinervia does not show any interac-

tion with the inhibitor [6]. Recently, we used the

molecular difference in the feedback inhibitor-binding

site of PEPC between C3 and C4 plants to develop

potential C4 plant-selective herbicides with a novel

mode of action [7,8]. These studies verified that the

identified chalcones, quinoxalines, and catechins are

Abbreviations

PEP, phosphoenolpyruvate; PEPC, phosphoenolpyruvate carboxylase; RMSD, root mean square deviation.

3369FEBS Letters 591 (2017) 3369–3377 ª 2017 Federation of European Biochemical Societies

potent inhibitors of C4 PEPC (low micromolar to sub-

micromolar range) and show selectivity factors over C3

PEPC up to 27. However, chalcones are known to

have a wide range of biological activities [9] such as

antibacterial [10], antioxidant [11], antifungal [12], and

anticancer [13] effects. Quinoxalines also display a

broad spectrum of biological activities [8], and cate-

chins are known to be antioxidants [14] and antifungal

molecules [15]. Hence, identifying alternative molecule

classes that serve as unique inhibitors for C4 PEPC

but show no/less effects on other targets remains an

important task. Here, we identified, by repetitive

rounds of ligand- and structure-based virtual screening

on the ZINC database, analog synthesis, and in vitro

validation, pyrazolidine-3,5-dione-based selective C4

PEPC inhibitors with a so far unexplored chemotype

for C4 PEPC inhibition. These inhibitors are compara-

ble in affinity and selectivity to the previous inhibitors

but, to our knowledge, no other biological activity has

yet been reported for them.

Materials and methods

Protein and ligand structure preparation

The crystal structures of PEPC of the C4 plant F. trinervia

(PDB ID 3ZGE) and the C3 plant F. pringlei (PDB ID

3ZGB) [16] were retrieved from the Protein Data Bank [17].

Both structures were preprocessed with the Protein Prepara-

tion Wizard [18] of Schr€odinger’s Maestro suite. Thereby,

bond orders were assigned, hydrogen was added, the H-

bond network was optimized, and missing side chains were

added using Prime [19,20]. Finally, the systems were energy

minimized using the OPLS 2005 force field, resulting in a

root mean square deviation (RMSD) of 0.3 �A with respect

to chain A of the crystal structure [21]. In both crystal struc-

tures, PEPC forms a homo 4-mer. Since the allosteric-bind-

ing site is not affected by the presence of the other

monomer, we only used one monomer for further analysis

[22,23]. 3D structures of ligands were generated with the

LigPrep [24] module of the Schr€odinger suite.

Virtual screening with ROCS and FRED

The ZINC database [25,26] was initially screened for inhibi-

tor candidates applying rules that describe favorable phy-

sico-chemical properties of herbicides [27]. The program

Omega [28] was used to generate up to 100 conformations of

each remaining compound. These conformations were

screened with the combined shape query created from previ-

ously identified active compounds [8] using ROCS [29]. The

default settings of ROCS were used for this screening step,

and only the best fitting conformation for each compound

was saved. Hits from this step were then docked by FRED

[30–33] into C3 and C4 PEPC. For this, the box was centered

on the cocrystallized aspartate of PDB ID 3ZGE, and the

dimensions of the box were set to 18.00, 14.67, and 17.00 �A.

Docking with AutoDock

AUTODOCK 3.0 [34] in combination with DrugScore [35] was

utilized to identify common binding modes of hits using

the Lamarckian genetic algorithm with default parameters.

The box was again centered on the cocrystallized aspartate.

The dimension of the box was set to 60 �A in each direc-

tion. Hundred independent docking runs were performed

for each compound, with a maximum of 500 000 energy

evaluations. A clustering RMSD cutoff of 2.0 �A was cho-

sen, and the docking solution with the lowest energy was

taken if the configuration was found within a cluster cover-

ing more than 20% of all docking solutions.

Synthesis of new pyrazolidine-3,5-diones

All compound syntheses were performed by Taros Chemi-

cals, Dortmund. The syntheses were performed according

to Koo et al. [36]. Analytic data of the final products are

provided as Data S1.

Chemicals for in vitro assays

Chemicals were acquired from Sigma-Aldrich (St Louis,

USA) if not stated otherwise.

Cloning, protein expression, and purification

His-tagged recombinant PEPC of C4 plant F. trinervia and

C3 plant F. pringlei cloned in bacterial expression vector

pETEV16b (Novagene, Darmstadt, Germany) were

expressed in Escherichia coli and purified by immobilized-

metal affinity chromatography (IMAC) on Ni2+-nitrilotria-

cetic acid agarose (GE Healthcare, Munchen, Germany)

according to the protocol described in Paulus et al. [6]. Pur-

ity of recombinant proteins was confirmed by SDS/PAGE

and Colloidal Coomassie Staining. Protein concentrations

were determined by UV absorption at 280 nm on a Tecan

Infinite M-200 (Tecan Group Ltd., M€annedorf, Switzerland)

using a molar absorptivity value of 119 930 M�1�cm�1 for

F. trinervia PEPC or 125 430 M�1�cm�1 for F. pringlei

PEPC, respectively. Purified proteins were concentrated to

5–15 mg�mL�1 and stored at �80 °C, respectively.

PEPC-coupled spectrophotometric assay

Inhibition of purified recombinant PEPCs by pyrazolidine-

3,5-diones was measured in a coupled spectrophotometric

assay at 25 °C in a Beckman DU-800 spectrophotometer as

described previously [37]. Reaction mixture contained

3370 FEBS Letters 591 (2017) 3369–3377 ª 2017 Federation of European Biochemical Societies

Pyrazolidine-3,5-diones as C4 plant herbicides M. Dick et al.

purified C3 or C4 PEPC (0.05 U), NADH (150 lM), malate

dehydrogenase (2 U), and the pyrazolidine-3,5-diones stud-

ied in 50 mM HEPES/KOH pH 7.5, 10 mM MgCl2, 10 mM

KHCO3. Catalytic turnover was started by adding PEP at

concentrations corresponding to two-fold Km of the recom-

binant PEPC. The IC50 of selected pyrazolidine-3,5-diones

was determined by varying the inhibitor concentration from

0.01 to 200 lM in the coupled spectrophotometric assay.

Data were analyzed using GRAPHPAD Prism (GraphPad Soft-

ware, Inc., La Jolla, CA, USA).

Results

We performed a ligand-based virtual screening on a

subset of 15 000 molecules of the ZINC database

filtered such that they comply with rules for favorable

physico-chemical properties of herbicides [27] (see

Fig. 1A). Our search query was built based on two

types of selective C4 PEPC inhibitors, catechins and

quinoxalines, identified previously by us [8]. To do so,

we used Rapid Overlay of Chemical Structures (ROCS

[29]) to create a 3D overlay considering molecular

shape and acceptor, donor, cationic, anionic, aromatic,

and hydrophobic properties of these active compounds

(Fig. 1B). Applying this query with ROCS to the gen-

erated conformations of the 15 000 molecules yielded

7000 candidates fulfilling the characteristics of the 3D

ROCS overlay. The candidates were then docked with

FRED [30–33] into the feedback inhibitor-binding

Fig. 1. Workflow for structure-based screening for PEPC inhibitors. (A) Based on 20 million compounds from the ZINC database and two

filtering steps, 7000 compounds were docked into the binding pocket of PEPC. Out of 700 hits, 10 were used for in vitro tests resulting in

one active compound. In a second run, 15 000 compounds that fulfill the rules for bioavailability were screened for ligands with a

pyrazolidine-3,5-dione core, and 20 hits were tested. In addition, ten potential inhibitors that only vary in ring B (compared to the initial hit)

were designed, synthesized, and tested. (B) Shape- and properties-based query for a ROCS search based on an overlay of known inhibitors

of C4 PEPC [8]. Colored spheres represent ring structures (green), acceptor (blue), or donor atoms (red).

3371FEBS Letters 591 (2017) 3369–3377 ª 2017 Federation of European Biochemical Societies

M. Dick et al. Pyrazolidine-3,5-diones as C4 plant herbicides

pocket of C4 PEPC from F. trinervia. From the first

700 best scored molecules, 10 were selected by visual

inspection for in vitro activity tests (Fig. 1A). One

active compound with an IC50 value of 2.79 lM for C4

PEPC and 13.98 lM for C3 PEPC (resulting in a selec-

tivity factor of 5.01) was identified (compound 1–1,Table 1, Fig. S3). It contains a pyrazolidine-3,5-dione

core with a phenyl substituent at position 2 and a ben-

zylidene one at position 4 (referred to as ring A and

ring B hereafter).

Based on this initial hit, we performed a similarity

search (2D using ROCS) to retrieve all pyrazolidine-

3,5-diones from the 15 000 compounds. Twenty-two

compounds were chosen for a second round of in vitro

Table 1. Inhibitory effect of pyrazolidine-3,5-diones on PEPC from Flaveria trinervia (C4 plant) and Flaveria pringlei (C3 plant).

Molecule IC50 values (lM)

Structure Vendor ID C4 C3 Ratio C3/C4

1–1a

NHN

O

O

Br

O

Amb3596945 2.79 � 0.04 13.98 � 0.03 5.01

2–1a

NHN

O

O

Br

Cl

O

Amb3455285 0.90 � 0.04 4.80 � 0.02 5.33

2–2

NHN

O

O

BrCl

O

Amb3455286 1.20 � 0.04 4.90 � 0.20 4.08

2–3

NHN

O

O

O Br Amb8583694 1.20 � 0.04 2.60 � 0.10 2.17

2–4

NHN

O

O

O

Br

Amb21992151 0.56 � 0.04 0.52 � 0.60 0.93

3–1

NHN

O

O

Br

O2N

Taro27483 24.09 � 2.14 103.07 � 6.81 4.28

3–2

NHN

O

O

BrO2N Taro27484 1.26 � 0.19 2.06 � 0.22 1.63

3–3

NHN

O

O

Br

O2N Taro27485 2.85 � 0.42 2.59 � 0.28 0.91

3372 FEBS Letters 591 (2017) 3369–3377 ª 2017 Federation of European Biochemical Societies

Pyrazolidine-3,5-diones as C4 plant herbicides M. Dick et al.

testing based on visual inspection. A first activity test

revealed that 10 of the 22 compounds had an inhibi-

tory effect (> 60%, Fig. S4). For the best four hits (in-

hibitory effect > 80%), the IC50 values were measured

for both C3 and C4 PEPC (Table 1, compounds 2–1 to

2–4; see also Fig. S4). All four molecules have IC50

values in the range of 0.56–4.9 lM and selectivity fac-

tors of 0.93–5.33. Compound 2–4 is the most potent

C4 PEPC inhibitor known to date but lacks selectivity.

Notably, of all five molecules identified so far, it is the

only one that does not have a bromide substitution at

ring A. Thus, in order to improve the selectivity, we

decided to perform a screen for differently substituted

aromatic rings B while keeping the 2-bromophenyl

substituent as ring A. To be not limited by the pool of

commercially available molecules, the compounds of

interest were newly synthetized (compounds 3–1 to 3–10 in Table 1). The best compound 3–5 had an IC50

value of 1.96 lM toward C4 PEPC, a selectivity factor

of 16.6, and a 2-hydroxyl-benzylidene substituent as

Table 1. (Continued).

Molecule IC50 values (lM)

Structure Vendor ID C4 C3 Ratio C3/C4

3–4

NHN

O

O

Br

HO

Taro27486 13.94 � 0.94 72.58 � 6.79 5.21

3–5a

NHN

O

O

BrHO Taro27487 1.96 � 0.17 32.59 � 1.94 16.63

3–6

NHN

O

O

Br

HO ChEMBLl1533094 18.99 � 2.12 72.53 � 2.22 3.82

3–7

NHN

O

O

Br

N

N Taro27489 39.57 � 9.23 131.39 � 47.83 3.32

3–8

NHN

O

O

Br

N

O ChEMBL1434855 63.95 � 14.97 149.00 � 28.29 2.33

3–9

NHN

O

O

Br

HO

OTaro27491 11.62 � 0.99 105.94 � 3.07 9.12

3–10

NHN

O

O

BrO

HO

Taro27492 11.46 � 1.53 144.98 � 68.24 12.65

a IC50 measurements for inhibition of C4 and C3 PEPC are shown as examples in Fig. S3.

3373FEBS Letters 591 (2017) 3369–3377 ª 2017 Federation of European Biochemical Societies

M. Dick et al. Pyrazolidine-3,5-diones as C4 plant herbicides

ring B. To investigate its binding mode, 3–5 was

docked into the binding sites of C3 and C4 PEPC

using Autodock with DrugScore potentials [38,39].

The predicted binding pose in C4 PEPC is shown in

Fig. 2. Ring B is located close to R641 and R888,

forming cation–p interactions and a hydrogen bond

between the 2-hydroxyl group and R641. A similar

interaction pattern has been described for chalcone-

based inhibitors of PEPC [7]. Ring A is buried in the

hydrophobic part of the binding pocket in between

I113, L881, and A132. The comparison with the dock-

ing result in C3 PEPC revealed a nearly identical bind-

ing mode. The observed selectivity, thus, cannot

simply be explained by a different shape matching.

Discussion

Phosphoenolpyruvate carboxylase is a crucial enzyme

for carbon fixation in photosynthesis. The structural

differences in PEPC of C3 and C4 plants make it an

attractive target for C4 plant-selective herbicides. In

this study, we used a repetitive ligand- and structure-

based virtual screening approach to identify and char-

acterize small molecules with a pyrazolidine-3,5-dione

core as selective inhibitors of C4 PEPC. Specifically,

we identified 2-(3-bromophenyl)-4-(3-hydroxybenzyli-

dene)-pyrazolidine-3,5-dione (compound 3–5) with an

IC50 value of 1.96 lM and a selectivity factor of 16.6

over C3 PEPC. The inhibitor furthermore complies

with rules for favorable physico-chemical properties of

herbicides [27]. Both the inhibitory potential and the

selectivity factor of 3–5 are comparable with those

identified for the most promising chalcone-based

inhibitors [7] and superior to those identified for

catechine-based inhibitors [8]. With respect to quinoxa-

line-based inhibitors [8], the inhibitory potential of 3–5is higher by a factor of at least 70. In addition, for the

identified pyrazolidine-3,5-diones, other biological

activities have rarely been mentioned so far: A similar-

ity search for compound 3–5 (Tanimoto similarity

score > 90%) using SciFinder [40] or ChEMBL [41]

did not reveal any literature entry related to a known

other biological activity (databases accessed in July

2017).

From our in vitro activity data and the predicted

binding mode, a qualitative structure–activity relation-

ship for the investigated pyrazolidine-3,5-diones can be

derived. Bulky and rigid substituents in ring B such as

4-morpholino have a negative effect on the inhibitory

potential (compounds 3–7, 3–8), which may be

rationalized by steric clashes as identified in the dock-

ing experiment. Interestingly, bulky but more flexible

residues such as benzyloxy (2–3) or phenylethyloxy

(2–4) lowered the IC50 to a submicromolar range.

However, this increase in inhibitory potential toward

C4 PEPC was accompanied by a loss in selectivity over

C3 PEPC. This loss is probably caused by a different

binding mode (shown for 2–4 in Fig. S5), where a new

Fig. 2. Predicted binding pose of compound 3–5 within C4 PEPC. The conformation with the lowest energy in the largest cluster is shown.

(A) 3D representation of the binding pocket. The bound inhibitor is represented in space-filling mode (carbon: green, nitrogen: blue, oxygen:

red, bromide: light red), while the residues of PEPC are shown as sticks (carbon: light blue, nitrogen: blue, oxygen: red, polar hydrogen:

white). (B) 2D scheme of the binding pose. Red arrows represent hydrogen bond formation. Protein residues are labeled in blue (pos.

charge), red (neg. charge), light blue (polar), or green (hydrophobic).

3374 FEBS Letters 591 (2017) 3369–3377 ª 2017 Federation of European Biochemical Societies

Pyrazolidine-3,5-diones as C4 plant herbicides M. Dick et al.

orientation of both ring A and ring B enable a cation–p interaction of the benzyloxy moiety with the nearby

R687. A 3-hydroxyl group at ring B yielded both a

good inhibitory potential and selectivity (3–5), likely

because it allows the formation of a hydrogen bond

with either R888 or N964, which appears unlikely to

form with hydroxyl groups in positions 2 or 4 due to a

larger distance to those amino acids (resulting in

higher IC50 values). In support of this hypothesis, a

similar effect was detected when the hydroxyl group

was exchanged by a nitro group (compounds 3–1 to 3–3) or ether moiety (3–9 and 3–10), which can function

as (weak) hydrogen bond acceptors [42].

Finally, as to selectivity-determining factors, the

binding pockets of our C3 and C4 PEPC differ only in

position 884 (arginine in C3 PEPC instead of glycine

in C4 PEPC). Multiple sequence alignments reveal that

amino acids are highly conserved in the feedback inhi-

bitor-binding pocket (Figs S1 and S2), with R884 in

C3 PEPC being fully conserved and glycine being the

most common amino acid in C4 PEPC at position 884

that is different from the arginine found in C3 PEPC.

Hence, the PEPC investigated here are good represen-

tatives of their families. As almost all our compounds

show a selectivity toward C4 PEPC, apparently, disfa-

vorable steric interactions of R884 in C3 PEPC with

ring B in the identified binding mode outweigh poten-

tially favorable cation–p or polar interactions.

As to in vivo experiments, only a preliminary study

has been carried out on the effect of the identified

pyrazolidine-3,5-dione inhibitors on C4 plants. No

obvious damage or growth-inhibitory effect was

observed with the C4 weed Amaranthus retroflexus

when sprayed at single dose at 3 mM concentration.

However, previous weed control and herbicide

research projects have already shown that herbicidal

activities critically depend on formulation and applica-

tion of the active compound [43]. Consequently, fur-

ther plant uptake studies will address if higher dose,

repetitive spraying, or the addition of surfactants can

promote foliar absorption and herbicidal activities of

pyrazolidine-3,5-diones.

In conclusion, we identified pyrazolidine-3,5-diones

(Fig. 3) as a new class of small molecules with inhibi-

tory potential in the low micromolar to submicromolar

range against C4 PEPC and a selectivity factor of up

to 16 with respect to binding to C3 PEPC. To the best

of our knowledge, no other biological activity has yet

been reported for our best compound, 3–5. A system-

atic variation of substituents at ring B allowed deriving

a qualitative structure–activity relationship. Together,

these findings make this compound class highly inter-

esting for further investigations toward generating

potent, C4 plant-selective herbicides with a low poten-

tial for unwanted effects.

Acknowledgements

This study was funded within the Bioeconomy Science

Center (BioSC) by the boost fund ‘C4-PSH’ granted to

HG and GG. The scientific activities of BioSC were

financially supported by the Ministry of Innovation,

Science, and Research within the framework of the

NRW Strategieprojekt BioSC (No. 313/323-400-002

13). We are grateful to the ‘Zentrum f€ur Informations-

und Medientechnologie (ZIM)’ at Heinrich Heine

University D€usseldorf for providing computational

support.

Authors contribution

HG and GG conceived and supervised the study; MD

and GE performed in silico experiments; GN and KF

performed in vitro tests; MD and HG wrote the manu-

script; GG, GN, KF, and GE made manuscript revi-

sions.

References

1 May RM (1981) The world’s worst weeds. Nature 290,

85–86.2 Tremmel DC and Patterson DT (1993) Responses of

soybean and five weeds to CO2 enrichment under two

temperature regimes. Can J Plant Sci 73, 1249–1260.3 Lepiniec L, Thomas M and Vidal J (2003) From

enzyme activity to plant biotechnology: 30 years of

research on phosphoenolpyruvate carboxylase. Plant

Physiol Biochem 41, 533–539.4 Sage RF (2004) The evolution of C4 photosynthesis.

New Phytol 161, 341–370.5 Paulus JK, Niehus C and Groth G (2013) Evolution of

C4 phosphoenolpyruvate carboxylase: enhanced

feedback inhibitor tolerance is determined by a single

residue. Mol Plant 6, 1996–1999.6 Paulus JK, Schlieper D and Groth G (2013) Greater

efficiency of photosynthetic carbon fixation due to

single amino-acid substitution. Nat Commun 4, 1518.

NHN

O

O

R2R1

R2 = H, BrR1 = NO2, OH, aryl, oxy,and other ring moieties

Fig. 3. Core structure of pyrazolidine-3,5-dione and substituents

that have been identified to yield potent inhibitors of C4 PEPC.

3375FEBS Letters 591 (2017) 3369–3377 ª 2017 Federation of European Biochemical Societies

M. Dick et al. Pyrazolidine-3,5-diones as C4 plant herbicides

7 Nguyen GT, Erlenkamp G, Jack O, Kuberl A, Bott M,

Fiorani F, Gohlke H and Groth G (2016) Chalcone-

based selective inhibitors of a C4 plant key enzyme as

novel potential herbicides. Sci Rep 6, 27333.

8 Paulus JK, Forster K and Groth G (2014) Direct and

selective small-molecule inhibition of photosynthetic

PEP carboxylase: New approach to combat C4 weeds in

arable crops. FEBS Lett 588, 2101–2106.9 Mahapatra DK, Asati V and Bharti SK (2015)

Chalcones and their therapeutic targets for the

management of diabetes: structural and

pharmacological perspectives. J Med Chem 92, 839–865.10 Abdullah MI, Mahmood A, Madni M, Masood S and

Kashif M (2014) Synthesis, characterization, theoretical,

anti-bacterial and molecular docking studies of

quinoline based chalcones as a DNA gyrase inhibitor.

Bioorg Chem 54, 31–37.11 Aoki N, Muko M, Ohta E and Ohta S (2008) C-

geranylated chalcones from the stems of Angelica

keiskei with superoxide-scavenging activity. J Nat Prod

71, 1308–1310.12 Lahtchev KL, Batovska DI, Parushev SP, Ubiyvovk

VM and Sibirny AA (2008) Antifungal activity of

chalcones: a mechanistic study using various yeast

strains. Eur J Med Chem 43, 2220–2228.13 Modzelewska A, Pettit C, Achanta G, Davidson NE,

Huang P and Khan SR (2006) Anticancer activities of

novel chalcone and bis-chalcone derivatives. Bioorg

Med Chem 14, 3491–3495.14 Morel I, Lescoat G, Cogrel P, Sergent O, Pasdeloup N,

Brissot P, Cillard P and Cillard J (1993) Antioxidant

and iron-chelating activities of the flavonoids catechin,

quercetin and diosmetin on iron-loaded rat hepatocyte

cultures. Biochem Pharmacol 45, 13–19.15 Hirasawa M and Takada K (2004) Multiple effects of

green tea catechin on the antifungal activity of

antimycotics against Candida albicans. J Antimicrob

Chemother 53, 225–229.16 Paulus JK, Niehus C and Groth G (2013) Evolution of

C4 phosphoenolpyruvate carboxylase: enhanced

feedback inhibitor tolerance is determined by a single

residue. Mol Plant 6, 1996–1999.17 Berman HM, Westbrook J, Feng Z, Gilliland G,

Bhat TN, Weissig H, Shindyalov IN and Bourne PE

(2000) The protein data bank. Nucleic Acids Res 28,

235–242.18 Schr€odinger Suite 2014-1, Protein Preparation Wizard.

Schr€odinger, LLC, New York (2014).

19 Jacobson MP, Friesner RA, Xiang Z and Honig B

(2002) On the role of the crystal environment in

determining protein side-chain conformations. J Mol

Biol 320, 597–608.20 Jacobson MP, Pincus DL, Rapp CS, Day TJ, Honig B,

Shaw DE and Friesner RA (2004) A hierarchical

approach to all-atom protein loop prediction. Proteins

55, 351–367.21 Jorgensen WL and Tirado-Rives J (1988) The OPLS

potential functions for proteins. Energy minimizations

for crystals of cyclic peptides and crambin. J Am Chem

Soc 110, 1657–1666.22 Matsumura H, Xie Y, Shirakata S, Inoue T, Yoshinaga

T, Ueno Y, Izui K and Kai Y (2002) Crystal structures

of C4 form maize and quaternary complex of E. coli

phosphoenolpyruvate carboxylases. Structure 10, 1721–1730.

23 Muramatsu M, Suzuki R, Yamazaki T and Miyao M

(2015) Comparison of plant-type phosphoenolpyruvate

carboxylases from rice: identification of two plant-

specific regulatory regions of the allosteric enzyme.

Plant Cell Physiol 56, 468–480.24 Schr€odinger Suite 2014-1, LigPrep. Schr€odinger, LLC,

New York, NY (2014).

25 Irwin JJ and Shoichet BK (2005) ZINC - a free

database of commercially available compounds for

virtual screening. J Chem Inf Model 45, 177–182.26 Irwin JJ, Sterling T, Mysinger MM, Bolstad ES and

Coleman RG (2012) ZINC: a free tool to discover

chemistry for biology. J Chem Inf Model 52, 1757–1768.

27 Tice CM (2001) Selecting the right compounds for

screening: does Lipinski’s Rule of 5 for pharmaceuticals

apply to agrochemicals? Pest Manag Sci 57, 3–16.28 Hawkins PCD, Skillman AG, Warren GL, Ellingson

BA and Stahl MT (2010) Conformer generation with

OMEGA: algorithm and validation using high quality

structures from the protein databank and Cambridge

structural database. J Chem Inf Model 50, 572–584.29 Hawkins PCD, Skillman AG and Nicholls A (2007)

Comparison of shape-matching and docking as virtual

screening tools. J Med Chem 50, 72–82.30 McGann M (2012) FRED and HYBRID docking

performance on standardized datasets. J Comput Aided

Mol Des 26, 897–906.31 McGann M (2011) FRED pose prediction and virtual

screening accuracy. J Chem Inf Model 51, 578–596.32 McGann MR, Almond HR, Nicholls A, Grant JA and

Brown FK (2003) Gaussian docking functions.

Biopolymers 68, 76–90.33 McGaughey GB, Sheridan RP, Bayly CI, Culberson

JC, Kreatsoulas C, Lindsley S, Maiorov V, Truchon JF

and Cornell WD (2007) Comparison of topological,

shape, and docking methods in virtual screening. J

Chem Inf Model 47, 1504–1519.34 Morris GM, Goodsell DS, Halliday RS, Huey R, Hart

WE, Belew RK and Olson AJ (1998) Automated

docking using a Lamarckian genetic algorithm and an

empirical binding free energy function. J Comput Chem

19, 1639–1662.

3376 FEBS Letters 591 (2017) 3369–3377 ª 2017 Federation of European Biochemical Societies

Pyrazolidine-3,5-diones as C4 plant herbicides M. Dick et al.

35 Velec HF, Gohlke H and Klebe G (2005) DrugScore

(CSD)-knowledge-based scoring function derived from

small molecule crystal data with superior recognition

rate of near-native ligand poses and better affinity

prediction. J Med Chem 48, 6296–6303.36 Koo KA, Kim ND, Chon YS, Jung M-S, Lee B-J, Kim

JH and Song W-J (2009) QSAR analysis of

pyrazolidine-3,5-diones derivatives as Dyrk1A

inhibitors. Bioorg Med Chem Lett 19, 2324–2328.37 Jacobs B, Engelmann S, Westhoff P and Gowik UDO

(2008) Evolution of C4 phosphoenolpyruvate

carboxylase in Flaveria: determinants for high tolerance

towards the inhibitor L-malate. Plant, Cell Environ 31,

793–803.38 Gohlke H, Hendlich M and Klebe G (2000)

Knowledge-based scoring function to predict protein-

ligand interactions. J Mol Biol 295, 337–356.39 Sotriffer CA, Gohlke H and Klebe G (2002) Docking

into knowledge-based potential fields: a comparative

evaluation of DrugScore. J Med Chem 45, 1967–1970.40 SciFinder Scholar, version 2004.2. Columbus, OH

(2004).

41 Bento AP, Gaulton A, Hersey A, Bellis LJ, Chambers

J, Davies M, Kruger FA, Light Y, Mak L, McGlinchey

S et al. (2014) The ChEMBL bioactivity database: an

update. Nucleic Acids Res 42, D1083–D1090.

42 Laurence C and Berthelot M (2000) Perspect Drug

Discovery Des 18, 39–60.

43 Liu Z (2004) Effects of surfactants on foliar uptake of

herbicides - a complex scenario. Colloids Surf B

Biointerfaces 35, 149–153.

Supporting information

Additional Supporting Information may be found

online in the supporting information tab for this

article:Fig. S1. C3 PEPC sequences according to Paulus et al.

[1] aligned with the C3 PEPC from Flaveria pringlei

Fig. S2. C4 PEPC sequences according to Paulus et al.

[1] aligned with the C4 PEPC from Flaveria triniervia.

Fig. S3. Assay for C4 and C3 PEPC inhibition by dif-

ferent pyrazolidine-3,5-dione-based inhibitors (A to D:

compounds 1–1, 2–1, 2–4 and 3–5).Fig. S4. Activity assay to determine the inhibitory

effect of selected compounds (second screening round

of Fig. 1) on the catalytic activity of PEPC from

Flaveria trinervia.

Fig. S5. 2D schemes of the predicted binding modes of

(A) compound 2–4 (selectivity < 1) and (B) 3–5 (selec-

tivity 16.6) within PEPC from Flaveria trinervia (C4

plant).

Data S1. Analytic data for compounds 3–1 to 3–10provided by Taros Chemicals, Dortmund.

3377FEBS Letters 591 (2017) 3369–3377 ª 2017 Federation of European Biochemical Societies

M. Dick et al. Pyrazolidine-3,5-diones as C4 plant herbicides