Glutathione reductase-catalyzed cascade of redox reactions to bioactivate potent antimalarial 1,4-naphthoquinones - A new strategy to combat malarial parasites

ARTICLE

pubs.acs.org/JACS

Glutathione Reductase-Catalyzed Cascade of Redox Reactions To

Bioactivate Potent Antimalarial 1,4-Naphthoquinones ꢀ A New

Strategy to Combat Malarial Parasites

Tobias M€uller,^†,#Laure Johann,^†,‡,#Beate Jannack,^†Margit Br€uckner,^†Don Antoine Lanfranchi,^‡

Holger Bauer,^†Cecilia Sanchez,^§Vanessa Yardley,^||Christiane Deregnaucourt,^{^}Joseph Schrꢀevel,^{^}

Michael Lanzer,^§R. Heiner Schirmer,^†and Elisabeth Davioud-Charvet*^,†,‡

^†Biochemie-Zentrum der Universit€at Heidelberg, Im Neuenheimer Feld 328, D-69120 Heidelberg, Germany

^‡European School of Chemistry, Polymers and Materials (ECPM), University of Strasbourg, UMR CNRS 7509, 25,

rue Becquerel, F-67087 Strasbourg, France

^§Department of Infectiology, University of Heidelberg, Im Neuenheimer Feld 324, 69120 Heidelberg, Germany

Department of Infections and Tropical Diseases, London School of Hygiene & Tropical Medicine, Keppel Street,

London WC1E 7HT, United Kingdom

^{^}Musꢀeum National d’Histoire Naturelle, FRE 3206 CNRS, BP 52, 61 rue Buﬀon, 75231 Paris cedex 05, France

S Supporting Information

b

ABSTRACT: Our work on targeting redox equilibria of malarial

parasites propagating in red blood cells has led to the selection

of six 1,4-naphthoquinones, which are active at nanomolar

concentrations against the human pathogen Plasmodium falci-

parum in culture and against Plasmodium berghei in infected

mice. With respect to safety, the compounds do not trigger

hemolysis or other signs of toxicity in mice. Concerning the

antimalarial mode of action, we propose that the lead benzyl

naphthoquinones are initially oxidized at the benzylic chain to

benzoyl naphthoquinones in a heme-catalyzed reaction within the digestive acidic vesicles of the parasite. The major putative

benzoyl metabolites were then found to function as redox cyclers: (i) in their oxidized form, the benzoyl metabolites are reduced by

NADPH in glutathione reductase-catalyzed reactions within the cytosols of infected red blood cells; (ii) in their reduced forms, these

benzoyl metabolites can convert methemoglobin, the major nutrient of the parasite, to indigestible hemoglobin. Studies on a

ﬂuorinated suicide-substrate indicate as well that the glutathione reductase-catalyzed bioactivation of naphthoquinones is essential

for the observed antimalarial activity. In conclusion, the antimalarial naphthoquinones are suggested to perturb the major redox

equilibria of the targeted infected red blood cells, which might be removed by macrophages. This results in development arrest and

death of the malaria parasite at the trophozoite stage.

source of essential nutrients (Figure 1).²Hemoglobin digestion

1. INTRODUCTION

takes place in acidic vesicles and in the food vacuole of the

Malaria is a major tropical parasitic disease aﬀecting 500

parasite;^3athis leads to the formation of iron(III) ferriprotopor-

million people worldwide and causing the death of 1ꢀ2 million

phyrin (FPIX(Fe^III) or FPIXꢀOH(Fe^III) or hematin) as a toxic

people per year, mostly children in Africa. Plasmodium is the

byproduct to the parasite.^3bꢀeIn the course of this prooxidative

causative agent of malaria; the most dangerous species is P.

falciparum, which is responsible for malaria complications such as

cerebral malaria or severe anemia. Multidrug-resistance of Plas-

modium toward broadly used antimalarial drugs (e.g., chloro-

process, the parasite is exposed to elevated ﬂuxes of reactive

oxygen species.

The parasite evades the toxicity of the released heme by

expressing at least two major detoxiﬁcation pathways, the

quine, sulphadoxine-pyrimethamine) has spread all over the

hemozoin formation in the food vacuole and an eﬃcient thiol

world in the last ﬁve decades. Resistance to artemisinin has

network in the cytosol. First, FPIX(Fe^III) is polymerized forming

now emerged.¹Therefore, new drugs are urgently needed,

crystals of hemozoin or malaria pigment.⁴Second, FPIX(Fe^III)

especially in poor countries. These drugs must be cheap and

is thought to undergo rapid peroxidative decomposition by

easy to synthesize, and exhibit mechanism(s) of action that

counteract parasite resistance to drugs in use.

The malarial parasite P. falciparum digests a large amount of its

host cell hemoglobin during the intraerythrocytic cycle as a

Received: February 24, 2011

Published: June 17, 2011

r

2011 American Chemical Society

11557

dx.doi.org/10.1021/ja201729z J. Am. Chem. Soc. 2011, 133, 11557–11571

|

Journal of the American Chemical Society

ARTICLE

Scheme 1. Redox-Active 1,4-Naphthoquinones as Subver-

sive Substrates of NADPH-Dependent Disulﬁde Reductases^a

Figure 1. Heme detoxiﬁcation pathways in the intraerythrocytic cycle

of P. falciparum and redox homeostasis in P. falciparum-infected red

blood cell. HMS, hexose monophosphate shunt. CQ, chloroquine is an

inhibitor of hemozoin formation (designated by the black X).

^a(a) Key structures are shown in Figure 3; (b) R⁵is the series variable.

reacting with H₂O₂, mainly in the food vacuole, that is, at pH 5.2

where autoxidation of hemoglobin(Fe^II) (Hb) takes place.⁵

Degradation of heme might also be signiﬁcant in the cytosol

via Fenton reactions, in the presence of glutathione⁶and

molecular oxygen following the generation of FPIX(Fe^II). A

number of antimalarial drugs such as chloroquine (CQ) and

related 4-aminoquinolines, quinine, as well as the phenothiazine

methylene blue (MB), are known to act as inhibitors of hemo-

zoin formation due to their binding to noncrystalline heme.³In

sensitive parasites, CQ prevents heme detoxiﬁcation resulting in

free heme accumulation, in the food vacuole and in membranes,

which in concert with oxygen species is thought to catalyze

oxidation reactions and protein damage causing parasite death.⁷

Malarial parasites digest methemoglobin(Fe^III) (metHb) faster

than Hb(Fe^II).⁸Therefore, the reduction of metHb can be used

to slow the parasite’s metHb digestion rate. Because FPIX(Fe^II)

is also an inhibitor of heme polymerization,⁹redox-active com-

pounds displaying the ability to reduce FPIX(Fe^III) to FPIX-

(Fe^II) can lead to the decrease of hemozoin formation and

increased oxidative stress in infected-red blood cells. Both events

would contribute to the arrest of trophozoite development.

Thus, agents with redox-cycling activity able to generate Fe^II

from Fe^IIIstates either at the hemoglobin or at the heme levels

might act as putative antimalarial agents.

parasites. They ranged from uncompetitive or catalytic inhibitors

(eq 1, Scheme 1),^11a,cto ﬂuorine-based suicide-substrates,^11band

their related prodrugs.¹²Considering the diﬀerent inhibitor

types, the most potent antimalarial eﬀects were indeed observed

for redox-cyclers acting as subversive substrates (eq 2 and

Scheme 1). These compounds are reduced via 1 or 2 electron-

(s)-transfer and regenerated in the presence of molecular oxygen,

with the concomitant production of superoxide anion radicals.

Starting from an identiﬁed lead antimalarial 3-benzylmena-

dione 1a (Scheme 1), we report here how we prepared the 3-

benzylmenadione analogues 1, 2, 5, 7 (benzylNQ), and 9 (azaNQ)

and oriented the investigations to the oxidative metabolism of the

potent antimalarial 3-(substituted-benzyl)menadiones 1, 2 (see

general structures in Scheme 1). Because the infected erythro-

cytes are the target of the lead compound, we hypothesized that

the benzylNQ 1a might be subjected to heme-catalyzed oxida-

tion reactions under oxidative conditions in the acidic vesicles or

in the food vacuole as part of the drug bioactivation mechanism

(Scheme 2 and Figure 2). On the basis of this hypothesis, we

prepared the putative metabolites oxidized in the benzyl position,¹³

that is, a menadione (or plumbagin, i.e., 5-hydroxymenadione)

substituted by a 3-benzoyl-chain (benzoylNQ) 3, 4, and the related

“benzhydrol” 6a (Scheme 1). We also synthesized a series of

(()-3-(R-methylbenzyl)NQ series 5a,5e (Scheme 1) as meta-

bolically resistant derivatives because oxidation might be ham-

pered. To correlate the antimalarial eﬀects of 1a with the redox

system, we envisioned that the GR-reduced naphthoquinone

(NQ) might also play a role both in the hemoglobin redox status

and the drug bioactivation process (Scheme 2 and Figure 2).

Therefore, as a link between the thiol network and the heme

metabolism, we showed that the metabolites I, the benzoylNQ,

might subsequently be reduced by GR in the cytosols (Scheme 2).

Taking into account the putative redox-active behavior of the

benzoylNQ metabolites, we then investigated if, in their reduced

form, they could convert metHb to Hb in vitro, as MB and

menadione do so in vivo.¹⁴As an expected outcome, we observed

The most important antioxidative system consists of thiols

that are regenerated by NADPH-dependent disulﬁde reductases;

these include the glutathione reductases (GR, E.C. 1.8.1.7) from

the parasite P. falciparum and of human erythrocytes. Both

enzymes are targets of antimalarial drugs.¹⁰These enzymes are

essential proteins for the survival of the malarial parasite infecting

red blood cells. They maintain the redox equilibrium in the

cytosol by catalyzing the physiological reaction, that is, the

reduction of glutathione disulﬁde (GSSG) to its thiol form

glutathione (GSH): NADPH + H⁺+ GSSG f NADP⁺

+

2GSH, in particular during the intensive production of reactive

oxygen species resulting from hemoglobin digestion.

Diﬀerent types of inhibitors (reversible, irreversible) were

designed in the 1,4-naphthoquinone series^11,12to evaluate the

impact of each inhibition mode on the propagation of the

11558

dx.doi.org/10.1021/ja201729z |J. Am. Chem. Soc. 2011, 133, 11557–11571

Journal of the American Chemical Society

ARTICLE

Scheme 2. Putative Redox Cascade Reactions Accounting for

the Observed Antimalarial Activity of 3-Benzylmenadione

Derivatives

Scheme 3. Synthesis of 3-Benzyl and 3-Benzoyl-Substitued

Derivatives of Menadione, Plumbagin, Diﬂuoromethylmena-

dione, and 6-Methylquinoline-5,8-dione^a

^aReaction conditions: (a) phenylacetic acid derivative (2.0 equiv),

AgNO₃(0.1 equiv), (NH₄)₂S₂O₈(1.3 equiv), MeCN/water 3/1, 85 ꢀC,

2 h (7ꢀ89%); (b) CrO₃(0.2 equiv), H₅IO₆(7.0 equiv), MeCN, room

temperature, 24 h (33ꢀ67%).

an inhibition of hemozoin formation and the arrest of tropho-

zoite development in experiments with synchronized parasites in

vitro (step 6 in Figure 2).

Finally, to oﬀer evidence for the concept that the GR-catalyzed

bioactivation step is essential for the observed antimalarial

activity of the 3-benzylmenadione series, the synthesis of a

diﬂuoromethyl analogue (7a, Scheme 1) of the lead 3-benzyl-

menadione 1a was designed to generate in situ (i.e., in the

parasite) the corresponding 3-benzoyl-diﬂuoromenadione me-

tabolite. This compound, the 3-benzoylmenadione 8a, when

synthesized (Scheme 1), was shown to act as a potent suicide-

substrate of isolated GRs. Hence, the GR inactivation by the

bioreducible alkylating NQ may be correlated with the abolish-

ment of the antimalarial activity of the parent diﬂuorinated

benzylNQ 7a. Our data support the essential role of an active

GR within a cascade of redox reactions to bioactivate potent

antimalarial 1,4-naphthoquinones. This GR-dependent drug

bioactivation process combined with the prodrug eﬀect speciﬁ-

cally present in Plasmodium via the hematin-catalyzed oxidation

reactions represents a new strategy to combat malarial parasites.

Figure 2. Putative model for lead 1,4-naphthoquinone bioactivation

aﬀecting redox homeostasis in P. falciparum through a cascade of redox

reactions accounting for the observed antimalarial activity of 3-benzyl-

menadione derivatives. HMS, hexose monophosphate shunt. Blue

arrows indicate reduction; red arrows oxidation. Dashed arrows repre-

sent uptake processes. The lead benzylNQ are proposed to be taken up

by the infected red blood cells (step 1), to be reduced in the cytosol of

the human red blood cell by human GR (step 2), and then to be oxidized

at the benzylic chain to benzoylNQ in the acidic vesicles or in the food

vacuole (step 3) in heme-catalyzed reactions. Subsequently, the

benzoylNQ are proposed to be transported into the cytosols and

reduced by GR of the human host cell (not shown) or of the parasite

(step 4) in a continuous redox cycle. The reduced species of

benzoylNQ are assumed to be transported through Fe^IIIcomplexation

into the acidic vesicles where the benzoylNQ_redtransfer the electrons

to oxidants (hematin or metHb, step 5). The ﬁnal result is an inhibition

of hemozoin formation (step 6) and the arrest of trophozoite devel-

opment as shown in Figure 8. Thus, the antimalarial benzylNQ would

act as prodrugs of redox-active principles, being cycled in and out of the

acidic vesicles in infected red blood cells, thereby oxidizing major

intracellular reductants (NADPH) and subsequently reducing oxi-

dants like hematin or metHb.

2. RESULTS

2.1. Chemistry. We first synthesized a series of menadione,

5-hydroxymenadione (plumbagin), and 5-azamenadione (6-meth-

yl-quinoline-5,8-dione) analogues 1, 2, 5, 7, and 9 functionalized at

11559

dx.doi.org/10.1021/ja201729z |J. Am. Chem. Soc. 2011, 133, 11557–11571

Journal of the American Chemical Society

ARTICLE

Scheme 4. Synthesis of the (()-3-(Phenylhydroxy-

methyl)menadione or “Benzhydrol” 6a^a

^aReaction conditions: (a) 2.5 equiv of SnCl₂, HCl, EtOH, room

temperature, 0.5 h; (b) 5.0 equiv of KOH, 3.0 equiv of Me₂SO₄, acetone,

60 ꢀC, 2 h (73%); (c) 1.0 equiv of Br₂, CH₂Cl₂, 0 ꢀC, 1 h, then room

temperature, 2 h (73%); (d) 1.05 equiv of ⁿBuLi, THF, ꢀ78 ꢀC, 0.5 h,

then 1.0 equiv of p-bromobenzaldehyde, THF, ꢀ78 ꢀC, 10 min, then

room temperature, 1 h (73%); (e) 3.0 equiv of CAN, CH₃CN, H₂O,

room temperature, 1 h (86%).

Scheme 5. Synthesis of r-Alkylated Carboxylic Acids 15a

and 15e and of 3-(R-Methylbenzyl)menadione Derivatives 5a

and 5e^a

Figure 3. Structures of key 3-substituted benzyl and benzoyl derivatives

of menadione, plumbagin, diﬂuoromenadione, and 6-methylquinoline-

5,8-dione.

C-3 by a benzyl-chain carrying a broad array of various substit-

uents and tested them for antimalarial activity (Scheme 3, see

selected key structures in Figure 3). The silver-catalyzed radical

decarboxylation reaction¹⁵provides a powerful tool for a fast and

easy access to 3-substituted-benzyl derivatives of menadione,

plumbagin, and 6-methylquinoline-5,8-dione in only one step in

moderate to good yields (46ꢀ89%, Scheme 3, Table S1 in the

Supporting Information) tolerating a wide spectrum of func-

tional groups except for the phenol and the carboxylic acid

derivatives (7% for 1d, and 12% for 1f). After flash chromatog-

raphy, all compounds were analytically pure and did not have to

be further purified.

The benzoylNQ derivatives 3, 4 were prepared as putative

metabolites of the benzylNQ 1, 2. The benzoylNQ were

obtained by oxidizing the corresponding benzylic compounds

using chromium(VI) oxide and periodic acid as terminal oxidants

(Scheme 3).¹⁶The oxidation reactions from benzyl derivatives 1,

2 were performed at room temperature for 24 h giving the

desired benzoylNQ products 3aꢀf, 4a, 8a in moderate to good

yields(33ꢀ67%, Scheme 3, Table S1 in the Supporting Information).

The “benzhydrol” 6a was also prepared as a putative metabo-

lite of 1a (Scheme 4). Its synthesis followed lithiation of the

2-methyl-3-bromo-1,4-dimethoxy-naphthalene 11 and p-bromo-

benzaldehyde addition to give the benzhydrol intermediate 12.

Oxidation with cerium ammonium nitrate (CAN) aﬀorded the

benzhydrol 6a.

^aReaction conditions: (a) lithium diisopropyl amide (LDA), THF, ꢀ78

ꢀC, then MeI; (b) EtOH, KOH, H₂O, RT; (c) menadione (1.0 equiv),

acid (2.0 equiv), CH₃CN/H₂O (v/v = 3/1), 0.1 equiv of AgNO₃, 1.3

equiv of (NH₄)₂S₂O₈, 85 ꢀC, 3ꢀ4 h.

Scheme 6. Preparation of 2-Diﬂuoromenadione^a

^aReaction conditions: (a) 2.0 equiv of DAST, under N₂, room tem-

perature, overnight (90%); (b) 3.0 equiv of CAN, CH₃CN, H₂O, room

temperature, 1 h (93%).

When the benzyl position of the phenylacetic acids was

R-alkylated, the yields of the silver-catalyzed decarboxylation

were signiﬁcantly lower (5a, 5e, yields 25ꢀ29%, Table S1 in the

Supporting Information). The compounds 5a and 5e were

synthesized via the silver-catalyzed decarboxylation of R-methyl

phenylacetic acids, 15a and 15e, prepared from phenylacetic

methyl esters (Scheme 5). In the case of 5e bearing a tert-butyl

group in para-position, the corresponding phenyl acetic acid was

synthesized from methyl-p-tert-butyl-phenylacetate (Scheme 5),

which was methylated in R-position to obtain the R-methylacid

methyl ester 14e.

The ester was ﬁnally hydrolyzed under basic conditions

to obtain the corresponding carboxylic acid 15e in excellent

yield. The same protocol was applied for the synthesis of the

11560

dx.doi.org/10.1021/ja201729z |J. Am. Chem. Soc. 2011, 133, 11557–11571

Journal of the American Chemical Society

ARTICLE

Figure 4. In vitro growth inhibition of malarial parasites (Dd2) by menadione derivatives and dose response curves for a selected number of 3-benzyl

representatives. Table: The IC₅₀values were determined using the multidrug resistant P. falciparum clone Dd2 in the ³H-hypoxanthine incorporation-

based assay.^18,19Values are the mean ( SEM of (n) independent determinations. The IC₅₀value of the antimalarial drug chloroquine is indicated as a

reference. Panels AꢀD: Highly synchronized P. falciparum cultures were incubated with increasing concentrations of the compounds indicated, and the

percent of parasite growth was determined after 72 h of drug exposure. The data points were ﬁtted using a three-parameter Hill function to calculate the

IC₅₀values given in the table. (A) Compounds 1a (black symbol) and 7a (gray symbol). (B) Compounds 1b (black symbol), 1e (gray symbol), and 1f

(white symbol). (C) Compounds 1c (black symbol) and 1g (gray symbol). (D) Compounds 1d (black symbol) and chloroquine (CQ, gray symbol).

The mean ( SEM of at least three independent determinations is shown.

R-methylphenylacetic acid 15a bearing a bromine atom in para-

position, but from ethyl 4-bromophenylacetate as starting material.

The synthesis of the diﬂuoromenadione derivative 7a was

achieved in ﬁve steps via the straightforward synthesis of

diﬂuoromenadione, prepared by direct ﬂuorination of 1,4-di-

methoxy-naphthalene-2-carbaldehyde 18 (Scheme 6). The in-

troduction of the aldehyde function was almost quantitatively

achieved (98%) by using 1,4-dimethoxy-naphthalene, dichloro-

methyl methyl ether, and tin(IV) chloride as reagents.¹⁷2-Formyl-1,

4-dimethoxy-naphthalene 18 was treated with diethyl-

aminosulfur triﬂuoride (DAST) leading to the diﬂuoromethyl-

substituted 1,4-dimethoxynaphthalene 19 with an excellent yield

(90%). A standard oxidation using CAN as oxidant gave the

diﬂuoromenadione in 93% yield (Scheme 6). The benzylation of

diﬂuoromenadione was performed through the standard silver-

catalyzed decarboxylation reaction of 4-bromo-phenylacetic acid

as coupling partner to give 7a (Scheme 3). Finally, the general

procedure for the oxidation of 3-benzyl derivatives using CrO₃/

H₅IO₆was applied to 7a and led to its benzoyl derivative 8a in

good yield (78%).

compounds were revealed with IC₅₀values below 100 nM (table

in Figure 4). Key structures are presented in Figure 3. The

compounds belong to the 3-benzylmenadione series bearing

halide- (1a, 1b, 1c), hydroxy- (1d) functions, nitrogen-based

(1g), and alkyl-substituent (1e). As previously seen, the car-

boxylic acid function in 1f is not a favorable parameter for cell

bioavalaibility.^11a,12aIn these assays, the standard drug CQ was

also tested as reference displaying an IC₅₀value of 110 nM. As

compared to the benzyl series, the corresponding benzoyl series

displayed lower antimalarial activities with IC₅₀values being

greater than 1 μM except in the case of the nitro derivative 3g.

Concerning the set of key compounds alkylated at the benzyl

position, they were designed to study the inﬂuence of the

“blocked” benzyl position toward the antimalarial activity. The

(()-3-(R-methylbenzyl)NQ 5 series displayed a 3ꢀ6-fold de-

creased antimalarial activity. This suggests an oxidative bioactiva-

tion of the parent benzylNQ involving the benzyl chain as part of

the mechanism of action. One explanation could be that the

benzoyl metabolites need to be generated in the parasite to

express their activity. Alternately, contrary to the parent ben-

zylNQ, they might not be recognized by a putative transporter to

be internalized by the parasite. This hypothesis remains to be

conﬁrmed in the future; we are preparing ¹³C-enriched ben-

zylNQ representatives to study their metabolism in P. falciparum-

infected erythrocytes by ¹H high resolution magic angle spinning

NMR as previously applied for ethionamide pro-drug activation

in mycobacteria.²⁰

2.2. Antimalarial Activities. The library of representative NQ

3

was tested for antimalarial effects using the H-hypoxanthine

incorporation-based assay (Figure 4).^18,19Inhibition of the

growth of P. falciparum by the compounds was evaluated by

determining the inhibitor concentration required for killing 50%

of the parasite (IC₅₀values). In a screening assay, all compounds

were tested against the CQ-resistant P. falciparum strain Dd2. Six

11561

dx.doi.org/10.1021/ja201729z |J. Am. Chem. Soc. 2011, 133, 11557–11571

Journal of the American Chemical Society

ARTICLE

Figure 5. Inhibition of the glutathione reductases from P. falciparum and human by 3-benzyl- and 3-benzoyl-substituted derivatives of menadione and

dose response curves for the selected 3-benzyl-menadione derivatives 3g and 3f as inhibitors of P. falciparum and human glutathione reductases under

steady-state conditions. Table: (a) The values were determined at pH 6.9 and 25 ꢀC in the presence of 1 mM GSSG; (b) in the presence of 5% dimethyl

sulfoxide (DMSO); (c) reprecipitation of the compound in the cuvette prevented IC₅₀determination; (d) in the presence of 1% DMSO;

(e) concomitant aggregation of the enzyme occurred; (f) data from ref 11b. nd: not determined. Panel A: Dose response curves for the selected

3-benzylmenadione 3g as inhibitor of P. falciparum (0) and human (9) glutathione reductases in the presence of 1 mM GSSG as disulﬁde substrate and

100 μM NADPH. Panel B: Dixon plot showing inhibition of P. falciparum GR-catalyzed GSSG reduction at a ﬁxed NADPH concentration (100 μM) in

the presence of GSSG as variable substrate (20 (Δ)ꢀ40 (2)ꢀ60 (O)ꢀ100 (()ꢀ400 (0)ꢀ1000 (b) μM) and increasing 3f concentration

(0ꢀ0.5ꢀ1ꢀ3 μM). 1% ﬁnal DMSO was present in experiments.

2.3. Glutathione Reductase Inhibition Studies. Determina-

tion of IC₅₀Values. As a consequence of their potential substrate

competence, the 3-benzyl- and 3-benzoylmenadione analogues

were first tested as inhibitors of both GRs in the standard

1 mM GSSG reduction assay (table in Figure 5). 2-Methyl-1,4-

naphthoquinone (menadione) was used as reference, displaying

IC₅₀values of 42.0 and 27.5 μM for P. falciparum GR and human

GR respectively. All compounds (Figure 5) displayed higher GR

inhibition potencies than did the unsubstituted menadione. In

addition, the 3-benzoylmenadiones 3, 4 behaved as the most

potent GSSG reductase inhibitors described so far, with IC₅₀

values up to 10-fold lower than for the 3-benzyl analogues 1, 2

(Figure 5A). In general, all compounds showed a 1.25ꢀ9-fold

stronger inhibition for human GR than for P. falciparum GR,

assuming a stronger interaction with human GR. Compound 3f

was then selected for detailed inhibition studies to determine the

inhibition type and constants in P. falciparum GR assays.

the appropriate equation with Kaleidagraph using a computer-

ized least-squares regression program. With GSSG (20ꢀ1000 μM)

as variable substrate and NADPH at a constant concentration of

100 μM, the K_mvalue of GSSG in the absence of inhibitor was

found to be 144.4 ( 15.7 μM. The inhibition of P. falciparum GR

by 3f (0ꢀ3 μM) was found uncompetitive with respect to GSSG.

The K_ivalue for 3f was determined as 1.0 ( 0.06 μM from the

secondary plot expressing 1 + [I]/K_ias a linear function of

inhibitor concentration. The uncompetitive type of inhibition of

P. falciparum GR by 3f was deduced from CornishꢀBowden,

LineweaverꢀBurk, and Dixon (Figure 5B) plots. In the human

GR assay also assuming uncompetitive inhibition, the K_ivalue for

3f was determined as 0.64 ( 0.02 μM with respect to GSSG.

Glutathione Reductase-Catalyzed Naphthoquinone Reduc-

tase Activity. The ability of P. falciparum GR to reduce the 1,4-

naphthoquinones was studied in vitro by following the oxidation

of NADPH in the presence of various 3-benzyl- and 3-benzoyl

menadione derivatives and was compared to menadione itself as

reference (Table 1). The NQ reductase activity of P. falciparum

GR was compared to the intrinsic NADPH oxidation activity of

the enzyme in the absence of the NQ. The catalytic parameters

of menadione against P. falciparum GR were shown^11ato be

Inhibition Type of P. falciparum and Human Glutathione

Reductase under Steady State. To investigate the type of P.

falciparum GR inhibition, kinetics were performed at various

concentrations of both GSSG and potent inhibitor 3f, in the

presence of a saturating NADPH level. The data were fitted to

11562

dx.doi.org/10.1021/ja201729z |J. Am. Chem. Soc. 2011, 133, 11557–11571

Journal of the American Chemical Society

ARTICLE

Table 1. Plasmodium falciparum Glutathione Reductase-Catalyzed Naphthoquinone Reduction^a

naphthoquinone reductase activity of PfGR^a

k_cat(min^ꢀ1k_cat/K_m(mM^ꢀ1s^ꢀ1

nd^bnd^b

2.53 1.62

series

benzyl

compd

1a

R⁵

K_m(μM)

)

Br

F

nd^b

26

1b

1d

OH

Br

67

2.54

2.31

0.64

6.31

4.96

1.32

5.26

7.76

1.4

benzoyl

3a

6.1

3b

F

56.1

42.4

84.2

18

16.7

3.36

26.6

8.38

111.6

3c

CF₃

OH

NO₂

COOH

Br

3d

3g

3f

1325

3.07

8.5

4a

2.3

12.5

benzhydrol

ref

6a

Br

2.17

9.6

4.25

1.99^c

menadione

MB

82.2

42.2

34.8

13.7

^aFor GSSG as a substrate of PfGR: K_m= 83 μM, k_cat= 9900 min^ꢀ1= 165 s^ꢀ1, k_cat/K_m= 1988 mM^ꢀ1s^ꢀ1. ^bPrecipitation of the compound was observed

above 10 μM; at 10 μM there is no inhibition. ^cFrom ref 11a.

82.2 μM and 9.6 min^ꢀ1for K_mand k_cat, respectively, leading to a

catalytic efficiency k_cat/K_mof 1.99 mM^ꢀ1s^ꢀ1(Table 1).

Scheme 7. Time-Dependent Inactivation of Human Glu-

tathione Reductase by Monoﬂuoromenadione Derivatives

As compared to menadione as reference, the tested com-

pounds 1aꢀd, 3a-g, and 4a bearing a various set of substituents

showed low K_mvalues ranging from 6.1 μM (3a) to 56.1 μM

(3b), indicating a tighter binding to the enzyme PfGR, while 3d

with a K_mvalue of 84.2 μM showed similar apparent Michaelisꢀ

Menten constant. From the catalytic eﬃciencies, expressed as

k_cat/K_mvalues, 3a, 3b, 3d, 3g, 6a, and 4a behaved as very eﬀective

substrates of PfGR, among the most eﬃcient subversive sub-

strates of P. falciparum GR described so far. In contrast, when

diﬀerent concentrations of 3f were used, catalytic constants

expressed as a k_catof 1.86 s^ꢀ1and a K_mvalue of 1324.6 μM

were determined assuming MichaelisꢀMenten kinetics. The

high K_mvalue for the carboxylic acid function in the napthto-

quinone reductase activity was previously observed for ana-

logues.^11aCompounds with two redox active moieties like 3g

(with a nitrophenyl group) showed a 3.9-fold increased catalytic

eﬃciency when compared to menadione. Also, an increased

oxidant character, expressed with the plumbagin derivative 4a,

led to 6.3-fold higher catalytic eﬃciency than that of menadione

and a value similar to that of MB.

2.4. Abolishment of the Antimalarial Activity by Glu-

tathione Reductase Inactivation. Interestingly, the 3-benzyl-

difluoromenadione 7a, an analogue of the active lead 1a, was

found to have almost no effect in antimalarial screening assays

(IC₅₀> 16 μM, Figure 4). This abolishment of the antimalarial

activity suggested that the presence of the fluorine atoms at the

methyl group inhibited the drug metabolism in the Plasmodium-

infected erythrocytes. Because its benzoyl metabolite was ex-

pected to be generated in the parasite in situ, the 3-benzoyl-

difluoromenadione 8a, analogue of the active lead 3a, was

synthesized and tested as a suicide-substrate of both GRs.

As shown earlier with monoﬂuoromenadione derivatives^11b

(Scheme 7), it has been conﬁrmed that reduction by NADPH is a

prerequisite for irreversible GR inhibition by the suicide-sub-

strate, 6-[2⁰-(3⁰-ﬂuoromethyl)-1⁰,4⁰-naphthoquinolyl]hexanoic

acid (ﬂuoroM₅), and time-dependent inactivation resulted in

alkylation of human GR at Cys58 of the redox active dithiol.^11b

Noteworthy is the previous observation made with these mono-

ﬂuorine-based suicide-substrates of glutathione reductases: irre-

versible GR inhibition did not lead to increased antimalarial

eﬀects but instead to increased toxicity toward human cells.

The presence of one, two, or three ﬂuorine atoms has an

inﬂuence on the oxidant character of the parent menadione core.

The electrochemical properties of the ﬂuoromethyl compounds

were investigated by cyclic voltammetry under the conditions

previously described.^11bBoth with the triﬂuoromenadione

(Eꢀ₁= ꢀ0.308 V/standard calomel electrode (SCE)) and the

diﬂuoromenadione (Eꢀ₁=ꢀ0.385 V/SCE), one-electron reduction

occurred at less negative values than with menadione used as

reference (Eꢀ₁= ꢀ0.650 V/SCE). These values make these two

ﬂuoro analogues of menadione strong oxidants and potential

eﬀective substrates of both glutathione reductases. In contrast to

menadione showing redox reactions kinetically reversible in the

cyclic voltammogram (data not shown), the formation of the

quinone methide species followed by polymerization in the case

of the diﬂuoromenadione might be responsible for the non

reversibility of the electron transfer reactions. These observations

are in agreement with earlier literature,^11b,21rendering the

diﬂuoromenadione more prone to act as a suicide-substrate.

When following NADPH consumption at 340 nm in the enzymic

assays, the diﬂuoromenadione was reduced by P. falciparum GR

with a catalytic eﬃciency k_cat/K_mof 155.9 mM^{ꢀ1 ꢀ1}, a value

s

12.7-fold lower than that determined for the physiological

substrate GSSG but 82-fold higher than that determined for

menadione (Table 2).

11563

dx.doi.org/10.1021/ja201729z |J. Am. Chem. Soc. 2011, 133, 11557–11571

Journal of the American Chemical Society

ARTICLE

Table 2. Kinetic Parameters of Human and P. falciparum

Glutathione Reductases with Fluoromethyl Analogues of

Menadione as Substrates

GR

k_cat/K_m

(mM^ꢀ1s^ꢀ1

)

compd

menadione

species

K_m(μM)

k_cat(s^ꢀ1

)

h

31.2

87.4

66.2

82.1

0.16

0.17

6.7

5.1

1.9

menadione

Pf

h

diﬂuoro menadione

101.4

155.9

Pf

12.8

Table 3. Selected Kinetic Parameters of Human Glutathione

Reductase Inactivation by Fluoro Menadione Analogues

compd

t_1/2(min) K_I(μM) k_i(min^ꢀ1

)

k_i/K_I(mM^ꢀ1s^ꢀ1

)

diﬂuoro menadione

0.1

66.2

16.8

8.9

7.0

1.8

0.025

16.9

Figure 6. Dose response curves for the 3-benzoyl-menadione 3a and -

diﬂuoromenadione 8a derivatives as human glutathione reductase

inhibitors under steady-state conditions. From the dose response curves

recorded at pH 6.9 and 25 ꢀC in the presence of 1 mM GSSG as disulﬁde

substrate, 100 μM NADPH, and 1% DMSO, the IC₅₀values for the

selected 3-benzoylmenadione derivatives 3a (0) and 8a (9) were

determined as 0.3 and 12 μM, respectively.

7a

8a

27.7

0.025

9.0

0.08

With the diﬂuoromenadione, 7a (benzylNQ) and8a (benzoylNQ),

the two diﬂuoromethyl analogues of the 3-benzyl- and the

3-benzoyl-menadione derivatives 1a and 3a, respectively, time-

dependent inhibition of the GRs from man and P. falciparum was

studied using the 1 mM GSSG reduction assay to evaluate the

residual activity of reacted enzymes. Upon incubation with

NADPH and the suicide-substrates, inactivation of human GR

by the three compounds followed pseudoﬁrst-order reaction

kinetics. The experimental data allowed the application of the

derivation by Kitz and Wilson²²for irreversible inactivation. A

semilogarithmic plot of the fraction of noninhibited enzyme

activity ln(v_i/v₀) versus incubation time yielded straight lines

with increasing slopes at short time periods, equivalent to the

apparent rate constant of irreversible inhibition (k_obs). The

secondary plot expressing k_obsas a function of inhibitor con-

centration followed eq 3:

derivatives in GSSG reduction assays in the presence of the

human GR are shown in Figure 6. They are thought to result

from two events: NADPH consumption due to GSSG reduction

catalyzed by free homodimeric human GR in solution and

polymerization of the diﬂuoro quinone methide generated from

the diﬂuoromenadione derivative in the presence of the reduced

enzyme. The high reactivity of the diﬂuoro quinone methide

intermediate is likely to act as a cross-linking reagent, by analogy

to acrolein and its diﬂuoro analogue in polyacroleins formation

(see below).²³

k_i½Iꢁ

k_obs

¼

ð3Þ

K_I+ ½Iꢁ

where K_Iis the dissociation constant of the inhibitorꢀenzyme

complex and k_iis the ﬁrst-order rate constant for irreversible

inactivation, respectively. The k_obsdata versus [8a] (or [7a] or

[diﬂuoromenadione]) were ﬁtted to eq 3, which resulted in the

hyperbolic curve and the determination of the dissociation

constant for the 8aꢀ, the 7aꢀ, or the diﬂuoromenadioneꢀ

GR complexes and the ﬁrst-order rate constant for irreversible

inactivation, respectively, as K_I= 8.9 ( 4.0 μM and k_ias 9.0 (

1.8 min^ꢀ1for 8a (versus 16.8 μM and 0.025 min^ꢀ1for 7a and

66.2 μM and 7.0 min^ꢀ1for the diﬂuoromenadione), allowing

estimation of a second-order rate constant k_i/K_Ias 16.9 mM^ꢀ1

s^ꢀ1for 8a versus 0.025 for 7a and 1.8 mM^ꢀ1s^ꢀ1for diﬂuor-

omenadione (Table 3). The resulting half-times (t_1/2) of in-

activation of the human GR were determined as 0.1 min and

below 0.1 min for the diﬂuoromenadione and 8a, respectively,

versus 27.7 min for 7a (Table 3).

The results are consistent with our concept of an irreversible

mechanism-based inhibition by 3-benzoyl-diﬂuoromenadione

derivatives. In additional kinetic experiments (data not shown),

we observed that Plasmodium GR was also quickly inactivated

upon incubation of the enzymes with NADPH and 8a. The ﬁrst-

order rate constant k_ifor irreversible inactivation could not be

determined because of the concomitant inactivation of the

parasitic enzyme by NADPH itself in open air.^11bThus,

the irreversible GR inactivation by a ﬂuorine-based suicide-substrate

of both GRs designed to be generated from 7a in the parasites,

the benzoylNQ 8a, was accompanied by a complete loss of the

antimalarial activity of both the 3-benzylmenadione 7a and the

3-benzoylmenadione 8a. These results suggest the essential

requirement of an active GR (either the human or the parasitic

enzyme, or both) in the drug bioactivation of the antimalarial

lead benzylNQ 1a.

Interestingly, the IC₅₀values of the diﬂuoro-naphthoquinones

(diﬂuoromenadione, 7a, 8a) determined in GR-catalyzed GSSG

reduction assays under steady conditions were found to be 1

order of magnitude higher than the values for the benzoylNQ

series, 3aꢀg (see Table in Figure 5). The dose response curves

for the 3-benzoyl-menadione 3a and -diﬂuoro-menadione 8a

2.5. Methemoglobin Reduction in the Presence of the

Redox-Active BenzoylNQ. To evaluate the reduction of metHb

to Hb, we set up an assay using the benzoylNQ, metHb(Fe^III),

and the glutathione reductase/NADPH-based system to regenerate

11564

dx.doi.org/10.1021/ja201729z |J. Am. Chem. Soc. 2011, 133, 11557–11571

Journal of the American Chemical Society

ARTICLE

Figure 7. Methemoglobin reduction by the glutathione reductase/NADPH-reduced redox-active agents, benzoylNQ 3d or methylene blue, or through

the direct reaction with the 3-benzoyl-dihydronaphthoquinone 20. UVꢀvis absorption spectra evidencing the reduction of metHb(Fe^III) to Hb(Fe^II) in

the coupled assay based on the human GR/NADPH system in the presence of the subversive substrate benzoylNQ 3d, or MB, and the direct reduction

of metHb(Fe^III) to Hb(Fe^II) by the dihydrobenzoylNQ 20. Panel A: The reaction of 200 μM 3d in the presence of 100 μM metHb, 400 μM NADPH,

1 μmol human GR, ﬁnal [DMSO] 1%, at 37 ꢀC caused a shift in the λ_maxof metHb (from 405 to 410 nm). Spectra were measured at diﬀerent times over a

15 min period following addition of 20 μL of the reaction mixture to a 1 mL cuvette containing GR buﬀer, pH 6.9. Panel B: Methylene blue was used as a

positive control under the same conditions; a clear shift of λ_maxof metHb was observed from 405 to 410 nm. Panel C: CQ was used as a negative control.

As expected, no shift was observed in the presence of CQ. The conditions used in the reaction mixture were 50 μM metHb, 100 μM CQ, 200 μM

NADPH, and 1.06 μmol hGR, ﬁnal [DMSO] 1%, 37 ꢀC. Panel D: The conditions used in the reaction mixture were 2 μM metHb, 40 μM

dihydrobenzoylNQ 20, ﬁnal [DMSO] 2% in GR buﬀer, pH 6.9, at 37.5 ꢀC. The reaction was started directly in the cuvette, and UVꢀvis absorption

spectra were recorded as a function of time.

the hydronaphthoquinone continuously. The UV spectrum of

metHb between 300 and 700 nm is characterized by a maximal

absorbance at 405 nm, and a broad band centered at 630 nm.

Figure 7A displays reduction of metHb in the presence of the

position to the methanone group, which likely stabilizes the

β-hydroxyketone bridge via hydrogen bonds. In that case, a complete

reduction of metHb(Fe^III) to Hb(Fe^II) was observed in 2 min in the

absence of the regenerating GR/NADPH system (Figure 7D).

2.6. Cytotoxic Activities against Human Cells in Vitro.

Noteworthy is the absence of hemolysis of infected and non-

infected red blood cells, at the highest micromolar drug con-

centrations tested (30 μM). In addition, all compounds were

tested for cytotoxicity against two human cell lines, the human

buccal carcinoma cell line cell (KB) and the human lung MRC-5

fibroblasts, by using the Alamar blue assay (Table S2 in the

Supporting Information). Most of the 1,4-naphthoquinone

derivatives exhibited a low cytotoxicity profile against both

human KB and MRC-5 cell lines as evidenced by the high IC₅₀

values above 32 μM. The toxicity of the two derivatives, 3-benzyl-

difluoromenadione 7a and the oxidant azamenadione 9a, prob-

ably stems from the higher lipophilicity due to the presence of

fluorine atoms and/or from their high oxidant character,^11c

respectively. Furthermore, when tested in Ames test, the lead

benzylNQ 1c, up to 100 μM, did not induce mutagenicity using

three strains of Salmonella thyphimurium TA98, TA100, and

TA1535 (the complete data of the genotoxicity study will be

published elsewhere).

benzoylNQ 3d, characterized by a k_cat/K_mof 5.26 mM^ꢀ1s^ꢀ1

,

continuously reduced by the human GR/NADPH-based system.

Upon metHb reduction, the spectrum of Hb(Fe^II) showed a shift

of the maximal absorbance from 405 to 410 nm and two weak

bands at 536 and 576 nm (Figure 7A). MB with a k_cat/K_mvalue of

13.7 mM^ꢀ1s^ꢀ1was used as a positive control of this effect

(Figure 7B). In the coupled assay, the shift of the maximal

absorbance from 405 to 410 nm is already visible at 20 μM MB,

or at 100 μM menadione (data not shown), or at 50 μM

benzoylNQ 3d. As expected, no band shift was observed in the

presence of CQ instead of the benzoylNQ nor in the presence of

MB (Figure 7C).

To conﬁrm the electron transfer from the reduced 3-benzyl-

naphthoquinones to MetHb(Fe^III), the direct reduction of

methemoglobin MetHb(Fe^III) by a 3-benzoyl-dihydronaphtho-

quinone derivative was evaluated (in the absence of the GR/

NADPH system). Because the 3-benzoyl-dihydronaphthoqui-

none derivatives are quite unstable in open air and very prone to

be quickly reoxidized into naphthoquinones, a “stabilized”

3-benzoyl-dihydronaphthoquinone derivative 20 (Figure 3) was

prepared. This compound possesses a ﬂuorine atom in ortho

2.7. Antimalarial Activities in Plasmodium berghei-Infected

Mice. Five of the most active compounds (1a, 1c, 1d, 1e, 1g),

11565

dx.doi.org/10.1021/ja201729z |J. Am. Chem. Soc. 2011, 133, 11557–11571

Journal of the American Chemical Society

ARTICLE

Table 4. Reduction of Parasitemia in CD1Mice Infected with

Plasmodium berghei Strain ANKA^a

dose

reduction po^b

(%)

compd

(mg/kg ꢂ 4) reduction ip^b(%)

p-value^c

untreated control

0

CQ^a

10

30

50

30

94.9

98.8

0

0.0079

ns

1a

35.6

20.7

28.4

42.4

43.4

0.8

0.0079

ns

1c

1d

1g

ns

10.4

34.9

ns

1e

0.0079

^aIn this assay, at 1.0, 3.0, and 10.0 mg/kg po, chloroquine displayed a

reduction of the parasitemia of 2.5%, 16.6%, and 94.9%, respectively. ^bip

and po stand for intraperitoneal and oral administration, respectively.

^cCalculated in the two-tailed MannꢀWhitney U test on the diﬀerences

of treated versus untreated mice. ns means non signiﬁcant.

which had successfully passed all studies with parasites in

cultures, were subsequently tested for their antimalarial action

in vivo using the murine CD1 malaria model infected by the CQ-

susceptible ANKA strain of P. berghei. Results of in vivo screens

for the five compounds conducted according to the Peters’s

4-day test are given in Table 4. For comparative purposes, data

for CQ and the five derivatives upon intraperitoneal administra-

tion (ip) acquired in the same screens were included. The

untreated control group showed no decrease in parasitemia after

ip or oral (po) administration. CQ as reference was applied at a

dose of 10 mg/kg ip and po; the parasitemia decreased by 94.9%

and by 98.8%, respectively. Compounds 1a, 1c, 1d, 1e, and 1g

showed significant activity following a 4-day treatment.

Compound 1a was administered at 50 and 30 mg/kg per day.

There was no signiﬁcant inhibition when the compound was

given orally, but there was a signiﬁcant eﬀect, 35.6% inhibition

(p-value = 0.0079), when given intraperitoneally at 50 mg/kg.

The two most active drugs 1e (t-Bu derivative) and 1g (nitro

derivative) caused 43.4% and 42.4% reduction of parasitemia,

using a daily dose of 30 mg/kg ip, respectively. At this dose level,

1c (CF₃derivative) and 1d (phenol) displayed a 20.7ꢀ28.4%

decrease of parasitemia, suggesting a poor bioavailability or a high

rate of drug clearance due to rapid elimination in the host. Upon

oral administration of the two most active drugs 1e and 1g at

30 mg/kg per day, the parasitemia was reduced to 34.9% and

10.4%, respectively. So far, the eﬃciency of 1a has been improved

by preparing a metabolically resistant analogue by introduction

of a t-Bu group in the critical position for biooxidation.²⁴The data in

Table 4 support an oxidative metabolism in vivo (see 1a data versus

1e data); that is, 1e is equally active by ip (43.4%) and po (34.9%)

administration, while this does not apply for 1a. More chemistry is

required to improve the eﬃciency of the 1a series in vivo.

Figure 8. Compound 1c-induced inhibition of P. falciparum tropho-

zoite development. Optical examination of compound 1c-induced

inhibition of P. falciparum trophozoites was performed microscopically.

Synchronized parasites at the trophozoite stage of the intraerythrocytic

cycle (i.e., 16ꢀ21 h posterythrocyte invasion) were grown either in

normal culture medium (no drug) or in culture medium supplemented

with 5 μM chloroquine (CQ) or 5 μM compound 1c. Diﬀ Quick-stained

smears were prepared from aliquots of the cultures at diﬀerent time

points. Details from smears at 0 h (16ꢀ21 h-old parasites), 12 h (28ꢀ33

h-old parasites), and 25 h (41ꢀ46 h-old parasites) of incubation are

presented. Hemozoin in the parasite food vacuole is indicated by an

arrow. Black arrowhead: erythrocyte. Note the multinucleated parasite

(empty arrowhead) after 25 h in normal culture (no drug). Scale bar:

2.5 μm. Photographs: Metamorph software, Power Point/Illustrator software.

1c inhibits FcB1 trophozoite growth with a 0.90 μM IC₅₀value

and an estimated 5 μM IC_{100 min}. Chloroquine IC₅₀and IC₁₀₀

values on FcB1 trophozoites were 0.25 and 2 μM, respectively. As

shown in Figure 8, maturation of young trophozoites was

dramatically altered in the presence of 5 μM compound 1c

(column 1c) or 5 μM chloroquine (column CQ).

In both cases, a 12 h application prevented the parasite

development, and no hemozoin was seen in the food vacuole

(Figure 8, 1c 12 h, CQ 12 h). Prolonged incubation with

compound 1c led to parasite condensation and appearance of

pycnotic forms (Figure 8, 1c 25 h) contrasting with the shapeless,

uncondensed forms observed with chloroquine (Figure 8, CQ 25 h).

Thus, our results show that compound 1c is inhibitory to tropho-

zoite development and prevents the formation of hemozoin.

2.8. Inhibition of the Development of P. falciparum Tro-

phozoites by Compound 1c. The intraerythrocytic cycle of P.

falciparum begins when the infective form of the parasite invades

an erythrocyte, and it ends 48 h later when multiple new infective

parasites escape from the host cell. In the meantime, Plasmodium

has developed through ring (0ꢀ20 h postinvasion (p.i.)), tropho-

zoite (20ꢀ36 h p.i.), and schizont (36ꢀ48 h p.i.) stages. Because

trophozoite is the most active stage at digesting host cell hemoglobin,

we assessed whether compound 1c might inhibit trophozoite

development. As determined from doseꢀresponse assays, compound

3. DISCUSSION

The role of redox enzymes in allowing the establishment of a

microenvironment for parasite development is well-known. It is

11566

dx.doi.org/10.1021/ja201729z |J. Am. Chem. Soc. 2011, 133, 11557–11571

Journal of the American Chemical Society

ARTICLE

clinically validated by the observation made in hemoglobinopa-

thies that increasing the oxidative stress in erythrocytes confers a

milieu hostile for parasite development. Prevalence of glucose-6-

phosphate dehydrogenase (G6PDH) deﬁciency with 400 million

people aﬀected coincides with that of malaria with a very similar

global distribution; this supports the so-called malaria protection

hypothesis.²⁵In particular, the parasites do not develop well in

G6PDH-deﬁcient red blood cells²⁶or in erythrocytes depleted in

GR activity²⁷or in de novo biosynthesis of glutathione.²⁸At the

biochemical level, G6PDH is the ﬁrst enzyme of the pentose

phosphate pathway (hexose monophosphate shunt, HMS in

Figure 1), which provides the reducing power of the cells in

the form of NADPH. As NADPH is the only reducing substrate

for GR in erythrocytes, G6PDH deﬁciency in red blood cells can

be regarded as a natural knock-down version of human GR.

G6PDH deﬁciency is not a serious condition for humans but

prevents a severe malaria attack because the low antioxidative

capacity in the red blood cells makes the milieu hostile for

Plasmodium. Consequently, rapid elimination of parasitized

G6PDH-deﬁcient cells from the circulation occurs by enhanced

phagocytosis via complement activation.²⁹This important ob-

servation was recently supported by data from Arese and Becker’s

groups.³⁰Human GR deﬁciency or drug-induced GR inhibition

may protect from malaria via inducing enhanced ring stage

phagocytosis by macrophages rather than by impairing parasite

growth directly. Thus, mimicking G6PDH deﬁciency by GR

redox-cyclers and other GR-targeting compounds represents a

challenge to design new preclinical candidates as antimalarial

drugs. However, these drugs must not trigger hemolysis both in

G6PDH-suﬃcient and in G6PDH-deﬁcient individuals.³¹This

essential prerequisite could be reached using a prodrug approach

or an enzyme-drug bioactivation process. The most relevant

examples of this strategy are being exempliﬁed by the ﬁrst-

discovered synthetic antimalarial drug, methylene blue (MB),³²

and, in the present work, by our lead antimalarial benzylNQ 1

(Figure 2).³³These compounds lead to the killing of the malarial

parasites, as reported here.

To develop this strategy, the synthesis and the screening of

more than 100 new menadione derivatives for antimalarial activity

allowed us to characterize a new series of potent antimalarial

benzylNQ analogues 1 that are active both in vitro (in the low

nanomolar range against the tested P. falciparum Dd2 strain) as

well as in vivo. The selected compounds displayed no observable

toxic eﬀects against the human KB and MRC-5 cell lines or in

mice. These lead benzylNQ are proposed to be metabolized to

benzoylNQ through a cascade of redox reactions, involving a

heme-catalyzed benzylic oxidation reaction under the speciﬁc

conditions found in the acidic vesicles/food vacuole of the

parasites, and then by a highly eﬀective GR-catalyzed reduction

in the cytosol. This last step was conﬁrmed to be essential for the

bioactivation of the benzylNQ series because GR inactivation by

a potent suicide-substrate generated in the parasites led to the

abolishment of the antimalarial activity of the parent diﬂuori-

nated benzylNQ. From the mechanistic point of view, the

benzoylNQ metabolites, in oxidized form (1,4-NQ core), were

shown to act as the most eﬃcient subversive substrates of

reduced GR described so far and, in reduced form (hydro-1,4-

NQ core), to convert metHb to Hb. The reduced species of

benzoylNQ are assumed to be transported through Fe^IIIcom-

plexation into the food vacuole where the electrons are trans-

ferred to oxidants (heme and metHb). Subsequently, in oxidized

forms, the benzoylNQ might be transported into the cytosol

where they are reduced by the cytosolic GR (either from human

erythrocyte or from the parasite) in a continuous redox cycle

(Figure 2). This work provides a new strategy based on the

particular situation of electrochemically active P. falciparum-infected

erythrocytes in the presence of redox-cyclers, transported as

charged species between two compartments, the anodic one

for oxidation reactions (acidic vesicle/food vacuole network),

and the cathodic one for reduction reactions (cytosol), possibly

via iron complexation.^11d,34Starting with P. falciparum-infected

red blood cells and glucose as energy source, the benzoylNQ

were designed to serve as membrane-permeable redox mediators

and a source of electrons to inhibit the trophozoite development.

Noteworthy is the fact that no harm is expected from human GR

subversive substrates in all other host cells containing a nucleus

because of the high ﬂux control in these cells, for instance, by the

de novo enzyme synthesis. Only host red blood cells, which have

no nucleus and no possibility to regulate the glutathione bio-

synthesis, can be sensitive to reversible glutathione reductase

inhibitors. This distinct situation between red blood cells and

other nucleated cells represent the Achilles’ heel on which we

cash in as the cotarget of our drug design strategy.

A second reason to target the reduction of metHb is that the

ferric form of hemoglobin is not capable of oxygen transport. A

decreased oxygen carrying capacity of blood due to anemia is

exacerbated by reduction in oxygen carrying capacity; this might

proceed from even a modest concentration of metHb leading to

an impaired supply of oxygen for the tissue. The reduction of

metHb(Fe^III) to Hb(Fe^II) is of great importance in the treatment

of malaria. High levels of metHb are observed in nonparasitized

human red blood cells during Plasmodium vivax infections,^35a

during Plasmodium yoelii nigeriensis infection in mouse erythrocytes,^35b

and also in primaquine (PQ)- or dapsone-treated patients with

G6PDH deﬁciency.^36,37In the case of PQ, methemoglobinemia

was recently shown to result from the transfer of one electron

both from iron of Hb and the metabolite of PQ, 5-hydroxyPQ, to

the π* orbital of O₂.³⁸This favors the formation of H₂O₂and the

accumulation of metHb. In the present work, a high beneﬁt of

our strategy originates from the fact that the expected metabo-

lites generated from the lead benzylNQ are eﬀective substrates of

GR; thus, the product, the reduced benzoylNQ, redox-cycles

metHb to Hb. This is not the case of PQ and other 8-aminoqui-

nolines. Nevertheless, optimizing and developing our lead ben-

zylNQ into preclinical candidates will involve several challenges

concerning the pathophysiology of hemolysis in G6PDH pa-

tients. As there is no pharmacologically validated model today for

hemolysis in G6PDH-deﬁcient populations, we are setting up

such a platform to determine and to predict any detrimental

eﬀects on G6PDH-deﬁcient red blood cells from patients at risk.

In summary, the antimalarial 1,4-naphthoquinones are proposed to

act catalytically as redox-active biosensors that are cycled in and out

the food vacuole in infected red blood cells, thereby oxidizing major

intracellular reductants (like NADPH in the cytosol) and subse-

quently reducing the oxidants (hematin and metHb). Ultimately, the

antimalarial 1,4-naphthoquinones are suggested to aﬀect the redox

equilibrium in the parasites, drowning the parasite in its own

metabolic products. They will serve as lead compounds for further

structural reﬁnement with improved solubility and/or bioavailability.

In the long term, the results from our interdisciplinary

approach will provide detailed insight into the understand-

ing of how redox-cyclers interfere with the putative negative

cooperativity of GR toward NADPH binding and how this

property can be exploited for disulﬁde reductase-catalyzed drug

11567

dx.doi.org/10.1021/ja201729z |J. Am. Chem. Soc. 2011, 133, 11557–11571

Journal of the American Chemical Society

ARTICLE

bioactivation as a general concept to open a new perspective in

medicinal chemistry. The mechanism of this putative negative

cooperativity toward NADPH binding remains to be proved and

orientates the future direction of our investigations.

(w), 3076 (w), 3049 (w), 3018 (w), 2936 (w), 1672 (vs), 1657 (vs), 1625

(s), 1594 (s), 1487 (vs), 1406 (m), 1329 (s), 1297 (vs), 1181 (m), 1123

(s), 1082 (s), 1071 (m), 1035 (vs), 1013 (s), 876 (m), 831 (s), 788 (s), 733

(s), 713 (m), 535 cm^ꢀ1(m). EI MS (70 eV, m/z (%)): 377.1 ([M]⁺, 21),

325.1 (11), 257.1 (10), 169.0 (100), 90.1 (18). Anal. Calcd for C₁₈H₁₁-

BrF₂O₂: C, 57.32; H, 2.94. Found: C, 57.01; H, 3.12.

4. EXPERIMENTAL SECTION

General Procedure for the Oxidation of Benzyl Derivatives to the

Corresponding Benzoyl Derivatives. H₅IO₆(1.40 g, 6.16 mmol) was

dissolved in 25 mL of acetonitrile by vigorous stirring, and then CrO₃

(17.6 mg, 0.18 mmol) was dissolved into the mixture to give an orange

solution. The benzyl-derivative (0.88 mmol) was added to the above

solution with stirring. The solution turned to an orange suspension

within a few seconds that turned yellow after a few minutes. The solution

was stirred at room temperature until all starting material was consumed

(TLC control). The solvent was removed in vacuo, and the residue was

purified by flash chromatography to give the corresponding benzoyl-

derivative.

2-(4-Bromobenzoyl)-3-methylnaphthalene-1,4-dione (3a). As

starting material, 1a was used. After chromatography on silica gel (PE:

CH₂Cl₂= 1:3, UV), 133 mg (0.38 mmol, 43% yield) of 3a was isolated as

a yellow solid. mp 170ꢀ171 ꢀC. ¹H NMR (300 MHz, CDCl₃):

δ 8.14ꢀ8.17 (m, 1H), 8.03ꢀ8.06 (m, 1H), 7.73ꢀ7.81 (m, 4H), 7.64

(t, ³J = 2.08 Hz, 1H), 7.61 (t, ³J = 1.95 Hz, 1H), 2.05 (s, CH₃). ¹³C NMR

(75 MHz, CDCl₃): δ 192.72 (C_q), 184.63 (C_q), 183.32 (C_q), 144.30 (C_q),

143.85 (C_q), 134.49 (C_q), 134.30 (CH), 134.19 (CH), 132.48 (CH),

131.87 (C_q), 131.49 (C_q), 130.52 (CH), 130.01 (C_q), 126.78 (CH),

126.44 (CH), 13.60 (CH₃). IR (KBr): ν = 3442 (b, m), 1669 (vs), 1653

(vs), 1627 (vs), 1586 (m), 1568 (m), 1398 (m), 1378 (m), 1329 (s),

1291 (vs), 1272 (s), 1241 (m), 1176 (m), 1069 (m), 1011 (m), 978 (s),

864 (m), 784 (s), 722 (m), 692 cm^ꢀ1(m). EI MS (70 eV, m/z (%)):

353.9 ([M]⁺, 41), 275.0 (100), 182.9 (71), 115.0 (50), 76.0 (41). Anal.

Calcd for C₁₈H₁₁BrO₃: C, 60.87; H, 3.12; Br, 22.50. Found: C, 60.96; H,

3.24; Br, 22.60.

4.1. Chemistry. Melting points were determined on a B€uchi

melting point apparatus and were not corrected. H (300 MHz) and

1

¹³C (75 MHz) NMR spectra were recorded on a Bruker DRX-300

spectrometer; chemical shifts were expressed in ppm relative to tetra-

methylsilane (TMS); multiplicity is indicated as s (singlet), d (doublet),

t (triplet), q (quartet), sep (septet), m (multiplet), cm (centered

multiplet), dd (doublet of a doublet), dt (doublet of a triplet), and td

(triplet of a doublet). C_qindicates a quaternary carbon in the ¹³C NMR

assignation. ¹⁹F NMR was performed using 1,2-difluorobenzene as

external standard (δ = ꢀ139.0 ppm). Intensities in the IR spectra are

indicated as vs (very strong), s (strong), m (medium), w (weak), and b

(broad). Elemental analyses were carried out at the Mikroanalytisches

Laboratorium der Chemischen Fakult€at der Universit€at Heidelberg.

Electron impact mass spectrometry (EI MS) and chemical ionization

mass spectrometry (CI MS), were recorded at facilities of the Institut f€ur

Organische Chemie der Universit€at Heidelberg. Analytical TLC was

carried out on precoated Sil G-25 UV₂₅₄plates from Macherey&Nagel.

Flash chromatography was performed using silica gel G60 (230ꢀ400

mesh) from Macherey&Nagel.

General Procedure for the Silver-Catalyzed Radical Decarboxyla-

tion Reactions of 1,4-Naphthoquinones with Carboxylic Acids. A

solution of menadione or plumbagin (5.81 mmol) and of phenylacetic

acid derivative (11.58 mmol) in 52.5 mL of acetonitrile and 17.5 mL of

water was heated to 85 ꢀC. AgNO₃(90 mg, 0.58 mmol) was added.

(NH₄)₂S₂O₈(1.72 g, 7.54 mmol) in 15 mL of acetonitrile and 5 mL of

water was added dropwise over a period of 45 min and then heated at

reflux for 2 h. The acetonitrile was removed in vacuo. The aqueous phase

was extracted with dichloromethane (4 ꢂ 10 mL), dried over MgSO₄,

and purified by flash chromatography.

2-(4-Bromobenzoyl)-3-(difluoromethyl)-naphthalene-1,4-dione

(8a). As starting material, 7a was used. 40 mg (0.10 mmol, 78% yield) of

1

8a was isolated as a yellow solid. mp = 135ꢀ137 ꢀC. H NMR (300

MHz, CDCl₃): δ 8.20 (dd, 1H, ³J = 7,1 Hz, ⁴J = 1,9 Hz), 8.10 (dd, 1H,

³J = 7,2 Hz, ⁴J = 1,8 Hz), 7.84ꢀ7.89 (m, 2H), 7.75 (d, 2H, ³J = 8,6 Hz),

7.65 (d, 2H, ³J = 8,6 Hz), 6.90 (t, 1H, ²J = 53,3 Hz). ¹³C NMR (75 MHz,

CDCl₃): δ 189.86 (C_q), 183.38 (C_q), 182.44 (C_q), 144.38 (C_q), 135.30

(C_q), 135.28 (CH), 135.18 (CH), 134.34 (C_q), 132.35 (CH), 131.91

(CH), 131.67 (CH), 130.91 (C_q), 130.42 (CH), 130.04 (C_q), 126.91

(CH), 126.87 (CH), 125.70 (C_q), 110.02 (CHF₂). EI MS (70 eV, m/z

(%)): 391.9 ([M]⁺, 26), 326.9 (35), 200.0 (40), 182.9 (100), 154.9 (30),

104.0 (26), 76.1 (25). Anal. Calcd for C₁₈H₉BrF₂O₃: C, 55.27; H, 2.32.

Found: C, 54.92; H, 2.85.

2-(4-Bromobenzyl)-3-methylnaphthalene-1,4-dione (1a). As

starting materials for the radical decarboxylation reaction, menadione

and 4-bromophenylacetic acid were used. After chromatography on

silica gel (PE:CH₂Cl₂= 1:1, UV), 3.10 g (9.12 mmol, 78% yield) of 1a

1

was isolated as a yellow solid. mp 121ꢀ122 ꢀC. H NMR (300 MHz,

CDCl₃): δ 8.03ꢀ8.10 (m, 2H), 7.66ꢀ7.71 (m, 2H), 7.36 (dt, ³J = 8.46

4

3

Hz, J = 1.95 Hz, 2H), 7.09 (d, J = 8.53 Hz, 2H), 3.96 (s, 2H), 2.22

(s, 3H). ¹³C NMR (75 MHz, CDCl₃): δ 185.20 (C_q), 184.54 (C_q), 144.75

(C_q), 144.57 (C_q), 137.06 (C_q), 133.58 (CH), 132.08 (C_q), 131.94

(C_q), 131.71 (CH), 130.32 (CH), 126.50 (CH), 126.35 (CH), 120.31

(C_q), 31.93 (CH₂), 13.31 (CH₃). IR (KBr): ν = 3449 (b, w), 3068 (w),

2962 (w), 1661 (vs), 1624 (m), 1618 (m), 1594 (s), 1486 (s), 1376 (m),

1332 (s), 1315 (s), 1294 (vs), 1071 (m), 1010 (s), 971 (w), 815 (m),

787 (s), 730 (m), 702 (m), 629 (w), 426 cm^ꢀ1(w). EI MS (70 eV, m/z

(%)): 340.1 ([M]⁺, 13), 325.0 (100), 246.1 (63), 215.1 (41), 202.1 (49),

128.1 (72), 76.0 (74). Anal. Calcd for C₁₈H₁₃BrO₂: C, 63.36; H, 3.84.

Found: C, 63.02; H, 3.84.

2-(Difluoromethyl)-1,4-dimethoxynaphthalene 19. The reaction

was conducted in a Teflon bottle under N₂-atmosphere. To a solution of

750 mg (3.47 mmol) of 1,4-dimethoxy-naphthalene-2-carbaldehyde in

10 mL of dry CH₂Cl₂were added 775 μL (950 mg, 5.90 mmol) DAST

and 10 μL (0.17 mmol) ethanol at 0 ꢀC. The reaction mixture was stirred

for 1 h at this temperature and then heated overnight to 40 ꢀC. To run

the reaction to completion, another 140 μL of DAST was added

followed by incubation at 40 ꢀC for additional 5 h. Ten milliliters of

saturated NaHCO₃solution was added in small portions to quench the

reaction. The organic phase was separated, and the aqueous phase was

extracted with CH₂Cl₂(2 ꢂ 20 mL), dried over MgSO₄, and purified by

flash-chromatography on silica gel (PE:CH₂Cl₂= 1:1, UV) to give 723

mg of 19 (3.03 mmol, 88%) of an almost colorless solid. mp 44ꢀ46 ꢀC.

¹H NMR (300 MHz, CDCl₃): δ 8.25ꢀ8.30 (m, 1H), 8.06ꢀ8.11 (m,

1H), 7.53- 7.62 (m, 2H), 7.16 (t, ¹J = 55.8 Hz, 1H, CHF₂), 6.90 (s, 1H),

4.02 (s, 3H), 3.97 (s, 3H). ¹³C NMR (75 MHz, CDCl₃): δ 152.3 (C_q),

148.5 (C_q), 127.9 (C_q), 126.9 (CH), 126.7 (CH), 122.6 (C_q), 122.5

2-(4-Bromobenzyl)-3-(difluoromethyl)naphthalene-1,4-dione (7a).

As starting materials for the radical decarboxylation reaction, difluorome-

nadione and 4-bromophenylacetic acid were used. After chromatography

on silica gel (PE:CH₂Cl₂= 1:1, UV), 132 mg (0.35 mmol, 73% yield) of 7a

1

was isolated as a yellow solid. mp 103ꢀ104 ꢀC. H NMR (300 MHz,

CDCl₃): δ 8.07ꢀ8.13 (m, 1H), 7.99ꢀ8.05 (m, 1H), 7.70ꢀ7.79 (m, 2H),

7.36 (d, ³J = 8.46 Hz, 2H), 7.24 (t, ¹J = 53.87 Hz, 1H, CHF₂), 7.18 (d, ³J =

8.42 Hz, 2H), 4.19 (s, 2H). ¹³C NMR (75 MHz, CDCl₃): δ 184.38 (C_q),

182.79 (C_q), 149.21 (C_q), 136.03 (C_q), 134.41 (CH), 131.76 (C_q), 131.63

(CH), 130.89 (CH), 126.87 (CH), 126.54 (CH), 120.67 (C_q), 110.45

(CHF₂, ¹J = 239.85 Hz), 31.66 (CH₂). IR (KBr): ν = 3436 (b, w), 3100

1

(CH), 122.2 (C_q), 122.0 (CH), 111.7 (t, J = 235 Hz, CHF₂), 98.6

11568

dx.doi.org/10.1021/ja201729z |J. Am. Chem. Soc. 2011, 133, 11557–11571

Journal of the American Chemical Society

ARTICLE

(CH), 63.8 (CH₃), 55.5 (CH₃). ¹⁹F NMR (CDCl₃): δ ꢀ112.37 (d, ¹J =

55.8 Hz). EI MS (70 eV, m/z (%)): 238.07 ([M]⁺, 95), 223.0 (100).

Anal. Calcd for C₁₃H₁₂F₂O₂: C, 65.54; H, 5.08. Found: C, 65.39;

H, 5.20.

(0ꢀ300 μM). The 1,4-naphthoquinone was first dissolved in DMSO,

and a final 1% DMSO concentration was kept constant in the 1,4-

naphthoquinone reductase assay. For the determination of K_mand V_max

values, the steady-state rates were fitted by using nonlinear regression

analysis software (Kaleidagraph) to the MichaelisꢀMenten equation.

From these values, the turnover number k_catand the catalytic efficiency

k_cat/K_mwere calculated.

4.4. Time-Dependent Inactivation of Glutathione Reduc-

tases by Difluoromenadione, 7a and 8a. For determining the

rate constants of human GR inactivation, the residual GR activity was

monitored over time by following an incubation protocol. All reaction

mixtures (final volume of 50 μL) contained 160 μM NADPH, varying

inhibitor concentrations (0ꢀ5 μM in the human GR assay; 0ꢀ50 μM in

the P. falciparum GR assay), GR (6.85 pmol of human GR; 21 pmol of P.

falciparum GR), 2% DMSO in GR buffer at 25 ꢀC. At different time

points (<10 min), 5 μL aliquots of each reaction mixture were removed,

and the residual activity was measured in the standard GSSG reduction

assay at 25 ꢀC (1 mM GSSG and 100 μM NADPH). 2% DMSO was

used in control assays.

4.5. Methemoglobin Reduction in the Presence of Ben-

zoylNQ 3a or 3d or MB upon Reduction by the NADPHꢀGR

System. In an Eppendorf tube containing 6.4 mg of human metHb

dissolved in 885 μL of GR buffer (47 mM potassium phosphate buffer

pH 6.9, 200 mM KCl, 100 mM EDTA), 10 μL of 20 mM 3a dissolved in

DMSO and 100 μL of NADPH dissolved in GR buffer were added. The

reaction was started by the addition of 5 μL of human GR (1.06 μmol)

and then incubated at 37 ꢀC. In a 1 mL cuvette, 20 μL of the reaction

mixture was diluted with 980 μL of GR buffer. UVꢀvis spectral

absorption changes induced by the benzoylNQ 3d, or the controls

MB or CQ, on methemoglobin chromophore were monitored from 300

to 700 nm on a Cary 50 (Varian) spectrophotometer maintained at

37 ꢀC after 0, 5, 10, 20, and 30 min incubation time. The final

concentrations in the reaction mixture were 100 μM metHb, 200 μM

benzoylNQ or MB, 400 μM NADPH, and 1.06 μmol hGR, final DMSO

concentration 1%, 37 ꢀC. No metHb reduction was observed in the

presence of CQ, which was used as negative control.

2-(Difluoromethyl)naphthalene-1,4-dione (or Difluoromenadione).

500 mg (2.10 mmol) of 2-difluoromethyl-1,4-dimethoxy-naphthalene 19

was dissolved in 10 mL of CH₃CN, and a solution of 3.45 g (6.30 mmol) of

CAN in 8 mL of H₂O was added. The reaction mixture was stirred for 15

min at room temperature, the acetonitrile was removed in vacuo, and the

product was extracted with CH₂Cl₂(5 ꢂ 10 mL), dried over MgSO₄, and

purified by flash-chromatography on silica gel (PE:CH₂Cl₂= 1:1, UV) to

give 405 mg of difluoromenadione as a yellow solid (1.95 mmol, 93%). mp

74 ꢀC. ¹H NMR (300 MHz, CDCl₃): δ 8.07ꢀ8.15 (m, 2H), 7.76ꢀ7.83

(m, 2H), 7.19 (m, 1H), 6.83 (t, ¹J = 53.9 Hz, CHF₂, 1H). ¹³C NMR (75

MHz, CDCl₃): δ 183.9 (C_q), 182.4 (C_q), 140.4 (t, ²J = 21.2 Hz, C_q), 135.1

(CH), 134.9 (CH), 134.3 (CH), 134.2 (CH), 131.6 (C_q), 131.3 (C_q),

126.4 (CH), 109.2 (t, ¹J = 240.4 Hz, CHF₂). ¹⁹F NMR (CDCl₃):

1

δ ꢀ123.85 (d, J = 53.9 Hz). EI MS (70 eV, m/z (%)) 208.02 ([M⁺],

100) calcd 208.03. Anal. Calcd for C₁₁H₆F₂O₂: C, 63.47; H, 2.91. Found: C,

63.58; H, 3.02.

4.2. In Vitro Antiparasitic Bioassays. P. falciparum in vitro

culture was carried out using standard protocols³⁹with modifications.^12b

Drug susceptibility of P. falciparum was studied using a modified

method¹⁸of the protocol described previously for the ³H-hypoxanthine

incorporation-based assay.¹⁹All assays included CQ diphosphate (Sigma,

UK) as standard and control wells with untreated infected and unin-

fected erythrocytes. IC₅₀values were derived by sigmoidal regression

analysis (Microsoft xlfit).

Determination of IC₅₀Values against Dd2 P. falciparum Strain. The

IC₅₀was tested by standard in vitro antiproliferation assays based on the

³H-hypoxanthine incorporation.¹⁹Infected erythrocytes in ring stage

(0.5% parasitemia, 2.5% hematocrit) in 96-well plates were exposed to

the compounds for 48 h and then to radioactive hypoxanthine for 24 h.

The amount of radioactivity in precipitable material served as an index of

cell proliferation. Chloroquine was added as reference and displayed an

IC₅₀value of 110 nM.

4.3. Enzymes. Recombinant human and Plasmodium falciparum

glutathione reductases were purified as previously reported.^32a,40One

unit of GR activity is defined as the consumption of 1 μmol of NADPH

per min ε_{340 nm}= 6.22 mM^ꢀ1cm^ꢀ1) under conditions of substrate

saturation. The enzyme stock solutions used for kinetic determinations

were >98% pure as judged from silver stained SDS-PAGE and had

specific activities of 200 U/mg (human GR) and 120 U/mg

(P. falciparum GR). All other reagents were of the highest available

purity and were purchased from Biomol, Boehringer, and Sigma.

Glutathione Disulfide Reduction Assays. The standard assay was

conducted at 25 ꢀC in a 1 mL cuvette. The assay mixture contained

100 μM NADPH and 1 mM GSSG in GR buffer (100 mM potassium

phosphate buffer, 200 mM KCl, 1 mM EDTA, pH 6.9). IC₅₀values were

evaluated in duplicate in the presence of seven inhibitor concentrations

ranging from 0 to 100 μM. Inhibitor stock solutions were prepared in

100% DMSO. 1% DMSO concentration was kept constant in the assay

cuvette. The reaction was started by adding enzyme (8 mU human GR,

6.5 mU P. falciparum GR), and initial rates of NADPH oxidation were

monitored at 340 nm (ε_{340 nm}= 6.22 mM^ꢀ1cm^ꢀ1). Inhibition of GSSG

reduction by 3f was measured as a function of substrate concentration,

and the data were fitted by using a nonlinear regression analysis software

(Kaleidagraph) to the equation for uncompetitive inhibition.^11a

1,4-Naphthoquinone Reductase Activity of P. falciparum Glu-

tathione Reductase. The reduction assay mixture consisted of

100 mM potassium phosphate buffer pH 6.9, 200 mM KCl, 1 mM

EDTA, and 100 μM NADPH in a total volume of 1 mL. 1,4-Naphtho-

quinone reductase activity was determined by recording the initial

velocities in the presence of increasing naphthoquinone concentrations

Direct Methemoglobin Reduction in the Presence of the 3-Benzoyl-

dihydronaphthoquinone 20. The UVꢀvis spectral absorption changes

induced by 40 μM dihydronaphthoquinone 20 on 2 μM methemoglobin

chromophore were directly monitored in the cuvette from 250 to

550 nm on a Cary 300 (Varian) spectrophotometer maintained at

37.5 ꢀC by the flow of a Lauda E200 thermostat. The Scanning Kinetics

application in FLWinLab software was used to monitor the spectral

changes versus time with the following instrumental parameters: spectral

window, 250ꢀ550 nm; number of cycles, 30; cycle time, 45 s; scanning

rate, 200 nm/min; interval, 1 nm.

4.6. Evaluation of Cytotoxicity of the Different Drugs

against Human Cell Lines. Evaluation of the Cytotoxicity against

Human KB Cells. Cytotoxicity on human KB cells (human oral pharyngeal

carcinoma) was evaluated using the resazurin (Alamar Blue) assay as

described.¹⁸The positive control drug was podophyllotoxin (Sigma).

IC₅₀values were calculated as compared to blanks and untreated controls.

Evaluation of the Cytotoxicity against Human MRC-5 Cells. Human

MRC-5_SV2cells are cultured in Earl’s MEM + 5% FCSi. Assays are

performed in 96-well microtiter plates, each well containing about 10⁴

cell/well. After 3 days incubation, cell viability is assessed fluorimetrically

after addition of resazurin and fluorescence is measured (λ_ex550 nm,

λ_em590 nm).¹⁸The results are expressed as % reduction in cell growth/

viability as compared to untreated control wells and IC₅₀is determined.

Compounds are tested at five concentrations (64ꢀ14ꢀ 4ꢀ1ꢀ0.25 μM

or mg/mL). When the IC₅₀is lower than 4 μg/mL or μM, the compound is

classified as toxic.

4.7. Determination of % Parasitemia in CD1Mice. The

in vivo tests in the mouse model were done according to a standard

11569

dx.doi.org/10.1021/ja201729z |J. Am. Chem. Soc. 2011, 133, 11557–11571

Journal of the American Chemical Society

ARTICLE

protocol.⁴¹Compounds were tested in the ANKA P. berghei model by

using the 4-day suppressive test, as indicated by Peters, and using

chloroquine as a positive control. Briefly, naive 18ꢀ20 g CD1 mice were

infected intravenously with 2 ꢂ 10⁶parasitized red cells on day +0. For

administration, compounds were freshly prepared in 10% DMSO in

sterile phosphate-buffered saline the day of use. Two hours postinfec-

tion, mice received the first treatment by the intraperitoneal route. Mice

were further treated on days +1ꢀ3. Blood films from tail blood were

prepared on day +4, and parasitemia was determined by microscopic

examination of Giemsa-stained blood films. Compounds were tested

with a daily dose of 50 or 30 mg/kg by the intraperitoneal or oral route.

Chloroquine treatment p.o. at 10 mg/kg/day was included as a positive

control and resulted in complete inhibition (data not shown). Intraper-

itoneous administrations of CQ have shown similar activity (98.9%

inhibition at 10 mg/kg i.p). Mice were treated, and levels of parasitemia

were determined as described.

Dr. Hervꢀe Vezin for the residual data of cyclic voltammetry

carried out in the experiments previously reported,^11band Prof.

Louis Maes, Antwerp University, for the primary cytotoxicity

assays with the human MRC-5 cell line.

’ REFERENCES

(1) White, N. J. Lancet 2010, 376, 2051–2052.

(2) Zarchin, S.; Krugliak, M.; Ginsburg, H. Biochem. Pharmacol.

1986, 35, 2435–2442.

(3) (a) Abu-Bakar, N.; Klonis, N.; Hanssen, E.; Chan, C.; Tilley, L.

J. Cell Sci. 2010, 123, 441–450. (b) Ncokazi, K. K.; Egan, T. J. Anal.

Biochem. 2005, 338, 306–319. (c) Cohen, S. N.; Phifer, K. O.; Yielding,

K. L. Nature 1964, 202, 805–806. (d) Egan, T. J. J. Inorg. Biochem. 2006,

100, 916–926. (e) Egan, T. J. J. Inorg. Biochem. 2008, 102, 1288–1299.

(4) Dorn, A.; Stoﬀel, R.; Matile, H.; Bubendorf, A.; Ridley, R. Nature

1995, 374, 269–271.

(5) Loria, P.; Miller, S.; Foley, M.; Tilley, L. Biochem. J. 1999, 339,

363–370.

(6) (a) Atamna, H.; Ginsburg, H. J. Biol. Chem. 1995, 270, 24876–

83. (b) Ginsburg, H.; Famin, O.; Zhang, J.; Krugliak, M. Biochem.

Pharmacol. 1998, 56, 1305–1313.

(7) Vippagunta, S. R.; Dorn, A.; Ridley, R. G.; Vennerstrom, J. L.

Biomed. Biochim. Acta 2000, 1475, 133–140.

(8) Hogg, T.; Nagarajan, K.; Herzberg, S.; Chen, L.; Shen, X.; Jiang,

H.; Wecke, M.; Blohmke, C.; Hilgenfeld, R.; Schmidt, C. L. J. Biol. Chem.

2006, 281, 25425–25437.

4.8. Morphological Examination of P. falciparum Tropho-

zoite Development Inhibition by Compound 1c. The FcB1

strain of P. falciparum was routinely grown in human erythrocytes and

RPMI medium supplemented with 8% heat-inactivated human plasma

under a 91% N₂, 6% O₂, and 3% CO₂atmosphere. Parasites were

synchronized on a 0ꢀ5 h time window by successive Plasmion⁴²and

sorbitol⁴³treatments. Doseꢀresponse assay based upon [³H]-hypox-

anthine incorporation by growing parasites¹⁹was used to determine the

inhibitory concentrations of compound 1c. Shortly, 16ꢀ21 h-old

parasites (trophozoites) were incubated for 48 h in the presence of

decreasing concentrations of compound 1c and 0.25 μCi [³H]-hypox-

anthine. IC₅₀and IC_{100 min}values were determined from the resulting

doseꢀresponse curves; they correspond to the drug concentration

required to observe 50% and 100% of the inhibition of the parasitic

growth, respectively (min means minimal drug concentration). For

analysis of compound 1c-induced morphological alterations, tropho-

zoites (16ꢀ21 h-old) were grown for 25 h at 1% parasitemia and 2%

(9) Monti, D.; Vodopivec, B.; Basilico, N.; Olliaro, P.; Taramelli, D.

Biochemistry 1999, 38, 8858–8863.

(10) (a) Schirmer, R. H.; Mueller, J. G.; Joachim, G.; Krauth-Siegel,

R. L. Angew. Chem., Int. Ed. Engl. 1995, 34, 141–154. (b) Krauth-Siegel,

R. L.; Bauer, H.; Schirmer, R. H. Angew. Chem., Int. Ed. 2005, 44,

690–715.

(11) (a) Biot, C.; Bauer, H.; Schirmer, R. H.; Davioud-Charvet, E.

J. Med. Chem. 2004, 47, 5972–5983. (b) Bauer, H.; Fritz-Wolf, K.;

Winzer, A.; K€uhner, S.; Little, S.; Yardley, V.; Vezin, H.; Palfey, B.;

Schirmer, R. H.; Davioud-Charvet, E. J. Am. Chem. Soc. 2006, 128,

10784–10794. (c) Morin, C.; Besset, T.; Moutet, J.-C.; Fayolle, M.;

Br€uckner, M.; Limosin, D.; Becker, K.; Davioud-Charvet, E. Org. Biomol.

Chem. 2008, 6, 2731–2742.(d) Davioud-Charvet, E.; Lanfranchi, D. A.

In Apicomplexan Parasites ꢀ Molecular Approaches toward Targeted Drug

Development; Becker, K., Selzer, P. M., Eds.; Wiley-VCH: Weinheim,

2011; Vol. 2 (Drug Discovery in Infectious Diseases).

(12) (a) Davioud-Charvet, E.; Delarue, S.; Biot, C.; Schw€obel, B.;

B€ohme, C. C.; M€ussigbrodt, A.; Maes, L.; Sergheraert, C.; Grellier, P.;

Schirmer, R. H.; Becker, K. J. Med. Chem. 2001, 44, 4268–4276.

(b) Friebolin, W.; Jannack, B.; Wenzel, N.; Furrer, J.; Oeser, T.; Sanchez,

C. P.; Lanzer, M.; Yardley, V.; Becker, K.; Davioud-Charvet, E. J. Med.

Chem. 2008, 51, 1260–1277.

hematocrit in culture medium supplemented with 5 μM (IC_{100 min}

)

compound 1 or 5 μM chloroquine. Control for normal growth was in

culture medium alone. Diff Quick-stained smears were made every 2 h

during 12 h and then after 25 h, and were analyzed by optical examination.

’ ASSOCIATED CONTENT

S

Supporting Information. Detailed experimental proce-

b

dures, spectroscopic data, and elemental analyses for the pre-

paration and the characterization of the new compounds 1ꢀ20,

and cytotoxicity activities against human cells. This material is

available free of charge via the Internet at http://pubs.acs.org.

’ AUTHOR INFORMATION

(13) (a) Subba-Rao, R. V.; Alexander, M. Appl. Environ. Microbiol.

1985, 49, 509–516. (b) Segrestaa, J.; Verite, P.; Estour, F.; Menager, S.;

Lafont, O. Chem. Pharm. Bull. 2002, 50, 744–748.

Corresponding Author

elisabeth.davioud@unistra.fr

(14) (a) Jaﬀꢀe, E. R.; Neurmann, G. Nature 1964, 202, 607–608.

(b) Beutler, E.; Baluda, M. C. Blood 1963, 22, 323–233.

(15) Kochi, J. K.; Anderson, J. M. J. Am. Chem. Soc. 1970, 92,

1651–1659.

Author Contributions

^#These authors contributed equally.

(16) Yamazaki, S. Org. Lett. 1999, 1, 2129–2132.

’ ACKNOWLEDGMENT

(17) Evans, P. A.; Brandt, T. A. J. Org. Chem. 1997, 62, 5321–5326.

(18) O’Brien, J.; Wilson, I.; Orton, T.; Pognan, F. Eur. J. Biochem.

2000, 267, 5421–5426.

(19) Desjardins, R. E.; Canﬁeld, C. J.; Haynes, J. D.; Chulay, J. D.

Antimicrob. Agents Chemother. 1979, 16, 710–718.

(20) (a) Hanoulle, X.; Wieruszeski, J. M.; Rousselot-Pailley, P.;

Landrieu, I.; Locht, C.; Lippens, G.; Baulard, A. R. J. Antimicrob.

Chemother. 2006, 58, 768–772. (b) Hanoulle, X.; Wieruszeski, J. M.;

Rousselot-Pailley, P.; Landrieu, I.; Baulard, A. R.; Lippens, G. Biochem.

Biophys. Res. Commun. 2005, 331, 452–458.

E.D.-C. is a delegate of Centre National de la Recherche

Scientiﬁque, in the frame of a FrenchꢀGerman cooperation with

Heidelberg University, Germany. E.D.-C. thanks DFG via the

SFB 544 (B14 project) and CNRS for ﬁnancial support. R.H.S.

and M.L. thank DFG via the SFB 544 (B2 and B12 projects).

T.M. thanks Prof. Thomas J. J. M€uller for fruitful discussions during

the P.h.D thesis work. We are grateful to Dr. Mourad Elhabiri,

Strasbourg University, for the direct metHb reduction experiment,

11570

dx.doi.org/10.1021/ja201729z |J. Am. Chem. Soc. 2011, 133, 11557–11571

Journal of the American Chemical Society

ARTICLE

(21) H€unig, S.; Bau, R.; Kemmer, M.; Meixner, H.; Metzenthin, T.;

Peters, K.; Sinzger, K.; Gulbis, J. Eur. J. Org. Chem. 1998, 335–348.

(22) Kitz, R.; Wilson, I. B. J. Biol. Chem. 1962, 237, 3245–3249.

(23) Gulino, D.; Pham, Q. T.; Golꢀe, J.; Pascault, J. P. Anionic

Polymerization; ACS Symposium Series; American Chemical Society:

Washington, DC, 1981; Vol. 166, Chapter 20, pp 307ꢀ326.

(24) Jain, M.; Vangapandu, S.; Sachdeva, S.; Singh, S.; Singh, P. P.;

Jena, G. B.; Tikoo, K.; Ramarao, P.; Kaul, C. L.; Jain, R. J. Med. Chem.

2004, 47, 285–287.

(25) Cappellini, M. D.; Fiorelli, G. Lancet 2008, 371, 64–74.

(26) Ginsburg, H.; Atamna, H.; Shalmiev, G.; Kanaani, J.; Krugliak,

M. Parasitology 1996, 113, 7–18.

(27) (a) Zhang, Y. A.; Hempelmann, E.; Schirmer, R. H. Biochem.

Pharmacol. 1988, 37, 855–860. (b) Zhang, Y.; K€onig, I.; Schirmer, R. H.

Biochem. Pharmacol. 1988, 37, 861–865.

(28) L€uersen, K.; Walter, R. D.; M€uller, S. Biochem. J. 2000, 346,

545–552.

(29) Cappadoro, M.; Giribaldi, G.; O’Brien, E.; Turrini, F.; Mannu,

F.; Ulliers, D.; Simula, G.; Luzzatto, L.; Arese, P. Blood 1998, 92, 2527–

2534.

(30) Gallo, V.; Schwarzer, E.; Rahlfs, S.; Schirmer, R. H.; van Zwieten,

R.; Roos, D.; Arese, P.; Becker, K. PLoS One 2009, 4, e7303.

(31) Baird, J. K.; Surjadjaja, C. Trends Parasitol. 2011, 27, 11–16.

(32) (a) F€arber, P. M.; Arscott, L. D.; Williams, C. H., Jr.; Becker, K.;

Schirmer, R. H. FEBS Lett. 1998, 422, 311–314.(b) Schirmer, M.;

Scheiwein, M.; Gromer, S.; Becker, K.; Schirmer, R. H. In Flavins and

Flavoproteins; Ghisla, S., Kroneck, P., Macheroux, P., Sund, H., Eds.;

Rudolf Weber, Agency for Scientiﬁc Publications: Berlin, 1999; pp

857ꢀ862. (c) Kasozi, D.; Gromer, S.; Adler, H.; Zocher, K.; Rahlfs,

S.; Wittlin, S.; Fritz-Wolf, K.; Schirmer, R. H.; Becker, K. Redox Rep.

2011, in press.

(33) Davioud-Charvet, E.; M€uller, T.; Bauer, H.; Schirmer, R. H.

Patent EP 08290278.4 (March 26, 2008); patent PCT/EP2009/053483

(March 25, 2009).

(34) Cohen, M. S.; Chai, Y.; Britigan, B. E.; McKenna, W.; Adams, J.;

Svendsen, T.; Bean, K.; Hassett, D. J.; Sparling, P. F. Antimicrob. Agents

Chemother. 1987, 31, 1507–1513.

(35) (a) Anstey, N. M.; Hassanali, M. Y.; Mlalasi, J.; Manyenga, D.;

Mwaikambo, E. D. Trans. R. Soc. Trop. Med. Hyg. 1996, 90, 147–151.

(b) Srivastava, S.; Alhomida, A. S.; Siddiqi, N. J. In Vivo 2000, 14,

547–550.

(36) Brabin, B. J.; Eggelte, T. A.; Parise, M.; Verhoeﬀ, F. Drug Saf.

2004, 27, 633–48.

(37) Vale, N.; Moreira, R.; Gomes, P. Eur. J. Med. Chem. 2009,

44, 937–953.

(38) Liu, H.; Walker, L. A.; Nanayakkara, N. P.; Doerksen, R. J. J. Am.

Chem. Soc. 2011, 133, 1172–75.

(39) Trager, W.; Jensen, J. B. Science 1976, 193, 673–675.

(40) Nordhoﬀ, A.; Bucheler, U. S.; Werner, D.; Schirmer, R. H.

Biochemistry 1993, 32, 4060–4066.

(41) Peters, W.; Robinson, B. L. Handbook of Animal Models of

Infection; Academic Press: London, 1999; pp 756ꢀ771.

(42) Pasvol, G.; Wilson, R. J.; Smalley, M. E.; Brown, J. Ann. Trop.

Med. Parasitol. 1978, 72, 87–88.

(43) Lambros, C.; Vanderberg, J. P. J. Parasitol. 1979, 65, 418–420.

11571

dx.doi.org/10.1021/ja201729z |J. Am. Chem. Soc. 2011, 133, 11557–11571

Article Doi

Glutathione reductase-catalyzed cascade of redox reactions to bioactivate potent antimalarial 1,4-naphthoquinones - A new strategy to combat malarial parasites

DOI: 10.1021/ja201729z

Source and publish data:

Authors:

Article abstract of DOI:10.1021/ja201729z

Full text of DOI:10.1021/ja201729z

Products guided by the article

R&D Labs maybe for 1189394-17-9

Relevant to this article

Hot Product