Synthesis and Antischistosomal Activity Profile of Several Sub-Series of Redox-Active Lead 3‑BenzylmenadionesInvestigations of the Modes of Action in Yeast

Abstract

Three sub-series of 3-benzylmenadiones were discovered as new antischistosomal agents. The most potent representatives, called schistodiones, exhibited IC₅₀ values <5 μM against larvae and ex vivo Schistosoma mansoni adult worms and moderate activity per os in the S. mansoni-infected mouse model in vivo. We synthesized the putative 3-benzoylmenadione metabolites of the most active compounds and explored the possible mode(s) of action using the yeast model and the recombinant drug target S. mansoni thioredoxin-glutathione reductase. Two drug targets were identified in yeast: the yeast NADH dehydrogenase and Cox15, involved in events responsible for respiratory growth defect in yeast, respectively: (i) the bioreductive activation step producing reactive oxygen species and (ii) the heme a synthase inhibition. These two drug targets are discussed in the context of S.mansoni biology, with S. mansoni thioredoxin–glutathione reductase redox-cycling and a potential source of oxidative stress in the parasite.

Introduction

Schistosomiasis, a neglected tropical disease caused by parasitic flatworms of the genus Schistosoma, affects over 250 million people worldwide, primarily in tropical and subtropical regions where basic sanitation is lacking. Also known as bilharziasis, schistosomiasis results in the second most morbidity and mortality from a human parasitic disease in the world, after malaria. This chronic and often debilitating illness causes as many as 200,000 deaths annually. The main Schistosoma species infecting humans are Schistosoma haematobium, Schistosoma japonicum, and Schistosoma mansoni. Chronic schistosomiasis can lead to significant health complications, including liver fibrosis, portal hypertension, hepatosplenomegaly, pulmonary arterial hypertension, and bladder cancer. Acute infections present as transient urticaria rash, allergic pneumonia, and Katayama syndrome. , The disease is transmitted when individuals are exposed to freshwater contaminated by schistosome cercariae, which are released by infected snails. Cercariae penetrate the skin of human hosts, enter the circulatory system, migrate to the liver via the lungs, and develop into adult worms in the mesenteric veins (S. mansoni and S. japonicum) or the venous plexuses around the urinary bladder (S. haematobium). These worms produce eggs, some of which become entrapped within host tissues, leading to significant pathology. , Currently, praziquantel (PZQ), a class II compound, is the sole clinically available drug for treating schistosomiasis. − While praziquantel shows high activity against adult worms, it lacks activity against juvenile worms, leading to concerns about the emergence of resistant strains. High doses are required due to low solubility and first-pass metabolism issues, which complicates administration, particularly in children. The bitter taste and large tablet size further complicate treatment adherence in children, although recently a pediatric formulation of PZQ has become available. , Moreover, high dosages of PZQ have been associated with side effects such as abdominal pain, nausea, and allergy. Alternatives to PZQ, such as (±)-oxamniquine and metrifonate, have limitations due to restricted activity against specific Schistosoma species (S. haematobium and S. mansoni, respectively), ineffectiveness, emerging resistance, and toxicity and hence are no longer marketed. It is thus crucial to develop cost-effective alternatives and complementary therapies, including novel drugs or analogs targeting schistosome-induced liver fibrosis. Artemisinin, a drug derived from the sweet wormwood plant and widely used in malaria treatment, has also shown potential against schistosome infections by generating reactive oxygen species (ROS) that attack multiple parasite targets, including SERCA, and trigger an antioxidant response. ,, Research indicates that artemisinin and its derivatives can complement praziquantel by targeting different stages of the parasite’s lifecycle, potentially offering a synergistic approach to treatment and addressing issues of drug resistance.

So far, our group has pioneered the biological chemistry and the molecular understanding of the 3-benzylmenadione (benzylMD) class of electrophores described as redox-active agents that catalyse a range of NADPH-consuming reactions responsible for the antiparasitic activities of the series. − Previous studies on the mode of action (MoA) of benzylMD revealed that this scaffold disturbs the redox balance of protozoan parasites by acting as a subversive substrate of NADPH-dependent oxido-reductases. The first success was achieved with the discovery of the potent early antimalarial lead benzylMD named plasmodione (PD), which is active against young asexual stages of Plasmodium falciparum across strains with varying degrees of multidrug resistance, as well as against sexual stages (early gametocytes). PD presents a safe toxicological profile for potential human use, including in glucose-6-phosphate-dehydrogenase (G6PD)-deficient populations. , We have shown that the antimalarial selectivity of benzylMD comes largely from its specific bioactivation within Plasmodium parasitized-red blood cells through a cascade of NADPH-dependent redox reactions generating several key metabolites (Figure ). Our approach of selectively inducing oxidative stress specifically in parasitized human erythrocytes has been validated as a novel and efficient strategy to combat malarial parasites, mimicking the conditions encountered by Plasmodium developing in G6PD-deficient red blood cells from populations protected against cerebral malaria.

1.

Mechanism of action of antiparasitic 3-benzylmenadiones via a cascade of redox reactions involving NADPH-dependent oxido-reductases in the bioactivation (step 1 leading to 1 e- or 2 e- reduced species NQH + NQH₂), redox-cycling, and interplay with hemoglobin catabolism leading to oxidative stress and death of parasites (step 2). With several representatives of antischistosomal benzylMD, called schistodiones (SD), and benzoylMD metabolites, called schistodione oxides (SDO), a study of the mode of action was investigated in yeast, which revealed mitochondrial NADH dehydrogenases as the key enzymes for SD bioactivation and enhanced oxidative stress. Furthermore, the heme a synthase Cox15 has been identified as a key enzyme in yeast inhibited by reduced schistodione oxide (SDO_red).

In parallel, our team and partners developed and tested libraries of benzylMD analogues against the blood-dwelling S.mansoni, which digests host hemoglobin and shares an analogous heme detoxification pathway with malarial parasites. By screening these libraries across a broad panel of assays, we identified 44 antischistosomal benzylMDs classified in three sub-series and identified preliminary structure-activity relationships. The tests included experimental characterization of the anti-S. mansoni activity (IC₅₀ values) in both the newly transformed schistosomula (NTS) and adult worm assays, along with cytotoxicity assessment (CC₅₀) and calculated pharmacokinetic properties of the new benzylMD derivatives. In addition, a peculiar phenotypic effect was observed in the worms exposed to these compounds.

In addition, we explored the MoA of the antischistosomal benzylMDs in yeast model to test whether these compounds share, at least in part, the MoA of a antiplasmodial PD. Using the yeast model, we observed that (i) the flavoenzyme NADH-dehydrogenases plays a key role in bioactivation of antischistosomal benzylMDs derivatives, and (ii) the heme a synthase Cox15 may represent a new potential target of these antischistosomal redox-active agents. Eleven benzoylMDs were also prepared as putative schistodione metabolites and tested as substrates of yeast mitochondrial NADH-dehydrogenases. Their effective interactions on the NADH-driven cytochrome c reduction independent of complex III revealed that redox-cycling reactions produced superoxide anion radicals, which would contribute to respiratory growth defect in yeast.

Finally, we observed that purified S. mansoni thioredoxin-glutathione reductase (SmTGR) could react with benzoylMDs in a redox-cycle regenerating hemoglobin (Fe²⁺) when methemoglobin was used as electron acceptor for reduced benzolMD metabolites.

Results and Discussion

Chemistry

Synthesis of 2,5-, 3,5-Dimethoxy-benzylmenadione Analogues and Control

Based on previous research, , two series of 3-benzylmenadiones with various dimethoxy-substituted aromatic groups were synthesized, starting from precursors with a hydrogen atom (menadione, 1a), a fluorine atom (1b) or a CF₃ group (1c) at the 6-position of the menadione moiety (Scheme , Route A, sub-series I). Fluorinated functions were introduced on the molecular scaffold to enhance metabolization resistance, , as demonstrated in antimalarial action studies in vivo, or to improve lipophilicity for better worm tegument penetration compared to their non-fluorinated derivatives.

1. Synthesis of 2,5- and 3,5-Dimethoxy Substituted 3-Benzylmenadiones 2a–c, 3a–c and Control 4c (Sub-Series I) and of 3-Fluoro-4-nitro Substituted 3-Benzylmenadiones 5a–c (Sub-Series II)­ .

^a Reagents and conditions: (a) phenylacetic acid, AgNO₃, (NH₄)₂S₂O₈, MeCN/H₂O (3:1 v/v), 95 °C, 3 h; (b) 3-fluoro-4-nitrobenzyl bromide, 2,6-lutidine, γ-terpinene, Fe­(acac)₃, MeCN, 90 °C, hv (365 nm), 24 h.

The 2,5- and 3,5-dimethoxy substituted benzylMDs series (compounds 2a–c and 3a–c, respectively) were obtained through a Kochi-Anderson radical decarboxylation using 2,5- or 3,5-dimethoxybenzoic acids, both commercially available, in the presence of AgNO₃ and (NH₄)₂S₂O₈, with yields ranging from 43 to 80%. , The mono-methoxy 2-substituted benzylMD 4c was synthesized, in the same way, as a control for derivatives 2c and 3c.

Synthesis of 3-Fluoro-4-nitro Analogues

The nitro group (NO₂) is essential for the antiparasitic activity of known nitroheteroaromatic drugs, such as oxamniquine and niridazole, with observed antischistosomal activity in vivo. , The NO₂ moiety induces redox reactions within cells, leading to toxicity and the subsequent death of protozoan parasites and multicellular organisms like parasitic helminths. This effect can also occur in humans, making nitro groups both pharmacophores and toxicophores. Additionally, the nitro group significantly influences both the polarity and the electronic properties of the resulting molecules, enhancing interactions with certain amino acids in proteins. In light of these observations and with the aim of combining the properties of nitro group with those of fluorinated groups, benzylMDs bearing these two functions were synthesized following synthesis methodologies previously described in the laboratory.

The synthesis of compounds 5a–c, based on a Fe­(III)-catalyzed photoredox benzylation of 1,4-naphthoquinones derivatives 1a–c, is shown in Scheme , Route B sub-series II. UV irradiation (365 nm) of a media composed of 3-fluoro-4-nitrobenzyl bromide, 2,6-lutidine, γ-terpinene, Fe­(acac)₃ and ambient oxygen afforded the corresponding benzylMD derivatives 5a–c with moderate yields.

Synthesis of 4-Nitro-3-(oxetan-3-yloxy) Analogues

In order to reduce the lipophilicity of the compounds with methoxy units while retaining the benefits of adding a nitro function to the benzylMD scaffold and its properties, the replacement of methoxy with an O-oxy-oxetanyl ether group was considered and benzylMDs bearing nitro and oxetanyl ether moieties were synthesized. In the last decade, four-membered heterocycles/rings have emerged as beneficial motifs in drug discovery because of their low molecular weight, high polarity, drug metabolism resistance, and tridimensionality, which can improve physicochemical properties including target affinity and aqueous solubility. − Substituted benzylMDs 8a–c were prepared via a convergent synthetic approach (Scheme ) starting from precursors 6a–c (see Scheme S1 in the Supporting Information). These intermediates were involved in a pallado-catalysed Suzuki-Miyaura coupling with boronic ester S4a, in basic media, leading to compounds 7a–c with yields ranging from 59 to 91% (Scheme ). Then, the final analogues 8a–c were obtained by oxidative demethylation with a cerium­(IV) ammonium nitrate (CAN) in excellent yields. The benzylMDs 10a–e, containing two oxetane moieties, were obtained in a similar manner using various boronic esters S4b–e (see Scheme S2 in the Supporting Information). Compounds 9a–e, obtained with yields ranging from 44% to 81%, were then oxidized using CAN to form the final products 10a–e with excellent yields exceeding 82%.

2. Synthesis of 4-Nitro-3-(oxetan-3-yloxy) Substituted 3-Benzylmenadiones 8a–c and Bis­(3-(oxetan-3-yloxy)) Substituted 3-Benzylmenadiones 10a–e .

^a Reagents and conditions: (a) S4a-d, Na₂CO₃, Pd­(PPh₃)₄, DME/H₂O (3:1 v/v), reflux, 1 h. (b) CAN, MeCN/H₂O, (5:1 v/v), rt, 30 min.

Synthesis of Various Substituted BenzylMD Derivatives 11a- o / m / p to 15a- o / m / p

Existing published data and reported results from us have highlighted the value of adding cyano and carboxylic acid functions to increase uptake by adult worms and improve their antischistosomal action. ,, However, their antischistosomal properties can be improved by masking the polarity of the carboxylic acid, e.g., by an ester or amide function, but also through an approach based on replacing the acid with a bioisosteric tetrazole. This modification would increase the compound’s bioavailability and metabolic stability, while maintaining a size and acidity similar to those of the previous analogue. , In the light of these observations, two series of benzylMDs involving various aromatic groups substituted by a tetrazole unit were synthesized. The latter was considered directly on the benzyl ring (‘aryl’ derivatives, Scheme , path A), but also spaced by a carbon atom from the same ring (bis­(benzyl) derivatives, Scheme , path B). In each family, the three ortho/meta/para (o/m/p) regioisomers were synthesized.

3. Synthesis of Aryl (Route A) and Benzyl (Route B) Tetrazole-Substituted 3-Benzylmenadiones and Intermediates .

^a (A) Reagents and conditions for compounds 11-15: (a) Benzoic acid, AgNO₃, (NH₄)₂S₂O₈, MeCN/H₂O (3:1 v/v), 95 °C, 3 h. (b) i. SOCl₂, reflux, 2 h. ii. 3-aminopropionitrile, NEt₃, THF, 25 °C, 18 h. (c) NaN₃, Tf₂O, DCM, 25 °C, 18 h. (d) LiOH·H₂O, THF/H₂O (3:1 v/v), 25 °C, 1 h. n.d.: not detected. (B) Reagents and conditions for compounds 18- and 19-o/m/p : (a) Diacetic acid, AgNO₃, (NH₄)₂S₂O₈, MeCN/H₂O (3:1 v/v), 95 °C, 3 h. (b) i. SOCl₂, DMF, DCM, 25 °C, 2 h. ii. 3-aminopropionitrile, NEt₃, DCM, 25 °C, 18 h. (c) NaN₃, Tf₂O, DCM, 25 °C, 18 h. (d) LiOH.H₂O, THF/H₂O (3:1 v/v), 25 °C, 1 h.

In this top-down approach, cyano-‘aryl’ derivatives 11a-o/m/p and 11c- o , benzoic acid derivatives 12-o/m/p were first synthesized through a Kochi Anderson reaction under the same conditions as described above. All benzylMDs 11 and 12 were obtained in excellent yields (over 77%), except for compound 12- o due to the proximity of the two acid functions and the resulting steric hindrance. Then, acids 12-o/m/p were converted to amides via the formation of an acyl chloride intermediate, more reactive than its carboxylic acid counterpart, using SOCl₂ at reflux. Amides 13-o/m/p , synthesized in yields ranging from 70 to 89%, were subsequently treated with NaN₃ under mild conditions (room temperature) to give the corresponding tetrazoles 14- m / p in yields below 10%. During this reaction step, 4 equiv of Tf₂O were used to limit the addition of sodium azide to the benzylMD ketones. Finally, basic β-elimination of the propionitrile moiety in tetrazoles 14- m / p with LiOH.H₂O afforded corresponding aryl tetrazole derivatives 15- m / p in excellent yields. To complement ‘aryl’ compounds 11–15, bis­(benzyl) derivatives 16-19-o/m/p (Scheme , path B) were prepared via the same linear synthetic sequence as previously described, with only minor modifications, starting from menadione 1a. After Kochi-Anderson radical decarboxylation of phenyl acetic acids with various dibenzoic acids and amide functions (following coupling reactions from acids 16-o/m/p with 3-aminopropionitrile in basic media), amide derivatives 17-o/m/p were obtained in moderate to good overall yields (27 to 66%). Then, tetrazole formation with NaN₃ and Tf₂O gave protected tetrazoles 18 in moderate yields except for compound 18- m exhibiting a high rate of degradation. Final basic β-elimination of propionitrile moieties afforded quasi-quantitatively target tetrazole derivatives 19-o/m/p .

Synthesis of Substituted (±) 4a-ψ-BenzylMD Derivative 24

Pseudo-1,4-naphthoquinones (Ψ-NQ) are non-quinoid compounds in which aromaticityfound in 1,4-naphthoquinonesis broken by the introduction of an angular methyl at C-4a or -8a. From a medicinal chemistry point of view, the loss of planarity of the scaffold improve the aqueous solubility of the final compound. This series was also designed to circumvent the bad reputation of quinones in the pharmaceutical industry by developing prodrugs of 1,4-naphthoquinones, bioactivable in an oxidative environment. Although there is one CYP450 gene in the S. mansoni genome, Schistosoma worms digest large amounts of hemoglobin and release free heme, crystallizing it as hemozoin granules in the gut. In principle, both CYP450 and free heme catalyze oxidation reactions and metabolize drugs, such as benzylMD, by oxidation at different loci in the benzylMD core. Our recent work on the design of 4a- and 8a-Ψ-plasmodione regioisomers has shown that these compounds possess an increased aqueous solubility (Table S1, supporting information) and very high antischistosomal action in the NTS and adult worm assays.

Based on our previous work about the regioselective synthesis of potential non-quinonoid prodrugs of plasmodione, 4a-Ψ-schistodione regioisomer 24 was synthesized through a linear five-step approach (Scheme ). Following the functionalization of the bromo-quinone 20 via a Kochi-Anderson benzylation reaction, leading to derivative 21 with 69% yield, and two subsequent reduction/protection steps, the dimethoxy-bromobenzene 22 was obtained with 75% yield. A halogen-to-boron exchange was then performed using n-BuLi and B­(OⁱPr)₃, directly generating a reactive intermediate, which, without prior isolation, underwent oxidative demethylation to afford the borylated compound 23 with 25% yield over two steps. Finally, the target Ψ-benzylMD, called 4a-Ψ-schistodione 24 was synthesized via a Diels-Alder cycloaddition between 23 and (3E)-4-methoxybuta-1,3-diene, with an excellent yield of 86%.

4. Synthesis of (±) 4a-ψ-Benzylmenadione Derivative 24 .

^a Reagents and conditions: (a) 3,5-dimethoxyphenylacetic acid, AgNO₃, (NH₄)₂S₂O₈, MeCN/H₂O (3:1 v/v), 95 °C, 3 h, 69%. (b) i. SnCl₂, HCl, MeOH, reflux, 30 min ii. DMS, KOH, Acetone/MeOH (5:1 v/v), 1 h, 75%. (c) n- BuLi, B­(OⁱPr)₃, THF, −78 to 25 °C, 16 h. (d) CAN, MeCN/H₂O (3:1 v/v), 25 °C, 1 h, 25% over 2 steps. (e) (3E)-4-methoxybuta-1,3-diene, DCM, −20 to 25 °C, 86%.

Synthesis of 3-BenzoylMD 25-30 as Potential Metabolites of 3-BenzylMD

The potential metabolites of schistodiones (SD) A_2,5 and SD A_3,5 (benzylMD 2a and 3a, respectively) were previously produced as pure benzoylMD references 25 and 26, respectively SDO A_2,5 and SDO A_3,5, for investigations on drug targets in the enzymic assays with recombinant pure SmTGR and/or docking studies (Figure ). In this study, we showed that benzoylMDs 25 and 26 act as subversive substrates of SmTGR.

To investigate the MoAs of antischistosomal benzylMD 13–14 and 17–18 in a cellular model (yeast), the benzoylMD analogues from the corresponding benzylMD precursors 13- o / m / p , 14- m / p , 17- o / m/p , and 18- o / m / p were oxidized to their benzoyl forms 27- m / p , 28- m / p , 29- o / m / p , and 30- o / m / p via a photoredox reaction developed in our laboratory. , Menadione is known to undergo reduction to dihydro-menadione through a semiquinone intermediate under photoreductive conditions in isopropanol (hydrogen donor), and benzylic oxidation can occur via O₂ insertion into the reduced species. After 72 h of irradiation at 350 nm, benzylation oxidation took place affording the benzoyl species 27- m / p , 28- m / p , 29- o / m / p , and 30- o / m / p , with yields ranging from 58% to 85% (Scheme , Route A). However, compound 27- o could not be obtained under these conditions, likely due to the electron-deficient benzylic position resulting from the presence of an electron-withdrawing amide group in the ortho position. To address this limitation, the synthesis of compound 27- o was carried out in a linear manner as described, starting from known material S1a through a Friedel-Crafts acylation (Scheme , Route B). In this bottom-up approach, a Friedel-Crafts acylation reaction led to compound 31 with a limited yield of 15%. The methyl ester was then hydrolyzed under basic conditions in the presence of LiOH.H₂O, leading to the corresponding acid 32 with a yield of 80%. The latter was subsequently oxidized using an excess of CAN, and the carboxylic acid moiety was converted into an amide via the formation of an acyl chloride intermediate using SOCl₂, ultimately affording compound 27- o with a very low yield of 9% over two steps.

5. Synthesis of Substituted 3-Benzoylmenadiones Derivatives 27–30 .

^a (A) Reagents and conditions for compounds 27- m / p , 28- m / p , 29- o / m / p and 30- o / m / p : (a) O₂, hv (350 nm), DCM/ⁱPrOH (3:1 v/v), 17 °C, 72 h. (B) Reagents and conditions for compound 27- o : (a) 2-(methoxycarbonyl)­benzoic acid, TFAA, TfOH, DCM, 25 °C, 16 h, 15%. (b) LiOH.H₂O, MeOH/H₂O (5:1 v/v), reflux, 18 h, 80%. (c) CAN, MeCN/H₂O (3:1 v/v), 25 °C, 1 h. (d) i. SOCl₂, reflux, 2 h. ii. 3-aminopropionitrile, NEt₃, DCM, 25 °C, 18 h, 9% over 2 steps.

Structure–Anti–S. mansoni Activity Relationships and Cytotoxicity

All compounds were classified in three sub-series: I to III (Figure ).

2.

Structures of the 3-benzylMD sub-series I to III prepared in this work.

Data from Tables S1 and S2 describe the SAR profile of the 3-benzylMD sub-series I to III in two parasitic S. mansoni assays, NTS (Figure A and Table S2 in the Supporting information) and against ex vivo adult worms (Figure B and Table S3 in the Supporting information).

3.

In vitro antischistosomal activity at various drug concentrations (at 1 or 1.25, 10 or 12.5, and 25 or 50 μM) versus DMSO (0.5%) control for the 3-benzylMD sub-series I to III in two parasitic S. mansoni assays, NTS (Figure 3A, and Table S2 in the Supporting information) and against ex vivo adult worms (Figure 3B, and Table S3 in the Supporting information). Adult schistosomes were incubated in culture media for 48 and 72 h at 37 °C and 5% CO₂. The viability of the worms was microscopically evaluated and scored at 48 h and 72 h. Results are from two independent experiments each performed in duplicate. Drug activities were normalized to the DMSO controls. Data represent mean ± standard deviation (sd) in percentage of DMSO control. A compound was classified as active when the drug effect was ≥50%.

Praziquantel displayed IC₅₀ values of 2.2 μM and 0.1 μM against S. mansoni NTS and adult worms, respectively (data from ref ) and a CC₅₀ value of 96 μM, attesting for a high selectivity. Early hits 2a and 3a, described as antimalarial compounds in ref , were discovered by D. L. Williams and collaborators and are now developed in this work. The newly re-synthesized compounds 2a and 3a were tested in parallel with the compounds from sub-series I to III developed in this study. All data from Tables S1 and S2 were measured at TPH, (J. Keiser). Assays were conducted in duplicate and triplicate and repeated once fulfilling all requirements of reproducibility.

The sub-series I includes the early hits, called schistodiones A_2,5 and A_3,5, from the 3-benzylMDs, in which the aromatic ring from the benzylic side chain was disubstituted by methoxy groups, in 2′,5′- or in 3′,5′ (compounds 2a–2c, 3a–3c versus control 4c). Innovative synthetic methodologies heteroaromatic analogues of compounds 2a and 3a hetero-Diels-Alder reactions. However, the final benzylated quinoline-5,8-diones behaved as very weak antischistosomal compounds, suggesting that the presence of a nitrogen atom in the menadione core was not favorable for efficient worm penetration. Furthermore, as none of the potential metabolites of these heteroaromatic analogues synthesized at that timethe 3-benzylMDs are poor oxidant molecules or subversive substrates of disulfide reductases (such as SmTGR)it was not surprising to find them as poor SmTGR inhibitors.

In this study, we introduced a fluorine atom or a trifluoromethyl group at C-6. The sub-series I was extended to derivatives in which both methoxy groups were replaced by oxy-oxetanyl groups (compounds 10a–10e). The influence of fluorinated substituents at C-6 of the menadione core was observed: in the NTS assay, derivatives from sub-series I. bearing a CF₃ substituent at position C-6 (2c, 3c, 4c, 10e) to the menadione core were generally more active in the NTS assay than the corresponding unsubstituted analogues (e.g., 2c vs. 2a, 10e vs 10a), but it was not the case in the adult worm assay (3c vs. 3a, 10e vs 10a), except for 6F-10d (IC₅₀ = 8.8 μM), 10e (IC₅₀ = 11.8 μM), vs 10a (IC₅₀ = 36.6 μM). The most potent compounds of the sub-series I. in the NTS assays were found to be the 3′,5′-disubstituted derivatives: 10e (6-CF₃)> 10d (6-F) > 3a (6-H) with IC₅₀ values below 1 μM for 10e (6-CF₃) and 10d (6-F). Against adult worms, both regioisomers 2a and 3a showed similar IC₅₀ values at 1.88 and 2.96 μM (72h), respectively. The replacement of the methoxy groups by oxy-oxetanyl groups (compounds 10a, 10d, 10e) did not improve the antischistosomal potencies in vitro. Within the p-H sub-series (10a, 10d, 10e), compounds bearing bis­(O-oxetanyl) substituents exhibited greater activity than the corresponding bis­(methoxy) analogues against NTS, whereas the reverse trend was observed for adult worms, where bis­(methoxy) derivatives were more potent.

The sub-series II is characterized by the presence of an electron-withdrawing group (NO₂, CF₃) in para of the benzyl chain of 3-benzylMDs (compounds 5a–5c, 8a–8c, 10b, 10c). Again, we introduced a fluorine atom or a trifluoromethyl group at C-6 (5b–5c), or we substituted the fluorine atom in ortho to the EWG group by an oxy-oxetanyl group (8a–8c), or we introduced two oxy-oxetanyl groups in ortho to the nitro group (10b). Also, because various reported compounds containing a trifluoromethyl group showed better antischistosomal action against adult worms, illustrating better penetration of the lipophilic group across the tegument, we also prepared a bis­(oxy-oxetanyl)-substituted 3-benzylMD (compound 10c) with a trifluoromethyl groupinstead of the nitro group between both oxy-oxetanyl groups to design a hybrid plasmodione-schistodione analogue. BenzylMD 5a (6-H) was the most active for killing NTS with an IC₅₀ of 0.86 μM. Introduction of a fluorine atom or a CF₃ group did not improve the antischistosomal action, but the activity remained high (IC₅₀ < 3 μM). Introduction of one or two oxy-oxetanyl groups in ortho to the nitro group (8a–8c, 10b) or in ortho to the CF₃ group (10c) was not favorable. This sub-series II of benzylMD was not very effective at killing adult worms.

The sub-series III is represented by 3-benzylMD substituted by cyano and carboxylic groups, amides, protected tetrazoles and deprotected tetrazoles (compounds 11–15), and by bis­(benzyl) groups functionalized by the same functions: cyano, carboxylic, amide groups, protected and deprotected tetrazoles (compounds 16–19). In each functionalization, we introduced the substituent in ortho (- o ), meta (- m ) or para (- p ) to compare the activity of regioisomers. In this sub-series III of benzylMD, the substituent in ortho (- o ) revealed to confer the highest potency to the benzylMD core with a cyano group (11c- o ), a carboxylic group (12- o ), or an amide group (13- o ). Replacement of the carboxylic group by bioisoteric tetrazoles did not improve the antischistosomal action (compounds 15). However, the protected tetrazoles functionalized by a cyanoethyl chain (compounds 14) were as active as the amides (compounds 13). Against adult worms, compound 13- o with an amide group in ortho- revealed to be the most potent schistosomicide with an IC₅₀ of 0.59 μM, justifying the in vivo testing in the murine model. In the bis­(benzyl) sub-series functionalized by the same functions: cyano, carboxylic, amide groups, protected and deprotected tetrazoles (compounds 16-19), none of the compounds displayed adequate potency in both NTS and adult worm assays. Again, introduction of a 6-CF₃ to the p-cyano-benzylMD 11a- o improved the potency of the unsubstituted 6-H analogue (11c- o vs. 11a- o ).

Furthermore, we also designed the prodrug of benzylMD 3a, the (±) 4a-ψ-schistodione 3a (24), anticipated to release the benzylMD 3a, in an oxidant environment. The reported strategy had been successfully demonstrated with the prodrug of plasmodione, the (±) 4a-ψ-PD, which exerted potent antischistosomal action in vitro. With the dimethoxy groups in 3′,5′ position of the benzylic chain, the (±) 4a-ψ-schistodione 3a (24) displayed potent antischistosomal activity against NTS with an IC₅₀ value of 0.96 μM, but moderate activity against adult worms.

To summarize, Figure and Table illustrate the structure-activity relationships among benzylMDs from the three sub-series I to III.

1. Selectivity Indexes of the Most Potent BenzylMDs and Reported Derivatives in NTS and Adult Worm Assays In Vitro and Antischistosomal Activities of Selected Compounds in the Patent S. mansoni-Mouse Model.

In Table , we calculatedthe selectivity indexes of compounds displaying IC₅₀ values below 3 μM in the NTS assay, and 4 μM in the adult worm assays. The cytotoxicity has been evaluated for all compounds against the rat L6 cell line. Noteworthy is to mention than the menadione scaffold is the essential pharmacophore responsible for redox-mediated activity. Replacing the aromatic nucleus of menadione by a non-quinonoid analogue with an angular methyl confers a prodrug effect, mainly expressed in adult worms when hemoglobin digestion is an important metabolic pathway in this parasitic stage (Table ). Also, replacing the benzyl linker of benzylMDs with either α- or β-triazole significantly reduced biological activity, indicating that the benzyl moiety is critical for maintaining potency. The benzylic chain confers a more potent antischistosomal action than a triazole function built by a CuAAC reaction (click), as exemplified by the most potent reported triazole, β-36 triazole (Table ). ,

In Vivo Antischistosomal Activities of the Most Potent BenzylMD in S. mansoni-Infected Mice

Table , which presents the selectivity indices derived from tests on NTS and adult worms, enabled us to select the most active compounds against adult worms for in vivo testing in the S. mansoni-infected mouse model (green cells). However, we also selected a few compounds (orange cells) into which oxetanes or tetrazoles groups were introduced as pharmacophore elements known to confer in vivo activity by oral route because of improved bioavailability. Thus, we tested two oxetane derivatives (10a, 10d) and the most potent protected tetrazole (18- o ) in the mouse model, despite their low activity in vitro. We also tested a 6-CF₃ p-cyano-benzylMD to explore more SAR in vivo.

The first hit compounds 3a and 2a of sub-series I, schistodiones A_2,5 and A_3,5, were initially tested in vivo at 200 mg/kg via administration of a single dose per os. Schistodiones A_2,5 2a and A_3,5 3a resulted in low worm burden reduction (WBR) of 18.5% and 11.1%, respectively (Table ). Then, a set of benzylMD and derivatives were selected based on their structural diversity and/or their effective antischistosomal actionderived from either NTS or adult worm assaysfor evaluation of their in vivo potency in the S. mansoni mouse model at 200 mg per os. Results are shown in Table . Because the previously described (±) 4a-ψ-PD showed moderate activity in vivo (Table ), the (±) 4a-ψ-SD-3a (24), which was less active than the reported ψ-PD in vitro, has not been tested in the S. mansoni mouse model.

Noteworthy, the benzylMD (13- o ) combining the highest antischistosomal action both in ex vivo NTS and adult worm assays, and suitable calculated physicochemical properties revealed to be the most efficient to kill adult worms in vivo per os in one dose. This result attests to the reliability of the used in vitro tests against NTS and adult worms before assessing the antischistosomal action in the patent S. mansoni-mouse model. Noteworthy, the substitution pattern of the 3-benzylMD core did not strongly impact the PK properties of final compounds; we know that the 3-benzylMD scaffold is highly lipophilic. The aim of this study was to identify a new sub-series of antischistosomal 3-benzylMDs and start to explore their potential targets, in a living model organism (yeast), and observe the flux of electrons generated in a cascade of redox reactions using the known validated target, SmTGR, and the major nutrient for worms, metHb, in the metHb reduction assay.

2,5- and 3,5-Dimethoxy-benzylmenadiones Induce a “Hairy” Phenotype in S. mansoni

Both 2,5- and 3,5-dimethoxy-benzylMDs 2a and 3a at 50 μM induce discernible alterations, the so-called “hairy phenotype”, 1 h after treatment (Figures and S1 in the supporting information). The hairy phenotype was found only at 50 μM, with an effect more intense in females. In males, a greater quantity of “hair” is found in the anterior sucker. In drug-treated worms versus control parasites the “hair” appeared as a crystalline material which, in some cases, is piercing through the muscle layer and tegument of the worm. However, it is important to note that similar crystalline structures also formed when benzylMD 2a was added to RPMI alone (without worms; Figure S1A) and around dead worms which had been fixed in 4% PFA prior to incubating with the either of the two drugs (Figures S1E,F). On NTS, both drugs were lethal at 50 μM for 2 days and at 20 μM for 3 days, as observed using light microscopy (data not shown), with little effect at 10 μM for 3 days.

4.

Phenotypic effects induced by benzylMD 2a at 50 μM in the presence of adult S. mansoni worms. Crystals protruding through the tegument of a male worm.

Various morphological alterations of the tegument were previously observed, but not clearly understood. One of these compounds was the natural dibenzofuran product rhodomyrtoxin B (Figure ). Phenotypic effects of rhodomyrtoxin B on S. mansoni tegument (Figure S2 in the supporting information) were published; however, they were different from those observed with schistodiones. In order to investigate the essential importance of the redox-active benzylMD core on the antischistosomal action, we introduced the same carboxylic acid/amide/tetrazole functions on the dibenzofuran core (blue colour displayed in Figure ) as found in rhodomyrtoxin B. Three compounds with the same chains introduced in the 3-benzylMD that confer the highest antischistosomal potency were synthesized (Scheme S3 in the supporting information), affording 4,6-dibenzofurandipropanoic acid S6, the amide S7 and the protected tetrazole S8 from the known starting material, S5. We can conclude that the redox-active menadione component was essential for the efficacy of sub-series III. Noteworthy, compounds S5–S9 did not induce a hairy phenotype.

5.

Structures of rhodomyrtoxin B and of related dibenzofuran derivatives.

Redox-Cycling Enzymes and Production of ROS in Yeast

The yeast model was used to show that anti-schisto compounds share (at least in part) the MoA of other studied bMDs: (1) a NADH dependent flavoenzyme for the reduction of bMDO and (2) a target such as Cox15. Enzymes of that kind are of course present in S. mansoni: (1) TGR, other flavoenzymes or even Complex I; (2) Sm Cox15 (but Sm Cox15 is not functional in yeast and its sensitivity to bMDs could not be tested) or other enzymes whose catalytic mechanism involves radical reaction.

We previously showed that the MoA of the antiplasmodial plasmodione (PD) involved NADH/NADPH consuming redox-cycles catalysed by flavoenzymes, which resulted in ROS production. In the yeast model, the mitochondrial NADH-dehydrogenases (together with other flavoenzymes as minor players) had a key role in the redox-cycling. Therefore, we used yeast to test whether NADH-dehydrogenase also played a role in the antischistosomal benzylMD MoAs.

We monitored the effect of 12 antischistosomal benzylMDs on yeast growth (Figure ). For the growth assays, we used the strain AD1-9 that lacks several efflux pumps and shown to be more sensitive to added compounds than standard laboratory strains (for instance, see ref ). From AD1-9, three strains were derived: (i) Cox15-Tc harbors a modified version of the heme a synthase Cox15 that renders the cells more sensitive to the benzylMDs, PD and cruzidione (CZ); , (ii) AD1-9Δnde1 lacks Nde1, the main NADH-dehydrogenase, which results in a significant decrease in NADH-dehydrogenase activity and in sensitivity to PD; (iii) Cox15-TcΔnde1 combines the modified Cox15 with the deletion of NDE1. The function of Cox15 and its possible role in benzylMD MoA will be discussed later. Here the mutant Cox15-Tc was used to increase the cell sensitivity and by consequence to test the effect of Δnde1 on a larger panel of compounds.

6.

Effect of antischistosomal benzylmenadiones on the growth of AD1-9 (Panel A) and three mutants Cox15-Tc (Panel B), AD1-9Δnde1 (Panel C) and Cox15-TcΔnde1 (Panel D), and growth inhibition induced by 13- m : effect of NADH-dehydrogenase level (Panel E) and of the absence of SOD2 (Panel F). The cells were grown in YPEth medium with different concentrations of the antischistosomal benzylMDs. Cell density (OD_600nm) were measured after 3 days culture. The experiments were repeated at least twice and the data averaged. Error bars represent standard deviation. Panels (A to D): The control (ctrl, black) is the OD_600 nm reached in absence of inhibitor. The compounds were added at 1 μM (dark blue), 10 μM (light blue), or 50 μM (grey). Panel (E): WT (AD1-9), line; Δnde1 (lower NADH-dehydrogenase level), dashed line; pNDE2 (higher NADH-dehydrogenase level), dotted line. Panel (F) Δsod2 (absence of the mitochondrial superoxide dismutase).

The effect of antischistosomal benzylMDs on AD1-9 (panel A) and Cox15-Tc (panel B) were tested at 1, 10, and 50 μM. The sensitivity of AD1-9 to the different compounds varied. For instance, no growth inhibition was observed with 3a or 13- o , whereas 13- m at 10 μM severely inhibited the growth. Cox15-Tc showed an increased sensitivity to all the tested compounds, except to 13- o . The insensitivity to 13- o could be caused by the compound not entering the cells and reaching its targets.

Then, we tested the effect of decreasing NADH-dehydrogenase activity (obtained by deleting NDE1) on growth inhibition. We previously found that cells with higher NADH-dehydrogenase content were more sensitive to PD whereas cells with lower NADH-dehydrogenase content were less sensitive. Further analysis confirmed that the NADH-dehydrogenase played a key role in PD mode of action. ,

Here, we compared the sensitivity of AD1-9Δnde1 (panel C) with that of its parental strain AD1-9 (panel A). On AD1-9Δnde1 compounds were tested at 10 and 50 μM. The compounds that did not inhibit AD1-9 growth were excluded. AD1-9Δnde1 showed a decreased sensitivity to all the tested compounds. For instance, 10 μM 13- m had no effect on AD1-9Δnde1 whereas it fully inhibited AD1-9. The same assay was performed with Cox15-TcΔnde1 (panel D) that was tested at 1 and 10 μM. Again, the loss of Nde1 resulted in a decreased sensitivity to all the tested compounds (compare panels D and B). Therefore, it seems that in the yeast model, the NADH-dehydrogenase activity is required for the antischistosomal benzylMD activity, as for PD.

As 13- m appeared to be one of the most efficient inhibitors of yeast growth, it was chosen to further examine growth assays (Figure , Panels E and F). We monitored again the growth inhibition of AD1-9 and its Δnde1 mutant and we tested pNDE2, a strain derived from AD-9 that possesses a higher number of NDE2 genes, which results in increased NADH-dehydrogenase activity and increased PD sensitivity. As previously reported, NADH-dehydrogenase was approximately four-fold decreased in Δnde1 and two-fold increased in pNDE2. As presented in panel E, pNDE2 showed an increased sensitivity to 13- m while Δnde1 was markedly less sensitive to the compound. We also tested the effect of decreasing the antioxidant defence system by deleting SOD2 gene (Δsod2) that encodes the mitochondrial superoxide dismutase. As observed for PD, Δsod2 showed a severely increased sensitivity to 13- m (panel F).

Finally, the role of NADH-dehydrogenase in the antischistosomal benzylMD MoAs was further studied by investigating whether the 3-benzoylmenadione derivatives of these compounds (11 of the 12 compounds) could interact with NADH-dehydrogenase. We previously reported that the PD metabolite, 3-benzoylmenadione PDO, acted as NADH-dehydrogenase substrate, which initiated redox-cycling reactions that produced superoxide anion radicals.

Here, we monitored the rate of cytochrome c reduction initiated by the addition of NADH using isolated mitochondria (Figure A). In the control, cytochrome c was reduced by the combined activity of NADH-dehydrogenase and complex III, the re-oxidation of cytochrome c by cytochrome oxidase being blocked by cyanide.

7.

Effect of 3-benzoylmenadione derivatives as antischistosomal benzylMD metabolites on the NADH-driven cytochrome c reduction independent of complex III. NADH c reductase activities were measured by monitoring the rates of cytochrome c reduction at 550–540 nm. Mitochondria were added at around 25 μg protein mL^–1. The reactions were initiated by the addition of 0.8 mM NADH. The measurements were repeated at least twice and the data averaged. Error bars represent standard deviation. Panel (A) Effect of 3-benzoylmenadiones on NADH-cytochrome c reduction rate. Black, control, complex III dependent rate; Grey, complex III independent rates, observed after addition of 10 μM of the complex III inhibitor atovaquone (ATV) and 0 or 50 μM of the compounds. Panel (B) Effect of 27- m concentration on the rate of NADH-driven cytochrome c reduction. 27- m was added at 0 to 100 μM to the atovaquone inhibited sample.

Addition of the well-known complex III inhibitor atovaquone (ATV) blocked the activity of the complex and the rate of cytochrome c reduction fell to 4.5% of the control rate. Addition of 50 μM of all the 3-benzoylmenadiones to the atovaquone-inhibited sample resulted in a marked increase in complex III-independent cytochrome c reduction rate, from approximately a 5-fold increase for 29- m to approximately 16-fold for 27- p . We also monitored the rate of NADH-driven cytochrome c reduction in the presence of atovaquone at different concentrations of 27- m (Figure B). A clear dose response curve was obtained. This could be explained by cytochrome c being reduced in a complex III-independent manner by the superoxide anions generated by the reaction of the benzoyMDs with NADH-dehydrogenase as shown for PDO It is known that superoxide anions can directly reduce cytochrome c.

Thus, these data indicated that all the tested 3-benzoylMDs could react with the NADH-dehydrogenase, producing superoxide anions, which in this experiment resulted in complex III-independent cytochrome c reduction. This agreed with the data presented in Figure and supported the role of NADH-dehydrogenase in benzylMD activity. As mentioned above, 13- o seemed inactive against Cox15-Tc. Therefore, it was not possible to test the effect of the Δnde1 mutation and obtain an information on the role of NADH-dehydrogenase in that compound’s activity. However, we found that 13- o metabolite, the 3-benzoylMD derivative 27- o , could directly react with the NADH-dehydrogenase similar to the other benzylMDs (Figure A).

In conclusion, from the data obtained with the yeast model, we can propose that all the antischistosomal benzylMDs from the sub-series III share at least part of the PD mode of action and that in yeast NADH-dehydrogenase play a key role as redox-cycling enzyme using the compounds 3-benzoylMD metabolites as substrates.

A Possible Role for SmTGR in ROS Production

Although we previously reported that in yeast NADH-dehydrogenases represent the main enzymatic activity involved in the redox-cycling of PD-generated benzoylMD metabolite, other enzymes may play a role, such as glutathione reductase (GR), albeit to a lesser extent, and also contribute to the process. We also showed that purified human GR (hGR) could catalyze redox-cycling using benzoylMD as substrate. Moreover, in vitro experiments using purified hGR showed that a PD-derived photoaffinity labelling probe could bind to a cysteine, Cys234, of that enzyme. Cys234 is located in proximity of the FAD domain of GR catalytic center in the so-called “doorstop pocket”. , Unlike yeast, S. mansoni does not possess a NADH-dehydrogenase type-2; instead, the parasites possess a mitochondrial complex I. − So, SmTGR, the sole pyridine nucleotide-disulfide oxidoreductase in parasitic platyhelminths and capable of reducing both thioredoxin and glutathione disulfides, , could be a good candidate to play a role in benzoylMD reduction and redox-cycling.

In adult worms, the druggable SmTGR, along with other flavoenzymes may contribute to the general mechanism of action of benzylMD (through benzoylMD metabolites) depicted in Figure . Recent studies have shown that benzoylMD 25 and 26 act as subversive substrates of purified recombinant SmTGR. Redox-cycling by SmTGR in the presence of oxygen produce ROS and may contribute to the death of adult worms. Morerover, docking simulations showed that SmTGR could bind to different schistodione metabolites, the benzoylMD 25 and 26, a benzylMD-derived quinone, and key 25 and 26-derived benzoxanthones, in its so-called “doorstop pocket”, a druggable cavity which has been demonstrated to bind different TGR inhibitors. ,

Furthermore, in this study both benzoylMD 25 and 26 were shown to reduce metHb to oxyHb in a SmTGR-NADPH-metHb coupled assay (Figure A). The reaction of 25 or 26 with metHb in the presence of NADPH and SmTGR caused a red shift in the λ_max of metHb (from 405 to 410 nm) characteristic of oxyHb formation (Figure B). Analysis of k _obs obtained from the spectra at different times and concentrations of 25 and 26 allowed the calculation of the kinetic parameters k _red and K _m characterizing the reduction of metHb: for 25 k _red = 0.0010 s^–1 and K _m = 30.78 μM, for 26 k _red = 0.0007 s^–1 and K _m = 2.25 μM. From k _red we calculated a t _1/2 = 11.5 min for 25 and t _1/2 = 16.5 min for 26. The same red shift was observed with hGR. On the other hand, neither benzoylMD 25 or 26 inhibited the formation of β-hematin in vitro directly under reported conditions (data not shown). As shown above, on the basis of in vitro assays and in silico analysis, SmTGR could be a target of the antischistosomal benzylMDs. The binding of the compounds, more specifically of their benzoylMD metabolites, would initiate redox-cycling, ROS production, and lead to depletion of that key enzyme activity.

8.

3-BenzoylMD reductase activity of SmTGR coupled to methemoglobin reduction. Panel A: The naphthoquinone reductase activity of recombinant SmTGR was tested by monitoring methemoglobin reduction in the coupled assay with NADPH-dependent SmTGR using SDO metabolite as substrate. The methemoglobin reduction assay principle coupled with the human glutathione reductase/NADPH-based system was described in ref but, in this work, SmTGR replaced the human glutathione reductase. Panel B: The reaction of benzoylMD 25 with metHb in the presence of NADPH and SmTGR at 25 °C, and pH 7.4 under steady-state conditions caused a shift from 405 (characteristic of metHb) to 410 nm (oxyHb formation) of the Soret band of hemoglobin. In this assay, final concentrations were 0.27 μM recombinant SmTGR, purified under reported conditions, 2% DMSO and 100 μM NADPH in TGR buffer (0.1 M potassium phosphate, 10 mM EDTA pH 7.4). BenzoylMD 25 became insoluble (turbidity is visible) at concentrations above 80 μM. The curves are the average of two experiments..

Putative Targets of Antischistosomal BenzylMDs in S. mansoni

The data obtained with the yeast model pointed to a second possible target, i.e., Cox15. Two mutants, namely cox15-Tc and cox15-S429F, with modified versions of Cox15 rendering yeast more sensitive to PD and CZ , were tested for their sensitivity to the antischistosomal benzylMDs. In cox15-S429F, a serine residue in Cox15 catalytic domain was replaced by a phenylalanine; in cox15-Tc, the native yeast Cox15 was replaced by Trypanosoma cruzi enzyme.

Cox15-Tc (Figure , panel B) and cox15-S429F (data not shown) also had an increased sensitivity to the antischistosomal benzylMDs, with the possible exception of 13- o . However, as already mentioned, that compound might not be able to enter the yeast cells and reach its targets. Alternatively, 13- o metabolite might bind to the native yeast Cox15, wt and S429F mutant, or with Cox15-Tc. Unfortunately, S. mansoni Cox15 was found not to be functional in yeast. Therefore, we could not test the sensitivity of yeast mutant expressing the worm enzyme.

Cox15 protein is the heme a synthase that converts heme o (produced by Cox10) to heme a by catalysing the oxidation of heme o C8-methyl group. Heme a is an essential co-factor of cytochrome c oxidase, the terminal enzyme of the respiratory chain. Cox15 catalytic mechanism, although not fully understood, is likely to involve radical reactions.

In our recent report on anti-T. cruzi benzylMDs, we hypothesized that reduced benzoylMDs produced by NADH-dehydrogenase could react with Cox15, more specifically with heme radicals generated during Cox15 catalytic activity. We propose here that the antischistosomal benzylMDs would act in a similar manner and interfere with Cox15 activity in the worms resulting in depletion of heme a and, by consequence, of cytochrome c oxidase and respiratory chain activity. We also speculate that other parasitic enzymes whose catalytic reaction involves the generation of radicals could be other targets of benzylMDs and/or its metabolites.

Conclusions

In conclusion, we designed and synthesized three sub-series of 3-benzylmenadiones displaying moderate to potent antischistosomal activity in vitro in NTS (13- o : IC₅₀ = 1.04 M) and adult worm (13- o : IC₅₀ = 0.59 μM) assays but low activity (13- o : WBR, 29.5%) after one single dose at 200 mg/kg per os in the S. mansoni-infected mouse model. Further, our investigations into the MoAs of antischistosomal benzylMDs using the yeast model indicate that these compounds share, at least in part, the MoA of the antiplasmodial agent PD and the anti-T. cruzi compound CZ, and that yeast NADH-dehydrogenases play a key role in the bioactivation of benz­(o)­ylMDs through bioreduction of the naphthoquinone moiety. In S. mansoni, SmTGR (among other parasitic flavoenzymes) may be a key enzyme involved in the continuous NADPH-dependent redox-cycling of benzoylMDs acting as subversive substrates. This process contributes to futile consumption of NADPH and the elevation of the steady flux of ROS in the parasite.

Another potential candidate target would be Cox15. Inhibition of the heme a synthase by the benzylMDs would result in respiratory defect, as a consequence of the depletion of the cytochrome oxidase essential co-factor heme a. It is well established that schistosomes consume large amounts of glucose, approximately 20% of their weight per hour, with the majority being converted to lactate and excreted. However, mitochondrial respiration is also important, and it is estimated that under aerobic conditions about one third of the energy production of adult worms results from mitochondrial electron transport. Therefore, Cox15 inhibition with consequent defect in respiratory function could have a severe impact on the worms.

One could suggest that the antischistosomal activity of the benzylMDs results from the combined oxidative stress and respiratory function impairment. Their impact might be broader as the compounds could target other enzymes and functions that remain to be identified.

Experimental Section

Protocols for in vitro drug assay against S. mansoni NTS and adult worms, phenotypic effects of benzyIMD 2a and 3a in adult S. mansoni worms, cytotoxicity assays with the rat L6 cell line, are described in the supporting information.

In Vivo Drug Assay in S. mansoni-Infected Mice

Animals were purchased, maintained and infected as described above. Approximately 49 days post-infection, mice were treated according to their individual body weights measured immediately prior to dosing 200 mg/kg. The test item was administered orally by gavage, with the administered volume adjusted based on the recorded individual body weight. The selected vehicle consisted of 7% Tween 80 (Fluka), 3% absolute ethanol (≥99.9%; Supelco), and tap water. Control mice were infected but not treated. Formulations were kept under continuous magnetic stirring for a maximum of 30 min and were administered immediately after preparation. Twenty-one to twenty-3 days after treatment, mice were euthanized by CO₂ asphyxiation and dissected. The mesenteric tissue and liver were collected, and adult worms were recovered from the hepatic portal system and mesenteric veins. The liver was flattened and examined for the presence of worms, and the distribution of worms in the liver and mesenteric veins was recorded. Worms were washed, sexed, and counted using a binocular microscope.

The worm burden (WB) of treated animals was calculated and compared with that of simultaneously infected but untreated control mice (n = 4–8). Worm burden reduction (WBR) was calculated as follows

$$\mathrm{W}\mathrm{B}\mathrm{R}(\%)=100\times [1-({\mathrm{W}\mathrm{B}}_{\mathrm{t}\mathrm{r}\mathrm{e}\mathrm{a}\mathrm{t}\mathrm{m}\mathrm{e}\mathrm{n}\mathrm{t}}/{\mathrm{W}\mathrm{B}}_{\mathrm{c}\mathrm{o}\mathrm{n}\mathrm{t}\mathrm{r}\mathrm{o}\mathrm{l}})]$$

Growth Media and Yeast Strains

The following growth media were used: YPD (1% yeast extract, 2% peptone, 3% glucose), YPEth (1% yeast extract, 2% peptone, 2% ethanol) and CSM-uracil medium (0.7% yeast nitrogen base, 2% glucose, 2% agar and 0.8 g/L of a complete supplement mixture minus uracil supplied by Bio 101 (San Diego, CA, USA)). The yeast strains derived from AD1-9 (α, ura3, his1, Δyor1, Δsnq2, Δpdr5, Δpdr10, Δpdr11, Δycf1, Δpdr3, Δpdr15, Δpdr1), which has been kindly provided by M. Ghislain, UCL, Belgium. AD1-9 lacks several membrane transporters rendering the cells more sensitive to compounds added in the growth medium. The deletion of NDE1 (nde1: kanMX4) was generated by PCR-based deletion. The overexpression of NDE2 was performed by cloning the gene with its own promoter and terminator on the multi-copy plasmid yEP352. Cox15-tc mutant was generated by replacing the yeast COX15 gene by T. cruzi cox15 gene.

Growth Assays and Drug Toxicity Tests

The drug sensitivity assays were performed by monitoring growth inhibition in the presence of the compounds. Yeast were grown in 1 mL culture medium YPEth with different concentrations of the compounds to be tested. Cultures were inoculated at an OD_600nm of 0.2 and incubated at 28°C with vigorous shaking for 3 days. OD_600nm were then measured. Each measurement was repeated at least twice, the data were averaged and plotted with error bars representing standard deviation.

NADH-Cytochrome c Reduction Measurements

Mitochondria were prepared as described. NADH-cytochrome c reductase activities were measured by monitoring the rate of reduction of cytochrome c in a spectrophotometer at 550–540 nm over 3 min time-course. Measurements were performed at room temperature in 1 mL of 10 mM potassium phosphate pH 7, 2 mM KCN and 20 μM cytochrome c. Mitochondria were added at 25 μg protein mL^–1. The reaction was started by the addition of 0.8 mM NADH.

Glossary

Abbreviations

ATV

atovaquone

benzoylMD

3-benzoylmenadione

benzylMD

3-benzylmenadione

CZ

cruzidione

DCM

dichloromethane

hGR

human glutathione reductase

mp

melting point

MeOH

methanol

MoA(s)

mode(s) of action

NEt₃

triethylamine

NTS

newly transformed schistosomula

n.d.

not detected

PZQ

praziquantel

ROS

reactive oxygen species

rt

room temperature

sd

standard deviation

SD

schistodione

SDO

schistodione oxide

SmTGR

Schistosoma mansoni thioredoxin–glutathione reductase

WBR

worm burden reduction.

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsomega.6c06911.

• Chemistry (pages S2–S31): general information, general procedures A-H, synthetic procedures (Schemes S1–S3) and analytical data of all compounds 2-32 and control compounds S6–S8; parasitology (pages S32–S39): antischistosomal activities (NTS, adult worms), calculated physicochemical parameters, and cytotoxicity of 3-benzyMDs and related derivatives in vitro: protocols and supplementary data in Tables S1 and S2; and phenotypic effects induced by antischistosomal compounds on S. mansoni adult worms: supplementary Figures S1 and S2; ¹H, ¹⁹F and ¹³C {¹H} NMR spectra (pages S42–S134) of new compounds 2–32 and intermediates (PDF)

∇.

Department Chemie, Johannes Gutenberg-Universität, Duesbergweg 10-14, 55128 Mainz, Germany

#.

J.P. and J.R. contributed equally as first co-authors. J.P., J.R., B.M., M.D., and C.H. generated and analyzed the experimental data. J.P., J.R., B.M., D.L.W., J.K., and E.D.-C. analyzed all data and wrote the paper.

The authors declare no competing financial interest.