In Vitro, In Vivo, and Absorption, Distribution, Metabolism, and Excretion Evaluation of SF₅-Containing N,N’-Diarylureas as Antischistosomal Agents

ABSTRACT

In recent years, N,N’-diarylureas have emerged as a promising chemotype for the treatment of schistosomiasis, a parasite-caused disease that poses a considerable health burden to millions of people worldwide. Here, we report a novel series of N,N’-diarylureas featuring the scarcely explored pentafluorosulfanyl group (SF₅). Low 50% inhibitory concentration (IC₅₀) values for Schistosoma mansoni newly transformed schistosomula (0.6 to 7.7 μM) and adult worms (0.1 to 1.6 μM) were observed. Four selected compounds that were highly active in the presence of albumin (>70% at 10 μM), endowed with decent cytotoxicity profiles (selectivity index [SI] against L6 cells >8.5), and good microsomal hepatic stability (>62.5% of drug remaining after 60 min) were tested in S. mansoni-infected mice. Despite the promising in vitro worm-killing potency, none of them showed significant activity in vivo. Pharmacokinetic data showed a slow absorption, with maximal drug concentrations reached after 24 h of exposure. Finally, no direct correlation between drug exposure and in vivo activity was found. Thus, further investigations are needed to better understand the underlying mechanisms of SF₅-containing N,N’-diarylureas.

INTRODUCTION

There are more than 200 million people enduring schistosomiasis and several million people currently living at risk for an infection. This neglected tropical disease is caused by the blood dwelling flatworm of the genus Schistosoma, with Schistosoma haematobium, Schistosoma japonicum, and Schistosoma mansoni being the main responsible species (1–2). Schistosomiasis causes a tremendous health burden, including malnutrition, anemia, learning deficiencies, and, in rare cases, death due to chronicity of the disease (2–3).

For more than four decades, treatment of schistosomiasis has relied on one drug only, praziquantel, patented in 1973 and widely used in preventive chemotherapy programs (4). It is inexpensive and safe and can be used in the treatment of all six Schistosoma species infecting humans. Major drawbacks include the inactivity of praziquantel against juvenile parasites and the rising concern of drug resistance (4–6). Therefore, there is an urgent need to develop new antischistosomal drugs able to effectively control this disease (7–8).

Over the past years, drug discovery efforts have focused on identifying new scaffolds through phenotypic screenings. Specifically, N,N’-diarylureas emerged as a new, promising chemotype, with the chlorinated, symmetrical urea MMV665852 (Fig. 1) standing out as an early lead for the discovery of potential antischistosomal agents (9). Of note, this compound was able to decrease worm viability in vitro with a 50% inhibitory concentration (IC₅₀) of 0.8 μM (only 4-fold less potent than praziquantel) and it reduced the worm burden of S. mansoni-infected mice by 53% after a single oral dose of 400 mg/kg. Furthermore, it presents an acceptable oral pharmacokinetic (PK) profile, characterized by a half-life (t_1/2) of 4.7 h and a maximum concentration in serum (C_max) of 4.4 μM at 46.4 mg/kg (9).

FIG 1.

Chemical structures of the symmetrical diarylurea MMV665852 and its CF₃-analog 1.

Worth mentioning is that the presence of N,N’-diarylurea compounds in medicinal chemistry is quite remarkable, owing to the versatility of this scaffold, and thus their broad spectrum of biological activities. Particularly, they have been widely studied as anticancer (10–12), insecticidal (13), and antimicrobial agents (14–15), as well as in immunology (16–17) and in other infectious diseases (18–19).

In view of these promising results, the past few years have witnessed an increase in structure-activity relationship (SAR) studies aimed at obtaining new antischistosomal analogs featuring the N,N’-diarylurea core (20–22). Interestingly, from the SAR perspective, the N,N’-diarylurea series seems to require electron-withdrawing groups on both aryl units for displaying schistosomicidal activity (20, 22).

Notwithstanding these results, the promising in vitro antischistosomal potency of these reported compounds has not to date been translated into significant worm-killing activity in vivo. In some cases, no direct correlation has even been found between compound exposure and in vivo activity (21). In this sense, future medicinal chemistry and pharmacology work is needed to understand and overcome these issues (23).

Among the different electron-withdrawing substituents that have been explored so far in this type of compounds, it is worth mentioning the trifluoromethyl group (CF₃), commonly used in medicinal chemistry as a bioisosteric replacement of chlorine atoms. Indeed, a representative example can be found in compound 1, rationally designed by replacing the meta chlorine atoms of MMV665852 by CF₃, which displays similar activity against S. japonicum as does the parent compound (Fig. 1) (22).

In recent years, a new bioisoster of the trifluoromethyl unit has been gaining attention in medicinal chemistry: the pentafluorosulfanyl group (SF₅) (24). The commercial unavailability of building blocks functionalized with this polyfluorinated substituent has often precluded its use in medicinal chemistry (24–26). Nevertheless, significant improvements in the synthetic accessibility of SF₅ in the past decades have boosted the use of this functional group in agrochemicals and chemistry of materials, along with intensive study of this group as a bioisoster of CF₃ in many drug candidates (27–29). Noteworthy, in 2017, the first clinical candidate containing an SF₅ unit (DSM265) reached Phase IIa clinical trials for the treatment of malaria (30).

The pentafluorosulfanyl moiety has often been referred as a “super-trifluoromethyl” group; however, these two substituents have remarkable differences, including distinct electron density profiles (a tetragonal bipyramidal shape for SF₅ as opposed to a tetrahedral for CF₃) and some advantageous properties for the SF₅. These advantages include high electronegativity (3.65 versus 3.36 for the trifluoromethyl group), high lipophilicity, and large steric volume (only slightly lower than that of tert-butyl). It also shows high chemical and hydrolytic stability (24, 27, 30). Indeed, it has been demonstrated that the exchange of CF₃ for SF₅ leads to higher lipophilicity, but it can also improve other parameters that have an important influence on in vivo activity. Designing bioactive compounds is a multifactorial process in which many parameters (e.g., solubility, permeability, and stability) need to be considered. For instance, the SF₅ introduction confers good membrane permeability while also being metabolically stable and chemically robust (31–32).

Taking into account the unique properties of this polyfluorinated unit and based on the positive outcomes obtained with the replacement of CF₃ groups in several well-known drugs (for instance fenfluramine and mefloquine) with SF₅ (33–34), the aim of the present work was to explore the introduction of this moiety in the diarylurea scaffold as both a Cl and a CF₃ replacement and investigate its impact on antischistosomal activity. Herein, we report the design, synthesis, and in vitro and in vivo evaluation of new diarylureas containing SF₅, CF₃, and Cl groups against S. mansoni in an attempt to expand SAR studies around this chemotype and provide useful insights for a next generation of optimized antischistosomal drugs (Fig. 2).

FIG 2.

General structure of the new N,N’-diarylureas presented in this work, designed as derivatives of MMV665852 and compound 1. SAR, structure-activity relationship.

RESULTS AND DISCUSSION

Chemistry.

The N,N’-diarylureas presented in this work were synthesized following a simple and straightforward procedure consisting of the coupling of phenyl isocyanates with the corresponding anilines under four slightly different reaction conditions (Fig. 3). In turn, the intermediate phenyl isocyanates were either commercially available or synthesized in situ from the reaction of the precursor anilines with triphosgene, in the presence of a base (triethylamine or saturated aqueous NaHCO₃ solution) in hot toluene or in dichloromethane (Fig. 4). Compounds were synthesized in low to moderate yields, since it was observed that dimerization products from the starting aromatic amines were often predominant. The structure of the target compounds was confirmed by infrared spectroscopy (IR), ¹H, ¹³C and ¹⁹F nuclear magnetic resonance (NMR), elemental analysis, and/or high-resolution mass spectrometry (HRMS) (please refer to the supplemental material for the synthesis pathways and further experimental details).

FIG 3.

Synthetic procedures for the preparation of novel diarylureas. Reagents and conditions: (a) pyridine, room temperature (rt), 1 h; (b) n-BuLi, anh. THF, –78°C to rt, overnight; (c) CH₂Cl₂, rt, overnight; (d) THF, rt, overnight. See Table 1 and supplemental material for details.

FIG 4.

Synthetic procedure for the preparation of the intermediate arylisocyanates. Reagents and conditions: (a) Triphosgene, Et₃N, toluene, 70°C, 2 h; (b) Triphosgene, sat. NaHCO₃, DCM, 30 min. See supplemental material for details.

In vitro antischistosomal and cytotoxicity evaluation.

Drugs were first tested for their ability to reduce the viability of newly transformed schistosomula (NTS) and subsequently against adult S. mansoni worms (Table 1 and Table S1 in the supplemental material). Of the 31 compounds tested at 10 μM against NTS, 26 compounds exhibited more than 75% activity and 22 compounds killed NTS within 72 h. The calculated IC₅₀ ranged from 0.6 to 7.7 μM. Of 31 compounds, 14 killed adult S. mansoni worms after 72 h of incubation at 10 μM. Of these, four compounds killed adult S. mansoni worms at 1 μM. IC₅₀ values ranging from 0.1 to >10 μM were calculated.

TABLE 1.

In vitro antischistosomal activities and cytotoxicity evaluation of the new N,N’-diarylureas

Compound	Structure	ClogPa	NTS			S. mansoni adult worms			L6a cells	SIa
			% Effect at 10 μM (SD)	% Effect at 1 μM (SD)	IC50, μMb	% Effect at 10 μM (SD)	% Effect at 1 μM (SD)	IC50, μMb	IC50, μM (SD)
1		4.8	100.0 (0.0)	31.9 (13.3)	NA	82.5 (3.3)	88.4 (5.5)	0.3	6.0 (2.2)	18.4
2		6.2	100.0 (0.0)	26.7 (12.8)	1.5	100.0 (0.0)	95.9 (0.0)	0.3	4.0 (2.3)	12.4
3		6.5	100.0 (0.0)	43.3 (2.1)	NA	91.8 (5.8)	88.4 (5.5)	0.3	3.8 (1.9)	12.0
4		7.3	100.0 (0.0)	67.3 (2.4)	0.6	100.0 (0.0)	86.5 (2.7)	0.2	3.2 (0.4)	15.4
5		6.2	100.0 (0.0)	24.1 (9.8)	NA	100.0 (0.0)	36.3 (8.2)	1.4	6.8 (1.2)	5.1
6		6.2	68.5 (7.9)	17.2 (24.4)	NA	61.2 (8.7)	59.5 (8.2)	1.6	24.0 (10.4)	15.2
7		5.5	100.0 (0.0)	25.9 (2.4)	NA	81.6 (2.9)	80.7 (0.0)	0.5	8.2 (1.0)	15.1
8		6.1	100.0 (0.0)	34.5 (0.0)	NA	91.8 (5.8)	86.5 (13.6)	0.4	5.5 (1.9)	12.7
9		5.5	95.9 (5.8)	19.7 (2.6)	NA	28.8 (0.0)		>10	17.6 (2.4)	NA
10		6.0	100.0 (0.0)	62.1 (9.8)	0.8	100.0 (0.0)	50.2 (2.6)	0.4	5.0 (2.9)	13.5
11		5.3	100.0 (0.0)	17.1 (21.3)	1.6	79.6 (0.0)	71.4 (5.8)	0.3	5.8 (2.0)	20.8
12		5.3	12.2 (14.5)	8.3 (3.4)	>10	NA	79.6 (5.8)	NA	73.6 (32.6)	NA
13		5.3	90.7 (2.6)	53.5 (2.4)	0.9	89.8 (2.9)	67.2 (8.2)	0.6	4.7 (1.1)	7.5
14		5.7	100.0 (0.0)	48.4 (23.0)	1.0	100.0 (0.0)	80.0 (5.7)	0.2	2.8 (1.3)	14.8
15		5.5	100.0 (0.0)	50.0 (7.3)	1.0	100.0 (0.0)	100.0 (0.0)	0.3	3.3 (0.2)	10.4
16		5.5	100.0 (0.0)	44.8 (0.0)	1.1	100.0 (0.0)	100.0 (0.0)	0.3	2.9 (0.7)	8.5
17		5.3	100.0 (0.0)	13.7 (11.0)	1.5	88.8 (0.0)	83.7 (5.8)	0.4	4.0 (2.3)	10.8
18		5.7	100.0 (0.0)	37.9 (12.0)	1.2	100.0 (0.0)	22.4 (5.8)	NA	2.0 (0.6)	NA
19		4.8	100.0 (0.0)	19.9 (7.9)	1.5	82.5 (3.3)	70.0 (8.5)	0.3	6.0 (2.2)	18.4
20		4.7	98.0 (2.9)	21.5 (0.1)	1.9	100.0 (0.0)	100.0 (0.0)	NA	6.8 (5.6)	NA
21		5.5	61.2 (54.8)	2.8 (1.1)	7.7	100.0 (0.0)	71.4 (0.0)	0.6	4.6 (1.4)	7.9
22		4.6	76.0 (0.7)	29.0 (0.6)	2.8	94.4 (7.9)	18.3 (5.8)	NA	50.7 (22.4)	NA
23		5.7	58.0 (1.6)	15.4 (13.5)	6.9	96.3 (5.3)	30.6 (5.8)	NA	50.3 (44.4)	NA
24		4.4	22.9 (2.9)		>10	52.1 (14.7)		>10	51.1 (16.8)	NA
25		4.5	100.0 (0.0)	27.9 (24.3)	1.5	100.0 (0.0)	74.0 (2.8)	0.5	8.6 (8.9)	16.7
26		4.7	100.0 (0.0)	59.7 (23.8)	0.7	94.6 (7.7)	67.9 (11.3)	0.2	4.0 (2.0)	24.3
27		5.1	100.0 (0.0)	43.1 (2.4)	NA	100.0 (0.0)	100.0 (0.0)	0.2	3.3 (0.7)	13.8
28		4.6	100.0 (0.0)	15.6 (8.2)	1.8	98.2 (2.6)	84.0 (5.7)	0.1	3.3 (1.2)	26.4
29		4.6	100.0 (0.0)	41.4 (4.9)	NA	100.0 (0.0)	98.1 (2.7)	0.4	8.5 (1.9)	20.6
30		4.3	100.0 (0.0)	58.6 (14.6)	0.8	91.8 (2.6)	80.7 (0.0)	0.4	2.9 (0.8)	8.0
31		4.3	100.0 (0.0)	20.7 (0.0)	NA	100.0 (0.0)	73.0 (0.0)	0.4	2.8 (0.7)	7.5
Podophyllotoxin		NA	NA		NA	NA		NA	0.005 (0.004)	NA

^aClogP values were calculated using ChemDraw 19.1; L6, rat myoblast cell line; SI, selectivity index (SI = IC₅₀ in L6 divided by IC₅₀ in adult S. mansoni).

^bNA, not applicable, no robust IC₅₀ values could be calculated. Please refer to Table S1 from the supplemental material for further details.

To evaluate the in vitro cytotoxicity, all compounds were tested on rat myoblast L6 cells. The antischistosomal selectivity of the compounds appeared to be moderate to high, with selectivity indexes (SI) ranging from 5.1 to 26.4 (Table 1).

From an SAR perspective, the studied diarylureas displayed variable activity against adult worms and cytotoxicity that depended on the lipophilicity, pattern of substitution, and halogen content (tetrasubstituted, trisubstituted, and disubstituted) of the phenyl rings. Starting from resynthesized compound 1 and considering that high lipophilicity seems to be an important feature for good in vitro activity, we decided to further explore the tetrasubstituted series. Ureas 2 and 3 were thus generated by changing the position of the chlorine on the left-hand side ring and, in 3, replacing the CF₃ by the SF₅ group, resulting in good potency and cytotoxicity similar to that of the former compound. This encouraged us to prepare compound 4, the SF₅ analog of compound 1, which displayed a similar profile to its counterpart but with a much higher ClogP. Moreover, isomers 5 and 6 did not improve the overall results of the CF₃-containing compounds 1 and 2 but were slightly less active, which reveals that the replacement of CF₃ by SF₅ is not always advantageous. Provided that the aforementioned ureas are highly lipophilic and that the potency of 1 could not be significantly improved, another strategy was followed, based on reducing the overall chlorine content. Thus, removal of a first chlorine atom gave rise to the trisubstituted series, with compounds 7–10 displaying likewise high ClogP, similar potency, and similar selectivity index. Next, from urea 8, a second Cl removal was explored, affording the disubstituted series of compounds. Of note, it was decided to remove the chlorine atom from the right-hand side ring, provided that nonsubstituted phenyl rings had previously shown to be poorly active in vitro (20–22). Satisfactorily, diarylurea 11 showed good potency and selectivity index. However, its isomers 12 and 13, featuring ortho and para SF₅ substitutions, respectively, were less potent, which suggested that the meta-SF₅ position was preferred when combined with a chlorine. Furthermore, from 11 and 13 we explored the replacement of the last Cl on the left-hand side ring for either CF₃ or SF₅, obtaining ureas 14, 15, and 16 that presented good potency and selectivity index, especially 14. These last results are remarkable, since it was demonstrated that the SF₅ group as the sole substitution on both phenyl rings is a promising approach. Additionally, based on the positive outcomes of the combination of SF₅ with Cl and CF₃ in 11 and 14, respectively, their meta,meta-derivatives were investigated, but both the SF₅,Cl-compound (compound 17) and the SF₅,CF₃-compound (compound 18) presented a slightly less adequate profile. Finally, the para,ortho disubstitution was explored, with compounds 19–24 featuring different combinations of Cl, CF₃, and SF₅ substituents. Overall, 19-24 presented significantly lower ClogP values coupled with slightly lower antischistosomal activities.

As previously mentioned, the N,N’-diarylurea chemotype can be found in many published works, particularly regarding its anticancer activity. Worth highlighting are compounds 25 and 26, reported as potent anticancer agents both in vitro and in vivo (12). In our work, both ureas, which present a benzothiadiazole ring and a differently halogen-substituted phenyl ring, were found to act as antischistosomal agents. Indeed, 26, with meta-CF₃ and para-Cl substitution on the right-hand side ring, displayed a very potent worm-killing activity in vitro. We therefore envisaged the synthesis of its SF₅ analog (compound 27) and its subsequent isomers (compounds 28 and 29), as well as the chlorine-free derivatives 30 and 31, which were endowed with very good activity. Outstanding results were obtained with urea 28, which was the most potent SF₅ compound of all the series, with the highest SI.

According to their in vitro activity against NTS and adult S. mansoni worms, as well as their selectivity toward S. mansoni rather than mammalian cells, seven SF₅ compounds (10, 11, 14, 15, 16, 19, and 28) were selected for further in vitro profiling, with the final aim of progressing the best ones to in vivo efficacy studies in mice with an established chronic S. mansoni infection.

In vitro activity of selected compounds in the presence of albumin and hepatic metabolic stability assay.

We wondered whether the antischistosomal activity of the selected compounds was affected by the presence of albumin, the main protein of human plasma. To estimate this impact, adult S. mansoni were incubated in albumin-enriched medium (45 g/liter) for 72 h in the presence of 10 μM of the respective diarylurea. Satisfactorily, four out of seven tested compounds remained highly active against adult S. mansoni worms with more than 70% activity. Hence, 11, 15, 16, and 28 were evaluated for their in vitro metabolic stability using mouse liver microsomes. The percentage of compound remaining after 60 min of coincubation with the microsomes was calculated. Urea 28 was moderately stable, with 62.5% of the compound remaining after the incubation period, while the other three diarylureas showed high stability with values over 70% (Table 2).

TABLE 2.

In vitro antischistosomal activities in the presence of albumin and metabolic stability of selected compounds

Compound	S. mansoni adult worms: % effect (SD) at 10 μM in albumin-suppl. medium (45 g/liter)	Metabolic stability: % compound remaining (SD) after 60 min of incubationa
10	49.0 (2.7)
11	71.7 (2.7)	80.1 (4.5)
14	58.4 (10.7)
15	77.3 (0.0)	70.3 (6.5)
16	71.7 (2.7)	104.8 (32.2)
19	52.8 (13.4)
28	86.8 (2.7)	62.5 (7.8)

^aEmpty cells indicate where compounds were not further analyzed for their metabolic stability in the presence of mouse liver microsomes.

In vivo oral efficacy and pharmacokinetic evaluation.

After the in vitro evaluation of synthesized ureas and the further profiling of the four promising compounds, they were progressed to in vivo efficacy studies in mice harboring chronic S. mansoni infections (Table 3). All four compounds were tested at an initial dose of 400 mg/kg, corresponding to the effective dose of the gold standard, praziquantel (35). Unfortunately, while this dose did not reduce the worm burden of mice treated with 11 and 28, it resulted in toxicity for 15 and 16. Both groups of mice (n = 2 × 4) died one or 2 days after receiving treatment. For 15 and, to a lesser extent, for 16, toxicity was observed as a starting necrosis in the spleen and parts of the liver. Lowering the dose from 400 to 100 mg/kg resulted in low to moderate worm burden reductions when mice were treated with 15 and 16. Huge differences in worm burden reductions of single mice treated with 100 mg/kg of 15, as well as the still-observed toxicity (two mice died shortly after treatment), suggest that this compound has a narrow therapeutic window. Although little is known in the literature of the toxicology of SF₅ compounds, nor their biological fate, we did not foresee such toxicity, especially considering the positive outcomes of our previous studies with SF₅ diarylureas both in vitro and in vivo, in which we also used high doses in mice with a compromised hepatic function (a diet-induced obesity model) (15, 36). Besides, during the clinical development of the SF₅-containing candidate DSM265, which required metabolite identification and profiling in animals, metabolism of the SF₅ group was not observed (37). Importantly, both diarylureas 15 and 16 feature the SF₅ group in the para position, in comparison to 11 and 28, where this polyfluorinated group is in meta. Therefore, a possible explanation for their toxicity may rely on this para-substitution and potential metabolites with this pattern that may be generated in animals with a compromised hepatic and metabolic function due to schistosomiasis. Nevertheless, further studies should be performed to fully confirm this observation.

TABLE 3.

In vivo oral efficacy of selected compounds

Compound	Dose mg/kg	No. of mice cured/no. investigated	No. of mice that died	Mean no. of worms (SD)	% Worm burden reduction (SD)	P valued
11	400	0 / 8b	0	16.0 (10.5)	15.1 (62.8)	0.2
15	100	0 / 4	0	18.3 (8.7)	24.0 (36.0)	0.2
16	100	1 / 8a	2	7.3 (10.6)	36.2 (70.0)	0.05
28	400	0 / 4	0	30.5 (19.6)	0.0 (0.0)	0.7
Control (C1)		0 / 8c	0	24.0 (6.8)	NAe	NA
Control (C2)		0 / 4	0	16.8 (3.6)	NA	NA

^aResults were averaged from two experiments in which C1 and C2 served as controls, and one mouse was not infected and was excluded from the analysis.

^bResults were averaged from two experiments, in which C1 and C2 served as controls.

^cC1 served as control for compounds 28 and 11.

^dP value calculated by Kruskal-Wallis test.

^eNA, not applicable.

Comparing the overall low and moderate worm burden reductions from this study with previously evaluated diarylureas, such as MMV665582 (9) and MMV665582 analogs described by Cowan et al. (20) and Yao et al. (22) (although this study was conducted in S. japonicum-infected mice), as well as Wu et al. (21), it could be concluded that despite the herein-reported finding that N,N’-diarylureas presented decreased lipophilicity and relatively low IC₅₀ and cyctotoxicity values against adult S. mansoni in vitro, this was not enough for getting acceptable in vivo activity.

Next, we analyzed the PK data generated from the in vivo studies, since it could help us better understanding the overall outcomes. Blood samples that were collected from infected mice alongside the in vivo efficacy study at time points 1, 3, 5, 7, and 24 h after oral treatment (100 or 400 mg/kg) with the four selected compounds, 11, 15, 16 and 28, were analyzed by liquid chromatography-tandem mass spectrometry (LC−MS/MS). Plasma concentrations of all four compounds increased until 24 h post drug administration, independent of the dosage (as can be seen in Fig. S1). We wondered whether the infection status could affect the PK behavior of the tested compounds. For that, a single 20 mg/kg oral dose of the most active compound, 16, was administered to noninfected mice to assess exposure over longer evaluation time points (Table 4). Plasma concentrations of compound 16 increased until 24 h post-dosing (Fig. 5), and a half-life of 112.9 h and an area under the concentration curve from zero to 72 h (AUC_0-72h) of 2.326 × 10⁵ ng/ml · h were calculated. Overall, this data indicates that compound 16 following oral administration was slowly absorbed and exhibited a relatively flat distribution and elimination profile. The profile likely reflects a poor aqueous solubility leading to a prolonged absorption. Moreover, the plasma concentration remained above 2,000 ng/ml over the 72 h sampling period, and the apparent total clearance of the drug from plasma after oral administration was 1.4 ml/min/h. Most intriguing was the fact that the maximal concentration of the administered compound was only reached at 24 h post-dosing. This, together with the very long half-life of the molecule, stands in contrast to previously analyzed antischistosomal diarylureas (9, 21). Finally, these findings clearly demonstrated that the infection status of the mice did not influence the plasma-concentration curve of the compounds in the in vivo study.

TABLE 4.

Pharmacokinetic parameters of compound 16 in noninfected mice^a

Oral dose	Cmax (ng/ml)	Tmax (h)	AUC0-72h (ng/ml·h)	t1/2 (h)	CL/F (ml/min/kg)
20 mg/kg (n = 4)	3685	24	2.326 × 105	112.9	1.4

^aData presented are based on the mean concentration versus time data. Mice were not fasted prior to drug administration. C_max, maximum concentration of drug in serum; T_max, time to maximum concentration of drug in serum; AUC, area under the concentration curve; t_1/2, half-life; CL/F, oral clearance.

FIG 5.

Plasma concentration-time curve of compound 16 in female NMRI mice following oral administration of 20 mg/kg.

In summary, 31 N,N’-diarylurea analogs featuring the scarcely explored pentafluorosulfanyl group were prepared and evaluated as antischistosomal agents. High in vitro activity against NTS (IC₅₀ = 0.6 to 7.7 μM) and adult S. mansoni (IC₅₀ = 0.1 to >10 μM), was observed. Seven target ureas were further characterized in vitro, with four of them showing no significant impact on binding to albumin and moderate to high hepatic metabolic stability. However, despite the promising in vitro worm-killing potency, none of them showed significant activity in vivo. Moreover, we observed in vivo toxicity for ureas 15 and 16, although the in vitro selectivity toward S. mansoni was more pronounced than toward mammalian L6 cells. Pharmacokinetic data showed that these compounds have a very long half-life and, strikingly, they reach the maximal concentration at 24 h post-dosing, which differs from previously reported studies with this type of compounds. Future studies should be focused on understanding both the lack of activity in vivo and the PK behavior of this family of N,N’-diarylureas, as well as the origin of the toxicity of our compounds. We believe that the continuous improvement in synthetic methodologies will likely result in increasing numbers of biologically active SF₅ derivatives, and, at the same time, contributions like ours will certainly reinforce the interest of studying the metabolism and toxicity of this type of compounds.

MATERIALS AND METHODS

Chemical synthesis. General methods.

Commercially available reagents and solvents were used without further purification unless stated otherwise; 2-chloro-3-(pentafluoro-λ⁶-sulfanyl)aniline, 2-chloro-5-(pentafluoro-λ⁶-sulfanyl)aniline and 4-chloro-3-(pentafluoro-λ⁶-sulfanyl)aniline, as well as 2-chloro-5-(trifluoromethyl)aniline, were synthesized according to a reported procedure (38). Likewise, benzo[d][1,2,3]thiadiazol-6-amine was synthesized following a reported synthetic sequence (12). Preparative normal phase chromatography was performed on a CombiFlash Rf 150 (Teledyne Isco) with prepacked RediSep Rf silica gel cartridges. Thin-layer chromatography was performed with aluminum-backed sheets with silica gel 60 F₂₅₄ (Merck, ref 1.05554), and spots were visualized with UV light and 1% aqueous solution of KMnO₄. Melting points were determined in open capillary tubes with an MFB 595010M Gallenkamp. Spectra 400 MHz ¹H, 100.6 MHz ¹³C, and 376.5 MHz ¹⁹F NMR were recorded on a Varian Mercury 400 or on a Bruker 400 Avance III spectrometers. The 500 MHz ¹H NMR spectra were recorded on a Varian Inova 500 spectrometer. The chemical shifts are reported in ppm (δ scale) relative to internal tetramethylsilane, and coupling constants are reported in Hertz (Hz). Assignments given for the NMR spectra of the new compounds have been carried out on the basis of DEPT, COSY ¹H/¹H (standard procedures), and COSY ¹H/¹³C (gHSQC and gHMBC sequences) experiments. IR spectra were run on Perkin-Elmer Spectrum RX I, Nicolet Avatar 320 FT-IR, or Perkin-Elmer Spectrum TWO spectrophotometers. Absorption values are expressed as wave-numbers (cm⁻¹); only significant absorption bands are given. High-resolution mass spectrometry (HRMS) analyses were performed with an LC/MSD TOF Agilent Technologies spectrometer. The elemental analyses were carried out in a Flash 1112 series Thermofinnigan elemental microanalyzator (A5) to determine C, H, N, and S. The structure of all new compounds was confirmed by elemental analysis and/or accurate mass measurement, IR, ¹H NMR, ¹³C NMR, and ¹⁹F NMR. The analytical samples of all the new compounds, which were subjected to pharmacological evaluation, possessed purity of ≥95% as evidenced by their elemental analyses or their high-pressure liquid chromatography-mass spectrometry (HPLC/MS). HPLC/MS were determined with an HPLC Thermo Ultimate 3000SD (Thermo Scientific Dionex) coupled to a photodiode array detector DAD-3000 (Thermo Scientific Dionex) and mass spectrometer LTQ XL ESI-ion trap (Thermo Scientific) with Xcalibur v2.2 acquisition software (Thermo Scientific) (HPLC-PDA-MS). Five microliters of sample 0.5 mg/ml in methanol was injected, using a ZORBAX Extend-C18 3.5 μm 2.1 × 50 mm column at 30°C. The mobile phase was a mixture of A = formic acid 0.05% in water and B = formic acid 0.05% in acetonitrile, with the method described as follows: flow 0.6 ml/min; gradient from 95% A–5% B to 100% B in 3 min, 100% B 3 min, from 100% B to 95% A–5% B in 1 min, 95% A–5% B 3 min. Purity is given as % of absorbance at 254 nm; UV-Vis spectra were collected every 0.2 s between 650 and 275 nm. Data from mass spectra were analyzed by electrospray ionization in positive mode every 0.3 s between 50 and 1,000 Da. Alternatively, some HPLC/MS were determined with an HPLC Agilent 1260 Infinity II LC/MSD coupled to a photodiode array and mass spectrometer. Five microliters of sample 0.5 mg/ml in methanol:acetonitrile was injected using an Agilent Poroshell 120 EC-C₁₈, 2.7 μm, 50 mm × 4.6 mm column at 40°C. The mobile phase was a mixture of A = water with 0.05% formic acid and B = acetonitrile with 0.05% formic acid, with the method described as follows: flow 0.6 ml/min; gradient from 95% A–5% B to 100% B in 3 min, 100% B 3 min, from 100% B to 95% A–5% B in 1 min, 95% A–5% B 3 min. Purity is given as % of absorbance at 254 nm.

Urea compounds 1, 3, 4, 7, 8, 9, 10, 11, 13, 15, 16, 25, and 26 were synthesized as previously reported (15). The synthesis of the remaining target compounds (2, 5, 6, 12, 14, 17, 18, 19, 20, 21, 22, 23, 24, 27, 28, 29, 30, and 31), including experimental details, can be found in the supplemental material.

Compounds and media.

For in vitro assays with NTS and adult S. mansoni, stock solutions (10 mM) of all synthesized diarylureas were prepared in dimethyl sulfoxide (DMSO) (Sigma-Aldrich, Switzerland). For in vivo efficacy and PK studies, selected test drugs were first dissolved in 5% (final volume) DMSO and then suspended in 2% (wt/vol) 2-hydroxypropyl-β-cyclodextrin solution (in tap water). The oral formulation was thoroughly vortexed and sonicated. Hank balanced salt solution (HBSS, 1×), M199 medium, RPMI 1640, and AlbuMax II were purchased from Gibco (Waltham MA, USA). Penicillin/streptomycin 10,000 U/ml and fetal calf serum (FCS) were purchased from Bioconcept AG (Allschwil, Switzerland). Podophyllotoxin (PTT) was commercially obtained from Sigma-Aldrich (Switzerland) and stock solutions (5 μg/ml) were prepared in L6 cells medium, supplemented with FCS and l-glutamine (Sigma-Aldrich, Switzerland).

Ethics.

In vitro and in vivo efficacy studies were carried out in accordance with Swiss national and cantonal regulations on animal welfare at the Swiss Tropical and Public Health Institute (Swiss TPH, Basel, Switzerland) under the permission number 2070.

Newly transformed schistosomula and adult Schistosoma mansoni.

The S. mansoni life cycle is maintained at the Swiss TPH, as described (39). To obtain NTS, cercariae were collected from infected Biomphalaria glabrata snails and were mechanically transformed. The NTS were kept in the incubator (37°C and 5% CO₂) in medium M199, supplemented with 5% FCS, 1% penicillin/streptomycin, and 1% (vol/vol) antibacterial/antifungal solution (39) until use. Adult S. mansoni worms were collected by dissecting the mesenteric veins of infected mice at day 49 postinfection. Worms were incubated in supplemented RPMI medium (5% FCS, 100 U/ml penicillin, and 100 μg/ml streptomycin) at 37°C and 5% CO₂ until use.

In vitro phenotypic screening assays.

For NTS and adult S. mansoni, transparent flat-bottom 96- and 24-well plates were used, respectively (Sarstedt, Switzerland). The activity of the test drugs was evaluated following an in-house-developed screening cascade. In brief, approximately 30 to 40 NTS were incubated with the respective test drug in 250 μl of M199 medium (Gibco, USA) supplemented with 5% (vol/vol) FCS (Bioconcept AG, Switzerland), 1% (vol/vol) penicillin/streptomycin solution (Sigma–Aldrich, Switzerland), and 1% (vol/vol) antibacterial/antifungal (39) solution for up to 72 h at 37°C and 5% CO₂. The drugs were initially tested at 10 μM in triplicate and repeated once; compounds that showed high activity (75 to 100% reduction of viability) against NTS were subsequently tested at 1 and 0.1 μM for IC₅₀ determination. Worms were judged via microscopic readout 72 h after incubation; they were scored according to phenotypic reference points such as motility, morphology, and granularity (scores from 0 to 3) (39). Identified hits from the NTS screening (>75% activity at 10 μM) were tested on S. mansoni adult worms. At least three worms (both sexes) were incubated in a final volume of 2 ml RPMI 1640 supplemented with 5% (vol/vol) FCS and 1% (vol/vol) penicillin/streptomycin at 37°C and 5% CO₂ for 72 h, at drug concentrations ranging from 10 to 0.05 μM. The experiment was conducted in duplicate. Finally, the possible influence of protein binding on the drug activity was tested at a 10 μM drug concentration by supplementing the medium with 45 g/liter BSA (AlbuMax II Lipid-Rich BSA, Gibco) (40). The experiment was conducted in duplicate and repeated once. For all in vitro assays, negative controls (using the highest concentration of DMSO) were included.

In vitro cytotoxicity assay with rat skeletal myoblast L6 cells.

Rat skeletal myoblast L6 cells were grown in RPMI 1640 supplemented with 10% FCS and 1.7 μM l-glutamine. Cell viability was monitored by microscopy with trypan blue. Cells were trypsinized with trypsin-EDTA (Sigma-Aldrich) and were then seeded at 2 × 10⁴ cells/ml in a transparent flat bottom 96-well plate for 24 h. The cells were drug treated (3-fold serial dilution, concentration range: 100 to 0.0017 μM) and incubated for 46 h at 37°C and 5% CO₂. Podophyllotoxin was used as a reference substance, starting at a concentration of 0.05 μg/ml. After 70 h, resazurin was added to the wells and, after another 2 h of incubation, the fluorescence was read using an excitation wavelength of 530 nm and an emission wavelength of 590 nm (SpectraMax, Molecular Devices; Softmax, version 5.4.6). Drugs were tested as singletons and repeated three times.

Metabolic microsomal stability.

Mouse liver microsomes (obtained from Thermo Fisher Scientific) were used to evaluate the metabolic stability of the test compounds 11, 15, 16, and 28. The test compounds were suspended in 0.1 M phosphate buffer (pH 7.4) and incubated at a final concentration of 1 μM at 37°C and a 0.5 mg/ml protein concentration. The metabolic reaction was initiated by the addition of an NADPH-regenerating system. Over a 60-min incubation period, the reaction was quenched at different time points by adding acetonitrile containing compound 26 (100 nM) as an internal standard, followed by centrifugation to pellet the precipitated material. Control samples (containing no NADPH) were included to monitor for potential degradation in the absence of the cofactor. The supernatant was evaporated (1.5 h, 45°C) and resuspended in a 80:20 acetonitrile:water mixture for the analysis. The method for analysis was adapted from Zarei et al. (36). Briefly, samples were analyzed by liquid chromatography-tandem mass spectrometry (LC-MS/MS), employing a 1260 infinity liquid chromatographic system connected to a 6460 triple quadrupole mass spectrometer (Agilent Technologies, Santa Clara, CA, USA) under positive electrospray ionization. Samples (10 μl) were injected and separated using a Waters Symmetry C₁₈ column (3.5 μm, 100 Å, 50 × 4.6 mm). A gradient cycle time of 2.6 min and a flow rate of 1 ml/min were employed. The mobile phase consisted of an acetonitrile/water gradient with 1% formic acid. The relative disappearance of parent compound over the course of the incubation was monitored, and results were reported as percentage (%) remaining after 60 min incubation. The experiment was conducted as singleton or duplicate and repeated once.

In vivo efficacy studies in a Schistosoma mansoni-infected mouse model.

For in vivo efficacy studies, female NMRI mice (age 3 weeks, ca. 20 to 22 g) were purchased from Charles River (Germany). Mice were kept under environmentally controlled conditions (temperature ∼25°C; humidity ∼70%; 12 h light and 12 h dark cycle) with access to water and rodent diet ad libitum. After an acclimatization phase of 1 week, mice were infected by subcutaneously injecting approximately 100 S. mansoni cercariae in phosphate-buffered saline (PBS). Cercariae of S. mansoni (Liberian strain) were obtained from infected intermediate host snails (B. glabrata) and mechanically transformed to NTS as described previously (39). Selected single doses (100 and 400 mg/kg) of test compounds were administered to groups of four to eight S. mansoni-infected mice by oral gavage 7 weeks after infection. Untreated mice served as controls. Control and treated mice were euthanized by the CO₂ method 3 weeks post drug administration. Worms were removed by picking, they were sexed and counted, and worm burden reduction was calculated (39).

Pharmacokinetic evaluation of four lead compounds.

For compounds 11, 15, 16, and 28, blood samples were collected alongside the in vivo efficacy study (100 and 400 mg/kg) at time points 1, 3, 5, 7, and 24 h post-dosing. Because some toxicity was observed in these studies, in a second step, urea 16 was selected for further evaluation of PK parameters at a lower dose (20 mg/kg); furthermore, the time frame for sample collection was extended. Compound 16 was orally administered to noninfected female NMRI mice (n = 4) at a single dose of 20 mg/kg. Blood samples were collected by puncture of the vein tail (approximately 50 μl, conscious sampling) at 1, 6, 24, 48, and 72 h. The micro-hematocrit capillaries (Marienfeld, Germany; containing heparin as an anticoagulant) were centrifuged (2,500 × g for 5 min) and plasma samples were transferred to Eppendorf tubes and stored at −20°C until analysis. Samples were assayed against a calibration curve (2,005 to 25,510 ng/ml for 11; 829 to 20,656 ng/ml for 15; 1,036 to 9,529 ng/ml for 16 at 20 mg/kg; 996 to 19,254 ng/ml for 16 at 100 and 400 mg/kg; and 2,286 to 24,624 ng/ml for 28) prepared in blank mouse plasma and processed together with the study samples. The drugs were extracted within 3 months of collection and samples were analyzed, as described in the metabolic stability method section, by liquid chromatography-tandem mass spectrometry (LC-MS/MS), employing a 1260 infinity liquid chromatographic system connected to a 6460 triple quadrupole mass spectrometer (Agilent Technologies, Santa Clara, CA, USA) under positive electrospray ionization.

Data analysis.

The in vitro activity against NTS and adult S. mansoni was calculated in Microsoft Excel using the mean viability values (± standard deviation [SD]) of drug-treated NTS or adult S. mansoni worms in relation to control parasite viability values. GraphPad Prism (Version 8.2.1(411)) was used to calculate IC₅₀ values from the NTS and adult S. mansoni screening based on the 2-parameter model (variable slope model: Y = 100/[1+[IC₅₀/X]^HillSlope]). IC₅₀ values against the mammalian cell line (L6) were generated using the Softmax software. Selectivity indexes (SI) were calculated by dividing the L6 cytotoxicity IC₅₀ by the adult S. mansoni IC₅₀. Averages and standard deviations of two microsomal stability assays were calculated using Microsoft Excel. The in vivo worm burden reductions were determined by calculating the mean values of living worms of each treatment group relative to the untreated mice; results were given in percentage (± SD). The Kruskal-Wallis test was employed for statistical significance in R (version 3.5.1). PK parameters were calculated using noncompartmental methods (PKSolver Version 2.0).