Antimalarial Trioxolanes with Superior Drug-Like Properties and In vivo Efficacy

Abstract

The emergence of artemisinin resistance, combined with certain sub-optimal properties of ozonide agents arterolane and artefenomel has necessitated the search for new drug candidates in the endoperoxide class. Our group has focused on trioxolane analogs with substitution patterns not previously explored. Here we describe the enantioselective synthesis of analogs bearing a trans-3” carbamate side chain and find these to be superior, both in vitro and in vivo, to the previously reported amides. We identified multiple analogs that surpass the oral efficacy of arterolane in the P. berghei model while exhibiting drug-like properties (logD, solubility, metabolic stability) similar or superior to next-generation clinical candidates like E209 and OZ609. While the preclinical assessment of new analogs is still underway, current data suggest the potential of this chemotype as a likely source of future drug candidates from the endoperoxide class.

Graphical Abstract

Despite significant recent progress in the prevention and treatment of malaria, infection by Plasmodium spp. parasites remains a cause of significant morbidity and mortality in sub-Saharan Africa and Southeast Asia.¹ Emerging artemisinin resistance resulting from mutations in the PfKelch13 (K13)^2–3 protein has spurred the search for new compounds that might replace current artemisinin analogs as the rapidly killing component in future antimalarial combination therapies. The 1,2,4-trioxolane artefenomel (OZ439)⁴ remains a strong candidate to play this role due to a prolonged exposure profile in human patients that when modelled in ring-stage survival assays^5–7 predicts for clinical efficacy against K13-mutant parasites. However, the 800 mg dose of artefenomel studied clinically has proven extremely challenging to formulate due to its poor aqueous solubility and lipophilic nature.⁸ Thus, despite promising pharmacodynamic (PD) properties, it remains unclear whether artefenomel will ultimately prove a suitable replacement for artemisinin in antimalarial combinations. Nor does a tetraoxane development candidate (E209),¹⁰ which bears an artefenomel-like structure and even higher reported LogD_7.4 value, seem likely to overcome these challenges. A recent resurgence of efforts aimed at identifying compounds with superior drug-like properties, notably enhanced solubility and stability to metabolism, has seen the nomination of OZ609⁶ (Figure 1) as the leading drug candidate from this class.

Figure 1.

Structure of arterolane¹¹ and newer trioxolane drug candidates artefenomel⁴ and OZ609⁶.

Arterolane¹¹ in fact exhibits excellent solubility and reduced lipophilicity as compared to the more recent drug candidates. However, a shorter half-life of the drug in infected patients predicts for inferior clinical efficacy, including against K13-mutant parasites where the more durable killing effect conferred by OZ439 and OZ609 is most desirable.⁷ It is worth noting that the distinct pharmacological properties of the various ‘OZ’ agents derive entirely from differences in the side chain emerging from the cyclohexane ring (Figure 1). Indeed, Vennerstrom and Charman first proposed¹² that the ferrous iron reactivity, and thus antiparasitic activity, of OZ compounds is governed by the conformational preferences of the cyclohexane ring, which are in turn determined by the nature of the pendant side chain. Following this reasoning, we considered that control over conformation might be equally well achieved with trans-R³ substitution (Figures 2 and 3). To test this notion, we devised a stereocontrolled synthesis of arterolane-like analogs bearing trans-R³ side chains¹³ and gratifyingly found that these compounds exhibited potencies and in vivo efficacy on par with or even superior to canonical OZ-like cis-R⁴ analogs (Figure 2).¹³

Figure 2.

Chemotypes explored by our laboratory previously¹³ and in the present communication.

Figure 3.

Iron reactivity is governed by conformational dynamics of the cyclohexane ring,¹² with the peroxide-exposed conformer (top right) comprising the ferrous iron-reactive species. 1,3-Diaxial interactions (bottom in red) disfavor the iron-reactive conformer in artefenomel, and similarly in the trans-R³ carbamate analogs described in herein.

These encouraging findings compelled further exploration of the trans-R³ chemotype as described herein. In particular, we considered that a carbamate side chain at R³ might swing conformational dynamics in a favorable direction with respect to iron reactivity. Thus, much as the bulky aryl R⁴ substituent of OZ439/OZ609 is thought to reduce iron reactivity by disfavoring the peroxide-exposed conformer, so we reasoned might lone-pair–lone-pair interactions of the carbamate sp³ O-atom and O-4 of the trioxolane ring (Figure 3). The ultimate goal of such exploration, naturally, is to identify new molecules that combine the PK/PD profile of artefenomel with the more favorable drug-like properties (reduced MW and logD, enhanced solubility) of arterolane.

Here we describe the enantioselective synthesis of trans-R³ carbamate-substituted trioxolanes. We generally find these analogs to exhibit superior in vitro potency and in vivo efficacy when compared to their amide congeners, consistent with our design hypotheses. The combination of excellent physiochemical properties and promising pharmacodynamics argues for the further study of this chemotype in the search for new antimalarial drug candidates with enhanced properties.

Results and Discussion

We devised an asymmetric synthesis of the desired (R,R) and (S,S) forms of carbamate analogs leveraging our previous findings¹⁴ that Griesbaum co-ozonolysis of 3-substituted cyclohexanones proceeds with good selectivity for the trans diastereomer. Controlling both stereocenters thus requires access to non-racemic cyclohexanone starting materials. A known asymmetric borylation¹⁵ reaction of cyclohexenone with B₂(pin)₂, mediated by a CuCl/(S)-Taniaphos complex, afforded intermediate (R)-2 in yields of 89–91% (Scheme 1). The use of (R)-Taniaphos in turn afforded (S)-2 in similarly high yield. Oxidation of 2 with NaBO₃ afforded intermediate 3, which was prone to undergo elimination and was best converted immediately to the silyl ether 4. Griesbaum co-ozonolysis reaction of 4 with two equivalents of adamantan-2-one O-methyloxime and ozone in CCl₄ afforded the expected 1,2,4-trioxolane intermediate 5 in yields of 77–94%. As expected, ¹H NMR analysis of 5 confirmed that the reaction proceeded with high diastereoselectivity (12:1 d.r). Deprotection of silyl ether 5 with TBAF in THF proceeded smoothly to give alcohol 6. Mosher ester analysis of intermediate alcohol 6 prepared via asymmetric and racemic routes confirmed the reported¹⁵ high enantioselectivity of the asymmetric borylation reaction (see Supporting Information).

Scheme 1.

Optimized enantiocontrolled synthetic route used to prepare analogs (R,R)-8a–n. The same route was used to prepare analogs (S,S)-9a–c, using (R)-Taniaphos in the first step.

To complete the synthesis of final analogs, alcohol 6 was activated as the p-nitrophenylcarbonate (7). Reaction of this key intermediate with primary or secondary amines afforded the final analogs (8a-n, 9a-c) in modest to excellent yields depending on the amine employed. A final, carefully controlled Boc deprotection step (AcCl, MeOH, 0°C) was required for analogs 8c/9c, 8j-k, and 8m-n. This new process is amenable to multi-gram laboratory scales and should serve to enable preclinical PK/PD and toxicological assessment of both enantiomeric forms. It is likely that drug candidates of this chemotype would be studied clinically as the racemate to reduce drug costs and leverage existing process routes to cis-R⁴ substituted agents like arterolane and artefenomel.

The new carbamate analogs were evaluated for antiplasmodial effect in the W2 strain using our standard protocol.¹⁶ With few exceptions, the carbamate analogs were found to be 2–3 fold more potent than amide-side chain congeners, modest differences that were nevertheless statistically significant (Chart 1). Carbamate enantiomer pairs 8a-c/9a-c were by contrast found to be essentially equipotent. The exceptional potency of analogs 8c and 8i bearing terminal primary amines inspired the synthesis of homologs 8j and 8k and ring contracted congeners 8m and 8n. These new analogs exhibited 2-fold weaker EC₅₀ values than 8c, with the exception of aminopyrrolidine 8n, which mostly retained the excellent potency of the aminopiperidine progenitor. While most of the analogs explored bore basic amines intended to enhance aqueous solubility, the presence of such functionality was not required for potent activity (e.g., 8f and 8l).

Chart 1.

In vitro activity of 1a–i,¹³ 8a–n, and 9a–c against W2 P. falciparum parasites (EC₅₀ ± SEM). Reported EC₅₀ values are the means of at least three determinations. EC₅₀ values for controls used in specific experiments are indicated at bottom of figure.

Next we evaluated the physiochemical and ADME properties of select carbamate analogs, using as a benchmark the measured or reported^{10, 17} values for current clinical candidates OZ439, OZ609, and E209 (Table 1). As expected, the new carbamate analogs exhibited much more favorable LogD/cLogD (pH 7.4) values compared to the current candidates, and this translated to superior aqueous solubility. Selected analogs were assessed for human liver microsome (HLM) stability and analogs 8a/9a, 8b/9b, and 9c showed notably showed superior T_1/2 and intrinsic clearance (CL_int) values when compared to measured or reported values for OZ439 and E209, though all were inferior to values for the OZ609 control.

Table 1.

In vitro ADME data for selected analogs and controls.

compound	HLM T1/2 (min)	HLM CLint (μL/min/mg)	MLM T1/2 (min)	MLM CLint (μL/min/mg)	Log D in octanol/PBS; pH 7.4 or cLogDd	Solubility in PBS pH 7.4 (μM)
OZ439a	21.8	63.7	165	8.4	5.81; 5.5c	0.181b
OZ609a	180.4	7.7	451	3.1	5.14	94.3b
OZ277b	>60	<7	n.t	n.t	2.6	504
E209c	68	25	132	13	6.4c; 5.5d	0.112b
8a	108.4	12.8	n.t.	n.t.	1.6c	n.t.
9a	>120	<11.5	n.t.	n.t.	1.6c	173.7
8b	96.8	14.3	n.t.	n.t.	2.4c	n.t.
9b	64.8	21.4	n.t.	n.t.	2.4c	164.1
8ca	11.3	123	9.0	155	2.44; 1.1c	144.5
9c	117.4	11.8	n.t.	n.t.	1.1c	118.0
8ea	35.2	39.4	4.3	321	3.25; 3.3c	157.0
8na	30.3	45.8	9.4	147	2.89; 1.6c	167.0
verapamila	9.4	147.6	5.2	268	n.t.	n.t.
diclofenaca	n.t.	n.t.	n.t	n.t	n.t.	199

^adata generated via Medicines for Malaria Venture.

^bdata from ref 17.

^cdata from ref 10.

^dcLogD (pH 7.4) calculated in Marvin Sketch v17.22 from ChemAxon.

We used the P. berghei mouse malaria model as a convenient means to assess the oral efficacy of selected new carbamate analogs, using arterolane (OZ277), 1a, or OZ439 as controls. The data presented in Table 2 reflects four separate studies examining different dosing paradigms in cohorts of five animals per compound/dosing arm, with animals judged to be cured if parasitemia could not be detected in blood smears at day 30. All regimens shown in Table 2 were well tolerated, with no overt signs of toxicity observed for any of the analogs studied. The data in experiment 1 is reproduced from our previous report,¹³ wherein we found that trans-R³ amide analogs such as 1a, and 1d, were at least as efficacious as OZ277 with four QD oral doses of 6 mg/kg. Experiment 2 was a bridge study designed to compare carbamate analogs 8a-e to arterolane and 1a, under a similar dosing schedule. We were pleased to find that the carbamate analogs were highly efficacious, the lone exception (8e) mirroring the similarly poor in vivo efficacy we observed¹³ for its direct amide congener 1e. Lowering the daily oral dose to just 2 mg/kg revealed the clear superiority of carbamate analogs 8a-d and 8i over arterolane and 1a. Reducing the number of daily doses of analog 8c led to fewer cures, although it was impressive that three, or even two QD doses at 4 mg/kg produced cures in some animals.

Table 2.

Oral efficacy of trans-R3 analogs and controls in P. berghei-infected female Swiss Webster mice.^a

compound	PO dose (days)	cures
Experiment 1b
OZ277	6 mg/kg/day (4 days)	4/5
1a		5/5
1b		0/5
1c		4/5
1d		5/5
1e		1/5
1g		0/5
Experiment 2
OZ277	4 mg/kg/day (4 days)	5/5
1a		5/5
8a		5/5
8b		5/5
8c		5/5
8d		5/5
8e		0/5
OZ277	2 mg/kg/day (4 days)	2/5
1a		1/5
8a		5/5
8b		4/5
8c		5/5
8d		5/5
8i		4/5
	4 mg/kg/day
8c	4 days	5/5
	3 days	4/5
	2 days	1/5
8c	40 mg/kg (1 day)	1/5
	80 mg/kg (1 day)	5/5
Experiment 3
OZ439	40 mg/kg (1 day)	5/5
8c		3/5
8j		1/5
8k		2/5
8n		3/5
Experiment 4
OZ439	40 mg/kg (1 day)	5/5
8c		5/5
9c		3/5

^aMice were judged cured if parasitemia was undetectable at day 30.

^bdata reported previously in ref 13

Compound 8c administered in a single 40 mg/kg or 80 mg/kg dose was partially or fully curative, respectively. The OZ439 control cured all animals at 40 mg/kg, and has been reported⁴ to be effective at a single dose of just 20 mg/kg in this model. In experiment 3 we explored the single-exposure efficacy of analogs with 8c-inspired side chains, employing 8c and OZ439 as controls. At a single 40 mg/kg dose, analog 8n was just as effective as 8c (3/5 cures), superior to 8j and 8k, but again inferior to OZ439 (5/5 cures). In experiment 4 we compared analog 8c with its (S,S) form 9c, and found the former to be more efficacious than the latter. Some variability in inoculation challenge likely explains the variable efficacy of 8c between experiments 2–4 (e.g. ranging between 1 and 5 cures at 40 mg/kg). Valid comparisons can nonetheless be made between compounds in contemporaneous experiments, as we have been careful to do here.

To better understand the results of these in vivo studies, we considered the in vitro microsome stability data and further, generated mouse pharmacokinetic data for selected analogs (8c and 8n as ‘actives’; 8e as ‘inactive’). The superior in vivo efficacy of 8c and 8n when compared to 8e was consistent with more rapid metabolism of the latter in mouse liver microsomes (MLM), although 8c and 8n were also prone to metabolism by MLM. In a mouse PK study with oral dosing, compound 8e achieved ~three-fold lower systemic exposure than 8c or 8n (Table 3 and Supporting Information), a difference that seemed insufficient to fully explain the poor efficacy of this analog relative to 8c/8n. In this regard our findings echo those of a previous report¹⁸ describing the SAR effort leading to OZ439, which noted that curative efficacy in the P. berghei model was not well correlated with exposure as determined in PK studies. Despite this challenge, superior PK and PD were ultimately married in the form of OZ439, whereas currently available data for carbamate analogs like 8c and 8n suggest compounds with superior PD properties but a sub-optimal PK profile, at least in mice. The much more rapid elimination of 8c/8e/8b vs. OZ439 is apparent when comparing plasma-concentration time courses (Supporting Information).

Table 3.

Selected pharmacokinetic parameters for analogs 8c, 8e, 8n, and OZ439 following a single dose of 50 mg/kg by oral gavage in male CD-1 mice.

compound	Tmax (hr)	Cmax (ng/mL)	AUClast (hr•ng/mL)
OZ439	2	5660 ± 732	52600 ± 4810
8c	2	631 ± 205	3430 ± 772
8e	2	223 ± 25	1070 ± 127
8n	2	594 ± 99	2800 ± 373

Ultimately PK in rodents is of secondary importance compared to PK in humans, and measures of human liver microsome (HLM) stability must be weighed alongside efficacy in mouse models. Although compound 8c became a focus of the in vivo studies detailed herein, compounds 8a/9a and 8b/9b in fact exhibit much more promising HLM CL_int values, similar or even superior to those of current clinical candidates (Table 1). This is noteworthy, considering that the carbamate analogs also exhibit highly favorable cLogD values and superior aqueous solubility, the very properties that are highly sought in next-generation endoperoxide drug candidates. Further optimization and preclinical assessment of the chemotypes described herein therefore seems advisable.

Conclusions.

Here we described the stereocontrolled synthesis of ozonide-class antimalarials bearing trans-3” carbamate substitution of the cyclohexane ring, informed by conformational design principles. We find that prototypical trans-R³ carbamate analogs like 8a/9a combine excellent in vitro activity and in vivo efficacy with high aqueous solubility and good stability toward cultured human liver microsomes. It remains to be seen whether extended, artefenomel-like, exposure profiles can be achieved in these arterolane-like trans-R³ carbamates. Overcoming current, formidable challenges in drug formulation and a higher than preferable human dose, remains our long-term goal and is the focus of ongoing efforts.

METHODS

Materials.

All chemical reagents were obtained commercially and used without further purification, unless otherwise stated. Anhydrous solvents were purchased from Sigma-Aldrich and used without further purification. Solvents used for flash column chromatography and reaction work-up procedures were purchased from either Sigma-Aldrich or Fisher Scientific. Column chromatography was performed on Silicycle Sili-prep cartridges using a Biotage Isolera Four automated flash chromatography system.

Instrumentation.

NMR spectra were recorded on either a Varian INOVA 400 MHz spectrometer (with 5 mm Quad-Nuclear Z-Grad Probe), or a Bruker AvanceIII HD 400 MHz (with 5mm BBFO Z-gradient Smart Probe), calibrated to CH(D)Cl₃ as an internal reference (7.26 and 77.00 ppm for ¹H and ¹³C NMR spectra, respectively). Data for ¹H NMR spectra are reported in terms of chemical shift (δ, ppm), multiplicity, coupling constant (Hz), and integration. Data for ¹³C NMR spectra are reported in terms of chemical shift (δ, ppm), with multiplicity and coupling constants in the case of C–F coupling. The following abbreviations are used to denote the multiplicities; s = singlet, d = doublet, t = triplet, q = quartet, m = multiplet, br = broad, app = apparent, or combinations of these. LC-MS and compound purity were determined using Waters Micromass ZQ 4000, equipped with a Waters 2795 Separation Module, Waters 2996 Photodiode Array Detector, and a Waters 2424 ELSD. Separations were carried out with an XBridge BEH C18, 3.5μm, 4.6 × 20 mm column, at ambient temperature (unregulated) using a mobile phase of water-methanol containing a constant 0.10 % formic acid.

Synthetic Procedures

(R)-3-((tert-butyldiphenylsilyl)oxy)cyclohexan-1-one (4).

A 200-mL round bottom flask equipped with a stirbar, rubber septum, and argon inlet was charged with (R)-3-hydroxycyclohexan-1-one¹⁵ (2.1 g, 18.4 mmol, 1.0 equiv), N,N-dimethylformamide (40 mL), and imidazole (2.51 g, 36.8 mmol, 2.0 equiv). The mixture was cooled at 0 °C while tert-butyl(chloro)diphenyl silane (5.3 mL, 20.2 mmol, 1.1 equiv) was added dropwise via syringe. The reaction mixture was allowed to slowly warm to rt. After stirring for 16 h, the reaction was judged complete based on TLC and LC/MS analysis. The reaction mixture was then diluted with EtOAc (100 mL) and DI H₂O (100 mL). The organic phase was separated and washed with brine (2 × 50 mL), dried (MgSO₄), filtered and concentrated to afford a colorless oil. The crude material was purified using flash column chromatography (330 g silica gel cartridge, 0–20% EtOAc–Hexanes, product eluted during 8% EtOAc–Hex) to give the desired ketone 4 (6.01 g, 17.05 mmol, 93%) as a colorless oil. ¹H NMR (400 MHz, CDCl₃) δ 7.68–7.72 (m, 4 H), 7.39–7.49 (m, 6 H), 4.23 (t, J = 4.9 Hz, 1 H), 2.47 (d, J = 5.0 Hz, 2 H), 2.35–2.42 (m, 1 H), 2.25–2.32 (m, 1 H), 2.17 (br dd, J = 8.5, 5.8 Hz, 1 H), 1.78–1.83 (m, 2 H), 1.64–1.71 (m, 1 H), 1.08–1.12 (s, 9 H); ¹³C NMR (100 MHz, CDCl₃) δ 210.0, 135.8, 135.7, 134.9, 133.9, 133.6, 129.9, 129.8, 127.7, 127.7, 71.1, 50.4, 41.2, 32.9, 26.9, 26.6, 20.6, 19.2; MS (ESI) calcd for C₂₂H₂₉O₂Si [M + H]⁺: m/z 353.19, found 353.46.

(1R,3”R)-tert-butyl((dispiro[adamantane-2,3’-[1,2,4]trioxolane-5’,1”-cyclohexan]-3”-yl)oxy) diphenylsilane (5).

A 200-mL recovery flask was charged with ketone (R)-4 (1.51 g, 4.28 mmol, ca. 1 equiv), carbon tetrachloride (100 mL), and O-methyl 2-adamantanone oxime (768 mg, 4.28 mmol, 1.0 equiv). The solution was cooled to 0 °C and sparged with O₂ for 10 minutes. The reaction was maintained at 0 °C while ozone was bubbled (2 L/min, 40% power) through the solution. After stirring for 90 mins, the reaction was judged to be incomplete based on TLC and LC/MS analysis so additional oxime (0.386 g, 2.14 mmol, 0.5 equiv) was added in a single portion to the reaction mixture followed by additional carbon tetrachloride (50 mL). Ozone was bubbled through the reaction mixture for another 45 mins, after which a third portion of oxime (0.386 g, 2.14 mmol, 0.5 equiv) was added and ozone again bubbled through the reaction for a final 45 mins. The solution was then sparged with O₂ for 10 minutes to remove any dissolved ozone, followed by sparging with argon gas for 10 minutes to remove any dissolved oxygen. The solution was then concentrated under reduced pressure to provide a viscous oil. The crude material was purified using flash column chromatography (120 g silica gel cartridge, 0–20% EtOAc–Hexanes, product eluted during 5% EtOAc–Hex) to give the desired trioxolane product 5 (2.01 g, 3.87 mmol, 91%, 12:1 dr) as a colorless semi-solid. ¹H NMR (400 MHz, CDCl₃) δ 7.69 (td, J = 7.7, 1.5 Hz, 4 H); 7.36–7.46 (m, 6 H), 3.89–3.96 (m, 1 H), 3.78–3.85 (m, 1 H), 1.95–2.05 (m, 3 H), 1.65–1.84 (m, 12 H), 1.46–1.65 (m, 3 H), 1.19–1.36 (m, 3 H), 1.08 (s, 9 H); ¹³C NMR (100 MHz, CDCl₃) δ 135.8, 135.8, 134.5, 134.4, 129.6, 129.5, 127.6, 127.5, 111.2, 109.2, 69.8, 43.8, 36.8, 36.3, 36.2, 34.8, 34.4, 33.8, 33.2, 27.0, 26.9, 26.5, 19.9, 19.2; MS (ESI) calcd for C₃₂H₄₂NaO₄Si [M + Na]⁺ m/z 541.28, found 541.56.

(1R,”R)-Dispiro[adamantane-2,3’-[1,2,4]trioxolane-5’,1”-cyclohexan]-3”-ol (6).

To a stirred solution of 5 (2.0 g, 3.86 mmol, 1.0 equiv) in THF (20 mL) was added a solution of tetrabutylammonium fluoride (1.0 M in THF, 19.2 ml, 19.3 mmol, 5.0 equiv) dropwise whilst stirring at 0 °C. The reaction mixture was allowed to slowly warm to rt and stirred for 12 h, at which point conversion was determined to be complete based on TLC and LC/MS analysis. The reaction was then diluted with brine (100 mL) and extracted with EtOAc (2 × 100 mL). The organic layer was then dried (Na₂SO₄), filtered, and concentrated under reduced pressure to afford a yellow oil. The crude material was purified using flash column chromatography (80 g silica gel cartridge, 0–50% EtOAc–Hexanes, product eluted during 20% EtOAc–Hexanes) to yield the desired product 6 (1.01 g, 3.60 mmol, 93%) as a colorless solid: ¹H NMR (400 MHz, CDCl₃) δ 3.94–4.14 (m, 1 H), 2.47 (br s, 1 H), 2.07 (d, J = 4.0 Hz, 1 H), 1.90–2.05 (m, 7 H), 1.69–1.88 (m, 12 H), 1.47–1.63 (m, 2 H); ¹³C NMR (100 MHz, CDCl₃) δ 111.9, 109.1, 67.9, 41.7, 36.7, 36.2, 36.2, 34.9, 34.9, 34.8, 34.7, 33.8, 33.1, 26.8, 26.4, 19.1; MS (ESI) calcd for C₁₆H₂₄O₄ [M + H]⁺ m/z 281.17, found 281.51.

(1R,3”R)-Dispiro[adamantane-2,3’-[1,2,4]trioxolane-5’,1”-cyclohexan]-3”-yl (4-nitrophenyl) carbonate (7).

To an oven-dried round bottom flask containing a magnetic stir bar under an Ar(g) atmosphere was added alcohol 6 (0.150 mg, 0.54 mmol, 1.0 equiv), dichloromethane (10 mL), N,N-diisopropylethylamine (0.30 mL, 1.74 mmol, 3.25 equiv), and 4-dimethylaminopyridine (0.078 g, 0.64 mmol, 1.2 equiv). The mixture was cooled to 0 °C while 4-nitrophenyl chloroformate (0.350 g, 1.74 mmol, 3.25 equiv) was added as a solid in two portions. The reaction mixture was allowed to warm to rt and stirred for 3 h. The reaction was diluted with DI H₂O (100 mL) and extracted with EtOAc (1 × 100 mL). The organic layer was washed repeatedly with 1 M aq K₂CO₃ solution until the aqueous layer was colorless and no longer yellow (indicating that most of the p-nitrophenol had been successfully removed from the organic layer). The organic layer was then dried (Na₂SO₄), filtered and concentrated under reduced pressure to yield a viscous yellow oil. The crude material was purified using flash column chromatography (80 g silica gel cartridge, 0–25% EtOAc–Hexanes, product eluted during 10% EtOAc–Hex) to yield the desired product 7 (208 mg, 0.467 mmol, 87%) as a pale yellow oil (93:7 dr). ¹H NMR (400 MHz, CDCl₃) δ 8.26 (d, J = 9.1 Hz, 2 H), 7.38 (br d, J = 9.1 Hz, 2 H), 4.94 (td, J = 9.2, 4.5 Hz, 1 H, minor diastereomer), 4.79–4.88 (m, 1 H), 2.32–2.42 (m, 1 H), 2.10 (br d, J = 8.8 Hz, 1 H), 1.69–2.00 (m, 1 H), 1.64–2.01 (m, 17 H), 1.40–1.64 (m, 3 H), 1.20–1.28 (m, 1 H), 0.83–0.96 (m, 1 H); ¹³C NMR (100 MHz, CDCl₃) δ 155.5, 151.6, 145.3, 125.3, 121.9, 121.7, 111.9, 108.3, 76.2, 39.5, 36.7, 36.3, 36.3, 34.8, 34.7, 34.7, 34.7, 33.5, 30.0, 26.8, 26.4, 19.5; MS (ESI) m/z [M+Na]+ calcd for C₂₃H₂₇NNaO₈: 468.16; found 467.99.

Representative procedure for synthesis of final analogues 8a-b and 8d-i. (1R,3”R)-Dispiro[adamantane-2,3’-[1,2,4]trioxolane-5’,1”-cyclohexan]-3”-yl l (2-amino-2-methylpropyl)carbamate (8a). To a solution of 7 (50 mg, 0.112 mmol, 1.0 equiv) in dichloromethane (1.5 mL) was added Et₃N (50 μL, 0.359 mmol, 3.2 equiv), followed by 1,2-diamino-2-methylpropane (60 μL, 0.561 mmol, 5.0 equiv) at rt. The bright yellow mixture was allowed to stir at rt for 3 h. The reaction was quenched with 1 M aq NaOH (20 mL) and diluted with EtOAc (30 mL). The organic phase was separated and washed with additional 1 M aq NaOH (4 × 30 mL) until the aqueous layer was colorless (indicating that p-nitrophenol had been successfully removed from the organic layer). The combined aqueous layers were then back extracted with EtOAc (1 × 30 mL). The combined organic layers were then washed with brine (20 mL), dried (MgSO₄), filtered and concentrated under reduced pressure. The crude residue was purified using flash column chromatography (12 g silica gel cartridge, 0–20% MeOH (containing 0.7 N NH₃)/CH₂Cl₂, with the desired product eluting during 15% MeOH (containing 0.7 N NH₃)/CH₂Cl₂). The fractions containing product were combined and lyophilized to give carbamate 8a (39.0 mg, 0.099 mmol, 88%) as a colorless solid. ¹H NMR (400 MHz, MeOD) δ 4.62–4.72 (m, 1H), 3.13–3.23 (m, 2H), 2.22–2.30 (m, 1H), 2.03 (br d, J = 12.7 Hz, 3H), 1.86–1.98 (m, 6H), 1.72–1.86 (m, 13H), 1.60–1.70 (m, 1H), 1.29–1.56 (m, 3H), 1.24 (s, 6H); ¹³C NMR (100 MHz, MeOD) δ 157.6, 111.4 (minor diastereomer), 111.2, 108.7, 108.5 (minor diastereomer), 71.4 (minor diastereomer), 71.1, 53.1, 49.6, 40.0, 39.6 (minor diastereomer), 36.4, 36.4, 36.4, 34.4, 33.4, 30.3, 29.4 (minor diastereomer), 27.0, 26.6, 23.7 (app d, J = 2.0 Hz), 19.5; MS (ESI) calculated for C₂₁H₃₅N₂O₅ [M + H]⁺: m/z 395.25, found 395.05.

(1R,3”R)-Dispiro[adamantane-2,3’-[1,2,4]trioxolane-5’,1”-cyclohexan]-3”-yl (3-amino-2,2-dimethylpropyl)carbamate (8b).

Prepared according to the standard procedure described above for 8a with the following modification: 2,2-dimethyl-1,3-propanediamine (59.1 mg, 0.561 mmol, 5.0 equiv) was added to the reaction as a solution in dichloromethane (0.5 mL). Purification via flash column chromatography (12 g silica gel cartridge, 0–75% EtOAc–Hexanes, followed by 0–20% MeOH (containing 0.7 N NH₃)/CH₂Cl₂, with the desired product eluting during 15% MeOH (containing 0.7 N ammonia)/CH₂Cl₂) afforded carbamate 8b (45.9 mg, 0.112 mmol, 100%) as a colorless oil. ¹H NMR (400 MHz, MeOD) δ 4.64 (tt, J = 10.0, 5.0 Hz, 1H), 2.97 (s, 2H), 2.41 (s, 2H), 2.29–2.14 (m, 1H), 2.08–1.98 (m, 2H), 1.98–1.87 (m, 5H), 1.87–1.68 (m, 11H), 1.65 (td, J = 12.8, 4.2 Hz, 1H), 1.48 (qt, J = 12.7, 3.3 Hz, 1H), 1.39–1.25 (m, 1H), 0.88 (s, 6H); ¹³C NMR (100 MHz, MeOD) δ 158.9, 112.5, 110.0, 72.1, 50.1, 48.6, 41.3, 37.8, 37.7, 35.8, 34.8, 31.7, 28.3, 27.9, 23.5, 20.8; MS (ESI) calcd for C₂₂H₃₆N₂O₅K [M + K]⁺: m/z 447.23, found 447.57.

(1R,3”R)-Dispiro[adamantane-2,3’-[1,2,4]trioxolane-5’,1”-cyclohexan]-3”-yl (2-aminoethyl) carbamate (8d).

Prepared according to the standard procedure described above for 8a with the following modification: ethylenediamine dihydrochloride (74.6 mg, 0.561 mmol, 5.0 equiv) was added to the reaction as a solution in dichloromethane (0.5 mL) and the reaction was stirred for 1 h. Purification was performed using flash column chromatography (12 g silica gel cartridge, 0–20% MeOH (containing 0.7 N NH₃)/CH₂Cl₂, with the desired product eluting during 15% MeOH (containing 0.7 N ammonia)/CH₂Cl₂) to give carbamate 8d (31.3 mg, 0.085 mmol, 76%) as a colorless oil. ¹H NMR (400 MHz, MeOD) δ 4.63 (tt, J = 9.9, 4.9 Hz, 1H), 3.19 (t, J = 6.2 Hz, 2H), 2.75 (t, J = 6.3 Hz, 2H), 2.28–2.14 (m, 1H), 2.08–1.98 (m, 2H), 1.98–1.87 (m, 5H), 1.87–1.68 (m, 11H), 1.63 (td, J = 12.8, 3.8 Hz, 1H), 1.54–1.40 (m, 1H), 1.40–1.27 (m, 1H); ¹³C NMR (100 MHz, MeOD) δ 158.6, 112.6, 110.0, 72.2, 43.4, 42.1, 41.3, 37.8, 37.8, 37.7, 35.8, 35.8, 34.7, 31.7, 28.4, 28.0, 20.8; MS (ESI) calcd for C₁₉H₃₀N₂O₅K [M + K]⁺: m/z 405.18, found 405.63.

(1R,3”R)-Dispiro[adamantane-2,3’-[1,2,4]trioxolane-5’,1”-cyclohexan]-3”-yl piperazine-1-carboxylate (8e).

Prepared according to the standard procedure described above for 8a with the following modification: piperazine (24.2 mg, 0.281 mmol, 2.5 equiv) was added to the reaction which was stirred at rt for 5 h. Purification via flash column chromatography (12 g silica gel cartridge, 0–75% EtOAc–Hexanes, followed by 0–20% MeOH (containing 0.7 N NH₃)/CH₂Cl₂, with desired product eluting during 10% MeOH (containing 0.7 N ammonia)/CH₂Cl₂) afforded carbamate 8e (33.8 mg, 0.086 mmol, 77%) as a colorless oil. ¹H NMR (400 MHz, CDCl₃) δ 4.86–4.96 (m, 1H, minor diastereomer), 4.72–4.85 (m, 1H), 3.35–3.52 (m, 4H), 2.81 (br s, 4H), 2.67 (br s, 1H), 2.09–2.25 (m, 1H), 1.86–2.02 (m, 8H), 1.62–1.81 (m, 13H), 1.46–1.56 (m, 1H), 1.31–1.42 (m, 1H); ¹³C NMR (100 MHz, CDCl₃) δ ppm 154.6, 111.6, 108.6, 71.2, 45.6, 44.4, 39.8, 36.7, 36.3, 36.2, 34.9, 34.8, 34.6, 34.6, 34.1, 30.6, 26.9, 26.4, 19.5; MS (ESI) calcd for C₂₁H₃₂N₂O₅K [M + K]⁺: m/z 431.19, found 431.53.

(1R,3”R)-Dispiro[adamantane-2,3’-[1,2,4]trioxolane-5’,1”-cyclohexan]-3”-yl morpholine-4-carboxylate (8f).

Prepared according to the standard procedure described above for 8a with the following modifications: morpholine (25 μL, 0.281 mmol, 2.5 equiv) was added to the reaction, which was stirred at rt for 3.5 h. Purification via flash column chromatography (12 g silica gel cartridge, 0–50% EtOAc–Hexanes, with desired product eluting during 15% EtOAc–Hexanes) afforded carbamate 8f (43.5 mg, 0.111 mmol, 99%) as a colorless oil. ¹H NMR (400 MHz, MeOD) δ 4.93–4.85 (m, 1H, minor diastereomer), 4.77 (tt, J = 9.1, 4.5 Hz, 1H), 3.71–3.55 (m, 4H), 3.54–3.35 (br s, 4H), 2.24–2.07 (m, 1H), 2.07–1.85 (m, 7H), 1.85–1.65 (m, 12H), 1.57–1.39 (m, 2H); ¹³C NMR (100 MHz, MeOD) δ 156.4, 112.7, 109.8, 73.1, 67.5, 40.5, 37.8, 37.7, 35.9, 35.8, 35.7, 35.7, 35.0, 31.4, 28.3, 27.9, 20.5; MS (ESI) calcd for C₂₁H₃₁NO₆K [M + K]⁺: m/z 432.18, found 432.53.

(1R,3”R)-Dispiro[adamantane-2,3’-[1,2,4]trioxolane-5’,1”-cyclohexan]-3”-yl piperidine-1-carboxylate (8g).

Prepared according to the standard procedure described above for 8a with the following modifications: piperidine (28 μL, 0.281 mmol, 2.5 equiv) was added to the reaction which was stirred at rt for 2 h. Purification via flash column chromatography (12 g silica gel cartridge, 0–50% EtOAc–Hexanes, with desired product eluting during 10% EtOAc–Hexanes) afforded carbamate 8g (43.3 mg, 0.111 mmol, 99%) as a colorless oil. ¹H NMR (400 MHz, CDCl₃) δ 4.90 (tt, J = 8.7, 4.4 Hz, 1H, minor diastereomer), 4.79 (tt, J = 9.3, 4.6 Hz, 1H), 3.52–3.25 (m, 4H), 2.25–2.11 (m, 1H), 2.06–1.85 (m, 7H), 1.85–1.61 (m, 12H), 1.61–1.43 (m, 7H), 1.43–1.32 (m, 1H); ¹³C NMR (100 MHz, CDCl₃) δ 154.7, 111.5, 108.6, 70.8, 44.6, 39.8, 36.7, 36.3, 36.2, 34.9, 34.8, 34.6, 34.5, 34.1, 30.6, 26.9, 26.4, 24.4, 19.5; MS (ESI) calcd for C₂₂H₃₃NO₅K [M + K]⁺: m/z 430.20, found 430.57.

(1R,3”R)-Dispiro[adamantane-2,3’-[1,2,4]trioxolane-5’,1”-cyclohexan]-3”-yl 4-methyl piperazine-1-carboxylate (8h).

Prepared according to the standard procedure described above for 8a with the following modifications: 1-methylpiperazine (31.0 μL, 0.281 mmol, 2.5 equiv) was added to the reaction. Purification via flash column chromatography (12 g silica gel cartridge, 0–75% EtOAc–Hexanes, followed by 0–20% MeOH (containing 0.7 N NH₃)/CH₂Cl₂, with desired product eluting during 5% MeOH (containing 0.7 N ammonia)/CH₂Cl₂) afforded carbamate 8h (24.8 mg, 0.061 mmol, 54%) as a colorless oil. ¹H NMR (400 MHz, CDCl₃) δ 4.91 (tt, J = 8.3, 4.2 Hz, 1H, minor diastereomer), 4.79 (tt, J = 9.2, 4.6 Hz, 1H), 3.58–3.35 (m, 4H), 2.42–2.26 (m, 7H), 2.23–2.13 (m, 1H), 2.05–1.84 (m, 7H), 1.84–1.61 (m, 12H), 1.58–1.47 (m, 1H), 1.43–1.33 (m, 1H); ¹³C NMR (100 MHz, CDCl₃) δ 154.5, 111.5, 108.5, 71.2, 54.6, 46.0, 39.7, 36.7, 36.2, 36.2, 34.9, 34.8, 34.6, 34.5, 34.1, 30.5, 26.8, 26.4, 19.4; MS (ESI) calcd for C₂₂H₃₄N₂O₅K [M + K]⁺: m/z 445.21, found 445.55.

(1R,3”R)-Dispiro[adamantane-2,3’-[1,2,4]trioxolane-5’,1”-cyclohexan]-3”-yl [(trans)-4-aminocyclohexyl]carbamate (8i).

Prepared according to the standard procedure described above for 8a with the following modifications: dichloromethane was substituted for THF (2 mL) and trans-1,4-cyclohexanediamine (32 mg, 0.281 mmol, 2.5 equiv) was added to the reaction which was stirred for 1.5 h at rt. The reaction was quenched with 1M aq NaOH (20 mL) and extractions were performed with dichloromethane (1 × 30 mL) rather than EtOAc. Purification via flash column chromatography (12 g silica gel cartridge, 0–100% EtOAc–Hexanes, followed by 0–20% MeOH (containing 0.7 N NH₃)/CH₂Cl₂, with the desired product eluting during 10–15% MeOH (containing 0.7 N ammonia)/CH₂Cl₂) afforded carbamate 8i (36.8 mg, 0.088 mmol, 78%) as a white foam. ¹H NMR (400 MHz, MeOD) δ 4.63 (tt, J = 9.8, 4.8 Hz, 1H), 3.45–3.27 (m, 1H), 2.78–2.60 (m, 1H), 2.26–2.13 (m, 1H), 2.10–1.99 (m, 2H), 1.99–1.88 (m, 9H), 1.88–1.70 (m, 11H), 1.64 (td, J = 12.7, 3.7 Hz, 1H), 1.56–1.42 (m, 1H), 1.40–1.19 (m, 5H); ¹³C NMR (100 MHz, MeOD) δ 157.6, 112.5, 110.0, 71.9, 54.8, 50.7, 50.7, 41.3, 37.8, 37.7, 35.8, 35.8, 34.8, 34.7, 32.5, 31.7, 28.3, 27.9, 20.8; MS (ESI) calcd for C₂₃H₃₇N₂O₅ [M + H]⁺: m/z 421.27, found 421.65.

(1R,3”R)-Dispiro[adamantane-2,3’-[1,2,4]trioxolane-5’,1”-cyclohexan]-3”-yl (2-amino-2-oxoethyl)carbamate (8l)

To a solution of carbonate 7 (54 mg, 0.121 mmol, 1.0 equiv) in dichloromethane (1.5 mL) was added glycinamide HCl (27.1 mg, 0.245 mmol, 2.0 equiv) followed by N,N-diisopropylethylamine (0.1 μL, 0.574 mmol, 4.74 equiv) at rt. The resulting suspension was stirred for 3 h, after which it was determined that the reaction was not progressing. The solution was then purged with Ar(g) for 10 min to evaporate all volatile organic materials from the mixture before DMSO (0.50 mL) and Et₃N (78 μL, 0.561 mmol, 4.6 equiv) were added to the reaction. The resultant homogenous yellow solution was stirred at rt for 3 h before being judged complete by TLC and LCMS. The reaction was then diluted with EtOAc (30 mL) and DI H₂O (75 mL). Following separation of the layers, the organic layer was washed with satd aq NaHCO₃ (2 × 30 mL) and the combined aqueous layers were back extracted with EtOAc (1 × 30 mL). The combined organic layers were washed with brine (1 × 20 mL), dried (Na₂SO₄), filtered and concentrated under reduced pressure to a crude residue. Purification via flash column chromatography (12 g silica gel cartridge, 0–50% EtOAc–Hexanes, followed by 0–20% MeOH (containing 0.7 N NH₃)/CH₂Cl₂, with desired product eluting during 15% MeOH (containing 0.7 N NH₃)/CH₂Cl₂) afforded carbamate 8l (40.4 mg, 0.106 mmol, 88%) as a white foam. ¹H NMR (400 MHz, MeOD) δ 4.65 (tt, J = 10.1, 5.0 Hz, 1H), 3.74 (s, 2H), 2.30–2.18 (m, 1H), 2.10–1.87 (m, 7H), 1.87–1.68 (m, 11H), 1.63 (td, J = 12.8, 4.0 Hz, 1 H), 1.54–1.26 (m, 2H); ¹³C NMR (100 MHz, MeOD) δ 174.9, 158.4, 112.6, 110.0, 72.6, 44.4, 41.2, 37.8, 37.7, 35.8, 35.7, 34.7, 31.6, 28.3, 27.9, 20.8; MS (ESI) calcd for C₁₉H₂₈N₂O₆K [M + K]⁺: m/z 419.16, found 419.50.

Plasmodium falciparum EC₅₀ determinations

The growth inhibition assay for P. falciparum was conducted as described previously¹⁶ with minor modifications. Briefly, P. falciparum strain W2 synchronized ring-stage parasites were cultured in human red blood cells in 96-well flat bottom culture plates at 37 °C, adjusted to 1% parasitemia and 2% hematocrit under an atmosphere of 3% O₂, 5% CO₂, 91% N₂ in a final volume of 0.1 mL per well in RPMI-1640 media supplemented with 0.5% Albumax, 2 mM L-glutamine and 100 mM hypoxanthine in the presence of various concentrations of inhibitors. Tested compounds were serially diluted 1:3 in the range 10,000 – 4.6 nM (or 1,000–0.006 nM for more potent analogs), with a maximum DMSO concentration of 0.1%. Following 48 hours of incubation, the cells were fixed by adding 0.1 ml of 2% formaldehyde in phosphate buffered saline, pH = 7.4 (PBS). Parasite growth was evaluated by flow cytometry on a FACsort (Becton Dickinson) equipped with AMS-1 loader (Cytek Development) after staining with 1 nM of the DNA dye YOYO-1 (Molecular Probes) in 100 mM NH₄Cl, 0.1% Triton x-100 in 0.8% NaCl. Parasitemias were determined from dot plots (forward scatter vs. fluorescence) using CELLQUEST software (Becton Dickinson). EC₅₀ values for growth inhibition were determined from plots of percentage control parasitemia over inhibitor concentration using GraphPad Prism software.

P. berghei Mouse Malaria Model

Female Swiss Webster Mice (~20 g body weight) were infected intraperitoneally with 10⁶ P. berghei-infected erythrocytes collected from a previously infected mouse. Beginning 1 hour after inoculation the mice were treated once daily by oral gavage for 1–4 days with 100 μL of solution of test compound formulated in 10% DMSO, 40% [20% 2-hydroxypropyl-beta-cyclodextrin in water], and 50% PEG400. There were five mice in each test arm. Infections were monitored by daily microscopic evaluation of Giemsa-stained blood smears starting on day seven. Parasitemias were determined by counting the number of infected erythrocytes per 1000 erythrocytes. Body weight was measured over the course of the treatment. Mice were euthanized when parasitemia exceeded 50% or when weight loss of more than 15% occurred. Parasitemias, animal survival, and morbidity were closely monitored for 30 days post-infection when experiments were terminated.

Animal Welfare

No alternative to the use of laboratory animals is available for in vivo efficacy assessments. Animals were housed and fed according to NIH and USDA regulations in the Animal Care Facility at San Francisco General Hospital. Trained animal care technicians provide routine care, and veterinary staff are readily available. Euthanasia was performed when malarial parasitemias top 50%, a level that does not appear to be accompanied by distress, but predicts progression to lethal disease. Euthanasia was accomplished with CO₂ followed by cervical dislocation. These methods are in accord with the recommendations of the Panel on Euthanasia of the American Veterinary Medical Association. Our studies are approved by the UCSF Committee on Animal Research.

Synopsis:

Here Blank, et. al. describe an antimalarial chemotype intended to combine the favorable physiochemical properties of arterolane with the superior pharmacodynamics of artefenomel. Design principles are detailed, as are in vitro and in vivo data for several analogs, including compound 8a which is curative in mouse malaria models with QD oral doses as low as 2 mg/kg.