Aminoalkoxycarbonyloxymethyl Ether Prodrugs with a pH-triggered Release Mechanism: A Case Study Improving Solubility, Bioavailability, and Efficacy of Antimalarial 4(1H)-Quinolones with Single Dose Cures

Abstract

Preclinical and clinical development of numerous small molecules is prevented by their poor aqueous solubility, severely limiting absorption and oral bioavailability. Herein, we disclose a general prodrug approach that converts promising lead compounds into aminoalkoxycarbonyloxymethyl (amino AOCOM) ether-substituted analogues that display significantly improved aqueous solubility and enhanced oral bioavailability, restoring key requirements typical for drug candidate profiles. The prodrug is completely independent of biotransformations and animal-independent because it activates via a pH-triggered intramolecular cyclization-elimination reaction. As a proof-of-concept, the utility of this novel amino AOCOM ether prodrug approach was demonstrated on an antimalarial compound series representing a variety of antimalarial 4(1H)-quinolones, which entered and failed preclinical development over the last decade. With the amino AOCOM ether prodrug moiety, 3-aryl-4(1H)-quinolone preclinical candidate was shown to provide single-dose cures in a rodent malaria model at an oral dose of 3 mg/kg, without the use of an advanced formulation technique.

Graphical Abstract

INTRODUCTION

Poor aqueous solubility, rate of dissolution, as well as poor permeability of lead candidates are the most frequent causes of low oral bioavailability.¹ Intriguingly, almost half of the drug candidates deriving from high-throughput campaigns have poor aqueous solubility of less than 10 μM.² Furthermore, the therapeutic use of lead candidates entering the drug development phase is significantly diminished because of limited aqueous solubility.³ Therefore, to overcome solubility barriers, numerous formulation techniques such as the use of solubilizing agents, the formation of salt forms, as well as various approaches to modify particle sizes have shown to improve oral bioavailability. Alternatively, the development of a prodrug – deliberately modified forms of the active ingredient that can undergo an enzymatic and/or chemical transformation in vivo – is a well-documented alternative approach to increase aqueous solubility and oral bioavailability.^{4, 5}

Only a few successful water-soluble prodrugs for oral administration have been developed thus far. Among the most promising ones are those based on phosphate esters, which undergo rapid hydrolytic transformations to the parent drug by endogenous alkaline phosphatases.⁵ Despite the satisfactory chemical stability of phosphate ester prodrugs and similar phosphate ester hydrolysis rates by alkaline phosphatases between different animal species, phosphates can also lead to a precipitation of the parent compound in the intestinal lumen.⁵ An alternative approach that can regenerate the parent compound via chemical methods has been identified as a superior method to avoid inter- and intraspecies variability with enzymatic activities altogether.⁶ Given that numerous solubility-limited molecules with therapeutic potential are required to overcome poor oral bioavailability to enter preclinical development, we present herein aminoalkoxycarbonyloxymethyl (amino AOCOM) ether as a solubilizing, pH-activated prodrug moiety to significantly increase aqueous solubility and oral bioavailability. As a proof-of-concept, the utility of this novel amino AOCOM ether prodrug approach was demonstrated on an antimalarial compound series representing a variety of antimalarial 4(1H)-quinolones and close analogues,⁷ which entered and failed preclinical development over the last decade.^8–13 The amino AOCOM prodrugs release the parent 4(1H)-quinolone in vivo via a pH-triggered intramolecular cyclization reaction, enhancing oral bioavailability and affording very potent in vivo efficacy.

Cyclization-activated prodrugs have been widely used to improve delivery of the drugs with low bioavailability mainly due to poor absorption or first pass metabolism.⁶ However, this is a first successful report of bioavailability improvement by using amino AOCOM ethers to release a solubility-limited parent drug in a pH controllable fashion. Delivery of lead 4(1H)-quinolone compound P4Q-391(2) from its prodrug was enhanced 18-fold (relative to the administration of the parent 4(1H)-quinolone), reaching a C_max of 9.1 μM in 4 h following oral administration (single dose, 10 mg/kg). Furthermore, in vivo efficacy of various 4(1H)-quinolones was significantly improved with the prodrug of 2 curing P. berghei-infected mice in a single oral dose of 3 mg/kg, without the use of an advanced formulation.

Antimalarial drug discovery mostly focuses on the erythrocytic stage of malaria, which cause the disease. In order to combat the pernicious problem of parasitic resistance and eradicate the disease, it would benefit the community to develop agents capable of blocking multiple stages of the parasite life cycle. The best-known antimalarial drug that kills dormant liver stages and gametocytes is the 8-aminoquinoline, primaquine, which was developed more than 20 years ago.^14–16 Unfortunately, primaquine causes hemolysis in individuals with a glucose-6-phosphate dehydrogenase deficiency (an estimated 400 million people worldwide).¹⁷ To guide the development of new antimalarials, the Malaria Eradication Research Agenda (malERA) initiative defined Target Product Profiles (TPPs) for antimalarial drugs to treat and prevent malaria infections, or to be used for radical cures (eradicate dormant exoerythrocytic (EE) stages of the parasite) of P. falciparum or P. vivax (see Supplementary section ‘Target Product Profiles’).¹⁰ These TPPs list important benchmark criteria, such as potent activity against resistant parasites, good oral bioavailability, a specific mechanism of action to effectively target multiple stages of the parasite life cycle, a shelf life of 5 years, low costs of active ingredients and formulations in the final medicine (< $0.25 per dose), and others.^{18, 19} An ambitious Single Exposure Radical Cure and Prophylaxis (SERCaP) treatment would require the ideal drug to be potent enough to work in a single, curative dose to treat P. falciparum and P. vivax infections.^{19, 20} A curative dose, in this context, is one which eliminates all persistent blood-stages, gametocytes and hypnozoites of the parasite. To the best of our knowledge, of the antimalarials currently in clinical trials, only ozonide OZ429,²¹ aminopyridine MMV390048,^{22, 23} 3,4-dihydro-1(2H)-isoquinolone (+)-SJ733,²⁴ spiroindolone KAE609²⁵ and triazolopyrimidine DSM265²⁶ have been reported to be a part of a single exposure radical cure initiative (PO dose 20 mg/kg for OZ429, 30 mg/kg for MMV390048, 100 mg/kg for KAE609, lowest single-cure dose data has not been reported for (+)-SJ733 and DSM265).

Recent evaluation and optimization studies of antimalarial 4(1H)-quinolones,^{8, 27–31} 4(1H)-pyridones,³² 1,2,3,4-tetrahydroacridones,^{33, 34} 4(1H)-quinolone esters,^{11, 35–37} and 2-aryl-4(1H)-quinolones³⁸ led to new agents with potent in vitro and in vivo erythrocytic stage activity and improved physicochemical properties.³⁹ Extensive development of the 3-aryl-4(1H)-quinolone⁹ chemotype series resulted in frontrunner compound ELQ-300 (1) and its close analogue 2 (Fig. 1).⁷ These compounds are potent and selective inhibitors of the parasite’s mitochondrial cytochrome bc₁ complex and efficiently target the blood, the liver, and the transmitting stages of the parasite in murine models. Spearheaded by the Medicines for Malaria Venture,1 entered preclinical development in 2013.⁷ Unfortunately, the advancement of 1 towards Phase I studies was deferred due to poor oral bioavailability, limiting preclinical safety and toxicity studies. Moreover, the absence of dose-proportionality impeded the determination of therapeutic index and in vivo toxicity.⁷ DMPK studies with lead 4(1H)-quinolone compounds suggested aqueous solubility to be the major reason for poor oral bioavailability in the series.⁴⁰ Therefore, this particular 4(1H)-quinolone compound series was considered to be ideally suited for developing a new solubilizing prodrug approach. Finally, this will prove the versatility and applicability of the developed amino AOCOM ether prodrug to the 4(1H)-quinolone scaffold and any other compound with an appropriate prodrug attachment site.

Figure 1.

Compounds of the antimalarial 3-aryl-4(1H)-quinolone³² chemotype series including frontrunner compounds ELQ-300 (1) and its close analogue P4Q-391 (2).

The conversion of 4(1H)-quinolones 5 into an ethylcarbonate prodrug, utilizing the reactivity of the hydroxy group of the respective tautomeric 4-quinolinol 6 (Fig. 2), was recently described by Riscoe and co-workers.⁴⁰ One of the major disadvantages of this approach, which may complicate the development, is the reliance of the carbonate prodrug on esterases for the release of the parent drug.⁴¹ For example, differences between specific esterase activities in various animal models possibly complicate dosing predictions for in vivo efficacy and pharmacokinetics in humans.⁵ More importantly, the reported carbonate prodrug of 1 required a dissolution step using neat PEG 400 prior to performing in vivo efficacy and PK studies.⁴⁰ Originally, this esterase-dependent prodrug approach was designed to overcome the poor oral bioavailability of antimalarial 4(1H)-quinolones 5 by reducing crystallinity and enhancing aqueous solubility. Unfortunately, no solubility data for the carbonate prodrug was reported. Thus, our studies toward a soluble quinolone prodrug began with investigating the altered crystal packing of 4(1H)-quinolone prodrugs and determining its effect on aqueous solubility and dissolution rate. We intended to prepare a series of easily accessible prodrugs by utilizing the hydroxy group of the tautomeric 4-quinolinols 6 as the attachment site of the prodrug moiety (Fig. 2). By linking the promoiety onto the oxygen of the 4-quinolinol 6, our prodrug approach is tailored to be applicable to any 4(1H)-quinolones independently of their core substitutions. Previous studies have shown that methylation of 4(1H)-quinolones provided mixtures of N- and O-methylated products in a 2:1 ratio, whereas alkylations with larger alkyl halides yielded predominantly O-substituted 4-quinolinols^{7–10, 30, 42} although the 4(1H)-quinolone tautomer is favored in both solid and solution states.⁴²

Figure 2.

Prodrug strategies for 4(1H)-quinolones 5.

RESULTS AND DISCUSSIONS

For investigation and optimization of a suitable promoiety, known 3-aryl-4(1H)-quinolones 5 were synthesized as previously described via a Conrad-Limpach cyclization of 2-aryl substituted ethyl acetoacetate with 3-chloro-4-methoxy aniline (Fig. 2).^{8, 31} The first set of carbamate, ester and carbonate prodrugs 8–11 (Scheme 1) was obtained in moderate to good yields by derivatization of the 4-hydroxy group of 7 with corresponding electrophiles and cesium carbonate in DMF. The parent 4(1H)-quinolone 3 was chosen, as it was shown to be slightly more soluble in comparison to the majority of the 3-aryl-4(1H)-quinolones.⁹ Furthermore, the preparation of compound 3 was straightforward in comparison to others such as 4(1H)-quinolone 2. All prodrugs were tested for in vitro activity against the clinically relevant multidrug-resistant malarial strains W2 (pyrimethamine and chloroquine resistant strains) and TM90-C2B (mefloquine, chloroquine, atovaquone, pyrimethamine resistant strains) following previously reported procedures (Tables 1 and 2).^{9, 35} Aqueous solubility at pH 7.4 was determined using a previously reported protocol.^{8, 33}

Scheme 1.

Synthesis of 6-chloro-7-methoxy-2-methyl-3-(o-tolyl)-4(1H)-quinolone 3 via Conrad-Limpach reaction²⁴ and conversion to first set of prodrugs (8-11) via O-acylation.

Table 1.

In vitro antimalarial activities and solubility of 7 and prodrugs 8–11

No		W2 pEC50 a	W2 SD pEC50 a	TM90-C2B EC50 a	TM90-C2B SD EC50 a	Aqueous solubility pH 7.4 [μM] b
7	-H	7.66	0.28	7.42	0.52	8.7
8		<5.64	0.68	<5.45	0.85	< 1.0
9		<5.64	0.68	<5.45	0.61	2.9
10		7.20	0.25	7.20	0.30	2.9
11		6.77	0.29	6.90	0.20	< 1.0

^aEC₅₀ data represents means for at least two independent experiments. Chloroquine (CQ), atovaquone (ATO) and dihydroartemisinin (DHA) are internal controls for each in vitro assay: CQ pEC₅₀ (W2) 6.38 and pEC₅₀ (TM90-C2B) 6.64; ATO pEC₅₀ (W2) 8.85 and pEC₅₀ (TM90-C2B) 4.74; DHA pEC₅₀ (W2) 8.26 and pEC₅₀ (TM90-C2B) 8.23.

^bAqueous solubility data represent means for four independent measurements. Standards for the solubility assay include carbamazepine and albendazole. Solubility for carbamazepine at pH 7.4 was 95 μM. Solubility for albendazole at pH 7.4 was 6.1 μM.

Table 2.

In vitro antimalarial activities and solubility

No			W2 pEC50	W2 SD pEC50	C2B pEC50	C2B pEC50	Aqueous solubility [mM]b
	R1	R2					pH 2.4	0.5% HEC
25		o-CH3	6.66	0.32	6.97	0.16	>0.1	>21
26		o-CH3	6.57	0.30	6.77	0.18	>0.1	>20
27		p-CF3	-c	-	-c	-	-c	>5.6
28		p-CF3	-c	-	-c	-	0.6	>5.5
13			<5.08	0.52	-c	-	-c	-c
14			inactive		inactive	-	-c	-c

^apEC₅₀ data represents means for at least two independent experiments. Chloroquine (CQ), atovaquone (ATO) and dihydroartemisinin (DHA) are internal controls for each in vitro assay: CQ pEC₅₀ (W2) 6.38 and pEC₅₀ (TM90-C2B) 6.64; ATO pEC₅₀ (W2) 8.85 and pEC₅₀ (TM90-C2B) 4.74; DHA pEC₅₀ (W2) 8.26 and pEC₅₀ (TM90-C2B) 8.23.

^bAqueous solubility data represent means for four independent measurements. Standards for the solubility assay include carbamazepine and albendazole. Solubility for carbamazepine at pH 2.4 was 100 μM. Solubility for albendazole at pH 2.4 was 100 μM. The aqueous solubility of the parent compounds 3 and 4 at pH 2.4 is 6.9 μM and <1.0 μM respectively.

^cnot determined.

In earlier studies, N- and O-methylated 4(1H)-quinolones displayed no antimalarial activity in vitro.²⁰ Similarly, carbamates 8 and 9 lacked in antimalarial activity with pEC₅₀ values smaller than 5.64 (Table 1). In contrast, ester 10 and carbonate 11were determined to be 10-fold less potent than reference compound 3. These results suggest that the ester and carbonate prodrugs are partially hydrolyzed into the active parent molecule 3 during the multiday incubation conditions, whereas compounds 8 and 9 are most likely inactive due to the high stability of the carbamate group. Despite the previous reported increased oral bioavailability utilizing an ethyl carbonate prodrug of 4(1H)-quinolone 1,⁴⁰ carbamate (8, < 1.0 μM and 9, 2.9 μM, Table 1), ester (10, 2.9 μM) and carbonate (11, <1.0 μM) linked promoieties do not noticeably improve the aqueous solubility of reference compound 3 (8.7 μM) and thus we did not consider these prodrug approaches to be adequately suited for further development.

To optimize the compromise between carbonate prodrug’s stability and the release rate of corresponding parent 4(1H)-quinolone, especially electron withdrawing 4(1H)-quinolones, we introduced a methylene bridge between the carbonate group and the 4-quinolinol’s oxygen, leading to an alkoxycarbonyloxymethyl (AOCOM) ether prodrug.⁴³ Using the AOCOM approach, the promoiety’s carbonyl group is electronically insulated from the 4(1H)-quinolone’s core, which aids to the stability of the prodrug. It was previously shown that linking of an ethoxycarbonyloxymethyl promoiety onto an aromatic hydroxy group of parent drugs lead to prodrugs that are stable to chemical hydrolysis at pH 1 (t_1/2 ≥ 50 h) and undergo slow hydrolysis (t_1/2 ≥ 8 h) at pH 6.0–8.5.⁴⁴

Further control over the release rate was gained by integrating an ionizable amino group into a particular position of the AOCOM residue, in which the amino group’s nucleophilicity can be utilized in a pH-dependent mechanism to release the parent compound from the corresponding prodrug.⁶ This amino group was also introduced to significantly improve the limited aqueous solubility of carbonate prodrugs. Protonation of the amino group at a low pH (e.g. upper GIT pH) should lead to a stable prodrug with a substantially enhanced aqueous solubility, whereas an increase in pH should gradually deprotonate the amino group. This will steadily increase the nitrogen’s nucleophilic character and accelerate the release of the parent compound via an intramolecular cyclization reaction (Fig. 3a). The introduced methylene bridge between the parent compound and the carbonate group ensures the applicability of this prodrug approach to any 4(1H)-quinolone without influencing the electrophilicity of the promoiety’s carbonate carbon, which is important for the release process. As a result of this pH-triggered release mechanism, the amino-based AOCOM prodrug approach is independent of any enzyme activity avoiding inter- and intra-species variabilities, which potentially complicate clinical studies or future therapeutic applications. During the release of the parent 4(1H)-quinolone, formaldehyde 14 as well as cyclic carbamates 12 and 13, whose ring size depends on the length of the promoiety, are formed as side products. Importantly, neither cyclic carbamates 12 and 13 nor formaldehyde 14 should lead to toxicity at the intended dose levels (see Supplementary section ‘Discussion on Toxicity of Formaldehyde’ for further details).

Figure 3.

Release of the parent compound. a, mechanism of the pH-activated parent compound release of amino AOCOM prodrugs via an intramolecular nucleophilic attack. b and c, in vitro parent compound release profiles of 4(1H)-quinolone amino AOCOM ether prodrugs 27 (b) and 28 (c) at pH 2.0, 4.0, 7.0, SGF (simulated gastric fluid; pH ~ 1.2) and SIF (simulated intestinal fluid; pH ~ 6.5). The stability of prodrugs in aqueous media and the release of parent compound was followed for 10 h using HPLC in triplicates. The error bars represent the standard deviation of triplicate measurements.

Analogous to known AOCOM ether prodrugs, AOCOM iodides were used to prepare the desired prodrugs.³⁹ Thus, the synthesis route to the prodrugs commences with conversion of Boc-protected amino alcohols into corresponding Boc-protected aminoalkyloxycarbonyloxymethyl (amino AOCOM) iodides (Scheme 2). Reaction of 4-quinolinols 7 and 15 with the Boc-protected amino AOCOM iodides, in the presence of cesium carbonate led to Boc-protected AOCOM derivatives 20–23, which were deprotected into the water-soluble HCl salts 25–28 (Table 2). 4(1H)-Quinolone 4 was chosen as parent compound as it was shown to possess enhanced microsomal stability over 3 due to the CF₃-substituted phenyl moiety in 4-position. The assessment of the aqueous solubility at pH 7.4 was challenging due to the limited stability of the prodrug moiety at neutral pH and was not performed. Aqueous solubility at pH 2.5 was therefore determined using a precedented UV-based assay.¹¹ For amino AOCOM ether prodrugs 25 and 26, an aqueous solubility of over 100 μM at acidic pH was determined. In addition, solid material was rapidly dissolved in 0.5% aqueous HEC solution and visually inspected after 60 seconds for material dissolution. The observed solubility of > 20 mM for amino AOCOM ether prodrugs of more soluble 3 and > 5 mM for amino AOCOM ether prodrugs of less soluble 4 illustrate the significant increase in solubility and dissolution rate enabled by the use of amino AOCOM ethers. Antimalarial activity for AOCOM prodrugs was decreased (EC₅₀ > 50 nM) compared to parent compounds 3 and 4 exhibiting single digit nanomolar EC₅₀ values (Table 2). Furthermore, cyclic carbamate byproducts 12 and 13 as well as formaldehyde 14 (Fig. 3a) were devoid of antimalarial activity (EC₅₀ > 4 μM), which confirms that antimalarial activity is predominantly derived from the prodrug’s parent compound.

Scheme 2.

Conversion of 4(1H)-quinolones 3 and 4 into Boc-protected AOCOM ether prodrugs 20–24 via O-alkylation and subsequent deprotection into amino AOCOM ether prodrugs 25–28, 30.

Prodrugs 27 and 28 show suitable aqueous stability-release profiles (Fig. 3b and c). Compound stability was assessed using quantitative HPLC in buffers at different pHs (2, 4, 7), in a simulated gastric fluid (SGF) and in a simulated intestinal fluid (SIF). Generally, in all tested solutions, 28 appeared to be more stable than 27. Compound 27 rapidly released parent compound 4 at pH 7.0 and pH 4.0 (> 90% parent compound released in 1 h at pH 7.0 or > 25% parent compound released in 10 h at pH 4.0). In comparison, prodrug 28 was stable at low pH values and decomposed slowly, releasing parent compound 4, at pH 7.0 (> 55% parent compound released in 5 h at pH 7.0). These results show that the parent compound release can be adjusted in terms of pH and rate. Of the two promoieties, the one containing a three-methylene spacer between the carbonate oxygen and the methylamino group possesses the most promising pH-stability profile. It should release the parent compound slowly in the intestine so that it can be absorbed continuously. In contrast, the prodrug with the two-methylene spacer could precipitate in the upper GIT because it would be released too early and too quickly.

Nevertheless, all amino AOCOM ether prodrugs of 3 and 4 (25, 27 with a two and 26, 28 with a three methylene groups-containing promoiety) were selected to undergo in vivo efficacy testing using a modified Thompson test model, which was previously reported (Table 3).^{7, 9, 36} However, parent compounds 3 and 4 and prodrugs 25–28 were administered in 0.5% aqueous HEC instead of neat PEG 400. The prodrugs were solubilized in 0.5% aqueous HEC approximately 1 minute before being dosed to the animals in order to avoid premature decomposition of the prodrug. Parent compounds displayed poor activity with 29% or lower suppression of parasitemia on day 6 post infection (PI). In contrast, at both doses (10 mg/kg and 50 mg/kg), the four amino AOCOM ether prodrugs 25–28 proved more efficacious in vivo than parent compounds 3 and 4, demonstrating the viability of amino AOCOM ether prodrugs in an in vivo setting (Table 3). Prodrugs 27 and 28, which contain a p-(trifluoromethyl) phenyl group in the parent 4(1H)-quinolone’s 3-position, are slightly more efficacious than prodrugs 25 or 26, which are substituted with a o-methyl phenyl moiety in 3-position. Compound 28, which contains the prodrug moiety that includes a three-methylene spacer between the carbonate and the methylamino group, was able to suppress parasitemia by 82% at a 10 mg/kg dose and 96% at a 50 mg/kg dose. The dose linearity for prodrug 28 was proven in another series of tests (see Supplementary section ‘Dose Linearity of Amino AOCOM Ether Prodrug’). Nevertheless, 28 does not significantly extend the life span of treated animals relative to untreated controls. These findings are not unexpected as parent 4(1H)-quinolones 3 and 4 were shown to exhibit rapid clearance in vivo.⁹ Nevertheless, these in vivo efficacy studies demonstrated that the longer chain prodrugs 26 and 28 with their increased pH-stability are easier in terms of handling and dosing. Furthermore, importantly for this study, treatment with any of the parent compounds and prodrugs administered (at the dose up to 50 mg/kg PO) did not provoke significant weight loss in the treatment group. Collectively, this data supports our thesis that prodrugs of the antimalarial 4(1H)-quinolones are well tolerated with no obvious gross toxicity issues.

Table 3.

In vivo efficacy of 7, 15 and their amino AOCOM ether prodrugs 25–28.^a

No			10 mg/kg (PO) a		50 mg/kg (PO) a
	R1	R2	Suppression [%] day 6 PI ± S.D. b	P value c	Suppression [%] day 6 PI ± S.D. b	P value c
7	-H	o-CH3	29 ± 2.1	n.s	28 ± 3.4	n.s
25		o-CH3	71 ± 2.5	≤ 0.01	46 ± 1.7	≤ 0.05
26		o-CH3	52 ± 2.6	≤ 0.05	72 ± 2.0	≤ 0.05
15	-H	p-CF3	11 ± 2.3	n.s.	10 ± 1.2	n.s.
27		p-CF3	75 ± 0.4	≤ 0.01	96 ± 0.3	≤ 0.01
28		p-CF3	82 ± 0.6	≤ 0.01	96 ± 0.2	≤ 0.01

^aMice were infected with 1·10⁶ P. berghei-GFP parasites and then orally treated once a day on days 3–5 PI with test compound in a 0.5% aqueous HEC solution. Experiments with untreated animals or with amodiaquine-treated animals (ADQ; 10 mg/kg PO) were controls for the in vivo efficacy study producing 0% suppression for the untreated animals and > 99% for the ADQ-treated animals. Furthermore, parent compounds 3 and 4 were administered as compound-specific controls using the same protocol.

^bOral administration as three daily doses (formulated in 0.5% aqueous HEC) on days 3–5 PI. Percent supression as compared to untreated control animals. S.D. – standard deviation.

^cIn an unpaired t test with Welch’s correction, percent suppression of prodrugs has been shown to be significant with P values of ≤ 0.05 or ≤ 0.01. n.s. – not significant. P value for the experiment with ADQ was ≤ 0.01.

Pharmacokinetic studies with the two most potent prodrugs 27 and 28 as well as their parent compound 4 were conducted to profile compound plasma exposure after single oral administration at a dose of 50 mg/kg in 0.5% aqueous HEC formulation (Fig. 4a). 4(1H)-quinolone 4 was slowly absorbed, reaching a maximum plasma concentration of 365 nM, 2 hours after oral dosing of 4. Importantly, the concentration of 4 remained above the lower limit of quantification (LLQ) of 1 nM over the course of the entire profiling experiment. The maximum plasma concentration of 4 significantly increased for prodrugs 27 and 28 to 12 μM and 21 μM, whereas AUC of 4 improved by 36-fold for 27 and 52-fold for 28 (Fig. 4c). As previously observed with orally dosed 4,⁹ parent compound 4 derived from both prodrugs 27 and 28 had a comparable apparent half-life t_1/2 of 7,7 and 6.4 hours.

Figure 4.

PK profiles of amino AOCOM prodrugs of 4(1H)-quinolones 2 and 4. a, plasma concentration of 4(1H)-quinolone 4 after single oral administration of 50 mg/kg of 4 and corresponding amino AOCOM ether prodrugs 27 and 28 in 0.5% aqueous HEC formulation. b, plasma concentration of 4(1H)-quinolone 2 after single oral administration of 2 and corresponding amino AOCOM ether prodrug 30 in 0.5% aqueous HEC formulation. c, PK parameters of parent 4, 2 and their amino AOCOM ether prodrugs 27, 28, 30 after oral administration of test compound in a 0.5% aqueous HEC solution.

Antimalarial compounds with potent activity and a long in vivo half-life have the potential to be curative single-dose agents. We hypothesized that installation of an amino AOCOM ether promoiety onto 2 would deliver a curative single-dose agent, because, in comparison to 4, 4(1H)-quinolone 2 was previously reported to display low in vivo clearance following oral administration in addition to excellent in vitro activity.^{7, 9} Of the two prodrug moieties, the one with a three methylene group spacer between the carbonate and the methylamino group was considered to be the most promising, due to the optimized pH-stability (Fig. 3c), the improved in vivo efficacy (Table 3), and the enhanced pharmacokinetic profile (Fig. 4a) when comparing to parent compound and other prodrugs. Prodrug 30 was synthesized in an analogous manner to 25–28 (Scheme 2).

Plasma exposure of prodrug 30 and parent compound 2 was determined following a single oral administration at 10 mg/kg and 3 mg/kg doses in 0.5% aqueous HEC (Fig. 4b). Solubility tests showed that prodrug 30 is soluble in 0.5% aqueous HEC at a concentration of greater than 5.5 mM. Overall, prodrug 30 performed better than parent compound 2 at both doses, increasing C_max and AUC of 2 approximately 20-fold. For example, at a 3 mg/kg dose, a C_max of 5.74 μM was determined for 30, whereas a C_max of a 0.31 μM was determined for 2. Importantly, 18-fold improvements of C_max and 21-fold improvements in AUC were achieved without the use of any advanced formulation techniques. This constitutes a major advancement when compared to the carbonate prodrug approach developed by Riscoe and co-workers in which the plasma exposure of parent 1 only enhanced by ~3-fold (PEG 400 formulation, 3.5 mg/kg PO dose of prodrug).⁴⁰ In a manner identical to previously reported pharmacokinetic studies with parent compound 2 in neat PEG 400,⁷ prodrug 30 exhibited low clearance with a long half-life of 10 hours or longer. This in conjunction with the increased C_max and AUC values implied prodrug 30 to be sufficiently bioavailable to produce single-dose cures at a dose as low as 3 mg/kg.

P. berghei-infected mice were treated orally with a single dose of parent compound 2 or prodrug 30 on day 3 PI. Prodrug 30 was administered in 0.5% aqueous HEC formulation at doses ranging from 0.01 mg/kg to 10.0 mg/kg whereas, for comparative reasons, parent 4(1H)-quinolone 2 was dosed at 10 mg/kg. Remarkably, prodrug 30 was curative at both 3 mg/kg and 10 mg/kg (at both does, 5 out of 5 mice cured), whereas parent compound 2 was completely inactive when administered as a single dose of 10 mg/kg using the same vehicle (Table 4). To the best of our knowledge, a single oral dose of 3 mg/kg is the lowest dose among all antimalarials that are currently in clinical trials (the closest candidate trioxolane OZ439, which is administered in combination with ferroquine, cures P. berghei-infected mice with a single oral dose of 20 mg/kg) producing curative activity. Furthermore, prodrug 30 at a 1 mg/kg oral dose showed 97% suppression of parasitemia on day 6 PI extending the average day of death for all 5 mice beyond day 13 (the average suppression of parasitemia on day 13 was 90%).

Table 4.

In vivo efficacy of 29 and its amino AOCOM ether prodrug 30.^a

No		Dose (PO) [mg/kg] b	Suppression [%] day 6 PI ± S.D. a	P value c
29	-H	10	< 1	n.s.
30		10	> 99 (number of cures: 5/5 mice)	≤ 0.01
		3	> 99 (number of cures: 5/5 mice)	≤ 0.01
		1	97 ± 0.1	≤ 0.01
		0.3	57 ± 2.0	n.s.
		0.1	< 1	n.s.
		0.03	< 1	n.s.
		0.01	< 1	n.s.

^aMice were infected with 1·10⁶ P. berghei-GFP parasites and then orally treated with one single dose on day 3 PI with test compound in a 0.5% aqueous HEC solution. Parent compound 2 was administered as control using the same protocol (one single dose, 10 mg/kg PO). Furthermore, experiments with untreated animals or with amodiaquine-treated animals (ADQ; three daily 10 mg/kg PO doses (formulated in 0.5% aqueous HEC) on days 3-5 PI) were controls for the in vivo efficacy study producing 0% suppression for the untreated animals and > 99% for the ADQ-treated animals. Percent suppression as compared to untreated control animals; S.D. – standard deviation.

^bOral administration as single dose (formulated in 0.5% aqueous HEC) on day 3 PI.

^cIn an unpaired t test with Welch’s correction, percent suppression of prodrugs has been shown to be significant with P values of ≤ 0.05 or ≤ 0.01. n.s. – not significant. P value for the experiment with ADQ was ≤ 0.01.

To demonstrate the versatility of the developed prodrug approach, the amino AOCOM prodrug moiety was successfully installed in ICI56,780, another poorly soluble 4(1H)-quinolone with antimalarial activity (see Supplementary section ‘Amino AOCOM Prodrug of ICI56,780’). As expected, the prodrug of ICI56,780 noticeably improved exposure and in vivo efficacy in comparison to parent compound ICI56,780.

CONCLUSIONS

In summary, the development and utility of a generally applicable amino AOCOM prodrug approach was demonstrated. The prodrug moiety design comprises an amino group, which in a pH-dependent manner not only improves aqueous solubility but also initiates the prodrug’s release mechanism rendering the prodrug activation to be completely independent of any enzymatic activity. The synthesis of the amino AOCOM prodrug moiety is straightforward as it can generally be attached to any parent compound containing an appropriate heteroatom. In a proof-of-concept study, the amino AOCOM prodrug moiety was installed in analogues 25–28 and 30 of antimalarial 3-aryl-4(1H)-quinolone series, whose clinical development was halted due to poor oral bioavailability. Significant improvements of exposure and in vivo efficacy was observed for all amino AOCOM 4(1H)-quinolones prodrugs with 30 producing single dose cures at a low oral dose of 3 mg/kg. This in combination with the previously reported potent in vivo efficacy against the liver stages (with single oral dose of ≤ 0.1 mg/kg) and the stages that are crucial to disease transmission (with single oral dose of 0.1 mg/kg), restore the 3-aryl-4(1H)-quinolones as an attractive class of antimalarials with potential for clinical development. Studies applying the prodrug approach to other compound classes are underway.

EXPERIMENTAL SECTION

General Material and Methods

All reagents and solvents were purchased from commercial sources and used without further purification unless noted otherwise. All reactions were carried out under an argon atmosphere using flame-dried glassware and standard Schlenk techniques unless indicated otherwise. Prior to use of solvents in reactions, they were purified by passing the degassed solvents through a column of activated alumina and transferred by an oven-dried syringe or cannula. The identity of all title compounds was verified via ¹H NMR, ¹³C NMR, and HRMS. The chemical purity of the title compounds was determined by LC/MS using the following instrumentation and the following analytical conditions: an Agilent 1100 series LC/MSD equipped with a Phenomenex Kinetex reversed phase column (50 mm × 4.6 mm, 2.6 μm, C18, 100Å); method: 10% (v/v) of acetonitrile (+0.1% formic acid) in 90% (v/v) of H₂O (0.1% formic acid), ramped to 100% acetonitrile (0.1% formic acid) over 5.5 min, and holding at 100% acetonitrile (0.1% formic acid) for 1 min with a flow rate of 1.3 mL/min; UV detector, 254 nm. The purity of each compound was ≥95% in this analysis. ¹H NMR spectra were recorded at ambient temperature on a 400 or 500 MHz Varian NMR spectrometer in the solvent indicated with the signal of the residual solvent (CDCl₃ δ 7.26 ppm; (CH₃)₂SO δ 2.50 ppm)⁴⁵ as internal standard. ¹³C NMR spectra were recorded with ¹H decoupled observation at ambient temperature on a Varian NMR spectrometer operating at 100 or 125 MHz in the solvent indicated with the signal of the residual solvent (CDCl₃ δ 77.1 ppm; (CH₃)₂SO δ 39.5 ppm)⁴⁵ as internal standard. ¹⁹F NMR spectra were recorded on a Varian NMR spectrometer operating at 376 MHz with 3-nitrofluorobenzene (−112.0 ppm) as added standard. Data for ¹H NMR are reported as follows: chemical shift (ppm), multiplicity (s = singlet, br = broad, d = doublet, t = triplet, q = quartet, sept = septet, m = multiplet), coupling constant (Hz), and integration. For the ¹³C and ¹⁹F NMR data chemical shifts (δ ppm) and multiplicities (if not a singlet) are given. NMR data was processed using MestReNova Software ver. 8.1. ¹³C signals arising from carbons next to the Boc-protected or protonated nitrogen are often broad or even split into two signals. In those cases, the central of the signal is given. The ¹H and ¹³C signals of compound 24 and 30 were assigned by using ¹H/¹H COSY, ¹H/¹³C HSQC and ¹H/¹³C HMBC. The O-alkylation instead of a possible N-alkylation was verified by that. By comparison of the spectra this verification is transferrable to all other prodrug compounds. High resolution mass spectra (HRMS) were performed on an Agilent LC/MS Q-TOF system 6540 UHD. Compounds were eluted using a gradient elution of 70/30 to 50/50 A/B over 30 minutes at a flow rate of 5.0 mL/min, where solvent A was 0.1% TFA in water and solvent B was 0.1% TFA in acetonitrile. Analytical thin layer chromatography (TLC) was performed on silica gel 60 F254 pre-coated plates (0.25 mm) from EMD Chemical Inc. and components were visualized by ultraviolet light (254 nm). Silica gel 60 from EMD Chemical Inc. with 230–400 (particle size 40–63 μm) mesh was used for all flash column chromatography. Purified compounds were further dried under high vacuum to remove residual solvent. Yields given in the Supplementary information refer to purified compounds. All 4(1H)-quinolones (6-chloro-3-(2-fluoro-4-(4-(trifluoromethoxy)phenoxy)phenyl)-7-methoxy-2-methyl-4(1H)-quinolone (2), 6-chloro-7-methoxy-2-methyl-3-(o-tolyl)-4(1H)-quinolone (3), and 6-chloro-7-methoxy-2-methyl-3-(4-(trifluoromethyl)phenyl)-4(1H)-quinolone (4) were prepared as previously described via a Conrad-Limpach cyclization of 2-aryl substituted ethyl acetoacetate with 3-chloro-4-methoxy aniline.^{8, 9} The Boc-protected aminoalcohols 2-((tert-butoxycarbonyl)(methyl)amino)ethanol⁴⁶ and 3-((tert-butoxycarbonyl)(methyl)amino)propanol⁴⁷ were prepared as described in the literature.

Synthetic Chemistry

General Procedure (GP A) for the Formation of Chloromethyl Carbonates.

A round-bottom flask was charged with a solution of the respective alcohol in DCM (2.5 mL/mmol alcohol). The stirred solution was cooled to 0 °C and pyridine was added. Then, chloromethyl chloroformate, dissolved in DCM (1.0 mL per mmol nucleophile), was added dropwise. The mixture was allowed to stir at room temperature for 10 h. The reaction was quenched with 3M HCl_(aq) (3.0 mL/mmol nucleophile) and extracted with DCM (3 × 2.0 mL/mmol nucleophile). The combined organic layers were washed with saturated NaHCO_3(aq) (4.0 mL/mmol nucleophile) and brine (4.0 mL/mmol nucleophile), dried over Na₂SO₄ and concentrated under reduced pressure. The pure product was obtained without any further purification.

Chloromethyl 2-((tert-butoxycarbonyl)(methyl)amino)ethyl carbonate (16).

According to GP A, chloromethyl carbonate was prepared reacting 2-((tert-butoxycarbonyl)(methyl)amino)ethanol (3.15 g, 18.0 mmol), pyridine (1.58 g, 20.0 mmol), and chloromethyl chloroformate (2.32 g, 18.0 mmol). Compound 16 was obtained as a colorless solid (4.22 g, 88%). ¹H NMR (400 MHz, CDCl₃) δ 5.69 (s, 2H), 4.30–4.26 (m, 2H), 3.48–3.46 (m, 2H), 2.87 (s, 3H), 1.41 (s, 9H). ¹³C NMR (101 MHz, CDCl₃) δ 155.4, 153.4, 80.0, 72.4, 67.0, 47.8, 35.3, 28.5 (three carbons).

Chloromethyl 3-((tert-butoxycarbonyl)(methyl)amino)propyl carbonate (17).

According to GP A, chloromethyl carbonate was prepared reacting 3-((tert-butoxycarbonyl)(methyl)amino)propanol (8.14 g, 43.0 mmol), pyridine (6.80 g, 86.0 mmol), and chloromethyl chloroformate (5.54 g, 43.0 mmol). Compound 17 was obtained as a colorless oil (10.3 g, 85%). ¹H NMR (400 MHz, CDCl₃) δ 5.67 (s, 2H), 4.18 (t, J = 6.4 Hz, 2H), 3.26 (t, J = 6.8 Hz, 2H), 2.79 (s, 3H), 1.86 (tt, J = 6.8 Hz, J = 6.4 Hz, 2H), 1.39 (s, 9H). ¹³C NMR (101 MHz, CDCl₃) δ 155.5, 153.2, 79.5, 72.1, 66.7, 45.3, 34.3, 28.3 (three carbons), 27.0.

General Procedure (GP B) for the Formation of Iodomethyl Carbonates.

A round-bottom flask was charged with a solution of the respective chloromethyl carbonate in acetone (0.4 mL/mmol chloromethyl carbonate). NaI was added in portions. The reaction mixture was heated to 45 °C and it was allowed to stir for 12 h. Subsequently, the mixture was filtered and concentrated under reduced pressure. Information regarding the purification are given for each product.

Iodomethyl 2-((tert-butoxycarbonyl)(methyl)amino)ethyl carbonate (18).

According to GP B, iodomethyl carbonate was prepared reacting chloromethyl 2-((tert-butoxycarbonyl)(methyl)amino)ethyl carbonate (2.68 g, 10.0 mmol) and NaI (4.50 g, 30.0 mmol). The crude product was dissolved in CH₃Cl (50 mL) and washed with saturated Na₂CO_3(aq) (30.0 mL), concd NaHCO_3(aq) (3 × 20.0 mL), and water (30 mL). The organic layer was dried over Na₂SO₄ and concentrated under reduced pressure. Compound 18 was obtained as a colorless oil (3.38 g, 94%) without any further purifications.¹H NMR (400 MHz, CDCl₃) δ 5.92 (s, 2H), 4.30–4.26 (m, 2H), 3.48–3.45 (m, 2H), 2.87 (s, 3H), 1.42 (s, 9H). ¹³C NMR (101 MHz, CDCl₃) δ 155.7, 153.3, 80.1, 67.2, 47.8, 35.7, 34.1, 28.6 (three carbons).

Iodomethyl 2-((tert-butoxycarbonyl)(methyl)amino)propyl carbonate (19).

According to GP B, iodomethyl carbonate was prepared reacting chloromethyl 3-((tert-butoxycarbonyl)(methyl)amino)propyl carbonate (10.0 g, 35.5 mmol) and NaI (6.92 mg, 46.1 mmol). Purification by flash silica gel chromatography (DCM–n-hexane–Et₂O 5:4:1, R_f = 0.58) afforded compound 19 as a yellowish oil (11.9 g, 90%). ¹H NMR (400 MHz, CDCl₃) δ 5.83 (s, 2H), 4.11 (t, J = 6.4 Hz, 2H), 3.19 (t, J = 6.8 Hz, 2H), 2.73 (s, 3H), 1.79 (tt, J = 6.8 Hz, J = 6.4 Hz, 2H), 1.32 (s, 9H). ¹³C NMR (101 MHz, CDCl₃) δ 155.4, 152.9, 79.3, 66.6, 45.3, 34.1, 28.3 (three carbons), 26.8.

General Procedure (GP C-a) for Alkylation/Acylation of 4(1H)-quinolones 5.

A round-bottom flask was charged with 4(1H)-quinolone (1.0 equiv) and Cs₂CO₃ (3.0 equiv) followed by addition of DMF (10.0–16.5 mL/mmol 4(1H)-quinolone). The suspension was cooled to 0 °C and stirred for 1 h. Subsequently, the respective alkylating/acylating agent (1.5–3.0 equiv) was added dropwise. The mixture was allowed to stir at room temperature for 18 h. The reaction was quenched with brine (5.0 mL/mmol 4(1H)-quinolone) and extracted with EtOAc (5 × 6.0 mL/mmol 4(1H)-quinolone). The combined organic layers were washed with water (3 × 5.0 mL/mmol 4(1H)-quinolone) and brine (5.0 mL/mmol 4(1H)-quinolone), dried over Na₂SO₄ and concentrated under reduced pressure. The crude product was purified by flash silica gel chromatography.

General Procedure (GP C-b) for Alkylation/Acylation of 4(1H)-quinolones 5.

A round-bottom flask was charged with a solution of the respective 4(1H)-quinolone (1.0 equiv) in DMF (5.0 mL/mmol 4(1H)-quinolone). The stirred solution was cooled to 0 °C and Cs₂CO₃ (3.0 equiv) was added in portions. The cooling bath was removed, and the reaction was allowed to stir for 2 h at room temperature. The mixture was then recooled to 0 °C and the respective alkylating agent (1.5 equiv), dissolved in DMF (1.0 mL/mmol 4(1H)-quinolone), was added dropwise. The mixture was allowed to stir at room temperature for 10 h. The reaction was quenched with brine (5.0 mL/mmol 4(1H)-quinolone) and extracted with EtOAc (3 × 6.0 mL/mmol 4(1H)-quinolone). The combined organic layers were washed with water (5.0 mL/mmol 4(1H)-quinolone) and brine (5.0 mL/mmol 4(1H)-quinolone), dried over Na₂SO₄ and concentrated under reduced pressure. The crude product was purified by flash silica gel chromatography.

6-Chloro-7-methoxy-2-methyl-3-(o-tolyl)-4-((diethylcarbamoyl)oxy)quinoline (8).

According to GP C-a, 4(1H)-quinolone 3 (200 mg, 0.64 mmol), Cs₂CO₃ (623 mg, 1.90 mmol), and diethyl carbamoylchloride (120 μL, 0.96 mmol, 1.5 equiv) were reacted in 6.4 mL DMF. Purification by flash silica gel chromatography (n-hexane–EtOAc 2:1, R_f = 0.30) afforded compound 8 as a colorless solid (171 mg, 65%). ¹H NMR (400 MHz, CDCl₃) δ 7.83 (s, 1H), 7.48 (s, 1H), 7.29–7.27 (m, 2H), 7.24–7.21 (m, 1H), 7.17 (d, J = 7.3 Hz, 1H), 4.05 (s, 3H), 3.17–2.99 (m, 4H), 2.39 (s, 3H), 2.06 (s, 3H), 0.94 (t, J = 7.1 Hz, 3H), 0.89 (t, J = 7.1 Hz, 3H). ¹³C NMR (126 MHz, CDCl₃) δ 160.3, 156.5, 152.7, 152.0, 149.0, 137.1, 134.5, 130.3, 130.2, 128.5, 126.4, 126.2, 124.5, 123.0, 118.0, 108.4, 56.8, 42.5, 42.1, 24.3, 20.0, 13.9, 13.3. HRMS (ESI-TOF) m/z: [M + H]⁺ calcd for C₂₃H₂₅ClN₂O₃ 413.1627; found 413.1616.

6-Chloro-7-methoxy-2-methyl-3-(o-tolyl)-4-((pyrrolidine-1-carbonyl)oxy)quinoline (9).

According to GP C-a, 4(1H)-quinolone 3 (175 mg, 0.56 mmol), Cs₂CO₃ (545 mg, 1.68 mmol), and 1-pyrrolidinecarbonyl chloride (185 μL, 1.68 mmol, 3.0 equiv) were reacted in 5.6 mL DMF. Purification by flash silica gel chromatography (n-hexane–EtOAc 2:1, R_f = 0.25) afforded compound 9 as a colorless solid (144 mg, 63%). ¹H NMR (500 MHz, CDCl₃) δ 7.93 (s, 1H), 7.47 (s, 1H), 7.29–7.28 (m, 2H), 7.25–7.21 (m, 3.0 Hz, 1H), 7.17 (d, J = 7.3 Hz, 1H), 4.04 (s, 3H), 3.20–3.13 (m, 3H), 2.92–2.87 (m, 1H), 2.40 (s, 3H), 2.05 (s, 3H), 1.74–1.64 (m, 4H). ¹³C NMR (126 MHz, CDCl₃) δ 160.2, 156.6, 152.2, 151.8, 149.0, 137.3, 134.5, 130.3, 130.2, 128.5, 126.2, 126.0, 124.5, 123.2, 118.0, 108.3, 56.8, 46.6, 46.4, 25.8, 25.1, 24.4, 20.0. HRMS (ESI-TOF) m/z: [M + H]⁺ calcd for C₂₃H₂₃ClN₂O₃ 411.147; found 411.146.

6-Chloro-7-methoxy-2-methyl-3-(o-tolyl)-4-(propionyloxy)quinoline (10).

According to GP C-a, 4(1H)-quinolone 3 (175 mg, 0.56 mmol), Cs₂CO₃ (545 mg, 1.68 mmol), and propionyl chloride (73.0 μL, 0.84 mmol, 1.5 equiv) were reacted in 8.0 mL DMF. Purification by flash silica gel chromatography (n-hexane–EtOAc 2:1, R_f = 0.29) afforded compound 10 as a yellowish solid (148 mg, 72%). ¹H NMR (400 MHz, CDCl₃) δ 7.74 (s, 1H), 7.49 (s, 1H), 7.31–7.29 (m, 2H), 7.25–7.22 (m, 1H), 7.09 (d, J = 7.3 Hz, 1H), 4.05 (s, 3H), 2.41 (s, 3H), 2.34–2.18 (m, 2H), 2.04 (s, 3H), 0.88 (t, J = 7.6 Hz, 3H). ¹³C NMR (126 MHz, CDCl₃) δ 171.8, 160.3, 156.6, 151.3, 148.9, 137.0, 134.1, 130.4, 130.1, 128.6, 126.2, 126.0, 124.8, 122.5, 116.9, 108.5, 56.7, 27.6, 24.3, 19.8, 9.1. HRMS (ESI-TOF) m/z: [M + H]⁺ calcd for C₂₁H₂₀ClNO₃ 370.1205; found 370.1201.

6-Chloro-7-methoxy-2-methyl-3-(o-tolyl)-4-((ethoxycarbonyl)oxy)quinoline (11).

According to GP C-a, 4(1H)-quinolone 3 (175 mg, 0.56 mmol), Cs₂CO₃ (545 mg, 1.68 mmol), and ethyl chloroformate (106 μL, 1.12 mmol, 2.0 equiv) were reacted in 8.0 mL DMF. Purification by flash silica gel chromatography (n-hexane–EtOAc 2:1, R_f = 0.42) afforded compound 11 as a colorless solid (164 mg, 76%). ¹H NMR (400 MHz, CDCl₃) δ 7.87 (s, 1H), 7.48 (s, 1H), 7.31–7.28 (m, 2H), 7.25–7.22 (m, 1H), 7.13 (d, J = 7.4 Hz, 1H), 4.06 (q, J = 7.1 Hz, 2H), 4.03 (s, 3H), 2.40 (s, 3H), 2.05 (s, 3H), 1.10 (t, J = 7.1 Hz, 3H). ¹³C NMR (101 MHz, CDCl₃) δ 160.7, 156.8, 152.2, 150.8, 149.1, 137.2, 133.5, 130.6, 130.2, 128.8, 126.2, 125.9, 125.2, 122.4, 116.6, 108.6, 65.7, 56.8, 24.4, 19.9, 14.2. HRMS (ESI-TOF) m/z: [M + H]⁺ calcd for C₂₁H₂₀ClNO₄ 386.1154; found 386.1136.

6-Chloro-7-methoxy-2-methyl-3-(o-tolyl)-4-(((2-((tert-butoxycarbonyl)(methyl)amino)ethoxycarbonyl)oxy)methoxy)quinoline (20).

According to GP C-a, 4(1H)-quinolone 3 (300 mg, 0.96 mmol), Cs₂CO₃ (935 mg, 2.87 mmol), and iodomethyl 2-((tert-butoxycarbonyl)(methyl)amino)ethyl carbonate (687 mg, 1.91 mmol, 2.0 equiv) were reacted in 16.0 mL DMF. Purification by flash silica gel chromatography (n-hexane–EtOAc 1:1, R_f = 0.48) afforded compound 20 as a colorless solid (354 mg, 68%). ¹H NMR (500 MHz, CDCl₃) δ 8.04 (s, 1H), 7.44 (s, 1H), 7.35–7.33 (m, 2H), 7.31–7.28 (m, 1H), 7.23 (d, J = 7.4 Hz, 1H), 5.25 (d, J = 6.1 Hz, 1H), 5.22 (d, J = 6.1 Hz, 1H), 4.18–4.13 (m, 2H), 4.03 (s, 3H), 3.41–3.38 (m, 2H), 2.82 (s, 3H), 2.37 (s, 3H), 2.08 (s, 3H), 1.40 (s, 9H). ¹³C NMR (126 MHz, CDCl₃) δ 160.9, 156.7, 156.3, 156.0, 154.3, 149.0, 137.3, 134.6, 130.9, 130.7, 128.8, 126.5, 124.4, 124.1, 123.4, 118.0, 108.3, 91.1, 80.2, 66.7, 56.8, 47.9, 35.6, 28.7 (three carbons), 24.7, 20.1. HRMS (ESI-TOF) m/z: [M + H]⁺ calcd for C₂₈H₃₃ClN₂O₇ 545.2049; found 545.2074.

6-Chloro-7-methoxy-2-methyl-3-(o-tolyl)-4-(((3-((tert-butoxycarbonyl)(methyl)amino)propyloxycarbonyl)oxy)methoxy)quinoline (21).

According to GP C-a, 4(1H)-quinolone 3 (500 mg, 1.59 mmol), Cs₂CO₃ (1.56 g, 4.78 mmol), and iodomethyl 2-((tert-butoxycarbonyl)(methyl)amino)propyl carbonate (1.19 g, 3.19 mmol, 2 equiv) in 0.06 molar DMF were reacted. Purification by flash silica gel chromatography (n-hexane–EtOAc 2:1, R_f = 0.22) afforded compound 21 as a colorless solid (678 mg, 76%). ¹H NMR (400 MHz, CDCl₃) δ 8.05 (s, 1H), 7.44 (s, 1H), 7.36–7.34 (m, 2H), 7.32–7.28 (m, 1H), 7.23 (d, J = 7.3 Hz, 1H), 5.24 (d, J = 6.1 Hz, 1H), 5.22 (d, J = 6.1 Hz, 1H), 4.08 (t, J = 6.5 Hz, 2H), 4.04 (s, 3H), 3.21 (t, J = 6.1 Hz, 2H), 2.79 (s, 3H), 2.38 (s, 3H), 2.09 (s, 3H), 1.84–1.78 (m, 2H), 1.41 (s, 9H). ¹³C NMR (126 MHz, CDCl₃) δ 160.9, 156.7, 156.3, 156.0, 154.5, 149.0, 137.3, 134.6, 131.0, 130.7, 128.9, 126.6, 124.4, 124.1, 123.4, 118.1, 108.1, 90.8, 79.6, 66.5, 56.6, 45.6, 34.4, 28.5 (three carbons), 27.2, 24.5, 19.9. HRMS (ESI-TOF) m/z: [M + H]⁺ calcd for C₂₉H₃₅ClN₂O₇ 559.2206; found 559.2197.

tert-butyl (3-(((((6-chloro-7-methoxy-2-methyl-3-(4-(trifluoromethyl)phenyl)quinolin-4-yl)oxy)methoxy)carbonyl)oxy)propyl)(methyl)carbamate (22).

According to GP C-a, 4(1H)-quinolone 4 (300 mg, 0.82 mmol), Cs₂CO₃ (797 mg, 2.45 mmol), and iodomethyl 2-((tert-butoxycarbonyl)(methyl)amino)ethyl carbonate (586 mg, 1.63 mmol, 2 equiv) were reacted in 13.7 mL DMF. Purification by flash silica gel chromatography (n-hexane–EtOAc 1:1, R_f = 0.32) afforded compound 22 as a colorless solid (340 mg, 70%). ¹H NMR (500 MHz, CDCl₃) δ 8.01 (s, 1H), 7.75 (d, J = 8.1 Hz, 2H), 7.52 (d, J = 8.1 Hz, 2H), 7.43 (s, 1H), 5.27 (s, 2H), 4.12–4.07 (m, 2H), 4.03 (s, 3H), 3.40–3.36 (m, 2H), 2.80 (br s, 3H), 2.46 (s, 3H), 1.40 (s, 9H). ¹³C NMR (126 MHz, CDCl₃) δ 159.7, 157.0, 156.6, 155.5, 154.2, 149.3, 139.1, 131.1 (two carbons), 130.5 (q, J = 32.7 Hz), 126.0 (q, J = 3.6 Hz; two carbons), 124.9, 124.3 (q, J = 272.0 Hz), 124.2, 123.2, 117.6, 108.4, 91.5, 80.2, 66.8, 56.8, 47.9, 35.5, 28.6 (three carbons), 25.2. ¹⁹F NMR (376 MHz, CDCl₃) δ −62.6 (three fluorines). HRMS (ESI-TOF) m/z: [M+H]⁺ calcd for C₂₈H₃₀ClF₃N₂O₇ 599.1766; found 599.1784.

6-Chloro-7-methoxy-2-methyl-3-(4-(trifluoromethyl)phenyl)-4-(((3-((tert-butoxycarbonyl)(methyl)amino)propyloxycarbonyl)oxy)methoxy)quinoline (23).

According to GP C-a, 4(1H)-quinolone 4 (300 mg, 0.82 mmol), Cs₂CO₃ (797 mg, 2.45 mmol), and iodomethyl 2-((tert-butoxycarbonyl)(methyl)amino)propyl carbonate (609 mg, 1.63 mmol, 2 equiv) were reacted in 13.7 mL DMF. Purification by flash silica gel chromatography (n-hexane–EtOAc 1:1, R_f = 0.39) afforded compound 23 as a colorless solid (339 mg, 68%). ¹H NMR (500 MHz, CDCl₃) δ 8.01 (s, 1H), 7.75 (d, J = 8.1 Hz, 2H), 7.52 (d, J = 8.1 Hz, 2H), 7.42 (s, 1H), 5.26 (s, 2H), 4.03 (s, 3H), 4.02 (t, J = 6.5 Hz, 5H), 3.20–3.17 (m, 2H), 2.78 (s, 3H), 2.46 (s, 3H), 1.79–1.76 (m, 2H), 1.40 (s, 9H). ¹³C NMR (126 MHz, CDCl₃) δ 159.7, 157.0, 156.7, 155.9, 154.3, 149.3, 139.2, 131.1 (two carbons), 130.5 (q, J = 32.8 Hz), 126.0 (q, J = 3.6 Hz; two carbons), 124.9, 124.3 (q, J = 272.2 Hz), 124.1, 123.3, 117.7, 108.4, 91.3, 79.8, 66.2, 56.8, 45.46, 34.7, 28.7 (three carbons), 27.3, 25.2. ¹⁹F NMR (376 MHz, CDCl₃) δ −62.6 (three fluorines). HRMS (ESI-TOF) m/z: [M+H]⁺ calcd for C₂₉H₃₂ClF₃N₂O₇ 613.1923; found 613.1931.

6-Chloro-7-methoxy-2-methyl-3-(2-fluoro-4-(4-(trifluoromethoxy)phenoxy)phenyl)-4-(((3-((tert-butoxycarbonyl)(methyl)amino)propyloxycarbonyl)oxy)methoxy)quinoline (24).

According to GP C-b, 4(1H)-quinolone 2 (1.50 g, 3.04 mmol), Cs₂CO₃ (2.97 g, 9.12 mmol), and iodomethyl 2-((tert-butoxycarbonyl)(methyl)amino)propyl carbonate (1.70 g, 4.56 mmol) were reacted. Purification by twofold flash silica gel chromatography (1. Et₂O–toluene–DCM 4:4:2, R_f = 0.42; 2. Et₂O–DCM–toluene–n-hexane 5:2:2:1, R_f = 0.32) afforded compound 24 as a colorless solid (1.79 g, 80%). ¹H NMR (400 MHz, CDCl₃) δ 8.01 (s, 1H, m-CH_Ar(N)), 7.46 (s, 1H, o-CH_Ar(N)), 7.32–7.28 (m, 3H, m-CH_Ar(F), 2 × m-CH_Ar(OCF₃)), 7.18–7.16 (m, 2H, 2 × o-CH_Ar(OCF₃)), 6.93 (dd, ³J_HH = 8.5 Hz, ⁴J_HH = 1.8 Hz, 1H, p-CH_Ar(F)), 6.88 (dd, ³J_HF = 10.5 Hz, ⁴J_HH = 1.8 Hz, 1H, o-CH_Ar(F)), 5.44 (d, ²J_HH = 6.0 Hz, 1H, OCH₂O), 5.31 (d, ²J_HH = 6.0 Hz, 1H, OCH₂O), 4.12 (t, ³J_HH = 6.5 Hz, 2H, OCH₂CH₂CH₂N), 4.06 (s, 3H, OCH₃), 3.26–3.23 (m, 2H, OCH₂CH₂CH₂N), 2.82 (s, 3H, NCH₃), 2.52 (s, 3H, C_q,ArCH₃), 1.86–1.83 (m, 2H, OCH₂CH₂CH₂N), 1.43 (s, 9H, C(CH₃)₃). ¹³C NMR (101 MHz, CDCl₃) δ 160.4 (d, ¹J_CF = 248.3 Hz, C_q,ArF), 160.3 (C_ArCH₃), 158.8 (d, ³J_CF = 10.2 Hz, m-C_q,Ar(F)), 157.5 (C_q,ArOCH₂), 156.6 (C_q,ArOCH₃), 155.5 (NC(O)O), 154.2 (p-C_q,Ar(OCF₃)), 153.9 (OC(O)O), 149.0 (p-C_q,Ar(Cl)), 145.3 (q, ³J_CF = 1.6 Hz, C_q,ArOCF₃), 132.9 (d, ³J_CF = 5.1 Hz, m-CH_Ar(F)), 124.2 (C_q,ArCl), 122.9 (m-CH_Ar(OCH₃)), 122.8 (2 × m-CH_Ar(OCF₃); two carbons), 120.8 (2 × o-CH_Ar(OCF₃); two carbons), 120.4 (q, ¹J_CF = 259.2 Hz, CF₃), 118.4 (o-C_q,Ar(CH₃)), 117.3 (m-C_q,Ar(Cl)), 117.1 (d, ²J_CF = 17.2 Hz, o-C_q,Ar(F)), 114.1 (d, ⁴J_CF = 3.0 Hz, p-CH_Ar(F)), 108.0 (o-CH_Ar(OCH₃), 106.2 (d, ²J_CF = 26 Hz, o-CH_Ar(F)), 91.1 (OCH₂O), 79.3 (C(CH₃)₃)), 66.3 (OCH₂CH₂CH₂N), 56.3 (OCH₃), 45.3 (OCH₂CH₂CH₂N), 34.2 (NCH₃), 28.2 (C(CH₃)₃), 26.8 (OCH₂CH₂CH₂N), 24.1 (C_q,ArCH₃). ¹⁹F NMR (376 MHz, CDCl₃) δ −59.2 (OCF_3; three fluorines), −112.1 (dd, ³J_HF = 10.4 Hz, ⁴J_HF = 8.2 Hz, C_q,ArF).

General Procedure (GP D-a) for Deprotection of Boc-Protected Prodrugs.

A round-bottom flask was charged with a solution of the respective Boc-protected 4-alkoxyquinoline in Et₂O (10.0 mL/mmol 4-alkoxyquinoline). The solution was cooled to 0 °C and in situ generated HCl_(g) was bubbled through for 10 min. After stirring for 10 min at 0 °C the cooling bath was removed, and the reaction was allowed to stir for additional 10 min at room temperature upon precipitation of the HCl salt. The mixture was then concentrated under reduced pressure. The pure product was obtained without any further purification.

General Procedure (GP D-b) for Deprotection of Boc-Protected Prodrugs.

A round-bottom flask was charged with the respective Boc-protected 4-alkoxyquinoline. At 0 °C HCl (2 M in Et₂O; 25.0 mL/mmol 4-alkoxyquinoline) was added. After stirring for 30 min at 0 °C the cooling bath was removed, and the reaction was allowed to stir for additional 10 h at room temperature upon precipitation of the HCl salt. The mixture was then concentrated under reduced pressure. The pure product was obtained without any further purification.

6-Chloro-7-methoxy-2-methyl-3-(o-tolyl)-4-(((2-(methyl ammonio)ethoxycarbonyl)oxy)methoxy)quinoline, chloride salt (25).

According to GP D-a, Boc-protected 4-alkoxyquinoline 20 (25 mg, 0.05 mmol) was deprotected and the HCl salt 25 was obtained as a colorless solid (21 mg, 95%). ¹H NMR (500 MHz, (CD₃)₂SO) δ 9.20 (br s, 2H), 8.28 (s, 1H), 8.03 (s, 1H), 7.47–7.43 (m, 2H), 7.40–7.37 (m, 1H), 7.35–7.34 (m, 1H), 5.44 (d, J = 6.1 Hz, 1H), 5.40 (d, J = 6.1 Hz, 1H), 4.28 (t, J = 5.2 Hz, 2H), 4.08 (s, 3H), 3.16–3.12 (m, 2H), 2.51–2.49 (m, 6H), 2.06 (s, 3H). ¹³C NMR (126 MHz, (CD₃)₂SO) δ 160.1, 159.0, 157.9, 152.9, 141.9, 136.9, 131.2, 130.6, 130.5, 129.4, 126.4, 125.4, 124.2, 123.7, 116.9, 102.3, 91.1, 63.6, 57.2, 46.4, 32.5, 20.7, 19.2. HRMS (ESI-TOF) m/z: [M]⁺ calcd for C₂₃H₂₆ClN₂O₅⁺ 445.1525; found 445.151.

6-Chloro-7-methoxy-2-methyl-3-(o-tolyl)-4-(((3-(methyl ammonio)propyloxycarbonyl)oxy)methoxy)quinoline, chloride salt (26).

According to GP D-a, Boc-protected 4-alkoxyquinoline 21 (200 mg, 0.36 mmol) was deprotected and the HCl salt 26 was obtained as a colorless solid (164 mg, 93%). ¹H NMR (500 MHz, (CD₃)₂SO) δ 8.85 (br s, 2H), 8.19 (s, 1H), 7.86 (s, 1H), 7.46–7.42 (m, 2H), 7.39–7.36 (m, 1H), 7.31 (d, J = 7.4 Hz, 1H), 5.40 (d, J = 6.2 Hz, 1H), 5.38 (d, J = 6.2 Hz, 1H), 4.08 (s, 3H), 4.06 (t, J = 6.5 Hz, 2H), 2.89–2.83 (m, 2H), 2.51 (t, J = 3.4 Hz, 3H), 2.44 (s, 3H), 2.05 (s, 3H), 1.93–1.88 (m, 2H). ¹³C NMR (101 MHz, (CD₃)₂SO) δ 159.4, 158.5, 153.2, 143.6, 136.6, 130.5, 130.3, 129.0, 126.3, 124.2, 123.2, 116.9, 90.9, 65.4, 57.1, 44.9, 32.3, 24.6, 19.3. HRMS (ESI-TOF) m/z: [M]⁺ calcd for C₂₄H₂₈ClN₂O₅⁺ 459.1681; found 459.1689. Note: Signals corresponding to the carbons C_q,ArCH_3, o-CH_Ar(OCH₃), C_q,ArOCH_2, o-C_q,Ar(CH₃) were too broad to be observable.

6-Chloro-7-methoxy-2-methyl-3-(4-(trifluoromethyl)phenyl)-4-(((2-(methyl ammonio)ethoxycarbonyl)oxy)methoxy)quinoline, chloride salt (27):

According to GP D-a, Boc-protected 4-alkoxyquinoline 22 (60 mg, 0.10 mmol) was deprotected and the HCl salt 27 was obtained as a colorless solid (50 mg, 93%). ¹H NMR (500 MHz, (CD₃)₂SO) δ 9.22–9.19 (m, 2H), 8.30 (s, 1H), 8.00 (s, 1H), 7.96 (d, J = 8.2 Hz, 2H), 7.71 (d, J = 8.2 Hz, 2H), 5.53 (s, 2H), 4.27–4.25 (m, 2H), 4.09 (s, 3H), 3.13–3.13 (m, 2H), 2.59 (s, 3H), 2.52–2.50 (m, 3H). ¹³C NMR (126 MHz, (CD₃)₂SO) δ 158.7, 157.8, 152.8, 136.6, 131.1 (two carbons), 129.1 (q, J = 31.6 Hz), 125.7 (q, J = 3.5 Hz; two carbons), 124.7, 124.2 (q, J = 272.2 Hz), 123.6, 116.8, 91.9, 63.6, 57.2, 46.6, 32.5, 21.5. ¹⁹F NMR (376 MHz, (CD₃)₂SO) δ −61.0 (three fluorines). HRMS (ESI-TOF) m/z: [M]⁺ calcd for C₂₃H₂₃ClF₃N₂O₅⁺ 499.1242; found 499.126. Note: Signals corresponding to the carbons C_q,ArCH_3, o-CH_Ar(OCH₃), p-C_q,Ar(Cl)_, o-C_q,Ar(CH₃) were too broad to be observable.

6-Chloro-7-methoxy-2-methyl-3-(4-(trifluoromethyl)phenyl)-4-(((3-(methyl ammonio)propyloxycarbonyl)oxy)methoxy)quinoline, chloride salt (28).

According to GP D-a, Boc-protected 4-alkoxyquinoline 23 (65 mg, 0.11 mmol) was deprotected and the HCl salt 28 was obtained as a colorless solid (56 mg, 95%). ¹H NMR (500 MHz, (CD₃)₂SO) δ 9.07–9.03 (m, 2H), 8.26 (s, 1H), 7.98 (s, 1H), 7.94 (d, J = 8.2 Hz, 2H), 7.71 (d, J = 8.2 Hz, 2H), 5.50 (s, 2H), 4.09 (s, 3H), 4.04 (t, J = 6.3 Hz, 2H), 2.87–2.81 (m, 2H), 2.58 (s, 3H), 2.49 (t, J = 5.5 Hz, 3H), 1.96–1.87 (m, 2H). ¹³C NMR (126 MHz, (CD₃)₂SO) δ 158.7, 157.7, 153.1, 136.6, 131.1 (two carbons), 129.0 (q, J = 31.6 Hz), 125.6 (q, J = 3.3 Hz; two carbons), 124.8, 124.1 (q, J = 272.2 Hz), 123.6, 116.9, 91.7, 65.4, 57.2, 45.8, 32.2, 24.5, 21.7. ¹⁹F NMR (376 MHz, (CD₃)₂SO) δ −61.1 (three fluorines). HRMS (ESI-TOF) m/z: [M]⁺ calcd for C₂₄H₂₅ClF₃N₂O₅⁺ 513.1399; found 513.1397. Note: Signals corresponding to the carbons C_q,ArCH_3, o-CH_Ar(OCH₃), p-C_q,Ar(Cl)_, o-C_q,Ar(CH₃) were too broad to be observable.

6-Chloro-7-methoxy-2-methyl-3-(2-fluoro-4-(4-(trifluoromethoxy)phenoxy)phenyl)-4-(((3-(methyl ammonio)propyloxycarbonyl)oxy)methoxy)quinoline, chloride salt (30).

According to GP D-a, Boc-protected 4-alkoxyquinoline 24 (300 mg, 0.41 mmol) was deprotected and the HCl salt 30 was obtained as a colorless solid (271 mg, 98%). ¹H NMR (500 MHz, (CD₃)₂SO) δ 9.30–9.24 (m, 2H, NH₂⁺), 8.31 (s, 1H, m-CH_Ar(N)), 8.07 (s, 1H, o-CH_Ar(N)), 7.57 (dd, ³J_HH = 8.5 Hz, ⁴J_HF = 8.4 Hz, 1H, m-CH_Ar(F)), 7.50–7.46 (m, 2H, 2 × m-CH_Ar(OCF₃)), 7.33–7.30 (m, 2H, 2 × o-CH_Ar(OCF₃)), 7.21 (dd, ³J_HF = 10.7 Hz, ⁴J_HH = 2.4 Hz, 1H, o-CH_Ar(F)), 7.07 (dd, ³J_HH = 8.5 Hz, ⁴J_HH = 2.3 Hz, 1H, p-CH_Ar(F)), 5.61 (d, ²J_HH = 6.5 Hz, 1H, OCH₂O), 5.58 (d, ²J_HH = 6.5 Hz, 1H, OCH₂O), 4.10 (t, ³J_HH = 6.6 Hz, 2H, OCH₂CH₂CH₂N), 4.09 (s, 3H, OCH₃), 2.85 (tt, ³J_HH = 7.3 Hz, ³J_HH = 6.1 Hz, 2H, OCH₂CH₂CH₂N), 2.47 (s, 3H, NCH₃), 2.68 (s, 3H, C_q,ArCH₃), 1.96 (tt, ³J_HH = 7.3 Hz, ³J_HH = 6.6 Hz, 2H, OCH₂CH₂CH₂N). ¹³C NMR (101 MHz, (CD₃)₂SO) δ 162.1 (C_q,ArOCH₂), 159.9 (d, ¹J_CF = 247.0 Hz, C_q,ArF), 158.9 (C_ArCH₃), 158.8 (d, ³J_CF = 11.5 Hz, m-C_q,Ar(F)), 158.4 (C_q,ArOCH₃), 154.3 (p-C_q,Ar(OCF₃)), 153.1 (OC(O)O), 144.5 (q, ³J_CF = 1.4 Hz, C_q,ArOCF₃), 141.6 (p-C_q,Ar(Cl)), 133.6 (d, ³J_CF = 4.2 Hz, m-CH_Ar(F)), 125.9 (C_q,ArCl), 123.8 (m-CH_Ar(OCH₃)), 123.2 (two carbons, 2 × m-CH_Ar(OCF₃)), 120.9 (two carbons, 2 × o-CH_Ar(OCF₃)), 120.1 (q, ¹J_CF = 256.0 Hz, CF₃), 119.5 (o-C_q,Ar(CH₃)), 116.9 (m-C_q,Ar(Cl)), 114.8 (d, ⁴J_CF = 2.4 Hz, p-CH_Ar(F)), 113.9 (d, ²J_CF = 17.7 Hz, o-C_q,Ar(F)), 102.0 (o-CH_Ar(OCH₃), 106.6 (d, ²J_CF = 25.9 Hz, o-CH_Ar(F)), 91.7 (OCH₂O), 65.7 (OCH₂CH₂CH₂N), 57.3 (OCH₃), 44.9 (OCH₂CH₂CH₂N), 32.2 (NCH₃), 24.6 (OCH₂CH₂CH₂N), 20.2 (C_q,ArCH₃). ¹⁹F NMR (376 MHz, CD₃)₂SO) δ −59.2 (OCF_3; three fluorines), −112.1 (dd, ³J_HF = 10.4 Hz, ⁴J_HF = 8.2 Hz, C_q,ArF).

3-Methyl-1,3-oxazinan-2-one (13).

3-Methyl-1,3-oxazinan-2-one (13) was synthesized from commercially available 1,3-oxazinan-2-one as previously described. ¹H NMR (500 MHz, CDCl₃) δ 4.24 (t, J = 5.0 Hz, 2H), 3.31 (t, J = 6.3 Hz, 2H), 2.98 (s, 3H), 2.07 – 2.00 (m, 2H). ESI-MS: [M+H]⁺ calcd for C₅H₁₀NO₂⁺ 116.1; found 116.1.

Antimalarial activity

In Vitro Antimalarial Activity

All synthesized prodrugs were tested, as previously reported, against the clinically relevant multi-drug-resistant P. falciparum clone W2/Indochina (chloroquine and pyrimethamine resistant) and TM90-C2B/Thailand (chloroquine, mefloquine, pyrimethamine, and atovaquone resistant).⁸ The source of parasites is from original isolations maintained in the Kyle Lab. Both W2 and TM90-C2B have been authenticated by phenotype and genotype analysis. Cultures were routinely checked for mycoplasma contamination.

P. falciparum clone W2 (Indochina) and TM90C2B (Thailand) were grown in continuous culture using RPMI 1640 media containing 10% heat-inactivated type A+ human plasma, sodium bicarbonate (2.4 g/L), HEPES (5.94 g/L) and 4% washed human type A+ erythrocytes. Cultures were gassed with a 90% N₂, 5% O₂ and 5% CO₂ mixture followed by incubation at 37 °C. Test compounds in DMSO at 10.0 mM concentration were diluted at least 1:400 and then serially diluted in duplicate or triplicate over 11 concentrations. P. falciparum cultures with >70% ring stage parasites were diluted to 0.5–0.7% parasitemia and 1.5% hematocrit in RPMI 1640 media. In 96-well plates a volume of 90 μL/well of parasitized erythrocytes was added on top of 10 μL/well of the test compound. A separate plate containing chloroquine, dihydroartemisinin and atovaquone was added to each set of assay plates as control drugs. A Beckman Coulter Biomek 3000 was used to dispense test compounds, control drugs and parasitized erythrocytes into the microtiter plates. Positive and negative controls were included in each plate. Positive controls consisted of drug-free parasitized erythrocytes and negative controls consisted of parasitized erythrocytes dosed with a high concentration of chloroquine or dihydroartemisinin that ensured 100% parasite death. Assay plates were placed into a plastic gassing chamber and equilibrated with 90% N₂, 5% O₂ and 5% CO₂ mixture then incubated at 37 °C for 48 hours; then ³H-hypoxanthine was added, and plates incubated another 24 hours. After 72 hours of incubation the assay plates were frozen at –80 °C until later processed for parasite growth determinations. Assay plates were removed from –80 °C and allowed to thaw at room temperature. Using a plate harvester, the contents of the plate were collected on filtermats and then CPMs counted in a Topcount liquid scintillation counter. Data analysis was performed using a custom database manager (Dataspects, Inc). Non-linear regression analysis was used to calculate EC₅₀. Given EC₅₀ values represent means of at least two independent experiments.

In Vivo Antimalarial Activity (Blood Stage Thompson Tests, P. berghei)

To determine in vivo efficacy, we utilized an unblinded “modified Thompson model”, which is accepted as a standard model in malaria chemotherapy research.⁴⁸ Infections were established by an intraperitoneal inoculation of 2 × 10⁶ P. berghei (GFP)-infected red blood cells. By day 3 post infection (PI), the first day of drug treatment, parasitemia typically reaches 1% (i.e. 1% of the RBC will be parasitized). Five mice (female, balb/c, 4–6 weeks of age, average weight 18 g, obtained from Harlan) per dosage group were used. Based upon our previous data with this model, a paired t-test power analysis for sample size determined five animals per group would allow us to detect p < 0.5 due to the small degree of variance in parasitemia in infected, untreated animals.

Compounds were administered orally (PO) from day 3 PI on once-daily (qd) for 3 days or on day 3 PI as a single dose with test compound in a 0.5% aqueous HEC solution. Treatment group assignments were allocated randomly. Atovaquone was included as a positive control for comparison. Infected mice, drug treated on days 3–5 or day 3 PI, were followed for 30 days total. Parasitemia was determined from blood collected from the tail vein on days 3, 6, 9, 12, 15, 18, 21, 24, 27, and 30. Parasitemia was quantified by flow cytometry and by microscopic examination of Giemsa-stained blood smears. Endpoints for the efficacy analysis were the percent suppression of parasitemia on day 6 post infection (PI), the no recrudescence dose (ie., no parasitemia following initial clearance of parasitemia), and percent cure at 30 days PI. The in vivo efficacy studies were conducted in compliance with the Guide for the Care and Use of Laboratory Animals of the National Research Council for the National Academies. The protocol (IS00748) was approved by the University of South Florida Institutional Animal Care and Use Committee.

ABBREVIATIONS

AOCOM

alkoxycarbonyloxymethyl

AUC

area under the curve

C_max

maximum serum concentration

EC₅₀

half maximal effective concentration

EE

exoerythrocytic

GIT

gastrointestinal tract

HEC

hydroxycellulose

malERA

malaria eradication research agenda

MMV

Medicines for Malaria Venture

PEG

polyethylene glycol

PI

post infection

PK

pharmacokinetics

PO

per os or oral administration

SERCaP

single exposure radical cure and prophylaxis

SGF

simulated gastric fluid

SIF

simulated intestinal fluid

TLC

thin layer chromatography

TPP

target product profile