2-hydroxy-1,4-naphthoquinones with 3-alkyldiarylether groups: synthesis and Plasmodium falciparum inhibitory activity

Abstract

Background

In 1948, the synthesis and Plasmodium lophurae activity of 2-hydroxy-1,4-naphthoquinones containing 3-alkyldiarylether side chains was reported.

Method/results

The synthesis of five related compounds, designed to be more metabolically stable, was pursued. The compounds were synthesized using a radical alkylation reaction with naphthoquinones. One compound had a lower IC₅₀ value against various strains of Plasmodium falciparum and assay data indicate that it binds to the Q_o site of cytochrome bc₁. With a low yield for the radical alkylation of the most active compound, a reductive alkylation method with used to improve reaction yields.

Conclusion

Further synthetic knowledge was obtained, and the assay data indicate that there are sensitivity differences between avian and human malarial parasites for these molecules.

Plain language summary

Malaria is a disease caused by a parasite that affects millions of people each year and results in many deaths. In 1948, 300 structurally related compounds were made and tested for antimalarial activity with the goal of finding a drug to treat the disease. From this work, promising compounds were identified and this work has served as a starting point for further investigations. Based on recent discoveries, this study made variations of promising 1948 compounds to investigate whether antimalarial activity could be improved. These compounds were made using two different methods. One derivative was found to be more potent than the original compound but was not the one expected based on the 1948 work.

Graphical abstract

In 1943, the antimalarial activity of a 3-alkyl-2-hydroxy-1,4-naphthoquinones (1), specifically 2, was discovered (Figure 1) [1]. This led Fieser and co-workers' reporting more than 300 structurally related compounds, including 3a–f (Figure 1) and the discovery of other promising compounds with antimalarial activity [1–4]. Compounds 3a–f had high in vitro activity (measured by antirespiratory activity of Plasmodium lophurae) relative to quinine and had promising in vivo antimalarial activity in an avian malaria model (P. lophurae in ducks) [3]. In both cases, 3c was the most potent. There has since been extensive work to further investigate the antimalarial activity of 3-alkyl-2-hydroxy-1,4-naphthoquinones and this area of research eventually led to the discovery, development and use of atovaquone ([4]; Figure 1) in the 1990s to treat malaria [5–8]. Resistance to atovaquone has developed and this compound has poor bioavailability, so there is still ongoing research to investigate the antimalarial activities of new quinones, including naphthoquinones [6]. Fieser's work is still used as a starting point for recent investigations into new antimalarial compounds [9,10]. In addition to antimalarial activity, 3-alkyl-2-hydroxy-1,4-naphthoquinones (1) are also known to have bactericidal [11], insecticidal [12], molluscicidal [13], herbicidal [14] and general antiprotozoal [8] properties.

Figure 1. . Structures of 3-substituted 2-hydroxy-1,4-naphthoquinone (1), hydrolapachol (2), 3a–f and atovaquone (4).

The synthesis of 1 can be challenging. Fieser and co-workers' general synthetic procedure is outlined in Figure 2A and utilizes a radical alkylation of 2-hydroxy-1,4-naphthoquinone (5) using a diacyl peroxide (6) to form 7 and other products [2]. The Hooker oxidation can be used to shorten the 3-alkyl group of 7 by a CH₂ group to form 1 (Figure 2A) [2]. The synthesis of 3c, 3e and 3f were accomplished from the necessary diacyl peroxides in 31–42%, 30% and 14% yields, respectively [2]. Two procedures for the Hooker oxidation were able to convert 3c to 3b in 21% and 91% yields depending on the procedure [2]. Next, 3b was converted to 3a in an 87.5% yield utilizing the higher-yielding Hooker oxidation procedure [2]. Similarly, 3d was synthesized via Hooker oxidation from 3e in a 56% yield [2]. Atovaquone (4) has been manufactured by a radical alkylation, and hydrolysis reaction, which is shown in Figure 2B [15]. The inefficiency of the radical alkylation results in a low overall yield of 4. Alternative syntheses of atovaquone have been developed [16] and a manufacturing route to atovaquone has been developed to avoid a radical alkylation reaction (Figure 2C) [15]. In this route, ketone 11 reacts with aldehyde 12 to form the α,β-unsaturated product 13, followed by hydrolysis and rearrangement to form the 2-hydroxy-1,4-naphthoquinone ring. Another method is to directly react 2-hydroxy-1,4-naphthoquinone (5) with an aldehyde in the presence of catalytic proline and 15 to form 16 and 17 (Figure 2D) [17].

Figure 2. . Four methods to synthesis 3-alkyl-2-hydroxy-1,4-naphthoquinones.

(A) Summary of Fieser’s synthetic route. (B) An industrial synthesis of Atovaquone. (C) A manufacturing route to Atovaquone. (D) Three-component reductive alkylation example.

Atovaquone antimalarial activity is due to binding to the Q_o site of the cytochrome bc₁ complex, which disrupts the mitochondrial electron transport chain [8]. Another binding site in the cytochrome bc₁ complex is the Q_i site and compounds that bind to this site have been discovered. Two Q_i site inhibitors are pyridone 18 (GSK932121) [18,19] and quinolone 19 (ELQ-300 [20]; Figure 3), which both contain trifluoromethoxy diarylether side chains. The development of 19 started from the reported activity of quinolones in the 1940s, and one of the main modifications was to utilize a diphenylether chain based on 3c and 18. In the development of 19, it was found that the addition of the OCF₃ group increased the potency by about twofold from its progenitor and stabilized the compound against liver microsomal enzymes [21]. Based on compounds 18 and 19, the trifluoromethoxy (OCF₃) derivatives of Fieser's compounds 3a–d needed to be synthesized and tested to determine the antimalarial activity of these compounds (3g–j; Figure 3). This paper reports the synthesis of compounds 3g–h via a radical alkylation step, the IC₅₀ values of these compounds against various drug-resistant Plasmodium falciparum strains that provide comparative inhibitory activity about the binding site and further work to improve the synthetic route via a three-component reductive alkylation procedure (Figure 2D).

Figure 3. . Structure of 18, 19 and 3g–j.

Materials & methods

General experimental information

Reagents were purchased and used without purification from commercial suppliers, unless otherwise indicated. Tetrahydrofuran and triethylamine were obtained from a Pure Solv Micro Solvent Purification System (Innovative Technology). Dichloromethane (CH₂Cl₂) was dried over heat-activated 3 Å molecular sieves [22]. Reactions were monitored using silica TLC plates. For the purification of compounds by flash chromatography, hexanes (95% n-hexane) were used. Melting points were measured with an SRS Digimelt apparatus. Infrared spectra were obtained with a Thermo Scientific Nicolet iS10 with attenuated total reflectance. NMR spectra were collected on a Varian or Bruker spectrometer (400, 100 and 376 MHz for ¹H, ¹³C and ¹⁹F, respectively). NMR chemical shifts (δ) are reported in ppm values relative to tetramethylsilane, solvent peaks or fluorobenzene (-113.15 ppm for ¹⁹F spectra) and coupling constants (J) are reported in Hz. The Mass Spectrometry Laboratory at the University of Illinois at Urbana–Champaign performed high-resolution mass spectrometry analyses. The purity of the final products was determined using HPLC and a Phenomenex Luna C8(2) reverse-phase column (5 μm, 50 × 2 mm) at 40°C and monitored at 254 nm [21]. HPLC condition A was with a mobile phase of CH₃OH/H₂O with 0.5% trifluoroacetic acid at 0.4 ml/min and the following gradient: 25–100% CH₃OH over 20 min, 100% CH₃OH for 4 min. HPLC condition B was with a mobile phase of CH₃CN/H₂O with 0.5% trifluoroacetic acid at 0.4 ml/min and the same gradient.

See the Supplementary Data for the synthesis and characterization of 4-hydroxyphenyl esters 20c and 20d; phenoxyphenyl esters 22a–d; phenoxyphenyl acids 23a–d; alcohols 28a–c; and final compounds 3b, 3c and 3h–k. Also included are ¹H and ¹³C spectra for all compounds.

Synthesis of 3g via a Hunsdiecker reaction & substitution [23]

2-Hydroxy-3-(4-[4-(trifluoromethoxy)phenoxy]benzyl)naphthalene-1,4-dione (3g)

A 350.9 mg (1.12 mmol, 1 equiv.) portion of 23a, 267.6 mg (1.12 mmol, 1 equiv.) of 2-bromo-1,4-naphthalenedione (24) and 35.3 mg (0.134 mmol, 0.12 equiv.) of silver nitrate and 2.8 ml of a 1:1 solution of acetonitrile and water (0.4 M solution of the carboxylic acid) were added to a round-bottom flask with a condenser. The reaction mixture was stirred and heated in a 65°C oil bath. A 2.4 M solution of ammonium persulfate (306.67 mg, 1.34 mmol, 1.2 equiv.) in 0.56 ml of water was added dropwise over an hour. The reaction mixture was then heated for an additional 30 min. The reaction mixture was then extracted twice with 10 ml of ethyl acetate, dried with MgSO₄ and concentrated to dryness. The resulting residue was used for the next step without purification. The residue was dissolved in 22.5 ml of methanol (20 ml/mmol) in a round-bottom flask with a condenser. Next, 9 ml of a 0.7 M solution potassium hydroxide (8 ml/mmol, 6.3 mmol, 5.6 equiv.) in water was added to the round bottom and refluxed at 90°C for 1 h. The reaction solution was allowed to cool to room temperature and acidified to a pH of 1 with a 2 M aqueous hydrochloric acid solution. The reaction solution was then extracted with chloroform (3 × 25 ml) and the combined organic layers were then dried with MgSO₄ and concentrated to dryness. The mixture was purified by silica chromatography (wet loaded and run using 1:1 hexanes:dichloromethane) to yield the desired product in 25% (122.7 mg, 0.28 mmol) as a yellow solid (mp = 124.8–125.7°C). IR: 3321, 3076, 2930, 1669, 1642, 1596, 1496. ¹H NMR (400 MHz, CDCl₃): δ 3.93 (s, 2H), 6.87–6.98 (m, 4 H), 7.13 (d, 2 H, J = 8.2), 7.37 (d, 2 H, J = 8.6), 7.44 (s, 1 H), 7.68 (dt, 1 H, J = 1.3, 7.6), 7.75 (dt, 1 H, J = 1.3,7.7), 8.08 (dd, 1 H, J = 1.2, 7.8), 8.13 (dd, 1 H, J = 1.2, 7.9). ¹³C NMR (100 MHz, CDCl₃): δ 28.4, 119.2, 119.3, 120.5 (q, J = 257), 122.5, 122.9, 126.2, 126.9, 129.4, 130.7, 132.8, 133.1, 134.5, 135.1, 144.3 (m), 153.0, 155.1, 156.1, 181.6, 184.4. ¹⁹F NMR (376 MHz, CDCl₃): -58.2. HRMS (ES-): m/z [M-H]^- calcd for C₂₄H₁₄F₃O₅^-: 439.0793; found: 439.0791. The retention time of the product for condition A was 15.7 min (>99% purity) and for condition B was 11.8 min (>99% purity).

General diacyl peroxide method for the synthesis of 3b, 3c & 3h–k [2,24]

A 0.62 M CH₂Cl₂ solution with 1 equiv. of a phenoxyphenyl acid, 1.25 equiv. of a 30% H₂O₂ solution, and a 0.062 M CH₂Cl₂ solution with 0.1 equiv. of 4-dimethylaminopyridine were added to a flame-dried round-bottom flask under argon atmosphere. The mixture was stirred in an ice bath for 10 min and then a 0.695 M CH₂Cl₂ solution with 1.12 equiv. of N,N’-dicyclohexylcarbodiimide was added. The reaction mixture was then stirred at 0°C for ∼3 h. The reaction mixture was subsequently filtered through a pad of silica gel, which was washed with cold CH₂Cl₂. The filtered solution was cooled in an ice-water bath and concentrated under reduced pressure to obtain the crude diacyl peroxide; 1 equiv. of 2-hydoxynaphthoquinone and glacial acetic acid to make a 0.33 M solution were added to a flame-dried round-bottom flask under argon atmosphere. The solution was then heated at 92°C for 10 min before a 0.33 M diethyl ether solution of crude diacyl peroxide (assumed 1 equiv.) was added dropwise over 5 min. The solution was heated for 1 h and was then concentrated to dryness. The residue was then dissolved in Et₂O and was washed with two portions of a saturated NaHCO₃ solution. Aqueous layers were combined and extracted with Et₂O. The combined organic layers were then dried using MgSO₄ concentrated to dryness, and the desired product was purified.

General procedure for alcohol oxidation & reductive alkylation reactions; Dess–Martin periodinane oxidation for 3h & 3k [25]

The alcohol (1 equiv.) and dry CH₂Cl₂ were combined to make a 0.6 M solution. The solution was cooled to 0°C and Dess–Martin periodinane (1.1 equiv.) was added. The reaction was stirred at room temperature for 1 h. The reaction was quenched with the addition of a portion, equal to the reaction volume, of a saturated sodium thiosulfate solution and then the same portion of a saturated sodium bicarbonate solution. The solution was extracted with three portions of CH₂Cl₂ and the combined organic layer was washed with H₂O and brine. The solution was dried with anhydrous MgSO₄, filtered over a short plug of silica gel using 20% hexanes and 80% CH₂Cl₂ and then concentrated and used in the next step.

Swern oxidation for 3i [26]

Freshly distilled oxalyl chloride (2 equiv.) and dichloromethane to make a 0.5 M solution was added to a flame-dried round-bottom flask. The solution was placed into an acetone/dry ice bath and stirred for 15 min. Over 5 min, a 4 M solution of DMSO (4 equiv.) in dichloromethane was added and then the solution was stirred for a further 10 min. Next, 1 equiv. of 28b was slowly added to the reaction flask in a 0.5 M dichloromethane solution. The reaction was stirred for 45 min and then triethylamine (6 equiv.) was added dropwise. The solution was then allowed to warm to 0°C, and then water (113 equiv.) was added to the reaction. The reaction was then diluted with chloroform and washed with three portions of water and three portions of brine. The solution was dried with anhydrous MgSO₄, filtered and then concentrated and used in the next step.

Reductive alkylation [27]

A 0.6 M dichloromethane solution of the crude aldehyde was added to a flame-dried condenser and round-bottom flask. Next, 1 equiv. of the 2-hydroxy-1,4-naphthoquinone or a derivative, 1 equiv. of diethyl 1,4-dihydro-2,6-dimethyl-3,5-pyrinedicarboxylate and 0.05 equiv. of L-proline were added to the reaction vessel. The solution was refluxed for 19 h, concentrated and purified to obtain the desired compound.

See the Supplementary Data for parasite culture and assay information.

Results & discussion

The initial synthetic route utilized to obtain 3g–j is shown in Figure 4. This route uses the traditional ways of adding an alkyl group to the 3-position of 2-substituted-1,4-naphthoquinones via radical alkylation using either a Hunsdiecker decarboxylation of a carboxylic acid using silver nitrate and peroxysulfate or conversion of a carboxylic acid to a diacyl peroxide with subsequent thermal decomposition [28]. Unlike Fieser's work, separate syntheses of 3g–j were pursued to avoid too much reliance on the synthesis of one compound (e.g., 3j) and subsequent Hooker oxidations to obtain compounds with fewer methylene groups (e.g., 3g–i).

Figure 4. . Initial method to synthesize 3g–j.

DCC: N,N’-dicyclohexylcarbodiimide; DMAP: 4-dimethylaminopyridine; THF: Tetrahydrofuran.

The planned route started from 4-hydroxyphenyl esters and 20a and 20b were commercially available (Figure 4). Methyl 4-(4-hydroxyphenyl)butanoate (20c) was synthesized from 4-(4-methoxyphenyl)butyric acid (25) in two steps based on the literature (Figure 5). In the first step, 25 was demethylated using hydrobromic acid [29], and then without purification, a Fischer esterification reaction with methanol yielded 20c in an 88% yield [30]. A Fischer esterification reaction with 5-(4-hydroxyphenyl)pentanoic acid (26) was used to produce 20d in an 89% yield (Figure 5) [30]. The 4-hydroxyphenyl esters 20a–d were then reacted separately in a copper coupling reaction with 4-(trifluoromethoxy)phenylboronic acid (21) [20] to provide 22a–d in 45%, 90%, 76% and 70% yields, respectively (Figure 4). Next, separate saponification reactions with sodium hydroxide were used to convert 22a–d to the phenoxyphenyl acids 23a–d (Figure 4) in good yields (82%, 100%, 91% and 87%, respectively) [32]. Recently, a new scalable procedure for ELQ-300 (19) reported a one-step reaction to synthesize 23a in a 78% yield by utilizing the Ullmann reaction between 4-bromophenylacetic acid and 4-trifluoromethoxyphenol [33]. This method could also be used to synthesize 23b–d.

Figure 5. . Synthesis of 20c and 20d.

Having obtained 23a–d, the next step was to utilize these compounds in a radical alkylation reaction to obtain the desired products. Radical alkylations of naphthoquinones are known to be challenging reactions [1,28]. Based on the literature [23], 23a was reacted with silver nitrate, ammonium peroxydisulfate and 2-bromo-1,4-naphthoquinone (24; Figure 4) in a Hunsdiecker reaction and the resulting product was subjected to a nucleophile substitution reaction with potassium hydroxide to yield 3g in 25% yield. Unfortunately, when acid 23b was reacted in the same manner with 24, the percent yield of 3h dropped to 2%. After investigating other methods for free radical alkylation reactions, compounds 3h–j were synthesized utilizing the thermal decomposition of diacyl peroxides. Specifically, carboxylic acids 23b–d were reacted with hydrogen peroxide, N,N’-dicyclohexylcarbodiimide and 4-dimethylaminopyridine to obtain a diacyl peroxide compound [24], which was then reacted with 2-hydoxy-1,4-naphthoquinone [2] (5) to form 3h, 3i and 3j in 6%, 28% and 20% yields, respectively (Figure 4). To determine the impact of the OCF₃ group, it was necessary to synthesize compounds 3b and 3c, which was accomplished using the same procedure as above and commercially available carboxylic acids 27a and 27b, respectively (Figure 6). These compounds were obtained in low yields and work was not done to optimize these yields. Compounds 3b and 3c were reported in 1948 [2] and further characterization of these compounds are reported here.

Figure 6. . Synthesis of 3b and 3c.

DCC: N,N’-dicyclohexylcarbodiimide; DMAP: 4-dimethylaminopyridine.

Next, compounds 3b, 3c, 3g–j and atovaquone (4) were tested for antimalarial activity against P. falciparum strains, and the results are shown in Table 1. The drug-sensitive and multidrug-resistant strains were D6 (sensitive to chloroquine and resistant to mefloquine), Dd2 (resistant to multiple antimalarial drugs), 7G8 (resistant to chloroquine), D1 (resistant to ELQ-300) and Tm90-C2B (resistant to atovaquone [4]). Based on Fieser's work with P. lophurae, 3i was expected to be the most potent compound from 3g–j; however, 3h was the most potent compound against P. falciparum strains D6, Dd2, 7G8 and D1 with low nM IC₅₀ values (entry 4, Table 1). This variance could be due to the difference between the cytochrome bc₁ complexes of avian (P. lophurae) and human (P. falciparum) malaria species. Interestingly, the activity of 3c (entry 2, Table 1) against strains D6, Dd2, 7G8 and D1 is similar to 3i (entry 5, Table 1), indicating that the OCF₃ group does not significantly change the IC₅₀ values. A slight improvement in potency is seen for 3h (entry 4, Table 1), with the OCF₃ group, over 3b (entry 1). The difference in activity between 3g (entry 3, Table 1) and 3h–j (entries 4–6) is quite striking, with 3g (one methylene group) having little activity and 3h–j (two to four methylene groups) having much lower IC₅₀ values. The comparative inhibitory activity of all the synthesized compounds against strains D6 and Tm90-C2B indicates that these compounds, like atovaquone (4), bind to the Q_o site of cytochrome bc₁ rather than the Q_i site, like 18 (GSK932121) and 19 (ELQ-300). In comparison with atovaquone (4; entry 8, Table 1), 3h (entry 4) has IC₅₀ values ∼73-times less potent against strain D6 and ∼3.4-times less potent against strain D1.

Table 1. . Activity of compounds 3b, 3c, 3g–k and 4 against multiple Plasmodium falciparum strains.

Entry	Compound	n	X	IC50 (nM)
				D6	Dd2	7G8	D1	Tm90-C2B
1	3b	2	H	15	ND	ND	15	>2500
2	3c	3	H	11	26	79	69	>2500
3	3g	1	OCF3	650	690	>2500	>2500	>2500
4	3h	2	OCF3	7.3	16	25	11	>2500
5	3i	3	OCF3	14	28	80	71	>2500
6	3j	4	OCF3	16	55	150	150	>2500
7	3k	2	F	14	ND	ND	ND	>2500
8	4: atovaquone			0.10	0.10	ND	3.2	>2500

ND: Not determined.

Further work was conducted to obtain the most active compound, 3h, in higher yields and to be able to synthesize derivatives of 3h. More recent work has reported the proline catalyzed three-component reductive alkylation reaction between aldehydes and 2-hydroxy-1,4-naphthoquinone to form 2-hydroxy-1,4-naphthoquinones with an alkyl substituent at the 3-position in high yields (Figure 2D) [17,27]. The necessary aldehydes for this reaction could be made by the reduction of esters 22a–d (Figure 4). Specifically, 22a could be converted to the aldehyde and then reacted with 2-hydroxy-1,4-naphthoquinone to obtain 3h.

Attempts to selectively reduce 22a with diisobutylaluminum hydride resulted in a mixture of the desired aldehyde and alcohol products, as indicated by proton NMR. Due to the significant amount of alcohol that was formed, ester 22a was completely reduced using LiAlH₄ [25] to alcohol 28a in a 95% yield (Figure 7). Next, 28a was oxidized with Dess–Martin periodinane [25]; the resulting aldehyde was quickly purified and then reacted with 2-hydroxy-1,4-naphthoquinone (5), diethyl 1,4-dihydro-2,6-dimethyl-3,5-pyrinedicarboxylate (15) and a catalytic amount of L-proline [17, 27] to obtain 3h in a 34% yield. This yield is lower than the 84% yield to synthesize 16 (Figure 2D); however, 2 equiv. of aldehyde 15 were used to make 16 and only about 1 equiv. of aldehyde was used to synthesize 3h. In a model reaction, using 2 equiv. of aldehyde improved the reaction yield but the yield did not double. With this improved yield for the C–C bond-forming reaction, 3i was synthesized again utilizing the reductive alkylation method (Figure 7). Specifically, 22b was reduced to alcohol 28b in an 87% yield. This alcohol was then oxidized via a Swern oxidation [26] and the resulting aldehyde was reacted with 2-hydroxy-1,4-naphthoquinone (5), diethyl 1,4-dihydro-2,6-dimethyl-3,5-pyrinedicarboxylate (15) and a catalytic amount of L-proline [17,27] to obtain 3i in a 33% yield. Another metabolically stable version of 3b with a fluorine atom (3k; Figure 7) was pursued with the reductive alkylation methods. This compound was synthesized by the reduction of commercially available carboxylic acid 29 to obtain alcohol 28c (Figure 7), which has previously been synthesized by a different method [34]. This alcohol was oxidized with Dess–Martin periodinane and then used in a reductive alkylation reaction with 5 and 15 to obtain 3k in 22% yield. Compound 3k (entry 7, Table 1) is about twofold less potent than 3h (entry 4) against malarial strain D6 and has nearly the same IC₅₀ values as 3b (entry 1).

Figure 7. . Synthesis of 3h, 3i and 3k.

THF: Tetrahydrofuran.

In summary, compounds 3g–j, the OCF₃ derivatives of 3a–d, were synthesized and tested for antimalarial activity to understand how the modifications influence the potency. Initially, 3g–j were synthesized utilizing radical alkylation reactions with naphthoquinones. The yield for 3h was lower than the other compounds, so a proline-catalyzed, three-component reductive alkylation reaction was used to synthesize 3h in improved yield. With this synthetic improvement, compound 3k was also synthesized. Based on the reported activity of 3a–d with P. lophurae (avian) in which 3c was the most potent, 3i (the OCF₃ analog of 3c) was expected to have the lowest IC₅₀ value against various drug-resistant P. falciparum (human) strains. However, assaying 3h–k revealed that compound 3h had the lowest IC₅₀ values, which could be due to differences between avian and human malaria species. The OCF₃ group does provide 3h with a slight lowering of the IC₅₀ values relative to 3b (the non-OCF₃ analog) and 3k (the F analog). The comparative inhibitory activity with strains D6 and Tm90-C2B indicates that the newly synthesized 3-alkyl-2-hydroxy-1,4-naphthoquinones also bind to the Q_o site of cytochrome bc₁, like the 3-alkyl-2-hydroxy-1,4-naphthoquinone atovaquone (4), rather than the Q_i site, which is targeted by OCF₃ diarylether-containing compounds 18 and preclinical candidate ELQ-300 (19).

Conclusion & future perspective

Malaria, caused by Plasmodium parasites, is a disease that continues to affect millions of people each year and results in many deaths. The ambitious goal to eradicate malaria continues and focuses on surveillance, prevention, case management and elimination. Antimalarial drugs will continue to be an important part of this effort. This may occur through the discovery of new compounds and molecular targets or via modifications to known compounds. Adding cytochrome bc₁ Q_i site inhibitors will be a useful expansion to the arsenal of antimalarial drugs.

Summary points.

• In 1948, Fieser and co-workers reported more than 300 structurally related compounds and identified a 2-hydroxy-1,4-naphthoquinone with an alkyldiarylether side chain as a promising compound that has Plasmodium lophurae activity.

• Recent antimalarial compounds GSK932121 and ELQ-300 contain diarylether groups with a trifluoromethoxy group.

• The goal of this study was to synthesis trifluoromethoxy derivatives of Fieser's 2-hydroxy-1,4-naphthoquinone compounds with alkyldiarylether side chains.

• Desired compounds were synthesized using a radical alkylation reaction with naphthoquinones.

• Inhibitory assays with drug-sensitive and multidrug-resistant strains of Plasmodium falciparum found one trifluoromethoxy derivative has lower IC₅₀ values, but it was not the one expected based on Fieser's work. This indicates that there are differences between avian (P. lophurae) and human malarial parasites (P. falciparum) in their sensitivity to these molecules. The most potent trifluoromethoxy derivative does have slightly lower IC₅₀ values relative to the non-OCF₃ and F analogs.

• The comparative inhibitory activity of these compounds indicates that these compounds bind to the Q_o site of cytochrome bc₁, like atovaquone, rather than the Q_i site, which is targeted by GSK932121 and ELQ-300.

• The most active compound was obtained via a low yield for the radical alkylation step, so the synthetic method was changed to utilize a proline-catalyzed, three-component reductive alkylation. This improved the yield for the key C–C bond-forming step.