Spirofused Tetrahydroisoquinoline-Oxindole Hybrids (Spiroquindolones) as Potential Multitarget Antimalarial Agents: Preliminary Hit Optimization and Efficacy Evaluation in Mice

ABSTRACT

Previous studies suggest that 3′,5′-dihydro-2′H-spiro[indoline-3,1′-isoquinolin]-2-ones (DSIIQs [spiroquindolones]) are multitarget antiplasmodial agents that combine the actions of spiroindolone and naphthylisoquinoline antimalarial agents. In this study, 12 analogues of compound (±)-5 (moxiquindole), the prototypical spiroquindolone, were synthesized and tested for antiplasmodial activity. Compound (±)-11 (a mixture of compounds 11a and 11b), the most potent analogue, displayed low-nanomolar activity against P. falciparum chloroquine-sensitive 3D7 strain (50% inhibitory concentration [IC₅₀] for 3D7 = 21 ± 02 nM) and was active against all major erythrocytic stages of the parasite life cycle (ring, trophozoite, and schizont); it also inhibited hemoglobin metabolism and caused extensive vacuolation in parasites. In drug-resistant parasites, compound (±)-11 exhibited potent activity (IC₅₀ for Dd2 = 58.34 ± 2.04 nM) against the P. falciparum multidrug-resistant Dd2 strain, and both compounds (±)-5 and (±)-11 displayed significant cross-resistance against the P. falciparum ATP4 mutant parasite Dd2 SJ733 but not against the Dd2 KAE609 strain. In mice, both compounds (±)-5 and (±)-11 displayed dose-dependent reduction of parasitemia with suppressive 50% effective dose (ED₅₀) values of 0.44 and 0.11 mg/kg of body weight, respectively. The compounds were also found to be curative in vivo and are thus worthy of further investigation.

INTRODUCTION

Malaria continues to present a major public health challenge, with about 241 million cases and 627 thousand deaths globally in 2020—an increase of 6% in cases and 12% in deaths over the previous year (1).

Before the 1980s, the antimalarial armamentarium consisted mainly of the quinolines (represented by quinine, chloroquine [CQ], mefloquine, primaquine, and amodiaquine) and antifolates (such as proguanil, pyrimethamine, and sulfadoxine). Other compounds, such as lumefantrine, pyronararidine, and atovaquone, helped to round out this inventory. The advent of artemisinin (ART) in the 1980s substantially changed the malaria control landscape, and the subsequent adoption of artemisinin combination therapies (ACTs) as the standard frontline treatment for malaria further provided an additional boost to malaria control (2, 3). However, over the years, reports of resistance to available drugs, including artemisinin, have continued to surface, thus underscoring the need for continued exploration of new antimalarials with novel modes of action. In response to this recurring need, several new compounds have been discovered and some have entered the development pipeline in recent years (4). Prominent among these are cipagarmin (KAE609) and SJ733, which are of particular interest because they act by inhibition of Plasmodium falciparum ATPase (Pf-ATP4), a Na⁺ efflux pump and an essential enzyme for maintaining osmotic balance in malaria parasites. The disruption of sodium efflux in the parasite using either KAE609 or SJ733 has been shown to induce profound physical changes in the infected cells, including cell swelling, increased membrane rigidity, and externalization of phosphatidylserine (5, 6). This mechanism of action, hitherto unknown among antimalarial drugs, has generated considerable interest.

Structurally, cipagarmin (KAE609) is noteworthy in that it is composed of two easily identifiable privileged scaffolds, oxindole and 1,2,3,4-tetrahydro-β-carboline, and may therefore be regarded as a molecular hybrid, defined as the product of the combination of two or more pharmacophoric units from different bioactive compounds (7). Privileged scaffolds, as defined by Evans et al. (8), are molecular frameworks that occur in numerous biologically active low-molecular-weight compounds. Indeed, both the oxindole and 1,2,3,4-tetrahydro-β-carboline fragments have been identified in many compounds that display diverse pharmacological activities, including antimalarial activity. The novel pharmacological profile of cipargamin therefore provides inspiration for the exploration of new molecular hybrids as potential antimalarial agents.

In a previous report (9), we described the antiplasmodial activity of 3′,5′-dihydro-2′H-spiro[indoline-3,1′-isoquinolin]-2-ones (DSIIQs), a recently described class of spirooxindoles (10). Our investigation of this class of compounds was prompted by the observation that the DSIIQ scaffold (compound 1 in Fig. 1) is a molecular hybrid of two privileged scaffolds, tetrahydroisoquinoline (THIQ) and oxindole (OX). Moreover, DSIIQ shares prominent structural features, specifically the tetrahydroisoquinoline (THIQ) and oxindole (OX) fragments, with two mechanistically dissimilar antimalarial scaffolds: the naphthylisoquinolines (NIQs [represented by compound 2]), a naturally occurring class of potent antimalarials, and the synthetic spiroindolones (represented by KAE609 [compound 3]) (Fig. 1). Spiroindolones inhibit Pf-ATP4, resulting in deleterious effects on all stages of the erythrocytic phase of the parasite life cycle (11), display transmission blocking activity (12), and are active in vivo (11). On the other hand, the naphthylisoquinolines display potent activity against the asexual erythrocytic stages of P. falciparum (13, 14), are active against the liver stages of Plasmodium berghei (15), and are curative (16). The mode of action of naphthylisoquinolines appears to be primarily inhibition of hemoglobin metabolism (17, 18). Consequently, we proposed that a compound that combines the structural attributes of the spiroindolones and NIQs should be a powerful and efficient multitarget antimalarial agent.

FIG 1.

3′,5′-Dihydro-2′H-spiro[indoline-3,1′-isoquinolin]-2-ones as molecular hybrids.

In the above-mentioned study (9), we identified and confirmed a hit, (±)-moxiquindole [compound (±)-5], that displayed low-micromolar activities against both the CQ-sensitive P. falciparum strain 3D7 (50% inhibitory concentration [IC₅₀], 1.865 μM) and the multidrug-resistant P. falciparum strain Dd2 (IC₅₀, 1.730 μM) and was therefore subjected to further in vitro pharmacological evaluation. We also discovered that compound (±)-5 is a rapid acting multistage antimalarial agent that inhibits hemoglobin metabolism and interferes with parasite lipid dynamics (9). In addition, the compound was found to cause abnormal vacuolation in the parasites, similar to that observed following treatment with the prototypical spiroindolone antimalarial KAE609 (11). Taken together, the evidence pointed to (±)-moxiquindole [compound (±)-5] and its analogues, which we refer to as spiroquindolones, as multitarget antimalarials that combine aspects of the modes of action of the naphthylisoquinolines and spiroindolones. This combination of actions by spiroquindolones presented an attractive pharmacological profile that motivated further medicinal chemistry investigation of this class of spirooxindoles as potential antimalarial agents.

Herein, we describe the synthesis and pharmacological characterization of 12 previously unreported (±)-moxiquindole [compound (±)-5] analogues with a particular focus on (±)-trans5-chloro-6′,7′-dimethoxy-3′-methyl-3′,5′-dihydro-2′H-spiro[indoline-3,1′-isoquinolin]-2-one [compound (±)-11] (Fig. 2), a highly active compound obtained during this hit optimization effort.

FIG 2.

Derivatives of 3′,5′-dihydro-2′H-spiro[indoline-3,1′-isoquinolin]-2-one.

RESULTS

Chemistry.

In the current study, 12 compounds (Fig. 2) were prepared by reacting either 3,5-dimethoxyphenethylamine or α-methyl-3,5-dimethylphenethylamine with the corresponding isatins in polyphosphoric acid (Fig. 3 and Fig. 4) as described earlier (10). The yields were moderate to good without optimization of the reaction conditions.

FIG 3.

Synthesis of spiroquindolones: part I.

FIG 4.

Synthesis of spiroquindolones: part II.

In the case where α-methyl-3,5-dimethylphenethylamine was used, a mixture of diastereomers was obtained, owing to the presence of two stereogenic centers (C-1′ and C-3′) in the resulting spiroquindolone. The mixture was separated by column chromatography on silica gel to yield two racemates, compounds (±)-11 and (±)-12. The racemate displaying higher chromatographic mobility was designated compound (±)-11 and assigned the (1′S,3′S/1′R,3′R), or trans, configuration, while the other was designated (±)-12 and assigned the (1′S,3′R/1′R,3′S), or cis, configuration, based on the analysis provided herein. The designations trans and cis refer to the relative dispositions of atoms on the two stereogenic centers C-1′ and C-3′. These designations also apply to the analogues (±)-13 and (±)-14.

All target compounds were characterized by liquid chromatography-mass spectrometry (LC-MS) and nuclear magnetic resonance (NMR) spectrometry. For compounds (±)-11 and (±)-12, which contain two stereogenic centers, assignment of structure was initially based on comparison with spectroscopic data reported for the related spiroindolones (19). As reported by these authors, the methine proton at the C-3 position of the tetrahydroisoquinoline fragment (Fig. 2) is deshielded, by 0.56 ppm, in the trans isomer relative to the cis isomer. In contrast, the ¹³C signal for the C-3 carbon of this same fragment is deshielded by 1.5 ppm in the cis isomer relative to the trans isomer. In the current study, a similar pattern was observed. The chemical shift of the C-3 methine proton of compound (±)-11 was observed at δ 3.83 to 3.95, while that of compound (±)-12 was found at δ 3.38, giving a difference of 0.51 ppm. However, the ¹³C signal for C-3 carbon of the same fragment was observed at 53.5 ppm for compound (±)-11 but at 55.0 ppm for compound (±)-12, yielding a cis/trans difference of 1.5 ppm. The assignment of configuration was unequivocally confirmed by the X-ray crystal structure of compound (±)-11 (Fig. 5), which shows that this compound does indeed possess the (1′S,3′S/1′R,3′R) configuration (see the supplemental material).

FIG 5.

ORTEP diagram of compound (±)-11 with 50% ellipsoids.

Biological studies.

(i) Antiplasmodial activity of moxiquindole analogues. As shown in Table 1, 8 out of the 12 newly synthesized compounds [(±)-6 to (±)-17] displayed antiplasmodial activity below the hit selection cutoff point of 2 μM. In spite of the small size of the data set, it is clear that small electron-withdrawing substituents are favored at position C-5 of the oxindole fragment [compare compounds (±)-4 and (±)-5 or (±)-8], but electron-donating substituents are poorly tolerated at the same position [compare compounds (±)-7 and (±)-5 or (±)-8]. Moreover, position C-6 of this oxindole fragment is capable of some bulk tolerance [compare compounds (±)-5 or (±)-8 and (±)-10]. Among the 5-halo-substituted analogues, the introduction of a methyl group at position C-3 of the THIQ fragment produced a large increase in antiplasmodial activity when the methyl group and the carbonyl fragment of the oxindole were in a trans orientation [compare compounds (±)-5, (±)-11, and (+)-12 and compounds (±)-5, and (±)-13, and (±)-14]. Compounds (±)-11 (IC₅₀ = 21 ± 02 nM) and (±)-13 (IC₅₀ = 71 ± 0.4 nM) thus emerged as the most potent analogues of the study. In view of its considerable activity, compound (±)-11 was chosen for further in vitro and in vivo pharmacological evaluation.

TABLE 1.

In vitro antiplasmodial activity of moxiquindole analogues for P. falciparum strains 3D7 and Dd2^a

Series I compound	S/Nb				IC50 (nM) for strain:
	R1	R2	R3	R4	3D7	Dd2
(±)-4	H	H	H	H	>10,000a	NT
(±)-5	H	H	Cl	H	1,865	1,673
(±)-6	H	H	NO2	H	4,962	5,367
(±)-7	H	H	NH2	H	>10,000	NT
(±)-8	H	H	F	H	1,517	NT
(±)-9	H	H	F	F	3,309	3,220
(±)-10	H	H	F	Piperidin-1-yl	1,750	>10,000
(±)-11 (trans)	CH3	H	Cl	H	23	62
(±)-12 (cis)	CH3	H	Cl	H	1,573	NT
(±)-13 (trans)	CH3	H	I	H	70.62	NT
(±)-14 (cis)	CH3	H	I	H	4,159	>10,000
(±)-15	H	CH3	NO2	H	925	NT
(±)-16	H	CH3	Cl	H	281.6	171.9
(±)-17 (trans)	CH3	CH3	Cl	H	1,115	1,079

^aSeries I compounds were tested in duplicates, and those with an IC₅₀ of <2 μM were confirmed in two separate experiments.

^bS/N, nucleophilic substitution.

(ii) Effect of compound (±)-11 on erythrocytic stages of P. falciparum. The effect of compound (±)-11 on the asexual erythrocytic stages of P. falciparum was assessed as earlier described (9). Compound (±)-11, like compound (±)-5, ART, CQ, and KAE609, was multistage active. The compound displayed inhibitory activity against ring-to-trophozoite development, trophozoite-to-schizont development, and schizont rupture, as indicated by an accumulation of rings, trophozoites and schizonts, respectively, in the drug-treated cultures relative to the solvent-treated controls (Fig. 6, Fig. 7, and Fig. 8). As expected, E64 was inactive against both trophozoite and schizont development, but it inhibited merozoite egress as shown in Fig. 8.

FIG 6.

Stage-specific activity against trophozoite development. Early ring-stage parasites (pretreatment) were treated with compounds and controls for 24 h, and the proportions of each parasite stage were calculated relative to that of the negative control. Data presented for each bar correspond to the mean from three experimental replicates.

FIG 7.

Stage-specific activity against schizont development. Early trophozoite-stage parasites (pretreatment) were treated with compounds and controls for 24 h, and the proportions of each parasite stage were calculated relative to that of the negative control. Data presented for each bar correspond to the mean from three experimental replicates.

FIG 8.

Stage-specific activity against schizont rupture. Early schizont-stage parasites (pretreatment) were treated with compounds and controls for 24 h, and the proportions of each parasite stage were calculated relative to that of the negative control. Data presented for each bar correspond to the mean from three experimental replicates.

As previously reported for KAE609 (11) and compound (±)-5 (9), compound (±)-11 was also found to induce vacuolation in all the parasites identified by light microscopy (Fig. 9). No such effect was observed in artemisinin-treated parasites, thus highlighting the similarities between the spiroindolones and spiroquindolones, while clearly distinguishing these two chemotypes from artemisinin.

FIG 9.

Giemsa-stained thin smears of drug-treated parasites showing vacuolation (red arrows) in both spiroquindolone-treated parasites [treated with compounds (±)-5 and (±)-11] and in spiroindolone-treated controls (KAE609) but not in the artemisinin-treated controls (ART). Parasites were treated at the early ring stage (ER), at the early trophozoite stage (ET), or at the early schizont stage (ES) for 24 h and then analyzed by Giemsa stain light microscopy. NC, negative control.

(iii) Effect of compound (±)-11 on hemoglobin metabolism. The effect of compound (±)-11 on hemoglobin metabolism was assessed using previously described methods (9, 20, 21), with each compound tested at either its IC₅₀ or at a maximum-effect concentration corresponding to 5-fold IC₅₀ or 10 μM. As shown in Fig. 10, a significant accumulation (more than 3-fold) of intraparasitic hemoglobin was observed in parasites treated with compounds (±)-11 and (±)-5 when tested at all three concentrations compared with the untreated controls. Similarly, a significant increase in intraparasitic hemoglobin level was observed in parasites treated with either chloroquine or the cysteine protease inhibitor E64 when used at concentrations near their IC₅₀ values. No significant accumulation of hemoglobin was recorded in parasites treated with ART as well as KAE609, clearly distinguishing the spiroquindolones from the spiroindolone KAE609.

FIG 10.

Differential effect of spiroquindolines and KAE609 (spiroindolone) on hemoglobin uptake and degradation. Parasites in the mid-to-late ring stages were treated for 24 h with compounds at their respective IC₅₀ values or at a maximum-effect concentration of 5× the IC₅₀ value or 10 μM, and the intraparasitic hemoglobin levels were determined by spectrophotometric measurements at the 410-nm wavelength. Fold changes relative to hemoglobin levels in the untreated controls are the means of two independent experiments. Error bars represent standard deviations (SD) calculated from the two independent experiments.

(iv) Cytotoxicity and cross-resistance profiles of compound (±)-11 against drug-resistant parasite lines. Cytotoxicity evaluation of both compounds against human liver (HepG2) cells showed no significant decrease in cell viability, with a calculated 50% cytotoxic concentration (CC₅₀) of >200 μM and selectivity indices (SIs) greater than 8,000-fold. Like parent compound (±)-5, compound (±)-11 exhibited equipotent activity (resistance index of <3) against chloroquine-sensitive 3D7 and the chloroquine/multidrug-resistant Dd2 strain (cf. Table 1 and Table 2), indicating significant differences in the drug target sites. In a bid to better understand the mode of action of spiroquindolones, specifically with respect to their interaction with the drug target of spiroindolones, cross-resistance studies were carried out on well-characterized mutant strains of P. falciparum that are resistant to KAE609 (NITD609-R^Dd2) or SJ733 (DD2-SJ16-D2), two antimalarials that target distinct sites on P. falciparum Na⁺-ATPase (5, 11). As seen in Table 2, KAE609 displayed about 5-fold reduction in activity in the KAE609-resistant line relative to the wild type. However, the activities of both compounds (±)-5 and (±)-11 declined only 2- to 3-fold between the two cell lines, indicating no cross-resistance between the spiroquindolones and KAE609. On the other hand, compound (±)-5 was inactive against the SJ733 resistant cell line (IC₅₀ > 10,000) and the activity of compound (±)-11 declined 116-fold, indicating high-level cross-resistance between the spiroquindolones and the dihydroisoquinolone SJ733. For comparison, the activity of SJ733 declined more than 150-fold in the SJ733 mutant line relative to the wild type.

TABLE 2.

Cytotoxicity and cross-resistance profiles of compounds (±)-5 and (±)-11 against wild-type Dd2 and the Dd2 KAE609 and SJ733 mutant lines^a

Compound	CC50 (μM) for Hep-G2 cellsb	IC50 (nM)			Fold shift in IC50
		Dd2 (a)	Dd2 KAE609 (b)	Dd2 SJ733 (c)	b/a	c/a
(±)-5	>200	580.2 ± 46.4	1,393 ± 47	>10,000	2.4	>17
(±)-11	>200	58.34 ± 2.04	159.9 ± 10.10	6776 ± 370	2.7	116
KAE609	NDc	1.361 ± 0.10	6.244 ± 0.01	ND	4.6	ND
SJ733	ND	65.15 ± 1.82	ND	>10,000	ND	>150

^aThe mutant strains used include Dd2-SJ16-D2 (SJ733 resistant) and NITD609-R^Dd2 clone 3 (5, 11).

^bPuromycin was used as the cytotoxicity positive control, and the obtained CC₅₀ (1.7 ± 0.32 μM) was consistent with previously reported values.

^cND, not determined.

(v) In vivo efficacy and safety of spiroquindolones (±)-5 and (±)-11. As shown in Fig. 11 and in Table 3, both candidate compounds (±)-5 and (±)-11 displayed significant in vivo antiplasmodial activities in a dose-dependent manner, with calculated 50% effective dose (ED₅₀) values of 0.44 and 0.11 mg/kg of body weight, respectively. As seen by light microscopy examination of thin smears, treatment of mice with either compound at doses of 3 mg/kg or more resulted in a significant reduction in blood parasitemia, comparable to that of chloroquine (Fig. 12).

FIG 11.

In vivo chemosuppression curves of compounds (±)-5 and (±)-11. Groups of P. chabaudi-infected mice (n = 3) were treated daily for 3 days with test compounds at various concentrations (0.03, 0.3, 3, 10, and 30 mg/kg body weight), and the day 4 parasitemias were quantified by light microscopy. The percentage of suppression refers to the percentage of reduction in parasitemia on day 4 relative to the untreated mice group. Error bars represent the SD (percentage) of parasitemia at each dose tested.

TABLE 3.

Chemosuppressive potential of compounds (±)-5 and (±)-11 in mice infected with P. chabaudi

Treatment group	Dose (mg/kg)	Parasitemia ona:					% of suppressionb
		Day 0	Day 1	Day 2	Day 3	Day 4
Untreated	1× PBS	0.00 ± 0.00	1.00 ± 0.57	1.77 ± 0.56	3.66 ± 0.82	6.89 ± 1.05	0
Chloroquine	10	0.00 ± 0.00	0.66 ± 0.35	1.11 ± 0.51	0.55 ± 0.39	0.11 ± 0.19	98.40
Compound (±)-5	30	0.00 ± 0.00	0.55 ± 0.19	0.66 ± 0.35	0.22 ± 0.38	0.00 ± 0.00	100.00
	10	0.00 ± 0.00	1.66 ± 0.67	0.77 ± 0.20	0.33 ± 0.33	0.22 ± 0.19	96.81
	3	0.00 ± 0.00	1.22 ± 0.77	1.00 ± 0.88	0.33 ± 0.33	0.11 ± 0.19	98.40
	0.3	0.00 ± 0.00	0.77 ± 0.20	1.89 ± 0.77	2.66 ± 1.00	3.11 ± 2.05	54.86
	0.03	0.00 ± 0.00	0.77 ± 0.77	1.77 ± 0.20	2.77 ± 0.20	3.22 ± 1.50	53.27
Compound (±)-11	30	0.00 ± 0.00	1.00 ± 0.35	2.11 ± 0.51	0.17 ± 0.19	0.00 ± 0.00	100.00
	10	0.00 ± 0.00	0.88 ± 0.39	1.89 ± 0.77	1.22 ± 0.51	0.22 ± 0.19	96.81
	3	0.00 ± 0.00	1.33 ± 0.67	2.00 ± 0.88	0.68 ± 0.38	0.33 ± 0.33	95.21
	0.3	0.00 ± 0.00	0.89 ± 0.51	1.77 ± 0.51	3.00 ± 0.67	1.66 ± 0.35	75.91
	0.03	0.00 ± 0.00	0.77 ± 0.51	1.55 ± 1.02	2.33 ± 0.33	3.88 ± 0.69	43.69

^aData are means ± SD of the percentages of parasitemia from each group of three mice determined by light microscopy.

^bShown are the percentages of reduction in mean parasitemia on day 4 relative to the mean parasitemia in the untreated group. Values corresponding to significant suppression of parasite growth are in bold.

FIG 12.

Dose-dependent chemosuppressive activity of compounds (±)-5 and (±)-11 against P. chabaudi infection in mice. Groups of P. chabaudi-infected mice (n = 3) were treated daily for 3 days with test compounds at variable concentrations, and the day 4 parasitemias were quantified by Giemsa-based light microscopy. The percentage of suppression refers to the percentage of reduction in mean parasitemia on day 4 relative to the untreated mice group. Error bars represent the SD (percentage) of treated groups relative to the mean parasitemia in the untreated group.

Evaluation of the curative activity of the compounds revealed that at both concentrations tested (ED₅₀ or ED₉₀), there was a progressive reduction in parasitemia by both compounds relative to the control (Fig. 13). This was observed from day 3 and sustained for the duration of the experiment. Additionally, both compounds demonstrated a significant protective potential in mice as was observed in the survival time curves, with 100% and 80% survival at ED₉₀ and ED₅₀ concentrations, respectively, on day 29 (Fig. 14).

FIG 13.

Curative effect of compounds (±)-5 (A) and (±)-11 (B) against P. chabaudi infection in Swiss albino mice. Groups of P. chabaudi-infected mice (n = 5) were treated daily with the compounds at their estimated ED₅₀ [0.44 versus 0.11 mg/kg body weight for compounds (±)-5 and (±)-11, respectively] or at ED₉₀ concentrations [0.75 versus 1.25 mg/kg body weight for compounds (±)-5 and (±)-11, respectively] for 5 consecutive days starting from day 3 postinfection and then monitored for up to day 29 postinfection for complete parasite clearance. Data represent the mean ± SD percentage of parasitemia on each specified day postinfection as determined by light microscopy.

FIG 14.

Survival time of P. chabaudi-infected Swiss albino mice following treatment with compound (±)-5 or (±)-11 at either ED₅₀ (A) or ED₉₀ (B).

DISCUSSION

Due to the emergence of resistance to even the newer antimalarials, which threatens the progress achieved in malaria control in recent years, research continues into the discovery of new antimalarial agents with novel modes of action. In the last decade, the spiroindolones provided new avenues for malaria drug discovery, following the identification of their novel mode of action as inhibitors of Plasmodium falciparum Na⁺-ATPase (11, 22). Used alone or in combination with other antimalarials, the spiroindolones (or other compounds acting in a similar manner) may offer powerful new tools in the fight against drug-resistant malaria. In a previous paper (9), we presented studies suggesting that 3′,5′-dihydro-2′H-spiro[indoline-3,1′-isoquinolin]-2-ones, represented by (±)-moxiquindole [compound (±)-5], are multitarget antiplasmodial agents that combine the modes of action of the spiroindolones and naphthylisoquinolines, two potent but mechanistically different classes of antimalarials. Buoyed by these results, we embarked on a search for more potent (±)-moxiquindole analogues through further structure modification. For this exercise, the structure of (±)-moxiquindole was divided into two major fragments—THIQ and oxindole—and the modifications were then targeted at specific positions of each fragment based on previously identified structure-activity trends (23). For the present study, the 6,7-dimethoxy substitution pattern of (±)-moxiquindole was maintained while modifications were carried out at the C-5 and C-6 positions of the oxindole fragment and the N-2 and C-3 positions of THIQ fragment. The structural modifications were based in part on previous studies on the spiroindolones that clearly show the impact of substitution, especially at the C-5 and C-6 positions of the oxindole fragment, on antiplasmodial activity and pharmacokinetic properties (23). Substituents were thus chosen to investigate both electronic and steric properties in order to gain a better insight into the structural determinants of antiplasmodial activity in the spiroquindolone series. The 5′-amino-substituted compound was synthesized partly for its ability to serve as an intermediate for other target molecules to be prepared subsequently. Compound 13 may also serve as precursor and could be transformed into other compounds using a variety of methods, including Suzuki and Heck coupling. As shown in Table 1, some modest gains in activity were obtained from substitution at the C-5 and N-2 positions of the oxindole and the N-2 position of the THIQ fragments, respectively. However, the most notable increase in activity was obtained by the insertion of a methyl group at the C-3 position of the THIQ fragment to produce a homologue of compound (±)-5; the resulting compound, (±)-11 [(±)-homoquindole], displayed low-nanomolar antiplasmodial activity (IC₅₀ for P. falciparum 3D7 = 21 ± 2 nM) and is thus 89 times more potent than the parent compound, (±)-moxiquindole. The methyl group may therefore play an important role in the interaction of these compounds with their respective targets. Indeed, the increase in activity obtained by the insertion of a methyl group at the C-3 position of the THIQ fragment is particularly noteworthy, as it reveals the following striking parallels between spiroquindolones, NIQs, and spiroindolones: (i) insertion of a methyl group at the C-3 position of the THIQ fragment results in increased antiplasmodial activity for both the spiroindolone series (24) and the spiroquindolones [compare compounds (±)-5 and (±)-11]; (ii) for both the spiroindolones and spiroquindolones substituted with this C-3-methyl group, the (1′S,3′S/1′R,3′R) isomer is generally more active than the (1′S,3′R/1′R,3′S) isomer (24) [also compare compounds (±)-11 and (±)-12] and (iii) the C-3-methyl group is found in all three compound series in question.

At the molecular level, spiroquindolones appear to target Pf-Na+-ATPase, as suggested by the cross-resistance studies. However, it would appear that these compounds do not interact with the Pf-ATP4 pump in the same manner as the spiroindolones, for both compounds (±)-5 and (±)-11 show low-level cross-resistance with KAE609 but much higher cross-resistance with the dihydroisoquinolone SJ733. This observation may be consistent with the fact that both the spiroquindolones and SJ733 are partially reduced isoquinolines. Additional studies are however needed to confirm the similarity of target binding sites.

The identification of the highly potent compound (±)-11 therefore provided a tool for further proof-of-concept studies, which were conducted with the primary hit compound (±)-5 and KAE609 as reference compounds. As shown in Fig. 6 to 8, (±)-11 proved to be multistage active, inhibiting all three asexual erythrocytic stages of P. falciparum and inducing vacuolization in the parasite (Fig. 9) as earlier reported for compound (±)-5. This behavior was also observed with KAE609, a spiroindolone currently in clinical studies for the treatment of malaria. In subsequent studies, compound (±)-11 was found to cause accumulation of hemoglobin in the parasite, similar to compound (±)-5; however, KAE609 had no significant effect, suggesting that compounds (±)-5 and (±)-11 may have a broader spectrum of activity than KAE609. The pharmacological profile of homoquindole [compound (±)-11] therefore consistently and qualitatively mirrors that of the parent compound, moxiquindole [compound (±)-5], in all aspects studied, while it differs from that of KAE609 in those aspects relating to hemoglobin metabolism. This is precisely what would be expected of a compound that combines the modes of action of the spiroindolones and the naphthylisoquinolines. Structurally, the spiroquindolones may be viewed from at least three perspectives: (i) as spiroindolone analogues in which the THβC scaffold has been replaced by the THIQ framework, (ii) as tetrahydroisoquinoline-oxindole hybrids, and (iii) as linearly abbreviated dimensional probes obtained by the deletion of the pyrrolo fragment of the spiroindolones. Given these multiple structural attributes, it is not surprising that spiroquindolones appear to have a more complex mode of action than that of the spiroindolones. Subsequent studies in comparison with other known spirooxindoles and NIQs would shed light on the relative effect of spiroquindolones on sodium homeostasis and hemoglobin metabolism, as well as confirm the lethality of both the PfATP4 and hemoglobin effects of the new compounds.

Both compounds (±)-5 and (±)-11 displayed chemosuppressive and curative activity in vivo; however, the large (80-fold) difference in potency observed in vitro was no longer apparent. Indeed, compound (±)-5 was almost as effective as compound (±)-11 under the study conditions, possibly implicating differences in the pharmacokinetics of the two compounds. Subsequent evaluation of the metabolism and disposition of these compounds may explain these observations.

We conclude that spiroquindolones display potent multistage antiplasmodial activity through a combination of actions, including inhibition of Pf-ATPase, inhibition of hemoglobin metabolism, and potential disruption of intraparasite lipid dynamics. The compounds displayed potent in vivo suppressive activities, were curative in mice, and therefore deserve further investigation as potential multitarget antimalarial compounds.

MATERIALS AND METHODS

Chemistry.

(i) General. The synthesis of compounds was carried out in the Pharmaco-chemistry Research Laboratory of the Department of Chemistry at the University of Buea. Chemicals were purchased from Sigma-Aldrich Chemicals Company and were used as supplied. All solvents were reagent grade. Solvent removal was carried out under reduced pressure using a Buchi rotatory evaporator at temperatures not greater than 60°C. Melting points were measured using a Mel-Temp II apparatus with the use of open capillaries and are uncorrected. The progress of all reactions was monitored by thin-layer chromatography (TLC) on aluminum-backed sheets with silica gel 60 F₂₅₄ plates obtained from Sigma-Aldrich; visualization was by UV light at λ 254 nm or by staining with iodine. Compounds were purified by medium-pressure liquid chromatography over silica gel 60-to-400 mesh, using solvent mixtures that are specified below. Nuclear magnetic resonance (NMR) spectra were obtained using a Brucker Avance III spectrometer operating at 600 MHz (H¹) and 150 MHz (¹³C). Spectra were recorded in deuterated solvents and referenced to residual solvent signals. Chemical shifts (δ) were measured in parts per million. Hydrogen and carbon assignments were done using gradient correlation spectroscopy (gCOSY), gradient heteronuclear single quantum correlation (gHSQC) spectroscopy, and heteronuclear multiple bond correlation (gHMBC) techniques. Multiplicities are reported as singlet (s), doublet (d), doublet of doublets (dd), doublet of triplets (dt), triplet (t), triplet of doublets (td) and multiplet (m). Coupling constants (J) are reported in hertz. For biological evaluation, all compounds were converted to the corresponding hydrochlorides by treatment of the free bases with methanolic HCl. All compounds are greater than 95% pure by high-performance liquid chromatography (HPLC) analysis.

(ii) Synthesis of spiroquindolones. The synthesis of compounds 6 to 14 was accomplished with previously described methods (9). Details are provided in the supplemental material.

Synthesis of (±)-(1′S,3′S/1′R,3′R)-5-chloro-6′,7′-dimethoxy-3′-methyl-3′,5′-dihydro-2′H-spiro[indoline-3,1′-isoquinolin]-2-one [compound (±)-11] and (±)-(1′S,3′R/1′R,3′S)-5-chloro-6′,7′-dimethoxy-3′-methyl-3′,5′-dihydro-2′H-spiro[indoline-3,1′-isoquinolin]-2-one [compound (±)-12] by method A.

To an ethanolic solution of 3,5-dimethoxybenzaldehyde (5 g, 30.1 mmol) was added nitroethane (3.5 g, 55 mmol) and a catalytic amount of n-propylamine (0.2 g, 3.0 mmol). The reaction mixture was placed under reflux for 8 h, during which the reaction went to completion. The resulting mixture was allowed to cool overnight to yield bright yellow crystals of the β-methyl-β-nitrostyrene product, which were collected by filtration, washed with cold ethanol, and air dried to a constant weight yield of 5.0 g (60%). The product was used without further purification.

To a cooled (ice bath) suspension of lithium aluminum hydride (5 g, 107.5 mmol) in tetrahydrofuran (THF) (30 mL) was added dropwise a THF solution (20 mL) of β-methyl-β-nitrostyrene compound 19b (5 g, 18 mmol). After complete addition of the nitrostyrene, the reaction flask and contents were placed under reflux for 3 h, leading to completion of the reaction as revealed by TLC. The reaction mixture was further allowed to cool to room temperature while undergoing continuous stirring. A saturated aqueous solution of sodium chloride (100 mL) was then added slowly dropwise to the reaction mixture with vigorous stirring and cooling in ice over a period of 2 h in order to deactivate any unreacted LiAlH₄. Thereafter, the aqueous phase was decanted and set aside, while the residue was washed several times with water and filtered through celite. The aqueous extracts were pooled and subsequently extracted with ethyl acetate (30 mL five times). Finally, the combined organic extracts were concentrated under reduced pressure to yield the product as a yellow viscous oil with a yield of 3.05 g (87%). The product was used in the next step without further purification.

A mixture of 5-chloroisatin (1 g, 5.5 mmol) and the crude α-methyl-phenethylamine compound 20b (1.3 g, 6.6 mmol,) in polyphosphoric acid (2 g) in a Pyrex beaker was heated in an oil bath at 100°C for ~8 h. Analysis of the reaction mixture by TLC (hexane-ethyl acetate at 60:40) revealed two products and no starting material. The mixture was allowed to cool to room temperature, diluted with distilled water (50 mL), adjusted to pH 10 by the addition of a saturated solution of sodium carbonate, and finally extracted with ethyl acetate (20 mL five times). The combined organic extracts were dried over anhydrous sodium sulfate and concentrated under reduced pressure. The resulting crude mixture was separated by column chromatography on silica gel using stepwise gradient elution with ethyl acetate and hexane (0 to 50% in increments of 5%) to provide two compounds. The more mobile compound was designated compound 11, while the less mobile compound was designated compound 12.

Analytical data for compound 11 are as follows. Yield, 0.3 g, 12% (white solid). ¹H NMR (DMSO-d₆, 500 MHz): δ ppm 1.07 (d, J = 6.3 Hz, 3H, -N-2′C-3′HCH₃), 2.56 to 2.55 (m, 1H, H5′a), 2.66 (dd, J = 12.7 Hz, 1H, H5′b), 3.51 (s, 3H, 7′-OCH₃), 3.72 (s, 3H, 6′-OCH₃), 3.95 to 3.83 (m, 1H, H-3′), 5.85 (s, 1H, H-8′), 6.71 (s, 1H, H-5′), 6.88 (d, J = 8.2 Hz, 1H, H-7), 6.96 (d, J = 2.2 Hz, 1H, H-5), 7.26 (dd, J = 8.3, 2.2 Hz, 1H, H-6), 10.29 (s, 1H, H-1). ¹³C NMR (DMSO-d₆, 100 MHz): δ ppm 22.2 (-CCH3), 36.9 (C-5′), 53.5 (C-3′), 55.9 (7′-OCH₃), 56.2 (6′-OCH₃), 65.9 (C-3/C-1′), 109.5 (C-8′), 111.3 (C-7), 112.8 (C-5′), 121.2 (C-3a), 125.1 (C-5), 126.2 (C-8′a), 129.0 (C-6), 129.9 (C-5′a), 137.8 (C-5), 121.8 (C-7a), 127.5 (C-7′), 128.6 (C-6′), 172.5 (C-2). FTMS + cESI: m/z 359.29 [M + 1]⁺.

Analytical data for compound 12 are as follows. Yield, 0.55 g, 27% (white solid). ¹H NMR (DMSO-d₆, 500 MHz): δ ppm 1.12 (d, J = 6.2 Hz, 3H, -N-2′C-3′HCH₃), 2.58 to 2.39 (m, 1H, H-5′a), 2.77 (dd, J = 16.0 Hz, 1H, H-5′b), 3.38 (m, 1H, H-3′), 3.55 (s, 3H, 7′-OCH₃), 3.72 (s, 3H, 6′-OCH₃), 5.93 (s, 1H, H-8′), 6.72 (s, 1H, H-5′), 6.93 (d, J = 8.3 Hz, 1H, H-7), 7.13 (d, J = 2.1 Hz, 1H, H-5), 7.25 (dd, J = 8.3, 2.2 Hz, 1H, H-6), 10.65 (s, 1H, H-1). ¹³C NMR (DMSO-d₆, 100 MHz): δ ppm 22.5 (-CCH₃), 36.6 (C-5′), 55.0 (C-3′), 55.9 (7′-OCH₃), 56.1 (6′-OCH₃), 66.2 (C-3/C-1′), 109.0 (C-8′), 111.9 (C-7), 112.7 (C-5′), 125.7 (C-5), 126.0 (C-3a), 126.2 (C-8′a), 128.6 (C-6), 129.3 (C-5′a), 139.2 (C-5), 120.6 (C-7a), 127.7 (C-7′), 128.7 (C-6′), 179.3 (C-2). FTMS + cESI: m/z 359.29 [M + 1]⁺.

X-ray crystallography.

For details on X-ray crystallography, see the supplemental material.

Biological evaluation.

(i) Parasite strains and culture conditions. Plasmodium falciparum 3D7 (chloroquine-sensitive) and Dd2 (multidrug-resistant) strains were obtained from the Biodefense and Emerging Infections (BEI) Research Resources (Manassas, VA) and maintained using a modified Trager and Jensen method (25). Briefly, parasites were grown in fresh O⁺ human red blood cells (RBCs) at 3% (vol/vol) hematocrit in complete RPMI 1640 medium containing GlutaMAX and NaHCO₃ and supplemented with 25 mM HEPES, 0.5% Albumax II, 1× hypoxanthine-thymidine, and 20 μg/mL gentamicin. Parasite cultures were incubated at 37°C in a humidified atmosphere with 5% CO₂. Spent culture media were changed daily, and the parasitemia was maintained at <5% by regular partial replacement of the culture with equivalent amounts of fresh uninfected RBCs. Giemsa-stained thin blood smears were examined microscopically under oil immersion to quantify parasitemia and observe parasite morphology. When needed, parasites were synchronized at the ring stage by sorbitol (5%) treatment (26) and further cultivated through one complete cycle (58 h) prior to drug activity studies.

(ii) Compound screening and confirmation analysis. All compounds were submitted as hydrochloride salts and later solubilized in dimethyl sulfoxide (DMSO) prior to storage at −80°C until needed. Compound screening was carried out in 96-well microtitration plates (Thermo Fisher Scientific) using the SYBR green I-based fluorescence method as described by Smilkstein et al. (27). In principle, the dye intercalates between double-stranded DNA bases, producing over a 1,000-fold increase in fluorescence emission when appropriately excited. Given that erythrocytes are enucleated and lack DNA, the fluorescence produced is proportional to parasite density and DNA content. Solutions of the compounds in DMSO were diluted in RPMI 1640 and cocultured with parasites (1% parasitemia and 1.5% hematocrit) in 96-well plates. The final drug concentrations for primary screening were 10 μM and 10 to 0.078 μM for the dose-response analyses, and the final DMSO concentration was 0.05% in each culture well. Artemisinin and chloroquine (Sigma-Aldrich) at 1 μM were used as positive drug controls, while the solvent-treated culture (0.05% DMSO) was used as negative drug control. The plates were incubated at 37°C in a humidified atmosphere with 5% CO₂ for 72 h. Thereafter, parasite growth was assessed by a 2-fold dilution of SYBR green lysis buffer and treated parasite cultures. In brief, 80 μL of parasitized erythrocytes was transferred to dark plates and 40 μL of SYBR green lysis buffer added. Plates were incubated in the dark for 30 min, and fluorescence was measured using a Fluoroskan Ascent multiwell plate reader with excitation and emission wavelengths at 485 and 538 nm, respectively. Mean half-maximal inhibitory concentrations (IC₅₀) were derived by plotting the percentage of growth inhibition against the log drug concentration and fitting the response data to a variable-slope sigmoidal curve-fit function using GraphPad Prism v.8.0.

(iii) In vitro cytotoxicity studies of bioactive compounds. Viability of the cells was assessed by the XTT [(2,3-bis-(2-methoxy-5-nitro-5-sulfophenyl)-2H-tetrazolium-5-carboxanilide)]-based cell proliferation assay (X12223) according to the manufacturer’s instructions. In principle, soluble XTT is reduced to an orange soluble formazan product by actively respiring cells. Human liver hepatocellular epithelial cells were maintained in MEM supplemented with 10% fetal bovine serum (FBS), 1 mM sodium pyruvate, and 1% penicillin-streptomycin.

Confluent cells were trypsinized and seeded at a density of 10,000 cells/well (total volume of 90 μL) in 96-well plates and then incubated for 24 h prior to drug treatment. Two-fold serial dilutions of the compounds (100 to 0.0588 μM) were added to the plates and incubated under humidified conditions at 37°C for 48 h. Absorbance of the formed formazan product was measured at a 450-nm wavelength using a Multiskan FC microtiter plate reader. Dose-response curves were plotted using GraphPad Prism v.8, and CC₅₀ values were obtained. Selectivity indices (SI = CC₅₀/IC₅₀) were calculated as an indication of toxicity relative to the observed antiplasmodial activity.

(iv) Activity against different developmental stages of P. falciparum parasites. Stage-specific activity of active compounds or reference drugs on parasite development were determined by quantitative light microscopy as previously described (9, 28). Briefly, synchronized cultures (1.5% hematocrit and 3% parasitemia) at different time points (early rings, early trophozoites, and early schizonts) were treated with either test compound or controls at in vitro maximum-effect concentrations (~10 μM final concentration) for 24 h under regular culture conditions. Giemsa-stained thin smears were prepared posttreatment, and parasites were counted in 1,000 erythrocytes per treatment. Stage proportions in test wells relative to solvent-treated control wells were calculated and used to assess the in vitro effects of each compound on trophozoites and schizont development as well as merozoite egress and invasion.

(v) Effect of compounds on hemoglobin metabolism. Drug effect on hemoglobin uptake or degradation was assessed by previously described methods (9, 20, 21) with some modifications. Briefly, ring-stage parasites (1.5% hematocrit, 5% parasitemia) were cocultured with test compounds or controls (0.05% DMSO, chloroquine, artemisinin, KAE609 and E64) at a final concentration corresponding either to their respective IC₅₀ values or at a maximum-effect concentration of 5× IC₅₀ or 10 μM. The cultures were incubated at 37°C and 5% CO₂ for 24 h. Following incubation, inhibitors were removed by centrifugation at 1,800 rpm for 5 min, and pellets washed in equal volumes of 1×phosphate-buffered saline (PBS) and then resuspended in the same volume of 0.1% saponin (in 1× PBS) for 3 min. Isolated parasites were pelleted at 3,000 rpm for 10 min and washed twice in an equal volume of 1× PBS. The resulting parasite pellets were permeabilized by treatment in 20 μL of 1% Triton X-100 (in 1× PBS) for 5 min followed by centrifugation at 10,000 rpm for 2 min at 4°C. Intraparasite hemoglobin content in the cleared supernatant was measured at 410 nm using a Nanodrop spectrophotometer, and the fold increase in hemoglobin content was calculated relative to the untreated controls.

In vivo efficacy studies in mice.

(i) Mouse maintenance and infection. Eight-week-old male Swiss albino mice weighing ~23 ± 3 g were housed in cages and maintained in a well-ventilated room under standard environmental conditions of temperature at 22 to 25°C, under a 12-h dark-light cycle, with food and water provided ad libitum.

The animals were allowed 1 week of acclimatization before commencement of the study. Three mice were infected with stocked blood containing Plasmodium chabaudi subsp. chabaudi parasites and used as donors. In brief, frozen blood containing the sulfadoxine/pyrimethamine-resistant P. chabaudi subsp. chabaudi [AS(50S/P)] parasites was thawed, about 200 μL of parasites was injected intraperitoneally (i.p.) into each of the three mice, and parasitemia monitored through blood smear until a 20% threshold was reached. These infected animals then served as donors to the experimental animals. Ethical approval for the study was obtained from the University of Douala Institutional Review Board (no. IEC-UD/1126/09/2017/A), Douala, Cameroon.

(ii) Evaluation of chemosuppressive activity of spiroquindolones. In vivo efficacies were conducted following a modification of the Peter’s 4-day suppressive test as previously described (28–30). Briefly, following parasite passage in donor mice, the mice were euthanized using diethyl ether and P. chabaudi subsp. chabaudi-infected blood was obtained by cardiac puncture and placed into heparinized tubes. The blood was then diluted with phosphate-buffered saline (PBS) and immediately used to infect the experimental mice. Mice were randomly divided into seven groups of 3 animals, and each mouse was injected with 1 × 10⁷ infected erythrocytes intraperitoneally (i.p.) (day 0). The mice were kept for 24 h to establish infection, and Giemsa-stained thin blood smears were prepared to quantify parasites (day 1). Thereafter, groups 1, 2, 3, 4, and 5 were treated i.p. with compounds dissolved in 1× PBS at dosages of 30, 10, 3, 0.3, and 0.03 mg/kg of body weight, respectively, while groups 6 and 7 were treated with vehicle or chloroquine at 10 mg/kg, respectively, for three consecutive days (days 1 to 3). At 24 h and 96 h postinfection, thin blood smears were prepared from each animal with blood obtained from the tail vein, fixed in methanol, and stained with 10% Giemsa.

Parasitemia was determined by light microscopy using a 100× lens objective and the following equation: % parasitemia = (no. of parasitized RBCs/total no. of RBCs counted) × 100. The average percentage of chemosuppression was calculated as 100 × [(A − B)/A], where A is mean parasitemia in negative control and B is mean parasitemia in test group.

Compound ED₅₀ (dose resulting in a 50% reduction in parasitemia) and ED₉₀ (dose resulting in a 90% reduction in parasitemia) were further calculated using the online tool ED₅₀ Calculator (https://www.aatbio.com/tools/ed50-calculator).

(iii) Evaluation of curative potential of spiroquindolones in mice. The curative potential of each selected compound was evaluated using a modification of the Rane’s test (31, 32). A standard inoculum of 10⁷ infected erythrocytes per mouse was injected i.p. A group of seven mice were left uninfected for the duration of the study in order to monitor for any behavioral changes due to infection with the parasite or treatment with the experimental compounds. Seventy-two hours later, the mice were randomly distributed into respective groups and dosed accordingly once daily i.p. for 5 days. The experimental compounds were dosed at their calculated ED₅₀ and ED₉₀ values, whereas the treated controls were dosed with chloroquine at 10 mg/kg body weight. A Giemsa-stained thin blood smear was prepared from the tail blood of each infected mouse on specific days up to day 29 postinfection to monitor the effect of treatment on blood parasitemia. The survival time for each group of mice was determined by calculating the average survival time (days) of the mice over the 30-day study duration (days 0 to 29).

Supporting information.

Additional experimental details on synthesized compounds, NMR data, details of X-ray crystallography acquisition parameters and related data, and HPLC-MS traces of all compounds reported are provided in the supplemental material.