Background
Between 2000 and 2015, malaria incidence rates fell by 37% and malaria mortality by 60%. These achievements can be largely attributed to the implementation of the artemisinin (ART)-based combination therapy (ACT) as first-line treatment of
Plasmodium falciparum malaria in endemic countries [
1]. Artemisinin and its semisynthetic derivatives (ARTs), e.g. artesunate (ATN) and dihydroartemisinin (DHA), display a much better pharmacologic profile than other anti-malarials in use and nowadays world-wide
P. falciparum malaria treatment relies on ACT, although resistance to ARTs is now evident in several Southeast Asian countries [
2‐
5]. Currently, reports of clinical failure of ARTs in areas coincident with partner drug resistance are increasing [
6,
7]. Resistance to ARTs has been associated with mutations in the Kelch protein, K13 (PF3D7_1343700) and manifested as delayed parasite clearance [
8].
Another relevant issue related to the widespread application of artemisinins (ARTs) is the difficulty in maintaining the drug supply; artemisinin is extracted from
Artemisia annua in low yield and ART-derivatives are obtained by a long and expensive semi-synthesis process [
9]. These circumstances pushed efforts towards the development of a next generation of potent anti-malarial endoperoxides, equally effective against ART-susceptible and -resistant strains of
P. falciparum, as well as safer and cheaper than ARTs. Synthetic trioxolanes [
10] and tetraoxanes [
11] are particularly promising in this context, exhibiting anti-malarial activity similar to ARTs. Amenable synthetic routes have been developed to both compound classes, enabling the preparation of chemically diverse libraries of analogues for further structure–activity relationship analysis, selection of leads, optimization and development into anti-malarial drugs or drug-candidates. This approach has yielded ozonides OZ277 (arterolane) [
12] and OZ439 (artefenomel) [
13], but recent studies reported that OZ277 and OZ439 are compromised by the presence of K13 mutations due to potential cross-resistance with DHA [
13‐
15].
In a recent study, three synthetic trioxolanes were tested in vivo against a mouse model infected with artemisinin resistant parasites and the compounds showed high efficacy in clearing the infection [
16]. These results inspired the expansion of the library of compounds and further investigation of their efficacy against artemisinin-resistant (ART-R)
P. falciparum strains.
The synthesis and anti-malarial activity of an endoperoxide-type library of compounds (trioxolanes and tetraoxanes) is reported. These can be easily synthesized from relatively cheap starting materials (preparation of the 21 compounds requires from 3 to 6 synthetic steps, depending on the cyclohexyl substitution).
Discussion
The new peroxide-type compounds tested have shown low or undetectable cytotoxicity against human hepatocellular carcinoma (HepG2) and hamster lung (V79) cell lines. It is generally accepted that, if SI > 10, the observed pharmacological activity is not due to cytotoxicity [
32,
33]. Since the SI values calculated for the compounds are considerably higher, that the activity exhibited by the compounds is unlikely due to general cellular toxicity, but rather due to specific antiplasmodial activity. The eight endoperoxides selected from our library that met the criteria—SI > 100 and IC
50 < 100 nM [
34] were evaluated for in vitro and in vivo efficacy, demonstrating IC
50 values in the range of those exhibited by ARTs and a strong curative effect in vivo, with a parasitaemia reduction on day 10 above 76% (Tables
1,
2). These 8 compounds were particularly active against intra erythrocytic stages, across all four
P. falciparum strains tested, with IC
50 values ranging from 0.3 to 71.1 nM and high selectivity for the parasite. Even though LC136 presented RIs > 1 (up to 2.43), the values fall within the range of those exhibited by ARTs for the same parasite strains (0.62–8.32; Table
1).
Results show that the nature of the cyclohexyl substituent affects antiplasmodial activity. For instance, LC92 exhibited IC
50 values up to 11 times higher than the control drugs (ART derivatives) while its counterpart MIS13 was shown to be as effective as ARTs across the four parasite strains, with no apparent cross resistance. This increased activity probably arises from an improvement in pharmacokinetic properties due to: (i) a more substituted amino functionality that increases the overall hydrophilicity and favours protonation in acidic environments; and (ii) to the BOC-protection of the side chain that facilitates the transport of MIS13 through the cell membrane (compared to LC92) [
35]. Although MIS13 induced a slight cytotoxicity in vitro (30% reduction in survival of HepG2 cells at 1 mM), no adverse effects were observed in vivo. Only 1/5 animals treated with MIS13 presented parasitaemia on day 5, which gradually decreased until 0% by day 10, whereas in the group treated with LC92 all animals presented residual parasitaemia during the 10 days.
Compounds LC131, LC132 and LC136, comprising a tetrazole moiety linked to the trioxolane pharmacophore, exhibited an excellent anti-malarial profile, both in vitro and in vivo. Tetrazole-containing drugs and drug-candidates are thriving in medicinal chemistry [
36,
37] and many are effective therapeutic agents for various pathogenesis, showing a wide range of pharmacological activities, such as antimicrobial, antibacterial, antifungal, and antitubercular [
38‐
41]. The tetrazole moiety has been described as metabolically stable, featuring in the structure of several drugs and drug-candidates, usually as a surrogate for the carboxylic acid functionality, and generally improving the pharmacokinetic profile [
42,
43]. The higher metabolic stability conferred by the tetrazole [
42‐
46] could be improving the pharmacological properties or increasing accumulation (due to the electron-withdrawing properties). In line with this, a better in vivo anti-malarial profile for the tetrazole conjugates LC131, LC132, LC136 and LC163, was observed, particularly for trioxolanes LC131 and LC136, which exhibited a curative effect in mice with total suppression of parasitaemia from day 5 to day 10. To assess the nature of the tetrazole moiety contribution to the anti-malarial activity, the tetrazole building blocks (LC133 and LC126II; Additional file
1) were separately tested and revealed to be devoid of antiplasmodial activity. The ether-linked tetrazolyl–trioxolane conjugate LC132 was up to 12 times less active than LC131 or LC136 and, in vivo, allowed recrudescence on day 10. Probably, this compound is less stable due to the possibility of thermally-induced 1,3-isomerization of the alkyl-trioxolane moiety [
47], as the C(alkyl)-
O bond is relatively weak due to the electron withdrawing effect of the tetrazole heterocycle. Comparatively, the amino-linked conjugates LC131 and LC136 are chemically more stable. LC131 and LC136 exhibited an exceptionally good in vitro anti-malarial profile against both ART-R and ART-S parasites, comparable to that of the control drugs ART, ATN and DHA. The ability of the amino group to protonate in the acidic environment of the food vacuole probably confers additional anti-malarial efficacy by facilitating accumulation, placing the two new compounds as promising drug candidates. In conjugates LC131 and LC136 the heteroaromatic substituent is a 2-substituted 5-aminotetrazole (in LC136 the tetrazole ring exhibits tautomerism, with the 2-
H tautomer being favoured in the gas phase) [
48]. An additional interest of this substitution pattern resides in the potential of 2,5-disubstituted tetrazoles to act as probes. It was demonstrated that, upon narrow-band UV-induced irradiation, 2,5-dissubstituted tetrazoles generate nitrile imine dipolar species in situ, via extrusion of N
2, which may then react with an alkene dipolarophile through a 1,3-dipolar cycloaddition reaction [
49,
50]. Thus, the trioxolane–tetrazole conjugates LC131 and LC136 could also be developed as tags for mechanistic studies.
Compound LC163 is a tetraoxane–tetrazole conjugate structurally related to the trioxolane LC136. The results show that replacement of the trioxolane by a tetraoxane ring, as the pharmacophore, leads to a decreased anti-malarial activity, both in vitro and in vivo. Even though the RIs for LC163 are low, the corresponding IC
50 values for the resistant strains IPC4912 and IPC5202 are, respectively, 3 and 42 times higher than those found for trioxolane LC136. The reason for this difference is still elusive [
51], so detailed structural studies on both compounds are planned.
The saccharyl substituent present in compounds LC129 and LC130 is also electron withdrawing and metabolically stable. Additionally, the saccharyl system is thermally and photochemically more stable than the tetrazole system and is generally found to improve the overall stability of the molecule [
27], although it is worth noting that LC129 is prone to Chapman-type isomerization [
27]. The anti-malarial profile exhibited by the trioxolane–saccharyl conjugates LC129 and LC130 was slightly inferior to that of their tetrazole analogues (Tables
1,
2). As for trioxolane–tetrazoles, the saccharyl building block was tested separately and revealed to be devoid of antiplasmodial activity (Additional file
1). Hence, the anti-malarial activity demonstrated by the trioxolane–saccharyl conjugates can be ascribed to the presence of the trioxolane pharmacophore.
Structural analogies between DHA, trioxolanes and tetraoxanes may accommodate similar modes of action, hence some level of cross resistance [
10,
28‐
31,
52]. In the light of this, it was investigated whether parasites expressing variant forms R539T (IPC502) and I543T (IPC4912) of K13 are cross-resistant to the newly synthesized trioxolanes and tetraoxanes.
ARTs resistance cannot be evidenced by the standard in vitro IC
50 assay [
13,
21,
27‐
31]. Critical for the development of anti-malarials is the evaluation of their activity against ART-R parasites, which has been defined as delayed parasite clearance in patients [
53]. Slow parasite clearance of
P. falciparum malaria in patients results from reduced ring-stage susceptibility [
53‐
56], this increasing the need for new compounds with a low RSA. The mutation R539T (present in IPC5202) is one of the k13 mutations that confer high levels of DHA resistance in vitro [
31,
52,
57]. We explored the susceptibility of the IPC5202 ring and mature-stage parasites (RSA and MSA) to the best performing compounds. As depicted in Fig.
2, the endoperoxides selected demonstrated higher activity than DHA, both against ring and mature stages. As expected, DHA exposure resulted in a higher survival rate of IPC5202 (up to 25%) than for 3D7 [
58,
59], though lower than in other reports [
60]. A survival rate RSA < 1% is generally considered as susceptible behaviour. The tested trioxolanes were able to reduce ring survival to less than 1% in both resistant and susceptible strains (IPC5202 and 3D7), hence no cross-resistance with DHA is apparent, even in parasites carrying the K13 mutation R539T. This notable observation indicates that our compounds have an improved range of activity, compared to other trioxolanes in use, namely the registered drug OZ277, which is compromised by K13 mutations [
31]. Regarding the tetraoxane LC163, both the K13 wild type (3D7) and mutant (IPC5202) parasites presented RSA > 1% (though not significantly different). These observations are in agreement with data recently published showing that a tetraoxane also allowed growth above 1% in RSA assay in a strain carrying K13 R538T mutation [
61]. The results indicate therefore that the anti-malarial activity of the newly synthesized endoperoxides is not compromised by the K13 mutation R539T, a mutation that confers high levels of in vitro resistance and has been associated with delayed parasite clearance in patients [
8,
62‐
65]. The mature-stage assay (MSA) evidenced nearly full susceptibility of the two strains to all of the eight compounds (Fig.
2). Thus, all compounds performed better than DHA. Typically, DHA allows a ≅ 1% of viable mature-stage [
13,
21]. ARTs resistance phenotypes are associated with a decreased susceptibility of the ring stage to enter dormancy, a decreased sensitivity of mature-stage parasites and a faster recovery from dormancy [
66,
67]. Currently these parameters are being addressed in order to gather more information regarding the mode of action of the more promising compounds.
As expected, no cross-resistance of the proposed compounds with quinolone-type anti-malarials is foreseeable, as the calculated RIs for Dd2 and 3D7 were ≅1 (Table
1), which is considerably lower than the RI determined for CQ (21.5) [
20,
68].
Authors’ contributions
LL carried out the in vitro and in vivo anti-malarial evaluation of the tested compounds, participated in the design of the study and drafted the manuscript. LC carried out the synthesis of most of the tested compounds, participated in the design of the study and of the synthetic routes and drafted the manuscript. MIS and BG carried out the synthesis of intermediate compounds and of the target MIS13 and reviewed critically the manuscript. SR participated in the design of the study, and reviewed critically the manuscript. VAN participated in the design of the study, coordinated the in vivo anti-malarial evaluation of the tested compounds, and reviewed critically the manuscript. MLC participated in the conception, design and coordination of the study. Proposed the range of compounds investigated and synthetic routes, coordinated the synthesis, the analysis and interpretation of the results and the manuscript preparation. FN participated in the conception, design and coordination of the study. Coordinated the in vitro anti-malarial evaluation of the tested compounds, the analysis and interpretation of the results and manuscript preparation. All authors read and approved the final manuscript.