Introduction

Malaria is a disease caused by a parasite of genus Plasmodium and is transmitted by a mosquito bite. In 2019, there were 229 million malaria cases with 435,000 associated deaths according to the WHO (WHO 2020). Most of the reported deaths occurred in Africa with pregnant women and children under 5 years of age being among the worst affected. A number of antimalarial drugs have been utilized, but resistance has emerged to all, including the peroxidic semisynthetic artemisinins, currently the primary drugs of choice (Ashley et al. 2014). Thus, there is a continuing need of new clinical agents for the treatment of malaria.

Iron is important for all life forms including the malaria parasite (Smith and Meremikwu (2003). Several metabolic processes rely on iron-containing enzymes such as ribonucleotide reductase which catalyzes the syntheses of deoxyribonucleic acids from ribonucleic acids (Munro and Silva 2012). Other essential enzymes that require iron include dihydroorotate dehydrogenase, which is involved in de novo pyrimidine synthesis (Phillips and Rathod 2010), enzymes in the pentose phosphate pathway (Bozdech and Ginsburg 2005), glycolysis (Bergeron and Brittenham 1993) and for energy transport in mitochondria (Bozdech and Ginsburg 2005). Thus, iron withholding from the malaria parasite could starve the parasite of this essential nutrient and stall related vital processes needed for its survival (Pradines et al. 2002; Thipubon et al. 2015).

Cerebral malaria (CM), which is characterized by neurological impairments, is one of the most severe complications of Plasmodium falciparum infections (Idro et al. 2010). Cognitive deficits in children with CM are more common and persist longer than the physical or neurological complications. Co-administration of chloroquine with antioxidants such as deferoxamine (DFO, 1, Fig. 1) was found to prevent the development of persistent cognitive damage in surrogate models of CM (Reis et al. 2010). DFO is a natural and very effective trihydroxamic acid containing siderophore (microbial iron sequestering agent). It is a microbial virulence factor produced by Streptomyces pilosus (Crichton 2002), that also has been used to treat iron overload diseases under the trade name Desferal®. As an iron chelator, DFO also can withhold iron from Plasmodium and thus inhibit essential metabolic processes (vide supra). Additionally, it also can inhibit iron dependent free radical reactions that are responsible for the cognitive impairment associated with CM (Percario et al. 2012).

Fig. 1
figure 1

Structures of natural siderophores deferoxamine (1), ferricrocin (2), and synthetic iron chelator, ICL-670 (3)

Full size image

Gordeuk et al. reported a clinical trial involving 83 Zambian children with CM that included 50 patients already in deep coma and found that administration of DFO with quinine decreased the median recovery time from 68 to 24 h and hastened the clearance of parasitemia (Gordeuk et al. 1992). These remarkable results indicated that iron chelation therapy may enhance recovery from deep coma in CM patients.

Subsequent to the discovery of the antimalarial properties of the semisynthetic artemisinins, drugs based on the natural product artemisinin (qinghaosu) (Chaturvedi et al. 2010), a number of other peroxide derivatives were synthesized with the goal of further elaborating SAR (structure-activity-relationships) and providing reliable sources of synthetically accessible peroxide-containing agents to help circumvent resistance (Borstnik et al. 2002; Tang et al. 2004). 1,2,4-Trioxolanes (ozonides), discovered by Vennerstrom et al. have emerged as frontline antimalarial agents (Vennerstrom et al. 2004). In 2012, OZ277 (arterolane), in combination with piperaquine phosphate was approved in India for the treatment of malaria (Burrows et al. 2014). Other ozonide based compounds are being intensively studied (Shackleford et al. 2021). We have also previously demonstrated that a synthetic conjugate of the antimalarial agent, artemisinin and an analog of mycobactin T, the obligate siderophore for Mycobacterium tuberculosis, has potent anti-TB activity and retains full antimalarial activity associated with the artemisinin component (Miller et al. 2011). Thus, conjugates of other siderophores, their metal complexes and antimalarial ozonides was of considerable interest.

As discussed earlier, iron withholding from Plasmodium by siderophores reportedly results in the enhancement of parasite clearance and their use in combination with quinine prevented cognitive damage in experimental models of malaria after CM infection. Based on the above premise, we set out to test the hypothesis that conjugation of non-iron-complexed siderophores and/or iron-chelators (e.g. DFO, ICL-670) with peroxide-based antimalarials (e.g. ozonides) would preserve the antimalarial activity of the latter but offer the advantage of withholding iron from malaria parasites. While an evaluation of a clinical benefit of such conjugation of siderophores with ozonides is beyond the scope of this work, it was relevant to explore and evaluate whether ozonide conjugates of pre-iron-complexed siderophores (e.g. ferricrocin), possesses any antimalarial activity. Thus, the synthesis and antimalarial evaluation of covalent siderophore—ozonide conjugates were of interest and are described herein.

Results and discussion

The syntheses of three types of siderophore-based ozonide conjugates are shown in Schemes 1, 2 and 3. Preparation of the DFO and ferricrocin conjugates involved coupling of the free carboxylic acid of the ozonides with the free amine of DFO (1, Scheme 1) or the free primary hydroxyl group of ferricrocin (2, Scheme 2). Since the last iron chelator (ICL-670, 3), also contains a carboxylic acid, a diamine spacer was incorporated to allow for conjugate formation with the ozonide carboxylic acids (Scheme 3). Thus, as shown in Scheme 1, the carboxylic acid groups of ozonides OZ72 (4) and OZ78 (5) were converted to the corresponding NHS-active esters (6 and 7, respectively) by reactions of the acids with N-hydroxysuccinimide (NHS) in the presence of EDC. Direct reaction of the NHS active esters with DFO mesylate in the presence of Et3N provided amides 8 and 9 in 27% and 35% yields, respectively.

Scheme 1
scheme 1

Synthesis of DFO-ozonide conjugates 8 and 9

Full size image
Scheme 2
scheme 2

Synthesis of FC-ozonide conjugates, 10 and 11 using Yamaguchi esterification

Full size image
Scheme 3
scheme 3

Synthesis of ICL670-OZ72 conjugate, 13

Full size image

Ferricrocin (2) is another natural siderophore available as the preformed iron (III) complex. It contains a serine in its peptide backbone and the free hydroxyl of the serine was anticipated to be appropriate for esterification with the ozonide carboxylic acids since the modification would be remote from the iron binding area of the siderophore. The complexed iron (III) was also anticipated to protect the constituent hydroxamates during and after the esterification whereas presence of the free hydroxamates was compatible with formation of the amide linkage in the DFO conjugates. Attempted ester formation between ferricrocin and the ozonide carboxylates using the EDC-mediated protocol that worked for the DFO conjugates provided products in very low yield. Alternatively, Yamaguchi esterification (Inanaga et al. 1979) by preformation of the mixed anhydride of the carboxylates with 2,4,6-trichlorobenzoyl chloride followed by reaction with ferricrocin gave the corresponding esters 10 and 11 in 50–55% yield (Scheme 2).

ICL-670 (3) is a synthetic, orally active tridentate iron chelator developed for the treatment of iron-overload diseases (Galanello et al. 2003). Goudeau et al., demonstrated that it is a stronger antimalarial agent than DFO, and the antimalarial effects of both of these compounds are exerted by withholding the iron from the critical Plasmodium targets (Goudeau et al. 2001). As indicated earlier, since both ICL-670 and the ozonides contain carboxylic acid functional groups, we decided to conjugate them using a diamine spacer. As shown in Scheme 3, ICL-670 was first coupled to mono Boc-protected ethylene diamine using EDC to give amide 12 in 75% yield. Boc removal with TFA proceeded uneventfully and the resulting new free amine was then coupled to the ozonide carboxylic acid 4 to give conjugate 13 (Scheme 3).

The ozonide-iron chelator conjugates 8–11 and 13, along with controls (ozonides 4, 5, artemisinin and siderophores 1 and 2), were tested for activity against well-characterized Plasmodium strains that are resistant to standard antimalarial drugs including Dd2 (origin: Indochina, resistant to chloroquine, quinine, pyrimethamine, and sulfadoxine), HB3 (origin: Honduras, resistant to pyrimethamine), FCB (origin: South East Asia, resistant to chloroquine, quinine and cycloguanil) and 7G8 (origin: Brazil, resistant to chloroquine, pyrimethamine, cycloguanil). Additionally, the conjugates were also evaluated against drug sensitive strains such as 3D7 (origin: Africa—country unknown) and GB4 (origin: Ghana) (Preston et al. 2014; Drakeley et al. 1996). More details of the experimental protocols and strains are provided in the Supporting Information. Overall, the results shown in Table 1 indicate that ozonide conjugates 8–11 and 13 with and without preformed iron-complexed siderophores or iron chelators, have potent antimalarial activities (0.01 µM to 0.59 µM), comparable to the parent ozonides 4 and 5 (0.02 µM to 0.1 µM). This may indicate that a strategy to conjugate non-iron complexed siderophores or iron chelators with ozonides holds the potential to offer clinical benefit in CM patients similar to the study by Gordeuk et al. The antimalarial activities of both the iron free (8, 9, and 13) and preformed iron-complexed siderophore conjugates, 10 and 11 was especially gratifying as the results suggest that, as planned, the conjugates are intended to have dual modes of action. While the iron free conjugates will also scavenge detrimental iron, the subsequently formed iron complex will retain its separate ozonide induced antimalarial activity. Overall, the results of our antimalarial structure activity relationship (SAR) study presented for the conjugation of ozonides with representative siderophores and an iron chelator, are encouraging. The results presented here indicate that the rationale of covalently combining the siderophores as antioxidants with antimalarial ozonides for the prevention of cognitive damages in CM patients certainly is promising and needs further research and clinical studies. The work described herein is especially significant because of the potential dual advantage of providing the antioxidant therapy (iron chelation) to prevent neurological damage while retaining potency against malaria (ozonides).

Table 1 In vitro activity (IC50 in µM) of the amide and ester conjugates of siderophores and ozonides against Plasmodium falciparum strains
Full size table