In vitro antiplasmodial activity, pharmacokinetic profiles and interference in isoprenoid pathway of 2-aniline-3-hydroxy-1.4-naphthoquinone derivatives

Malaria still remains a major parasitic disease in the tropical and subtropical regions of the world due to its economic impact and high morbidity. In 2015, according to the latest estimates from the World Health Organization (WHO), there were 214 million new cases resulting in 438,000 deaths, mostly children under 5 years of age [1]. Although several anti-malarials drugs are available, their efficacies are limited by the existence of drug-resistant parasites worldwide (including Brazil), especially in the case of Plasmodium falciparum [2, 3]. In this context, it is needed either the discovery of new drugs associated with new targets or to improve determined anti-malarial drug class.

The antiparasitic activity of hydroxy-naphthoquinones derivatives is already known, with studies confirming its efficacy against Leishmania braziliensis and Leishmania amazonensis [4], Trypanosoma cruzi [5] and P. falciparum [6‐10]. Its mechanism of action was already proposed by means of the inhibition of the mitochondrial electron carrier chain [11].

The biosynthesis of isoprenoids pathway in P. falciparum are excellent therapeutic targets because they are different or even absent in the human host. Their many functions are quite important for the parasite’s survival [12]. Vitamins E and K, belonging to the family of isoprenoid, has an isoprenic chainsaw from MEP pathway. These vitamins are essential components for the cellular machinery found in all organisms. The menaquinone (MQ) (vitamin K₂) is employed as electron carriers required for the mitochondrial respiratory chain [13‐15], with the α-tocopherol (TC) representing more bioactivity of vitamin E, protecting the membranes against peroxidation [16].

According to the Resolution CNS 251/97, pre-clinical research is the first step in the study on the development of new drugs. This approach should provide information for possible therapeutic application, besides to predict some risks such as toxicity and adverse effects [17]. The absorption, distribution, metabolism, and elimination (ADME) characteristics of a drug are conventionally viewed as an important part of the drug development [18]. Unfavourable ADME is the leading cause of costly and clinical failures in the development of new chemical products during the drug development projects [19‐21]. It is believed that around 40–50% of study in clinical phases fail due to toxicity and pharmacokinetic difficulties [18, 22]. Therefore, there is an increasing interest in the early ADME prediction of drug candidates. Besides of drug-likeness and ADME data predictions, it has been demonstrated that the anti-malarial activity and interference with the synthesis of isoprenic precursors of three derivatives of hydroxy-naphthoquinones against P. falciparum.

All physicochemical descriptors and ADME predictions calculated for compound 4a and 4c are shown in Table 1. Figure 1 depicts the PhysChem filters positioning of candidates in (a) Lipinski RO5 [36], (b) Lead-like soft, (c) Drug-like soft, (d) REOS and (e) ZINC rules designed by combining several drugs’ physicochemical parameters. The R05 and Veber rules area obtained with the following descriptors ranges: logP (− 2 to 5), molecular weight (150 to 500), tPSA (20 to 150), rotatable bonds (0 to 10), H-bonds acceptors (0 to 10) and donors (0 to 5). The complexity, permeability/metabolic stability (Golden Triangle Rule), oral absorption estimation and the oral bioavailability evaluation of compounds are showed in Fig. 2.

The IC₅₀ for 4a and 4c were determined by in vitro growth inhibition assay using P. falciparum 3D7 and Dd2 strains. Table 2 and Fig. 3 present the mean IC50 values for 4a, 4c, and chloroquine (reference anti-malarial drug).

Change in MQ and TC biosynthesis resulting from treatment with the hydroxy-naphthoquinone derivatives is shown in Fig. 4. The activity (in c.p.m.) of the fractions corresponding to the retention times of MQ 15 and TC 16 were compared with those of the corresponding fractions from untreated controls. Biosynthesis of MQ and TC were inhibited in 71.2% and 62.7% (58.1% and 49.6%) by derivative 4a (4c).

The physicochemical property filters are used in drug discovery and drug development to narrow down the scope of molecules [37, 38]. They estimate the drug-likeness profile (Druglikeness prediction) of the active compounds by considering their physical and chemical properties [39‐41]. After analysing the physicochemical descriptors values calculated, it was observed that the components 4a and 4c meet the criteria for the Lipinski’s R05, Lead-like soft, Drug-like soft, REOS and ZINC rules [42] (see Table 1 and Fig. 1).

Figure 2 shows that both compounds reach the criteria for the number of system ring, stereocentres, rotatable and rigid bonds, as well as the maximum size of the system rings (number of system ring, stereocentres, rotatable and rigid bonds, the flexibility and the maximum size of system rings) successfully [43]. In this study, compound values are superimposed on a database comprising currently 1790 chemical structures of drugs [44]. By Golden Triangle Rule, both compounds have an optimal permeability (low clearance) and a good metabolic stability. They are believed to be the designed drug from the most ligand and lipophilic efficient lead into the centre of the Golden Triangle and should provide the maximum potency, stability, and permeability [45].

To further estimate the druggability of these naphthoquinones derivatives were report their ADME descriptors and profiles and potential biological activities using an in silico approach are depicted in Table 1. While the likelihood of drug absorption has been described in compliance to the Ro5, Veber rule, human intestinal absorption (HIA), skin permeability (logKp), colorectal carcinoma (Caco2) and Madin-Darby canine kidney (MDCK) cells permeability, the drug distribution was modeled using the blood/brain partition coefficient (BBB) and the coefficient of binding to human serum albumin (PPB).

Human skin has a low permeability for most of the foreign substances, which are unable to penetrate and diffuse through the skin. Pre ADME predicts in vitro skin permeability and the result value is given as logKp (permeability coefficient). In this work, the compounds 4a and 4c show logkp de − 3.59 e − 3.69 respectively, delimited within the scale of − 8 to − 1, common drug 95% [46]. As a comparative test, the water, for example, has predicted ranging from logkp − 7.39 to − 6.27 [47].

The oral absorption evaluation is one of the most influential ADME characteristics in the early stages of lead discovery and optimization [48]. The compounds 4a and 4c are well absorbed in the human intestine, with values for HIA being 93.67% and 93.85% respectively. The permeability coefficient in human intestinal epithelial (Caco-2) and renal (MDCK) cell presented intermediate values for 4a (4c), lying within the recommended ranges of 95% for known drugs [49‐51].

The RO5 and Veber rules comprise a set of rules that attempt to predict if a molecule could be administered orally [52]. As can be seen in Fig. 1, compounds 4a and 4c fail within R05 and Veber rules area obtained with the analyzed descriptors. In addition, the compounds examined successfully meet the proposed criteria for Egan et al. [53] and Lobell et al. [54], which states that for a compound be transcellularly absorbed from the gastrointestinal tract into systemic circulation it must be reasonably soluble in aqueous solution, although should not be too polar (PSA), too lipophilic (CLOGP), too large (MW) or too flexible (rotatable bonds) to pass cellular membranes.

A good oral bioavailability reduces the amount of an administered drug necessary to achieve a desired pharmacological effect and, therefore, could reduce the risk of side-effects and toxicity. A compound positioning within 466 and 916 orally bioavailable compounds extracted from the DrugBank and e-Drug3D databases respectively [43, 55] are presented in a graph obtained by applying the PCA (Principal Component Analysis) of the 15 main physicochemical descriptors of these molecules.

The plasma protein binding (PPB) and blood–brain barrier (BBB) penetration [56, 57] were calculated for detecting its distribution into the human body. The PPB values obtained for 4a and 4c indicate that the chemicals are moderately bound to albumin. An analysis of the PPB’s percentage distribution among some therapeutic drugs showed that chemotherapeutics presenting PPB > 90% binding can be classified as therapeutic drugs [58]. Further, a BBB penetration value determined in this work indicates a middle (low) absorption to central nervous system-CNS by studied molecules [59]. This is crucial in the pharmaceutical sphere because CNS-active compounds must pass across it, while CNS-inactive compounds do not, in order to avoid any CNS side effects.

According to the System Biopharmaceutics Classification (BCS), which ranks drugs based on their aqueous solubility and intestinal permeability through the correlation between in vitro dissolution and bioavailability of the drug in vivo [60, 61], atovaquone is classified in class II, which includes drugs with low solubility, having high permeability and good absorption [62].

Atovaquone, as well as the tested compounds, is a 3-substituted-2-hydroxy-1.4-naphthoquinone complex. It presents an excellent anti-malarial activity but has poor pharmacological properties, such as low bioavailability and high binding to plasma proteins due to its lipid solubility [63]. To improve its bioavailability, some analogs of atovaquone were created whose changes were made in the naftoquinoidal group, especially the alkyl side chain. It is known that modification of this chain can alter the activity of the drug [64]. The 4a and 4c compounds tested in vivo [9] presented interesting antiplasmodial activity, using the murine Plasmodium berghei model. The output of these experiments may result in low bioavailability and/or high plasma protein binding, and atovaquone. The low bioavailability of atovaquone, and probably of our compounds, is due to the limited solubility in aqueous media, although conferring good intestinal permeability. Tests that include solubility enhancement techniques through physical changes, chemical or by various methods has been not only already studied, but also be employed, since approximately 70% of new drug candidates and more than 40% of the newly developed chemical formulas in the pharmaceutical industry are practically insoluble in water [65, 66].

The antiplasmodial in vitro activity of derivatives hydroxy-naphthoquinones was satisfactory against the 3D7 strain (sensitive chloroquine), being classified as high activity anti-malarial compounds. For the Dd2 strain (multidrug resistant), the IC₅₀ values were fivefold higher; being classified as moderately active against this strain.

In a study which tested over 10 modified molecules from hydroxy-naphthoquinones, the IC₅₀ values were greater than 50.0 μM [8]. Another study tested five derivatives hydroxy-naphthoquinone with radical aquil, and only in one of the derivatives, whose radical was a phenyl grouping, was very active in the nanomolar order of magnitude [67]. Derivatives of 1.4-naphthoquinones, which was a radical trifluormethylbenzene, showed IC₅₀ in the nanomolar order of magnitude. This high anti-malarial activity is probably due to the instability of the radical trifluormethylbenzene, which leads to the formation of free radicals that are toxic to the parasite [11]. However, a study with three derivatives hydroxy-naphthoquinone aminated, showed that only one was effective against the P. falciparum NF54; strain sensitive to most anti-malarials [7]. Their radical was a phenylpiperazine on carbon 6, with the same root added to other lead compounds. This demonstrates antiplasmodial effectiveness in these other changes, confirming a relationship with this radical activity. The phenylpiperazine is attached to a piperazine phenylaniline, showing similarity with radicals used in structural changes, as in the present study, which is aniline derivatives and associated with the effectiveness of that radical. In another study, eight derivatives of 2-hydroxy-3-methylamino-1.4-naphthoquinone were tested, and five were active, exhibiting IC₅₀ values less than 30 μM against P. falciparum W2 (chloroquine-resistant strain) [10].

Pyrimethamine, an anti-malarial already used in combination with sulfadoxine, has a higher IC₅₀ when compared to present work and narrow therapeutic window [68], confirming the importance of obtained results. Several studies have shown that the structural changes in the molecule have reduced its cytotoxic effect while improving either the biological activity [69, 70] or the pharmacokinetic and pharmacodynamic properties [69].

Concerning the likely target of the derivatives of hydroxy-naphthoquinones tested, the results obtained in this study suggest the biosynthesis of isoprenoids as a possible target, since the two products of this pathway had a significantly decreased production: the vitamin K₂ (MQ) that act as an important electron receiver of the respiratory chain [34] and the α-tocopherol that protects membranes from lipid peroxidation [32]. The α-tocopherol also acts as an antioxidant and free-radical scavenger, can efficiently quench O₂ and scavenge various radicals released during oxidative stress [71]. A candidate for an anti-malarial drug that acts simultaneously in the synthesis of these two molecules is interesting because it can trick the parasite resistance mechanism, both interfering in the cellular respiration when in plasma membrane integrity. Furthermore, due to its absence in humans, the MEP pathway can be considered as an important drug target for malaria, although additional studies should be performed to confirm which enzyme is inhibited, after the addition of isopentenyl-PP [12] or geranylgeranyl pyrophosphate synthase (GGPP) to the cultures [72]. Due to their structural similarity with atovaquone, these novel compounds may probably act through a similar mechanism. Atovaquone is believed to inhibit electron flow in aerobic respiration, by binding to cytochrome b in place of ubiquinone in the parasite mitochondria. Such inhibition hampers the activity of dihydroorotate dehydrogenase, an enzyme linked to the mitochondrial electron transport system that is required in the de novo synthesis of pyrimidines [73]. Additionally, the redox potential of hydroxy-naphthoquinones is suggested to cause an oxidative stress which may also be responsible for some antiparasitic activity of this class of compounds. However, one recent study found that 3-[4-(trifluoromethyl)benzyl]-menadione do not inhibit the mitochondrial electron transport chain, but exert their anti-malarial activity as redox-active subversive substrates [11].

The compound 4c presents better in vitro results, since it has the lowest IC₅₀ and interfered in the biosynthesis of the two products of the isoprenoid pathway. In this sense, it provides a better activity of the compound, preventing escape mechanisms by the host avoiding the oxidative stress process, despite as much as tocopherol acts as an antioxidant. In silico computations have been carried out for the calculation of the physicochemical and subsequent drug-likeness and ADME predictions. Satisfying a number of rules and physicochemical filter (see “Methods”), the compounds present acceptable drug-like profile, specifically structural alerts, good oral bioavailability, permeability, and distribution. Modifications in the structure of compound 4c as the salt formation and complexation with β-cyclodextrin can be interesting because solubility enhancing techniques can promote the efficacy of the compound leading to an increase in bioavailability as a consequence of improved solubility.

Thus, the molecules tested in the present study, although presenting lower anti-malarial activity than some of the usual class of anti-malarials, could be used as prototypes for future programmes that could develop safe and effective drugs using their pharmacophores with those of other drugs into synergistic conjugates.