Antiplasmodial Activity Evaluation of a Bestatin-Related Aminopeptidase Inhibitor, Phebestin

ABSTRACT

The increasing burden and spread of resistant malaria parasites remains an immense burden to public health. These factors have driven the demand to search for a new therapeutic agent. From our screening, phebestin stood out with nanomolar efficacy against Plasmodium falciparum 3D7. Phebestin was initially identified as an aminopeptidase N inhibitor. Phebestin inhibited the in vitro multiplication of the P. falciparum 3D7 (chloroquine-sensitive) and K1 (chloroquine-resistant) strains at IC₅₀ values of 157.90 ± 6.26 nM and 268.17 ± 67.59 nM, respectively. Furthermore, phebestin exhibited no cytotoxic against human foreskin fibroblast cells at 2.5 mM. In the stage-specific assay, phebestin inhibited all parasite stages at 100 and 10-fold its IC₅₀ concentration. Using 72-h in vitro exposure of phebestin at concentrations of 1 μM on P. falciparum 3D7 distorted the parasite morphology, showed dying signs, shrank, and prevented reinvasion of RBCs, even after the compound was washed from the culture. An in silico study found that phebestin binds to P. falciparum M1 alanyl aminopeptidase (PfM1AAP) and M17 leucyl aminopeptidase (PfM17LAP), as observed for bestatin. In vivo evaluation using P. yoelii 17XNL-infected mice with administrations of 20 mg/kg phebestin, once daily for 7 days, resulted in significantly lower parasitemia peaks in the phebestin-treated group (19.53%) than in the untreated group (29.55%). At the same dose and treatment, P. berghei ANKA-infected mice showed reduced parasitemia levels and improved survival compared to untreated mice. These results indicate that phebestin is a promising candidate for development as a potential therapeutic agent against malaria.

INTRODUCTION

Malaria is a persistent life-threatening disease caused by Plasmodium parasites, with 241 million cases occurring in 2020 globally (1). Female Anopheles mosquitoes spread Plasmodium parasites during blood feeding on humans and develop complex life cycles in both humans (the hosts) and mosquitoes (the vectors) (2). Even though antimalarial medications and vector control measures have helped reduce the malaria burden over the last decade, malaria still affects half of the world’s population, and the link between efficacy and transmission prevention is not always clear (3). Additionally, commercially available antimalarial medications are susceptible to parasite chemoresistance, including the gold-standard artemisinin combination therapy, which poses a serious concern (4). Therefore, new targets and pathways vulnerable to chemotherapy must be continuously investigated to produce the next generation of antimalarial medicines.

The blood stages of Plasmodium spp. are the main targets of most antimalarial drugs because these stages are responsible for the clinical symptoms of malaria and because of the significant parasite population amplification by recurrent rounds of invasion, growth, division, and egress from erythrocytes (2, 5). In this stage, peptidases have essential roles in the degradation of hemoglobin, which provides amino acids for parasite protein synthesis as well as metabolic energy; intervention in this process could be an ideal target for chemotherapeutic strategies (6). Metalloaminopeptidase enzymes (MAPs) are exopeptidases that catalyze the hydrolysis and cleavage of a single N-terminal amino acid from a peptide substrate (7). Two MAPs of the P. falciparum M1 alanine aminopeptidase (PfM1AAP) and the P. falciparum M17 leucine aminopeptidase (PFM17LAP) were identified as targets of the natural aminopeptidase inhibitor Phe-Leu dipeptide analog, bestatin, which exhibited killing properties against P. falciparum in vitro culture and was also effective against P. chabaudi in vivo (6, 8). Based on our screening of a compound library from the Institute of Microbial Chemistry (https://www.bikaken.or.jp/research/compounds/index.php, Table S1), one bestatin structurally related compound, phebestin, exhibited in vitro inhibitory activity against P. falciparum 3D7 at the nanomolar scale.

Phebestin is an aminopeptidase N (CD13 or alanyl aminopeptidase) inhibitor isolated from Streptomyces sp. MJ716-m3 in 1996 (9). Its structure resembles bestatin ([2S,3R]-3-amino-2-hydroxy-4-phenylbutanoyl)-l-leucine with the addition of one phenylalanine moiety and one less methylene moiety in the side chain ([2S,3R]-3-amino-2-hydroxy-4-phenylbutanoyl)-l-valyl-l-phenylalanine (Fig. 1). The bestatin scaffold was used to exploit the common catalytic mechanism by coordinating one or two Zn ions in the active site of MAPs (7). Considering the P. falciparum 3D7 and aminopeptidase inhibitory activities along with the bestatin scaffold of phebestin, we investigated the in vitro inhibitory activity of phebestin, against the deadliest human malaria, P. falciparum, both for chloroquine-sensitive (3D7) and multidrug-resistant (K1, resistant to chloroquine, sulfadoxine, pyrimethamine, and cycloguanil [10]) strains. Furthermore, the in vitro inhibitory activities were then confirmed by in silico docking simulations and in vivo inhibitory evaluations against P. yoelii 17XNL (nonlethal malaria model) and P. berghei ANKA (lethal malaria model) (11).

FIG 1.

The structures of (A) phebestin and (B) bestatin.

RESULTS

Phebestin is more potent against P. falciparum in vitro than bestatin.

The antiplasmodial activities of phebestin in vitro, along with bestatin, artemisinin, and chloroquine, were determined using P. falciparum 3D7 and K1. The data are presented in Table 1. Phebestin inhibited the growth of both P. falciparum 3D7 and K1 with IC₅₀ values of 157.90 ± 6.26 and 268.17 ± 67.59 nM, respectively. The cytotoxicity of phebestin against HFFs was not detected at 2,500 μM, resulting in predicted selectivity indices (SIs) of >15,832.8 (3D7) and >9,322.44 (K1). Furthermore, at a concentration of 100 μM, phebestin exhibited a low erythrocyte hemolysis rate of 0.90 ± 0.79%.

TABLE 1.

Summary of the in vitro activities of phebestin compared with the positive controls^a

Compound	IC50 P. falciparum (nM)		Resistance index	CC50 HFF (μM)	SI for P. falciparum		RBC hemolysis rate (%) at 100 μM
	3D7	K1			3D7	K1
Phebestin	157.90 ± 6.26	268.17 ± 67.59	1.69	>2500	>15,832.80	>9,322.44	0.90 ± 0.79
Bestatin	3220 ± 168.00	4795.67 ± 424.82	1.49	>2500	>776.39	>521.30	1.74 ± 0.27
Artemisinin (31)	13.18 ± 2.66	19.89 ± 1.51	1.4	153.00 ± 30.76	10,983.49	7692.31	1.03 ± 0.46
Chloroquine (31)	26.20 ± 3.66	740.07 ± 95.67	28.24	20.71 ± 6.80	790.46	27.98	0.71 ± 0.35

^aValues are presented as the average ± SD of three independent experiments. IC₅₀, half-maximal inhibitory concentration. CC₅₀, half-maximal cytotoxic concentration. Resistance index, the ratio between the IC₅₀ values of P. falciparum K1 and 3D7. SI, selectivity index (ratio between IC₅₀ and CC₅₀). HFF, human foreskin fibroblast; RBC, red blood cell.

Effects of phebestin and bestatin on the morphology and parasitemia level of P. falciparum 3D7 in vitro.

Microscopy observation of thin blood smears of P. falciparum 3D7 in the presence of phebestin and bestatin was performed to investigate the effects of the compounds on morphological changes and parasitemia levels during the incubation periods. The first observation investigated the stage-specific action of the tested compounds. In the ring, trophozoite, or schizont stages. In treatments of all compounds using concentrations of 100 and 10-fold IC₅₀, parasites were clearly arrested in the progression to the next cycle stages, compared to no treatment group (Fig. 2). The treatments of artemisinin and chloroquine caused the shrinkage of all parasite stages; meanwhile, phebestin and bestatin caused hollow-like phenotypes for ring and trophozoite stages and shrinkage phenotype for the schizont stage. Furthermore, the treatments with 1-fold IC₅₀ caused only partial inhibition of the life cycle stages, except for artemisinin which still showed high inhibition of the parasite’s progressions (Fig. 2B).

FIG 2.

(A) Representative morphologies of parasites after incubation for 24 h under no treatment (NT), artemisinin (ART), chloroquine (CHQ), phebestin (PHE), and bestatin (BES), at 100, 10, and 1-fold their respective IC₅₀. Each compound was added at ring, trophozoite, or schizont stage. Scale bar: 5 μm. (B) The percentage of drug-affected parasites, including abnormal and retained morphologies, after 24 h exposure of ART, CHQ, PHE, and BES, at 100, 10, and 1-fold their respective IC₅₀. Percentage levels are the average of triplicate tests; the error bars represent standard deviations.

The second observation was to investigate the effect of phebestin or bestatin exposures on the parasite’s morphology and parasitemia level, during and after 72 h of incubation, with continued incubation up to 144 h with compound (unwashed group) or without compound (washed group). The life cycles of parasites were retained in the trophozoite stage during 24 to 72 h of incubation with 1 μM phebestin (Fig. 3A). Phebestin treatment at 1 μM for 72 h induced morphological alterations, including shrinkage of trophozoites. The parasitemia levels during this experiment were also not increased during the incubation period (Fig. 3B) compared to control and bestatin treatment, indicating that reinvasion of the erythrocytes did not occur in the treatment of phebestin. The parasite’s growth inhibition was continued in both the unwashed and washed groups after 72 h of phebestin exposure in the culture. After this period, the parasites were continuously observed as shrunken dot-like spots within erythrocytes (Fig. 4A), and there was no increase in parasitemia levels (Fig. 4B). In the case of bestatin, even though the suppression was significantly lower than the control, the parasites appeared in typical morphologies with increasing parasitemia levels, indicating that bestatin at a concentration of 1 μM showed inhibition activity and live cycle delay but did not prevent the reinvasion of erythrocytes.

FIG 3.

Representative morphologies and parasitemia levels of parasites under no treatment (control), phebestin (1 μM), and bestatin (1 μM). (A) Parasite morphology after incubation for 1, 24, 48, and 72 h. Scale bar: 5 μm. (B) Parasitemia levels after incubation for 1, 24, 48, and 72 h. Parasitemia levels are the average of triplicate tests; the error bars represent standard deviations. *, significant (P < 0.05). The significance of differences in the level of parasitemia of the phebestin- and bestatin-treated cultures compared with the control was analyzed by two-way ANOVA with Tukey’s multiple-comparison.

FIG 4.

Representative morphologies and parasitemia levels of parasites after 72 h of growth exposure to no treatment (control), phebestin (1 μM), and bestatin (1 μM). (A) Parasite morphologies after 96, 120, and 144 h of incubation in the presence (unwashed) or absence of phebestin and bestatin (washed). Scale bar: 5 μm. (B) Parasitemia levels after 96, 120, and 144 h of incubation in the presence (unwashed) or absence of phebestin and bestatin (washed). Parasitemia levels are presented as the average of triplicate tests, and the error bars represent standard deviations. The significance of differences (*, significant, P < 0.05) in the level of parasitemia of phebestin- and bestatin-treated cultures compared with the control was analyzed by two-way ANOVA and Tukey’s multiple-comparison test.

Docking simulations and drug-likeness predictions of phebestin.

An in silico study was conducted to assess the possible interaction of phebestin with PfM1AAP and PfM17LAP. Bestatin, a validated inhibitor of both proteins (7), was used as a positive control. The interactions of phebestin and bestatin from docking simulation are shown in Table 2. Phebestin exhibited a higher free energy of binding (binding affinity) to both receptors than bestatin, as well as convincing CNN scores. Phebestin shared the same interactions with bestatin by coordinating their α-hydroxy group to Zn ion(s) in the active site of both receptors. The receptor–ligand interactions within 4 Å of the ligand with the lowest binding affinity are shown in Fig. 5 for PfM1AAP and Fig. 6 for PfM17LAP.

TABLE 2.

Interactions of phebestin and bestatin from docking simulations^a

Receptor	Ligand
	Phebestin	Bestatin
PfM1AAP (3T8V)
Minimized affinity	−9.698 kcal/mol	−9.163 kcal/mol
CNN score	0.9835	0.9909
Ion complex	Zn	Zn
Hydrogen bond interaction	Glu319, Gly460, Glu463, Arg489, Glu497, Glu519, Tyr580	Glu319, Gly460, Glu463, Arg489, Glu497, Glu519, Tyr580
Hydrophobic interaction	Glu319, Val459, Gly460, Met462, Val493, Tyr575, Tyr580	Glu319, Val459, Gly460, Met462, Val493, Tyr575, Tyr580
PfM17LAP (3T8W)
Minimized affinity	−8.808 kcal/mol	−8.555 kcal/mol
CNN score	0.9928	0.9935
Ion complex	Zn-1, Zn-2	Zn-1, Zn-2
Hydrogen bond interaction	Lys374, Asp379, Lys386, Glu461, Leu487, Ala490, Gly489, Ser554	Lys374, Asp379, Lys386, Arg463, Leu487, Gly489,
Hydrophobic interaction	Met392, Met396, Phe398, Ala460, Thr486, Leu487, Thr488, Gly489, Tyr493, Ala577	Met392, Met396, Phe398, Ala460, Thr486, Leu487, Thr488, Gly489, Asn457, Ala577

^aThe observed different interactions are shown in italics.

FIG 5.

Binding modes and predicted receptor–ligand interaction within 4 Å of the ligand for PfM1AAP and phebestin in (A) stick and (B) surface models and PfM1AAP and bestatin in (C) stick and (D) surface models. The atom colors are defined as follows: receptor carbon in gray, phebestin carbon in magenta, bestatin carbon in salmon, oxygen in red, nitrogen in blue, sulfur in yellow, and zinc in orange. The dashed yellow lines are the polar interactions.

FIG 6.

Binding modes and predicted receptor–ligand interaction within 4 Å of the ligand for PfM17LAP and phebestin in (A) stick and (B) surface models and PfM17LAP and bestatin in (C) stick and (D) surface models. The atom colors are defined as follows: receptor carbon in gray, phebestin carbon in magenta, bestatin carbon in salmon, oxygen in red, nitrogen in blue, sulfur in yellow, and zinc in orange. The dashed yellow lines are the polar interactions.

Phebestin and bestatin were subjected to the SwissADME web server (http://www.swissadme.ch/, accessed on June 24, 2022) and pkCSM-pharmacokinetics web server (http://biosig.unimelb.edu.au/pkcsm/, accessed on June 24, 2022) to predict their drug-likeness characteristics, ADME, and toxicity of the compound of interest. Table 3 summarizes the results of the drug-likeness properties of phebestin and bestatin.

TABLE 3.

Predicted pharmacokinetic and pharmacodynamic properties of phebestin and bestatin

Parameter	Compound
	Phebestin	Bestatin	Desired value
mol wt (g/mol)	441.52	308.37	≤500
No. H-bond acceptors	6	5	≤10
No. H-bond donors	5	4	≤5
No. rotatable bond (rotors)	13	9	≤10
Topological polar surface area (TPSA, Å2)	141.75	112.65	≤140
Log P octanol/water partition coefficient	0.95	0.66	≤5
Intestinal absorption (% absorbed)	31.264	35.928	≥30%
Total clearance (log mL/min/kg)	0.466	0.403
LD50 oral rat acute toxicity (mol/kg)	2.512	1.942
Hepatoxicity	Yes	Yes
AMES toxicity (act as a carcinogen)	No	No
Max tolerated dose in human (log mL/min/kg)	−0.123	0.239	≥0.477 high tolerance
Drug likeness
Lipinski (Pfizer); violation	Yes; 0	Yes; 0
Ghose; violation	Yes; 0	Yes; 0
Veber (GSK); violation	No; 2: rotors >10, TPSA >140	Yes; 0
Egan (Pharmacia); violation	No; 1: TPSA >140	Yes; 0
Muegge (Bayer) filter; violation	Yes; 0	Yes
Abbott Bioavailability Score	0.55	0.55

Effect of phebestin and bestatin on P. yoelii 17XNL-infected mice.

A dose of 10 mg/kg phebestin exhibited no alteration of parasitemia levels in P. yoelii 17XNL-infected mice compared to the control (Fig. 7). In another experiment, 20 mg/kg phebestin and bestatin significantly reduced parasitemia compared with the control (Fig. 7B). No mortality was observed in either control, phebestin, or bestatin-treated mice; additionally, significantly lower parasitemia peaks were observed in phebestin-treated (19.53%) and bestatin-treated (23.00%) groups than in the untreated group (29.55%).

FIG 7.

Effects of 7 days of phebestin and bestatin treatments on C57BL/6 mice infected with P. yoelii 17XNL. Parasitemia levels after treatment with (A) 10 mg/kg phebestin and (B) 20 mg/kg phebestin and bestatin (representative from two independent trials with similar results) following the inoculation of 1 × 10⁷ infected erythrocytes. Each group consisted of six mice. (*) The significance of differences in the level of parasitemia in the phebestin-treated mice compared with the control mice, (#) the significance of differences in the level of parasitemia in the bestatin-treated mice compared with the control mice, and (■) the significance of differences in the level of parasitemia in the phebestin-treated mice compared with the bestatin-treated mice. The significance of differences was analyzed by two-way ANOVA followed by Tukey’s multiple-comparison test (P < 0.05).

Effect of phebestin and bestatin on P. berghei ANKA-infected mice.

The same dose of 20 mg/kg/day of phebestin and bestatin was used in mice infected with P. berghei ANKA based on the efficacy treatments of P. yoelii 17XNL-infected mice. Treatment with phebestin for 7 days, once daily, prolonged survival by 16.67% up to 23 dpi (Fig. 8). At 20 mg/kg, bestatin treatment also exhibited prolonged survival at the same level as phebestin treatment (16.67%, up to 23 dpi). Although the effect of the 7-day regimen on survival rates was statistically insignificant (P > 0.05), the impact of this regimen on parasitemia generally exhibited significant differences compared to the control group during the treatment period.

FIG 8.

Effects of 7 days of treatment with 20 mg/kg phebestin and bestatin. (A) Parasitemia levels after treatment with phebestin and bestatin. (B) The survival graph of mice infected with P. berghei ANKA after treatment with phebestin and bestatin. The data are representative of two independent trials with similar results. Mice were infected by inoculation with 1 × 10⁷ P. berghei ANKA-infected erythrocytes. (*) The significance of differences in the level of parasitemia in the phebestin-treated mice compared with the control mice, (#) the significance of differences in the level of parasitemia in the bestatin-treated mice compared with the control mice, and (■) the significance of differences in the level of parasitemia in the phebestin-treated mice compared with the bestatin-treated mice. The significance of differences was analyzed by two-way ANOVA followed by Tukey’s multiple-comparison test (P < 0.05). Survival rates were calculated using the log rank test (Mantel–Cox).

DISCUSSION

Plasmodium spp. relies on exogenous amino acids from the host cell and the external environment for fast amplification and repeated rounds of invasion, growth, division, and egress (12). For protein anabolism, hemoglobin is an essential supplier of these nutrients. Endopeptidases, such as falcipains and plasmepsins, work within the acidic digestive vacuole to generate peptides that are transferred to the cytosol and processed by aminopeptidase (13). PfM1AAP and PfM17LAP are neutral aminopeptidases that act in the parasite cytosol, which have neutral pH environments resulting in the release of amino acids from hemoglobin-derived peptides. Inhibitors of PfM1AAP and PfM17LAP, such as bestatin, have antimalarial action in vitro and in vivo (8). Additionally, the MAPs were identified to be highly expressed in trophozoites, the intraerythrocytic stage where protein production is at its peak, showing the most susceptible stage to the MAP inhibitor (8).

Phebestin was first isolated from Streptomyces sp. MJ716-m3 and identified to have aminopeptidase N inhibitory activity (9). This inhibitory property of phebestin may also be essential in inhibiting malaria parasites as bestatin. Phebestin, in this study, exhibited in vitro inhibitory activity against both P. falciparum 3D7 and K1 with IC₅₀ values of 157.90 ± 6.26 nM and 268.17 ± 67.59 nM, respectively (Table 1). The resulting IC₅₀ values of phebestin were better than those of the aminopeptidase inhibitor bestatin (3,220 ± 168.00 nM for P. falciparum 3D7 and 4,795.67 ± 424.82 nM for K1). Compared to the positive control, chloroquine and phebestin exhibited lower IC₅₀ values against P. falciparum K1, consequently showing a lower resistance index (14). Although phebestin exhibited much weaker IC₅₀ values than artemisinin, the cytotoxicity of HFFs was not observed at a concentration of 2,500 μM, allowing the estimation SI values of >15,832.80 for P. falciparum 3D7 and >9,322.44 for K1, which are higher than those of artemisinin and chloroquine. Regarding the cytotoxicity against erythrocytes, phebestin exhibited a comparable hemolysis rate value to the antimalarial drugs, with 0.90 ± 0.79% at a concentration of 100 μM, approximately >300-fold of the IC₅₀ values of phebestin, ensuring its safe use in vitro. Collectively, phebestin meets the criteria for the hit of an antimalarial drug defined by the Japanese Global Health Innovative Technology (GHIT) Fund (15) with IC₅₀ values <1 μM for both sensitive and resistant strains of P. falciparum and SI values >10. Furthermore, the present study is targeting the molecules that potently clear asexual blood-stage parasitemia, Target Candidate-1 (TCP-1), based on the criteria given by Medicine for Malaria Venture (16).

To investigate the stage-specific inhibition of phebestin to three main asexual blood stages of P. falciparum 3D7, phenotypic analysis was performed. Using the concentration of 100, 10, and 1-fold of the IC₅₀ of phebestin, all parasite stages were inhibited in a similar tendency to the positive controls (Fig. 2A), indicating that phebestin inhibited all stages of the parasites, under 24 h exposure. Additionally, the inhibitions appeared to exhibit higher abnormal morphologies percentage in a dose-dependent manner rather than retained life cycle (Fig. 2B). Further observation of the inhibition activity of phebestin on in vitro culture of P. falciparum 3D7 was performed to identify the effects on the parasitemia levels and morphology of the parasite. Although the parasite cultures exposed to 1 μM phebestin can develop into trophozoites from the ring stage on the initial culture for the first 24 h of incubation (Fig. 3A), after the following 48 and 72 h of incubation, the parasites were retained at the trophozoite stage and showed shrunken morphology at 72 h of incubation, indicating that the 1 μM concentration of phebestin had no effect on the ring stage, but affected the trophozoite stage due to higher metabolism rate, which consequently altered parasite development and the life cycle. These results are consistent with the altered morphology of aminopeptidase inhibitor treatments reported by Harbut et al. (7) and Bounaadja et al. (17). Furthermore, under treatment with phebestin, the parasitemia levels did not increase, indicating that erythrocyte reinvasion did not occur (Fig. 3B). In contrast, bestatin at the same concentration did not exhibit altered morphologies or life cycles, merely causing lower levels of parasitemia than the control (Fig. 3B).

Following the 72-h exposure to phebestin, parasites were incubated up to 144 h from initial infection; the compounds were provided in the culture (unwashed group) or removed (washed group). Under these conditions, parasite growth remained inhibited in the unwashed and washed groups for the phebestin- and bestatin-treated cultures. Notably, in the washed group of phebestin, the parasites were still unable to grow even after the compound's removal (Fig. 4A), indicating that the parasites were impaired and showed dying signs, as observed by the shrunken morphology within the erythrocytes after 72 h of exposure to phebestin. On the other hand, in the case of bestatin, the parasites were still growing even under unwashed or washed conditions; nevertheless, the growth of the parasites was observed to be significantly lower than that in the control groups.

The structure of phebestin resembles that of bestatin ([2S,3R]-3-amino-2-hydroxy-4-phenylbutanoyl)-l-leucine with the addition of one phenylalanine moiety and one less methylene moiety in the side chain ([2S,3R]-3-amino-2-hydroxy-4-phenylbutanoyl)-l-valyl-l-phenylalanine. These additional side groups resulted in a greater inhibitory effect on the in vitro culture of P. falciparum 3D7. From the screening results of bestatin derivatives at 1 μM (Table S1), the antimalarial activity of phebestin (AHPA-Val-Phe) was 92.00% and 34.79% for bestatin (AHPA-Leu), and none was found active in between (Fig. S1). AHPA-His-Phe, in which the Val of phebestin was replaced with His (prepared in the expectation that it competes with His in the zinc-binding HEXXH motif, which coordinates to the active center metal Zn of aminopeptidase), gave only 2.06% inhibition. AHPA-Phe-His, the reverse, showed 4.94% inhibition, while AHPA-Val-His, in which the Phe is replaced by His, decreased to 2.45%. On the other hand, the inhibition values of AHPA-d-Leu and (2S,3R)-AHPA-l-Ala, in which the second Leu of bestatin was replaced with d-form or Ala, decreased the activity to 1.23% and 2.11%, respectively. Based on these experimental results, it is important to note that the bestatin-based scaffold for the antiplasmodial agent should contain AHPA at the N terminus, the second amino acid preferably Val instead of Leu, and the third amino acid preferably an amino acid with a high hydrophobic moiety such as Phe, are sufficient conditions for antiplasmodial activity. Furthermore, the interactions assessment of phebestin with the targeted receptors in comparison with bestatin was performed in silico.

We performed an in silico study of phebestin against PfM1AAP and PfM17LAP to assess the structural difference in the binding properties. Both phebestin and bestatin are bound to the main active site of both receptors, Zn ions, by ion complex interactions via α-hydroxyl and/or carbonyl groups, with convincing CNN scores. In the case of the interaction of ligands and receptors in PfM1AAP, both phebestin and bestatin were found to bind with the same site at the binding pocket. However, observation of the surface model (Fig. 5B and D) showed that phebestin had a more hydrophobic interaction with Tyr-580 (red circle) due to the phenylalanine moiety compared to bestatin, resulting in a stronger binding affinity. Furthermore, the hydrogen bond interactions were related to the different bonding motives in the PfM17LAPs (Table 2, Fig. 6), with the presence of amino acid residues Glu461, Ser554; phebestin/Arg463; bestatin, which is characteristic of both compounds involved in hydrogen bonding in PfM17LAP, these interactions may also contribute to the characteristics of the antiplasmodial activities. Overall, phebestin exhibited more affinity for both receptors than bestatin due to the addition of a phenylalanine moiety, resulting in more contact sites for hydrogen bond interactions and/or hydrophobic interactions (18) (Table 2, Fig. 5 and 6). These results also suggest that phebestin possibly acts as a dual inhibitor both for PfM1AAP and PfM17LAP, as observed for bestatin.

The drug-likeness prediction was performed using the SwissADME web server and pkCSM-pharmacokinetics web server (Table 3). From these analyses, phebestin was not carcinogenic based on the AMES toxicity profile but may have hepatotoxicity; these properties were also observed for bestatin. The oral rat acute toxicity of phebestin showed a high possible concentration at LD₅₀ value of 2.512 mol/kg, equal to 1,109.118 g/kg, but low dose tolerance used in humans (−0.123, desired range >0.477 log mL/min/kg), along with predicted intestinal absorption of 31.264% and total clearance 0.466 mL/min/kg. Collectively, phebestin had drug-likeness properties based on Lipinski (19), Ghose (20), and Muegge (21) rules but categorized as no drug-likeness based on Veber (22) and Egan (23) rules because of the topological polar surface area (TPSA, >140) and the number of rotatable bonds (>10). Based on these results, phebestin is likely ineffective when orally administered.

The in vivo rodent malaria model was used to confirm the in vitro inhibitory activity. P. yoelii 17XNL and P. berghei ANKA were used in this study. Mice were treated using phebestin i.p. at doses of 10 and 20 mg/kg/day for 7 days; bestatin was also used at a dose of 20 mg/kg/day for comparison purposes. In the nonlethal rodent malaria P. yoelii 17XNL infection, no inhibition was observed at a dose of 10 mg/kg/day phebestin (Fig. 7A). However, by increasing to a 2-fold dose at 20 mg/kg/day, partial inhibition of parasitemia levels was observed and exhibited significantly lower parasitemia peaks in the phebestin-treated (19.53% on day 16) and bestatin-treated (23.00% on day 17) groups than in the untreated group (29.55% at day 17) (Fig. 7B). In a different set of experiments of a lethal rodent malaria model, P. berghei ANKA infection (Fig. 8) treated using the same dose and phebestin treatment prolonged survival by 16.67% (up to 23 dpi). Treatment with bestatin resulted in prolonged survival at the same level as phebestin treatment (16.67%, up to 23 dpi). The rapid clearance of phebestin and bestatin may be one of the reasons why both compounds did not reach adequate levels of parasite clearance. Meanwhile, in reported study, the treatments of artemisinin and chloroquine with a dose range of 10 to 50 mg/kg/day for 3 consecutive days were required to cure the infections completely (24). As reported previously, bestatin enters erythrocytes at a very low rate (0.3%/min) and is rapidly eliminated from the serum with a half-life of ~2 h (25). Even though there is no reported study related to phebestin, the prediction of the pharmacokinetic and pharmacodynamic properties (Table 3) indicates similar properties to that of bestatin. Furthermore, the effects of bestatin and phebestin on parasitemia levels were significantly different compared to the control group during the 7-day treatment period and prevented early death occurrences compared to the control group.

In summary, phebestin exhibited more potent and selective antiplasmodial activity than the well-known aminopeptidase inhibitor bestatin. However, the in vivo study was limited to showing only partial inhibition of the parasite’s growth; accordingly, investigation using different administration methods, different doses, or structure-activity relationship studies to achieve better pharmacokinetics may give further insight related to the efficacy and inhibitory effects of phebestin or its derivatives. Additionally, the current research is focused on the blood stage of Plasmodium spp., and further investigation into other life stages may also be useful for the broader use of phebestin as an antiplasmodial agent.

MATERIALS AND METHODS

Compounds.

Phebestin (molecular weight [MW], 441.5 g/mol, CAS 187402-73-9) was provided by the Institute of Microbial Chemistry (BIKAKEN). The positive control for aminopeptidase inhibitor, which also exhibited antiplasmodial activity, bestatin (MW 308.4 g/mol, CAS 58970-76-6), was provided by the Institute of Microbial Chemistry (BIKAKEN). The positive controls for the antiplasmodial drug artemisinin (MW, 282.33 g/mol, CAS 63968-64-9) and chloroquine diphosphate (MW, 515.86 g/mol, CAS 50-63-5) were purchased from Sigma–Aldrich (St. Louis, MO, USA). DMSO (MW 78.13 g/mol, CAS 67-68-5, Wako, Osaka, Japan) was used as a negative control. The maximum final concentration of DMSO used for in vitro study was 0.01%; at this concentration, the solvent does not affect the proliferation of the parasites (26).

Parasites.

Two human malaria parasites, chloroquine-sensitive (3D7) and multidrug-resistant (K1) strains of P. falciparum, were used for the in vitro study. In addition, P. yoelii 17XNL (nonlethal malaria model) and P. berghei ANKA (lethal malaria model) were used for in vivo studies in murine models.

In vitro culture of P. falciparum.

P. falciparum 3D7 and K1 were cultivated in complete medium, which consisted of Roswell Park Memorial Institute (RPMI)-1640 (Sigma–Aldrich, St. Louis, MO, USA), 25 mM HEPES (CAS 7365-45-9, Sigma–Aldrich, St. Louis, MO, USA), 0.5% (wt/vol) AlbuMax II (lipid-rich BSA, Gibco, Waltham, MA, USA), 24 mM NaHCO₃ (CAS 144-55-8, Wako, Osaka, Japan), 184 μM hypoxanthine (CAS 68–94-0, Wako, Osaka, Japan), 0.025% (vol/vol) gentamicin (50 mg/mL, CAS 1403-66-3, Gibco, Waltham, MA, USA), and supplemented with 2% washed human O-(+) erythrocytes (supplied by Japanese Red Cross Society, Hokkaido, Japan) in an incubator at 37°C, 5% CO₂, and 5% O₂. In addition, parasitemia level was monitored using Giemsa-stained thin blood smears (CAS 51811-82-6, Merck, Darmstadt, Hesse, Germany).

In vitro inhibition assay of P. falciparum.

The 5% d-sorbitol method (CAS 50–70-4, Wako, Osaka, Japan) was used for parasite synchronization to obtain ≥90% ring-stage parasites and used within 2 h. Next, 50 μL of synchronous parasites (0.5% parasitemia and 2% hematocrit) were seeded in a 96-well plate filled with 50 μL of test compounds. Each compound/drug was prepared from 20 mM stock (in DMSO) by two times dilutions, using complete medium, to eight concentrations from 2,000 nM (final concentration of 1000, 500, 250, 125, 62.5, 31.25, 15.625, and 7.8125 nM). Wells containing only test compounds and erythrocytes were used to correct background signals. The plate was incubated for 72 h at 37°C, 5% CO₂, and 5% O₂. As previously described, the SYBR green I-based fluorescence assay was used for the growth detection of P. falciparum 3D7 and K1 (27, 28). Following 72 h of incubation, 100 μL of lysis buffer with 0.02% (vol/vol) SYBR green I (SYBR Green I, Lonza, Basel, Switzerland) was added to each well mixed by pipetting and then incubated at room temperature for 2 h in the dark. A Fluoroskan Ascent instrument (Thermo Fisher Scientific, Waltham, MA, USA) was used to measure the fluorescence intensities at an excitation of 485 nm and emission wavelengths of 518 nm. This assay was performed in quadruplicate for each concentration, and the resulting inhibition values are the average of three independent trials.

Effects of in vitro treatments of phebestin and bestatin on morphology and parasitemia level of P. falciparum 3D7.

Before being used for the experiment, the parasites were tightly synchronized using a two-step sorbitol treatment protocol. The synchronized parasites were used within 2 h for the ring stage, 8 to 10 h for the trophozoite stage, and 24 to 26 h for the schizont stage. The synchronization of the parasite culture resulted in >90% ring-stage parasites. Following this sorbitol treatment, early trophozoites can be obtained after 6 to 8 h of incubation from the first sorbitol treatment (up to 20 h old), and the early schizonts can be obtained after following 16 h of incubation (up to 36 h old). After 36 h, the schizonts will be matured, rupture the RBC, and then release merozoites which readily infect new RBC. Until this time point, at 48 h post synchronization, the ring stage can again be observed, and at 72 h, trophozoites can be observed (29, 30).

For the stage-specific test, the parasites were used at the ring, trophozoite, or schizont stages with 0.5% parasitemia and 2% hematocrit. In this assay, phebestin, bestatin, artemisinin, and chloroquine (each 100, 10, and 1-fold of their respective IC₅₀) were designated treated groups, and cultures with complete medium only defined as no treatment (NT) group. This assay was performed in triplicate for each concentration. Subsequently, the thin blood smears of each treatment group were prepared after 24 h of incubation. The percentage of the drug-affected parasites, including abnormal and retained morphologies, were determined by counting the total of 100 parasites and the number of parasites that exhibit abnormal development. Then, the number of abnormal parasites was divided by the total number and multiplied by 100. The criteria of the normal and abnormal morphologies are provided in Table S2 and Fig. S2.

For the 72 h compound exposure with compound wash-unwashed assay, the parasites were used at the ring stage with 0.5% parasitemia and 2% hematocrit. In this assay, phebestin and bestatin (each 1 μM, equal to IC₉₀ of phebestin) were designated treated groups, and cultures with complete medium only designated untreated groups. Using IC₉₀ for the compound exposure allowed to observe the morphologies of the parasites under incomplete inhibition, the possibility of the occurring abnormal morphology, and the conditions of parasites after the compound's removal. This assay was performed in triplicate for each concentration and time point. Subsequently, the thin blood smears of each treatment group were prepared after 1, 24, 48, or 72 h of incubation. In different culture sets, after 72 h of incubation, the medium was changed, the compounds were washed three times using a new complete medium, and the incubation was continued in either the presence (unwashed group) or absence of tested compounds (washed group) for up to 144 h. The thin blood smears of each treatment group were consecutively prepared at 96, 120, and 144 h. The smears were visualized by a BZ-900 all-in-one microscope (Keyence BioRevo, Tokyo, Japan). The parasitemia levels were determined by enumerating the number and ratio of infected erythrocytes relative to uninfected erythrocytes.

In vitro cytotoxicity in human cells and hemolysis rate in human erythrocytes.

The cytotoxicity of the tested compounds was evaluated using a cell viability assay of human foreskin fibroblasts (HFFs), as described previously (27, 31), and a human erythrocyte hemolysis assay was performed at 100 μM each tested compound as previously described (31).

Docking simulation and prediction of drug-likeness properties.

The initial ligand coordinates (phebestin, bestatin) used for docking were generated using RDkit's rdDistGeom. ETKDGv3 module (32) for 300 conformers, and then optimized conformers were calculated by MMFF (AllChem. MMFFGetMoleculeForceField module). These conformers were then clustered into 13 (phebestin) and 9 (bestatin) conformers using DBSCAN (33, 34). Each conformer was used to dock with Gnina (35–38). The center of docking was set within 5 Å with reference to the ligands of each crystal structure. Protein crystals were obtained from the Protein Data Bank (https://www.rcsb.org/, accessed on June 1, 2022) in .pdb format with protein IDs 3T8V (PfM1AAP) and 3T8W (PfM17LAP) (7). Only the highest CNN (convolutional neural network) score results from docking simulations will be used for further interaction analysis. The interaction analysis was performed using Ligplot⁺ (39), and the models were visualized using the PyMOL Molecular Graphic System (version 2.0 Schrödinger, LLC). Furthermore, phebestin was analyzed using SwissADME (40) and pkCSM pharmacokinetics (41) to predict its drug-likeness properties, including pharmacokinetic and pharmacodynamic profiles; bestatin was also analyzed for comparison purposes.

Mice and in vivo infections.

Eight-week-old male C57BL/6J mice (Japan CLEA, Tokyo, Japan) were used for in vivo experiments and set as six mice per cage per group. The mice were maintained at 24°C, with 50% relative humidity and lighting from 8 a.m. to 8 p.m.; additionally, drinking water and food were provided ad libitum (CLEA Rodent Diet CE-2; Japan CLEA, Tokyo, Japan). Before being used for the experiment, P. yoelii 17XNL and P. berghei ANKA were recovered from frozen packed erythrocyte stocks via a passage in mice. Infection experiments were performed with intraperitoneal (i.p.) injections of freshly infected erythrocytes from donor mice (1 × 10⁷ parasites/mouse; designated 0 days postinfection [dpi]). The parasitemia levels were checked after 2 h of infection by taking 2 μL of blood from the mouse tail and applying thin blood smears. The treatments started when the parasitemia level reached 1%. The infected mice were intraperitoneally administered vehicle (1× PBS), phebestin (10 or 20 mg/kg), or bestatin (20 mg/kg) once daily for 7 days (0 to 6 dpi) (42). The choice of 7 days for the suppressive test is based on several factors. First, it is long enough to allow for multiple cycles of parasite replication and growth in the animal's blood, which is necessary to assess the drug's effectiveness accurately. At the same time, it is not too long that the test becomes overly cumbersome or time-consuming. Additionally, a 7-day treatment regimen is typically sufficient to achieve a high level of drug exposure and ensure that any parasites remaining in the animal's blood are likely to be susceptible to the drug, allowing for a clear assessment of drug efficacy (43). The daily observation was performed for survival and clinical signs. Parasitemia of infected mice was observed daily using Giemsa-stained thin blood smears (up to 30 dpi). Mice that displayed severe clinical symptoms (such as an arched back, immobility, >20% body weight loss, and pain signs) were immediately euthanized. The animals used in this study were cared for and treated based on the Guide for the Care and Use of Laboratory Animals of the Ministry of Education, Culture, Sports, Science and Technology, Japan. The protocol for malaria parasite culture in human blood was used with permit number: #2013-04-3 by the ethical committee of Obihiro University of Agriculture and Veterinary Medicine. The animal experiment protocols were used with permit number 21-35 by the ethical committee of Obihiro University of Agriculture and Veterinary Medicine.

Statistical analysis.

The IC₅₀ or CC₅₀ values were analyzed from three independent experiments using a nonlinear regression fit to the logarithm concentration value of the compound versus inhibition percentage of parasite or cell growth, respectively, using GraphPad Prism 8 (GraphPad Software, Inc., La Jolla, CA, USA). Group comparison analyses were performed using one-way or two-way ANOVA followed by Tukey’s multiple-comparison test. The survival percentage was calculated using the log rank (Mantel–Cox) test. A P value of <0.05 was considered statistically significant and is shown as an asterisk or a symbol, defined in each figure legend, together with the name of the test used.