Efficacy of the Antimalarial MMV390048 against Babesia Infection Reveals Phosphatidylinositol 4-Kinase as a Druggable Target for Babesiosis

ABSTRACT

The present study aimed to evaluate the anti-Babesia effect of MMV390048, a drug that inhibits Plasmodium by targeting the phosphatidylinositol 4-kinase (PI4K). The half inhibitory concentration (IC₅₀) of MMV390048 against the in vitro growth of Babesia gibsoni was 6.9 ± 0.9 μM. In immunocompetent mice, oral treatment with MMV390048 at a concentration of 20 mg/kg effectively inhibited the growth of B. microti (Peabody mjr strain). The peak parasitemia in the control group was 30.5%, whereas the peak parasitemia in the MMV390048-treated group was 3.4%. Meanwhile, MMV390048 also showed inhibition on the growth of B. rodhaini (Australia strain), a highly pathogenic rodent Babesia species. All MMV390048-treated mice survived, whereas the mice in control group died within 10 days postinfection (DPI). The first 7-day administration of MMV390048 in B. microti-infected, severe combined immunodeficiency (SCID) mice delayed the rise of parasitemia by 26 days. Subsequently, a second 7-day administration was given upon recurrence. At 52 DPI, a parasite relapse (in 1 out of 5 mice) and a mutation in the B. microti PI4K L746S, a MMV390048 resistance-related gene, were detected. Although the radical cure of B. microti infection in immunocompromised host SCID mice was not achieved, results from this study showed that MMV390048 has excellent inhibitory effects on Babesia parasites, revealing a new treatment strategy for babesiosis: targeting the B. microti PI4K.

INTRODUCTION

Babesiosis is a zoonotic disease caused by the tick-transmitted Babesia species, and it leads to huge economic losses and poses a serious health risk worldwide (1). Most species responsible for Babesia infections are host-specific, and over 100 Babesia species have the ability to infect many types of mammalian hosts, including domestic animals and humans (2). With the rapid increase in the number of human babesiosis cases caused by Babesia microti reported over the past years, particularly in the northeastern and northern midwestern regions of the United States (3), babesiosis has been attracting interest as an emerging zoonosis (2). Clinical symptoms of babesiosis include fever, anemia, hemoglobinuria, anorexia, and emaciation (2). Some severe cases of human babesiosis include multiple organ failure, which leads to death (3). The current treatment strategy for human babesiosis in severely ill patients is atovaquone (ATO) combined with azithromycin (AZI) or clindamycin combined with quinine (4). ATO combined with AZI is recommended for all B. microti-infected patients and shows fewer side effects than clindamycin plus quinine. This combination is also used for B. gibsoni, an agent of canine babesiosis (5, 6). However, mutations in the cytochrome b and ribosomal protein subunit L4 genes confer ATO and AZI resistance to the parasites, resulting in the failure of treatment (7). However, clindamycin combined with quinine, an alternative for human babesiosis treatment, shows serious side effects to humans (8, 9). Furthermore, monotherapy of either clindamycin or quinine displayed poor efficacy against the parasites (10–12). Hence, the diversity of Babesia species, the wide range of hosts, and the low effectiveness of currently used drugs highlight the urgency to discover new drug targets and produce new drugs with anti-Babesia activity (13). Due to the shared biological features of Babesia with Plasmodium, some of the currently used anti-Babesia drugs were initially developed as anti-Plasmodium drugs. Therefore, the screening of drugs with known action modes, based on their efficacy against Plasmodium, will reduce the cost and save time in anti-Babesia drug development.

Phosphatidylinositol 4-kinase (PI4K) is a ubiquitous eukaryotic lipid kinase and is involved in the synthesis of phosphatidylinositol 4-phosphate (PI4P), a member of the phosphoinositide family (14). PI4P is involved in the architecture of the Golgi apparatus and the trans-Golgi network, and it also regulates trafficking to and from the Golgi (15). It plays a key role in the synthesis of membrane polyphosphoinositides and in the regulation of multiple cellular functions (16, 17). PI4Ks are classified into two classes based on enzymatic differences: type II (PI4KII) and type III (PI4KIII). Each class contains α and β isoforms (18). Recently, Plasmodium PI4K (PfPI4K) type III β was demonstrated to be a promising druggable target for the elimination of malaria (15). So far, three novel anti-Plasmodium compound classes that target PfPI4K have been reported to inhibit the multiple-stage development of Plasmodium, namely, imidazopyrazine, quinoxaline, and 2-aminopyridine (15, 19, 20). MMV390048 (Fig. 1A), also known as MMV0048, is a representative of the 2-aminopyridine class and was developed based on the hits of a phenotypic high-throughput screen from the commercial BioFocus library (21). A previous report showed that MMV390048 could inhibit all Plasmodium life cycle stages except hypnozoites in the late liver stage (19). Given its strong inhibitory effect and known action mode on Plasmodium, we investigated whether MMV390048 has the same inhibitory effect on Babesia species. Therefore, this study aimed to evaluate the efficacy of MMV390048 against the in vitro growth of B. gibsoni and the in vivo growth of the human babesiosis causative agent, B. microti (Peabody mjr strain), and a highly pathogenic rodent species, B. rodhaini. Likewise, we elucidated the potential mechanisms of MMV390048 activity against Babesia parasites.

FIG 1.

MMV390048 demonstrates potent inhibition on B. gibsoni in vitro and B. microti in vivo. (A) Chemical structure of MMV390048. (B) Dose-dependent inhibition curve of MMV390048 on B. gibsoni in vitro. Each value represents the mean ± standard deviation (SD) of three independent experiments carried out in triplicate. (C) Inhibitory effects of MMV390048 (daily dose of 20 mg/kg from 4 to 10 DPI) and atovaquone (ATO) plus azithromycin (AZI) (daily dose of 20 mg/kg plus 20 mg/kg from 4 to 10 DPI) on the growth of B. microti in BALB/c mice. (D) Changes of hematocrit (HCT) values in mice treated with MMV390048 or ATO plus AZI compared to those of vehicle-treated mice. The arrows indicate time of treatment. The asterisks indicate significant differences (P < 0.05) between the drug-treated groups and the vehicle-treated control group.

RESULTS

Inhibitory efficacy of MMV390048 on B. gibsoni in vitro and B. microti in vivo.

The in vitro activity of MMV390048 against B. gibsoni showed a steep inhibition curve with a half inhibitory concentration (IC₅₀) value of 6.9 ± 0.9 μM (Fig. 1B). In B. microti-infected BALB/c mice, parasitemia significantly increased in the vehicle-treated control group and reached the highest parasitemia (average 30.5%) at 8 days postinfection (DPI) (Fig. 1C). In contrast, the parasites were significantly inhibited by MMV390048 (daily dose of 20 mg/kg from 4 to 10 DPI) and by ATO plus AZI (daily dose of 20 mg/kg plus 20 mg/kg from 4 to 10 DPI) (P < 0.05). The peak parasitemia of the MMV390048-treated and ATO plus AZI-treated groups were 3.4% and 4.0%, indicating growth inhibition of 88.9% and 86.9%, respectively, compared with the control group (peak parasitemia 30.5%) (Fig. 1C). However, the ATO plus AZI-treated group eventually relapsed and maintained a low parasitemia (<1%) until 24 DPI, whereas the blood smears of the control group mice were negative from 28 DPI. Hematocrit (HCT) changes were monitored as the index for anemia in B. microti-infected mice. There was no significant HCT reduction noted in either the MMV390048-treated group or the ATO plus AZI-treated group. In contrast, significant HCT reductions were observed in the vehicle-treated group from 8 DPI to 28 DPI (P < 0.05) (Fig. 1D), indicating that MMV390048 treatment could block the development of anemia in B. microti-infected mice.

Inhibitory efficacy of MMV390048 on B. rodhaini-infected mice.

To confirm whether MMV390048 has the ability to inhibit the growth of other Babesia species, the lethal B. rodhaini was used for further study. In the vehicle-treated group, mice infected with B. rodhaini showed high parasitemia (82%) (Fig. 2A) and severe anemia (data not shown). The B. rodhaini-infected mice died within 10 DPI (Fig. 2B), whereas the administration of MMV390048 (daily dose of 20 mg/kg from 2 to 8 DPI) or tafenoquine (TAF; dose of 20 mg/kg once at 2 DPI) inhibited the rapid growth of the parasite, thereby preventing anemia development and resulting in no animals succumbing to the lethal B. rodhaini infection. Additionally, a relapse was observed in the TAF-treated group at 16 DPI, and parasitemia was eventually eliminated by immunity, a result similar to those presented in previous reports (22, 23).

FIG 2.

Efficacy of MMV390048 against lethal B. rodhaini infection in BALB/c mice. (A) MMV390048 (daily dose of 20 mg/kg from 2 to 8 DPI) and tafenoquine (TAF; dose of 20 mg/kg once at 2 DPI) prevented B. rodhaini growth in mice, compared with vehicle-treated mice. (B) Survival rates of MMV390048-treated, TAF-treated, and vehicle-treated mice. The arrows indicate time of treatment. The asterisks indicate significant differences (P < 0.05) between the drug-treated groups and the vehicle-treated group.

Identification of a B. microti PI4K mutation in a MMV390048-treated, immunodeficient mouse with relapsed babesiosis.

Following the promising results described above, MMV390048 was further evaluated in B. microti-infected, severe combined immunodeficiency (SCID) mice with 2-fold higher inoculum (2 × 10⁷ infected red blood cells [iRBCs]) to test a more rapidly progressive babesiosis model. In the vehicle-treated mice, the highest parasitemia reached 40% to 50% at 10 to 14 DPI (Fig. 3A). Treatment with MMV390048 (daily dose of 20 mg/kg from 4 to 10 DPI) prevented the rise of parasitemia, with no parasites being observed on blood smears starting at 8 DPI (Fig. 3B). However, the first 7-day treatment of MMV390048 failed to prevent relapse. Notably, two (MMV390048 no. 1 and MMV390048 no. 4) of five MMV390048-treated mice relapsed at 26 DPI. Thus, a second 7-day treatment (daily dose of 20 mg/kg from 28 to 34 DPI) was given to all mice, but an unexpected relapse in the MMV390048 no. 2 mouse at 52 DPI, with increased parasitemia reaching 18.9%, was observed (Fig. 3B). As the PI4K mutation in P. falciparum confers resistance to MMV390048 (15, 19), we confirmed whether this was also present in the SCID mouse (MMV390048 no. 2) that had reemergent parasitemia by sequencing the B. microti PI4K (BmPI4K; PiroplasmaDB: BMR1_03g03920; GenBank: XP_012649395), a homologue of PfPI4K obtained by a BLASTP search. A single nucleotide variant (SNV) of Bmpi4k with a substitution at position 2,237 (from T to C) was found, and this led to a nonsynonymous coding change from leucine to serine (L746S) (Fig. 3C and D).

FIG 3.

Efficacy of MMV390048 against B. microti in SCID mice and BmPI4K mutation as the target of MMV390048. (A and B) Parasitemia changes of vehicle-treated or MMV390048-treated (daily dose of 20 mg/kg from 4 to 10 DPI and 28 to 34 DPI, respectively) B. microti-infected SCID mice. The arrows indicate time of treatment. (C and D) Representative sequencing chromatogram of recrudescent parasites from MMV390048-treated B. microti-infected mice. The parasite DNA of MMV390048 no. 2 was extracted from a blood sample at 60 days postinfection. The DNA was used to amplify the B. microti PI4K gene and was sequenced.

B. microti PI4K L746S as the resistance-conferring mutation against MMV390048 action.

The BmPI4K L746S mutant strain was isolated and passaged to a donor SCID mouse that was subjected to uninterrupted monotherapy with MMV390048. Then, BALB/c mice were infected with either the purified BmPI4K mutant isolate or the B. microti wild-type (BmWT) to confirm the association of the L746S mutation with the resistance of B. microti to MMV390048. Consistent with the above results, the BmWT-infected mice treated with MMV390048 (daily dose of 20 mg/kg from 4 to 10 DPI) exhibited decreased parasitemia compared with the vehicle-treated group (Fig. 4A). On the other hand, the BmPI4K L746S mutation resulted in the ineffectiveness of MMV390048 (daily dose of 20 mg/kg from 4 to 10 DPI) against parasite growth (Fig. 4B). Meanwhile, the TAF-treated (dose of 20 mg/kg once at 4 DPI) and ATO plus AZI-treated groups (daily dose of 20 mg/kg plus 20 mg/kg from 4 to 10 DPI) showed significant inhibition on the BmPI4K L746S mutant strain (Fig. 4B). The morphological changes of parasites in all groups are shown in Fig. 4C. MMV390048-treated BmWT showed severe degenerative changes at 6 DPI (Fig. 4C) compared with vehicle-treated BmWT. At 6 DPI, the TAF-treated BmPI4K L746S mutant parasites showed abnormal morphology, as described in previous reports (23, 24), whereas the ATO plus AZI treatment caused severe degeneration of BmPI4K L746S mutant parasites. No significant morphological changes were observed in MMV390048-treated BmPI4K L746S mutant parasites compared with vehicle-treated parasites.

FIG 4.

Efficacy of MMV390048, TAF, and ATO plus AZI against the B. microti PI4K L746S mutant strain. (A) Growth of B. microti wild-type (WT) parasites in vehicle-treated and MMV390048-treated (daily dose of 20 mg/kg from 4 to 10 DPI) groups (as control groups; n = 3 per group). (B) Growth of the B. microti PI4K L746S mutant strain in vehicle-treated, TAF-treated (dose of 20 mg/kg once at 4 DPI), ATO plus AZI-treated (daily dose of 20 mg/kg plus 20 mg/kg from 4 to 10 DPI), and MMV390048-treated (daily dose of 20 mg/kg from 4 to 10 DPI) BALB/c mice (n = 5 per group). The arrows indicate time of treatment. The asterisks indicate significant differences (P < 0.05) between the drug-treated groups and the vehicle-treated group. (C) Light micrographs of B. microti-infected mice during vehicle and MMV390048 treatment as well as light micrographs of B. microti PI4K mutant-infected mice during vehicle, TAF, ATO plus AZI, and MMV390048 treatment. Bar = 10 μm.

Multiple sequence alignment of Babesia PI4K and molecular docking study.

The full-length amino acid sequence of BmPI4K shared identity values of 41.3%, 62.8%, 59.0%, 60.5%, and 44.3% with human (uniprot: Q9UBF8; GenBank: NP_001185702), P. falciparum (PlasmoDB: PF3D7_0509800; GenBank: XP_001351656), B. bovis (PiroplasmaDB: BBOV_IV009520; GenBank: XP_001610874), B. bigemina (PiroplasmaDB: BBBOND_0107050; GenBank: XP_012766582), and B. gibsoni (unpublished data) PI4K, respectively. The results revealed that PI4K is evolutionarily conserved across Babesia species, especially the C-terminal of PI4K (Fig. 5A). The kinase domain of PfPI4K (ranging from residues 1,261 to 1,559) was assigned in a previous report (15). Herein, we predicted the structure of the catalytic domain of BmPI4K and performed molecular docking with MMV390048 (Fig. 5B). The docking model showed that the MMV390048 is embedded in the kinase domain of BmPI4K and positioned in the binding pocket (Fig. 5B). Likewise, the hydrogen bonds (H-bonds) between MMV390048 and BmPI4K at residues LYS741, LEU790, and SER795 were predicted, whereas the mutation site found in this study (LEU746) showed van der Waals interaction with the MMV390048 (Fig. 5C).

FIG 5.

Multiple sequence alignment of Babesia PI4K and molecular docking study. (A) Multiple sequence alignment of PI4K kinase domains. The mutation site of BmPI4K found in this study is indicated by a red inverted triangle, while the sites related with P. falciparum drug resistance are marked with blue inverted triangles. The key site that blocks MMV390048 affinity with human PI4K (25) is labeled with a green inverted triangle. (B) Docking representation of the binding modes of MMV390048 to the pocket of BmPI4K. (C) 2D representation of BmPI4K and MMV390048 interactions. The mutation site is labeled in red.

DISCUSSION

MMV390048 is a new compound that inhibits Plasmodium development by targeting PfPI4K. A previous clinical trial of MMV390048 demonstrated the promise of this compound as a single drug for Plasmodium treatment (25). Such results provide ideas for screening or developing new compounds for treating babesiosis. Indeed, our study confirmed that MMV390048 showed potent inhibition against Babesia species. The IC₅₀ value of MMV390048 against B. gibsoni in vitro was 6.9 ± 0.9 μM (Fig. 1B), which is lower than the previously reported IC₅₀ value of TAF (IC₅₀ = 20 ± 2.4 μM) (22) but higher than the IC₅₀ value of ATO (IC₅₀ = 89.0 ± 17.3 nM) (26). Likewise, the current MMV390048 IC₅₀ value was higher than the IC₅₀ value against the intraerythrocytic stage of P. falciparum (NF54 drug-sensitive strain, 28 nM) (19). Although Plasmodium and Babesia are closely related, some variation in its features, such as the solute permeability of infected erythrocytes, may have resulted in lower sensitivities of the Babesia species to drugs compared with the Plasmodium species (27). A previous study showed that peak concentrations (Cmax) of MMV390048 reached 5.4 μg/mL (13.7 μM) in mouse plasma and 3.9 μg/mL (9.9 μM) in monkey plasma after a single oral dose administration of 20 mg/kg (19). In another clinical trial, the peak Cmax of MMV390048 was 1.1 μg/mL (2.8 μM) in humans after a single oral dose administration of 120 mg (25, 28). Therefore, the recommended dose for treating human babesiosis should be higher than 120 mg to reach a high plasma concentration, which is expected to eventually eliminate parasites. Nonetheless, the concomitant safety issues should also be considered.

In the current study, the in vivo inhibitory effects of MMV390048 on B. microti were comparable with those of ATO plus AZI, which is the drug treatment recommended by the U.S. Centers for Disease Control and Prevention (CDC) (29), and anemia development was significantly prevented. In subsequent experiments, mice infected with the lethal B. rodhaini showed a high parasitemia of 82% and died within 10 DPI, whereas MMV390048 effectively inhibited the rapid growth of the parasite, thereby protecting the mice from death. TAF is more effective than MMV390048, as only one dose is required. Despite this, the limitation of TAF use is its risk of inducing severe hemolytic anemia in some G6PD-deficient patients (30), whereas MMV390048 is deemed safe as a treatment for G6PD-deficient patients (19). The first-line drug for babesiosis treatment is ATO plus AZI. A recent report showed that B. microti-infected SCID mice that were consecutively treated with ATO plus AZI relapsed on 28 DPI with no response to the subsequent treatment (31). Indeed, acquired drug resistance has been reported in immunocompromised patients, who will experience a relapse of disease if initial antibabesial therapy fails (32). In the present study, we used B. microti-infected SCID mice to evaluate the effect of MMV390048 on an immunocompromised host. Babesia microti was potently inhibited by MMV390048 after a 7-day treatment, but parasite recurrence was observed after 2 weeks, prompting us to give a second 7-day treatment. Although the twice-given treatment did not achieve the radical cure of babesiosis, the tolerance to MMV309948 was better than the tolerance to ATO plus AZI. Moreover, subsequent experiments proved that the potential drug target of MMV390048 was BmPI4K.

In this study, a single nucleotide mutation was detected in relapsed B. microti infection by sequencing the BmPI4K gene, and a nonsynonymous coding change of BmPI4K L746S resulted in a lessened sensitivity of parasites to MMV390048. In the case of P. falciparum, whole-genome sequencing of the MMV390048-resistant strain revealed that the PfPI4K S743T or A1319V mutation conferred the resistance to the parasites (19). Moreover, the H1484Y and S1320L mutations in PfPI4K also conferred some degree of cross-resistance to MMV390048. The BmPI4K L746S mutation was characterized by parasite growth and by morphology phenotypes similar to those of the wild-type parasites after MMV390048 treatment. On the other hand, the BmPI4K L746S mutation did not affect the inhibitory effects of other antibabesial drugs, such as TAF and ATO plus AZI, on parasites, further validating that the mutation in the BmPI4K renders MMV390048 inefficacious against B. microti infection. Altogether, similar to that observed in P. falciparum, these data demonstrate that the PI4K is the target of MMV390048 action in B. microti.

PI4K is ubiquitous in eukaryotes, and it regulates intracellular signaling and trafficking via the phosphorylation of lipids (15). Kinase domain alignment of PI4K orthologues showed high identity among the different Babesia species and with its human orthologue. However, MMV390048 had been proven to have no affinity with human kinase, apart from human PIP4K2C (uniprot: Q8TBX8), ATM (uniprot: Q13315), and TNIK (uniprot: Q9UKE5). Human PIP4K2C is the main target of MMV390048, and its inhibition on the host is unknown (19). A docking study explained that the most significant difference in the binding cavities between HuPI4K and PfPI4K is the change of Q606 in HuPI4K to S1365 in PfPI4K and that the sidechain of HuPI4K Q606 clashes with the CF3-pyridyl of MMV390048, resulting in a low potency of MMV390048 to HuPI4K (33). However, this conflict does not appear to exist in Babesia species, as evidenced by the different residues (B. microti F798, B. gibsoni V873, B. bovis V949, and B. bigemina V944) and the significant inhibition on parasites. In addition, we observed that the residue LEU746 was spatially close to the LYS741, which generates H-bonds and contributes to anchor the MMV390048 onto the binding pocket. Hence, we speculate that the BmPI4K L746S mutation may affect the H-bonds between LYS741 and MMV390048 to prevent compound binding to the pocket. BmPI4K plays a key role for parasite survival and could develop resistance under drug pressure. This resistance confers some degree of fitness advantage for parasite survival but may also incur a fitness cost (34). However, we did not observe prominent differences on parasite growth or virulence between the BmPI4K L746S mutant strain and the BmWT strain. Therefore, this mutation may not be enough to markedly affect catalytic function.

Despite the excellent inhibition shown by MMV390048 on Babesia spp., there are some limitations in the present study. For instance, the twice 7-day regimen of MMV390048 did not eradicate Babesia infection in immunocompromised mice and did not halt the emergence of resistant parasites, suggesting that the evaluation of various reasonable modes of administration and combination with currently used drugs is necessary. Collectively, our results demonstrate that MMV390048 is a promising drug for babesiosis treatment and that Babesia PI4K is a druggable target for babesiosis.

MATERIALS AND METHODS

Ethics statement.

The protocols performed in the current study were carried out according to the ethical guidelines approved by the Obihiro University of Agriculture and Veterinary Medicine (permit numbers: animal experiment, 21-133; DNA experiment, 1723-5 and 1724-5; pathogen, 201712-5).

Chemicals.

MMV390048 was purchased from MedChem Express (NJ, USA). ATO and AZI were purchased from Sigma-Aldrich (Tokyo, Japan). The drugs were dissolved in a vehicle (sesame oil) (Sigma-Aldrich) to make a stock solution of 40 mg/mL and stored at 4°C until use.

Mouse strains.

6-week-old female BALB/c mice and SCID mice (CLEA, Tokyo, Japan) were used for the in vivo studies. All animal experiments followed the regulations of the Animal Care and Use Committee and the Biological Safety Committee of the Obihiro University of Agriculture and Veterinary Medicine, Hokkaido, Japan.

Parasites.

In vitro cultures of B. gibsoni (Oita strain) were used for the experiment. Parasites were grown in canine RBCs suspended with RPMI 1640 containing 20% normal canine serum (22). The culture was maintained in an atmosphere of 5% CO₂ and 5% O₂. For the in vivo study, the frozen stocks of B. microti Peabody mjr strain (ATCC PRA-99) and B. rodhaini Australia strain were thawed and intraperitoneally injected in BALB/c mice. When parasitemia levels were 10 to 20% in donor mice, the iRBCs were collected and diluted with phosphate-buffered saline (PBS). The B. microti challenge infection dose was 1.0 × 10⁷ iRBCs in BALB/c mice and 2.0 × 10⁷ iRBCs in SCID mice, while the B. rodhaini dose was 1.0 × 10⁷ iRBCs in BALB/c mice.

In vitro anti-Babesia efficacy of MMV390048.

The in vitro anti-Babesia activity was measured using a SYBR green I (SG1) proliferation assay (35). Briefly, 95 μL of medium with various concentrations (100, 50, 25, 10, 5, and 1 μM) of MMV390048 were added in triplicate into 96-well plates. Then, 5 μL of diluted parasite-infected canine RBCs (1% parasitemia) were added to each well and incubated for 96 h. Next, 100 μL of lysis buffer containing 2 × SG1 nucleic acid stain were added to each well. The fluorescence values were measured using a fluorescence spectrophotometer (485 nm and 518 nm wavelengths, Fluoroskan Ascent, Thermo Fisher Scientific, USA) after 2 h of incubation. IC₅₀ was calculated based on the fluorescence values using GraphPad Prism 8 (GraphPad Software Inc., USA).

In vivo anti-Babesia efficacy of MMV390048.

The efficacy of MMV390048 on Babesia infection was evaluated in B. microti- and B. rodhaini-infected mice. For the treatment of B. microti-infected mice, 15 mice were divided into 3 groups. The dosages for treatment were determined based on in vitro results and previous reports. Group I (n = 5) was orally treated with 20 mg/kg MMV390048, while group II (n = 5) was orally treated with 20 mg/kg ATO plus 20 mg/kg AZI (31). As the solvent control, group III (n = 5) was orally treated with 0.2 mL vehicle. All of the treatments were given for 7 consecutive days, starting at 4 DPI. For the treatment of B. rodhaini-infected mice, 15 mice were divided into 3 groups. Groups I and III (n = 5 per group) were treated as described above, whereas group II (n = 5) was treated with 20 mg/kg TAF once (24).

For the treatment of B. microti-infected SCID mice, 10 mice were divided into 2 groups. Group I (n = 5) was orally treated with 20 mg/kg MMV390048 for 7 consecutive days, starting at 4 DPI, while group II (n = 5) was orally treated with 0.2 mL vehicle. Due to the relapse of parasites, a second 7-day treatment was given from day 28. The parasitemia in mice was monitored by examining 3,000 RBCs on Giemsa-stained thin smear slides. Anemia development was monitored by the change of HCT values using an automated veterinary hematology analyzer (MEK-6550 Celltac α, Nihon Kohden, Tokyo, Japan).

Efficacy of MMV390048 on BmPI4K mutant strain.

To isolate the BmPI4K L746S mutant strain, the parasites from the SCID mouse that relapsed after two regimens of MMV390048 were collected and passaged in a naive SCID mouse. Then, to evaluate the efficacy of MMV390048, TAF, and ATO plus AZI on the BmPI4K L746S mutant strain, 26 BALB/c mice were divided into 6 groups: groups I and II (n = 3 per group) were infected with BmWT, while groups III to VI (n = 5 per group) were infected with the BmPI4K L746S mutation strain. Groups II and III were orally treated with MMV390048, groups I and IV were orally treated with 0.2 mL vehicle, and groups V and VI were treated with TAF and ATO plus AZI, respectively.

Detection of B. microti PI4K gene variants.

At 60 DPI, the DNA of the relapsed parasites were extracted, and sequencing was performed. Briefly, 10 μL blood were collected from the tail vein and diluted in 90 μL phosphate-buffered saline (PBS). Samples were incubated at 100°C for 5 min and then centrifuged at 10,000 rpm for 5 min. The supernatants were collected and used for the following PCR assay. The B. microti PI4K gene (3,212 bp) was amplified using KOD FX Neo DNA polymerase (Toyobo; catalog no. KFX-201), using the primers listed in Table S1. The final reaction volume of 25 μL consisted of 4.5 μL of double-distilled water, 12.5 μL of 2 × PCR buffer, 5 μL of 2 mM dNTP, 0.75 μL of 10 μM forward and reverse primers, 1 μL of DNA sample, and 0.5 μL of KOD FX Neo DNA polymerase. The following thermocycling conditions were used: initial denaturation at 94°C for 2 min, 35 cycles of 98°C for 10 sec denaturation, 55°C (BmPI4K-U-F and BmPI4K-U-R2) or 58°C (BmPI4K-D-F1 and BmPI4K-D-R) for 30 sec annealing, 68°C for 1 min; and the final extension at 68°C for 7 min. The amplicons were purified using the QIAquick PCR Purification Kit (28106; Qiagen) and were subjected to Sanger sequencing. The genetic variants of each amplicon were confirmed by alignment to the WT sequence as the reference. The obtained sequence was deposited in the GenBank database with accession no. ON191810.

PI4K sequence alignment and homology modeling.

B. microti RI strain PI4K (100% identity with Peabody mjr strain) (PiroplasmaDB: BMR1_03g03920; GenBank: XP_012649395), B. gibsoni PI4K (unpublished data), B. bigemina PI4K (PiroplasmaDB: BBBOND_0107050; GenBank: XP_012766582), B. bovis PI4K (PiroplasmaDB: BBOV_IV009520; GenBank: XP_001610874), and human PI4K (uniprot: Q9UBF8; GenBank: NP_001185702) were obtained by a homology search using P. falciparum PI4K (PlasmoDB: PF3D7_0509800; GenBank: XP_001351656). Sequence alignment was done using MUSCLE and was analyzed using Jalview v2.8 software. The three-dimensional structures prediction of BmPI4K was done using AlphaFold as described by Jumper et al. (36), and the docking poses were generated by Smina (37). The molecular docking results of MMV390048 with BmPI4K, the hydrogen bonds, and the van der Waals interaction were evaluated using Discovery Studio 3.5 Visualizer.

Statistical analysis.

Data analysis was performed using GraphPad Prism (La Jolla, CA, USA) version 8. The differences in parasitemia and HCT values between the control and the treated groups were analyzed using a one-way analysis of variance. Survival rates were calculated using the Kaplan-Meier method with regard to the log-rank test. A P value of <0.05 was considered to be indicative of a statistically significant result.