Orally Bioavailable Endochin-Like Quinolone Carbonate Ester Prodrug Reduces Toxoplasma gondii Brain Cysts

Toxoplasmosis is a potentially fatal infection for immunocompromised people and the developing fetus. Current medicines for toxoplasmosis have high rates of adverse effects that interfere with therapeutic and prophylactic regimens. Endochin-like quinolones (ELQs) are potent inhibitors of Toxoplasma gondii proliferation in vitro and in animal models of acute and latent infection. ELQ-316, in particular, was found to be effective orally against acute toxoplasmosis in mice and highly selective for T. gondii cytochrome b over human cytochrome b.

ABSTRACT

Toxoplasmosis is a potentially fatal infection for immunocompromised people and the developing fetus. Current medicines for toxoplasmosis have high rates of adverse effects that interfere with therapeutic and prophylactic regimens. Endochin-like quinolones (ELQs) are potent inhibitors of Toxoplasma gondii proliferation in vitro and in animal models of acute and latent infection. ELQ-316, in particular, was found to be effective orally against acute toxoplasmosis in mice and highly selective for T. gondii cytochrome b over human cytochrome b. Despite its oral efficacy, the high crystallinity of ELQ-316 limits oral absorption, plasma concentrations, and therapeutic potential. A carbonate ester prodrug of ELQ-316, ELQ-334, was created to decrease crystallinity and increase oral bioavailability, which resulted in a 6-fold increase in both the maximum plasma concentration (C_max) and the area under the curve (AUC) of ELQ-316. The increased bioavailability of ELQ-316, when administered as ELQ-334, resulted in efficacy against acute toxoplasmosis greater than that of an equivalent dose of ELQ-316 and had efficacy against latent toxoplasmosis similar to that of ELQ-316 administered intraperitoneally. Treatment with carbonate ester prodrugs is a successful strategy to overcome the limited oral bioavailability of ELQs for the treatment of toxoplasmosis.

INTRODUCTION

Toxoplasma gondii infection is highly prevalent in humans, parasitizing billions of people (1). In the great majority of infections, symptoms are not appreciable; however, infected individuals are at risk of developing toxoplasmosis if they become immunologically compromised by HIV infection, cancer, or immunosuppressive therapies. Infection during pregnancy can cause fetal demise and severe congenital neurological and ocular damage. Outside of immunocompromised populations, otherwise healthy people may develop eye disease from T. gondii infection that can progress to blindness. Additionally, T. gondii causes severe infection in wild and domesticated animals and may threaten endangered species (2, 3).

Current multidrug regimens for toxoplasmosis have had high rates of adverse events, leading to discontinuation in 30% of patients (4–6). The most common adverse events were rash, diarrhea, and hematologic and hepatic toxicity (5). The high rates of hematologic toxicity are related to pyrimethamine inhibition of the host dihydrofolate reductase enzyme and high rates of allergic reactions and overlapping toxicities with the medications used in populations that are susceptible to toxoplasmosis. These groups also have higher rates of severe reactions than other patient groups, such as toxic epidermal necrolysis and Stevens-Johnson syndrome, which may be fatal (7). The toxicities of current therapies are made worse by the prolonged exposure that patients must undergo in the initial treatment phase and the suppressive secondary prophylaxis phase. In part, a long duration of treatment is required because T. gondii tissue cysts are not eradicated by current regimens. New treatments that are less toxic and that diminish or eliminate latent T. gondii tissue cysts would greatly improve outcomes for patients with toxoplasmosis.

Endochin-like quinolones (ELQ) are potent inhibitors of apicomplexan pathogens, including T. gondii, Plasmodium falciparum, and Babesia microti. ELQ-316 was identified as a lead compound for toxoplasmosis based on efficacy and specificity for T. gondii cytochrome b over human cytochrome b (8). Despite efficacy when administered orally in systemic mouse models of toxoplasmosis, malaria, and babesiosis, ELQ-316 plasma and brain concentrations were not adequate to eliminate acute T. gondii brain infection (9). Increasing the dose of ELQ-316 to attain adequate tissue concentrations was limited by the crystallinity of ELQs, because the bioavailability of ELQs decreases as the dose increases due to precipitation in the gastrointestinal tract at higher concentrations (10). In order to improve the oral efficacy of ELQ-316, we created a carbonate ester prodrug of ELQ-316, ELQ-334, and determined its pharmacokinetic parameters and efficacy against acute and latent toxoplasmosis.

RESULTS

Extended treatment of latent T. gondii brain infection with ELQ-316 and ELQ-271.

Previously, ELQ-316 and ELQ-271 were shown to reduce established brain cysts when administered intraperitoneally (i.p.) for 15 days (8). Based on these results, we tested the effects of a longer duration of treatment against established tissue cysts. In these experiments, CBA/J mice were infected with the MS-ME49 T. gondii strain for 5 weeks prior to treatment to establish chronic brain tissue cysts. ELQs were administered to the mice for 5 weeks i.p. at 5 mg/kg of body weight/day. This course rapidly reduced the number of brain tissue cysts within the first week, with a continued reduction compared to that for the control being seen at 5 weeks (Fig. 1A and B). Following treatment with ELQs, the mice were given dexamethasone continuously to evaluate the viability of the remaining cysts. Survival was prolonged in the majority of mice treated with ELQ-316; however, all mice eventually succumbed to infection, indicating that viable cysts remained after treatment (Fig. 1C). No signs of toxicity in mice treated with ELQs were observed. Together, these findings suggest that prolonged treatment substantially reduces the number of viable cysts but does not completely eliminate them.

FIG 1.

Extended treatment with ELQ-316 and ELQ-271 reduces cysts but does not eliminate all viable T. gondii brain cysts. (A) Mice received daily treatment via i.p. injection for 5 weeks after being infected for 5 weeks with the MS-ME49 strain of T. gondii. (B) Cysts were counted weekly in each group. Treatment with ELQ-316 and ELQ-271 decreased but did not eliminate T. gondii brain cysts. Cyst levels in DMSO-treated mice were significantly different (P < 0.05, Mann-Whitney test) from those in ELQ-271- and ELQ-316-treated mice at all time points except week 3 between the DMSO- and ELQ-316-treated mice. Initial group sizes were 5 mice per group, except for those treated with ELQ-271 and ELQ-316 for 5 weeks, which consisted of 13 and 10 mice each, respectively, because of uneven attrition due to infection. One group of 4 uninfected mice was included as a control for the dexamethasone treatments. All time points and treatment groups contain cyst values from 4 to 13 mice, except for the DMSO-treated group at 4 weeks, which contains data from 3 mice. (C) Mice were treated with dexamethasone continuously following treatment with the ELQs. All mice succumbed to infection. ELQ-treated mice survived longer than the controls (for DMSO- versus ELQ-271-treated mice, P = 0.017; for DMSO- versus ELQ-316-treated mice, P = 0.016; P values were determined by the log-rank Mantel-Cox test). The uninfected and DMSO-, ELQ-271-, and ELQ-316-treated groups contained 4, 5, 13, and 9 mice, respectively. IP, intraperitoneal; Dex., dexamethasone; PO, per os; error bars, standard deviations.

Chemical synthesis of ELQ-334.

A prodrug form of ELQ-316 was synthesized to increase oral bioavailability. ELQ-334 has been reported previously without detail of its synthesis and structural characterization (11). In this report, we describe ELQ-334 synthesis with full characterization of its structure, including the results of X-ray diffraction analysis of ELQ-334 crystals. Following a procedure modified from the procedure described by Miley et al. (12), 6-fluoro-7-methoxy-2-methyl-3-(4-(4-(trifluoromethoxy)phenoxy)phenyl)quinolin-4-yl ethyl carbonate (ELQ-334) was obtained by reacting ethyl chloroformate with 6-fluoro-7-methoxy-2-methyl-3-(4-(4-(trifluoromethoxy)phenoxy)phenyl)quinolin-4(1H)-one (ELQ-316) (13) in the presence of sodium hydride at 60°C in a 95% yield as a white solid (Fig. 2). Whereas ELQ-316 decomposes at 314°C, ELQ-334 exhibits a melting point of 140°C, consistent with a significant loss of crystallinity.

FIG 2.

(A) Synthesis and presumed in vivo conversion of ELQ-316 and ELQ-334. (B) Oak Ridge thermal ellipsoid plot (ORTEP) of ELQ-334. (C) A fragment of the crystal structure of ELQ-334. Hydrogen bonds are shown by dashed lines. Thermal ellipsoids are drawn at the 30% probability level.

The Oak Ridge thermal ellipsoid plot (ORTEP) of ELQ-334 and the X-ray crystal structure of ELQ-334 containing two symmetrically independent molecules are shown in Fig. 2B and C. The central planar quinoline ring and the first phenyl ring of the diaryl ether is twisted with a 68° torsion angle around the C-2—C-15 bond. Weak intermolecular hydrogen bonding networks between both CH-N and CH-OCH₃ are also observed in the crystal structures. The packing does not provide any evidence that the molecules in the crystal are connected via π-π stacking interactions. The diaryl ether group is significantly twisted from the central quinoline aromatic group to disrupt possible intermolecular π-π interactions. The absence of π-π stacking is consistent with the low melting point of 140°C.

Noncompartmental pharmacokinetic analysis of ELQ-316 and ELQ-334.

ELQ-316 and ELQ-334 were administered orally to mice, and then plasma and brain concentrations were measured to compare the maximum plasma concentration (C_max), the time to C_max (T_max), the area under the curve from time zero to 96 h (AUC_0–96), and the half-life (t_1/2) of ELQ-316 from ELQ-316 and ELQ-334 (Fig. 3). The compounds were dissolved in polyethylene glycol 400 (PEG 400) at a single dose of 10 mg/kg of ELQ-316. ELQ-334 yielded a plasma C_max from ELQ-316 of 4,378 ng/ml, whereas the C_max of ELQ-316 from ELQ-316 was 721 ng/ml. The T_max of both compounds was 4 h, and the AUC_0–96 was also increased approximately 6-fold to 115,195 ng · h/ml by the carbonate ester promoiety. The ELQ-316 plasma concentration of 1,665 ng/ml from ELQ-334 at 24 h and the t_1/2 of 11.6 h predicts efficacy with once-daily dosing. The brain tissue concentrations of ELQ-316 from ELQ-334 were 1,543 ng/ml at 4 h and 405 ng/ml at 24 h. The brain tissue concentration-to-plasma concentration ratio was 0.35 at the plasma T_max. By comparison, the brain concentrations achieved from the oral dose of ELQ-316 were 165 ng/g at T_max and 75 ng/g at 24 h. The mean plasma concentration of ELQ-334 did not exceed 238 ng/ml (5% of C_max), indicating that conversion from the prodrug (ELQ-334) to active compound (ELQ-316) occurs very quickly. Pharmacokinetic analysis was performed with PKSolver software (14).

FIG 3.

Pharmacokinetic study of ELQ-316 and ELQ-334 from oral administration of ELQ-316 or ELQ-334. Molar equivalents of 10-mg/kg ELQ-316 and ELQ-334 were administered in a single dose to mice via oral gavage in PEG 400. (A) Plasma concentrations of ELQ-316 and ELQ-334 over time after administration of ELQ-316 and ELQ-334. (B) Brain tissue concentrations of ELQ-316 after administration of ELQ-316 or ELQ-334. Error bars, standard errors of the means.

Efficacy of ELQ-334 against acute toxoplasmosis.

ELQ-334 at 5 mg/kg increased the survival of mice against a virulent T. gondii strain that is uniformly fatal in mice (Fig. 4A). ELQ-334 was given orally to mice at 1 or 5 mg/kg/day for 5 days following infection with the RH strain of Toxoplasma gondii for 24 h. Treatment with 5 mg/kg/day prolonged survival in all mice, with 2 out of 4 mice surviving through the conclusion of the experiment at 33 days. Control mice and mice treated with 1-mg/kg/day ELQ-334 were euthanized 6 and 7 days after displaying overt signs of infection, respectively. In the group treated with 5 mg/kg/day of ELQ-334, the mice were euthanized at day 17 and day 20 after infection. The survival of the mice treated with ELQ-334 at 5 mg/kg was statistically greater than that of the control mice (P = 0.008). In previously published experiments of ELQ-316 at 5 mg/kg/day in the same model, mice were euthanized at 13 days, and in these mice, systemic infection was cleared, but brain infection progressed (15). The survival of acute infection that is presented here suggests that the increased plasma concentration of ELQ-316 from ELQ-334 either prevents brain infection from becoming established or results in brain concentrations of ELQ-316 that are adequate to treat the acute brain infection. The burden of T. gondii infection in mice, as measured by T. gondii bioluminescence, was decreased after 3 days of treatment with ELQ-334 at 5 mg/kg/day compared to that after 3 days of treatment with ELQ-334 at 1 mg/kg/day and for the controls (P < 0.0001) (Fig. 4B). Bioluminescence became undetectable in the surviving mice. No signs of toxicity in mice treated with the ELQs were observed.

FIG 4.

Efficacy of oral treatment with ELQ-334 against acute toxoplasmosis. Mice were infected with a type I T. gondii strain expressing firefly luciferase on day 0, followed by daily treatment for 5 days starting on day 1. Each group consisted of 4 mice. (A) The survival of mice treated with 5-mg/kg ELQ-334 was statistically greater than that of the controls (P = 0.008, calculated by the log-rank test). (B) Luminescence in mice measured during and after treatment. Luminescence was significantly reduced by day 4 in mice treated with ELQ-334 at 5 mg/kg/day compared to that in the controls (P < 0.00001). LOD, limit of detection.

Treatment of latent T. gondii brain infection with ELQ-334.

The activity of ELQ-334, the carbonate ester prodrug of ELQ-316, against latent T. gondii infection with the MS-ME49 T. gondii strain was tested (Fig. 5). ELQ-334 given orally at 10 mg/kg for 2 weeks reduced the number of T. gondii cysts 67% compared to the number in control mice that received vehicle alone and reduced the number of parasite genomes per brain 82% compared to the number in control mice (Fig. 5B and C). No signs of toxicity in mice treated with the ELQs were observed. T. gondii brain cysts from mice that were treated with ELQ-334 were smaller in diameter than cysts from control mice (Fig. 5D), thus providing a potential explanation for the greater reduction in the number of parasite genomes per brain than in the number of cysts per brain due to treatment.

FIG 5.

Efficacy of oral treatment with ELQ-334 against established T. gondii brain cysts. (A) Mice were infected for 5 weeks with MS-ME49 prior to daily treatment for 2 weeks. (B) Number of T. gondii brain cysts in mice treated with 3 mg/kg and 10 mg/kg of ELQ-334. The indicated P values are from an ordinary analysis of variance with multiple comparisons and Tukey’s correction. One outlier was removed from the vehicle-treated group based on ROUT analysis (Q value, 0.1%). The cyst values in the groups treated with the vehicle, ELQ-334 at 3 mg/kg, and ELQ-334 at 10 mg/kg are from 12, 15, and 16 mice, respectively. (C) Number of genomes per brain in mice treated with ELQ-334. The indicated P values are from an ordinary analysis of variance with multiple comparisons and Tukey’s correction. One outlier was removed from the vehicle-treated group based on ROUT analysis (Q value, 0.1%). Data are from 7 vehicle-treated mice and 8 ELQ-334-treated mice. (D) Diameter of cysts from mice treated with ELQ-334. The P value is from a Mann-Whitney test. Data are from 5 brain samples per group, with 13 to 21 cysts being measured per sample. (E) Distribution of T. gondii in the brains of infected mice. Data are from 8 mice per group for the vehicle- and ELQ-316-treated groups. One mouse was used for the uninfected control group. (F) Regraphing of the distribution of T. gondii in the brains of infected mice treated with ELQ-334 to visualize the pattern and its similarity to that for mice treated with the vehicle, as shown in panel E. (G) Mice were infected with cysts from the ELQ-334-treated mice for 5 weeks prior to retreatment with ELQ-334 for 2 weeks. (H) ELQ-334 treatment of mice infected with cysts from mice that had previously been treated with ELQ-334. P values are from a Mann-Whitney test. Cyst levels in the vehicle-/vehicle-treated (Veh/Veh), vehicle-/ELQ-334-treated (Veh/ELQ), ELQ-334-/vehicle-treated (ELQ/Veh), and ELQ-334-/ELQ-334-treated (ELQ/ELQ) groups are from 9, 6, 6, and 9 mice, respectively. (I) Comparison of the cytochrome b Q_o site sequence of the MS-ME49 strain to that of the CH-ME49 strain. (J) Mice infected with T. gondii CH-ME49 cysts were treated with ELQ-334 (n = 13) or vehicle only (n = 8). The P value is from a Mann-Whitney test. PO, per os; error bars, standard deviations.

The presence of brain cysts remaining after treatment raised the possibility of inherently resistant populations of cysts or acquired drug resistance. The distribution of the cysts in the brain was examined by mechanically slicing the brains into even sections from the anterior to the posterior direction and estimating the total number of T. gondii genomes per milligram of brain tissue using quantitative reverse transcription-PCR of genomic DNA; however, no difference between treated and control mice was observed, indicating that the susceptibility of the cysts to ELQ-334 was not related to cyst location (Fig. 5E and F). T. gondii parasites exposed to ELQ-316 were tested for acquired resistance by infecting naive mice with cysts from ELQ-334-treated mice and then treating them again with ELQ-334. The number of cysts after treatment was equivalent to the number in the original experiment in the mice that received ELQ-334 and the mice that received only the vehicle (Fig. 5H).

The cytochrome b gene was sequenced to determine if acquired ELQ-316 resistance mutations resulted in drug resistance in the remaining tissue cysts. The MS-ME49 strain that was used in the latent toxoplasmosis experiments was found to have a single nucleotide substitution resulting in a change from isoleucine to leucine at position 262 in the Q_o site of cytochrome b compared to the sequence of strain CH-ME49 (a different lineage of ME49; see Materials and Methods), the type I RH strain used in the acute infection model, and the cytochrome b sequence in the ToxoDB database (accession number TGME49_330000) (16). This mutation has previously been associated with atovaquone resistance and is located adjacent to the highly conserved PEWY region in the Q_o site (Fig. 5I) (17). The I262L substitution was discovered after the above-described experiments were completed. This substitution was determined to be present prior to these experiments by sequencing the cytochrome b gene of T. gondii MS-ME49 parasites that were not exposed to ELQ-316, ELQ-334, or other compounds.

ELQ-316 resistance from the cytochrome b Q_i site mutations in T. gondii and B. microti has established the Q_i site to be the primary target of ELQ-316 (11, 15). The susceptibility of the CH-ME49 strain was tested to determine if the I262L mutation resulted in resistance or increased susceptibility. Treatment with ELQ-334 reduced the mean number of T. gondii CH-ME49 cysts 83% (Fig. 5J). This reduction in cyst number is similar to the 67% reduction in MS-ME49 cysts achieved with ELQ-334 at 10 mg/kg (Fig. 5B), indicating that this substitution did not cause substantial resistance to ELQ-334. No signs of toxicity in mice treated with the ELQs were observed.

DISCUSSION

A drug capable of eradicating T. gondii tissue cysts from infected individuals would prevent recurrent toxoplasmosis in people who have suffered acute infection and potentially could be used to prevent toxoplasmosis in immunocompromised individuals with evidence of latent infection. A drug that eradicates cysts may also be effective over a shorter duration than current regimens and prevent relapses of ocular toxoplasmosis, which cause scarring and vision loss. Specifically, the 1-year treatment of congenital toxoplasmosis and indefinite secondary prophylaxis in immunocompromised patients could be shortened to limit drug toxicity.

Over the last decade, a number of compounds with efficacy against brain tissue cysts have been identified (8, 18–20). These compounds reduced tissue cysts but did not eliminate cysts in mouse models with intact immunity. The cause of inherent drug refractivity in latent tissue cysts has not been determined. It is also not clear if drug refractivity is specific to individual inhibitors or is a general mechanism that affects many drugs. A lack of metabolic vulnerability in nonreplicating bradyzoites and decreased drug penetration through the cyst wall are among the possible explanations for drug resistance.

Extended treatment with ELQ-316 via i.p. injection for 5 weeks did not eradicate latent infection. Although the number of cysts was greatly reduced, the cysts remaining after the 5-week treatment were viable. T. gondii brain tissue cysts are dynamic over the time course of infection, with the cysts exhibiting various degrees of replication and cyst packing density (21). T. gondii bradyzoites express isoforms of the carbon metabolism enzymes lactate dehydrogenase and enolase that favor glycolysis (22–25). It has been reported that bradyzoites also lack a functional tricarboxylic acid (TCA) cycle (26, 27). Considering these shifts in bradyzoite metabolism, cysts with replicating parasites would be vulnerable to ELQs and cysts that are fully committed to dormancy and not reliant on oxidative phosphorylation would survive. On the other hand, cytochrome b inhibition induces the stage transition to bradyzoites (28). ELQs may drive cysts with replicating parasites into latency, stopping the growth of cysts. The smaller cyst size in the ELQ-treated mice may result from this growth-limiting effect or be associated with an inherently recalcitrant population. That being said, the large and continued reduction in the number of cysts indicates that the great majority of brain cysts are vulnerable to cytochrome b inhibition.

Intermittent treatment was given both to test the efficacy of a prolonged duration of treatment and to determine if this strategy would allow cysts to become vulnerable by relieving drug pressure and allowing replication. Although the cysts were not eliminated, the number of cysts in the treatment groups and the controls continued to decline over time. The question remains whether the ultimate nadir of cyst numbers in mice was further reduced by ELQ treatment or whether the cyst number in control mice would eventually reach the low numbers achieved with ELQ-316 treatment. Interestingly, previous experiments that evaluated brain cysts over 8 weeks showed an increase in the number and size of cysts (21). Here we observed that cyst numbers began to decrease after 7 weeks in control mice and continued to decrease out to 10 weeks. In the subsequent ELQ-334 experiment, surviving cysts did not differ from those from the controls in their distribution or acquired cytochrome b resistance mutations, further underscoring the inherent drug-refractive properties of cysts.

An oral agent to treat toxoplasmosis is optimal for prolonged courses of therapy and in settings with limited health care resources. The carbonate ester promoiety of ELQ-334 disrupted π-π stacking, leading to the 6-fold enhanced oral bioavailability of ELQ-316, resulting in plasma concentrations of 9.5 μM at C_max and 3.6 μM at 24 h and brain concentrations to 3.3 μM at C_max and 0.88 μM at 24 h. These concentrations are more than 1,000-fold greater than the 50% inhibitory concentration (IC₅₀) against T. gondii. The carbonate ester promoiety is also beneficial, in that ELQ-334 was rapidly metabolized to ELQ-316, which limits the host exposure to the prodrug form. ELQ-334 treatment was more effective than ELQ-316 treatment, with 50% survival in mice, whereas in previous studies of orally administered ELQ-316, the mice did not survive more than 8 days after the completion of treatment (9). Moreover, oral treatment with 10-mg/kg ELQ-334 was effective at decreasing the number of tissue cysts to a similar degree as intraperitoneal injections at 5 mg/kg. Despite not eradicating cysts, ELQ-334 quickly decreased the number of cysts that were established over 5 weeks under conditions in which meningeal inflammation would be minimal, and penetration through the blood-brain barrier was important. The survival of mice during the prolonged administration of ELQ-316 both orally and i.p. also demonstrated that these doses are tolerated well and that cumulative toxicity does not limit efficacy in mice.

The complete elimination of tissue cysts would likely have a significant clinical benefit; however, a drug that achieves this goal remains elusive. That being said, further research to understand the effect of ELQ-334 on tissue cysts and their inherent resistance may provide a means to exploit the vulnerabilities of bradyzoites in deep latency and find drug combinations that eliminate cysts. More urgently, drugs for toxoplasmosis that are more effective and better tolerated are needed. Overall, ELQ-334 is a highly promising candidate for the treatment of toxoplasmosis. ELQ-334 is more effective than atovaquone, a well-tolerated, clinically used drug, and achieves higher brain concentrations and markedly less inhibition of human cytochrome b than atovaquone (8, 29). ELQ-316 is also more potent than the other currently used drugs, pyrimethamine, sulfadiazine, clindamycin, and trimethoprim-sulfamethoxazole, and does not show fetal toxicity in mice (30). Current regimens for toxoplasmosis are significantly limited by side effects that at times are severe. Given the need for better-tolerated drugs, the efficacy of ELQ-334 described in this report warrants further investigation of carbonate ester prodrugs of ELQ-316 for toxoplasmosis.

MATERIALS AND METHODS

Experimental compounds.

Unless otherwise stated, all chemicals and reagents were obtained from Sigma-Aldrich Chemical Company in St. Louis, MO (USA), or Combi-Blocks in San Diego, CA, and were used as received. ELQ-316 was synthesized as described by Nilsen et al. (13). Melting points were obtained in an OptiMelt automated melting point system from Stanford Research Systems, Sunnyvale, CA, USA. Gas chromatography-mass spectrometry (GC-MS) was performed using an Agilent Technologies 7890B gas chromatograph (30 m; the Agilent DB-5 column was set at 200°C for 2 min and then at 30°C/min to 300°C, with the inlet temperature set at 250°C) with an Agilent Technologies 5977A mass-selective detector operating at 70 eV. Silica gel chromatography was performed using an automated flash chromatography system (Biotage Isolera One, Uppsala, Sweden). ¹H nuclear magnetic resonance (NMR) spectra were obtained using a Bruker AMX-400 NMR spectrometer operating at 400.14 MHz. Raw NMR data were analyzed using iNMR Spectrum Analyst software. Chemical shifts were reported in units of parts per million (ppm; δ) relative to either tetramethylsilane (TMS) as the internal standard or residual solvent proton (7.26 ppm for deuterated CDCl₃). Coupling constant values are reported in hertz.

6-Fluoro-7-methoxy-2-methyl-3-(4-(4-(trifluoromethoxy)phenoxy)phenyl)quinolin-4-yl ethyl carbonate (ELQ-334).

To a stirred suspension of ELQ-316 (3.21 g, 7.0 mmol) in dry tetrahydrofuran (THF; 100 ml) was added a 60% mineral oil suspension of NaH (560 mg, 14.0 mmol, 2.0 eq), and the mixture was heated at 60°C for 30 min. Ethyl chloroformate (1.51 g, 14.0 mmol, 2 eq) in THF (5 ml) was added, and the reaction mixture was heated at 60°C for 5 h. It was then cooled to room temperature and water (10 ml) was added. The resulting mixture was filtered and separated, and the organic layer was concentrated in vacuo to give 4.1 g of a white solid. The product was purified by silica gel flash chromatography using 40% ethyl acetate in hexanes to give 3.52 g (95% yield) of ELQ-334 as a white solid. GC-MS showed one peak with 531 M⁺ (32%), 281 (100%). Melting point, 140.1 to 140.5°C. ¹H NMR (400 MHz; CDCl₃): δ 7.52 (d, J = 8.0 Hz, 1-H), 7.48 (d, J = 11.2 Hz, 1-H), 7.30 to 7.28 (m, 2-H), 7.24 to 7.21 (m, 2-H), 7.11 to 7.06 (m, 4-H), 4.15 (q, J = 7.1 Hz, 2-H), 4.04 (s, 3-H), 2.53 (s, 3-H), 1.22 (t, J = 7.1 Hz, 3-H). For X-ray analysis, ELQ-334 was recrystallized by slow evaporation from an ethanol solution.

X-ray crystallography.

Diffraction intensities for ELQ-334 were collected at 100 K on a Rigaku XtaLAB SynergyS diffractometer using CuKα radiation and λ equal to 1.54184 Å. The space group was determined based on intensity statistics. Absorption correction was applied by use of the SADABS program (31). The structure was solved by direct methods and Fourier techniques and refined on F² using full-matrix least-squares procedures. All non-H atoms were refined with anisotropic thermal parameters. The H atoms in methyl groups were refined in calculated positions in a rigid group model without restrictions on its rotation around C—C bonds (the HFIX 138 instruction in the SHELXTL program) (32). Other H atoms were found on the residual density map and refined without any restrictions with isotropic thermal parameters. In the crystal, there are two symmetrical molecules. All calculations were performed using the Bruker SHELXL-2014 package (32).

Pharmacokinetic analysis of ELQ-316 and ELQ-334.

The plasma and brain concentrations of the compounds were evaluated in CF-1 mice, which had access to food and water ad libitum at all times. ELQ-334 and ELQ-316 were dissolved in PEG 400 to a dose of 10 mg/kg, with the dose of ELQ-334 being adjusted to the molar equivalency of the ELQ-316 dose. These solutions were administered orally by gavage (0.1 ml per mouse). Blood samples were collected at 0.5, 1, 2, 4, 8, 24, 48, 72, and 96 h postdose (n = 3 mice per time point), with a maximum of two samples being obtained from each mouse via tail poke. Blood was collected directly into heparinized polypropylene tubes containing a cocktail of protease inhibitor, potassium fluoride, 1 M acetic acid, and EDTA to minimize the potential for the ex vivo degradation of the compounds in blood/plasma samples. All plasma samples were snap-frozen on dry ice and then stored at −80°C until analysis within 6 weeks. The plasma samples were thawed at room temperature, and proteins were precipitated with freshly prepared 10% dimethylformamide in acetonitrile (DMF-ACN) containing 100-ng/ml ELQ-331 as the internal standard in a ratio of 5 μl of sample to 95 μl of DMF-ACN solution. Clarified supernatants were analyzed via liquid chromatography-tandem mass spectrometry (LC-MS/MS), using an Applied Biosystems (Foster City, CA) Qtrap 4000 mass spectrometer interfaced to a Shimadzu (Columbia, MD) SIL-20AC XR autosampler followed by two LC-20AD XR pumps. The mass spectrometer was operated in the positive electrospray ionization multiple-reaction monitoring mode. Analyte concentrations were determined relative to the concentrations on calibration curves prepared in blank mouse plasma.

After the blood was collected for the 4-h, 24-h, and 96-h time points, 3 mice for each time point were euthanized and the brains were collected. The brains were rinsed with phosphate-buffered saline (PBS) and then blotted dry and flash frozen for storage at −80°C until they could be analyzed concurrently with the plasma samples. The frozen brain was weighed and placed in a homogenizing tube with 3 beads, and then PBS was added at a ratio of 1 g tissue to 5 ml PBS. The sample was homogenized for 30 s using an Omni Bead Ruptor apparatus at room temperature. A 5-μl aliquot of that homogenate was then used for the analysis as described above for plasma; i.e., 5 μl to 95 μl of the DMF-ACN solution containing the internal standard was vortex mixed for 3 min and then centrifuged to clarify and remove the precipitated protein at room temperature at 10,000 × g. The supernatant was transferred to sample vials, preincubated for approximately 45 min at 35°C, and then injected for analysis.

Parasite strains and passage in mice.

ME49 (genotype II) strain parasites were used for all chronic infection experiments but were obtained from two different laboratories. MS-ME49 was obtained from Michael Shaw (University of Michigan), who previously received it from Alan Sher’s lab (National Institutes of Health). CH-ME49 was obtained from Christopher Hunter’s lab (University of Pennsylvania). Both versions of ME49 were maintained in Swiss Webster or CBA/J mice by serial passage at 8- to 12-week intervals. A type I RH T. gondii strain expressing luciferase and green fluorescent protein (GFP) was used for the acute infection model.

Treatment of latent toxoplasmosis.

Unless otherwise noted, all experiments involved 7- to 8-week-old recipient CBA/J female mice infected with 18 cysts of ME49 strain parasites in brain homogenate from a CBA/J donor mouse that had been infected for 5 weeks. Treatment of recipient mice commenced at 5 weeks postinfection and consisted of various schemes, as described below and in the Results section. Experimental compounds were administered either by the intraperitoneal route (in 0.1 ml dimethyl sulfoxide [DMSO]) or via oral gavage (in 0.1 ml PEG 400). The mice were humanely euthanized at 2 weeks following the final injection. The mouse brains were placed in 1 ml sterile PBS and individually minced with scissors, vortexed, and homogenized by passage 3 to 4 times through a 22-gauge needle and syringe. Three 10-μl samples of each brain homogenate were placed under 24- by 24-mm coverslips, and the cysts were enumerated by phase-contrast microscopy in a blind manner, i.e., without the individual doing the enumerating having knowledge of the sample identifications.

Seventeen groups of mice were included in the time course treatment experiments (Fig. 1A and B). Brain cysts in group 1 were enumerated at 5 weeks postinfection (0 weeks of treatment) as a reference for the time course. Groups 2, 5, 8, 11, and 14 were each treated with the vehicle (DMSO) for 1 to 5 weeks. Groups 3, 6, 9, 12, and 15 were each treated with 5-mg/kg ELQ-271 for 1 to 5 weeks. Groups 4, 7, 10, 13, and 16 were each treated with 5-mg/kg ELQ-316 for 1 to 5 weeks. Treatment was administered daily via intraperitoneal injection. The mice in groups 2 to 13 were euthanized 1 day following their last treatment for enumeration of brain cysts. The groups consisted of 5 mice each, except for groups 14 to 16, which included 5, 13, and 10 mice, respectively, because of uneven attrition due to infection. Group 17 contained 4 uninfected mice. Following 6 weeks of treatment, groups 14 to 17 received dexamethasone (10 mg/ml) in their drinking water (changed daily) to elicit immunosuppression for assessing the viability of residual cysts posttreatment.

To assess the distribution of residual cysts in the brain and to test the efficacy of secondary treatment (Fig. 5), groups 1 to 3, consisting of 20 mice each, were infected for 5 weeks and orally treated with vehicle (PEG 400) or 3-mg/kg or 10-mg/kg ELQ-334 daily for 2 weeks. As uninfected negative controls, groups 4 to 6 (5 mice each) were treated with vehicle (PEG 400) or 3-mg/kg or 10-mg/kg ELQ-334 daily for 2 weeks. At 2 weeks following the last treatment, the mice in all groups were humanely euthanized and each brain was split into right and left hemispheres. The right hemisphere was homogenized for cyst counts, measurement of the cyst size by microscopy, and extraction of genomic DNA, as described below. To assess the viability of residual cysts, brain homogenates were pooled within groups and injected into naive 7-week-old female CD1 mice (3 mice per group) at 1, 0.1, 0.01, and 0.001 brain equivalents by mixing with 0, 0.9, 0.99, or 0.999 brain equivalent homogenates, respectively, from groups 4 to 6, i.e., uninfected, treated negative-control mice, to ensure injection of equivalent brain material. The CD1 mice were examined for brain cysts at 5 weeks posttransfer. The left hemisphere was processed as described below. Eighteen cysts in pooled homogenates from groups 1 and 3 were also used to infect naive CBA/J mice (10 mice per group) for secondary treatment. At 5 weeks postinfection, these mice were orally treated with vehicle (PEG 400) or 10-mg/kg ELQ-334 daily for 2 weeks. At 2 weeks following the last treatment, the brain cysts were quantified to measure the efficacy of the secondary treatment.

For determining the sensitivity of CH-ME49 to ELQ-334 treatment (Fig. 5J), mice were infected with CH-ME49. Each group contained 15 mice. At 5 weeks postinfection, the surviving mice were given oral treatment with vehicle (PEG 400) or 10-mg/kg ELQ-334. The brain cysts were enumerated 2 weeks after the last treatment. Statistical analyses of the cyst burden were performed using Mann-Whitney tests. Outliers identified with the Grubb’s test were excluded.

Quantitative analysis of brain slices.

The brains of infected and uninfected mice were dissected, weighed, flash frozen, and stored at −20°C. The tissue was then thawed and sliced into sequential 1-mm sections using a model 51425 tissue slicer (Stoelting, Wood Dale, IL). The brain slices were homogenized with 50 μl of PBS in 1.5-ml Eppendorf tubes using a Kontes pellet pestle (catalog no. KT749521-1500; VWR), followed by passaging through a 17-gauge needle attached to a 1-ml syringe. Standard samples were generated by adding T. gondii tachyzoites to 25-mg samples of uninfected brain tissue samples at concentrations ranging from 0.1 to 5 × 10⁸ parasites/mg tissue. Whole genomic DNA was extracted from the samples and standards and purified using a DNeasy blood and tissue kit from Qiagen (catalog no. 69504) according to the manufacturer’s instructions. Fifty nanograms of genomic DNA was used per sample to run quantitative PCR (qPCR) using primers against a Toxoplasma-specific 529-bp repeat element (primer Tox-9 Forward [5′-AGGAGAGATATCAGGACTGTAG] and primer Tox-11 Reverse [5′-GCGTCGTCTCGTCTAGATCG]). All samples were quantified by qPCR using a Bio-Rad SYBR green master mix reagent (catalog no. 4309155; Thermo Fisher Scientific) on an Applied Biosystems Step One Plus qPCR instrument. Treated and control sample threshold cycle (C_T) values were compared against the values on a standard curve generated from qPCR analysis of samples with defined numbers of parasites with the same primers to extrapolate the number of T. gondii genomes per brain slice. Samples with C_T values above 40 were considered to have no detectable T. gondii genomes present.

Treatment of acute toxoplasmosis.

CF-1 mice that were 4 to 6 weeks old were inoculated intraperitoneally with 10,000 virulent type I RH T. gondii tachyzoites that expressed firefly luciferase and GFP for bioluminescence imaging using a previously established methodology (33, 34). After 24 h, the compounds were dissolved and administered in PEG 400 via oral gavage daily for 5 days. The vehicle-only control groups and treatment groups consisted of 4 mice per group. Mice were monitored for signs of infection and underwent bioluminescence imaging on days 4, 6, 13, and 29. Mice were injected i.p. with a dose of 0.1 ml of d-luciferin (150 mg substrate/kg of body weight) dissolved in PBS. At 3 min after luciferin injection, the mice were anesthetized using inhaled isoflurane and positioned ventral side up on a heated platform. Bioluminescent images were obtained using an IVIS SpectrumCT imaging system and processed using Living Image software (Perkin Elmer). Mice were humanely euthanized if they developed signs of severe infection, such as a >10% weight loss, lethargy, or a lack of self-grooming, or at 33 days. Analysis of survival and the differences in the tissue burden of T. gondii infection was performed using a log-rank test and an unpaired t test, respectively. GraphPad Prism (version 7.0) software was used for statistical analysis.

Cytochrome b gene sequencing.

DNA was isolated from 6 samples of T. gondii MS-ME49 cysts, CH-ME49 cysts, and RH strain tachyzoites using a DNeasy blood and tissue purification kit (Qiagen). The cytochrome b-coding sequences were amplified from genomic DNA and cDNA by PCR with primers 5′-ATGGTTTCGAGAACACTCAGT and 3′-GTATAAGCATAGAACCAATCCGGT and Phusion DNA polymerase, yielding a single PCR product visualized on an agarose gel. Control PCRs without DNA did not yield PCR products. Amplicons were sequenced using sequencing primers 5′-CTACCATGGGGACAAATGAGTTTCTGGGGTGCTACAGT and 3′-ACCATTCTGGTACGATATGAAGTGGTGTTAC. Protein alignment was performed with the MUSCLE (multiple-sequence comparison by log expectation) program.

Ethics.

All animal procedures and protocols were carried out in strict accordance with the Public Health Service policy on the humane care and use of laboratory animals and Association for Assessment and Accreditation of Laboratory Animal Care guidelines. The University of Michigan Committee on the Use and Care of Animals (animal welfare assurance number A3114-01, protocol number PRO00008638) and the Institutional Animal Care and Use Committee (protocol number 3276) of the Portland Veterans Administration Medical Center approved the animal protocol used for this study. All efforts were made to minimize pain and suffering.