Potent Tetrahydroquinolone Eliminates Apicomplexan Parasites

ORIGINAL RESEARCH

published: 09 June 2020

doi: 10.3389/fcimb.2020.00203

Potent T e trahydroquinolone

Eliminates Apicomplexan

Parasites

Martin J. McPhilli e^1†, Ying Zho u^2†, Mark R. Hickman^3†‡, James A. Gordon^1†‡

Christopher R. Weber^4†, Qigui Li³, Patty J. Lee³, Kangsa Amporndanai⁵,

Rachel M. Johnson^6‡, Heather Darby¹, Stuart Wood s⁷, Zhu-hong L i⁸,

Edited by:

Jeroen P. J. Saeij,

,

University of California, Davis,

United States

Reviewed by:

Louis Weiss,

Richard S. Priestley⁹, Kurt D. Ristrop h¹⁰, Scott B. Bierin g^4‡, Kamal El Bissat i²,

Seungmin Hwan g^4‡, Farida Esaa Hakim², Sarah M. Dovgin^2‡, Joseph D. Lykin s^2‡

Lucy Roberts⁷, Kerrie Hargrav e^7‡, Hua Con g^1‡, Anthony P. Sinai¹¹, Stephen P . M uenc h⁶,

Jitender P. Dubey¹², Robert K. Prud’homme¹⁰, Hernan A. Lorenz i^13†, Giancarlo A. Biagin i⁹,

,

Albert Einstein College of Medicine,

United States

Renato Augusto DaMatta,

State University of the North Fluminense

Darcy Ribeiro, Brazil

8

7

Silvia N. Moreno , Craig W. Roberts , Svetlana V. Antonyuk , Colin W. G. Fishwick and

5

1

*

Rima McLeo d^2,14

*

*Correspondence:

Silvia N. Moreno

¹School of Chemistry, The University of Leeds, Leeds, United Kingdom, ²Department of Ophthalmology and Visual

Sciences, The University of Chicago, Chicago, IL, United States, ³Experimental Therapeutics Branch, Walter Reed Army

Institute of Research, Silver Spring, MD, United States, ⁴Department of Pathology, The University of Chicago, Chicago, IL,

United States, ⁵Department of Biochemistry and Systems Biology, Faculty of Health and Life Sciences, Institute of Systems,

Molecular and Integrative Biology, The University of Liverpool, Liverpool, United Kingdom, ⁶School of Biomedical Sciences,

Faculty of Biological Sciences, and Astbury Centre for Structural Molecular Biology, The University of Leeds, Leeds,

United Kingdom, ⁷Strathclyde Institute of Pharmacy and Biomedical Sciences, The University of Strathclyde, Glasgow,

United Kingdom, ⁸Department of Cellular Biology, Center for Tropical and Emerging Global Diseases, University of Georgia,

Athens, GA, United States, ⁹Department of Tropical Disease Biology, Research Center for Drugs and Diagnostics, The

Liverpool School of Tropical Medicine, Liverpool, United Kingdom, ¹⁰Department of Chemical and Biological Engineering,

Princeton University, Princeton, NJ, United States, ¹¹Microbiology, Immunology and Molecular Genetics, The University of

Kentucky College of Medicine, Lexington, KY, United States, ¹²Animal Parasitic Diseases Laboratory (APDL), USDA-ARS,

Beltsville, MD, United States, ¹³Department of Infectious Diseases, J Craig Venter Institute, Rockville, MD, United States,

¹⁴Department of Pediatrics (Infectious Diseases), Institute of Genomics, Genetics, and Systems Biology, Global Health

Center, Toxoplasmosis Center, CHeSS, The College, University of Chicago, Chicago, IL, United States

smoreno@uga.edu

Svetlana V. Antonyuk

S.Antonyuk@liverpool.ac.uk

Colin W. G. Fishwick

C.W.G.Fishwick@leeds.ac.uk

Rima McLeod

rmcleod@bsd.uchicago.edu

†

These authors have contributed equally

to this work

‡

Present address:

Mark R. Hickman,

Fort Detrick, Frederick, MD, United States

James A. Gordon,

Charles River, Cambridge, United Kingdom

Rachel M. Johnson,

Monash University, Melbourne, VIC,

Australia

Apicomplexan infections cause substantial morbidity and mortality, worldwide. New,

improved therapies are needed. Herein, we create a next generation anti-apicomplexan

lead compound, JAG21, a tetrahydroquinolone, with increased sp3-character to improve

parasite selectivity. Relative to other cytochrome b inhibitors, JAG21 has improved

solubility and ADMET properties, without need for pro-drug. JAG21 signiﬁcantly reduces

Toxoplasma gondii tachyzoites and encysted bradyzoites in vitro, and in primary and

established chronic murine infections. Moreover, JAG21 treatment leads to 100%

survival. Further, JAG21 is efﬁcacious against drug-resistant Plasmodium falciparum

in vitro. Causal prophylaxis and radical cure are achieved after P. berghei sporozoite

infection with oral administration of a single dose (2.5 mg/kg) or 3 days treatment at

reduced dose (0.625 mg/kg/day), eliminating parasitemia, and leading to 100% survival.

Enzymatic, binding, and co-crystallography/pharmacophore studies demonstrate

selectivity for apicomplexan relative to mammalian enzymes. JAG21 has signiﬁcant

promise as a pre-clinical candidate for prevention, treatment, and cure of toxoplasmosis

and malaria.

Scott B. Biering,

University of California at Berkeley,

Berkeley, CA, United States

Seungmin Hwang,

VIR Biotechnology, San Francisco, CA,

United States

Sarah M. Dovgin,

Case Western Reserve University,

Cleveland, OH, United States

Joseph D. Lykins,

Virginia Commonwealth University Health

System, Richmond, VA, United States

Kerrie Hargrave,

University of Glasgow, Glasgow,

United Kingdom

Hua Cong,

Shandong University, Shandong, China

Specialty section:

This article was submitted to

Parasite and Host,

a section of the journal

Frontiers in Cellular and Infection

Microbiology

Keywords: Toxoplasma gondii, Plasmodium falciparum, cytochrome bc1, tetrahydroquinolone, nanoformulation,

structure-guided design, transcriptomics, RPS131

Frontiers in Cellular and Infection Microbiology | www.frontiersin.org

1

June 2020 | Volume 10 | Article 203

McPhillie et al.

Potent Tetrahydroquinolone Eliminates Apicomplexan Parasites

INTRODUCTION

have trigonal planar and tetrahedral geometries, respectively.

Flat aromatic rings (“sp2-rich”) are ubiquitous in drug discovery

campaigns, but molecules with more “sp3 character” are often

more speciﬁc for their protein target and can have better

physicochemical properties. Further, we reasoned that the larger

binding pocket in the parasite enzymes (McPhillie et al., 2016),

compared to their mammalian counterparts, would provide

room for bulkier substituents to minimize eﬀect on the human

enzyme. Within this new series of compounds, we aimed to

identify a mature lead compound with both anti-Plasmodium and

anti-T. gondii activity.

Our work developed as follows: We recently found markedly

increased expression of cytochrome b in the currently untreatable

T. gondii bradyzoite life-cycle stage (McPhillie et al., 2016).

Thus, we set out to develop a compound that would inhibit

tachyzoites, bradyzoites, and three life cycle stages of even

drug-resistant Plasmodia. We sought to do this without a

need for a pro-drug as has been needed in other attempts to

target apicomplexan cytochrome b (Frueh et al., 2017). Our

aim was to improve upon the physicochemical properties of

napthoquinones and endochin-like quinolones (ELQs) targeting

cytochrome b, including poor aqueous solubility and toxicity

(Khan et al., 1998; Doggett et al., 2012; Capper et al., 2015; Miley

et al., 2015; McPhillie et al., 2016). The intent was further to

provide potential solutions for limitations of other compounds

active against apicomplexan parasites (Waxman and Herbert,

1969; Caumes et al., 1995). Our concurrent crystallographic

studies also enable better understanding of the interactions

between ligand and the binding pocket of the Q_isite (McPhillie

et al., 2016).

Malaria results in the death of ∼0.5 million children a year,

with drug resistance impacting the usefulness of successive

generations of new medicines (www.who.int/malaria/

publications/world-malaria-report-2017/en/).

The

parasitic infection of humans in the world. It plays a signiﬁcant

role in food-borne associated death in the USA, destruction

of the human retina (Phan et al., 2008), and death and illness

from recrudescent disease in the immune compromised or

immunologically immature (McLeod et al., 2006; McLeod and

Boyer, 2019). It has been estimated that there are 1.9 million

new cases of this congenital T. gondii infection globally over

a 10 year period, causing 12 million disability adjusted life

years (Torgerson and Mastroiacovo, 2013) from damage to

the fetal brain and eye. Toxoplasmosis is an often neglected,

untreated, or mistreated disease. There are ∼2 billion people

throughout the world who have this parasite in their brain

lifelong, some with known, severe, adverse consequences (Delair

et al., 2011; Wallon et al., 2013; Lykins et al., 2016). There are

possible additional, harmful eﬀects for a substantial number of

chronically infected people as this parasite modulates signature

pathways of neurodegeneration, motor diseases, epilepsy, and

malignancies (Ngô et al., 2017). No medicine eliminates this

chronic, encysted form of the parasite. New and improved

medicines are greatly needed to cure Toxoplasma and Plasmodia

infections (McLeod et al., 2006). These parasites often share the

same molecular targets for medicines due to a relatively close,

apicomplexan, phylogenetic relationship (McPhillie et al., 2016).

Thus, medicine development for each of these parasites can

inform development of medicines that beneﬁt treating the other

(Muench et al., 2007; Fomovska et al., 2012a).

One such shared molecular target is the mitochondrial

cytochrome bc1 complex that is important for the survival

of apicomplexan parasites such as Plasmodia and T. gondii.

Cytochrome b is a subunit of the cytochrome bc₁complex,

an inner mitochondrial membrane protein that is part of the

electron transport chain. Activity of this complex is integral to

oxidative phosphorylation and generation of ATP (Vercesi et al.,

1998). Cytochrome b activity appears to be necessary for the

replication and persistence of the parasite (McPhillie et al., 2016),

and is the site of action of atovaquone (McPhillie et al., 2016).

Cytochrome b is the target for quinolone-based compounds, but,

signiﬁcant problems with solubility, and toxicity have been noted

with earlier cytochrome b inhibitors. In an attempt to design

novel quinolone-like inhibitors with improved solubility, and

lower toxicity, compared to known compounds in the literature,

we synthesized a series of tetrahydroquinolinones (THQs). Our

preliminary eﬀorts were described in McPhillie et al. (2016). We

reasoned that the increased “sp3” character of the THQs (i.e.,

moving from rod-like to sphere-like 3D space) could provide

the required improvement in solubility that would allow for

optimal pharmacokinetic properties. Molecules with an increased

percentage of “sp3 character” tend to be more three-dimensional,

than their planar (“sp2-rich”) counterparts. The terms “sp2” and

“sp3” refer to the shape of their hybridized atomic orbitals, which

Herein, we have identiﬁed a preclinical lead candidate

based on potent and selective inhibition of Plasmodium

falciparum, Plasmodium berghei, and T. gondii cytochrome bc1

for the treatment of malaria and toxoplasmosis. The candidate

compound demonstrates high eﬃcacy in relevant in vitro and

in vivo models of the diseases, and has considerable potential

for broad-spectrum use (i.e., against T. gondii tachyzoites

and encysted bradyzoites, and drug resistant Plasmodia). The

data which follow present the creation and characterization of

this novel, broad-spectrum, anti-apicomplexan lead compound

which has promise for deﬁnitive treatment of these infections.

MATERIALS AND METHODS

Syntheses of Compounds

Synthesis of Tetrahydroquinolones (THQs)

Compounds

The THQ compounds were synthesized at the University of

Leeds as described below. Ten millimolars stock solutions were

made with 100% Dimethyl Sulfoxide (DMSO) [Sigma Aldrich]

and working concentrations were made with IMDM-C (1x, [+]

glutamine, [+] 25 mM HEPES, [–] Phenol red, 10% FBS)[Gibco,

Denmark]). Compounds are shown in Figure 1A. Compound

name with “0” or no “0” between letters and number, e.g., JAG21

or JAG021, refer to the same compound. This is throughout

the manuscript. Final compounds had >95% purity determined

Frontiers in Cellular and Infection Microbiology | www.frontiersin.org

2

June 2020 | Volume 10 | Article 203

McPhillie et al.

Potent Tetrahydroquinolone Eliminates Apicomplexan Parasites

by high performance liquid chromatography (HPLC), high

resolution mass spectrometry, and NMR spectrometry. Liquid

chromatography-mass spectrometry (LC-MS) and NMR

spectrometry were used to determine the integrity and purity

of all intermediates. THQ compounds were synthesized as

described in Schemes 1, 2, which describe compounds MJM170

and JAG21 as exemplars. Building blocks 1, 8, 9, and 14 were

varied to create the complete series (Figure 1A).

Synthesis of 2-Methyl-3-Iodo-4-Ethoxy-5,6,7,8-

Tetrahydroquinoline (4)

Potassium carbonate (1.53 g, 11.1 mmol, 2.00 eq) was added

to

a heterogeneous mixture of 2-methyl-3-iodo-5,6,7,8-

tetrahydroquinolin-4-one (3, 1.60 g, 5.56 mmol, 1.00 eq) in DMF

(15.0 mL), and the reaction heated to 50^◦C for 30 min. The R.B.

ﬂask was removed from the heating mantle and ethyl iodide

(0.67 mL, 8.33 mmol, 1.50 eq) was added dropwise. The reaction

was then heated at 50^◦C for 18 h. The reaction was cooled to

R.T., quenched with water (40 mL). The resulting emulsion

formed which was extracted with EtOAc (50 mL). EtOAc layer

were washed with water (3 × 30 mL), brine (3 × 30 mL), dried

(Na₂SO₄) and concentrated to give a pale yellow oil (1.09 g, 3.44

mmol, 61%); R_f0.88 (1:1 Pet–EtOAc); HPLC (RT = 1.67 min);

LCMS (Method A), (RT = 1.6 min, m/z (ES) Found MH⁺318.0);

Synthesis of

2-Methyl-5,6,7,8-Tetrahydroquinolin-4-one (2)

Platinum oxide (0.100 g, 10 mol %) was added to a solution of

4-hydroxy-2-methylquinoline (1, 1.00 g, 6.28 mmol, 1.00 eq) in

glacial acetic acid (10.0 mL). The heterogeneous mixture was

catalytically hydrogenated under a balloon of hydrogen. After

22 h, TLC (10% MeOH–DCM) conﬁrmed complete reaction.

The mixture was ﬁltered through celite under vacuum, washing

thoroughly with EtOAc (50 mL). The ﬁltrate was concentrated

and the resulting residue puriﬁed by column chromatography

(10% MeOH–DCM) to give the desired product as a pale yellow

δ_H(500 MHz, CDCl₃) 1.49 (3H, t, J 7.0, ethoxy CH₃), 1.73–1.78

(2H, m, CH₂) 1.84–1.88 (2H, m, CH₂), 2.78–2.69 (5H, m, CH₂

& CH₃), 2.84 (2H, t, J 6.5, CH₂), 3.97 (2H, q, J 7.0, OCH₂); δ

C

(125 MHz, CDCl₃) 15.6 (CH₃), 22.3 (CH₂), 22.8 (CH₂), 23.6

(CH₂), 29.3 (CH₃), 32.0 (CH₂), 68.4 (OCH₂), 90.9 (Cq), 124.5

(Cq), 158.3 (Cq), 158.9 (Cq), 163.9 (Cq).

Synthesis of 2-Methyl-3-(4-Phenoxyphenyl)-4-

Ethoxy-5,6,7,8-Tetrahydroquinoline (6)

oil (0.917 g, 5.65 mmol, 89%); R 0.14 (10% MeOH–DCM); δ

(300 MHz, CDCl₃) 1.74–1.76 (4H, m, CH₂), 2.29 (3H, s, Me),

f

H

2.49–2.52 (2H, m, CH₂), 2.67–2.70 (2H, m, CH₂), 6.16 (1H, s,

Ar-H); δ (125 MHz, CDCl₃) 19.0 (Me), 21.8 (CH₂), 22.1 (CH₂),

C

27.1 (CH₂), 112.5 (CH), 122.4 (Cq), 146.4 (Cq), 147.0 (Cq),

178.3 (Cq); Spectroscopic data consistent with literature values

(Bradbury et al., 1993).

2-Methyl-3-iodo-4-ethoxy-5,6,7,8-tetrahydroquinoline

(4,

0.266 g, 0.839 mmol, 1.00 eq), Pd(PPh₃)₄(0.048 g, 0.0419 mmol,

5 mol%) and 4-phenoxyphenylboronic acid (5, 0.270 g, 1.26

mmol, 1.50 eq) were charged to a R.B. ﬂask under N₂(g).

Degassed DMF (10.0 mL) was added to the ﬂask followed by 2M

K₂CO₃(1.60 mL). The ﬂask was heated to 85^◦C under N₂(g).

After 15 min, TLC (4:1 Pet–EtOAc) conﬁrmed reaction was

complete. The reaction was cooled and diluted with EtOAc

(15 mL), ﬁltered through celite and partitioned between EtOAc

(10 mL) and H₂O (25 mL). Combined organics were washed

with H₂O (3 × 30 mL), then brine (3 × 30 mL), dried (Na₂SO₄)

and concentrated to give a red oil which was puriﬁed by column

chromatography (3:1 Pet–EtOAc), to give the desired product

Synthesis of

2-Methyl-3-iodo-5,6,7,8-Tetrahydroquinolin-4-One (3)

ⁿButylamine (6.20 mL, 62.8 mmol, 10.0 eq) was added to

a suspension of 2-methyl-5,6,7,8-tetrahydroquinolin-4-one (2,

1.02 g, 6.28 mmol, 1.00 eq) in DMF (10.0 mL). To this

heterogeneous mixture was added I₂(1.60 g, 6.28 mmol, 1.00 eq)

in a saturated solution of KI (6.00 mL). After 20 h stirring at R.T.,

a precipitate formed in the orange solution, excess iodine was

quenched with 0.1 M sodium thiosulfate solution (10.0 mL). The

precipitate was ﬁltered by vacuum ﬁltration, washed with distilled

H₂O and dried (Na₂SO₄) to give the desired product as a colorless

as a pale yellow oil (0.235 g, 0.655 mmol, 78%); R 0.31 (3:1

f

Pet–EtOAc); HPLC (RT = 3.08 min); δ (300 MHz, CDCl₃)

H

1.04 (3H, t, J 7.0, ethoxy CH₃), 1.76–1.93 (4H, m, 2xCH₂), 2.32

(3H, s, CH₃) 2.72 (2H, t, J 6.0, CH₂), 2.91 (2H, t, J 6.5, CH₂),

3.50 (2H, q, J 7.0, OCH₂), 7.05–7.16 (5H, m, Ar-H), 7.20–7.29

(2H, m, Ar-H), 7.31–7.43 (2H, m, Ar-H); δ (125 MHz, CDCl₃)

C

solid (1.76 g, 6.09 mmol, quantative yield); δ (300 MHz, DMSO-

d₆) 1.61–1.70 (4H, m, CH₂), 2.29 (2H, t, J 6.0, CH₂), 2.43 (2H, s,

CH₂), CH₃under DMSO peak.

15.7 (CH₃), 22.5 (CH₂), 23.0 (CH₃), 23.3 (CH₂), 23.4 (CH₂),

32.7 (CH₂), 68.2 (OCH₂), 118.6 (CH), 118.9 (CH), 123.4 (CH),

126.8 (Cq), 129.8 (CH), 131.5 (CH), 154.9 (Cq), 156.5 (Cq), 157.1

H

Frontiers in Cellular and Infection Microbiology | www.frontiersin.org

3

June 2020 | Volume 10 | Article 203

McPhillie et al.

Potent Tetrahydroquinolone Eliminates Apicomplexan Parasites

SCHEME 1 | Synthesis of hit compound 7, also known as MJM170 (McPhillie et al., 2016). Synthetic scheme inspired by the route to endochin-like quinolones (ELQs)

reported by Doggett et al. (2012).

(Cq), 157.3 (Cq); m/z (ES) (Found: MH⁺, 360.1973. C₂₄H₂₆NO₂

requires MH, 360.1964).

(CH₂), 21.8 (CH₂), 21.9 (CH₂), 26.2 (CH₂), 117.8 (CH), 118.6

(CH), 121.2 (Cq), 123.3 (CH), 123.7 (Cq), 130.0 (CH), 131.4 (Cq),

132.3 (CH), 142.3 (Cq), 143.2 (Cq), 155.0 (Cq), 156.8 (Cq), 175.4

(Cq); m/z (ES) (Found: MH⁺, 332.1654. C₂₂H₂₂NO₂requires

MH, 332.1645).

Synthesis of 2-Methyl-3-(4-Phenoxyphenyl)-4-

Ethoxy-5,6,7,8-Tetrahydroquinoline (7,

MJM170)

Synthesis of

1-(4-Bromophenyl)-4-(Triﬂuoromethoxy)Benzene (10)

Aqueous hydrobromic acid (>48%) (1.00 mL) was added to

a solution of 2-methyl-3-(4-phenoxyphenyl)-4-ethoxy-5,6,7,8-

tetrahydroquinoline (6, 0.226 g, 0.630 mmol, 1.00 eq) in glacial

acetic acid (2 mL). The reaction was stirred at 90^◦C for 5 days,

monitoring by LMCS. The reaction was cooled to R.T. and the pH

adjusted to pH 5 with 2M NaOH. The precipitate was collected

by vacuum ﬁltration and recrystallized from MeOH:H₂O to give

the desired product as an oﬀ-white solid (0.155 g, 0.467 mmol,

Copper (II) acetate (0.435 g, 2.39 mmol, 1.00 eq) was added to

a suspension of 4-bromophenol (8, 0.414 g, 2.39 mmol, 1.00 eq),

4-triﬂuoromethoxybenzeneboronic acid (9, 0.983 g, 4.79 mmol,

2.00 eq) and 4 Å molecular sieves (0.566 g) in DCM (12 mL)

at R.T. A solution of triethylamine (1.7 mL, 11.9 mmol, 5.00

eq) and pyridine (1 mL, 11.9 mmol, 5.00 eq) was added and

the reaction was stirred for 16 h, open to the atmosphere. After

18 h, the reaction was quenched with 0.5 M HCl (20 mL) and the

organic layer washed with water (20 mL), brine (20 mL), dried

(Na₂SO₄), and concentrated to give a red oil which was puriﬁed

by column chromatography (hexane) to give the desired product

74%); HPLC (RT = 2.56 min); δ (500 MHz, DMSO-d₆) 1.66–

H

1.72 (4H, m, 2xCH₂), 2.08 (3H, s, CH₃) 2.31 (2H, t, J 6.0, CH₂),

2.56 (2H, t, J 6.0, CH₂), 6.99 (2H, d, J 8.5, Ar-H), 7.06 (2H, d, J

7.5, Ar-H), 7.14–7.18 (3H, m, Ar-H), 7.40–7.43 (2H, m, Ar-H),

11.0 (1H, s, NH); δ (125 MHz, DMSO-d₆) 17.7 (CH₃), 21.5

as a colorless oil (0.582 g, 1.75 mmol, 73%); R 0.58 (hexane).

C

f

Frontiers in Cellular and Infection Microbiology | www.frontiersin.org

4

June 2020 | Volume 10 | Article 203

McPhillie et al.

Potent Tetrahydroquinolone Eliminates Apicomplexan Parasites

SCHEME 2 | Synthetic route to analogs of 7 (MJM170) via route A or route B. Route A is the original route to analogs but is linear and involves a tricky Suzuki step to

intermediate 12 from intermediates 4 and 10. Route B allows quicker access to analogs since intermediate 15 can be made in larger quantities and derivatives can be

synthesized via the Chan-Lam reaction to give ﬁnal intermediate 12 by varying the boronic acid 16.

Synthesis of 2-Methyl-3-(4-Hydroxyphenyl)-4-

Ethoxy-5,6,7,8-Tetrahydroquinolin-4-One (15)

mmol, 1.50 eq) were charged to a R.B. ﬂask under N₂(g).

Degassed DMF (10.0 mL) was added to the ﬂask followed by 2M

K₂CO₃(3.00 mL). The ﬂask was heated to 85^◦C under N₂(g).

After 3 h, TLC (EtOAc) conﬁrmed reaction was complete. The

reaction was cooled to 50^◦C, diluted with EtOAc (15 mL) and

activated charcoal was added. After stirring for 30 min, the

mixture was ﬁltered through celite and partitioned between

EtOAc (10 mL) and H₂O (25 mL). Combined organics were

washed with H₂O (3 × 30 mL), then brine (3 x 30 mL), dried

(Na₂SO₄) and concentrated to give a brown solid which was

triturated with diethyl ether to give the desired product as a

2-Methyl-3-iodo-4-ethoxy-5,6,7,8-tetrahydroquinoline

0.400 g, 1.26 mmol, 1.00 eq), Pd(PPh₃)₄(0.073 g, 0.06 mmol,

5 mol%) and 4-hydroxylphenylboronic acid (14, 0.260 g, 1.89

(4,

Frontiers in Cellular and Infection Microbiology | www.frontiersin.org

5

June 2020 | Volume 10 | Article 203

McPhillie et al.

Potent Tetrahydroquinolone Eliminates Apicomplexan Parasites

FIGURE 1 | Characteristics and effects of compounds on inhibition of Toxoplasma gondii replication and enzyme activity, and Structure Activity Relationship analysis.

(A) Cytochrome b/c inhibitor code, Chem Draw structure, solubility in PBS 7.4, toxicity against HFF, predicted half-life, and inhibitory effect of compounds on RH strain

tachyzoites and EGS strain bradyzoites in vitro and saffarine O assay enzyme activity. PBS Sol/Toxicity pH7.4 refers to solubility of the compound in Phosphate

Bufferred Saline (PBS) at pH 7.4. Toxicity refers to the highest concentration tested that does not show toxicity to Human Foreskin Fibroblast (HFF) in tissue culture in

WST assay; T1/2 (H) refers to the predicted half-life in human liver microsomes; T1/2 (M) refers to the predicted half-life in mouse liver microsomes. Tachy/Brady IC50

(Continued)

Frontiers in Cellular and Infection Microbiology | www.frontiersin.org

6

June 2020 | Volume 10 | Article 203

McPhillie et al.

Potent Tetrahydroquinolone Eliminates Apicomplexan Parasites

FIGURE 1 | was determined in studies in which cultures of parasites in HFF were treated with varying concentrations of the compound and there was 50% inhibition

of the replication (number) of parasites. Parasites were RH-YFP expressing tachyzoites (Tachy) and EGS (Brady) strains. Studies of effects of inhibitors on HFF or on T.

gondii tachyzoites were performed with triplicate wells in at least 2 biological replicate experiments. Studies of effects on bradyzoites were performed at least twice in

at least 2 biological replicate experiments. Compounds with much less inhibition of mammalian than T. gondii cytochrome bc, relative to JAG21 effect on parasite

enzyme, in the saffarine enzyme assay (indicated by **) provide potential to further develop compounds, if unanticipated toxicity occurs from JAG21. (B) Structure

Activity Relationship analysis (SAR). The effects of changing R1 as 7-Et, 7-Me, 6-CF₃, or 6-Me on activity against T. gondii RH strain tachyzoites, solubility, and stability

were compared in the SAR. Color Key in (B) Activity: Green <50 nM, Red > 1 µM; Solubility in 100 mM Phosphate Buffer (pH 7.4): Amber>10 µM, Red <10 µM;

Metabolic Stability: Green >120 min, Amber 60–120 min, red < 60 min. SAR panel displays only representative structures and trends within the JAG compound

series. JAG21 (blue font) is highly active, has the longest predicted half-life for humans of initial compounds tested (green), combined with improved solubility, no

hERG liability, and predicted capacity to cross the blood brain barrier (BBB). Deﬁnitions of ADMET terminology are in the Materials and Methods. In summary, in the

SAR overall, nitrogen atoms were not tolerated in aryl ring marked by green c, and the 4-position was optimal for phenol substituent. Compound name with “0” or no

“0” between letters and number, e.g., JAG21 or JAG021, refer to the same compound. This is throughout the manuscript.

pale red crystalline solid (0.220 g, 0.777 mmol, 60%); R 0.22

132.0 (CH), 126.7 (Cq), 123.0 (CH), 119.8 (CH), 119.0 (CH),

68.0 (OCH₂), 32.5 (CH₂), 23.0 (CH₂), 22.9 (CH₃), 22.7 (CH₂),

22.5 (CH₂), 15.05 (CH₃); m/z (ES) (Found: MH⁺, 444.1792.

C₂₅H₂₄F₃NO₃requires MH, 444.1781).

f

(EtOAc); m.p. 225–226^◦C (EtOAc); δ (500 MHz, MeOD-d₄)

H

7.07 (d, J = 8.6 Hz, 2H, H-3 & 5), 6.86 (d, J = 8.6 Hz, 2H, H-2 &

6), 3.51 (q, J = 7.0 Hz, 2H, CH₃CH₂O), 2.83 (t, J = 6.3 Hz, 2H,

H-8’), 2.72 (t, J = 6.1 Hz, 2H, H-5’), 2.23 (s, 3H, Me), 1.95–1.72

(m, 4H, H-6’ & 7’), 1.00 (t, J = 7.0 Hz, 3H, CH₃CH₂O); δ (125

Synthesis of

C

MHz, MeOD-d₄) 164.0 (Cq), 158.1 (C-1), 157.4 (C_q), 156.1 (C_q),

132.2 (C-3 & 5), 129.1 (Cq), 127.9 (Cq), 124.9 (Cq), 116.2 (CH),

69.1 (OCH₂), 32.7 (CH₂), 23.9 (CH₂), 23.4 (CH₃), 22.9 (CH₂),

22.3 (CH₂), 15.7 (CH₃); m/z (ES) (Found MH⁺, 284.1664,

C₁₈H₂₁NO₂requires MH, 284.1651).

2-Methyl-3-(4-(4-(Triﬂuoromethoxy)Phenoxy)Phenyl)-

5,6,7,8-Tetrahydroquinolin-4-One (13,

JAG21)

Synthesis of 2-Methyl-3-(4-Hydroxyphenyl)-4-

Ethoxy-5,6,7,8-Tetrahydroquinolin-4-One

(12)

Aqueous hydrobromic acid (>48%) (1.00 mL) was added to

a solution of 2-methyl-3-(4-phenoxyphenyl)-4-ethoxy-5,6,7,8-

tetrahydroquinoline (12, 30.0 mg, 0.07 mmol, 1.00 eq) in glacial

acetic acid (2 mL). The reaction was stirred at 90^◦C for 3 days,

monitoring by LMCS. The reaction was cooled to R.T. and the pH

adjusted to pH 5 with 2M NaOH. The precipitate was collected by

vacuum ﬁltration and recrystallized from MeOH:H₂O to give the

desired product as a colorless solid (25.0 mg, 0.06 mmol, 68%);

1-(4-bromophenyl)-4-(triﬂuoromethoxy)benzene (10, 0.100 g,

0.30 mmol, 1.00 eq), bis(pinacolato) diboron (1.10 eq), potassium

acetate (3.00 eq) and Pd(dppf)Cl₂(0.03 eq) were added to a

oven-dried ﬂask under inert (N₂) atmosphere. Anhydrous DMF

(6 mL) was added and the reaction heated to 80^◦C under N₂(g).

After 22 h, the reaction was cooled to R.T., fresh Pd(dppf)Cl₂

(0.03 eq) added, followed by 2-methyl-3-iodo-4-ethoxy-5,6,7,8-

tetrahydroquinoline (4, 0.400 g, 1.26 mmol, 2.00 eq) and 2M

Na₂CO₃(2.9 mL). The reaction was heated to 80^◦C for 20 h,

cooled, diluted with EtOAc (20 mL), ﬁltered through celite

and partitioned between EtOAc (20 mL) and H₂O (20 mL).

Combined organics were washed with brine (3 × 20 mL), dried

(Na₂SO₄) and concentrated to give a brown solid which was

puriﬁed by column chromatography (3:1 Pet–EtOAc) to give

the desired product as a colorless oil (30 mg, 0.07 mmol, 23%);

m.p. >250^◦C; HPLC (RT = 2.78 min); δ (500 MHz, DMSO-d₆)

H

11.07 (s, 1H, NH), 7.40 (d, J = 8.5 Hz, 2H, H-2^′& 6^′), 7.19 (d, J =

8.6 Hz, 2H, H-3^′′& 5^′′), 7.13 (d, J = 9.0 Hz, 2H, H-3^′& 5^′), 7.02

(d, J = 8.5 Hz, 2H, H-2^′′& 6^′′), 2.54 (t, J = 6.0 Hz, 2H, H-8), 2.28

(t, J = 5.9 Hz, 2H, H-5), 2.07 (s, 3H, Me), 1.71 (m, 2H, H-7), 1.65

(m, 2H, H-6); δ (125 MHz, DMSO-d₆) 175.7 (Cq), 155.9 (Cq),

C

154.5 (Cq), 143.5 (Cq), 143.2 (Cq), 142.2 (Cq), 132.5 (CH), 132.2

(Cq), 123.6 (Cq), 123.0 (CH), 121.3 (Cq), 119.6 (CH), 118.2 (CH),

26.2 (CH₂), 21.9 (CH₂) 21.8 (CH₂), 21.5 (CH₂), 17.7 (CH₃);

m/z (ES) (Found: MH⁺, 416.1492. C₂₃H₂₀F₃NO₃requires

MH, 416.1473).

Toxoplasma gondii

Parasite Strains (Isolates)

HPLC (RT = 2.41 min); δ (500 MHz, acetone) 7.28 (d, J =

H

8.7 Hz, 2H, H-2^′& 6^′), 7.26 (d, J = 9.1 Hz, 2H, H-2^′′& 6^′′),

7.09 (d, J = 9.1 Hz, 2H, H-3^′′& 5^′′), 7.07 (d, J = 8.7, 2H, H-3^′

& 5^′), 3.52 (q, J = 7.0 Hz, 2H, CH₃CH₂O), 2.85 (t, J = 6.5 Hz,

2H, H-8), 2.78 (t, J = 6.2 Hz, 2H, H-5), 2.26 (s, 3H, Me), 1.89–

1.81(m, 2H, H-7), 1.81–1.72 (m, 2H, H-6), 0.93 (t, J = 7.0 Hz,

RH-YFP tachyzoites (Gubbels et al., 2003; Fomovska et al., 2012a;

McPhillie et al., 2016), EGS strain (Vidigal et al., 2002; Paredes-

Santos et al., 2013, 2018; McPhillie et al., 2016), Pru-luciferase,

Me49, and RPS13ꢀ on the RH strain background (Hutson et al.,

2010) were prepared and passaged in human foreskin ﬁbroblasts

[HFF] as described.

3H, C H₃CH₂O); δ (125 MHz, acetone) δ 161.9 (Cq), 157.1

C

(Cq), 156.5 (Cq), 156.0 (Cq), 154.5 (Cq), 145.3 (Cq), 132.5 (Cq),

Frontiers in Cellular and Infection Microbiology | www.frontiersin.org

7

June 2020 | Volume 10 | Article 203

McPhillie et al.

Potent Tetrahydroquinolone Eliminates Apicomplexan Parasites

T. gondii in vitro

In vitro Challenge Assay for T. gondii

incubator for 72 h, checked once each 24 h. If tachyzoites were

visible in the control before 72 h, cells were ﬁxed and stained.

Cells were ﬁxed using 4% paraformaldehyde and stained

with Fluorescein-labeled Dolichos Biﬂorus Agglutinin, DAPI,

and antibody to BAG1. Disks were removed, mounted on

glass slides, and visualized using microscopy (Nikon Tl7).

Slides were scanned using a CRi Panoramic Scan Whole Slide

Scanner and viewed using Panoramic Viewer Software. Eﬀects

of compounds were quantitated by counting cysts in controls

and treated cultures. Dolichos staining delimited structures and

single organisms that remained were counted in a representative

ﬁeld of view. This was then multiplied by a factor determined by

the total area of the cover slip in order to estimate the number

of cysts and organisms in each condition. When the following

forms were observed: “true cysts” with a dolichos-staining wall,

“pseudocysts” or tight clusters of parasites, and small organisms,

if there were fewer than four parasites visible in a cluster,

organisms were counted individually (as “small organisms”). The

entire scanned coverslip with all ﬁelds was also reviewed by 3

observers to conﬁrm consistency.

RH strain YFP Tachyzoites. Protocol was adapted from

Fomovska et al. (2012a,b) for HFF. HFF were cultured on a

ﬂat, clear-bottomed, black 96-well plate to 90–100% conﬂuence.

IMDM (1x, [+] glutamine, [+] 25 mM HEPES, [+] Phenol red,

10% FBS [Gibco, Denmark]) was removed and replaced with

IMDM-C(1x, [+] glutamine, [+] 25 mM HEPES, [–] Phenol red,

10% FBS)[Gibco, Denmark]). RH-YFP, lysed from host cells by

passing twice through a 27-gauge needle, were counted, then

diluted to 32,000/mL in IMDM-C. HFF were infected with 3200

RH-YFP, then returned to 37^◦C, CO₂(5%) incubator for 1–2 h

for infection. Various concentrations of the compounds in 20 µL

IMDM-C were added to each well. There were triplicates for each

condition. Controls were pyrimethamine/sulfadiazine (standard

treatment), 0.1% DMSO only, HFF only, and untreated cultures

of HFF infected with 2-fold dilutions of YFP expressing parasites

(called “YFP gradient” to establish amount of color from known

numbers of YFP expressing parasites). Cells were incubated at

37^◦C for 72 h. Plates were read using a ﬂuorimeter (Synergy

H4 Hybrid Reader, BioTek) to ascertain amount of relative

ﬂuorescence units (RFU) YFP, to measure parasite burden after

treatment. Data were collected using Gen5 software with IC₅₀

calculated by graphical analysis in Excel.

Synergy Studies With RH Strain YFP Tachyzoites

Atovaquone and pyrimethamine were used to test whether they

are synergistic with JAG21. Serial dilutions of the combination

of JAG21 and either atovaquone or pyrimethamine were used

in an in vitro challenge assay as described above. The EC50

of each compound and the combination of two compounds

were determined. The eﬀect of the combination of drugs was

calculated with the following formula:C= [A]c/[A]a+[B]c/[B]a.

If C is lower than 1, the two compounds tested have synergistic

eﬀect; if C is >1, the two compounds tested have antagonist eﬀect

and if C is 1 they are additive.

An initial screening assay of 10 µM, 1 µM, 100 nM, and

10 nM of the compounds was performed. Compounds were not

considered eﬀective or pursued for further analysis if there was

no inhibition of tachyzoites at 1 µM. If compounds were eﬀective

at 1 µM, another experiment was performed to assess eﬀect at

1 µM, 500, 250, 125, 62.5, and 31.25 nM.

Cytotoxicity Assays in Parallel With RH Strain

T. gondii in vitro Studies

T. gondii and HFF Mitochondrial Membrane

Potential Measurements

Toxicity assays used WST-1 cell proliferation reagent (Roche)

as in Fomovska et al. (2012a). HFF were grown on a ﬂat,

clear-bottomed, black 96-well plate. Conﬂuent HFF were treated

with inhibitory compounds at concentrations of 10 and 50 µM.

Compounds were diluted in IMDM-C, and 20 µL were added

to each designated well, with triplicates for each condition. A

gradient with 2-folddecreasing concentrations of DMSO from 10

to 0% in colorless, translucent IMDM-C was used as a control.

The plate was incubated for 72 h at 37^◦C. Ten microlitre WST-1

reagent (Roche) were added to each well. Cells were incubated for

30–60 min. Absorbance was read using a ﬂuorimeter at 420 nm.

A higher degree of color change (and absorbance) indicated

mitochondrial activity and cell viability.

The mitochondrial membrane potential was measured by the

safranine method according to Vercesi et al. (1998). Freshly

egressed T. gondii tachyzoites were ﬁltered and washed twice

with intracellular buﬀer (125 mM sucrose, 65 mM KCl, 10 mM

HEPES-KOH buﬀer, pH 7.2, 1 mM MgCl₂, and 2.5 mM

potassium phosphate). After washing, the parasites were

resuspended in the same buﬀer at 10⁹/mL. An aliquot of 50 µL

of this suspension was added to a cuvette containing Safranin

O, 2.5 µM and Succinate 1 mM in ﬁnal volume of 2 mL of

the intracellular buﬀer. The ﬂuorescence was measured with

a Hitachi 7000 spectroﬂuorometer with settings Ex. 495/Em.

586. Once the baseline ﬂuorescence was stable, 30 µM digitonin

was added to permeabilize the parasites. Eighty ﬁve seconds

after permeabilization, the THQ derivatives, dissolved in DMSO,

were added. Five micromolars of FCCP (Carbonyl cyanide-4-

(triﬂuoromethoxy) phenylhydrazone) was used as an uncoupler

reference for calculations and its eﬀect was considered 100%.

We used similar conditions for measuring the mitochondrial

membrane potential of mammalian cells with the following

changes: the mammalian cells were resuspended at 10⁸/mL. We

also used 50 µL of this suspension for each experiment in a total

volume of 2 mL. The substrate used for mammalian cells was

In vitro Challenge Assay for EGS Strain Bradyzoites

HFF cells were grown in IMDM on removable, sterile glass

cover slips in the bottom of a clear, ﬂat-bottomed 24-well plate.

Cultures were infected with 3 × 10⁴EGS strain parasites per

well, in 0.5 mL media. The plate was returned to incubator at

37^◦C overnight. The following day, the media was removed.

Colorless IMDM and compounds were added to make various

concentrations of the drug. Total volume was 0.5 mL. Two wells

had media only, as a control. Plates were returned to the 37^◦C

Frontiers in Cellular and Infection Microbiology | www.frontiersin.org

8

June 2020 | Volume 10 | Article 203

McPhillie et al.

Potent Tetrahydroquinolone Eliminates Apicomplexan Parasites

5 mM glutamate and 5 mM malate. A higher concentration of

digitonin (50 µM) was used to permeabilize the mammalian cells.

The compounds were added at ∼400 s after permeabilization.

Each experiment was repeated at least three times in duplicates.

Statistical analysis, unpaired student t-test, was performed using

GraphPad Prism 8.0 (GraphPad Software, Inc., San Diego, CA).

lobe biopsy (Walton et al., 2006) were infected with either wild-

type or RPS13ꢀ RH tachyzoites using biological duplicates at

a multiplicity of infection of 2:1 and incubated as previously

described Ngô et al. (2017). Eighteen hours post-infection,

extracellular parasites were washed out with cold PBS before

total RNA extraction. Further isolation of the mRNA fraction

was carried out with miRNeasy Mini Kit columns (Qiagen)

following manufacturer instructions and Illumina barcoded

mRNA sequencing libraries were constructed with TruSeq RNA

Sample Preparation Kits v2 (Illumina). Libraries were sequenced

as 100 bp single reads with Illumina HiSeq 2000 apparatus at a

sequencing depth of ∼3 Gbp per sample. Sequencing reads were

mapped to the human (release GRCh38) and T. gondii ME49

strain (ToxoDB release 13.0) reference genome assemblies with

hisat2 (Kim et al., 2015) and raw read counts were per gene

were estimated with HTSeq (Anders et al., 2015). Identiﬁcation

of parasite genes that were diﬀerentially expressed between wild-

type and RPS13ꢀ parasites was performed with the R package

DESeq2 (Love et al., 2014) using a generalized linear model

likelihood ratio test. Identiﬁcation of orthologous genes between

T. gondii and P. cynomolgi was carried out by best-reciprocal

matches between T. gondii and P. cynomolgi proteomes using

Blastp and a e-value cutoﬀ of 1 × 10⁻³. The list of Genes that are

diﬀerentially expressed between P. cynomolgi hypnozoites and

the liver-schizont stage was extracted from a previously published

study by Cubi et al. (2017). Gene set enrichment analysis was

carried out with the GSEA tool (Subramanian et al., 2005) using

T. gondii Gene Ontology and cell cycle gene sets developed by

Croken et al. (2014) and visualized with the Enrichment Map

application in Cytoscape (Su et al., 2014).

Structure Activity Relationship (SAR) and

Comparison of Effect on Toxoplasma gondii and HFF

Enzyme Activity

The eﬀects of changing R1 as 7-Et, 7-Me, 6-CF₃, or 6-Me

on activity against RH strain tachyzoites, kinetic solubility,

and metabolic stability were compared. Kinetic solubility and

metabolic stability in human or murine liver microsomes were

measured. The hERG (human Ether-à-go-go-Related) liability

was also determined. The hERG gene (KCNH2) encodes

a protein K_v11.1, the alpha subunit of a potassium ion

channel. This channel conducts the rapid component of the

delayed rectiﬁer potassium current, IKr, which is critical for

repolarization of cardiac action potentials. A reduction in

hERG currents from adverse drug eﬀects can lead to long

QT interval syndromes. These syndromes are characterized by

action potential prolongation, lengthening of the QT interval

on surface EKG, and an increased risk for “torsade de pointes”

arrhythmias and sudden death. The MDCK-MDR1 Permeability

Assay was also performed. MDCK-MDR1 refers to the ability

of a compound to permeate across membranes of MDCK-

MDR1 (Madin Darby canine kidney [MDCK] cells with the

MDR1 gene [ABCB1], the gene encoding for the eﬄux protein,

P-glycoprotein (P-gp)) in vitro. Assessing transport in both

directions (apical to basolateral and basolateral to apical) across

the cell monolayers enables an eﬄux ratio to be determined. This

provides an indication as to whether a compound undergoes

active eﬄux (mediated by P-gp). This provides a prediction of

blood brain barrier (BBB) penetration potential/permeability and

eﬄux ratio. Eﬀect in CACO-2 (Colon Adenocarcinoma cells) as a

permeability assay and on cytochrome P450 (CYP 450) were also

determined. CYP enzymes catalyze oxidative biotransformation

(phase 1 metabolism) of most drugs. CYP enzymes, bind to

membranes in a cell (cyto) and contain a heme pigment (chrome

and P) that absorbs light at a wavelength of 450 nm when exposed

to carbon monoxide. Metabolism of a drug by CYP enzymes is a

major source of variability in drug eﬀect. These were measured by

Chem Partners. The relative eﬀect on HFF and parasite enzymes

also were compared.

Toxoplasma gondii in vivo

Type II Parasites in vivo

IVIS

Balb/C mice were infected intraperitoneally (IP) with 20 × 10³

T. gondii (Prugneaud strain expressing luciferase) tachyzoites.

Treatment began 2 h later with JAG21 (5 mg/kg) which was

dissolved in DMSO, administered IP in a total volume of 0.05 mL.

Mice were imaged every second day starting on day 4 post

infection using an IVIS Spectrum (Caliper Life Sciences) for

minute exposures, with medium binning, 20 min post injection

with 150 mg/kg of D-luciferin potassium salt solution.

Brain cysts

Brain cysts were searched for in paraﬃn imbedded tissue of

the surviving Prugneaud strain infected treated Balb/C mice in

the IVIS study, 30 days after infection which was 16 days after

treatment had been discontinued. All treated mice had survived.

There were no surviving untreated mice in those experiments.

In separate experiments, Balb/C mice were infected IP with

20 × 10³T. gondii Me49 strain tachyzoites. In these separate

studies of mice with established chronic infection, after 30 days,

IP treatment with JAG21 was begun each day for 14 days. JAG21

was dissolved in DMSO and administered IP in a total volume

of 0.05 mL. In experiments when tafenoquine was administered

alone or with JAG21 in some groups 3 mg/kg tafenoquine

was administered once on day−1 from when JAG21 treatment

RPS131 Tachyzoites in Human Primary Brain

Neuronal Stem Cells in vitro for Transcriptomics and

Transcriptomics Analyses

Culture of Human Primary Brain Neuronal Stem Cells (NSC)

was as described (McPhillie et al., 2016; Ngô et al., 2017); T.

gondii RPS13ꢀ on RH strain background (Hutson et al., 2010)

was used to infect the NSC as described (McPhillie et al., 2016;

Ngô et al., 2017). RNA was isolated and prepared and used for

transcriptomic experiments as described (McPhillie et al., 2016;

Ngô et al., 2017). Brieﬂy, NSC, initially isolated from a temporal

Frontiers in Cellular and Infection Microbiology | www.frontiersin.org

9

June 2020 | Volume 10 | Article 203

McPhillie et al.

Potent Tetrahydroquinolone Eliminates Apicomplexan Parasites

was initiated. Cysts in brain were quantitated on day 30, 16

days after discontinuing JAG21. Immunoperoxidase staining was

performed. Parasite burden was quantitated in two ways. The ﬁrst

was using a positive pixel count algorithm of Aperio ImageScope

software. Positive pixels were normalized to tissue area (mm²).

Brieﬂy, automated quantitation was done by counting positive

pixels per square area. The entire brain in one section was

scanned for each mouse. The Cyst burden was quantitated as

units of positive pixels per mm². The average S.E.M. numbers

of mm²per slide quantitated was 30.2 1.6 square mm per mouse

for this quantiﬁcation. Each highpower ﬁeld of view shown in

Figure 5C is ∼0.02 mm²per ﬁeld of view. Cysts on each slide for

each condition in two biological replicate experiments were also

quantitated by 2 separate observers independently and results

compared with automated counting, separately.

room temperature. Aliquots of the homogeneous dispersion

were frozen and lyophilized using a VirTis AdVantage freeze

drier. These aliquots were stored at room temperature for 5–

6 months. Prior to dosing, aliquots were reconstituted using

water. Controls containing no JAG21 were also prepared.

Following reconstitution with water, the dispersion was imaged

using a Nikon ECLIPSE E200 optical microscope set to 40x

magniﬁcation. The average particle size of the JAG21 dispersion

in HEC/Tween 80 was determined using an in-house image

analysis program. This novel method to stably formulate JAG21

was discovered after all other studies were completed and

this was the last experiment in this manuscript performed as

a consequence.

RH YFP challenge

For studies of the nano formulated JAG 21, this was administered

for 1 or 3 days by gavage in the doses shown in the results section.

These C57BL6 background mice received 2000 RH tachyzoites

IP. on day the ﬁrst day of the experiment and peritoneal ﬂuid

was collected 5 days later to quantitate ﬂuorescence and numbers

of parasites.

RPS13 1 in vivo

This G1 arrested parasite persists in tissue culture for prolonged

times in the absence of tetracycline (Hutson et al., 2010), but in

immune competent mice it cannot be rescued with teteracycline,

or LNAME (L-N^G-Nitro arginine methyl ester, an analog of

arginine) used as an antagonist of nitric oxide synthase (NOS)

that inhibits NO production, or both together (Hutson et al.,

2010).

In pilot studies, herein, interferon γ receptor knockout mice

that were not treated were observed following infection. At 7 and

at 14 days following infection, spleen, and liver were removed

and immune peroxidase stained. At 14 days a group of mice

were treated with anhydrotetracycline and when a subset of these

mice died, their spleen and liver were removed and immune

peroxidase stained.

As in the pilot studies, this RPS13 ꢀ parasite also was used to

infect interferon γ receptor knockout mice in a treatment study.

The design of this experiment with these immune compromised

mice is shown in Figure 6. In this separate study, groups of mice

were infected with RPS13 ꢀ. They were treated with tafenoquine

on day−1, or JAG21 for 14 days 2 h after infection, or the two

together with tafenoquine on day−1 and JAG21 for the ﬁrst

14 days, or with diluent only for 14 days, as described above.

For the initial 14 days, no tetracycline was administered. After

that time tetracycline was administered. Mice were observed

each day. At the time they appeared to have substantial illness

or at the termination of the experiment they were euthanized,

tissues ﬁxed in formalin and stained with hematoxylin and

eosin or immunoperoxidase stained and parasite burden

was assessed.

Malaria Assays

Enzyme Assays

Methods for enzyme assays: Materials

P. falciparum 3D7 strain were obtained from the Liverpool

School of Tropical Medicine. Protease cocktail inhibitor was

obtained from Roche. Bradford protein assay dye reagent was

obtained from Bio-Rad. All other reagents were obtained from

Sigma-Aldrich. Decylubiquinol was produced as per Fisher et al.

(2009). In brief, 25 mg of decylubiquinone were dissolved in 400

µL of nitrogen-saturated hexane. An equal volume of aqueous

1 M sodium dithionite was added, and the mixture vortexed

until colorless. The organic phase containing the decylubiquinol

was collected, the solvent was evaporated under N₂and the

decylubiquinol ﬁnally dissolved in 100 µL of 96% ethanol

(acidiﬁed with 10 mM HCl). Concentrations of decylubiquinol

was determined spectrophotometrically on a Cary 300 Bio

UV/visible spectrophotometer (Varian, UK) from absolute

spectra, using ε_288−−320= 8.1 mM⁻¹.cm⁻¹. Decylubiquinol was

stored at −80^◦C and used within 2 weeks.

P. falciparum culture and extract preparation

P. falciparum strain 3D7 blood-stage cultures were maintained

by the method of Trager and Jensen (1976). Cultures contained

a 2% suspension of O+ human erythrocytes in RPMI 1640

medium containing L-glutamine and sodium carbonate, and

supplemented with 10% pooled human AB+ serum, 25 mM

HEPES (pH 7.4) and 20 µM gentamicin sulfate. Cultures were

grown under a gaseous headspace of 4% O₂and 3% CO₂in N₂at

37^◦C. Cultures were grown to a parasitemia of 5% before use.

The protocol for the preparation of parasite extract was

adapted from Fisher et al. (2009). Free parasites were prepared

from infected erythrocytes pooled from ﬁve T75 ﬂasks, by

adding 5 volumes of 0.15% (w/v) saponin in phosphate-buﬀered

saline (137 mM NaCl, 2.7 mM KCl, 1.76 mM K₂HPO₄, 8.0 mM

Na₂HPO₄, 5.5 mM D-glucose, pH 7.4) for 5 min, followed by

RH Challenge in a Study of Oral Administration of a

Novel Nano Formulation of JAG21

Nanoformulation of JAG21 for oral administration in

T. gondii studies

JAG21 was prepared using hydroxyethyl cellulose (HEC) and

Tween 80. Brieﬂy, this dispersant solution containing 5 mg/mL

HEC and 2 mg/mL Tween 80 in water was prepared. Solid

JAG21 was added to 20 mg/mL, and the dispersion was

sonicated for 60 s using a Sonics vc50 probe-tip sonicator

set to 20 kHz to homogenize. Sonication was performed at

Frontiers in Cellular and Infection Microbiology | www.frontiersin.org

10

June 2020 | Volume 10 | Article 203

McPhillie et al.

Potent Tetrahydroquinolone Eliminates Apicomplexan Parasites

three washes by centrifugation in RPMI containing HEPES

(25 mM), and a ﬁnal resuspension in potassium phosphate

buﬀer (50 mM K₂HPO₄, 50 mM KH₂PO₄, 2 mM EDTA, pH7.4)

containing a protease inhibitor cocktail (Complete Mini; Roche).

Parasite extract was then prepared by disruption with a

sonicating probe for 5 s, followed by a 1 min rest period on ice

to prevent the sample overheating. This process was performed

three times. The parasite extract was used immediately. The

protein concentration of the parasite extract was determined by

Bradford protein assay (Bio-Rad).

P. berghei Causal Prophylaxis in vivo Model

P. berghei sporozoites were obtained from laboratory-reared

female Anopheles stephensi mosquitoes which were maintained

at 18 degrees C for 17–22 days after feeding on a luciferase

expressing P. berghei infected Swiss CD1ICR. Using a dissecting

microscope, the salivary glands were extracted from malaria-

infected mosquitoes and sporozoites were obtained. Brieﬂy,

mosquitoes were separated into head/thorax and abdomen.

Thoraxes and heads were triturated with a mortar and pestle

and suspended in medium RPMI 1640 containing 1% C57BL/6

mouse serum (Rockland Co, Gilbertsville, PA, USA). 50–80

heads with salivary glands were placed into a 0.5 mL Osaki

tube on top of glass wool with enough dissection media to

cover the heads. Until all mosquitoes had been dissected, the

Osaki tube was kept on ice. Sporozoites that were isolated

from the same batch of mosquitoes were inoculated into

C57BL/6, 2D knock-out, and 2D knock-out/2D6 knock-in

C57BL/6 mice on the same day to control for biological

variability in sporozoite preparations. On day 0, each mouse was

inoculated intravenously in the tail vein with ∼10,000 sporozoites

suspended in 0.1 mL volume. They were stained with a vital

dye containing ﬂuorescein diacetate (50 mg/mL in acetone)

and ethidium bromide (20 µg/mL in phosphate buﬀered saline;

Sigma Chemical Co, St. Louis, MO, USA) and counted in a

hemocytometer to ensure that inoculated sporozoites were viable

following the isolation procedure. Viability of the sporozoites

ranged from 90 to 100%.

Pfbc₁native assay

P. falciparum bc₁complex cytochrome c reductase (Pfbc₁)

activity was measured by monitoring cytochrome c reduction

at 550 vs. 542 nm using

a Cary 300 Bio UV-Visible

Spectrophotometer (Varian, UK), using a protocol adapted

from Fisher et al. (2009).The assay was performed in potassium

phosphate buﬀer in a quartz cuvette and in a ﬁnal volume of 700

µL. Potassium cyanide (10 µM), oxidized cytochrome c (30 µM),

parasite extract (100 µg protein), and compound/DMSO were

added sequentially to the cuvette, with mixing between each

addition. Test compounds were added to a ﬁnal concentration

of 1 µM. DMSO (0.1% v/v) and atovaquone (1 µM), a known

malarial cytochrome bc₁complex inhibitor, were used as

negative and positive controls, respectively. The reaction was

initiated by the addition of 50 µM decylubiquinol and allowed to

proceed for 3 min.

Malaria Parasite in vitro Studies

Animals

Malaria potency testing in vitro was performed using 4 diﬀerent

P. falciparum strains, D6, TM91-C235, W2, and C2B. The D6

strain is a drug sensitive strain from Sierra Leone, the TM91-

C235 strain is a multi-drug resistant strain from Thailand, the

W2 strain is a chloroquine resistant strain from Thailand, and the

C2B strain is a multi-drug resistant strain with resistance against

atovaquone. These assays were performed as described below.

The mice used in these experiments were albino C57BL/6 female

mice which were housed in accordance with the current Guide

for the Care and Use of Laboratory Animals (1996) under an

IACUC approved protocol. All animals were quarantined for 7

days upon arrival, and the animals were fed standard rodent

maintenance food throughout the study.

Test compounds, homogenization of JAG21 creating a

nanoformulation, and administration

Compound Activity against Plasmodium falciparum

Compound activity against P. falciparum, was tested using

the Malaria SYBR Green I–Based Fluorescence (MSF) Assay.

The complete method for performing this microtiter assay

is described in previous work published by Plouﬀe et al.

(2008) and Johnson et al. (2007). In brief, this assay uses

the binding of the ﬂuorescent dye SYBR Green I to malaria

DNA to measure parasite growth in the presence of 2-fold

diluted experimental or control. The relative ﬂuorescence of the

intercalated SYBR Green I proportional to parasite growth, and

inhibitory compounds will result in lower observed ﬂuorescence

compared to untreated parasites.

Animals were dosed with experimental compounds based on

body weight. The suspension solution of orally administered

drugs were conducted in 0.5% (w/v) hydroxyethyl cellulose and

0.2% Tween 80 in distilled water. To insure the size of the

compounds in the dosing solution were under 50 µM (measured

they were 4–6 µM), the suspension was homogenized using a

homogenizer (PRO Scientiﬁc Inc, Monroe, CT, USA) with a

10 mm open-slotted generator running at 20,000–22,000 rpm for

5 min in an ice bath. The compounds were made fresh each day

and used immediately (always in <1/2 h). Stability beyond that

time was not determined. It was not anticipated that they would

be stable beyond that time.

Cytotoxicity assays in parallel with P.falcipaum assays in

vitro

Compounds were administered on 3 consecutive days (−1,

0, +1) relative to sporozoite infection or a single dose on day

0. Drug suspensions were administered to mice by oral gavage

using an 18 gauge intragastric feeder. For the 3 day dosing

regimen, compounds were administered at 0.625 mg/kg and for

the single dose regimen administered on day 0, compounds were

administered at 2.5 mg/kg.

Toxicity studies also were performed with HepG2 cells (human

liver cancer immortal cell line derived from the liver tissue of a

15-year-old African American, ATCC ^RHB-8065^TM) in parallel

with the studies of P. falciparum, with inhibitors in vitro, as

described in McPhillie et al. (2016).

Frontiers in Cellular and Infection Microbiology | www.frontiersin.org

11

June 2020 | Volume 10 | Article 203

McPhillie et al.

Potent Tetrahydroquinolone Eliminates Apicomplexan Parasites

In vivo imaging

pellet was washed in 50 mM KPi (pH 7.5); 0.1 mM PMSF buﬀer

before second centrifugation under the same condition. The

pellet was collected and sored at −80^◦C for further use.

All of the in vivo bioluminescent imaging methods utilized

have been described previously. Brieﬂy, JAG21 was administered

orally on days −1, 0, and 1 with respect to sporozoite inoculation.

All inoculated mice were imaged using the Xenogen IVIS-

200 Spectrum (Caliper Life Sciences, Hopkinton, MA, USA)

IVIS instrument at 24, 48, and 72 h post-sporozoite infection.

The bioluminescent imaging experiments were conducted by

IP injection of the luciferase substrate, D-Luciferin potassium

salt (Xenogen, California and Goldbio, St Louis, MO, USA),

into mice at a concentration of 200 mg/kg 15 min before

bioluminescent images were obtained. Three minutes after

luciferin administration the mice were anesthetized using

isoﬂurane, and the mice were positioned ventral side up

on a 37^◦C platform with continual anesthesia provided

through nose cone delivery of isoﬂurane. All bioluminescent

images were obtained using 5 min exposures with f-stop =

1 and large binning setting. Photon emission from speciﬁc

Solubilization of Membrane Proteins

The frozen mitochondria were thawed and re-suspended in

50 mM KPi (pH 7.5); 250 mM NaCl; 0.5 mM EDTA; 0.1 mM

PMSF buﬀer; a small sample was taken for quantiﬁcation of total

mitochondrial proteins by BCA assay. The remaining sample was

centrifuged at 180,000 g in Beckman Ti70 rotor for 60 min. The

pellet was re-suspended in the same wash buﬀer with the addition

1 mg DDM per 1 mg of protein and then centrifuged under the

same conditions for 60 min. The pellet was discarded and the

supernatant was collected for ion exchange chromatography.

Puriﬁcation of Cytochrome bc1

During puriﬁcation the presence of protein was detected using

280 nm absorbance and the presence of heme was detected

using 415 nm Soret band peak and 562 nm absorbance. The

solubilized protein solution was applied on DEAE-Sepharose CL-

6B column (ca. 50 mL, GE Healthcare) pre-equilibrated with

buﬀer A [50 mM KPi (pH 7.5); 250 mM NaCl; 0.01% w/v DDM;

0.5 mM EDTA] and washed with 3 CV of buﬀer A. The protein

was eluted by linear gradient with buﬀer B [50 mM KPi (pH

7.5); 500 mM NaCl; 0.01%w/v DDM; 0.5 mM EDTA]. Fractions

containing cytochrome bc₁were pooled and concentrated to

0.5 mL using an Amicon Ultra-15 (Amicon, MWCO 100,000)

concentrator. Concentrated sample was applied to a Sephacryl-

S300 gel ﬁltration column (ca. 120 mL) pre-equilibrated in buﬀer

C [20 mM KMOPS (pH 7.2); 100 mM NaCl; 0.01%w/v DDM;

0.5 mM EDTA] and eluted at a ﬂow rate of 0.5 mL/min. Puriﬁed

cytochrome bc₁fractions were collected and concentrated to

40 mg/m. PEG fractionation with increasing concentration of

PEG4000 was used to precipitate cytochrome bc₁. Precipitating

solution (100 mM KMES pH 6.4; 10% PEG4000; 0.5 mM EDTA)

was mixed with the protein to a desired PEG concentration. The

precipitated protein pellet was re-solubilised in buﬀer D (25 mM

KPi pH 7.5, 100 mM NaCl, 0.5 mM EDTA, 0.015% DDM) and

dialysed in the same buﬀer in a centrifugal ultraﬁlter to remove

residual PEG. Five micromolars cytochrome bc₁was incubated

at 4^◦C for 12 h with 50 µM JAC21 (10-fold molar excess) diluted

from 20 mM solution stock in DMSO.

R

regions was quantiﬁed using Living Image^ꢀ3.0 software

(Perkin Elmer).

Additionally, blood stage parasitemia was assessed 3 days

after imaging was completed by treating small quantities of

blood obtained from tail bleeds with the ﬂuorescent dye Yoyo-

1 measured by using a ﬂow cytometry system (FC500 MPL,

Beckman Coulter, Miami, FL, USA) (Pybus et al., 2013; Marcsisin

et al., 2014).

Methods for Co-crystallization and Binding

Studies

Bovine Cytochrome bc₁Activity Assays

Bovine cytochrome bc₁inhibition assay was carried out in 50 mM

KPi pH 7.5, 2 mM EDTA, 10 mM KCN, 30 µM equine heart

cytochrome c (Sigma Aldrich), and 2.5 nM bovine cytochrome

bc₁at room temperature. 20 mM inhibitors dissolved in DMSO

were added to the assay at a desired concentration without

prior incubation. The working concentration of DMSO in the

assay did not exceed 0.3% v/v. The reaction was initiated by

the addition of 50 µM decylubiquinol (Abcam). The reduced

cytochrome c was monitored by the diﬀerent absorption between

550 and 542 nm using extinction coeﬃcient of 18.1 mM⁻¹cm⁻¹

in a SPECTRAmax Plus 384 UV-visible Spectrometer. The initial

kinetic rate is determined as a zero-order reaction and used as the

speciﬁc activity of cytochrome bc₁.

Crystallization, Data Collection, and Reﬁnement of

Cytochrome bc1–JAG21 Complex

Bovine Cytochrome bc1 Puriﬁcation Protocol

Preparation of crude mitochondria

The inhibitor-bound cytochrome bc₁was mixed with 1.6%

HECAMEG to the ﬁnal protein concentration of 40 mg/mL.

Hanging drop method was used for crystallization. Two

microliter of ﬁnal protein solution with 2 µL of reservoir

solution (50 mM KPi pH 6.8, 100 mM NaCl, 3 mM NaN₃,

10–12% PEG4000) was equilibrated over reservoir solution at

4^◦C. The crystals were grown to 100 µm within 4 days. The

single crystal was transferred in reservoir solution containing

increasing to 50% concentrations of ethylene glycol prior to

cryo-cooling in liquid nitrogen. X-ray data were collected from

single crystal PROXIMA2 beamline, SOLEIL light source, France

using DECTRIS EIGER X 9M detector at 0.9801Å wavelength

Whole fresh bovine heart was collected after slaughter and

transported in ice. All work was carried out at 4^◦C. Lean

heart muscle was cut into small cubes and homogenized in the

buﬀer composed from 250 mM sucrose; 20 mM K₂HPO₄; 2 mM

succinic acid; 0.5 mM EDTA. Buﬀer was added at a ratio of

2.5 L per 1 kg of muscle tissue. Ph of resulting homogenate was

adjusted to 7.8 using 2 M Tris and PMSF protease inhibitor

was added to 0.1 mM concentration. The homogenate was then

centrifuged in a Sorvall GS-3 rotor at 5,000 g for 20 min. The

resulting supernatant was then transferred to a Sorvall GSA rotor

and centrifuged at 20,000 g for 20 min. Obtained mitochondrial

Frontiers in Cellular and Infection Microbiology | www.frontiersin.org

12

June 2020 | Volume 10 | Article 203

McPhillie et al.

Potent Tetrahydroquinolone Eliminates Apicomplexan Parasites

up to 3.45Å resolution. Data were indexed and integrated RESULTS

using iMosﬂm (Battye et al., 2011), and scaled using Aimless

THQ Compounds Are Potent in vitro

(Evans, 2011). The starting model for reﬁnement was 5OKD. All

ligands except co-factors were removed from the model prior to

reﬁnement. Jelly-body reﬁnement was carried out with Refmac5

(Murshudov et al., 2011). The inhibitor model was generated

by Jligand (Lebedev et al., 2012). The model was manually

edited in COOT (between cycle reﬁnements. Data collection and

reﬁnement statistics are shown in Supplemental Table 1A).

Initially, a small library of seven compounds (Figure 1 [blue

and green font, Figure 1A] and Figure 2) were tested, and each

compound was tested at least twice against T. gondii tachyzoites.

JAG21 and JAG50 demonstrated eﬀect below 1 µM, and were

tested at lower concentrations. JAG50 and JAG21 were identiﬁed

as lead compounds given the IC₅₀values obtained were 33

and 55 nM, respectively. Correlation between concentration of

compound and inhibition of parasite replication (as measured by

ﬂuorescence) was observed for all compounds except JAG46. The

relative eﬀect on HFF and parasite enzymes were also compared,

with those marked ^∗∗in Figure 1A having the most eﬀect on the

parasite enzyme activity relative to host HFF enzyme activity as

shown below in Figure 3.

A representative graph of these in vitro data is shown in

Figure 2A. Subsequently, a larger library of 54 compounds

was synthesized to ascertain structure-activity relationships

(SAR) (Figure 1B). Our primary aims were to block putative

metabolism of the terminal phenol ring of MJM170 and improve

the solubility across the compound series. Substituents were

generally tolerated at the meta and para positions on the phenol

ring (R₁), similar to the trends observed in the ELQ series

(Vidigal et al., 2002; Doggett et al., 2012; McPhillie et al., 2016).

The incorporation of heteroatoms into the aryl rings of the

biphenyl moiety did not lead to improvements in solubility and

biological activity. Small substituents were tolerated at the 7-

position of the THQ bicyclic ring (Figure 1B; R₁), improving

selectivity (see below, SAR) but not at the 6-position unlike

the ELQ series. In summary, overall, nitrogen atoms were

not tolerated in aryl ring (C) and the 4-position was optimal

for phenol substituent. Ultimately, no other compound had

all the advantages of JAG21, although some of these were

identiﬁed as potential back up compounds (marked with ^∗∗),

with greater selectivity for the parasite relative to the mammalian

enzyme activity. Compound JAG21 displayed synergy against

RH strain tachyzoites with atovaquone (Figure 2C) but not with

pyrimethamine, although no antagonism was observed (data

not shown).

Cytotoxicity assays performed in parallel using HFF, WST-1

(Fomovska et al., 2012a,b), and HEP G2 cells demonstrated a

lack of toxicity at concentrations substantially in excess of the

concentrations eﬀective against tachyzoites. Because T. gondii

grows inside cells, if a compound were toxic to host HFF

then it would make the compound appear to be spuriously

eﬀective (Fomovska et al., 2012a,b), when in actuality only

toxicity for the host cell would be measured. Cytotoxicity to HFF

was therefore assessed for all compounds at 10 µM. Results of

this experiment are in Figure 1A, toxicity column. A two-way

ANOVA and subsequent pairwise comparison found none of

the diﬀerences in absorbance, compared to the media-DMSO

vehicle controls, to be statistically signiﬁcant (p > 0.05). Most of

these compounds are not toxic at 10 µM (the limit of solubility)

and that cytotoxicity to cells can be attributed to DMSO in the

solution, not the compound. Dose response testing (IC50) was

performed with HEP G2 cells as described and the observed

Cryo Electron Microscopy

Electron Microscopy and Image Processing

Cryo-EM was carried out as described in Amporndanai et al.

(2018). Brieﬂy, 3 µL of sample at 5 mg/mL concentration were

applied to Quantifoil Cu R1.2/1.3, 300 mesh holey carbon grids

and plunge frozen using an FEI Vitribot (blot time 6 s, blot

force 6). Data were collected on an FEI Titan Krios with a

Falcon III direct electron detector operated in integrating mode

at 300 kV. Automated data collection was carried out using

EPU software with a defocus range of −1 to −3.5 µm, and a

magniﬁcation of 75,000 × which yielded a pixel size of 1.065

Å. Data were collected for 72 h resulting in 5,356 micrographs.

The total dose was 66.4 e⁻/Å over a 1.5 s exposure which

was split into 59 frames. All of the processing was performed

in RELION 2.1 unless otherwise stated. The initial drift and

CTF correction was carried out using MOTIONCORR2 (Zheng

et al., 2017) and Gctf (Zhang et al., 2016), respectively. The

micrographs were examined and those with crystalline ice were

initially removed resulting in 2,960 micrographs. A subset of

∼2,000 particles were manually picked to generate 2D references

to facilitate auto-picking resulting in 439,009 particles. These

particles underwent an initial round of 2D classiﬁcation with

those classes that displayed clear secondary structure detail being

taken forward to 3D classiﬁcation and split into three classes.

Two of the three classes generated a high-quality cytochrome

bc₁reconstruction with secondary structure information clearly

visible. The particles from these two classes were recombined

to form the ﬁnal datasets consisting of 211,916 particles in the

ﬁnal reconstruction. The particles were 3D reﬁned using C2

symmetry to produce a map with resolution 3.8 Å. The particles

also underwent movie reﬁnement and particle polishing which

further improved the resolution of the map to 3.7 Å. A previously

reﬁned EM structure for SCR0911 (pdb 6FO6) was ﬁt into the

map using UCSF chimera and subsequently reﬁned using phenix

with the correct ligand. The maps were then inspected manually

in COOT (Emsley and Cowtan, 2004) and the model corrected

for any errors in reﬁnement and the placement of residues.

Statistical Analysis

A Pearson test was used to conﬁrm a correlation between

increasing dose and increasing inhibition. An ANOVA and

subsequent pairwise comparison with Dunnett correction was

used to determine whether or not inhibition or toxicity at a given

concentration was statistically signiﬁcant. Stata/SE 12.1 was used

for this analysis.

Frontiers in Cellular and Infection Microbiology | www.frontiersin.org

13

June 2020 | Volume 10 | Article 203

McPhillie et al.

Potent Tetrahydroquinolone Eliminates Apicomplexan Parasites

FIGURE 2 | JAG21 is potent in vitro against Toxoplasma gondii, tachyzoites and bradyzoites, and multiple drug resistant strains of P. falciparum. (A) JAG21 is

effective against RH-YFP tachyzoites, and does not harm human cells. Potent effect of JAG50 is also shown. A representative experiment is shown. N = triplicate

wells in at least 2 biological replicate experiments. Relative ﬂuorescence units are shown on the vertical axis, where decrease in ﬂuorescence compared to diluent

DMSO in media control indicates parasite inhibition (*p < 0.05). Horizontal axis indicates different treatment conditions: This shows results of testing of ﬁbroblasts in

media (HFF), DMS0 control, positive control pyrimethamine and sulfadiazine(P/S), and concentrations of JAG21 and JAG50 utilized. Differences were not statistically

(Continued)

Frontiers in Cellular and Infection Microbiology | www.frontiersin.org

14

June 2020 | Volume 10 | Article 203

McPhillie et al.

Potent Tetrahydroquinolone Eliminates Apicomplexan Parasites

FIGURE 2 | signiﬁcant in the cytotoxicity assay (data not shown). (B) JAG21 is effective against EGS bradyzoites. Effect of JAG21 in reducing bradyzoites in HFF by

parasite strain EGS. HFF were infected by EGS and treated with JAG21 at concentrations indicated. Slides were stained with Dolichos Biﬂorus Agglutinin conjugated

with FITC (which stains the cyst wall) and DAPI, and observed with ﬂuorescence microscopy. The red arrows point to the Dolichos enclosed organisms formed in

tissue culture. These were eliminated with treatment with JAG21. This experiment was performed >4 times. These experiments were performed with 3 different

observers reviewing slides at the microscope quantitating ﬁelds for each condition. Slides were also scanned and the scans of the slides were reviewed so all ﬁelds in

the entire slide were noted to be consistent. (C) Synergy of JAG21 and atovaquone against Rh-YFP tachyzoites in vitro. Isobologram comparing JAG21, atovoquone,

and JAG21 plus atovaquone demonstrates synergy. (D) THQs effective against drug resistant P. falciparum. Dose-response phenotypes of a panel of P. falciparum

parasite lines. IC50 values were calculated using whole-cell SYBR Green assay and listed as mean standard deviation of three biological replicates, each with

triplicate measurements. The D6 strain is a drug sensitive strain from Sierra Leone, the TM91-C235 strain is a multi-drug resistant strain from Thailand, the W2 strain is

a chloroquine resistant strain from Thailand, and the C2B strain is a multi-drug resistant strain with resistance against atovaquone. (E) Solubility and Stability in human

and mouse liver microsomes comparing MJM 170, JAG21, and JAG50. Performed by Chem Partners. (F) JAG21 CYP450 Inhibition, CACO-2, hERG, PPB, BBB

(MDCK-MDK1) efﬂux analyses. These were performed by Chem Partners and are as deﬁned in the section Materials and Methods. RG38 is a structurally related

inactive THQ analog.

toxicity was: HEP G2 IC50 17.70 µM (r²= 0.97) for JAG21;

7.1 µM (r²= 0.98) for JAG50.

across MDCK-MDR1 cell membranes (in vitro measure of

blood-brain barrier (BBB) penetration potential/permeability).

Solubility, half-life, HERG, and BBB permeability/eﬄux results

are shown in Figure 2F. The aqueous solubility (PBS, pH 7.4)

of amorphous compounds JAG21 and JAG50 was 7 and 16 µM,

respectively, which is improved over MJM170 (2 µM) and ELQ-

271 (0.2 µM). We also tested solubility of the microcrystalline

form of JAG21 and found that the solubility was 3.5 µM. JAG21

was the most metabolically stable compound in human liver

microsomes (>99% remaining after 45 min) compared with

other THQs and ELQ-271, although it displayed a much shorter

half-life of 101 min in mouse liver microsomes. All THQs tested

in the MDCK-MDR1 system for blood brain barrier (BBB)

permeability (including MJM170, JAG21, and JAG50), exhibited

high permeability (P_app>10 × 10⁶cm/s) and low eﬄux (eﬄux

ratio <1.5).

Lead compounds JAG50, JAG21, and others were tested

against EGS strain (Vidigal et al., 2002; Paredes-Santos et al.,

2013, 2018; McPhillie et al., 2016) tachyzoites and encysted

bradyzoites using methods described earlier (McPhillie et al.,

2016). We found a number of these compounds including JAG21

were highly eﬀective against tachyzoites (RH-YFP; Fomovska

et al., 2012a) (Figures 1A, 2A,C) and bradyzoites of EGS (Vidigal

et al., 2002; Paredes-Santos et al., 2013, 2018; McPhillie et al.,

2016) (Figure 2B). For example, in a separate experiment

(data not shown) using immunoﬂuorescence microscopy, the

following forms were observed: “true cysts” with a dolichos-

staining wall, “pseudocysts” or tight clusters of parasites, and

small organisms. If there were fewer than four parasites visible

in a cluster, organisms were counted individually (as “small

organisms”). A statistically signiﬁcant reduction in the number of

true cysts and small organisms was observed at 1 and 10 µM for

both compounds (p < 0.05, p < 0.005). Five hundred nanomolars

JAG21 treatment results in cultures where we do not see EGS

bradyzoites (e.g., Figure 2B).

Results against P. falciparum using methodology described

earlier (Trager and Jensen, 2005; Johnson et al., 2007; Plouﬀe

et al., 2008; McPhillie et al., 2016) also are shown in Figure 2D.

JAG 21 is potent against P. falciparum with IC50 values ranging

from 14 to 61 nM against a variety of drug sensitive and resistant

strains (McPhillie et al., 2016) including D6, TM91-C235, W2,

and C2B. The D6 strain is a drug sensitive strain from Sierra

Leone, the TM91-C235 strain is a multi-drug resistant strain

from Thailand, the W2 strain is a chloroquine resistant strain

from Thailand, and the C2B strain is a multi-drug resistant strain

resistant to atovaquone. Eﬀects of other comparison compounds

are also shown in this table and range from 31 to 20,000 nM

(Figure 2D).

THQs Potently Inhibit Parasite Cytochrome

bc1 (Cytbc1) Enzyme Activity

JAG21 is the most active of the initially tested THQs against T.

gondii Cytbc1, which also showed selectivity for the parasite over

the mammalian mitochondrial membrane potential (Figure 3).

Following the full SAR testing in vitro against tachyzoites, the

full set of compounds was tested against HFF; then the initial

compounds also were tested against the T. gondii and HFF

enzyme benchmarked against atovaquone, and ultimately the full

set of compounds was compared for eﬀect against the T. gondii

and HFF enzymes.

Mitochondrial membrane potential measurements were

performed with permeabilized T. gondii tachyzoites in

suspension using safranin O, which loads into polarized

membranes [see section Materials and Methods in the

Supplemental Materials (Vercesi et al., 1998)]. T. gondii

tachyzoites were permeabilized with digitonin to allow the

mitochondrial substrate succinate to cross the membrane and

energize the mitochondrion. The ﬂuorescence of safranin O,

which loads into energized mitochondria was used to measure the

membrane potential. The energized state of the mitochondrion

is observed by a decrease in ﬂuorescence (Figures 3A,C,E).

Triﬂuoromethoxy carbonylcyanide phenylhydrazone (FCCP)

was used to depolarize the membrane, which is observed as

an increase in ﬂuorescence (Figures 3A,B). JAG21 depolarized

the membrane potential even at concentrations as low as 2 nM

ADMET Superiority of JAG21

In vitro absorption, distribution, metabolism, excretion, and

toxicity (ADMET) analyses of the THQ compounds were

outsourced to ChemPartner Shanghai Ltd. ELQ-271 (synthesized

in-house) was tested as a comparison. THQs which were

potent inhibitors of T. gondii tachyzoites were assessed for their

kinetic solubility, metabolic stability in human, and mouse liver

microsomes (Figure 2E), hERG, and their ability to permeate

Frontiers in Cellular and Infection Microbiology | www.frontiersin.org

15

June 2020 | Volume 10 | Article 203

McPhillie et al.

Potent Tetrahydroquinolone Eliminates Apicomplexan Parasites

FIGURE 3 | Effect of JAG21, and other THQ compounds on mitochondrial functions of Toxoplasma gondii, Plasmodium falciparum and HFF-hTE RT (A). Maximum

mitochondrial membrane depolarization of JAG21, JAG39, JAG46, JAG47, JAG50, and Atovaquone (4 µM) and FCCP (5 µM). Digitonin was added where indicated

by the arrow to permeabilize cells and permit a necessary mitochondrial substrate (Succinate) to reach intracellular organelles. The addition of the indicated

compounds is shown by the second arrow. (B) Quantiﬁcation of the depolarization shown in (A). The relative depolarization of each compound was normalized to the

depolarization by FCCP which was considered 100% depolarization. (C) Effect of various concentrations of JAG21 and Atovaquone on the mitochondrial membrane

(Continued)

Frontiers in Cellular and Infection Microbiology | www.frontiersin.org

16

June 2020 | Volume 10 | Article 203

McPhillie et al.

Potent Tetrahydroquinolone Eliminates Apicomplexan Parasites

FIGURE 3 | potential measured as in (A). The ﬁrst arrow indicates digitonin addition and the second arrow indicates the addition of compounds at the speciﬁed

concentration. (D) Quantiﬁcation of the depolarization measured in (C). The relative depolarization of each compound was normalized to the depolarization by FCCP

(100%). (E) Mitochondrial membrane depolarization of HFF-hTERT in suspension by JAG21 and atovaquone. The ﬁrst arrow indicates the addition of digitonin, and

the second arrow indicates addition of the indicated compounds at the indicated concentration. (F) Quantiﬁcation of the depolarization measured in (E). The relative

depolarization of each compound was normalized with the depolarization by FCCP, which was considered 100%. (B,D,E) X S.D., N = 3 independent experiments.

Statistical analysis (unpaired student t-test) was performed using GraphPad Prism 8.0 (GraphPad Software, Inc., San Diego, CA). **P < 0.01. ***P < 0.001. (G)

JAG21, JAG99, and MJM210 (1 µM) inhibited P. falciparum cytochrome c reduction. Vehicle (DMSO)/atovaquone (1 µM) were negative/positive controls, 1,290

respectively. X S.D., N = 4 independent experiments.

(Figures 3C,D). JAG21 and Atovaquone had similar eﬀects

on the mitochondrial membrane potential (Figure 3D). Other

compounds like JAG46 and 47 showed almost no eﬀect at doses

as high as 4 µM (Figures 3A,B). JAG50 showed depolarizing

activity at doses of 200 nM and higher. The eﬀect of these THQ

compounds against the T. gondii mitochondrial membrane

potential was greater than the eﬀect on the human foreskin

ﬁbroblast mitochondrial membrane potential (Figures 3E,F).

This is consistent with the observation that JAG21 is less toxic

against human Telomerase reverse transcriptase immortalized

(hTERT) HFF cells than atovaquone. We had newly created

THQ compounds, not yet characterized fully, that show even less

toxicity to the human ﬁbroblast cytochrome b/c complex marked

with ^∗∗in Figure 1A. These could be developed in a second phase

of our program, were reductions in toxicity needed. However,

as data presented herein demonstrates, there are signiﬁcant

advantages in the ADMET properties of JAG21, and its dramatic

eﬃcacy in vivo, without toxicity. There may be no need to further

develop any of those potential additional leads.

Enzyme reduction of cytochrome c by P. falciparum parasite

extract (Fisher et al., 2004, 2009) is mediated by P. falciparum

bc₁complex cytochrome c reductase (Pfbc₁) enzyme. All three

compounds tested (1 µM) signiﬁcantly inhibited the reduction

of cytochrome c by the P. falciparum parasite extract (JAG21 =

86.4 3.2; JAG99 = 81.3 6.0; MJM170 = 69.7 11.3% of the

atovaquone response, Figure 3G. Additional data demonstrated

selective eﬀect on P. falciparum enzyme compared with bovine

enzyme (data not shown).

positioning of the inhibitor (Figure 4B). No additional electron

density was found within the Q_osite. After the reﬁnement, 2F_o-

F_celectron around JAG21 becomes clearer (Figure 4C). The

second aromatic ring in the tail group of the compound is less

deﬁned due to high ﬂexibility introduced by the oxygen linker.

The quinolone head of JAG21 is held between Asp228 and

His201 and adapted the same conformation as 4(1H)-pyridone

(GSK932121) (Capper et al., 2015) (Figure 4D) and tetrahydro-

4(1H)-quinolone (MJM170) (McPhillie et al., 2016) (Figure 4E)

by directing the NH group to His201 and the carbonyl group to

Asp228. The carbonyl of the quinolone head and OG1 atom of

Ser35 are within 3.0 Å distance that allows hydrogen bonding

and enhances the binding aﬃnity to the bovine enzyme. The

3-diarylether tail extends along a hydrophobic channel deﬁned

by Gly38, Ile39, and Ile42. The triﬂuoromethoxy group at the

phenoxy ring points toward Met190 and Met194 (Figure 4F).

CryoEM studies of the complex also demonstrate reasons for

selectivity. In Figure 4F, the density suggests that the inhibitor

can adopt two diﬀerent binding poses as observed previously in

the cryo-EM structure of GSK932121 (Capper et al., 2015). The

binding pose shown in yellow, which has the strongest density,

agrees with the crystal structure and has the triﬂuoromethoxy

group pointing toward Met194. However, there is additional

density which could result from a second binding pose (green)

in which the triﬂuoromethoxy group points toward Asp228

(McPhillie et al., 2016). Figure 4F shows GSK932121 pyridone

(PDB:4D6U) (Figure 4G) MJM170 quinolone (PDB:5NMI). The

EM map has been deposited at the EMDB (EMDB-11002).

JAG21 Is Potent in vivo

Binding, Co-crystallography,

Pharmacophore, and Cryo-electron

Microscopy Studies Demonstrate

In vivo studies of JAG21 against T. gondii demonstrated

high eﬃcacy in a variety of settings. JAG21 at 5 mg/kg/day

administered IP improves well-being and eliminates illness and

T. gondii Type II Prugneaud luciferase tachyzoites completely in

luminescence studies (Figure 5A). Further, treatment beginning

on day one after infection results in no cysts being found in

brains of these mice treated for 14 days with 5 mg/kg/day of

JAG21, when brains were evaluated 30 days after stopping JAG21

treatment in two replicate experiments. Treatment beginning

on day 30 after initiation of infection with Type II Me49

parasites results in marked, statistically signiﬁcant reduction in

normal appearing cysts, free organisms, and immunoperoxidase

stained cysts detected by automated imaging of scanned slides

(Figures 5B,C, p < 0.03 experiment 1: p < 0.01 experiments

1 and 2 together, Supplemental Figure 1). The automated

analysis conﬁrmed results from the blinded microscopic visual

quantitation of cysts and free organisms in slides by two

Selectivity

In binding assays and in co-crystallography (Emsley and Cowtan,

2004; Emsley et al., 2010; Battye et al., 2011; Laskowski

and Swindells, 2011; Murshudov et al., 2011; Lebedev et al.,

2012; Capper et al., 2015; McPhillie et al., 2016; Zhang

et al., 2016; Zheng et al., 2017; Amporndanai et al., 2018),

JAG21 has lower binding aﬃnity to bovine cytochrome bc

in comparison with previous compounds that we have tested.

JAG21 “inhibits” Cytbc1 but not fully, indicating that it will

be less toxic for mammalian (bovine/human) cyt bc1 than

the apicomplexan enzymes (Figure 4A). The electron density

map in the Qi site of bovine cytochrome bc₁complex with

JAG21 (Supplemental Table 1, Data Collection Statistics) reveals

an additional electron density, which allowed unambiguous

Frontiers in Cellular and Infection Microbiology | www.frontiersin.org

17

June 2020 | Volume 10 | Article 203

McPhillie et al.

Potent Tetrahydroquinolone Eliminates Apicomplexan Parasites

FIGURE 4 | Binding studies of JAG21 to bovine bc1. (A) Bovine Cytbc1 activity assays showing 36 and 63% inhibition at 0.1 and 1 µM concentration of JAG21,

respectively. N = at least 2 biological replicate experiments with similar results. (B) The Cytbc1 structure presented in cartoon style with clear omit (Fo-Fc) electron

density map for the bound JAG21 compound only in the Q_isite showing selectivity within the binding pocket. Qi and Qo sites are marked by black ellipsoids. (C) The

bound JAG21 compound (orange) within the Qi site with corresponding (2Fo-Fc) electron density map contoured at 1 σ level as gray mesh. The residues which make

close interactions with the bound inhibitor are shown in stick format and labeled. (D) 2D pharmacophore analysis of JAG21 binding pocket produced using Ligplot+

LS-2011. Hydrophobic interactions are shown as red spikes, hydrogen bond with Ser35 is shown by green dashes. (E) Cryo-EM derived structure of the Cytbc1

bound JAG021 structure with corresponding density map contoured at 3 σ level suggesting two different positions for the head group represented by two regions of

density shown as yellow mesh. The Cytbc1 structure bound to the pyridone GSK932121 (PDB:4D6U) (F) and quinolone MJM170 (PDB:5NMI) (G) in the Q_isite. Haem

and compounds are shown as colored sticks, Fe ion as orange sphere and hydrogen bonding as black lines. Hydrogen bonding with Ser35 is shown as black dashes.

Terms JAG021 and JAG21 used interchangeably for this same compound.

observers. Adding tafenoquine or primaquine to treatments of

active plus dormant malarias (St Jean et al., 2016; Lacerda

et al., 2019; Llanos-Cuentas et al., 2019) is partially eﬀective

against both active and dormant phase plasmodia, when neither

treatment of active nor dormant disease alone is eﬀective

for either in vivo. We developed experiments based on these

Frontiers in Cellular and Infection Microbiology | www.frontiersin.org

18

June 2020 | Volume 10 | Article 203

McPhillie et al.

Potent Tetrahydroquinolone Eliminates Apicomplexan Parasites

observations where experiments with tafenoquine alone or with

JAG21 alone was used in the experiments with established cysts

with immune competent mice. This was to determine whether

tafenoquine might add to eﬃcacy of JAG21. The eﬃcacy of

treatment with JAG21 alone was so robust (Figure 5B), that

no additive eﬀect was seen, or could have been detected, by

adding Tafenoquine to JAG21. Eﬃcacy was shown when data

were analyzed as separate groups, i.e., control vs. JAG21 alone

(p < 0.03) or control vs. JAG21 plus tafenoquine, or grouping

the JAG21 and JAG21 plus tafenoquine results as “untreated” vs.

“treated” (p < 0.01). Analysis shown combining both treatment

groups from two replicate experiments showed similar results (p

< 0.01, Figure 5B), and when results from replicate experiments

were grouped (Supplemental Figure 1). In Figure 5C the control

mice had cysts with usual morphology (Top two panels), whereas

treated mice had very few morphologically recognizable usual

cysts that were immunostained (bottom panels).

A nano formulation homogenized (<6 µM) was used

eﬀectively orally for the P. bergheii experiments, further,

importantly, was eﬀective in the single oral dose causal

prophylaxis in 5 C57BL/6 albino mice at 2.5 mg/kg dosed on

day 0, 1 h after intravenous administration of 10,000 P. berghei

sporozoites was completely protective. In addition, 3 dose causal

prophylaxis treatment in 5 C57BL/6 albino mice at 0.625 mg/kg

dosed on days −1, 0, and +1 also was completely protective. A

representative experiment at a higher dose (5 mg/kg) is shown,

but all experiments with the oral dosing regimen with the

nanoformulation speciﬁed above showed 100% survival 30 days

post infection with P. berghei, where all liver and blood stage

parasites were eliminated (Figures 5D,E) demonstrates not only

eﬃcacy of JAG21 against the three life cycle stages of P. berghei,

but also demonstrates the eﬃcacy of oral administration of the

nanoformulation when used immediately, at a low dose.

et al., 2010). The parasite could be rescued from its dormant—

ATc state by adding ATc, months after removing tetracycline

from infected HFF cultures, although it could not be rescued

in immunocompetent mice with LNAME and ATc when tested

1 week after infection (Hutson et al., 2010). We wondered if

this type of dormant organism could form in vivo, whether it

could contribute in a biologically relevant way to dormancy and

recrudescence, similar to the malaria hypnozoite (Cubi et al.,

2017; Muller et al., 2019), and whether JAG21 might be able to

eliminate it, or whether a companion compound eﬀective against

this form might be needed or work in conjunction with JAG21 if

needed. To begin to address these questions and to investigate

how close the T. gondii ꢀRPS13 -ATc phenotype might be to

the malaria hypnozoite, we compared the transcriptome of T.

gondii ꢀRPS13 in human, primary, brain, neuronal stem cells

+/– ATc to the recently published P. cynomolgi hypnozoite

transcriptome, established with single cell RNA sequencing in

laser captured organisms (Cubi et al., 2017). This analysis

identiﬁed 28 orthologous genes with similar expression pattern

in both T. gondii ꢀRPS13 -ATc and P. cynomolgi hypnozoites,

including the downregulation of rps13 and upregulation of

the eukaryotic initiation factor-2α kinase IF2K-B, a protein

involved in translational control in response to stress (Cubi

et al., 2017) (Figure 6A). Further. assessment of the T. gondii

ꢀRPS13 transcriptome in the absence or presence of ATc showed

upregulation of additional IF2K members, 25 Apetela (AP) 2

transcription factors and a number of genes that participate as

protein ubiquitin ligases, and in traﬃcking as well as in RNA

binding, and GCN1 (Supplemental Table 2). None of them,

except for AP2VIIa-7, have been shown to be upregulated nor

downregulated during diﬀerentiation to bradyzoites. Gene set

enrichment analysis showed that in the absence of ATc, the

T. gondii ꢀRPS13 transcriptome is enriched in genes typically

expressed during G1, conﬁrming previous results indicating that

downregulation of rps13 arrests the parasite at this stage of

the cell cycle (Figure 6B) (Hutson et al., 2010). Moreover, a

number of biological processes are downregulated without ATc,

including protein synthesis and degradation as well as energy

metabolism (Figure 6B). Noteworthy, some gene ontology (GO)

terms enriched in T. gondii ꢀRPS13 -Tc are also overrepresented

in the P. cynomolgi hypnozoite (stars in Figure 6D). Further,

without ATc the transcriptome of T. gondii ꢀRPS13 is compatible

with a parasite transitioning from an active replicating form

to a dormant stage, reﬂected by the downregulation of genes

typical of the S and M stages of the cell cycle, and of

genes that participate in energy metabolism and virulence

(Figures 6B,D, Supplemental Table 2, Supplemental Figure 2).

It has been reported that with treatment of active forms of

malaria, hypnozoites still persist, and recrudesce later (Hutson

et al., 2010; St Jean et al., 2016; Cubi et al., 2017; Lacerda

et al., 2019; Llanos-Cuentas et al., 2019). Also, compounds

that target cytochrome b/c were not eﬀective against malaria

hypnozoites. If primaquine or tafenoquine, which do not treat

the active Plasmodium vivax parasites, were added in vivo,

hypnozoites have been shown not to recrudesce, or do so less

often (St Jean et al., 2016; Lacerda et al., 2019; Llanos-Cuentas

et al., 2019). Testing with primaquine or tafenoquine could

G1 Arrest, Persisters, Companion

Compounds

In mice that were treated with JAG21 early after infection

(Figure 5A) we could ﬁnd no residual immunostaining for T.

gondii in brain tissue of any mice. This suggests that very

early treatment could prevent established, chronic infection, for

example in epidemics such as those that occurred in Victoria,

Canada, the U.S.A., and Brazil. In mice with established cysts,

following treatment with JAG21, we occasionally saw a small

number of cysts (Figure 5B) and amorphous immunostained

structures (Figure 5C, bottom panels). This was reminiscent of

persistence in some malaria infections (Cubi et al., 2017) and

abnormal immunostained structures we previously identiﬁed

with a conditional, tetracycline-on regulatable, mutant T. gondii

(Hutson et al., 2010 and Supplemental Figure 2). In this

ꢀRPS13 tachyzoite, small ribosomal protein 13 can be regulated,

depending on whether anhydrotetracycline (ATc) is absent or

present, leading the ATc responsive repressor to be on or oﬀ

response elements engineered into the promoter (Hutson et al.,

2010). ꢀRPS13 replicates with ATc present and is arrested in

G1 when ATc is absent in HFF cultures (Hutson et al., 2010).

The dormant parasite could persist for extended periods (Hutson

Frontiers in Cellular and Infection Microbiology | www.frontiersin.org

19

June 2020 | Volume 10 | Article 203

McPhillie et al.

Potent Tetrahydroquinolone Eliminates Apicomplexan Parasites

FIGURE 5 | JAG21 is a mature lead that protects against Toxoplasma gondii and Plasmodium berghei in vivo. (A) JAG21 treatment for 14 days protects against T.

gondii tachyzoites in vivo. Tachyzoite challenge with Prugneaud luciferase parasites imaged with leuciferin using IVIS demonstrates that treatment with JAG21

eliminates leuciferase expressing parasites and leads to 100% survival of JAG21 treated infected mice. No cysts were found in brains of mice at 30 days after infection

when they have been treated with JAG21 for the ﬁrst 14 days after infection. There were 2 biological replicate experiments with 5 mice per group with similar results.

(B) JAG21 and JAG21 plus tafenoquine markedly reduce Me49 strain brain cyst numbers in vivo in Balb/C mice at 30 days after infection. Parasites were quantitated

by scanning the entire immunoperoxidase stained slide in an automated manner and by two observers blinded to the experimental treatment using microscopic

evaluation. In each of two experiments, the numbers of mice per group were as follows: Experiment 1 had 4 diluent controls, 5 JAG21, 4 JAG21/Tafenoquine treated

mice; and Experiment 2 had 5 diluent controls, 5 JAG21, 3 JAG21/Tafenoquine treated mice. Immunoperoxidase staining was performed. Parasite burden was

quantitated using a positive pixel count algorithm of Aperio ImageScope software. Positive pixels were normalized to tissue area (mm²). Quantiﬁcation was by

counting positive pixels per square area. The entire brain in one section was scanned for each mouse. The parasite burden was quantitated as units of positive pixels

per mm². The average S.E.M. numbers of mm²per slide quantitated was 30.2 1.6 mm²per mouse for this quantiﬁcation. Each high power ﬁeld of view shown in

(Continued)

Frontiers in Cellular and Infection Microbiology | www.frontiersin.org

20

June 2020 | Volume 10 | Article 203

McPhillie et al.

Potent Tetrahydroquinolone Eliminates Apicomplexan Parasites

FIGURE 5 | C is ∼0.02 mm²per ﬁeld of view. A representative single experiment is presented and the data from the two experiments analyzed together also

demonstrated signiﬁcant differences between the untreated and treated groups (p < 0.01; Supplemental Figure 1). (C) Microscopic evaluation of the slides reveal

effect of JAG21 and JAG21 plus tafenoquine having the same pattern as the automated quantitation of immunoperoxidase stained material. There are usual

appearing cysts in the DMSO control untreated mice as shown in the top panels, and rare cysts in the treated mice with most of the brown material appearing

amorphous (bottom panels). (D) JAG21 nanoformulation dosages administered to P. berghei infected C57Bl6/albino mice compared with vehicle control. Design of

single dose and 3 day dose experiments. (E) JAG21 nanoformulation cures P. berghei sporozoites (left panel), blood (middle panel), and liver stages, leading to 100%

survival (right panel). This is with oral administration of a single dose of 2.5 mg/kg or 3 doses at 0.625 mg/kg. Single dose causal prophylaxis in 5 C57BL/6 albino

mice at 2.5 mpk dosed on day 0, 1 h after intravenous administration of 10,000 P. berghei sporozoites. Shown is 3 dose causal prophylaxis treatment in 5 C57BL/6

albino mice at 0.625 mpk dosed on days −1, 0, and +1. Representative ﬁgure showing survival (right panel), luminescence (left panel), and parasitemia quantitated by

ﬂow cytometry (middle panel) for 5 mg/kg.

only be performed in vivo, as activity against the hypnozoite

requires hepatic metabolism of primaquine or tafenoquine (St

Jean et al., 2016; Lacerda et al., 2019; Llanos-Cuentas et al.,

2019). Tafenoquine is not active in tissue culture which is

consistent with the ﬁndings that these compounds require

hepatic metabolism. To establish a parallel in vivo system, we

studied immune compromised mice (Interferon g receptor

knockout mice with the knockout in the germline) infected with

ꢀRPS13 herein. Although in immune competent mice ꢀRPS13

does not recrudesce with ATc treatment initially, beyond 3

days after infection, we found that when ATc was added after

treatment of the immune compromised mice with JAG21 dosed

intraperitoneally for 14 days, the dormant ꢀRPS13 parasite

could still recrudesce after JAG21 treatment was discontinued

and tetracycline added (Figure 6C and Supplemental Figure 2).

Consistent with adding tafenoquine to treatment of P.vivax

malaria with chloroquine where both medicines together were

partially eﬀective against the active and hypnozoite forms, the

combination of JAG21 and tafenoquine had a modest eﬀect

together on transiently improving survival time when ATc

was added when compared with JAG21 or tafenoquine alone

(Figure 6C and Supplemental Figure 2). The trend in the result

seems similar to the malaria infections where hypnozoites form,

although protection was not as robust, as in the malaria model,

and we did not achieve complete, durable protection against

ꢀRPS13. These results in Figure 6C and Supplemental Figure 2

suggest: (a) In G1 arrested organisms that begin as tachyzoites,

they can persist in vivo even if their morphology as parasites is

diﬃcult to discern; (b) Treatment with JAG21+Tafenoquine can

prolong time to death more robustly than other treatments; (c)

But, in these immune compromised mice at this dosage regimen

this treatment did not robustly, durably protect these mice from

death later; (d) In these immune compromised mice, whether this

lack of complete protection was because of immune compromise,

or less than optimal duration of treatment, or suboptimal dose or

timing of treatments, or that this G1 arrested organism is harder

to treat, remains to be determined in future studies. The modest

eﬃcacy of the two compounds, administered together, suggests

that treating both tachyzoites and the G1 arrested organisms

is important. This seems similar to P. cynomogli and P. vivax

treatment with tafenoquine and chloroquine studies, which also

showed eﬃcacy but was not completely successful in preventing

relapse. At the time this study was performed, formulation

and dosing (including duration and timing) had not yet been

optimized formally for the T. gondii model. P. vivax treatment

requires chloroquine to treat blood schizonts and tafenoquine

to treat hypnozoites. Treatment in man, per the FDA approved

label, consists of a single dose of 300 mg on day 1 co-administered

with chloroquine treatment on days 1 or 2. Both medicines have

long half-lives in humans. This treatment was relatively eﬀective

in humans, with about a 30% recurrence rate.

Sinai et al. have demonstrated heterogeneity in the phenotypes

of organisms within established cysts. Their work found

bradyzoites within cysts are not uniform with regard to

their replication potential (Watts et al., 2015), mitochondrial

activity (Sinai, unpublished), and levels of the glucose storage

polymer amylopectin (Sinai, unpublished). These properties of

bradyzoites (Watts et al., 2015), and properties of tissue cysts

that vary during the course of infection, demonstrated that

there are unappreciated levels of complexity in the progression

of chronic toxoplasmosis (Watts et al., 2015). The analysis

(Figure 6D) of the ꢀRPS13 infected NSC suggests molecular

targets that are modiﬁed in this G1 arrested ꢀRPS13 parasite

as shown in Figure 6D and Supplemental Table 2. In the future,

with formulation and pharmacokinetics of JAG21 optimized, it

will be of interest to determine whether JAG21 can eliminate

these organisms and any residual structures as in Figure 6C,

or whether adding synergistic compounds such as atovaquone

(Figure 4B) or antisense eﬀective against these upregulated

molecular targets, such as kinases, ATPases, AP2s (Figure 6D

and Supplemental Table 2), or a newly recognized bradyzoite

master regulator of diﬀerentiation might be eﬀective alone or

might be synergistic with JAG21 against this ꢀRPS13, as well

as the conventional recognized tachyzoite and bradyzoite life

cycle stages. Chen et al. reported in the transcriptomes of

established bradyzoite in vivo cysts that EIF2kinase of stressed

parasites is present (Chen et al., 2018), but we have not found

other overlap of Chen’s transciptome with P cynomogli or

ꢀRPS13 transcriptomes. Others have described EIF2kinase and

stress granules only in transitioning or extracellular parasites

(Watts et al., 2015). Bradyzoites within tissue cysts are not

monolithic. Thus, in future studies, single cell RNA sequencing

of bradyzoites obtained by laser capture of bradyzoites in

vivo, deﬁned on the basis of their physiological state, may be

needed to determine whether a transcriptome signature similar

to ꢀRPS13 is sometimes present. This could be linked to

morphologic/immunostaining features that might functionally

distinguish them to deﬁne the character of a hypnozoite-like

state in T. gondii. We noted heterogeneity of parasite phenotype,

even in the same vacuoles, in our earlier IFA and electron

microscopic characterization of G1 arrested ꢀRPS13 in HFF

(Hutson et al., 2010). Heterogeneity also was found very recently

Frontiers in Cellular and Infection Microbiology | www.frontiersin.org

21

June 2020 | Volume 10 | Article 203

McPhillie et al.

Potent Tetrahydroquinolone Eliminates Apicomplexan Parasites

FIGURE 6 | Toxoplasma gondii ꢀRPS13 transcriptome during Primary Human Brain Neuronal Stem Cell (NSC) infection and in-vivo susceptibility to JAG21 and TAF

treatment are reminiscent of literature ﬁndings with malaria hypnozoites. (A) P. cynomolgi-T. gondii best reciprocal match genes signiﬁcantly upregulated (red) or

downregulated (blue) in P. cynomolgi hypnozoites compared to liver-schizont stage and in ꢀRPS13 after downregulation of rps13 gene expression (p ≤ 0.05, FDR ≤

0.2). (B) Gene-set enrichment analysis of ꢀRPS13 +/– Tc. Blue and red nodes denote gene-sets enriched in presence or absence of Tc, respectively. Node diameters

are proportional to number of genes belonging to corresponding gene-sets. Edge thickness is proportional to number of genes shared between connected nodes. (C)

Survival rate of mice infected with 100,000 ꢀRPS13 followed by treatment with diluent, JAG21, tafenoquine (TAF) or the two together (JAG21/TAF). Then tetracycline

was added at 14 days. The combination of the two compounds resulted in improved time of survival (p < 0.03, Experiment 1; p = 0.08 Experiment 2, p = 0.002

Experiment 1+2). The full data are presented in the box below the image in (C). (C) Rx refers to treatment of mice with diluent (DMSO), Tafenoquine (TAF), or JAG21,

or JAG21 and TAF. ꢀRPS13 is referred to as RhRPS13ꢀ in the title of the box in (C). In Supplemental Figure 2, histological preparations that are

immunoperoxidase stained for T. gondii antigens from a pilot study were prepared (Supplemental Figure 2). These are images, in Supplemental Figure 2 of liver

and spleen from IFN γ receptor knock out mice without treatment on days 7 and 14 after infection. In those mice without any treatment there was amorphous brown

immunoperoxidase stained material in Supplemental Figure 2A. When tetracycline (aTet) was administered on day 14 after infection in drinking water, with tissues

obtained and immunostained for T. gondii antigens from mice that died or became very ill, organisms that were clearly recognizable could be seen

(Supplemental Figures 2B–E). Design of the treatment experiment with control DMSO diluent, JAG21 alone, Tafenoquine alone (TAF) or the two together (JAG/TAF)

with full data for each of the groups and with the composite analysis from replicate experiments, including numbers of mice, are shown in

Supplemental Figures 2C,D. Supplemental Figures 2C,D shows prolongation of survival time, but there is not durable protection against ꢀRPS13 in these

immune compromised mice treated with JAG21/TAF as described. This is summarized in C to demonstrate early prolongation of survival time with the detailed data in

Supplemental Figure 2. (D) Gene ontology enrichment analysis of ꢀRPS13 +/–Tc. Node and edge conventions are the same as in (B). There were at least 2

biologic replicates of each experiment.

Frontiers in Cellular and Infection Microbiology | www.frontiersin.org

22

June 2020 | Volume 10 | Article 203

McPhillie et al.

Potent Tetrahydroquinolone Eliminates Apicomplexan Parasites

FIGURE 7 | Oral nanoformulation of JAG21 potently protects against 2000 highly virulent RH strain tachyzoites given intraperitoneally. (A) Following sonication

produces nanoparticles of ∼2.86 µM. (B) Single oral dose of 10 mg/kg reduced intraperitoneal tachyzoites measured by RH YFP expression and counting with

hematocytometer (p < 0.03). (C) Three daily 10 mg/kg doses markedly and signiﬁcantly reduces intraperitoneal parasite burden measured as ﬂuorescence and by

hematocytometer on the ﬁfth day (p < 0.001). No compound was administered after the third day. N = at least 2 biological replicate experiments with 5 mice per

group with similar results.

in tachyzoites and bradyzoites created by alkaline conditions in

culture across the cell cycle in vitro in HFF, using single cell RNA

sequencing (Xue et al. Biorx, 6/3/2019 in press). These authors

also noted that what had been interpreted as “noise” earlier was

found actually to be signal in a more complex environment.

These authors suggest that such heterogeneity might make

developing curative treatments more complex. Our analysis of

JAG21 eﬀects and the ꢀRPS13-ATet knockdown herein begin

to help address this question: We noted that consistent with

heterogeneity in our IFAs, in our comparison with the Xue et al.’s

heterogeneous P1-6 clusters analysis, we found that most of

the up- or down-regulated genes are within P3-P5 tachyzoite

clusters. Also, consistent with the heterogeneity we observed in

our G1 arrested ꢀRPS13 -ATet comparison, ꢀRPS13 has a drop

in SAG1 and elevated SRS44 that is consistent with a brady-like

phenotype. BAG1 expression was too low overall to draw any

Frontiers in Cellular and Infection Microbiology | www.frontiersin.org

23

June 2020 | Volume 10 | Article 203

McPhillie et al.

Potent Tetrahydroquinolone Eliminates Apicomplexan Parasites

conclusion about BAG1. It is also noteworthy that in our -ATet

relative to +Atet conditions in primary, human, brain, neuronal

stem cells, the master regulator of bradyzoite diﬀerentiation is

slightly overexpressed (Log₂Fold Change=0.7, adjusted p =

0.043). Although JAG21 is highly potent against tachyzoites and

bradyzoites, it did not eliminate every long-established encysted

bradyzoite or -ATet ꢀRPS13 completely either in vitro or in IFNγ

knock-out mice in vivo. Consistent with heterogeneity, herein

JAG21 treatment of ꢀRPS13 and transcriptomics analyses deﬁne

a metabolically quiescent, persister, “stasis” state that is reversible

even after substantial periods of dormancy. These observations

contribute to conceptual and functional understanding of both

Plasmodia species and T. gondii infections and molecular

mechanisms whereby “persisters” might be eliminated.

against tachyzoites would be of considerable clinical usefulness.

Further, no medicines are active against the encysted stage or

deﬁnitively curative. In addition, malaria is lethal for 1 child

every 11 s and a threat to travelers going to endemic areas.

Development of drug resistance also increases the need for new

anti-malarial compounds. Our goal in this work herein was

to identify compounds highly eﬀective against T. gondii and

P. falciparum, and we believe we have achieved our goal by

developing lead compounds with dual activity.

To further develop the THQ series, 54 compounds

were synthesized to improve kinetic solubility, solubility in

physiologically-relevant media (FaSSIF, FeSSIF), and metabolic

stability (microsomes and hepatocytes), and other ADMET

properties. Compounds JAG50 and JAG21 were identiﬁed as

lead compounds, demonstrating potent inhibition on both

tachyzoites and bradyzoites life stages and were not toxic to

human cells in our in vitro model (HFF). In addition, both

compounds displayed low nanomolar eﬃcacy against multiple

drug resistant strains of P. falciparum in vitro. JAG50 and JAG21

demonstrate promising ADMET properties, with JAG21 slightly

superior due to the compound’s longer metabolic stability in

human and mouse microsomes.

A striking result with JAG21 in our in vivo parasite studies

is the compound’s high eﬃcacy against T. gondii tachyzoites

and bradyzoites. In our P. berghei in vivo model for malaria,

we observed that a single dose causal prophylaxis in 5 C57BL/6

albino mice at 2.5 mpk dosed on day 0, 1 h after intravenous

administration of 10,000 P. berghei sporozoites was achieved.

Causal prophylaxis was also observed after a 3-dose treatment in

5 C57BL/6 albino mice at 0.6 mpk dosed on days −1, 0, and +1. A

representative ﬁgure at a higher dose (5 mg/kg) is shown, and all

experiments with the amounts mentioned above demonstrated

identical high eﬃcacy in luminescence, parasitemia, and survival

results. This demonstrates that JAG21 functions better in this in

vivo model than the ELQ 300 series where prodrug formulation

is required to achieve solubility and eﬃcacy, in contrast to the

eﬃcacy of JAG21 at 2.5 mg/kg in a single oral dose model

resulting in cure without a prodrug. ELQ 300 (not the prodrug)

was not eﬀective at doses between 1 and 20 mg/kg although the

prodrug was more eﬀective (Doggett et al., 2012; Frueh et al.,

2017).

JAG50 and JAG21 are lead compounds, with JAG21 being

a superior compound due to its favorable predicted ADMET

properties, potency, eﬃcacy, and lack of toxicity. JAG21

demonstrates increased solubility and potential for advanced

formulation. There also is potential for improving solubility

and reducing toxicity further because of the larger binding

pocket in the apicomplexan Cytbc1 enzyme compared with the

mammalian Cytbc1 enzyme. This was determined by modeling

occupancy of the structure, enzyme assays and empirically.

We have created and tested additional compounds that take

advantage of these properties, although none at present, have

the proven ADMET and marked in vivo eﬃcacy we found to

be advantageous in our proof of principle studies of JAG21. At

present, however, our mature lead compound has suﬃcient drug

like properties to move to advanced formulations, suggesting

increased bulk will not be needed to reduce toxicity. It has

An Oral Nanoformulation Is Potent Against

Virulent RH

To further develop JAG21 for practical, clinical use, our next step

was to make a formulation that is stable at room temperature,

and would be eﬀective when administered orally: Following a

number of unsuccessful alternative methods (data not shown),

a dispersion of JAG21 was prepared using hydroxyethyl cellulose

(HEC) and Tween 80. This new formulation method is described

in the Materials and Methods. When this dispersion was imaged

using a Nikon ECLIPSE E200 optical microscope set to 40x

magniﬁcation, the average particle size of the JAG21 dispersion in

HEC/Tween 80 was determined using an in-house image analysis

program and was found to be 2.85 µm (Figure 7A). Material

was re-sonciated the same way just prior to administration after

being stored for 6 months and retained the same properties

(Figure 7A) when imaged. Following administration of 2,000

highly virulent RH Strain tachyzoites intraperitoneally, the oral

nanoformulation was administered by gavage using a 21 gauge

needle. This was given either (1) as a single dose of 5, 10, or 20

mg/kg, or (2) three daily doses of 10 mg/kg given for the ﬁrst

3 days after infection. After 5 days the RH strain tachyzoites in

peritoneal ﬂuid of each mouse were quantitated by measurement

of YFP they expressed using a ﬂuorimeter and by quantitating

parasites present in peritoneal ﬂuid using a hemocytometer.

Parasite burden was reduced by ∼60% 5 days later following the

single doses of 10 and 20 mg/kg (representative experiment with

10 mg/kg shown in Figure 7B; p < 0.03) and markedly reduced

with three doses of 10 mg/kg administered on each of the ﬁrst

3 days after intraperitoneal injection of the virulent RH strain

tachyzoites (Figure 7C, representative experiment, p < 0.001).

This is the proof of principle that will facilitate media milling,

dispersant, and a self disintegrating tablet in the future. JAG21

has real promise as a mature lead compound to treat both T.

gondii and Plasmodium species infections.

DISCUSSION

T. gondii infections are highly prevalent and the impact of this

disease can be devastatingly severe. Current treatments have

toxicity or hypersensitivity side eﬀects. New compounds that are

without toxicity or hypersensitivity, and that are highly active

Frontiers in Cellular and Infection Microbiology | www.frontiersin.org

24

June 2020 | Volume 10 | Article 203

McPhillie et al.

Potent Tetrahydroquinolone Eliminates Apicomplexan Parasites

selectivity as demonstrated by our enzymatic, binding, and

structure studies, although there are additional compounds that

show even greater selectivity. It is highly eﬀective in an oral nano

preparation against P. berghei’s three life cycle stages, and with

early treatment appears to be capable of curing toxoplasmosis in

immunocompetent mice. This work demonstrates the promising

nature of this novel tetrahydroquinolone scaﬀold and mature

lead compound. JAG21 has the potential to become an orally

administered medicine or with partners, part of a medicine

combination that is curative for toxoplasmosis and is a

single dose cure for malaria. It is suitable for partnering

with other compounds to obviate problems with selection of

resistant mutants. We have demonstrated earlier that the parent

compound with this new scaﬀold is synergistic with atovaquone

and additive with cycloguanil (in proguanil) against P. falciparum

(McPhillie et al., 2016). Herein, we also found synergy between

JAG21 and atovaquone against T. gondii tachyzoites in vitro.

This compound is a mature lead compound to treat both T.

gondii and Plasmodium species infections. If utility and safety

retained, and no toxicity appears in next stage studies, this

compound may become suitable for treatment of T. gondii and

P. falciparum infections.

and Prevention of Toxoplasmosis with approval by the University

of Strathclyde ethical review board.

AUTHOR CONTRIBUTIONS

MM, RM, MH, HL, SNM, CF, CR, SPM, GB, SA, KR, RKP, and

YZ conceptualized and designed the overall study. MM, RM,

MH, and HL wrote manuscript. All authors wrote subparts of

manuscript, performed experiments, and/or analyzed data. All

authors reviewed and edited manuscript in ﬁnal form.

FUNDING

This work was supported by NIAID NIH DMID U01 AI082180

(RM) and the National Institute of Diabetes and Digestive and

Kidney Diseases (NIDDK) Grant #5T35DK062719-28 to FH.

This was also supported by National Institutes of Health (NIH)

contract number HHNS272200900007C, NIH. National Institute

of Allergy and Infectious Diseases of the National Institutes

of Health (NIAID) award numbers R01AI071319(NIAID)

and R01AI027530 (NIAID) (RM); NIAID contract Number

HHNS272200900007C; NIAID award number U19AI110819

(HL): NIAID award number U01 AI077887(NIAID) and Defense

Threat Reduction Agency award number 13-C-0055, and

Department of Defense award numbers W911NF-09-D0001 and

W911SR-07-C0101(MH). This work was also supported by R01

AI128356 to SNM. The work also was supported by the Bill

and Melinda Gates Foundation (BMGF, OPP1150755) (RKP).

This material is based upon work supported by the National

Science Foundation Graduate Research Fellowship under Grant

No. #DGE-1656466 awarded to KR. RJ was funded by a Ph.D

scholarship by the Wellcome Trust (109158/B/15/Z).

CONCLUSION

JAG21 has real promise as a mature lead compound to treat both

T. gondii and Plasmodium species infections, demonstrated in

vitro and in vivo. It has high eﬃcacy against T. gondii tachyzoites

and bradyzoites, and established encysted organisms. Treatment

with a single low oral dose is eﬀective for causal prophylaxis and

radical cure of P. berghei infections. JAG 21 has complete eﬃcacy

against three life cycle stages of P. berghei. In terms of companion

inhibitors, JAG21, a Q_iinhibitor, synergizes against tachyzoites

with atovaquone (a Q_oinhibitor) in vitro. It appears able to

contribute modestly to protection of immune compromised mice

in conjunction with tafenoquine against an initially replicating,

then G1 arrested, T. gondii parasite that shares key transcriptomic

components with P. cynomolgi hypnozoites. Our mature lead

compound has suﬃcient selectivity and drug-like properties

to support ongoing eﬀorts to further develop this compound

through preparation of advanced formulations and testing in

additional studies of pharmacokinetics, eﬃcacy, and safety.

ACKNOWLEDGMENTS

We thank and gratefully also acknowledge the support of

the Cornwell Mann family, the Ramirez, Musillami, Quinn,

Rosenthal, Greenberg, Morel, Rooney and Engel families, Taking

out Toxo, and The Toxoplasmosis Research Institute. We would

like to thank beamline scientists at Proxima2, Synchrotron

Soeil, France, Proposal 20161037 and iNEXT (Proposal 1728)

for Financial support for synchrotron access. We thank Abigail

Spedding and Nisha Pokar for their assistance in helping

JG at Leeds University with compound synthesis, and Ryan

Gonciarz (‘RG’) for his compound RG38. We also acknowledge

the assistance of Leon Wang, Ph.D of Princeton University

with size determination for the ﬁnal step of nanoformulation

for this proof of principle study. Dennis Steindler kindly

provided the primary human brain neuronal stem cells used

in this study.

DATA AVAILABILITY STATEMENT

The EM map has been deposited at the EMDB (EMDB-11002).

Structure coordinates are deposited and available from Protein

Data Bank under the accession code: 6XVF.

ETHICS STATEMENT

This animal study was reviewed by, approved by, and carried

out in accordance with regulations of the University of Chicago

IACUC and IBC and of The Home Oﬃce of the UK Government

under the Animals [Scientiﬁc Procedures] Act 1986. All work in

the UK with mice was covered by License PPL60/4568, Treatment

SUPPLEMENTARY MATERIAL

The Supplementary Material for this article can be found

online at: https://www.frontiersin.org/articles/10.3389/fcimb.

2020.00203/full#supplementary-material

Frontiers in Cellular and Infection Microbiology | www.frontiersin.org

25

June 2020 | Volume 10 | Article 203

McPhillie et al.

Potent Tetrahydroquinolone Eliminates Apicomplexan Parasites

REFERENCES

Frueh, L., Li, Y., Mather, M., Li, Q., Pou, S., and Nilsen, A. (2017).

Alkoxycarbonate ester prodrugs of preclinical drug candidate ELQ-300

for prophylaxis and treatment of malaria. ACS Infect. Dis. 10, 728–735.

doi: 10.1021/acsinfecdis.7b00062

Gubbels, M.-J., Li, C., and Striepen, B. (2003). High-throughput growth assay

for Toxoplasma gondii using yellow ﬂuorescent protein. Antimicrob. Agents

Chemother. 47, 309–316. doi: 10.1128/AAC.47.1.309-316.2003

Hutson, S. L., Mui, E., Kinsley, K., Witola, W. H., Behnke, M. S., and El

Bissati, K., et al. (2010). T. gondii RP promoters and knockdown reveal

molecular pathways associated with proliferation and cell-cycle arrest. PLoS

ONE 5:e14057. doi: 10.1371/journal.pone.0014057

Amporndanai, K., Johnson, R. M., O’Neill, P. M., Fishwick, C. W. G., Jamson,

A. H., Rawson, S., et al. (2018). X-ray and cryo-EM structures of inhibitor-

bound cytochrome bc 1 complexes for structure-based drug discovery. IUCrJ

5, 200–210. doi: 10.1107/S2052252518001616

Anders, S., Pyl, P. T., and Huber, W. (2015). HTSeq–a Python framework

to work with high-throughput sequencing data. Bioinformatics 31, 166–169.

doi: 10.1093/bioinformatics/btu638

Battye, T. G., Kontogiannis, L., Johnson, O., Powell, H. R., and Leslie,

A. G. (2011). iMOSFLM:

a new graphical interface for diﬀraction-

image processing with MOSFLM. Acta Crystallogr. Sect. D 67, 271–281.

doi: 10.1107/S0907444910048675

Johnson, J. D., Dennull, R. A., Gerena, L., Lopez-Sanchez, M., Roncal, N. E., and

Waters, N. C. (2007). Assessment and continued validation of the malaria SYBR

green I-based ﬂuorescence assay for use in malaria drug screening. Antimicrob.

Agents Chemother. 51, 1926–1933. doi: 10.1128/AAC.01607-06

Khan, A. A., Nasr, M., and Araujo, F. G. (1998). Two 2-hydroxy-3-alkyl-1,4-

naphthoquinones with in vitro and in vivo activities against Toxoplasma gondii.

Antimicrob. Agents Chemother. 42, 2284–2289. doi: 10.1128/AAC.42.9.2284

Kim, D., Langmead, B., and Salzberg, S. L. (2015). HISAT: a fast spliced aligner with

low memory requirements. Nat. Methods 12, 357–360. doi: 10.1038/nmeth.3317

Lacerda, M. V. G., Llanos-Cuentas, A., Krudsood, S., Lon, C., Saunders,

D. L., Mohammed, R., et al. (2019). Single-dose Tafenoquine to prevent

relapse of Plasmodium vivax Malaria. N. Engl. J. Med. 380, 215–228.

doi: 10.1056/NEJMoa1710775

Laskowski, R. A., and Swindells, M. B. (2011). LigPlot +: multiple ligand à

protein interaction diagrams for drug discovery. J. Chem. Inform. Model. 51,

2778–2786. doi: 10.1021/ci200227u

Lebedev, A. A., Young, P., Isupov, M. N., Moroz, O. V., Vagin, A. A., Murshudov,

G. N. (2012). JLigand: a graphical tool for the CCP4 template-restraint library.

Acta Crystallogr. Sect. D 68, 431–440. doi: 10.1107/S090744491200251X

Llanos-Cuentas, A., Lacerda, M. V. G., Hien, T. T., Vélez, I. D., Namaik-

Larp, C., Chu, C. S., et al. (2019). Tafenoquine versus primaquine to

prevent relapse of Plasmodium vivax malaria. N. Engl. J. Med. 380, 229–241.

doi: 10.1056/NEJMoa1802537

Bradbury, R. H., Allot, C. P., Dennis, M., Girdwood, J. A., Kenny, P. W., Major,

J. S., et al. (1993). New nonpeptide angiotensin II receptor antagonists. 3.

Synthesis, biological properties, and structure-activity relationships of 2-alkyl-

4-(biphenylylmethoxy)pyridine derivatives. J. Med. Chem. 36, 1245–1254.

Capper, M. J., O’Neill, P. M., Fisher, N., Strange, R. W., Moss, D., Ward, S. A.,

et al. (2015). Antimalarial 4(1H)-pyridones bind to the Qi site of cytochrome

bc1. Proc. Natl. Acad. Sci. U.S.A. 112, 755–760. doi: 10.1073/pnas.14166

11112

Caumes, E., Bocquet, H., Guermonprez, G., Rogeaux, O., Bricaire, F., Katlama,

C., et al. (1995). Adverse cutaneous reactions to pyrimethamine/sulfadiazine

and pyrimethamine/clindamycin in patients with AIDS and toxoplasmic

encephalitis. Clin. Infect. Dis. 21, 656–658. doi: 10.1093/clinids/21.

3.656

Chen, L. F., Han, X. L., Li, F. X., Yao, Y. Y., Fang, J. P., and Liu, X. J., et al.

(2018). Comparative studies of Toxoplasma gondii transcriptomes: insights into

stage conversion based on gene expression proﬁling and alternative splicing.

Parasites Vect. 11:402. doi: 10.1186/s13071-018-2983-5

Croken, M. M., Qiu, W., White, M. W., and Kim, K. (2014). Gene set enrichment

analysis (GSEA) of Toxoplasma gondii expression datasets links cell cycle

progression and the bradyzoite developmental program. BMC Genomics

15:515. doi: 10.1186/1471-2164-15-515

Cubi, R., Vembar, S. S., Biton, A., Franetich, J. F., Bordessoulles, M., and Sossau,

D. (2017). Laser capture microdissection enables transcriptomic analysis

of dividing and quiescent liver stages of Plasmodium relapsing species. P

cynomogli. Cell. Microbiol. 19. doi: 10.1111/cmi.12735

Delair, E., Latkany, P., Noble, A. G., Rabiah, P., McLeod, R., and Brézin, A. (2011).

Clinical manifestations of ocular toxoplasmosis. Ocul. Immunol. Inﬂamm. 19,

91–102. doi: 10.3109/09273948.2011.564068

Doggett, J. S., Nilsen, A., Forquer, I., Wegmann, K. W., Jones-Brando, L., Yolken,

R. H., et al. (2012). Endochin-like quinolones are highly eﬃcacious against

acute and latent experimental toxoplasmosis. Proc. Natl. Acad. Sci. U.S.A. 109,

15936–15941. doi: 10.1073/pnas.1208069109

Emsley, P., and Cowtan, K. (2004). Coot: model building tools for molcular

graphics. Acta Cryst. D60, 2126–2132. doi: 10.1107/S0907444904019158

Emsley, P., Lohkamp,B., Scott, W. G., and Cowtan, K. (2010).

Love, M. I., Huber, W., and Anders, S. (2014). Moderated estimation of fold

change and dispersion for RNA-seq data with DESeq2. Genome Biol. 15:550.

doi: 10.1186/s13059-014-0550-8

Lykins, J., Wang, K., Wheeler, K., Clouser, F., Dixon, A., and El Bissati, K., et al.

(2016). Understanding Toxoplasmosis in the United States through “Large

Data” analyses. Clin. Infect. Dis. 63, 468–475. doi: 10.1093/cid/ciw356

Marcsisin, S. R., Sousa, J. C., Reichard, G. A., Caridha, D., Zeng, Q., and Roncal,

N. (2014). Tafenoquine and NPC-1161B require CYP 2D metabolism for anti-

malarial activity: implications for the 8-aminoquinoline class of anti-malarial

compounds. Malar. J. 13:2. doi: 10.1186/1475-2875-13-2

McLeod, R., and Boyer, K. (2019). “Toxoplasmosis (Toxoplasma gondii),” in Nelson

Textbook of Pediatrics, 21st Edn (Philadelphia, NY; Oxford: Elsevier), 528–540.

McLeod, R., Khan, A. R., Noble, G. A., Latkany, P., Jalbrzikowski, J., and Boyer,

K. (2006). Severe sulfadiazine hypersensitivity in a child with reactivated

congential toxoplasmic chorioretinitis. Pediatr. Infect. Dis. J. 25, 270–272.

doi: 10.1097/01.inf.0000202070.59190.9a

Features and development of coot. Acta Crystallogr.

D 66, 486–501.

doi: 10.1107/S0907444910007493

Evans, P. R. (2011). An introduction to data reduction: space-group determination,

McPhillie, M., Zhou, Y., El Bissati, K., Dubey, J., Lorenzi, H., and Capper, M.,

et al. (2016). New paradigms for understanding and step changes in treating

active and chronic, persistent apicomplexan infections. Sci. Rep. 6:29179.

doi: 10.1038/srep29179

Miley, G. P., Pou, S., Winter, R., Nilsen, A., Li, Y., and Kelly, J. X., et al. (2015).

ELQ-300 prodrugs for enhanced delivery and single-dose cure of Malaria. AAC

59, 5555–5556. doi: 10.1128/AAC.01183-15

Muench, S. P., Prigge, S. T., McLeod, R., Raﬀerty, J. B., Kirisits, M. J., and Roberts,

C. W. (2007). Studies of Toxoplasma gondii and Plasmodium falciparum

enoyl acyl carrier protein reductase and implications for the development of

antiparasitic agents. Acta Crystallogr. D Biol. Crystallogr. 63(Pt 3), 328–338.

doi: 10.1107/S0907444906053625

Muller, I., Jex, A. R., Kappe, S. H. I., Mikolajczak, S. A., Sattabongkot, J.,

Patrapuvich, R., et al. (2019). Transcriptome and histone 1 epigenome of

Plasmodium vivax salivary-gland sporozoites point to tight regulatory control

and potential mechanisms for liver-stage diﬀerentiation. Int. J. Parasitol. 49,

501–513. doi: 10.1016/j.ijpara.2019.02.007

scaling and intensity statistics. Acta Crystallogr. Sect.

D 67, 282–292.

doi: 10.1107/S090744491003982X

Fisher, N., Castleden, C. K., Bourges, I., Brasseur, G., Dujardin, G., and Meunier,

B. (2004). Human disease-related mutations in cytochrome b studied in yeast.

J. Biol. Chem. 279, 12951–12958. doi: 10.1074/jbc.M313866200

Fisher, N., Warman, A., Ward, S. A., and Biagini, G. A. (2009). “Chapter 17

Type II NADH: quinone oxidoreductases of Plasmodium falciparum and

Mycobacterium tuberculosis: kinetic and high-throughput assays. Methods

Enzymol. 456, 303–320. doi: 10.1016/S0076-6879(08)04417-0

Fomovska, A., Huang, Q., El Bissati, K., Mui, E. J., Witola, W. H., Cheng,

G., et al. (2012b). Novel N-Benzoyl-2-hydroxybenzamide disrupts unique

parasite secretory pathway. Antimicrob. Agents Chemother. 56, 2666–2682.

doi: 10.1128/AAC.06450-11

Fomovska, A., Wood, R. D., Mui, E., Dubey, J. P., Ferreira, L. R., Hickman, M. R.,

et al. (2012a). Salicylanilide inhibitors of Toxoplasma gondii. J. Med. Chem. 55,

8375–8391 doi: 10.1021/jm3007596

Frontiers in Cellular and Infection Microbiology | www.frontiersin.org

26

June 2020 | Volume 10 | Article 203

McPhillie et al.

Potent Tetrahydroquinolone Eliminates Apicomplexan Parasites

Murshudov, G. N., Skubák, P., Lebedev, A. A., Pannu, N. S., Steiner, R.

A., Nicholls, R. A., et al. (2011). REFMAC5 for the reﬁnement of

macromolecular crystal structures. Acta Crystallogr. Sect. D 67, 355–367.

doi: 10.1107/S0907444911001314

Wallon, M., Peyron, F., Cornu, C., Vinault, S., Abrahamowicz, M., and Kopp, C. B.

(2013). Congenital toxoplasma infection: monthly prenatal screening decreases

transmission rate and improves clinical outcome at age 3 years. Clin. Infect. Dis.

56, 1223–1231. doi: 10.1093/cid/cit032

Ngô, H. M., Zhou, Y., Lorenzi, H., Wang, K., Kim, T. K., and Zhou,

Y., et al. (2017). Toxoplasma modulates signature pathways of

human epilepsy, neurodegeneration and cancer. Sci. Rep. 7:11496.

doi: 10.1038/s41598-017-10675-6

Walton, N. M., Sutter, B. M., Chen, H. X., Chang, L. J., Roper, S. N., Scheﬄer,

B., et al. (2006). Derivation and large-scale expansion of multipotent astroglial

neural progenitors from adult human brain. Development 133, 3671–3681.

doi: 10.1242/dev.02541

Paredes-Santos, T. C., Martins-Duarte, E. S., de Souza, W., Attias, M., and

Vommaro, R. C. (2018). Toxoplasma gondii reorganizes the host cell

architecture during spontaneous cyst formation in vitro. Parasitology 145,

1027–1038. doi: 10.1017/S0031182017002050

Paredes-Santos, T. C., Martins-Duarte, E. S., Vitor, R. W., de Souza, W.,

Attias, M., and Vommaro, R. C. (2013). Spontaneous cystogenesis in vitro

of a Brazilian strain of Toxoplasma gondii. Parasitol. Int. 62, 181–188.

doi: 10.1016/j.parint.2012.12.003

Watts, E., Zhao, Y., Dhara, A., Eller, B., Patwardhan, A., and Sinai, A. P. (2015).

Toxoplasma gondii bradyzoites within tissue cysts are dynamic and replicating

entities in vivo. Novel approaches reveal that Toxoplasma gondii bradyzoites

within tissue cysts are dynamic and replicating entities in vivo. MBio 6:e01155–

e01115. doi: 10.1128/mBio.01155-15

Waxman, S., and Herbert, V. (1969). Mechanism of pyrimethamine-

induced megaloblastosis in human bone marrow. N. Engl.

10.1056/NEJM1969061228

Phan, L., Kasza, K., Jalbrzikowski, J., Noble, A. G., Latkany, P., Kuo, A., et al.

(2008). Longitudinal study of new eye lesions in children with toxoplasmosis

who were not treated during the ﬁrst year of life. Am. J. Ophthalmol. 146,

375–384. doi: 10.1016/j.ajo.2008.04.033

Plouﬀe, D., Brinker, A., McNamara, C., Henson, K., Kato, N., Kuhen, K.,

et al. (2008). In silico activity proﬁling reveals the mechanism of action of

antimalarials discovered in a high-throughput screen. Proc. Natl. Acad. Sci.

U.S.A. 105, 9059–9064. doi: 10.1073/pnas.0802982105

Pybus, B. S., Marcsisin, S. R., Jin, X., Deye, G., Sousa, J. C., and Li, Q. (2013). The

metabolism of primaquine to its active metabolite is dependent on CYP 2D6.

Malar. J. 12:212. doi: 10.1186/1475-2875-12-212

St Jean, P. L., Xue, Z., Carter, N., Koh, G. C., Duparc, S., and Taylor, M. (2016).

Tafenoquine treatment of Plasmodium vivax malaria: suggestive evidence that

CYP2D6 reduced metabolism is not associated with relapse in the Phase 2b

DETECTIVE trial. Malar. J. 15:97. doi: 10.1186/s12936-016-1145-5

Su, G., Morris, J. H., Demchak, B., and Bader, G. D. (2014). Biological

network exploration with Cytoscape 3. Curr. Protoc. Bioinform. 13, 1–24.

doi: 10.1002/0471250953.bi0813s47

02402

Zhang, J., Tan, P., Guo, L., Gong, J., Ma, J., Li, J., et al. (2016).

p53-dependent autophagic. Cryo-EM of cellular machineries. Proc.

Natl. Acad. Sci. U.S.A. 113, 11519–11524. doi: 10.1073/pnas.160948

2113

Zheng, S. Q., Palovcak, E., Armache, J. P., Verba, K. A., Cheng, Y., and

Agard, D. A. (2017). MotionCor2 - anisotropic correction of beam-induced

motion for improved cryo-electron microscopy. Nat. Methods 14, 331–332.

doi: 10.1038/nmeth.4193

Conﬂict of Interest: RM, MM, CF, CR, KE, MH, QL, and HL are inventors on

an International patent application PCT/US2016/067795 pertinent to the work in

this study. RM has completed an unrelated literature review for Sanoﬁ-Pasteur.

The remaining authors declare that the research was conducted in the absence of

any commercial or ﬁnancial relationships that could be construed as a potential

conﬂict of interest.

Subramanian, A., Tamayo, P., Mootha, V. K., Mukherjee, S., Ebert, B. L., Gillette,

M. A., et al. (2005). Gene set enrichment analysis: a knowledge-based approach

for interpreting genome-wide expression proﬁles. Proc. Natl. Acad. Sci. U.S.A.

102, 15545–15550. doi: 10.1073/pnas.0506580102

Torgerson, P. R., and Mastroiacovo, P. (2013). The global burden of congenital

toxoplasmosis: a systematic review. Bull. World Health Organ. 91, 501–508.

doi: 10.2471/BLT.12.111732

Received: 20 February 2020; Accepted: 16 April 2020; Published: 09 June 2020

Citation: McPhillie MJ, Zhou Y, Hickman MR, Gordon JA, Weber CR, Li Q,

Lee PJ, Amporndanai K, Johnson RM, Darby H, Woods S, Li Z, Priestley RS,

Ristroph KD, Biering SB, El Bissati K, Hwang S, Hakim FE, Dovgin SM, Lykins JD,

Roberts L, Hargrave K, Cong H, Sinai AP, Muench SP, Dubey JP, Prud’homme RK,

Lorenzi HA, Biagini GA, Moreno SN, Roberts CW, Antonyuk SV, Fishwick CWG

and McLeod R (2020) Potent Tetrahydroquinolone Eliminates Apicomplexan

Parasites. Front. Cell. Infect. Microbiol. 10:203. doi: 10.3389/fcimb.2020.00203

Trager, W., and Jensen, J. B. (1976). Human malaria parasites in continuous

culture. Science 193, 673–675.

Trager, W., and Jensen, J. B. (2005). Human malaria parasites in

continuous culture. 1976. J. Parasitol. 91, 484–486. doi: 10.1645/0022-

3395(2005)091[0484:HMPICC]2.0.CO;2

Vercesi, A. E., Rodrigues, C. O., Uyemura, S. A., Zhong, L., and Moreno,

S. N. J. (1998). Respiration and oxidative phosphorylation in the

apicomplexan parasite Toxoplasma gondii. J. Biol. Chem. 273:31040.

doi: 10.1074/jbc.273.47.31040

Vidigal, P. V. T., Santos, D. V. V., Castro, F. C., de Couto, J. C. F., de

Vitor, R. W. A., and Brasileiro Filho, G. (2002). Prenatal toxoplasmosis

diagnosis from amniotic ﬂuid by PCR. Rev. Soc. Bras. Med. Trop. 35, 1–6.

doi: 10.1590/S0037-86822002000100001

Johnson, Darby, Woods, Li, Priestley, Ristroph, Biering, El Bissati, Hwang, Hakim,

Dovgin, Lykins, Roberts, Hargrave, Cong, Sinai, Muench, Dubey, Prud’homme,

Lorenzi, Biagini, Moreno, Roberts, Antonyuk, Fishwick and McLeod. This is an

open-access article distributed under the terms of the Creative Commons Attribution

License (CC BY). The use, distribution or reproduction in other forums is permitted,

provided the original author(s) and the copyright owner(s) are credited and that the

original publication in this journal is cited, in accordance with accepted academic

practice. No use, distribution or reproduction is permitted which does not comply

with these terms.

Frontiers in Cellular and Infection Microbiology | www.frontiersin.org

27

June 2020 | Volume 10 | Article 203

Article Doi

Potent Tetrahydroquinolone Eliminates Apicomplexan Parasites

DOI: 10.3389/fcimb.2020.00203

Source and publish data:

Authors:

Article abstract of DOI:10.3389/fcimb.2020.00203

Full text of DOI:10.3389/fcimb.2020.00203

Products guided by the article

R&D Labs maybe for 873203-36-2

Relevant to this article

Hot Product