An in vitro screen reveals tachy- and bradyzocidal properties of the MMV Pathogen Box compounds.

(A) Experimental scheme of the MMV compound screen on Toxoplasma gondii tachyzoites in human foreskin fibroblast cells (B) Growth curves of tachyzoites treated with 10µM of each MMV compound for 7 days (shades of grey), a solvent control (green), and an inhibition control (red) (C) Experimental scheme of the MMV compound screen on bradyzoites matured in myotubes (D) Growth curves of bradyzoites treated with 10µM of each compound (shades of grey) or a solvent control (green) for 7 days, followed by a tachyzoite regrowth phase (E) Pie chart displaying the number of Pathogen Box compounds and their effects on our in vitro cultures.

Confirmed Pathogen Box compounds that target both tachyzoites and mature bradyzoites.

(A) shows the growth curve of MMV689480 (Buparvaquone), one of the 16 dually active compounds, in a tachyzoite growth assay (n=8). The infected cultures were treated for 7 days and regrowth in absence of the inhibitor was observed for another week. Depending on the development of the DMSO control, the half-maximal inhibitory concentration (IC50) was determined at either day 4 or 7 (arrow). (B) shown is the respective data plotted as relative fluorescence units depending on the concentration at day 7. (C) With 4 weeks old, mature bradyzoite cultures (n=6), the half-maximal lethal concentration (LC50) was determined at the time point which ideally reflected the dose-response curve (in the case of Buparvaquone at day 31; arrow). (D) The LC50 of Buparvaquone on encysted bradyzoites determined at day 31. All IC50 and LC50 of dually active compounds are summarized in (E) for tachyzoite and (F) for bradyzoite cultures. (G) Comparison of the minimal lethal concentration of each compound on tachyzoites and bradyzoites.

Untargeted metabolomics of tachyzoites treated with 6 screening hits effective against both stages of T. gondii.

Volcano plots showing the parasitic, intracellular metabolic phenotype caused by MMV Pathogen Box compounds that inhibited tachyzoites and bradyzoites during the screen. The metabolome was measured with two columns of complementary chemistries to enhance metabolite coverage. Blue dots indicate metabolites which were significantly reduced compared to DMSO treated parasites and red dots indicate significantly accumulated metabolites (p < 0.05, log2 fold change <-1 or = 1, 3x biological replicates with 3x technical replicates each, n = 9). Abbreviations: AcCoA = Acetyl-CoA, AcCys = Acetylcysteine, AcPhe = Acetylphenylalanine, AMP = Adenosinemonophosphate, ATP = Adenosinetriphosphate, CarbAsp = Carbamoylaspartate, cGMP-AMP = cyclic Guanosinemonophosphate- adenosinemonophosphate, CMP = Cytosinemonophosphate, DHO = Dihydroorotate, GABA = gamma-Aminobutyratic acid, GluCys = Glutamylcysteine, GluGln = Glutamylglutamine, Gro1P = Glyceraldehyde 1-phosphate, Gro3P = Glyceraldehyde 3-phosphate, GSH = Glutathione oxidized, GSSG = Glutathione reduced, GTP = Guanosinetriphosphate, HP4 = forth chromatographic peak of 260.02972, HydKyn = Hydroxylkynurenine, MEP = 2-C-Methyl-D- erythritol 4-phosphate, NAcAla = N-Acetylalanine, NAcCys = N-Acetylcysteine, NAcMet = N- Acetylmethionine, NAcPhe = N-Acetylphenylalanine, OAcSer = O-Acetylserine, PalmCarn = Palmitoylcarnitine, PEP = Phosphoenolpyruvate, PseudoU = Pseudouridine, Trp = Tryptophane, UMP = Uridinemonophosphate, UTP = Uridinetriphosphate

The metabolic response to electron transport chain inhibitors and MMV1028806.

(A) Scheme of the affected pathways upon bc1-complex inhibition. Intracellular tachyzoites were labeled with U-13C-glucose (B,C) or 15N-amide-glutamine (D,E) and treated with atovaquone, buparvaquone and MMV1028806 for 3h. Showing (B,D) the log2-fold changes to a DMSO treated sample of ETC related metabolites and (C,E) their average isotopologue distribution. Abbreviations: aKG = alpha-Ketoglutarate, ATQ = atovaquone, BPQ = buparvaquone, CarbAsp = Carbamoyl-aspartate, DHO = Dihydroorotate, DHODH = Dihydroorotate dehydrogenase, GABA = gamma-Aminobutyrate, MMV = MMV1028806, PPP = Pentose phosphate pathway, SSA = Succinic semialdehyde

An electron-potential dependent mitochondrial stain shows reduced mitochondrial capacity in atovaquone (ATQ), buparvaquone (BPQ) and MMV1028806 treated tachyzoites and bradyzoites.

(A) HFF cells were infected with tachyzoites and treated with DMSO, atovaquone, buparvaquone and MMV1028806 for 24 h, and mitochondria were stained with Mitotracker (red) and DNA was stained with DAPI (blue). scale bar = 5 µm. (B) The ratios of fluorescence intensities between mitotracker (MT) and DAPI dyes were calculated per vacuole (DMSO n = 56, ATQ n = 45, BPQ n = 49, MMV1028806 n = 40; Blue lines represent the median, two-sided Mann-Whitney test). (C) 4 weeks old, myotube derived ME49 in vitro cysts were treated with DMSO, atovaquone, buparvaquone and MMV1028806 for 24 h, and mitochondria were stained with mitotracker (red), the cyst wall was stained with Dolichos Biflorus Agglutinin and Streptavidin-Cy2 (green), and DNA was stained with DAPI (blue). scale bar = 20 µm. (D) The ratios of fluorescence intensities between mitotracker (MT) and DAPI dyes were calculated per vacuole (DMSO n = 56, ATQ n = 42, BPQ n = 45, MMV1028806 n = 38; Blue lines represent the median, two-sided Mann-Whitney test).

Untargeted metabolomic analysis of bradyzoites treated with bc1- complex inhibitors shows an energy disbalance and hurdling mechanisms.

(A) Four weeks old in vitro cysts were treated for three hours. The cell monolayer was scraped off and the cysts were syringe-released. With Dolichos biflorus agglutinin (DBA) coated magnetic beads, the isolated cysts were captured and washed with PBS. The metabolites were extracted with 80% acetonitrile (AcN) in water and sonication, and analyzed via hydrophilic- interaction ultra-high pressure liquid chromatography coupled mass spectrometry (HILIC- UHPLC-MS). The results are shown in three separate volcano plots: Metabolic phenotypes of bradyzoites treated with Atovaquone (B), Buparvaquone (C), and MMV1028806 (D). (n = 3, significance analyzed with a two-sided U-Mann-Whitney test) Abbreviations: ATQ = Atovaquone, BPQ = Buparvaquone, DMSO = Dimethylsulfoxide, LA-carnitine = Linoleylcarnitine, MMV = MMV1028806, NADH = reduced Nicotinamide adenine dinucleotide, PA-carnitine = Palmitoylcarnitine, SA-carnitine = Stearoylcarnitine

Differential efficacy and ATP depletion by atovaquone and HDQ in tachyzoites and bradyzoites.

Atovaquone and HDQ were tested against T. gondii Prugeniaud parasites. (A) Showing the normalized half-maximal inhibitory concentration (IC50) against tachyzoites in fibroblasts at day 4, n=8. (B) 4-week-old bradyzoites were treated for a week with Atovaquone or HDQ. Cultures were observed for tachyzoite re-growth for three weeks. The half-maximal lethal concentration (LC50) was determined at the time point which ideally reflected the dose-response curve, in this case day 14, n=6. 105 tachyzoites and bradyzoites were lysed and the ATP concentration was determined with an enzymatic luciferase assay. (C) Showing the calculated concentration of ATP within an average parasite. Lines reflect the median, error bars represent SEM, significance test Mann-Whitney, n=12. (D) Parasites were treated with 1μM Atovaquone or HDQ. RFUs were normalized to the untreated control of either the tachyzoite or bradyzoite dataset. Shown are means, error bars represent SD, significance test is pairwise Mann- Whitney, n=6.

Longitudinal Screening Procedure

All 400 compounds of the MMV Pathogen Box were screened against (A) tachyzoites (n = 4) and (B) in vitro tissue cysts (n = 3). Plotted fluorescence values were background-subtracted and reflect the abundance of parasites per treatment over time (black line = solvent control, green line = inhibitory compound, red line = ineffective compound, dashed line = growth threshold).

Properties of dually active screening hits.

(A) Table showing the relative amount of tachyzocidal, bradyzocidal and dually-active compounds per originally indicated target species. E.g. 11 compounds were indicated by MMV to be active against Cryptosporidium and 36% and 27% of these are bradyzocidal and dually active, whereas 67% of the 15 compounds indicated as active against Toxoplasmosis possessed activity against tachyzoites but none was active against against bradyzoites. (B) Estimated lipophilicity approximated by predicted partition coefficient suggest that active compounds generally exhibit an increased lipophilicity compared to inactive compounds and that bradyzocidal compounds are more lipophilic than tachyzocidal compounds. (C) A Venn diagram showing intersections of tachy- and bradyzocidal compounds as well as their predicted gastrointestinal absorption capability (GIA) and blood brain barrier (bbb) permeability. (D) Tabulation of compound IC50 and LC50 data of confirmed active compounds. The calculated IC50s and LC50s of 16 dually active compounds from figure 2 are summarized in this table. The tachyzoite IC50s represent 8 replicates, the bradyzocidal one where generated from 6 replicates.

Principal component analysis of LCMS metabolomes of treated parasites using two chromatographic colums.

(A) Principal component analysis of metabolites from extracellular tachyzoites that were treated with indicated MMV compounds. Metabolites were separted using BEH-Amide column -HILIC chromatography and detected by mass spectrometry. (B) The same metabolite extracts have been re-analysed using pHILIC-column HILIC chromatography for separation instead.

Intracellular tachyzoites exhibit differential 15N incorporation after treatment with ETC inhibitors.

RH-KU80 were treated with atovaquone, buparvaquone and MMV1028806, labeled with 15N- amide-L-glutamine for 3h, followed by their isolation from the HFF cells. The heatmap shows the overall isotopic labeling (100-"M+0") of each replicate.

bc1-complex inhibitors partially reduce mitochondrial profiles of bradyzoites.

Thin section electron microscopy of cysts that were treated with Atovaquone, Buparvaquone or MMV1028806 for 24 h (A). (B) Summary of measured number and area of mitochondrial profiles of 28 cells per condition.

Luminescence-based ATP assay

(A) The linear response of the BacTiter-Glo™ assay has a certain range, depending on the ATP concentration of the sample. The assay is reliable between 100fM and 1nM ATP. (B) Increasing numbers of freshly isolated parasites were analysed to determine a minimal parasite density necessary for steady readouts. 105 parasites per sample were both manageable and accurate.