Evolutionary remodeling of ubiquinone biosynthesis in Toxoplasma gondii reveals an essential bi-functional monooxygenase Authors: Baihetiya Baierna, Taufiq Rahman, Scott Latimer, Gilles J Basset, Silvia N J Moreno Published: 2026 ABSTRACT AbstractApicomplexan parasites like Toxoplasma gondii harbor a divergent mitochondrial proteome, much of which remains uncharacterized despite its essentiality for parasite survival. One such essential pathway is ubiquinone (UQ) biosynthesis. Here, we characterize the UQ synthesis machinery in T. gondii and show that conserved enzymes, TgCoq3 and TgCoq5, are essential for growth and mitochondrial function, and associate in a multi-protein complex. Using proximity labeling and subcellular fractionation, approaches suited to detect low-abundance proteins, we identify TgCoqFMO, a FAD-dependent monooxygenase required for UQ synthesis. Unlike canonical eukaryotic systems that employ multiple monooxygenases to modify specific carbons on the UQ aromatic ring, TgCoqFMO catalyzes two distinct hydroxylation steps, an activity not previously reported in eukaryotes. Molecular docking and chemical screening identified TgCoqFMO inhibitors that impair tachyzoite growth and reduce bradyzoite viability. These findings reveal a divergent UQ biosynthesis pathway with fewer components in apicomplexans and establish TgCoqFMO as a promising antiparasitic target.Subject terms: Parasite physiology, Protein-protein interaction networks Introduction Toxoplasma gondii is an obligate intracellular parasite of the phylum Apicomplexa, which also includes major human pathogens such as Plasmodium and Cryptosporidium spp., and infects nearly one-third of the global population1. Following infection, T. gondii progresses through an acute stage characterized by rapid replication of tachyzoites, which can disseminate systemically and cause tissue damage. This is followed by a chronic stage, during which the parasite persists as a latent bradyzoite in tissue cysts primarily in muscle and brain tissue2. Reactivation of latent infection in immunocompromised individuals can result in severe tissue damage in the brain, heart and lungs3–5. Current treatment for toxoplasmosis, such as pyrimethamine combined with sulfadiazine6, are limited by significant toxicity and frequent relapses upon cessation6–8. Consequently, there is a pressing need for novel therapeutics and parasite-specific targets that are essential for both acute and chronic stages of infection. The mitochondrion of T. gondii is an important therapeutic target due to its central role in parasite viability and the presence of essential enzymes that are highly divergent from those of the host9. Notably, atovaquone, a drug used to treat malaria, toxoplasmosis and babesiosis, targets the mitochondrial electron transport chain (ETC) at the ubiquinone (UQ): cytochrome c reductase (complex III)10–12. This underscores the potential of the UQ biosynthesis as a promising target for therapeutic intervention. Ubiquinone (UQ), also known as coenzyme Q, is a redox-active lipid composed of a water-soluble quinone head derived from para-hydroxybenzoate (PHBA) and a polyisoprene tail that anchors it to the inner mitochondrial membrane13,14. In the T. gondii ETC, UQ functions as the central electron carrier linking multiple dehydrogenases to complex III. In this context, because the parasite lacks a canonical complex I, electron input into the UQ pool instead occurs through succinate dehydrogenase (complex II) and additional enzymes, including malate-quinone oxidoreductase (MQO), dihydroorotate dehydrogenase (DHODH), type II NADH dehydrogenase and glycerol-3-phosphate dehydrogenase (G3PDH)15–17. Beyond its role in oxidative phosphorylation, UQ is also required for pyrimidine synthesis, fatty acid β-oxidation, and redox homeostasis18–23. Most of our understanding of UQ biosynthesis derives from studies in yeast, mammals, and bacteria24, where a largely conserved pathway involving a series of enzymes has been defined25 (Supplementary Fig. 1). In eukaryotes, the first two enzymes of this pathway, Coq1, a soluble mitochondrial matrix protein, and Coq2 (COQ1 and COQ2 in mammals), an inner mitochondrial membrane bound protein, synthesize the isoprene tail and attach it to the benzoquinone ring, respectively25. Subsequent ring modifications include decarboxylation, followed by a series of hydroxylation and methylation steps to ultimately yield the fully substituted UQ molecule (Fig. 1). The hydroxylation of the benzoquinone ring differs between organisms. In yeast and mammals, C5 hydroxylation is carried out by an FAD-dependent monooxygenases (Coq6/COQ6 respectively), while C6 hydroxylation is catalyzed by a diiron hydroxylase (Coq7/COQ7 respectively)25. In contrast, bacteria and plants use FAD-dependent monooxygenases to perform hydroxylation at different carbon positions24,26. In yeast, the enzymes responsible for these ring decorations, Coq3-Coq9 and Coq11, assemble into a large multi-protein complex known as the UQ synthome25. A similar protein complex is also present in mammals (the UQ complex)27, and in bacteria (the UQ metabolome)28. These complexes are thought to enhance the catalytic efficiency and sequester potentially toxic intermediates, preventing oxidative damage. Despite the central role of UQ in cellular metabolism, the mechanism of UQ ring modifications and the presence and organization of a UQ synthesis complex have not been investigated in T. gondii or other apicomplexan parasites to date. Predicted ubiquinone biosynthesis pathway and evolutionary analysis of Coq enzymes in T. gondii. A Schematics of the predicted ubiquinone synthesis pathway in T. gondii adapted from the yeast pathway25. Enzymes homologous to those of the yeast pathway are shown in orange, and steps for which no gene has been predicted are indicated by question marks. B Structure of UQ and the enzymes involved in the ring group modification that have homologs in T. gondii (orange). Question marks indicate unidentified genes. C Predicted domain arrangement of the TgCoq3 (TGGT1_266850) protein with the mitochondrial signal peptide indicated in blue and the methyltransferase motifs (Mtase) in red. Conserved methyltransferase motif sequences93 were compared with the H. sapiens and S. cerevisiae Coq3 protein motifs. D Predicted domain arrangement of the TgCoq5 (TGGT1_295690) protein with the predicted mitochondrial signal peptide indicated in blue and methyltransferase motifs (Mtase) in red. Conserved methyltransferase motif sequences93 were compared with H. sapiens and S. cerevisiae Coq5 protein motifs. E Phylogenetics of Coq3 in different organisms. UbiE (bacterial homolog of Coq5) and COMT (Catechol-O-methyltransferase) are used as outgroups. Sequences used are listed in Supplementary Table 2. F Phylogenetics of Coq5 in different organisms. UbiG (bacterial homolog of Coq3) and COMT are used as outgroups. Sequences used are listed in Supplementary Table 2. Our previous work demonstrated that TgCoq1 can be targeted by lipophilic bisphosphonates29, commonly used to treat osteoporosis30, supporting the druggability of this pathway. However, bioinformatic analyses have failed to identify all the enzymes required for UQ ring modifications in T. gondii, suggesting that the parasite may use novel and divergent enzymes for these steps. Given the divergence of other mitochondrial protein complexes in T. gondii and the presence of parasite-specific subunits identified within each ETC complex31,32, we hypothesized that UQ synthesis in apicomplexans also involves unique enzymes assembled into a stable protein complex. In this study, we present the first biochemical characterization of the UQ biosynthesis machinery in T. gondii. We investigate the potential assembly of these enzymes into a multi-protein complex, with a focus on identifying parasite-specific enzymes that differ from those found in mammals and yeast. Our work aims to expand the understanding of mitochondrial metabolism in apicomplexan parasites and to explore novel targets for therapeutic intervention. Results The ubiquinone ring modifying enzymes of T. gondii To identify potential enzymes involved in the UQ synthesis (Fig. 1A), we first used the Saccharomyces cerevisiae UQ synthesis protein sequences and performed a BLASTP33 and HHPRED34 search of the T. gondii proteome35. For some of the corresponding proteins, strong homologs were found in T. gondii (Supplementary Table 1). Our laboratory had previously characterized TgCoq1 and found that it is responsible for the synthesis of the isoprenoid unit, heptaprenyl diphosphate29. However, bioinformatic searches could not identify obvious homologs for Coq6, Coq7 or Coq9 (Fig. 1Aand Supplementary Table 1), which carry out essential enzymatic steps of the synthesis of UQ or Coq11 (Fig. 1Band Supplementary Fig. 1). Interestingly, we identified homologs of Yah1 and Arh1, which provide electrons to Coq6 in S. cerevisiae25. Based on this, we hypothesized that T. gondii possesses a monooxygenase capable of catalyzing the required hydroxylation steps (Fig. 1B), but that it is likely highly divergent from its fungal and mammalian counterparts. To characterize UQ synthesis, we focused on two genes TGGT1_266850 and TGGT1_295690, annotated as 3-demethylubiquinone-9 3-O-methyltransferase (TgCoq3), and ubiquinone/menaquinone biosynthesis methyltransferase subfamily protein (TgCoq5) respectively (Supplementary Table 1). These enzymes catalyze key methylation steps in the UQ biosynthesis pathway: Coq3 functions as an O-methyltransferase, acting at the C5 and C6 positions of the aromatic ring, while Coq5 serves as a C-methyltransferase (Fig. 1B). Sequence alignment with yeast (Coq3 and Coq5) and human (COQ3 and COQ5) homologs revealed low identity (TgCoq3: 37.31% human, 27.71% yeast; TgCoq5: 40% human, 32.51% yeast). Despite this divergence, both protein sequences retain conserved methyltransferase motifs (Fig. 1C, D), supporting their predicted enzymatic function. Phylogenetic analyses including representative animal, plant, yeast, and apicomplexan homologs showed that TgCoq3 and TgCoq5 cluster with UQ synthesis O- and C-methyltransferases, respectively, and are most closely related to other apicomplexan proteins predicted to function in UQ synthesis (Fig. 1E, F). Both enzymes remain distant from mammalian homologs, highlighting their evolutionary divergence. TgCoq3 and TgCoq5 are important for T. gondii growth and UQ synthesis Subcellular localization of TgCoq3 (TGGT1_266850) and TgCoq5 (TGGT1_295690) was initially predicted to be targeted to the mitochondrion by LOPIT (Localization of Organelle Proteins by Isotope Tagging) analysis36 (Supplementary Table 1). To validate these predictions, we tagged the C-terminus of both genes with a triple HA tag and isolated clonal lines expressing TgCoq3-HA and TgCoq5-HA. Immunofluorescence microscopy analysis (IFA) showed partial co-localization of both proteins with the mitochondrial marker TOM4037 (Supplementary Fig. 2A). However, because the signal was relatively weak, the localization was not fully convincing. To improve detection, we generated polyclonal antibodies against both proteins. For this, we cloned both TgCoq3 and TgCoq5 genes in bacterial expression plasmids and purified the respective recombinant proteins and used them to immunize animals. We generated and purified αTgCoq3 in mouse and αTgCoq5 in guinea pigs, which were used to confirm the mitochondrial localization of both proteins (Supplementary Fig. 3). To study the role of TgCoq3 and TgCoq5 in T. gondii, we generated conditional knockdown lines by inserting a tetracycline (ATc) response element at the promoter region38 of each gene, followed by pyrimethamine selection and clonal isolation (iΔTgCoq3 and iΔTgCoq5) (Fig. 2A). Expression of both proteins was reduced 24 h after the addition of ATc to the cultures and completely ablated by 48 h as shown by western blot analysis and IFAs (Supplementary Fig. 2B, Cand Fig. 2B). We next studied the role of both proteins in the T. gondii lytic cycle with plaque assays, in which the parasite engages in repetitive lytic cycles resulting in host cell monolayer lysis forming plaques that can be visualized by staining with crystal violet. Both mutants, iΔTgCoq3 and iΔTgCoq5 formed significantly smaller plaques when grown in the presence of ATc (Fig. 2C, D). To facilitate measurement of replication and growth over several days, we expressed a cytosolic red fluorescent protein (tdTOMATO) in both mutants (iΔTgCoq3-RFP and iΔTgCoq5-RFP). This allowed us to quantify growth by measuring red fluorescence over time. Parasite numbers were significantly reduced in the iΔTgCoq3-RFP and iΔTgCoq5-RFP mutants when grown in the presence of ATc (Fig. 2E, F, + ATc). In a separate replication assay, ATc treatment of the mutants for three days resulted in a significant replication defect (Supplementary Fig. 2D). TgCoq3 and TgCoq5 are essential for the lytic cycle, mitochondrial function and UQ synthesis. A Schematics showing the genetic modification of the conditional mutants. The 5’ cassette includes a DHFR drug-resistant marker, a tetracycline response element and a SAG4 promoter. The primers used are listed in Supplementary Table 5. B IFAs of i∆TgCoq3 and i∆TgCoq5 cell lines. Parasites were incubated with ATc for 0, 1, or 2 days. αTOM40 was used as mitochondrial marker, and αTgCoq3 or αTgCoq5 were used to visualize TgCoq3 or TgCoq5 respectively. Bars are 5 µm. C Representative plaques of the i∆TgCoq3 and i∆TgCoq5 lines cultured in the presence or not of ATc. Plates were fixed and stained at day 7. D Quantification of plaque areas from three biological replicates. Statistical significance was evaluated by two-way ANOVA. Each point represents a biological replicate, and error bars indicate SEM. E Growth assay of i∆TgCoq3-RFP -ATc: no ATc; +ATc: parasites pretreated with ATc for 3 days before infection and cultured with ATc; + ATc + UQ6: same as +ATc, plus 10 µM UQ6 added to the culture. One-way ANOVA analysis was performed using three biological replicates, n = 3. Data is presented as mean ± SEM. F Growth assay with the i∆TgCoq5 mutant. Conditions were like the ones in E. One-way ANOVA analysis was performed using three biological replicates, n = 3. G Scheme of the T. gondii Electron Transport Chain. H Representative trace of membrane potential measurements in iΔTgCoq3 parasites using 2.5 µM safranin O, with 1 mM succinate present in the buffer as the respiratory substrate. -ATc: no ATc (brown trace); +ATc: parasites cultured with ATc for 2 days (gray trace). Oligo: 1 µg/mL oligomycin. I iΔTgCoq3 mitochondrial membrane potential quantification of the differences between the maximum fluorescence (after depolarization with 5 µM FCCP) and the minimum fluorescence (after 30 µM digitonin permeabilization). Two-tailed Student’s t test was used, n = 3. Each point represents a biological replicate, and the error bars represent SEM. J iΔTgCoq5 mitochondrial membrane potential quantification, + ATc: ATc for 3 days, other conditions are the same as the ones in (I). K Representative tracing of the oxygen consumption rate of the iΔTgCoq5 mutant, -ATc: no ATc (orange trace); + ATc: parasites cultured with ATc for 2 days (gray trace). L iΔTgCoq3 mOCR quantification. Basal, ADP-stimulated (10 µM), and oligomycin-inhibited mOCR were quantified from three biological replicates. Two-way ANOVA was used for statistical analysis. Each point represents a biological replicate, and bars indicate SEM. M mOCR was quantified from three biological replicates. Two-way ANOVA was used for statistical analysis. Each point represents a biological replicate, and error bars indicate SEM. For (D–F) and (I, J, L, and M), significance was considered when the p-value is less than 0.05. Data are provided in the Source Data file. To validate the function of TgCoq3 and TgCoq5 in UQ synthesis, we added ATc to the cultures to knockdown the respective genes and supplemented the medium with 10 µM UQ6, the end product of the yeast UQ synthesis pathway (Fig. 2E, F, + ATc + UQ6). Importantly, the growth repression of both mutants by ATc, was rescued by UQ6, supporting the role of TgCoq3 and TgCoq5 in UQ synthesis. However, the recovery was not 100%, possibly due to the limited permeability of UQ6 or a preference of T. gondii for UQ7 in mediating its cellular functions. TgCoq3 and TgCoq5 are required for mitochondrial function UQ transfers electrons from complex II to complex III, which then reduces cytochrome c in the intermembrane space (Fig. 2G). This electron transfer is linked to proton transport from the matrix to the intermembrane space, generating a membrane potential. We used safranin O29,39, a fluorescent dye that inserts into energized mitochondrial membranes, to measure membrane potential. Permeabilizing parasites in suspension with digitonin allowed safranin to enter the cells. In the presence of succinate, the mitochondrial membrane becomes energized, leading to accumulation of safranin and a corresponding decrease in fluorescence (Fig. 2H, brown trace). The addition of ADP to stimulate ATP synthesis resulted in depolarization of the membrane, which is observed as an increase in fluorescence. Inhibition of ATP synthase by oligomycin restored the membrane potential, observed as a decrease in fluorescence. Lastly, the addition of Carbonyl cyanide-p-trifluoromethoxy phenylhydrazone (FCCP) dissipated the membrane potential (Fig. 2H, brown trace). We next measured the membrane potential of the iΔTgCoq3 (+ATc) mutant, which showed to be significantly unresponsive to all additions (Fig. 2H, I, gray trace and bar). Analysis of the membrane potential of iΔTgCoq5 (+ATc) revealed a similar phenotype (Fig. 2J, gray bar). The ETC uses oxygen as its terminal electron acceptor, so we next assessed oxygen consumption of the mutants using an Oroboros O2K respirometer, which measures the rate of mitochondrial oxygen consumption (mOCR). Permeabilizing the plasma membrane with digitonin allowed for the substrate succinate to access and energize the mitochondria, enabling the measurement of basal mOCR (Fig. 2K, orange trace). The addition of ADP increased the mOCR as it stimulated oxidative phosphorylation. Lastly, inhibition of ATP synthesis with oligomycin reduced the rate of oxygen consumption to a minimal mOCR. All of the characteristics of healthy mitochondrial oxygen consumption were defective in the iΔTgCoq3 and iΔTgCoq5 mutants when grown in the presence of ATc (Fig. 2K–M, gray trace and bars). Taken together, both TgCoq3 and TgCoq5 are essential for T. gondii mitochondrial functions. The presence of an UQ synthesis complex in T. gondii To investigate the organization and potential interaction between the UQ synthesis enzymes (Fig. 3A), we performed Blue Native PAGE (BN PAGE) analysis. BN PAGE was used for the detection of mitochondrial membrane complexes, and this method has been used to visualize ETC complexes in T. gondii31,40. Lysates from TgCoq3-3HA and parental strain, prepared under native conditions, were separated by native PAGE, transferred to a membrane, and probed with αHA antibody, which showed no signal. However, protein transfer was confirmed by probing the membrane with the F1β antibody, which recognizes a subunit of the ATP synthase, revealing a clear reaction corresponding to Complex V (Supplementary Fig. 4A). Considering that both TgCoq3 and TgCoq5 are expressed at very low levels, we thought that this could explain the absence of detectable signals. To overcome this issue, we next enriched mitochondrial membranes through subcellular fractionation, using a differential centrifugation protocol that separates organelles primarily based on size (Supplementary Fig. 4B). To identify enriched mitochondrial fractions, we measured succinate cytochrome c reductase activity, a mitochondrial marker (Supplementary Fig. 4C, D). The P2 fraction showed the highest enrichment of activity, compared to the S1 (total lysate) fraction (Supplementary Fig. 4C). We then performed BNPAGE using the P2 fraction, however, the signal was still undetectable. We considered the possibility that the epitopes of TgCoq3 and TgCoq5 could be hidden within the hypothetical complex of T. gondii. To address this issue, protein complexes separated by BN-PAGE were subjected to second-dimension SDS-PAGE to resolve individual subunits. Under these conditions, both TgCoq3 and TgCoq5 were detected as components of a high molecular weight (> 1000 kDa) complex (Fig. 3B). UQ synthesis enzymes in T. gondii form a high molecular weight complex. A Schematic of the putative UQ synthesis complex adapted from the yeast UQ synthome model25. Proteins with homologs in T. gondii are shown in solid colors, while proteins lacking identifiable homologs are shown in light blue. B 2D BN/SDS-PAGE analysis (details in the methods section) using 100 μg of solubilized P2 fraction obtained using the protocol shown in Supplementary Fig. 4B. Membranes probed with αTgCoq3 and αTgCoq5 antibody. C 2D BN/SDS-PAGE analysis using TgCoq4-3HA and TgCoq8-3HA lysates (protocol detailed in the method section). Membranes probed with αHA antibody. Blue arrowhead indicates the UQ synthesis complex, and blue brackets indicate complex or disassembly intermediates. For (B and C), each experiment was repeated two to three times. To further validate the presence of a UQ complex and to test physical interactions among its components, we focused on TgCoq4 (TGGT1_275980) and TgCoq8 (TGGT1_203940), two conserved Coq proteins predicted to function as structural and regulatory elements of the UQ biosynthetic machinery. We generated TgCoq4-HA and TgCoq8-3HA lines (Supplementary Fig. 5A, B) and confirmed their mitochondrial localization by IFA (Supplementary Fig. 5C, D). Reciprocal co-IP experiments revealed interactions between TgCoq4, TgCoq5, and TgCoq8 (Supplementary Fig. 5E, F), supporting the existence of a multi-protein UQ synthesis complex in T. gondii. To further characterize the organization of the complex, we performed 2D BN/SDS-PAGE using lysates from TgCoq4-3HA and TgCoq8-3HA (Fig. 3C). Contrary to TgCoq3, which could only be detected using the P2 fraction, both TgCoq4 and TgCoq8 were detected in a protein complex larger than 1 MDa using whole cell lysates, likely due to their higher abundance or epitope orientation. However, each protein displayed different lower molecular weights species (indicated with blue brackets), which may indicate different assembly intermediates or partial disassembly during sample preparation. Together, these data provide the first evidence that T. gondii contains a multi-subunit UQ synthesis protein complex. Novel enzymes of the UQ synthesis pathway identified by TurboID pulldown and subcellular fractionation The proteome database failed to reveal several enzymes involved in the UQ synthesis pathway, including counterparts to key yeast enzymes responsible for catalyzing essential steps. We hypothesize that these enzymes are present in T. gondii but are highly divergent, making sequence analyses unrewarding. We initially performed αHA immunoprecipitation assays using whole cell lysates from TgCoq3-3HA and TgCoq5-3HA, however, the results were unproductive, likely due to the low expression levels of these proteins. Since we confirmed that TgCoq3 and TgCoq5 are both part of a large protein complex, we tagged their C-terminus with a biotin ligase, TurboID41, and isolated the clonal mutants TgCoq3-TID and TgCoq5-TID (Fig. 4A). This approach allows us to identify proteins in close proximity to the baits, while direct protein-protein interactions requires further experimental validation. We first validated that both TgCoq3-TID and TgCoq5-TID mutants were able to biotinylate mitochondrial proteins via IFAs (Fig. 4B). We performed a timepoint IFA by incubating the parasites with biotin for various lengths of time and saw labeling of the mitochondrion after 2 h of biotin incubation (Supplementary Fig. 6A). Identification of candidate UQ synthesis enzymes by TurboID and subcellular fractionation. A Scheme showing the tagging of TgCoq3 and TgCoq5 with TurboID. CAT: Chloramphenicol Acetyl Transferase gene for selection. B IFA of tagged cell lines incubated with DMSO or 50 μM biotin for two hours. Biotinylated proteins were probed with avidin green, and TOM40 was used as a mitochondrial marker. Scale bars are 5 µm. C Western blot analysis of eluates from streptavidin co-immunoprecipitation of TatiΔku80, TgCoq3-TID, and TgCoq5-TID parasites incubated with biotin for 2 h. Blots were probed with αHA (top) and αTgCoq3 (bottom). The experiment was repeated two times. D Pie chart showing the predicted localization by LOPIT36 of the significantly (p < 0.05) enriched hits with a fold-change (TgCoq5-TID/FbxO14-TID) higher than or equal to 2. Mito: mitochondria, N: nucleus, DG: dense granules, AP: apicoplast, PM: plasma membrane, IMC: inner membrane complex, MN: microneme, RP: rhoptry. E Volcano plots of the peptides detected in the eluates graphed based on their fold-change (TgCoq5-TID/FbxO14-TID) and their p-values. Red dots indicate peptides significantly enriched (at least 2-fold) in the eluate of TgCoq5-TID. Statistical analysis was done using Fisher’s exact test with Benjamini-Hochberg multiple test correction. The black arrowhead points at TgCoq3, TgCoq5, TgCoq8, Arh1 homolog and TGGT1_232540 peptides. The data used to generate this volcano plot are included in Supplementary Data 1. With the aim to find interacting proteins, we performed immunoprecipitation using streptavidin magnetic beads with total lysates of the parental and TgCoq3-TID and TgCoq5-TID mutants previously incubated with biotin. The eluates were probed with αHA antibody, confirming successful pull down of both TgCoq3-TID and TgCoq5-TID (Fig. 4C, top panel). We also probed the eluate with αTgCoq3, which confirmed the presence of TgCoq3 in the TgCoq3-TID lysate as expected. Interestingly, TgCoq3 was also detected in the TgCoq5-TID pull down, validating the close proximity of these two proteins (Fig. 4C, bottom panel). Note that the size of TgCoq3 appears larger in the TgCoq3-TID eluate because of the addition of the TurboID tag, which increases the molecular weight by ~ 35 kDa. Mass spec analysis of the eluate of the TgCoq5-TID mutant (total lysate) revealed that only 20% of the enriched peptides (TgCoq5-TID eluate/parental eluate) were predicted to be mitochondrial by LOPIT36, while 25% were predicted to be cytosolic (Supplementary Fig. 6B). Since the UQ synthesis enzymes are predicted to be expressed at low levels, their signals may be masked by peptides labeled by immature TgCoq5-TID or by nonspecifically associated cytosolic proteins. With the aim to increase enrichment of mitochondrial proteins and minimize cytoplasmic contamination, we developed a protocol that consisted in obtaining mitochondrial membranes enriched fractions of the TgCoq5-TID (+biotin), as it was expressed at higher levels than TgCoq3 (Fig. 4C). In addition, we performed a similar subcellular fractionation of a cell line expressing TID in the cytosol. Because expression of TID by itself in the cytosol proved to be toxic to the parasites, we instead used a C-terminally tagged TID of the cytoplasmic protein FBXO1442. Mitochondrial membrane enriched fractions (P2 as shown in the scheme shown in Supplementary Fig. 4B) were isolated from both cell lines and used as inputs for streptavidin IP. The eluates were analyzed by Mass Spectrometry, revealing high specificity of the approach, as 90% of the enriched peptides (fold change of peptides detected in TgCoq5-TID P2 over FBXO14-TID P2) were predicted to be mitochondrial (Fig. 4D). These enriched peptides are shown in red on the volcano plot (Fig. 4E). Importantly, the top hits from the pulldown (defined as those with more than a 4-fold enrichment and p-value < 0.0001) were all predicted to be mitochondrial (Table 1). The most significantly enriched proteins included TgCoq3, TgCoq5, TgCoq8, and an Arh1-like protein, a putative UQ synthesis enzyme predicted to act as an electron carrier for Coq6 (Supplementary Table 1). We also identified several uncharacterized proteins predicted to be essential. The complete list of proteins obtained are shown in Supplementary Data 1. Table 1Protein characteristics of top hits from streptavidin pulldown masspectrometryProteins highly enriched in the TgCoq5-TID P2 fraction and their characteristics.List of proteins with p < 0.0001 and fold change (TgCoq5-TID/ FbxO14-TID) higher than 4 identified by the mass spec analysis shown in Fig. 4E. GeneIDs, annotation in ToxoDB35, predicted localization from LOPIT36 (mm: mitochondrial membranes, soluble: mitochondrial lumen) are shown. Homology is indicated by the BLASTP expected value (e-value) between each T. gondii protein and its closest match in Pf (Plasmodium falciparum), Cp (Cryptosporidium parvum), Cu (Cryptosporidium ubiquitum), Sc (Saccharomyces cerevisiae), and At (Arabidopsis thaliana). Boxes marked with X indicate no homolog was identified. Protein characteristics of top hits from streptavidin pulldown masspectrometry Proteins highly enriched in the TgCoq5-TID P2 fraction and their characteristics. List of proteins with p < 0.0001 and fold change (TgCoq5-TID/ FbxO14-TID) higher than 4 identified by the mass spec analysis shown in Fig. 4E. GeneIDs, annotation in ToxoDB35, predicted localization from LOPIT36 (mm: mitochondrial membranes, soluble: mitochondrial lumen) are shown. Homology is indicated by the BLASTP expected value (e-value) between each T. gondii protein and its closest match in Pf (Plasmodium falciparum), Cp (Cryptosporidium parvum), Cu (Cryptosporidium ubiquitum), Sc (Saccharomyces cerevisiae), and At (Arabidopsis thaliana). Boxes marked with X indicate no homolog was identified. In summary, we developed a protocol for enrichment of mitochondrial proteins for the identification of peptides proximal to TgCoq5. This strategy reduced nonspecific labeling by TurboID from ~ 70% to less than 10%. This significant improvement enabled the discovery of several new candidates, likely part of the UQ synthesis pathway, which were undetectable in our initial experiments. We believe that combining biotinylation with subcellular fractionation offers a powerful strategy to uncover low-abundance, functionally important proteins that are otherwise difficult to detect in total lysates. Discovery of a novel UQ modifying enzyme of plant origin that is absent in mammals We reasoned that the combined proximity biotinylation and subcellular fractionation strategy would enable the identification of highly divergent proteins that form part of the UQ synthesis protein complex. To demonstrate this, we C-terminally tagged seven hypothetical proteins, identified by mass spectrometry and predicted to be mitochondrial, with a triple HA tag (Supplementary Fig. 7A), and confirmed their mitochondrial localization via IFA (Supplementary Fig. 7B). We next generated conditional knockdown mutants for the TgCoq4 gene and four hypothetical genes that had a negative CRISPR score and were not predicted to be ribosomal proteins (Supplementary Fig. 7C). We first confirmed by plaque assays that two hypothetical genes, TGGT1_288710 and TGGT1_251890, are essential for parasite growth, while TGGT1_239610 and TGGT1_236510, predicted to be essential by CRISPR-based screens, were dispensable (Supplementary Fig. 7D, E).To assess their potential involvement in the UQ synthesis, we performed host cell lysis assays in a 96 well plates, culturing the mutants in media without ATc, with ATc, or with ATc and UQ6. The growth defect of the i∆TgCoq4 mutant (+ ATc) was rescued by UQ6 supplementation, validating the function of TgCoq4 in UQ synthesis. In contrast, none of the mutants deficient in the hypothetical genes were rescued by UQ6 supplementation, indicating that although these proteins were biotinylated by TgCoq5-TID due to proximity, they are unlikely to be part of the UQ synthesis pathway (Supplementary Fig. 7F). One of the top hits from the mass spec analysis, TGGT1_232540 (hereafter referred to as TgCoqFMO), was annotated in ToxoDB35 as a “FAD binding domain containing protein”, and HHpred34 analysis of its sequence identified it as a “Flavin-dependent monooxygenase” (Table 1). We wondered if this gene and the corresponding protein could represent one of the missing activities for UQ synthesis, specifically Coq6, which is a flavin-dependent monooxygenase in yeast and mammals. Based on bioinformatic analysis, TgCoqFMO possesses three FAD binding domains (Fig. 5A). In fungi and animals, the enzyme that carries out C6 hydroxylation is Coq7, a carboxylate-bridged diiron hydroxylase43. Interestingly, in plants, this enzymatic step is carried out by a flavin-dependent monooxygenase (FMO)44. By constructing a maximum likelihood tree, we discovered that TgCoqFMO is divergent from Coq6, Coq7, and bacterial UQ-hydroxylases, and is closely related to the plant FMO responsible for C6 hydroxylation. Moreover, other apicomplexans, such as Plasmodium (PF3D7_0815300), also possesses a homolog of TgCoqFMO, that is absent in yeast and mammals (Fig. 5B). Phylogenetic analysis using an expanded sampling of TgCoqFMO homologs revealed a clear pattern: alveolate sequences cluster with deltaproteobacteria (Desulfobacteria), while stramenopiles group with plants and green algae. In contrast, red algal sequences fall outside this clade and instead group with more distant homologs or outgroups (Supplementary Fig. 8). TgCoqFMO is essential for parasite growth and mitochondrial function and forms part of the UQ complex. A Predicted domains by Interpro of TGGT1_232540 (TgCoqFMO). B Phylogenetics of TgCoqFMO, UQ-hydroxylases, Coq7, and Coq6 from different organisms. Sequences are listed in Supplementary Table 3. C IFA showing TgCoqFMO localization to the mitochondrion using αTOM40 as a mitochondrial marker and αHA to probe for TgCoqFMO. TgCoqFMO’s expression is downregulated by ATc. Scale bars represent 5 µm. D Western blot showing downregulation of TgCoqFMO with ATc incubation for three days. αHA was used to detect TgCoqFMO-3HA, and αSAG1 was used as a loading control. The experiment was repeated three times. E Representative plaques showing that knockdown of TgCoqFMO resulted in reduced plaque size. F Quantification of plaque areas from three biological replicates. Statistical significance was evaluated by two-way ANOVA. Each point represents a biological replicate, and error bars indicate SEM. G Representative tracing of membrane potential of iΔTgCoqFMO using Safranin O. + ATc: parasites incubated with ATc for 3 days. Oligo: oligomycin. H iΔTgCoqFMO mitochondrial membrane potential quantification of the differences between maximum fluorescence (after depolarization with FCCP) and minimum fluorescence (after digitonin permeabilization). Each point represents a biological replicate, and the bars represent SEM. Two-tailed Student’s t test was used to evaluate significance, n = 3. I 2D BN/SDS-PAGE of lysates from the i∆TgCoqFMO-3HA mutant. The membrane was probed with αHA and αTgCoq5. Blue arrowhead indicates the UQ synthesis complex, and blue bracket indicates complex or disassembly intermediates. The experiment was repeated three times. J Reciprocal co-IPs were performed using iΔTgCoqFMO-3HA lysates. TgCoqFMO-3HA, TgCoq3, and TgCoq5 were immunoprecipitated using αHA-, αTgCoq3-, and αTgCoq5-coupled beads, respectively. Equivalent amounts of input and IP eluates were analyzed by western blotting. αTOM40 was used as an IP control. See detailed protocol in the Methods section. The experiment was performed three times. K Volcano plot of peptides detected in the IP eluates plotted by fold change (TgCoqFMO-3HA/TgCoq1-3HA) and p-value. Red dots indicate peptides significantly enriched (at least 2-fold) in the eluate of TgCoqFMO-3HA that are involved in UQ synthesis. Statistical analysis was done using Fisher’s exact test with Benjamini-Hochberg multiple test correction. The blue arrowhead indicates TgCoq1, which was used as bait for TgCoq1-3HA IP. Data used to generate this volcano plot is included in Supplementary Data 2. For F and H, significance was considered when p-value was less than 0.05. Source data is provided in the Source Data file. To characterize TgCoqFMO, we tagged its C-terminus with HA and generated conditional knockdown lines in the HA-tagged background (iΔTgCoqFMO-3HA). We confirmed its mitochondrial localization (Supplementary Fig. 9A), and ATc regulation of expression via IFA and western blot analysis (Fig. 5C, D). To identify additional UQ synthesis enzymes of plant origin, we used the Arabidopsis Coq6 sequence (Q9LRM9) as a query and searched the T. gondii proteome using HHpred34. Interestingly, the top hits included TgCoqFMO and TGGT1_220360, the latter predicted to localize to the cytosol. HA tagging of TGGT1_220360 confirmed its cytosolic localization (Supplementary Fig. 9A). To characterize the role of TgCoqFMO, we performed plaque assays in the presence and absence of ATc, revealing significantly smaller plaques with ATc treatment compared with the parental line (Fig. 5E, F). Lastly, we investigated the role of TgCoqFMO in mitochondrial activity. Downregulation of TgCoqFMO expression by ATc caused a significant depolarization of the mitochondrial membrane in the mutant parasites (Fig. 5G, H). Importantly, using 2D BN/SDS-PAGE analysis of i∆TgCoqFMO-3HA whole cell lysates, we demonstrated that TgCoqFMO forms part of a large protein complex (Fig. 5I). We also probed the same membrane with αTgCoq5 and observed the same high MW complex (Fig. 5I). Note that the banding pattern differs from that observed in Fig. 3B, likely due to partial complex disassembly during sample preparation, as Fig. 3B was generated using the P2 mitochondrial fraction, while Fig. 5I was obtained from whole-cell lysates. Next, using reciprocal co-IPs, we validated the interactions between TgCoqFMO, TgCoq3 and TgCoq5 (Fig. 5J). By performing αHA IPs using the mitochondrial fractions from TgCoqFMO-3HA and TgCoq1-3HA (used as control), mass spectrometry analysis identified TgCoq3, TgCoq4 and TgCoq5 enriched when TgCoqFMO-3HA was used as bait (Fig. 5K), which further confirms the involvement of TgCoqFMO in the UQ synthesis complex. TgCoqFMO is a multifunctional monooxygenase essential for UQ synthesis To validate the role of TgCoqFMO in UQ synthesis, we performed growth assays using iΔTgCoqFMO-RFP parasites. Culturing the mutant with ATc resulted in a significant growth defect (Fig. 6A, + ATc, gray line and bar). Importantly, supplementation of the media with UQ6 rescued this defect (+ATc + UQ6, purple line and bar), demonstrating that TgCoqFMO participates in the UQ synthesis pathway of T. gondii (Fig. 6A). Importantly, HPLC analysis of extracts of the iΔTgCoqFMO mutant (±ATc) demonstrated that UQ7 levels in the ATc-treated parasites were reduced by more than ten-fold (Fig. 6B), accompanied by the accumulation of the demethoxy-ubiquinone7 (DMQ7) precursor (Supplementary Fig. 9B, C). Together, these metabolic data provide direct biochemical evidence that TgCoqFMO is essential for UQ biosynthesis in T. gondii. Genetic and chemical complementation reveal TgCoqFMO as a bifunctional monooxygenase. A Growth of i∆TgCoqFMO expressing RFP ± ATc or +ATc and 10 µM UQ6. Parasites were preincubated in ATc for three days prior to the infection. Data are presented as mean ± SEM. One-way ANOVA was used to evaluate significance, n = 4 biological replicates. The bar graph shows the number of parasites at day 4. B Quantification of UQ7 from the i∆TgCoqFMO cell line incubated ± ATc (6 days). Data is presented as mean ± SEM. Two-tailed Student’s t test was used, n = 3 biological replicates. C Structure of vanillic acid. D Structure of UQn and nomenclature of the eukaryotic hydroxylases involved in UQ ring modifications. E Growth of i∆TgCoqFMO-RFP complemented with the Arabidopsis thaliana FMO (cAtFMO). -ATc: i∆TgCoqFMO with ethanol control; + ATc + 10 µM VA: i∆TgCoqFMO grown with ATc supplemented with 10 µM Vanillic acid; cAtFMO: i∆TgCoqFMO complemented with AtFMO grown + ATc; cAtFMO + 10 µM VA: i∆TgCoqFMO complemented with AtFMO grown with ATc supplemented with 10 µM VA. Parasites were preincubated with ATc for three days prior to the infection. Data are presented as mean ± SEM. Two-way ANOVA was used to evaluate significance, n = 3 biological replicates. The bar graph represents the number of parasites at day 4. F Representative tracing of the oxygen consumption rate of the iΔTgCoqFMO mutant cultured using similar conditions to the ones used in (E) and presented using similar colors. G iΔTgCoqFMO mOCR quantification. Basal, ADP-stimulated, and oligomycin-inhibited mOCR were quantified from three biological replicates. Two-way ANOVA was used for statistical analysis, with the −ATc condition as the control group. Each point represents a biological replicate, and the bars represent SEM. For A, B, E and G, significance was considered when p-value was less than 0.05. Source data are provided as a Source Data file. To investigate the enzymatic activity of TgCoqFMO, we used analogs of UQ precursors to bypass the effects of TgCoqFMO repression. Vanillic acid, which contains a methoxy group at the C3 position, required for C5 hydroxylation, has been used to bypass COQ6 activity (Fig. 6C)45. Similarly, 2,4 dihydroxybenzoic acid, which is hydroxylated at both C2 and C4, required for C6 hydroxylation has been used to bypass COQ7 activity46. Although vanillic acid contains a methoxy group at the C3 position, and 2,4-dihydroxybenzoic acid carries hydroxyl groups at the C2 and C4 positions, the C2 and C3 substitutions correspond to the C5 and C6 positions, respectively, in the final UQ ring after prenylation and decarboxylation. Consequently, supplying these compounds bypasses the requirement for COQ6 mediated C5 hydroxylation and COQ7 mediated C6 hydroxylation respectively (Supplementary Fig. 9D). However, neither precursor analog was able to rescue the growth defect caused by TgCoqFMO knockdown. This suggests that TgCoqFMO is not a monofunctional enzyme solely hydroxylating the C5 or C6 position but rather may perform multiple enzymatic functions that cannot be bypassed by these modified ring precursors (Supplementary Fig. 9E, F). We also tested if 2,3,4-trihydroxybenzoic acid, which contains hydroxyl groups at C2, C3, and C4 positions, can bypass TgCoqFMO knockdown, however, this compound was toxic to the host cells, and parasites were unable to grow (Supplementary Fig. 9D–F). To confirm that the inability of benzoquinone analogs to bypass the pathway is not due to disassembly of the UQ complex upon TgCoqFMO knockdown, we assessed the expression levels of other UQ synthesis enzymes and evaluated the integrity of the UQ complex following TgCoqFMO depletion. Western blots and 2D BN/SDS-PAGE analysis showed that expressions of TgCoq3 and TgCoq5 levels were not affected (Supplementary Fig. 10A). Moreover, the UQ complex is still present when knocking down TgCoqFMO (Supplementary Fig. 10B). Since phylogenetic analysis indicates that Arabidopsis thaliana flavin-dependent monooxygenase (AtFMO) is the closest homolog to TgCoqFMO (Fig. 5B), and its function has been previously characterized44, we thought to complement the TgCoqFMO conditional knockdown cell line with AtFMO to infer the function of TgCoqFMO. This approach enables the evaluation of their functional conservation in UQ synthesis. To ensure correct mitochondrial localization of AtFMO, we generated a plasmid encoding a chimeric protein consisting of TgCoqFMO’s mitochondrial targeting pre-sequence and a codon-optimized mature sequence coding AtFMO. We validated the expression and mitochondrial localization of AtFMO (Supplementary Fig. 10C). However, growth kinetics revealed that AtFMO alone could not compensate for the loss of TgCoqFMO, as the mutant parasites displayed the same growth defect in the presence of ATc (Fig. 6E, cAtFMO + ATc, blue line and bar). Interestingly, supplementing the growth medium of the cAtFMO + ATc mutant with vanillic acid fully restored its growth (Fig. 6E, cAtFMO + ATc + VA, purple line and bar). Vanillic acid provides a 3-methoxy group, which bypasses the requirement for the C5 hydroxylation step. Under these conditions, AtFMO, which catalyzes only the C6 hydroxylation44, becomes sufficient to support UQ synthesis. Full rescue of growth occurred only when AtFMO expression (providing C6 hydroxylation) was combined with vanillic acid supplementation (providing C5 modification) (Fig. 6C, D). These results demonstrate that TgCoqFMO normally performs both the C5 and C6 hydroxylation reactions during UQ biosynthesis. To further validate TgCoqFMO’s function, we measured mOCR in the mutants. Consistent with the mitochondrial membrane potential defect in the TgCoqFMO knockdown mutant, we observed a significant reduction of the mOCR when TgCoqFMO was depleted (Fig. 6F, G: -ATc and +ATc). Interestingly, in the cAtFMO + ATc + VA mutant, basal mOCR was fully restored, while ADP-stimulated and minimal mOCR (after oligomycin treatment) were increased relative to the +ATc condition but did not reach −ATc levels (Fig. 6F, G: cAtFMO + ATc + VA). Together, these functional data further support that TgCoqFMO likely catalyzes both C5 and C6 hydroxylation steps, identifying TgCoqFMO as the first identified bi-functional monooxygenase described in eukaryotic ubiquinone biosynthesis. TgCoqFMO can be targeted by small molecules in both tachyzoites and in vitro differentiated bradyzoites The results presented highlight the discovery of TgCoqFMO, a unique enzyme absent in mammals that is integral to UQ synthesis, essential for mitochondrial function in T. gondii, and critical for parasite growth. These features underscore TgCoqFMO’s potential as a chemotherapeutic target. As a proof of concept for its targetability, we first predicted potential druggable pocket(s) on the AlphaFold-generated structure of TgCoqFMO using FTSite47. The latter identified few potential pockets that were proximal to the putative FAD binding site (Supplementary Fig. 11). Using a GRID box large enough to encompass these pockets, we virtually screened an FDA-approved drug library (DrugBank approved library48) using FRED (OpenEye), which led to the identification of three structurally distinct compounds as best candidates for experimental validation, given their predicted modes of interaction (Fig. 7A, B). Indeed, all three hits, namely fenticonazole, flibanserin and redafamdastat, inhibited parasite growth at low µM concentrations (Fig. 7C). To validate TgCoqFMO as the target of these compounds, we tested their EC50s in a TgCoqFMO overexpression line (TgCoqFMO-OE). Since the conditional knockdown line uses the SAG4 promoter, which is predicted to drive higher expression than the endogenous promoter49, we first confirmed the overexpression of TgCoqFMO in the i∆TgCoqFMO line via western blot (Supplementary Fig. 12A, B). We observed more than a threefold increase in the EC50 values for fenticonazole and flibanserin, and over a fourfold increase for redafamdastat, in the TgCoqFMO-OE line, whereas the EC50 for atovaquone, used as a control, remained unchanged. Moreover, host cells are still viable at 10X their respective EC50s, indicating the inhibition of parasite growth was not due to unhealthy host cells (Fig. 7C). Since the clinically used drug combinations, such as pyrimethamine and sulfadiazine, are only effective against tachyzoites, but lack activity against bradyzoites50, we next tested these compounds against in vitro differentiated bradyzoites. At 3X their respective EC50s, fenticonazole and redafamdastat significantly reduced bradyzoite viability (Fig. 7D, E, and Supplementary Fig. 12C), supporting the targetability of TgCoqFMO in tachyzoites and its potential against bradyzoites. The limited activity of flibanserin against in vitro-derived bradyzoites may reflect reduced permeability into tissue cysts, an issue that will need to be addressed in future studies. Molecular docking identifies TgCoqFMO inhibitors active against tachyzoites and bradyzoites. A The TqCoqFAD model (uniprot id: A0A139Y5I3) generated shown in a cartoon representation and colored based on the pLDDT score ranges shown. Global and close-up views of the predicted complex showing the proposed binding model of FAD and three DrugBank hits: fenticonazole, flibanserin, and redafamdastat, are shown on the left and right, respectively. All ligands are shown as sticks. FAD pose was modeled using AlphaFold3, while the poses of the three drugs are typical of the most reproducible and highest-ranked poses predicted through GOLD-based focused docking done in 10 independent runs for each drug. The figure was generated in PyMol (open source, version 2.5). B 2D ligand interaction diagrams of the poses of the three drugs shown in A. These diagrams were generated using PoseView accessed through the Protein Plus server. C EC50 values of the compounds against parental (Tati∆ku80 RFP), TgCoqFMO-OE (i∆TgCoqFMO-3HA RFP), and fold change between TgCoqFMO-OE over parental. Cytotoxicity was assessed at 10x the EC50 determined for the parental line. The assay was conducted with 3 biological replicates, each containing at least 2 technical replicates or 3 technical replicates for cytotoxicity. EC50s were presented as mean ± SEM. D Representative images showing the viabilities of bradyzoites treated with 3X EC50 of the inhibitors. DMSO and Atovaquone treatments were used as negative and positive controls, respectively. E Bar graph shows the quantification of relative bradyzoite viability, percentage of plaques normalized to DMSO treatment. Fent: fenticonazole, Flib: flibanserin, Reda: redafamdastat. Data are presented as mean ± SEM (n = 3 biological replicates). One-way ANOVA was used to test for statistical significance, which was defined as p < 0.05. Source data are provided as a Source Data file. In summary, these data support the potential of TgCoqFMO as a therapeutic target. Although the three compounds we tested exhibited EC50 values that are not outstanding, they nonetheless provide proof of concept that TgCoqFMO, an enzyme absent in mammals, could be effectively targeted for therapeutic intervention. Discussion In this work, we characterized key components of the ubiquinone (UQ) synthesis pathway in Toxoplasma gondii, identifying TgCoq3, TgCoq4, TgCoq5, TgCoq8, and a novel bifunctional flavin-dependent monooxygenase, TgCoqFMO, which is present in apicomplexan parasites but absent in mammals and yeast. We show that these proteins assemble into a mitochondrial UQ synthesis complex. Conditional knockdown of each component impaired parasite growth and mitochondrial activity, and the growth defects were rescued by exogenous UQ6 supplementation, validating their roles in UQ biosynthesis. Interestingly, iΔTgCoq3, iΔTgCoq5, and i∆TgCoqFMO were still able to proliferate following conditional knockdown, though at a significantly slower rate. This is most likely due to the long half-life of UQ51,52 and its ability to be recycled through the Q cycle, allowing residual UQ to support minimal growth even after knockdown. A key function of UQ is to shuttle electrons from various dehydrogenases (such as complex II) to complex III, which then transfers electrons to cytochrome c in the mitochondrial ETC. In agreement with this role, iΔTgCoq3, iΔTgCoq5 and iΔTgCoqFMO showed mitochondrial dysfunction, as evidenced by depolarization of their mitochondrial membrane potential and reduced oxygen consumption rates. Unlike other organisms, the T. gondii UQ synthesis complex migrates at over 1 MDa by two-dimensional BN/SDS-PAGE analysis, which is larger than the yeast UQ synthome (~690 kDa) and the mammalian complex Q (less than 480 kDa). This size difference resembles the distinctive features of other T. gondii mitochondrial complexes, which are larger compared to the yeast and human counterparts31,53. TgCoq5 was previously detected in the complexome profiling study by Maclean et al.31, however, none of the other UQ synthesis proteins were detected, likely due to their low abundance. The use of the biotin ligase41 approach facilitates the identification of low-abundance proteins, but it is prone to false positives, especially for mitochondrial proteins. Most mitochondrial proteins are encoded in the nucleus54, and they transiently associate with cytosolic proteins prior to their import into the mitochondrion. To overcome this limitation, we combined biotinylation with subcellular fractionation to separate mitochondrial proteins from cytosolic contaminants and included a cytosolic TurboID as a spatial control. This strategy significantly improved the mitochondrial yield from 20% in total lysates to 90% in mitochondrial fractions. Our mass spec results revealed that TgCoq5-TID is in proximity to TgCoq3, Coq8 and Arh1 homologs. Although TgCoq4 was not detected in the proximity labeling dataset, we independently validated its role in UQ synthesis: conditional knockdown of TgCoq4 caused a growth defect that was rescued by UQ6 supplementation, and 2D BN/SDS-PAGE analysis confirmed its association with the UQ complex. We hypothesize that TgCoq4 functions as an oxidative decarboxylase similar to its mammalian homolog COQ455. Since the UQ complex exceeds 1,000 kDa, we hypothesize that the biotin ligase fused to TgCoq5 could be oriented away from TgCoq4, limiting its biotinylation. Another explanation could be that the peptide signal is masked by more abundant subunits of the ETC complex. Interestingly, several UQ synthesis components that are part of the canonical CoQ synthome in yeast, including Coq6, Coq7, Coq9, and Coq11, do not have identifiable homologs in T. gondii. In yeast, Coq9 facilitates Coq7 activity, and Coq6 and Coq7 together carry out essential hydroxylation steps required for Coq3-mediated O-methylation. In contrast, our data indicate that T. gondii has replaced these activities with the bifunctional enzyme, TgCoqFMO, which performs the roles of both Coq6 and Coq7. In this context, the absence of Coq9 in T. gondii likely reflects functional streamlining of the pathway. Although the T. gondii UQ synthesis complex is larger in apparent molecular weight than those described in other organisms, it is composed of fewer distinct protein components, highlighting an evolutionary reorganization of the UQ synthesis machinery in apicomplexan parasites. TgCoqFMO lacks clear homologs in yeast or mammals, explaining why it was not detected in previous bioinformatic analyses. Phylogenetic analysis revealed that TgCoqFMO is most closely related to a plant flavin-dependent monooxygenase (FMO), known to catalyze C6 hydroxylation. Previous work suggests that the plant demethoxyubiquinone hydroxylase, a FAD-dependent monooxygenase, originated via horizontal gene transfer from a bacterium harboring the ancestral gene to green algae44. Considering that apicomplexans and plants both belong to Diaphoretickes, whereas animals and fungi belong to Amorphea56, it is not surprising that T. gondii utilizes an enzyme homologous to plant FMO for this essential mitochondrial pathway. However, this evolutionary pattern is not consistent with inheritance from the red algal endosymbiont that gave rise to the apicoplast and instead suggests that TgCoqFMO was likely acquired through other evolutionary routes, such as horizontal gene transfer. This raises the intriguing possibility that additional metabolic enzymes in T. gondii may have been acquired in a similar manner, contributing to the unique organization of its mitochondrial pathways. Using comparative analysis of mitochondrial proteomes, we identified only 25 mitochondrial proteins of presumed plant origin, including TgCoqFMO. Replacement of the iron-dependent Coq7 with an FAD-dependent monooxygenase may provide a selective advantage by reducing dependence on labile iron and limiting iron-catalyzed reactive oxygen species (ROS) production57,58. In T. gondii, which faces fluctuating iron availability during its life cycle, such an adaptation may help maintain mitochondrial function. In Plasmodium spp., despite replication within heme-rich red blood cells, iron is tightly regulated, representing a major source of oxidative stress rather than a freely available cofactor59. Thus, replacing the iron-dependent Coq7 with TgCoqFMO could provide significant advantages for apicomplexan parasites. Because TgCoqFMO is absent in the host and essential for parasite replication, it represents an attractive chemotherapeutic target for apicomplexan diseases such as toxoplasmosis and malaria. Moreover, although ATP synthase is dispensable in blood stage Plasmodium parasites60, UQ remains essential throughout all stages of the parasite’s life cycle. Apart from its role in oxidative phosphorylation, UQ also functions as an electron acceptor for dihydroorotate dehydrogenase, a key enzyme in the essential pyrimidine biosynthesis pathway61. UQ’s involvement in multiple critical biological processes, combined with the parasite inability to use host UQ1029, further supports TgCoqFMO as a promising drug target. To investigate whether a second FMO might participate in UQ synthesis, we queried the T. gondii proteome using the plant Coq6 sequence (Q9LRM9). The top hits were TgCoqFMO and TGGT1_220360, which did not localize to the mitochondrion. While eukaryotic monooxygenases involved in UQ synthesis, including the plant FMO, are strictly monofunctional, bi- and tri- functional monooxygenases, such as UbiL and UbiM, are present in certain bacteria species, providing precedent for a single enzyme catalyzing multiple hydroxylation steps (Supplementary Fig. 1). Interestingly, UbiL and UbiM are found in genomes that encode multiple UQ-hydroxylases, while in genomes encoding a single UQ hydroxylase, UbiM is typically the sole enzyme present, supporting a model in which bifunctionality compensates for reduced pathway complexity. UbiM has been identified in alpha-, beta-, and gamma-proteobacteria, but it is less distributed and conserved than other UQ-hydroxylases, and its origin remains a mystery57. The bioinformatic analysis, combined with the results of the genetic and chemical complementation experiments, provides strong evidence that TgCoqFMO catalyzes both the C5 and C6 hydroxylation steps of UQ biosynthesis. The inability of plant AtFMO to rescue growth unless vanillic acid is supplied demonstrates that TgCoqFMO performs a function that cannot be replaced by a mono-functional FMO. We propose that TgCoqFMO may have originated from an ancestral UbiM-like enzyme acquired through horizontal gene transfer. Our results also underscore the therapeutic potential of TgCoqFMO. Virtual screening guided by an AlphaFold predicted model identified structurally diverse compounds with strong predicted binding affinity for TgCoqFMO. The in vitro efficacy of these compounds, combined with their reduced potency in a TgCoqFMO overexpression line, provides functional evidence that they likely act on this target. Notably, this effect was not observed with atovaquone12, indicating that TgCoqFMO is not only required for long-term parasite growth but also druggable. This positions TgCoqFMO as a compelling target for future drug development efforts aimed at treating toxoplasmosis and potentially other apicomplexan infections. Despite these advances, we were unable to directly assess TgCoqFMO’s enzymatic activity in vitro. This limitation is shared with many UQ biosynthesis enzymes, which often require integration into multiprotein complexes for catalytic function62. Reconstituting such complexes under in vitro conditions remains challenging. In addition, while AlphaFold based molecular docking offers a valuable starting point for structural insights, the inherent uncertainty of predicted models must be acknowledged. Therefore, these docking results should be interpreted as preliminary and serve as proof of concept of TgCoqFMO’s potential clinical relevance. Future work aimed at resolving the high-resolution cryo-EM or crystal structure of TgCoqFMO will be critical for advancing drug design. Our in vitro bradyzoite drug testing data suggest that TgCoqFMO may represent a promising target in this stage of the parasite. However, additional work will be required to confirm the essentiality of the UQ synthesis genes for bradyzoites. Interestingly, some azoles have been reported to have modest anti-Toxoplasma activity63, although no studies have previously shown that fenticonazole inhibits apicomplexan parasite growth, making this study the first indication of its activity. It is important to note that these compounds might have pleiotropic effects. Future medicinal chemistry efforts and target-focused compound optimization could build on these initial hits to develop more potent and specific TgCoqFMO inhibitors. In summary, we uncover the unique UQ biosynthesis machinery in T. gondii, a key component of the parasite mitochondrial function (Fig. 8). We identified key components of the UQ complex, TgCoq3, TgCoq4, TgCoq8 and TgCoq5, and characterized TgCoqFMO, an apicomplexan-specific bi-functional flavin-dependent monooxygenase likely catalyzing both C5 and C6 hydroxylation. To our knowledge, this is the first example of a bi-functional UQ monooxygenase in a eukaryotic system. This unprecedented bifunctionality likely provides an evolutionary advantage in coping with iron fluctuations and redox stress. Importantly, TgCoqFMO is essential for parasite long-term proliferation, absent in hosts, and druggable, positioning it as a highly promising antiparasitic target. Our findings lay the ground for developing targeted therapies against toxoplasmosis, malaria, and related diseases. Proposed model for UQ biosynthesis and complex organization in T. gondii. A Proposed aromatic ring modification steps of UQ in T. gondii, highlighting the enzymes characterized in this study (TgCoq3, TgCoq4, TgCoq5, and TgCoqFMO) and the proposed electron donors for TgCoqFMO (Arh1 and Yah1 homologs). Question marks indicate steps that have not been experimentally determined. B Proposed model of the T. gondii ubiquinone synthesis complex. Methods Culture hTERT (human telomerase reverse transcriptase) cells were cultured in Dulbecco’s modified minimal essential medium (DMEM) supplemented with 10% BCS (Gemini Bio-Products, 100-506-500). Cell cultures were maintained in 5% CO2 at 37 °C. T. gondii tachyzoites were cultured in hTERT cells in DMEM with 1% BCS and purified as described previously29. The C-terminally tagged line mixed populations were cultured in DMEM with 1% BCS with 6.8 µg/mL chloramphenicol (Sigma, C0378). The conditional knockdown mixed populations were cultured in DMEM with 1% BCS containing 1 µM pyrimethamine (Sigma, P7771). Complementation lines mixed populations were cultured in DMEM with 1% BCS with 10 µM fluorodeoxyuridine (Sigma, F0503). The clonal lines were cultured in DMEM with 1% BCS. For downregulation of the genes, 0.5 µg/mL anhydrotetracycline (ATc) (Cayman Chemical Co Inc, 10009542-500) was added. Phylogenetic analysis Protein sequences were obtained from the National Center for Biotechnology Information (NCBI)33 database. The T. gondii TgCoq3 (TGGT1_266850), TgCoq5 (TGGT1_295690) and TgCoqFMO (TGGT1_232540) sequences were used to query the NCBI database using the genomic Basic Local Alignment Search Protein (BLASTP) tool by selecting organisms in the apicomplexan phylum. Other sequences for Coq3 and Coq5 were obtained through UniProt64 by searching Coq3 and Coq5 in eukaryotes and UbiG and UbiE in prokaryotes. The phylogenetic tree from Latimer et al.44 was used as a guide to search for Flavin-dependent monooxygenases in plants and green algae. Coq7 sequences were obtained by searching the sequences in UniProt. For supplementary Fig. 8, sequences were obtained from BLASTP using Blossum45 setting. The evolutionary history was inferred by using the Maximum Likelihood method and the JTT matrix-based model65. The bootstrap consensus tree inferred from 100 replicates66 is taken to represent the evolutionary history of the taxa analyzed66. Branches corresponding to partitions reproduced in less than 50% bootstrap replicates are collapsed. The percentage of replicate trees in which the associated taxa clustered together in the bootstrap test (100 replicates) are shown next to the branches66. Initial tree (s) for the heuristic search were obtained automatically by applying Neighbor-Join and BioNJ algorithms to a matrix of pairwise distances estimated using the JTT model and then selecting the topology with the superior log likelihood value. Evolutionary analyses were conducted in MEGA X67. Generation of mutant parasites For C-terminus tagging, we used CRISPR/Cas9-based genome editing68 to introduce a 3xHA tag and a chloramphenicol drug- resistant cassette into the 3’ end of the gene of interest (GOI) upstream to the stop codon. We used a single guide RNA targeting the 3’ end of the open reading frame into pSAG1-Cas9-U6-sgUPRT69 using Q5-site directed mutagenesis (NEB, M0491L). The primers used for all mutants are listed in the Supplementary Table 5, and we confirmed the insertion by Sanger sequencing. For each GOI, we amplified a donor DNA sequence encoding 3xHA from pLic-3HA-CAT38,70 in tandem with 40 bp flanks of the GOI as listed in Supplementary Table 5. For the generation of TurboID-tagged lines, we amplified a donor DNA sequence encoding TurboID-3HA from the 3xHA TurboID pLIC plasmid (modified from Addgene #107169) in tandem with 40 bp flanks of the 3’ end of TgCoq3 or TgCoq5. For the generation of conditional knockdown cell lines, we used a single guide RNA targeting the 5’ open reading frame of the GOI into either pSAG1-Cas9-U6-sgUPRT71 using Q5-site directed mutagenesis or the pSS01368 using restriction enzymes as previously described68. The primers used are listed in Supplementary Table 5, and the sequences were confirmed by sequencing. For each GOI, we amplified the donor DNA sequence encoding Tet7SAG4 and a pyrimethamine drug resistance cassette from the pDTS4myc plasmid38 in tandem with 40 bp flanks of the GOI using primer pairs listed in Supplementary Table 5. For each cell line, we transfected the modified Cas9 plasmid and donor DNA by electroporation into Tati∆ku80 or the C-terminus tagged line, as mentioned in the results. For generation of the complementation cell lines, we used a single guide RNA targeting the UPRT locus using pSAG1-Cas9-U6-sgUPRT71 and amplified a donor DNA sequence encoding the SAG4 promoter, TgCoqFMO full-length cDNA without a stop codon (synthesized by Twist Bioscience, TGGT1_232540), a 3xTy tag and a stop codon in tandem with 40 bp flanks of UPRT. For complementation with AtFMO, we cloned the codon-optimized AtFMO (synthesized by Twist Bioscience, UniprotID: Q8GYJ7) into the pSAG4-TgCoqFMO-3Ty plasmid using primers listed in Supplementary Table 5. To ensure correct localization, we retained the endogenous mitochondrial targeting sequence (MTS) from TgCoqFMO and removed the plant MTS from the AtFMO sequence. Transfected parasites were selected with 10 µM fluorodeoxyuridine. The generation of RFP expressing parasites was done as described72. In brief, the plasmid expressing tdTomato73 was transfected into the conditional knockdown mutants. The mixed populations were sorted two times using a Bio-RadS3 cell sorter, selecting for red fluorescent-expressing parasites. Positive clones were obtained through serial dilution. Cloning and antibody generation cDNA from TgCoq3 (TGGT1_266850) or TgCoq5 (TGGT1_295690) was cloned into the pQE80L (Qiagen) plasmid and transformed into E. coli BL21 codon plus. Protein expressions were induced with 1 mM IPTG, and both recombinant proteins were purified from the insoluble fraction. For TgCoq3, 100 µg of purified protein was first mixed with Freund’s complete adjuvant and the mix used to inoculate CD-1 strain female mice (intraperitoneal) followed by three boosts 2 weeks apart with 50 µg of protein in Freund’s incomplete adjuvant74,75. For TgCoq5, 200 µg of purified protein was first mixed with Freund’s complete adjuvant and used to inoculate Hartley strain female guinea pigs subcutaneously, followed by 3 boosts 3 weeks apart, each with 100 µg of protein in Freund’s incomplete adjuvant. The final serums from mice and guinea pigs were tested by western blots with total parasite lysates and validated with lysates of the corresponding conditional knockdown cell lines (validation is shown in supplementary Fig. 3C, F). Both antibody serums were affinity-purified prior to use for IFAs and westerns. Work with mice and guinea pigs were carried out in strict accordance with the Public Health Service Policy on Humane Care and Use of Laboratory Animals and Association for the Assessment and Accreditation of Laboratory Animal Care guidelines. The animal protocol was approved by the University of Georgia’s Committee on the Use and Care of Animals (protocol A2021 03-005-A5). All efforts were made to humanely euthanize the mice and guinea pigs after collecting blood. Animals housing conditions include 12/12 h light/dark cycles, temperatures of 22 ± 2 °C, with 45%–65% humidity. Immunofluorescence assay (IFA) and western blots For IFAs with intracellular tachyzoites, sub-confluent hTERT cells seeded onto coverslips were infected for 20–24 h and the coverslips fixed with 3% paraformaldehyde for 15 min and permeabilized with 0.25% Triton X100 for 10 min. The slides were blocked in PBS pH 8.0 containing 3% bovine serum albumin (BSA) for 30 min prior to incubation with the respective antibody. Primary and secondary antibodies were incubated for 1 h each. Antibodies were diluted in PBS pH 8.0 with 3% BSA. Primary purified mouse αTgCoq3 was used at 1:25, purified guinea pig αTgCoq5 was used at 1:50, rabbit αTOM4037 (gift from the van Dooren lab) was used at 1:20,000, and mouse αHA (bioLegend) was used at 1:100. Secondary antibodies include Avidin Green (Invitrogen, A6374), GαM 546 (Invitrogen, A21123), GαGP 546 (Invitrogen, A11074), GαRab 488 (Invitrogen, A11008), were used at 1:1000. Coverslips were mounted on glass slides with DAPI fluoromount-G. Images were taken with a Delta Vision microscope (Olympus, IX-71) with a Photometric Coolsnap camera. Deconvolved images were obtained by using the softWoRx deconvolution software. For western blots, PAGE-separated proteins were transferred onto nitrocellulose membranes and blocked in 5% non-fat milk in PBS-T (0.1% Tween-20 in PBS) for 30 min. Incubation with primary antibodies was done overnight at 4 °C, and secondary antibodies for 1 h at room temperature. Antibodies were diluted in PBS-T. Primary purified mouse αTgCoq3 was used at 1:250, primary purified guinea pig αTgCoq5 was used at 1:500, mouse αHA (bioLegend) was used at 1:1000, mouse αHA (gift from Christopher West’s lab) at 1:10,000, mouse αTubulin at 1:30,000 (Sigma, T6199), mouse αF1β (Agrisera) at 1:10,000, and mouse αSAG176 (gift from the Carruthers lab) was used at 1: 10,000. Secondary antibodies used were GαM CW800 (LI-COR, 925- 32210), GαGP 680RD (Neta Scientific, LIC-925-68077), GαRab HRP (Bio- Rad, 1706515), GαM HRP (Invitrogen, 31430), GαGP HRP (Invitrogen, A18769) and Streptavidin CW800 (LI-COR, 926-32230), used at 1:10,000. The protein bands were imaged with a Licor Odyssey CLx equipment (LI-COR, 9140) or imaged with a ChemiDoc system (Bio-Rad, ChemiDocTM Imaging system) using HRP secondary antibodies. Lytic cycle assays, UQ6 rescue, and analog bypass hTert cells were seeded onto 6-well plates and cultured until confluent. A total of 150 parasites were plated onto the wells for each condition. The parasites were incubated for 7 days without disruption, with or without ATc. At day 7, infected host cells were fixed with 100% ethanol and stained with gentian violet as previously described77. ImageJ software was used to measure the size of the plaques. 30 plaques were randomly selected and measured for each condition and averaged for each biological repeat. For the 96 well cell lysis assay, hTert cells were seeded onto 96 well plates and cultured until confluent. A total of 4000 parasites were plated in each well in triplicate. The parasites were incubated for 5 days either without ATc, with ATc, or with ATC plus 10 µM UQ6. Infected host cells were then fixed and stained as described above, and the absorbance of gentian violet was measured at 592 nm. Replication assays were carried out as previously described78. hTert cells were seeded on coverslips in 12-well plates and cultured until sub-confluent. A total of 5 × 105 RFP-expressing parasites were used to infect the host cells and cultured with or without ATc for 24 h before fixing using the same protocol as for IFA. A fluorescence microscope was used for counting parasites per vacuoles. Two coverslips were infected per condition, and 100 parasitophorous vacuoles were randomly selected per coverslip. Three biological experiments were conducted. Growth assay experiments were carried out using parasites expressing tdTomato (red fluorescent protein)79 with a previously described modification80. RFP-expressing parasites were preincubated with ATc for two or three days prior to the growth assay. Preincubated and non-preincubated parasites were used to infect confluent hTert cells in 96-well plates. Parasites were cultured in clear DMEM media without ATc, with ATc, with ATc plus 10 µM UQ6, or with ATc plus vanillic acid or 2,4 di-hydroxybenzoic acid (concentrations as indicated in the figure legend). Fluorescence from RFP was measured daily with a Biotek synergy H1 plate reader (Agilent, H1MFDG). Standard curves were used to convert the fluorescence values to number of parasites72. Mitochondrial membrane potential and oxygen consumption i∆TgCoq3, i∆TgCoq5, and i∆TgCoqFMO-3HA were used to infect confluent hTert in T75 flasks and cultured with or without ATc for two days. Freshly egressed parasites were collected, washed with BAG (116 mM NaCl, 5.4 mM KCl, 0.8 mM MgSO4, 50 mM HEPES, pH 7.3 with 5.5 mM Glucose) and counted. For mitochondrial membrane potential and oxygen consumption, we used protocols previously described16,29,39,81. For mitochondrial membrane potential, 5 × 107 parasites was added to a cuvette previously prepared with intracellular buffer (125 mM sucrose, 65 mM KCl, 1 mM MgCl2 and 2.5 mM KH2PO4 in 10 mM HEPES-KOH buffer pH 7.2) containing 1 mM succinate, used as mitochondrial substrate. Safranin O (2.5 µM) was used as a fluorescent probe, and its fluorescence was measured in a Hitachi F-7000 fluorescence spectrophotometer. Digitonin (30 µM) was used to selectively permeabilize the plasma membrane, followed by 10 µM ADP, to stimulate ATP synthesis, thus causing depolarization of the mitochondrial membrane. Addition of 1 µg/mL of oligomycin would repolarize the membrane, and lastly, addition of 5 µM FCCP depolarized the mitochondrial membrane. The mitochondrial oxygen consumption rate was obtained using an Oroboros O2K respirometer (https://www.oroboros.at/). A total of 108 parasite was added to one of the chambers containing intracellular buffer and succinate. Parasites were permeabilized with digitonin, and the basal mCOR was measured. The addition of 200 µM ADP would stimulate the mOCR. Lastly, 1.5 µM of oligomycin was added to obtain the minimal mOCR. Subcellular fractionation The protocol for subcellular fractionation and mitochondrial membrane isolation was adapted from the one previously used for the purification of the plant like vacuolar compartment or acidocalcisomes82,83. Mutant parasites, either tagged with 3HA (TgCoq5-3HA) or with the modified biotin ligase turbo ID (TgCoq5-TID), were used to infect 10 T75 flasks per cell line (3-4 x109 parasites in total). When using the TgCoq5-TID or the FBXO14-TID mutant, the growth media was supplement with 50 µM biotin for 2 h (TgCoq5-TID) or 1 h (FBXO14-TID) prior to collection. The pellet from the last wash was mixed with silicon carbide and grinding followed for approximately 2 min using an ice-cold mortar and pestle. When approximately 90% of the parasites were lysed, the mixture was transferred to an open top tube with lysis buffer (125 mM sucrose, 50 mM KCl, 4 mM MgCl2, 0.5 mM EDTA, 20 mM HEPES-KOH pH7.2, mini complete protease inhibitor, 12 µg/mL DNase, 12 µg/mL RNase, and 8 µg/mL nocodazole). Differential centrifugation speeds and times are shown in Supplementary Fig. 4B. A total of 500 µL of each soluble fractions were collected, and the pellets were resuspended in intracellular buffer. Aliquots from different fractions were then subjected to mitochondrial activity assay and western blots. The P2 fractions (higher mitochondrial activity) were used for blue native PAGE, 2D BN/SDS-PAGE, and for immunoprecipitation with streptavidin beads. Blue native and 2D BN/SDS-PAGE The P2 fractions from the subcellular fractionation was solubilized in mitochondrial lysis buffer containing 2% dodecyl maltoside, 250 mM sucrose, 20 mM Tris HCl pH 8.0, 2 mM EDTA, and 750 mM amin-n-caproic acid as previously described84. A total of 100 µg of solubilized protein or the lysate from 108 parasites was loaded onto the first dimension BNPAGE gel, and solubilized bovine heart mitochondria40,85 was loaded as molecular weight marker for mitochondrial complexes. At the end of the run, the lane containing the bovine heart mitochondria was cut off and stained with Coomassie. The lane containing the solubilized P2 fraction or parasite lysate was cut and denatured with 1X Laemmli buffer with 100 mM DTT and heated for 20 s in a microwave. After the gel lane was cooled to RT, it was transferred and run in a SDS PAGE gel (second dimension). The protein gel was then transferred onto a PVDF membrane at 25 V, 4 °C, overnight. The membrane was fixed with acetic acid and de-stained with methanol as previously described84. The membrane was then probed with the antibodies mentioned in the western blot section. Proteins were then imaged with a chemiluminescence using ChemiDoc. Each BNPAGE and BN/SDS-PAGE were performed at least twice with the representative images shown in the figures. Succinate cytochrome c reductase activity assay The assay was carried out using a previously published protocol86. Two µL of each fraction was used for each technical replicate (3 in total). After the addition of reaction buffer (0.3 mM succinate, 0.3 mM KCN, 10 mM HEPES) containing succinate as a substrate, the reaction was initiated after the addition of cytochrome c (0.1 mM final). The absorbance of reduced cytochrome c was measured at 550 nm using a Biotek synergy H1 plate reader (Agilent, H1MFDG) for 10 min while shaking at 30 °C. The activity of each fraction was then normalized based on its protein concentration measured by BCA. Pulldown of biotinylated proteins Fractions were lysed in RIPA buffer (150 mM NaCl, 0.1 % SDS, 0.5 % sodium deoxycholate and 1% NP-40 in 50 mM HEPES pH 7.5 buffer) supplemented with mini complete protease inhibitor, and soluble protein concentrations were measured using BCA. A total of 100 µg of proteins were used as input and bound to streptavidin magnetic beads overnight at 4 °C. The experiment was carried as previously described41. In brief, the beads were washed with RIPA buffer, 1 M KCl, 0.1 M Na2CO3, 2 mM urea, and then twice with RIPA buffer. Proteins were eluted with 20 mM Biotin in 1% SDS and 25 mM Tris, pH 8 by heating at 75 °C for 30 min. Elutaes were sent for mass spectrometry analysis. Reciprocal co-immunoprecipitation and identification of TgCoqFMO interactors For reciprocal co-IPs, IPs were carried out as described previously31 with modifications. Protein-G magnetic beads (Thermo Fisher, 88847) were incubated with mouse αTgCoq3 or guinea pig αTgCoq5 in coupling buffer (10 mM NaH2PO4, 150 mM NaCl pH 7.2). Antibodies were crosslinked to the beads using 220 mM dimethyl pimelimidate (DMP, Sigma, D8388) in 200 mM sodium borate buffer (pH 9.0), and the reaction was quenched with 200 mM ethanolamine (pH 8.5). Parasites were lysed in buffer containing 1% digitonin, 20 mM HEPES (pH 7.4), 200 mM NaCl, and 2 mM MgCl2, supplemented with a protease inhibitor cocktail (Roche), for 1 h on ice. Lysates were clarified by centrifugation at 21,000 x g for 15 min at 4 °C. An aliquot of the clarified lysates was saved as input, and the remainder was incubated overnight at 4 °C with αHA magnetic beads (Thermo Fisher, 88836) or with αTgCoq3- or αTgCoq5-coupled beads. Beads were washed twice with buffer containing 0.05% digitonin, 20 mM HEPES (pH 7.4), and 200 mM NaCl, followed by one wash with the same buffer containing 500 mM NaCl, and a final wash with 200 mM NaCl. Bound proteins were eluted by boiling in sample buffer for western blot analysis. For identification of TgCoqFMO interactors, TgCoqFMO-3HA and TgCoq1-3HA (control) were subjected to subcellular fractionation to enrich mitochondrial proteins. To preserve complex integrity, 10 µM UQ6 was included in the lysis buffer. P2 fractions were solubilized in buffer containing 1% digitonin, 20 mM HEPES (pH 7.4), 200 mM NaCl, and 10 µM UQ6, supplemented with protease inhibitors, and incubated for 2 h at 4 °C. Lysates were clarified by centrifugation at 21,000 × g for 30 min at 4 °C. Clarified lysates (200 µg, as determined by BCA assay) were incubated with αHA magnetic beads and washed as described above. Beads were additionally washed with PBS (pH 7.4) prior to on-bead digestion and mass spectrometry analysis. Protein identification using LC-MS/MS Samples were sent to the Proteomics and Metabolomics Facility, Nebraska Center for Biotechnology, University of Nebraska-Lincoln, and analyzed by mass spectrometry (MS) as previously described87. Briefly, an aliquot of 37.5 µL of sample was added to 12.5 µL of 4X reducing NuPAGE LDS gel (Thermo Fisher Scientific, Waltham, MA) sample buffer containing 5 mM dithiothreitol (DTT) and incubated at 95 °C for 10 min. The samples were loaded and run on a Bolt 12% Bis-Tris-Plus gel (Thermo Fisher Scientific) in MES SDS running buffer to clean up the samples and concentrate the proteins into the top of the gel. The gel was then fixed in methanol:acetic acid:water (40:10:50) and stained with Colloidal Coomassie blue G-250. The gel containing proteins was excised and destained in 50% acetonitrile (ACN), 50 mM ammonium bicarbonate. The proteins were reduced in 100 mM ammonium bicarbonate with 10 mM DTT. The reducing buffer was removed, and proteins were alkylated with 10 mM iodoacetamide. Proteins were digested with 250 ng of trypsin overnight at 37 °C. Peptides were extracted from the gel pieces, dried down, and re-dissolved in 5% acetonitrile, 0.2% formic acid. Each digest was run by nanoLC-MS/MS using a 2 h gradient on a Waters CSH 0.075 mm × 250 mm C18 column (Waters Corp, Milford, MA) feeding into a Thermo Orbitrap Eclipse mass spectrometer. All MS/MS samples were analyzed using Mascot (Matrix Science, London, UK; version 2.7). Mascot was set up to search the cRAP_20150130.fasta (125 entries); and ToxoDB-59_TgondiiGT1_AnnotatedProteins_20221003 (8460 sequences), assuming the digestion enzyme trypsin. Mascot was searched with a fragment ion mass tolerance of 0.060 Da and a parent ion tolerance of 10.0 PPM. Carbamidomethyl of cysteine was set as a fixed modification. Deamidation of asparagine and glutamine, and oxidation of methionine was specified in Mascot as variable modifications. Scaffold (version Scaffold_4.8.9; Proteome Software Inc., Portland, OR) was used to validate MS/MS-based peptide and protein identifications. Hits were identified with a 95% protein threshold and a peptide threshold, and the minimum number of peptides detected was set to 2. Statistical analysis was done using Fisher’s exact test with Benjamini-Hochberg multiple test correction to identify significantly enriched proteins. To obtain a numeric fold-change value, peptides with values of 0 were adjusted to 0.0001. Analysis of quinone content Parasite pellets (109) were resuspended in 0.4 ml of 95% (vol/vol) ethanol, transferred to a 5 ml pyrex tissue grinder, and homogenized for 1 min. Homogenates were transferred to a 10 ml pyrex tube containing ~ 1 ml of 0.1 mm zirconia/silica beads (Biospec Products # 11079101z). The tissue grinder was rinsed twice with 0.3 ml of 95% (vol/vol) ethanol, and the washes were combined to the original extracts. Samples were vortexed vigorously for 3 min and then heated at 70 °C for 10 min. Samples were vortexed again for 3 min while cooling down and then mixed with 0.5 ml of water. The extracts were then partitioned three times with 5 ml hexane. Hexane layers were combined, evaporated to dryness with gaseous nitrogen, and resuspended in 0.3 ml of methanol:dichloromethane (10:1). Extracts were analyzed by HPLC on a 5 µM Supelco Discovery C-18 column (250 × 4.6 mm, SigmaAldrich), thermostated at 30 °C and developed isocratically at a flow rate of 1 ml/min with methanol:hexane (95:5; vol/vol). DMQ7 (10.5 min), UQ7 (11.3 min) and UQ10 (32.2 min) were detected spectrophotometrically at 275 nm and quantified according to external calibration curves. The corresponding quinol versions were fully re-oxidized during the heating step and were quantified as part of the quinone pool. UQ10 standard was from Sigma-Aldrich; UQ7 standard was extracted from Candida utilis and purified via reverse-phase chromatography. DMQ7 was HPLC-purified from iΔTgCoqFMO cell extracts (+ ATc) and its identity was verified using a hybrid quadrupole orthogonal time of flight spectrometer (Agilent Technologies) using the same conditions as those described in Latimer et al (2021)44. Ammonium adduct of DMQ7 was detected at m/z 782.65. Computational modeling, virtual screening and docking A model of the full-length mature (i.e., devoid of the first 149 amino acids that are cleaved upon mitochondrial targeting) TgCoqFMO (Uniprot accession no. A0A139Y5I3) was modeled along with FAD as a cofactor using AlphaFold 388. The best-ranked model was further refined through GalaxyRefine server89, which significantly improved the quality of the model further, and the final model had the Molprobity score of 1.568, clash score of 11.3, 0.1% poor rotamer and 99.1% residues as Ramachandran favored. Three potential druggable pockets proximal to the FAD molecule onto the refined TgCoqFMO model were predicted by FTSite47. Using a GRID box spanning these potential druggable pockets, virtual screening (VS) was performed with DrugBank 6.0 (approved) library48 using FRED implemented in OEDocking 4.3.2 suite (OpenEye, Cadence Molecular Sciences, Inc., Santa Fe, NM) at its default setting. From the top 100 hits, molecules with total polar surface area (TPSA) > 80 were excluded, and the remaining ones were blindly docked to the TgCoqFMO model using AutoDock Vina version 1.1.290 with exhaustiveness of 16. The hits preferentially going to the desired sites were subjected to focused docking using GOLD 5.3 suite (CCDC, Cambridge). Finally, the best 3 molecules were chosen for experimental testing, considering pose and docking scores used by Vina and GOLD. All ligand molecules for docking were obtained from PubChem. For Fig. 7B, 2D ligand interaction diagrams were generated using PoseView through the Protein Plus server91. Black dotted lines indicate hydrogen bonding; green dotted lines indicate p-p stacking, whilst green contour lines indicate hydrophobic and van der Waals interactions. Small molecule inhibition and host cell viability EC50 of the compounds were obtained by following the protocol as previously described80. In brief, hTERT cells were seeded onto black 96-well plates and cultured for two days prior to the addition of 4000 Tati∆ku80 RFP or i∆TgCoqFMO-3HA RFP parasites per well and various concentrations of the small molecules. Fluorescence of RFP was measured on the plate reader, and the EC50 values were calculated using GraphPad Prism. hTERT cells viability was assayed using Alamar Blue as described80 using 10X EC50 of the compounds against Tati∆ku80. In vitro bradyzoite viability assay The assay was carried out as described previously92. In brief, confluent HFF monolayers were preincubated with 3 µM compound 1, 4-[2-(4-fluorophenyl)-5-(1-methylpiperidine-4-yl)-1H-pyrrol-3-yl] pyridine, for 2 h before infection with ME49∆ku80. After 24 h, the medium was replaced, and cells were cultured at 37 °C under ambient CO2 conditions for 3 days to allow differentiation. Differentiated bradyzoites were then treated with drugs at 3X EC50 for 4 days. Bradyzoites were liberated by acid/pepsin digestion (170 mM NaCl, 60 mM HCl, and 0.1 mg/mL pepsin) for 10 min in a 37 °C water bath and neutralized with 94 mM Na2CO3. A total of 10,000 bradyzoites were plated per well in a fresh confluent HFF six-well plate and allowed to grow in a 37 °C incubator with 5% CO2 for 16 days. Plates were then fixed with 100% ethanol and stained with crystal violet as described in the plaque assay section. Plaques were counted using a light microscope (Nikon Eclipse TS100) with a 4X objective. Statistical analysis Experimental data were expressed as the mean with standard error of the mean (SEM) from at least three biological repeats unless indicated otherwise. Statistical analyses were performed using GraphPad PRISM (version 10). Student’s t test, one-way ANOVA, or two-way ANOVA tests were used when appropriate and indicated in the figure legends. Significance was considered when the p-value was less than 0.05. Reporting summary Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article. Supplementary information Supplementary Information Description Of Additional Supplementary File Supplementary Data 1 Supplementary data 2 Reporting summary Transparent Peer Review file Source data Source Data Acknowledgements We thank the CTEGD FACS sorting and biomedical microscopy cores for the use of their facilities and equipment. Julie Nelson and Muthugapatti K Kandasamy provided training and support. We thank the Proteomics & Metabolomics Facility (RRID:SCR_021314), at the Nebraska Center for Biotechnology at the University of Nebraska-Lincoln for the mass spectrometry analysis. The facility and instrumentation are supported by the Nebraska Research Initiative. We thank John Smith, a veterinary scholar for performing the initial BLASTP search during the summer of 2020. We thank Melissa Ann Sleda for initial training on the use of the established Toxoplasma protocols, Mayara Bertolini for training on the use of the Oroboro and protein purification, Jessica Kissinger for help with bioinformatics, Elisabet Gas Pascual from Chris West lab for training on co-immunoprecipitation, and Madelaine Usey for training on BlueNative PAGE. Andrea Hortua Triana directed the generation of the αTgCoq3 and αTgCoq5 antibodies. The FbxO14-TID cell line was a gift from Diego Huet lab. The αTOM40 antibody was a gift from Giel van Dooren, and the monoclonal αHA antibody was a gift from Chris West. We thank Justin Wiedeman from the Etheridge lab for sharing his protocol for the elution of biotinylated proteins from streptavidin beads. TR acknowledges OpenEye (Cadence Design Systems) for granting an academic license for Omega and FRED (OEDocking). This work was funded by NIH grant AI169846 to S.N.J.M. and NSF grant MCB-2216747 to G.J.B. B.B. was partially funded by an AHA predoctoral fellowship 24PRE1192541. Author contributions B.B performed most of the experiments, analyzed the data and wrote the manuscript. T.R. performed molecular docking, virtual screening, and edited the manuscript. S.L. and G.J.B. performed lipid extraction and quinone measurements via HPLC and edited the manuscript. S.N.J.M. supervised the project and wrote and edited the manuscript. Peer review Peer review information Nature Communications thanks Nishith Gupta, and the other anonymous reviewer(s) for their contribution to the peer review of this work. A peer review file is available. Data availability The proteomics data generated in this study have been deposited in the ProteomeXchange Consortium (http://proteomecentral.proteomexchange.org) via the PRIDE partner repository with the identifier numbers: PXD059833 (Fig. 4E) and PXD071065 (Fig. 5K). All data are included in the article or the Supplementary Information. Source data are provided in this paper. Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Supplementary information The online version contains supplementary material available at 10.1038/s41467-026-71902-1. References Associated Data Supplementary Materials Supplementary Information Description Of Additional Supplementary File Supplementary Data 1 Supplementary data 2 Reporting summary Transparent Peer Review file Source Data Data Availability Statement The proteomics data generated in this study have been deposited in the ProteomeXchange Consortium (http://proteomecentral.proteomexchange.org) via the PRIDE partner repository with the identifier numbers: PXD059833 (Fig. 4E) and PXD071065 (Fig. 5K). All data are included in the article or the Supplementary Information. Source data are provided in this paper.