A myzozoan-specific protein is an essential membrane-anchoring component of the succinate dehydrogenase complex in Toxoplasma parasites

Succinate dehydrogenase (SDH) is a protein complex that functions in the tricarboxylic acid cycle and the electron transport chain of mitochondria. In most eukaryotes, SDH is highly conserved and comprises the following four subunits: SdhA and SdhB form the catalytic core of the complex, while SdhC and SdhD anchor the complex in the membrane. Toxoplasma gondii is an apicomplexan parasite that infects one-third of humans worldwide. The genome of T. gondii encodes homologues of the catalytic subunits SdhA and SdhB, although the physiological role of the SDH complex in the parasite and the identity of the membrane-anchoring subunits are poorly understood. Here, we show that the SDH complex contributes to optimal proliferation and O2 consumption in the disease-causing tachyzoite stage of the T. gondii life cycle. We characterize a small membrane-bound subunit of the SDH complex called mitochondrial protein ookinete developmental defect (MPODD), which is conserved among myzozoans, a phylogenetic grouping that incorporates apicomplexan parasites and their closest free-living relatives. We demonstrate that TgMPODD is essential for SDH activity and plays a key role in attaching the TgSdhA and TgSdhB proteins to the membrane anchor of the complex. Our findings highlight a unique and important feature of mitochondrial energy metabolism in apicomplexan parasites and their relatives.


Introduction
Mitochondrial energy metabolism, which consists of processes like the tricarboxylic acid (TCA) cycle and the electron transport chain (ETC), is conserved throughout eukaryotic evolution and is important for the survival of many organisms [1,2].Apicomplexans are a eukaryotic phylum of intracellular parasites that inflict a major burden on human societies, both through impacting human health and the health of important livestock.Apicomplexans include Plasmodium species, the causative agents of malaria, and Toxoplasma gondii, a ubiquitous parasite of humans and livestock that causes the disease toxoplasmosis [3,4].Apicomplexans belong to a group of single-celled eukaryotes called the myzozoans, which also include chrompodellids and dinoflagellates [5,6].Considerable evidence indicates that the mitochondrial energy metabolism of myzozoans has diverged considerably from well-studied organisms such as animals and yeast [7][8][9].For example, although they contain canonical ETC complexes such as the cytochrome bc 1 and cytochrome c oxidase complexes, many of the proteins that comprise these complexes are unique to myzozoans [8][9][10][11][12].The essential function of these complexes has made the ETC a prime drug target in disease-causing apicomplexans [13][14][15][16].
The mitochondria of many myzozoans, including those of T. gondii and Plasmodium falciparum, harbour a functional TCA cycle.Like with the ETC, some of the enzymes that comprise this pathway have diverged considerably from the equivalent enzymes in animals [7,8,[17][18][19].The TCA cycle is dispensable for the disease-causing erythrocytic stage of the P. falciparum life cycle, probably reflecting the reliance of these parasites on ATP derived from glycolysis [20,21].However, the TCA cycle becomes essential following the transmission of these parasites into the insect stage of the life cycle [20].The importance of the TCA cycle in the disease-causing tachyzoite stage of T. gondii is less clear.Treatment with the aconitase inhibitor sodium fluoroacetate impairs the flux of carbons through the TCA cycle and inhibits parasite proliferation [18].By contrast, depletion of the succinyl CoA synthetase enzyme of the TCA cycle, which catalyses the synthesis of succinate from succinyl-CoA, leads to an apparently mild defect in parasite proliferation [22], with its non-essentiality proposed to result from a γ-aminobutyric acid (GABA) shunt that enables this step of the TCA cycle to be bypassed [18].
Succinate dehydrogenase (SDH), or complex II of the ETC, is a protein complex that participates in both the TCA cycle and the ETC.It catalyses the oxidation of succinate to fumarate in the TCA cycle and transfers the resulting electrons to the oxidized form of coenzyme Q (ubiquinone) in the ETC.In animals, fungi and bacteria, the protein complex typically comprises the following four subunits: SdhA and SdhB form the matrix-localized, catalytic core of the complex, while SdhC and SdhD anchor the complex in the mitochondrial membrane and facilitate electron transfer to coenzyme Q [23].SdhA is a large flavoprotein that catalyses the oxidation of succinate to fumarate in the TCA cycle, transferring electrons to a covalently attached flavin adenine nucleotide (FAD), forming FADH 2 [23,24].Electrons from FADH 2 are transferred via three iron-sulfur clusters coordinated at SdhB to coenzyme Q. SdhB is bound to the transmembrane proteins SdhC and SdhD, with the high-affinity coenzyme Q binding site of the complex comprising residues from each of these proteins [25].The reduced form of coenzyme Q then shuttles the electrons to complex III in the ETC from where they are donated to the terminal oxidase, complex IV, where O 2 is reduced [1].
The catalytic SdhA and SdhB proteins of the SDH complex are highly conserved in eukaryotes and prokaryotes, while the membrane-spanning proteins are more divergent.In plants and trypanosomes, the SdhC and SdhD proteins have diverged considerably from equivalent subunits in other organisms, and extra subunits found in the SDH complex may contribute to the functions of the membrane component of the complex [26][27][28].While the T. gondii genome encodes homologues of the SdhA and SdhB proteins, it lacks clear homologues of the membrane subunits [9,29].This poses the question of what could be anchoring the T. gondii complex in the mitochondrial membrane.A recent 'complexome'-based proteomic analysis of mitochondrial protein complexes from T. gondii parasites identified seven previously uncharacterized proteins that co-purified with the TgSdhA and TgSdhB proteins [12], although none were characterized further.A similar complexome study from P. falciparum demonstrated co-migration of homologues of several of the candidate SDH complex proteins identified in T. gondii with PfSdhA and PfSdhB [10,30].One of the candidate SDH complex proteins was previously identified as a myzozoan-specific, mitochondrial protein in the rodent malaria Plasmodium berghei [31].This protein was found to be essential for parasites to develop into the ookinete form during the mosquito stage of the parasite life cycle and therefore termed 'mitochondrial protein ookinete developmental defect' (MPODD) [31].This mirrors the essentiality of the canonical SDH complex protein SdhA for ookinete development in Plasmodium parasites and the increased importance of mitochondrial energy metabolism in insect stages of these parasites [20,32,33].
In this study, we set out to characterize the importance of the SDH complex for the disease-causing tachyzoite stage of T. gondii parasites and explore its role in mitochondrial physiology.We demonstrate that the SDH complex is important but not essential for parasite proliferation and ETC activity in vitro.We further demonstrate that the T. gondii homologue of MPODD (TgMPODD) is a bona fide member of the membrane component of the SDH complex, important for enabling the attachment of the catalytic SdhA and SdhB subunits to the membrane subunits of the complex, and therefore essential for SDH enzyme activity.Together, our data provide functional insights into a divergent and important mitochondrial protein complex in T. gondii parasites.

TgSdhB is an essential component of the SDH complex in T. gondii
To begin to explore the role of the SDH complex in T. gondii biology, we set out to characterize TgSdhB, the Fe-S protein of the complex.To facilitate its characterization, we introduced a haemagglutinin (HA) epitope tag at the 3′ end of the TgSdhB open reading frame of T. gondii, creating a parasite line termed TgSdhB-HA.SDS-PAGE western blotting and immunofluorescence assays of the HA-tagged protein revealed that, as described previously [12,34], TgSdhB exists as a single species of ~40 kDa (figure 1a) and localizes to the mitochondrion (figure 1b).To test whether TgSdhB exists in a protein complex, we extracted proteins from TgSdhB-HA parasites in the mild detergent digitonin, separated them by BN-PAGE and then performed western blotting.We found that TgSdhB is present in a main complex of ~660 kDa (figure 1c, magenta arrow), similar in size to the TgSdhB-containing complex reported previously [12].Additionally, we observed TgSdhB in two less abundant complexes of ~530 kDa and >720 kDa (figure 1c, green and blue arrows, respectively).We conclude that TgSdhB is a component of protein complexes in the mitochondrion.
Next, we wanted to explore the importance of TgSdhB and the SDH complex generally, in parasite biology.We replaced the native promoter of TgSdhB with an ATc-regulated promoter in the TgSdhB-HA parasite line wherein another candidate SDH protein, TgSdh18 [12], had been tagged with a FLAG epitope (electronic supplementary material, figure S1).We termed the resulting parasite line 'regulatable (r)TgSdhB-HA/TgSdh18-FLAG' (hereafter referred to as 'rTgSdhB').To measure the extent of TgSdhB knockdown upon the addition of ATc, we cultured rTgSdhB parasites for 0-3 days on ATc, extracted parasite proteins, separated them by SDS-PAGE and performed western blotting.This revealed that TgSdhB abundance was reduced in the rTgSdhB line after 2 days on ATc and barely detectable after 3 days (figure 1d).By contrast, addition of ATc to the parental TgSdhB-HA parasite line (hereafter referred to as wild type, WT) did not result in an appreciable change in TgSdhB abundance after 3 days on ATc (figure 1d).
We investigated the importance of TgSdhB for SDH enzyme activity using an absorbance-based enzymatic assay.In comparison to WT parasites, rTgSdhB parasites cultured in the absence of ATc had approximately twofold lower SDH activity (figure 1e), possibly due to differences in the timing of expression of TgSdhB from the ATc-regulatable promotor compared to the native promoter.Notably, we found that SDH activity decreased by 86% upon knockdown of TgSdhB by the addition of ATc for 3 days, while SDH activity in the WT parasite line was unaffected by the addition of ATc (figure 1e).We also measured the activity of malate:quinone oxidoreductase (MQO), an unrelated enzyme that mediates the oxidation of malate in the TCA cycle and contributes electrons to the ETC [1], in rTgSdhB parasites cultured in the absence or presence of ATc.We observed no significant change in MQO activity upon TgSdhB knockdown (figure 1f), implying that the loss of SDH activity associated with TgSdhB depletion is not the result of general impairment in TCA cycle or mitochondrial functions.
We complemented rTgSdhB parasites with Ty1-tagged TgSdhB expressed constitutively from the α-tubulin promoter, generating a line that we termed cTgSdhB/rTgSdhB.We extracted proteins from these parasites, separated them by SDS-PAGE and performed western blotting to demonstrate that the resulting TgSdhB-Ty1 protein was expressed (figure 1d).We cultured cTgSdhB/rTgSdhB in the absence or presence of ATc for 3 days and measured SDH activity.We found that SDH activity in the complemented strain was unchanged upon the addition of ATc, and indistinguishable from SDH activity in WT parasites (figure 1e).Together, these experiments indicate that TgSdhB-HA is critical for SDH activity in T. gondii parasites.

SDH is important but not essential for parasite proliferation and ETC activity in T. gondii
Having established the importance of TgSdhB for SDH activity, we next set out to determine the role and importance of SDH in the biology of the disease-causing tachyzoite stage of T. gondii parasites.We first investigated the contribution of TgSdhB, and by extension the SDH complex, to overall parasite proliferation.We cultured WT and rTgSdhB parasites with or without ATc for 7 days and compared sizes of plaques (zones of clearance in the host cell monolayer created by proliferating parasites).Plaque sizes decreased significantly in rTgSdhB but not in WT parasites cultured in the presence of ATc (figure 2a,b), indicating that rTgSdhB parasites cultured in the presence of ATc exhibit reduced proliferation.Complementation with constitutively expressed TgSdhB (cTgSdhB/rTgSdhB) restored parasite proliferation in the presence of ATc (figure 2a,b).We next compared the importance of SDH for parasite proliferation with other components of the ETC.We performed plaque assays using a parasite line termed rTgQCR11 wherein we could knockdown expression of TgQCR11, a key protein of complex III, by the addition of ATc [11].We found that plaque sizes were considerably smaller upon TgQCR11 knockdown than upon knockdown of TgSdhB (figure 2a,b).Although these data indicate that TgSdhB is less important for proliferation than complex III of the ETC, it is conceivable that the reduced proliferation defect we observe in the rTgSdhB strain upon ATc addition is because of incomplete knockdown of TgSdhB.To further test the essentiality of TgSdhB for parasite proliferation in vitro, we therefore attempted to generate a TgSdhB knockout.We integrated a phleomycin-resistance cassette into the open reading frame of the TgSdhB genomic locus in a Cas9-expressing strain of T. gondii (electronic supplementary material, figure S2a,b).The resulting strain is expected to harbour a functional knockout of the TgSdhB gene, and we therefore termed it ∆sdhB.We validated that ∆sdhB parasites were defective in SDH activity but not MQO activity via an enzymatic assay (electronic supplementary material, figure S2c,d).To test the effects of TgSdhB knockout on parasite proliferation, we undertook a plaque assay.We found that TgSdhB parasites were impaired in proliferation compared to the corresponding parental parasite strain but that plaques were still visible (figure 2c), similar to the moderate defect in proliferation observed in the rTgSdhB line cultured in the presence of ATc.Taken together, these data indicate that although it is not essential, TgSdhB, and by extension, the SDH complex, is important for optimal parasite proliferation in vitro.
Given the role of SDH in the mitochondrial ETC in other eukaryotes, we next explored the importance of TgSdhB and the SDH complex for parasite respiration.We measured the effects of TgSdhB knockdown on the basal mitochondrial O 2 consumption rate (mOCR) in the parasite, using a previously established Seahorse XFe96 Flux analyser assay [9,11,35].In this assay, we incubate parasites in medium containing the carbon sources glucose and glutamine.The resulting metabolism of these carbon substrates by the parasite, including by reactions in the TCA cycle, results in the transfer of electrons to coenzyme Q in the inner mitochondrial membrane (figure 2d).Electrons then transport via complexes III and IV and are ultimately donated to molecular oxygen, the final electron acceptor in the ETC.The rate at which oxygen is consumed by the parasite, therefore, correlates to ETC activity (figure 2d).We cultured WT, rTgSdhB, rTgSdhB/cTgSdhB or rTgQCR11 parasites in the absence of ATc or the presence of ATc for 1-3 days and then measured basal mOCR.We observed that basal mOCR decreased upon the knockdown of TgSdhB, depleting by ~70% 3 days after the addition of ATc (figure 2e).In contrast, the addition of ATc had no effect on mOCR in either WT parasites or the complemented line (cTgSdhB/rTgSdhB) (figure 2e).Notably, knockdown of TgQCR11 resulted in a more severe depletion of basal mOCR than knockdown of TgSdhB (figure 2e), reminiscent of the more severe proliferation defect observed upon TgQCR11 knockdown (figure 2a,b).
Taken together, our data indicate that the SDH complex from T. gondii contributes to parasite proliferation and mOCR but is less important for these processes than complex III of the ETC.

TgMPODD is a myzozoan-specific component of the SDH complex of T. gondii
The genome of T. gondii encodes homologues of the soluble, matrix-localized SdhA and SdhB subunits of SDH, but lacks clear homologues of the membrane subunits SdhC and SdhD.This raises the following question: what is anchoring the SDH complex in the mitochondrial membrane and facilitating electron transfer to ubiquinone?A recent complexome-based proteomic analysis of ETC complexes in T. gondii identified seven putative subunits of SDH that are restricted to apicomplexans and their nearest relatives such as chrompodellids (chromerids and colpodellids) and dinoflagellates, a group of organisms collectively referred to as myzozoans (figure 3) [12,30].Six of these seven novel putative subunits were identified in an analysis of the mitochondrial proteome of T. gondii [9] and five were detected in the mitochondrial membrane fraction in a localization of organelle proteins by isotope tagging (LOPIT) approach (figure 3) [40].Homologues of several of the candidate T. gondii SDH proteins were identified as candidate SDH complex proteins in a complexome analysis of P. falciparum [10,30].Several of the subunits are predicted to have at least one transmembrane domain (TMD) (figure 3), making them candidates for being part of the missing membrane component of the SDH complex.At the outset of our study, none of the identified proteins had been validated as SDH subunits (neither in T. gondii nor in other organisms).However, the homologue of one of the proteins, TgMPODD, was found to localize to the mitochondrion and be essential for transmission of P. berghei into its mosquito host [31].We hypothesized that TgMPODD was a myzozoan-specific subunit of SDH in T. gondii, and set out to characterize it.
To enable the detection of the TgMPODD protein, we integrated a FLAG epitope tag at the 3′ end of the open reading frame of TgMPODD in TgSdhB-HA parasites, creating a line termed TgMPODD-FLAG/TgSdhB-HA (electronic supplementary material, figure S3a,b).TgMPODD-FLAG migrated to ~15 kDa when assessed by SDS-PAGE western blotting (figure 4a) and localized to the mitochondrion in immunofluorescence assays (figure 4b), consistent with the localization of its homologue in P. berghei [31].To establish whether TgMPODD is part of a protein complex, we extracted proteins from TgMPODD-FLAG/ TgSdhB-HA parasites, separated them by BN-PAGE and performed western blotting to detect TgMPODD-FLAG.We observed that TgMPODD-FLAG is part of a main complex of ~660 kDa, and two less abundant complexes at ~530 and >720 kDa (figure 4c, magenta, green and blue arrows, respectively), similar to the masses of the TgSdhB-containing complexes (figure 1c).Additionally, we found that TgMPODD was part of a fourth complex at ~430 kDa (figure 4c, orange arrow).As a direct test of whether TgMPODD is in the same complex as TgSdhB, we solubilized proteins from TgMPODD-FLAG/TgSdhB-HA parasites in 1% (w/v) digitonin and performed co-immunoprecipitation experiments.We found that immunoprecipitation of TgSdhB-HA with anti-HA antibodies partially co-purified TgMPODD-FLAG but not the unrelated mitochondrial protein TgTom40 (figure 4d).Similarly, we found that immunoprecipitation of TgMPODD-FLAG with anti-FLAG antibodies co-purified TgSdhB-HA but not TgTom40 (figure 4d).These data indicate that TgMPODD is a component of the same protein complex(es) as TgSdhB.While immunoprecipitation of TgMPODD-FLAG resulted in co-purification of most of the TgSdhB-HA protein, immunoprecipitation of TgSdhB-HA co-purified proportionally less of the TgMPODD protein (figure 4d).Together with the BN-PAGE data, these experiments are consistent with the hypothesis that TgMPODD and TgSdhB are both components of the major ~660 kDa SDH complex in addition to the two less prominent complexes of ~530 and >720 kDa and that TgMPODD is a component of a fourth ~430 kDa complex from which TgSdhB is absent.
The finding that TgMPODD is a component of the SDH complex raises the possibility that it could contribute to the anchoring of this complex into the mitochondrial inner membrane.Previous bioinformatic analysis of the P. berghei homologue of MPODD predicted the presence of a single TMD in the protein [31], although our analysis of the TgMPODD protein with some TMD prediction algorithms (including CCTOP [39]) does not predict the presence of a TMD figure 3. We therefore set out to experimentally test whether TgMPODD is an integral membrane protein.To do this, we extracted proteins from TgMPODD-FLAG/TgSdhB-HA parasites in either 1% (v/v) Triton X-100 or in alkaline sodium carbonate, an approach that extracts non-membrane and peripheral membrane proteins into the soluble fraction while maintaining integral proteins in the membrane pellet [42].We found that the bulk of both TgMPODD-FLAG .Candidate subunits of the SDH complex in T. gondii parasites.Information on the candidate SDH complex proteins identified in a recent complexome analysis of T. gondii mitochondria [12]; the table is modelled after [11,36].ID: ToxoDB gene identification number [37].Annotation: the name of the listed protein used in this manuscript.PS: phenotype score, a prediction of the importance of the listed gene for parasite proliferation, with fitness conferring genes typically having phenotype scores of <−2 [38].Mass: the predicted molecular mass of the listed protein.TM domains: the predicted number of TMDs in the listed protein, as determined by CCTOP [39].Mito proteome: detection of the listed protein in the mitochondrial proteome of T. gondii [9].LOPIT: predicted subcellular localization of the listed protein in a spatial proteomic analysis of the T. gondii proteome using a localization of organelle proteins by isotope tagging approach [40].MM, mitochondrial membranes; MS, mitochondrial soluble; N/D, not determined.Homology: the presence of homologues of the listed gene in the genomes of Plasmodium falciparum (Pf), Plasmodium berghei (Pb), Vitrella brassicaformis (Vb), Eimeria tenella (Et), Homo sapiens (Hs), Saccharomyces cerevisiae (Sc), Arabidopsis thaliana (At), Trypanosoma cruzi (Tc) and Escherichia coli (Ec).Homology searches were performed using HMMER for all species [41], with the exception of Eimeria tenella, which was searched using BLAST on the www.toxodb.orgwebsite [37].Expected values (E-values) are indicated by coloured circles, with the intensity of red shading increasing in proportion with the −log 10 (E-value).Black circles indicate the absence of a detectable homologue in the target species.and TgSdhB-HA were soluble in Triton X-100, indicating that these proteins are not inherently insoluble (figure 4e).Notably, we found that most of the TgMPODD-FLAG protein was retained in the alkaline sodium carbonate pellet along with the known integral membrane protein TgTom40, whereas both TgSdhB-HA and the known peripheral inner mitochondrial membrane protein TgATPβ were fully extracted into the sodium carbonate supernatant (figure 4e).These data indicate that TgMPODD-FLAG is an integral membrane protein, whereas TgSdhB-HA is not.The existence of a proportion of TgMPODD-FLAG in the sodium carbonate supernatant suggests that the TMD of TgMPODD may be only mildly hydrophobic, as has been observed in proteins from other organisms [43] and as we have observed previously for other inner mitochondrial membrane proteins in T. gondii [44].

TgMPODD is essential for SDH activity in T. gondii
Having established that TgMPODD is a transmembrane subunit of the T. gondii SDH complex, we next wanted to investigate its role and importance in the complex.We replaced the native promoter of the TgMPODD gene with an ATc-regulatable promoter in the TgMPODD-FLAG/TgSdhB-HA line, creating a parasite line termed 'rTgMPODD' (electronic supplementary material, figure S3c,d).To determine the extent of TgMPODD knockdown, we cultured parasites in the absence of ATc or presence of ATc for 1-3 days, separated proteins by SDS-PAGE and measured TgMPODD-FLAG abundance by western blotting.We found that TgMPODD-FLAG was mostly depleted after 2 days on ATc (figure 5a).To test whether TgMPODD is important for SDH activity, we grew rTgMPODD parasites with or without ATc for 3 days and measured SDH activity via enzymatic assays.SDH activity decreased by 96% when TgMPODD was knocked down by the addition of ATc in the rTgMPODD line (figure 5b).In contrast, SDH activity of the parental line (WT, TgMPODD-FLAG/TgSdhB-HA) was not affected by the addition of ATc (figure 5b), and we observed no significant change in MQO activity upon TgMPODD knockdown (figure 5c).These data imply that depletion of TgMPODD leads to a specific loss of SDH activity that mirrors the loss in activity we observed upon TgSdhB depletion (figure 1e).Given that loss of TgSdhB led to defects in ETC activity (figure 2e), it is likely that loss of TgMPODD will lead to a similar depletion in ETC activity, although this is not something we tested.Finally, we tested whether TgMPODD is important for parasite proliferation.Depletion of TgMPODD resulted in a proliferation defect that was rescued by complementing rTgMPODD-FLAG/TgSdhB-HA parasites with a constitutively expressed copy of TgMPODD (figure 5d,e).The significant but partial proliferation defect observed when TgMPODD was depleted resembles the proliferation defect observed when depleting or knocking out TgSdhB (figure 2a-c).We conclude that TgMPODD is crucial for SDH function in T. gondii parasites and that it is important but not essential for parasite proliferation.

2.5.
TgMPODD is important for the attachment of the catalytic subunits of the SDH complex to the integral membrane anchor The SDH complex of T. gondii and related organisms appears to contain numerous novel subunits.To begin to elucidate the architecture of the SDH complex in T. gondii, we aimed to investigate how the loss of the matrix-localized TgSdhB protein affects TgMPODD and other putative subunits of the complex.We introduced c-myc epitope-tags into the genomic locus of TgMPODD, as well as the loci of three additional subunits identified in the complexome analysis of the T. gondii SDH complex [12], TgSdh11, TgSdh15 and TgSdh31, in the rTgSdhB-HA/TgSdh18-FLAG line (electronic supplementary material, figure S4).We validated that the previously uncharacterized candidate SDH proteins all localized to the mitochondrion of T. gondii (electronic supplementary material, figure S5a-d).
We first asked whether the abundances of the candidate SDH complex proteins changed upon depletion of TgSdhB.We cultured parasites in the absence or presence of ATc for 3 days, separated proteins by SDS-PAGE, and measured protein abundances by western blotting.TgMPODD-c-myc, TgSdh11-c-myc, TgSdh15-c-myc, TgSdh18-FLAG and TgSdh31-c-myc proteins were observed at approximately their predicted masses (figure 6a; electronic supplementary material, figure S6a).TgSdh15-c-myc was also observed in a second band of ~32 kDa, roughly double its expected mass, which could represent a dimer (figure 6a).Abundances of the TgMPODD-c-myc, TgSdh11-c-myc, TgSdh18-FLAG and TgSdh31-c-myc proteins did not change appreciably upon the knockdown of TgSdhB, while abundance of the lower mass band of TgSdh15-c-myc appeared to decrease (figure 6a; electronic supplementary material, figure S6a).
We next performed BN-PAGE on the tagged parasite strains cultured in the absence of ATc to test whether TgSdh11-c-myc, TgSdh15-c-myc, TgSdh18-FLAG and TgSdh31-c-myc existed in protein complexes.We found that, as observed previously (figure 4c), TgMPODD-c-myc was present in a major complex of ~660 kDa (figure 6b, magenta arrow), in addition to two smaller complexes of ~530 and ~430 kDa (figure 6b).TgSdh11-c-myc, TgSdh15-c-myc and TgSdh18-FLAG were present in complexes of approximately the same masses as TgMPODD-c-myc (figure 6b and electronic supplementary material, figure S6b).TgSdh31-c-myc was present in the ~660 and ~530 kDa complexes, but less clearly in the smaller ~430 kDa complex observed with the other proteins (figure 6b).Strikingly, when TgSdhB was knocked down by the addition of ATc for 3 days, the ~660 kDa complex observed in the −ATc condition for all five subunits was absent.Instead, each protein migrated in a complex at ~430 kDa (figure 6b, orange arrow; electronic supplementary material, figure S6b) and also in a less abundant complex of >720 kDa (figure 6b).Taken together with our previous BN-PAGE analysis of TgSdhB and TgMPODD (figures 1c and 4c), these data imply that TgSdh11-c-myc, TgSdh15-c-myc, TgSdh18-FLAG and TgSdh31-c-myc are likely components of the ~660 kDa SDH complex.In addition, TgSdh11-c-myc, TgSdh15-c-myc and TgSdh18-FLAG are, together with TgMPODD, all part of a smaller ~430 kDa complex from which TgSdhB is absent.Given that TgMPODD, TgSdh11 and TgSdh15 are integral membrane proteins royalsocietypublishing.org/journal/rsob Open Biol.14: 230463 (figure 4e) [45] and that TgSdh18 is predicted to contain TMDs (figure 3), we hypothesize that this smaller complex represents the membrane-anchoring domain of the SDH complex in these parasites.
To further explore SDH complex architecture, we investigated the importance of TgMPODD for complex integrity.We cultured rTgMPODD-FLAG/TgSdhB-HA parasites in the absence or presence of ATc for 3 days and measured protein abundance by SDS-PAGE western blotting.We found that the abundance of TgSdhB-HA does not change upon TgMPODD depletion (figure 7a).We then performed BN-PAGE western blotting on proteins extracted from rTgMPODD-FLAG/TgSdhB-HA parasites cultured for 0-3 days on ATc.As observed previously, we found that in the absence of ATc, both TgMPODD-FLAG and TgSdhB-HA were present in a major complex of ~660 kDa, in addition to less abundant complexes at ~530 and >720 kDa (figure 7b, magenta, green and blue arrows, respectively).TgMPODD-FLAG was additionally present in the ~430 kDa complex (figure 7b, orange arrow).Upon TgMPODD-FLAG knockdown, we observed disappearance of the 430 kDa complex after 1 day on ATc and depletion of the other TgMPODD-FLAG-containing complexes to undetectable levels after 3 days (figure 7b).Notably, the major 660 kDa TgSdhB-HA-containing complex depleted concomitantly with depletion of the equivalent 660 kDa TgMPODD-FLAG complex, with TgSdhB-HA appearing in a new, smaller complex of ~150 kDa upon TgMPODD-FLAG knockdown (figure 7b, purple arrow).These data indicate that depletion of TgMPODD results in the loss of TgSdhB from the SDH complex.
Having established that loss of TgMPODD affects SDH complex integrity and TgSdhB integration into this complex, we investigated what happens to the TgSdh11-, TgSdh15-, TgSdh18-and TgSdh31-containing complexes upon TgMPODD knockdown.We introduced c-myc tags into the loci of TgSdh11, TgSdh15, TgSdh18 or TgSdh31 in the rTgMPODD-FLAG/ TgSdhB-HA line (electronic supplementary material, figure S7a-h).We cultured parasites in the absence or presence of ATc for 3 days and then performed BN-PAGE western blotting on extracted proteins.As expected, all four c-myc-tagged proteins localized predominantly to the ~660 kDa complex in the −ATc condition, when parasites expressed TgMPODD-FLAG (figure 7c, magenta arrow).Upon TgMPODD-FLAG knockdown by the addition of ATc, this 660 kDa complex was depleted in all lines (figure 7c).In the TgMPODD-FLAG knockdown condition, both TgSdh15-c-myc and TgSdh18-c-myc migrated in a major complex of ~500 kDa and a less abundant complex of ~400 kDa (figure 7c).TgSdh31-c-myc migrated in a major complex of ~400 kDa, whereas the TgSdh11-c-myc protein appeared to be depleted from all protein complexes (figure 7c).These data indicate that knockdown of TgMPODD leads to defects in the integrity of the major ~660 kDa SDH complex, with the concomitant formation of smaller complexes containing some of the other SDH subunits.In SDH complexes from other organisms, the SdhB protein links the matrix-localized, catalytic SdhA flavoprotein to the membrane anchoring SdhC and SdhD proteins [46].Our finding that depletion of TgMPODD resulted in a detachment of the TgSdhB protein from other components of the SDH complex prompted us to examine the effects of TgMPODD depletion on TgSdhA.To test this, we introduced a c-myc-tag into the TgSdhA locus in the rTgMPODD-FLAG/TgSdhB-HA parasite line (electronic supplementary material, figure S7i,j) and demonstrated that TgSdhA-c-myc localized to the mitochondrion (electronic supplementary material, figure S5e).We cultured the resultant line in the absence or presence of ATc for 3 days, then separated protein complexes via BN-PAGE and probed for TgSdhA-c-myc.When TgMPODD was present, TgSdhA was detected in a major complex of ~660 kDa (figure 7c), consistent with this complex being the catalytically active SDH complex.Upon TgMPODD knockdown, the TgSdhA-c-myc was not detected at any higher molecular mass complex (figure 7c), suggesting that it is entirely lost from the SDH complex.
Taken together, these data indicate that the loss of TgMPODD causes major defects in the integrity of the SDH complex, with the matrix-localized catalytic TgSdhA and TgSdhB subunits dissociating from the complex entirely, and many of the other components (including TgSdh15, TgSdh18 and TgSdh31) forming smaller subcomplexes.

Discussion
In this study, we explored the physiological importance of the SDH complex in the disease-causing tachyzoite stage of the apicomplexan parasite T. gondii.We found that loss of TgSdhB, the catalytically essential Fe-S component of the complex, resulted in reduced mOCR and parasite proliferation.A parallel study by Silva et al. reached similar conclusions about other subunits of the T. gondii SDH complex [45].Although substantial, the observed defects in mOCR and proliferation upon TgSdhB knockdown were milder than defects observed upon knockdown of TgQCR11, an essential component of ETC complex III [11].These data suggest that the parasite ETC remains functional in the absence of the SDH complex, with electrons likely donated to coenzyme Q via other inner membrane dehydrogenases that are found in the parasite mitochondrion (figure 2d) [1,47,48].In addition to its role in the ETC, SDH catalyses a key reaction in the TCA cycle.Our data therefore imply that the TCA royalsocietypublishing.org/journal/rsob Open Biol.14: 230463 cycle is important but not absolutely essential for parasite proliferation, at least in the nutrient rich conditions in which we culture parasites in vitro.These data are somewhat surprising, given that treatment of parasites with sodium fluoroacetate, an inhibitor of the TCA cycle enzyme aconitase, fully inhibits tachyzoite proliferation [18,49].A previous study of the succinyl-CoA synthetase enzyme in T. gondii, which catalyses the synthesis of succinate in the TCA cycle, found that knockdown of the ScsA protein from this enzyme complex led to only a mild proliferation defect [22].The dispensability of succinyl-CoA synthetase was proposed to result from the existence of a GABA shunt, which bypasses the α-ketoglutarate dehydrogenase-and succinyl-CoA synthetase-catalysed reactions of the TCA cycle [18].The SDH-catalysed reaction, however, occurs downstream of the proposed GABA shunt, suggesting that its dispensability cannot be explained in this manner.Future experiments that assess changes to parasite metabolism associated with SDH loss and a broader characterization of the essentiality of TCA cycle enzymes will be key to resolving the role and importance of the TCA cycle in T. gondii tachyzoites.
Our findings on the physiological role of the SDH complex fit with a broader narrative that T. gondii tachyzoites exhibit considerable metabolic flexibility that may contribute to their ability to infect a broad range of host organisms and tissues [50,51].Future studies that examine the effects of SDH and TCA cycle impairment on the ability of parasites to cause disease in whole animal infection models will be crucial for understanding the role of mitochondrial energy metabolism in this proposed flexibility.The SDH complex, and in particular, the membrane anchor of the complex, is a major target of pesticides that target fungal pathogens in plants [52].Given the novelty in protein composition of the SDH complex in apicomplexans, particularly in the membrane anchoring subunits, the SDH complex is an attractive target for inhibitors.Elucidating the contribution of the  SDH complex for parasite virulence is therefore also of high importance for determining whether the SDH complex could be developed as a drug target in these parasites.
A bevy of recent studies has established that the mitochondrial ETC complexes of myzozoans, including T. gondii and other apicomplexans, contain numerous novel and divergent subunits compared to the equivalent protein complexes in well-studied organisms such as fungi and animals [9][10][11][12]44,45].Given that the common ancestor of myzozoans and opisthokonts (fungi and animals) probably existed at the dawn of eukaryotic evolution [53], this suggests that considerable novelty in ETC protein complexes arose over the course of myzozoan evolution.A major focus is now on elucidating the functional roles of the novel proteins in these complexes [1,44].Building on two recent 'complexome' studies [10,12], our data establish that MPODD is a bona fide component of the SDH complex of T. gondii.A previous study established that MPODD is a myzozoan-specific protein that is essential for development of ookinetes in the early insect stages of Plasmodium parasite development [31].Other studies have established that the SDH complex and other proteins of the TCA cycle are important for early insect stage development in Plasmodium parasites, possibly because of an increased reliance on oxidative phosphorylation or products of TCA cycle metabolism in these stages [20,32].Our findings that MPODD is essential for SDH function in these parasites provide a rationale for its importance in ookinete development.
When solubilized in the mild detergent digitonin, the SDH complex exists primarily in a ~660 kDa complex.We propose that this complex represents the functional unit of the SDH complex in parasites, since it contains the matrix-localized SdhA and SdhB that are essential for the catalytic activity of the complex (figure 8a).Although the stoichiometry of proteins within the complex remains unclear, the sum of the SDH subunits adds to ~230 kDa (figure 3).This suggests that the 660 kDa complex may represent a trimer, as has been proposed for the equivalent complex in P. falciparum parasites [10].We also observe a large complex of >720 kDa that could either represent a larger order arrangement of SDH complexes or an ETC supercomplex, an arrangement of multiple ETC complexes in a single larger complex that has been observed in many other eukaryotes [54] (figure 8a).Although the existence of SDH in such supercomplexes is not as common [55], proteomic analysis of a putative ~745 kDa SDH complex in the related parasite Eimeria tenella identified complex III and IV proteins [56], and the SDH complex of the ciliate Tetrahymena (a sister taxon to the myzozoans) exists in a massive ETC supercomplex [57].Curiously, our BN-PAGE experiments on numerous SDH subunits revealed the presence of subcomplexes of ~530 and ~430 kDa.These smaller complexes contained membrane-bound proteins such as TgMPODD, TgSdh15 and TgSdh18 but lacked the soluble catalytic TgSdhA in the 530 kDa complex and the TgSdhA and TgSdhB subunits in the 430 kDa complex (figure 8a).Following TgSdhB depletion, a ~430 kDa complex containing the proposed membrane-bound subunits remains (figure 8b), and we propose that this complex represents the membrane anchor to which TgSdhB and TgSdhA attach to form the catalytically active SDH complex, although it is also conceivable that this represents an assembly intermediate.Interestingly, the 430 kDa complex was not observed in the Silva et al. study, which used n-dodecyl-β-d-maltoside (DDM) to solubilize proteins for BN-PAGE analysis [45].DDM is a 'harsher' detergent than the digitonin we used in our study, which may reflect that the proposed 430 kDa membrane anchor is particularly labile.Future structural studies of the various SDH complexes in the parasite may provide insights into the nature and function of these subcomplexes.
Our experiments reveal that TgMPODD is an integral membrane protein that is essential for SDH complex function.What then is the role of TgMPODD in the complex?Strikingly, depletion of MPODD results in loss of the catalytic TgSdhB and TgSdhA subunits from the SDH complex and the formation of smaller complexes of ~400 and ~500 kDa that contain many of the proposed membrane anchoring subunits (figure 8c).We propose that TgMPODD plays a crucial role in recruiting and/or tethering the TgSdhB and TgSdhA subunits to the membrane anchoring proteins of the complex.Given that the channelling of electrons from the Fe-S clusters of TgSdhB to ubiquinone at the membrane anchor is critical for SDH function, this model explains the essentiality of TgMPODD for SDH activity.Future structural studies that elaborate on the position of TgMPODD in the SDH complex of apicomplexans and related organisms may provide insights into how this recruitment or tethering is mediated.
Our study explored the role and architecture of the SDH complex in T. gondii, demonstrating that TgMPODD is an essential, membrane-anchored subunit of the SDH complex, with an important role in maintaining complex integrity and tethering the matrix-localized SdhA and SdhB to the membrane anchor.Our data therefore provide insights into the previously unknown function of this protein.Future studies should examine whether the functions of MPODD holds true in other myzozoans, and further elucidate the structural architecture of the SDH complex in these organisms.

Genetic modifications of T. gondii
Most of the genetically modified parasite lines described in the study were generated in the TATi∆ku80 strain of T. gondii [58] and were cloned before subsequent characterization.All transfections and drug selections were performed as described previously [59].
We introduced a 3′ haemagglutinin (HA) into the TgSdhB locus of the TATi∆ku80 strain using a selectable marker-based 3′ replacement strategy described previously [34], generating a strain we termed TgSdhB-HA.We next introduced FLAG tags into the TgSdh18 or TgMPODD loci of this strain using a selection marker-less CRISPR/Cas9 genome editing approach described previously [9].To do this, we generated a vector expressing a single guide (sg)RNA targeting the region around the predicted stop codon of TgSdh18 or TgMPODD by modifying the pSAG1::Cas9-U6::sgUPRT vector (Addgene plasmid, 54467 [60]) using Q5 mutagenesis with the TgSdh18 3′ CRISPR fwd or TgMPODD 3′ CRISPR fwd primers together with the universal CRISPR rvs primer (electronic supplementary material, table S1) as per the manufacturer's instructions.We then PCR amplified a FLAG tag with 50 bp homology arms to the TgSdh18 or TgMPODD loci at either end using the primers TgSdh18 tag fwd and rvs or TgMPODD tag fwd and rvs and a FLAG gBlock (Integrated DNA Technologies, (IDT)) as template (electronic supplementary material, table S1).We co-transfected the resulted PCR product and sgRNA-expressing plasmid into TgSdhB-HA parasites, selected and cloned GFP-positive parasites (which express the Cas9-GFP encoded on the sgRNA-expressing vector) by fluorescence-activated cell sorting (FACS) using a FACSMelody cell sorter (BD Biosciences) 2 or 3 days after transfection, then screened for parasites that had incorporated the FLAG tag into the TgSdh18 or TgMPODD loci using the TgSdh18 or TgMPODD 3′ scrn fwd and rvs primers (electronic supplementary material, table S1).This generated strains we termed TgSdhB-HA/TgSdh18-FLAG or TgSdhB-HA/TgMPODD-FLAG.
Next, we introduced an ATc-regulatable promoter upstream of the TgSdhB or TgMPODD start codons of the TgSdhB-HA/TgSdh18-FLAG or TgSdhB-HA/TgMPODD-FLAG lines, respectively, using a selection marker-less CRISPR/Cas9 genome editing-based promoter insertion strategy described previously [9].We performed Q5 mutagenesis to modify the The functional T. gondii SDH complex comprises a ~660 kDa complex that contains all proposed subunits of the complex and may exist as a trimer (magenta arrow).The complex also exists in a higher-order structure of >720 kDa (blue arrow).Smaller complexes observed represent the membrane anchor lacking TgSdhA and with the TgSdhB protein tethered (~530 kDa complex; green arrow) or not-tethered (~430 kDa complex; orange arrow).(b) Depletion of TgSdhB leads to the formation of a ~430 kDa complex that contains the membrane-bound subunits of the complex.We propose that this represents the membrane anchor of the SDH complex in T. gondii.(c) Depletion of TgMPODD leads to the loss of TgSdhA and TgSdhB from the complex (grey and light brown arrows), with remnant complexes of ~400 and ~500 kDa containing some of the membrane anchoring subunits of the complex (dark brown arrows).These data suggest a role for TgMPODD in tethering the matrix components TgSdhB and TgSdhA to the membrane anchoring subcomplex.BN-PAGE data in the figure are derived from experiments depicted in figures 1, 3, 5 and 6.
pSAG1::Cas9-U6::sgUPRT vector to express sgRNAs targeting the 5′ region of the TgSdhB or TgMPODD genes close to the start codon using the TgSdhB 5′ CRISPR fwd or TgMPODD 5′ CRISPR fwd primer together with the universal CRISPR rvs primer (electronic supplementary material, table S1).We also amplified the ATc-regulatable teto7/sag4 promoter [61] and an associated spacer region along with 50 bp homology arms targeting the TgSdhB or TgMPODD loci using the primers SdhB prorep (promoter replacement) fwd and rvs or MPODD prorep fwd and rvs (electronic supplementary material, table S1) and the pPR2-HA 3 vector [62] as template.We co-transfected the PCR products and sgRNA-expressing plasmid into TgSdhB-HA/TgSdh18-FLAG or TgSdhB-HA/TgMPODD-FLAG parasites, sorted and cloned parasites 2 or 3 days after transfection as described above and screened clones for integration of the ATc-regulatable promoter using the SdhB 5′ scrn fwd and rvs or MPODD 5′ scrn fwd and rvs primers (electronic supplementary material, table S1).We referred to the resulting ATc-regulatable (r) TgSdhB-HA/TgSdh18-FLAG line as 'rTgSdhB' throughout the manuscript and the resulting rTgMPODD-FLAG/TgSdhB-HA line as 'rTgMPODD'.
To complement the rTgSdhB and rTgMPODD lines with constitutively expressed, Ty1-tagged TgSdhB and TgMPODD, respectively, we PCR amplified the open reading frames of TgSdhB or TgMPODD using the primers SdhB comp fwd or MPODD comp fwd together with Ty1 universal rvs (electronic supplementary material, table S1), and using gBlocks encoding the entire open reading frames of either TgSdhB or TgMPODD fused to a 3 × Ty1 epitope tag as templates (electronic supplementary material, table S1; IDT).We digested the resulting PCR products with BglII and XmaI and ligated these into the equivalent sites of a vector termed TgQCR11 in pUDT-Ty1 [11].The resultant vector expresses the TgSdhB-Ty1 or TgMPODD-Ty1 transgenes from the constitutive α-tubulin promoter and contains a pyrimethamine-resistant TgDHFR selectable marker and a UPRT flanking sequence for integration into the non-essential uracil phosphoribosyltransferase (TgUPRT) locus of the parasite.The resultant TgSdhB-Ty1 or TgMPODD-Ty1 expressing vectors were linearized in their UPRT flank with MfeI, transfected into rTgSdhB or rTgMPODD lines, respectively, selected on pyrimethamine and then cloned by limiting dilution.Although we anticipate that the vector should integrate into the UPRT locus of the parasite, we did not check for this.
To introduce c-myc tags into the TgMPODD, TgSdh11, TgSdh15, TgSdh18, TgSdh31 or TgSdhA loci of the rTgSdhB and TgMPODD lines, we used the same selection marker-less CRISPR/Cas9 genome editing approach described above.Briefly, we generated vectors expressing sgRNAs targeting near the predicted stop codons of the target genes using Q5 mutagenesis with the gene-specific '3′ CRISPR fwd' primer and the universal CRISPR rvs primer (electronic supplementary material, table S1).We also amplified a 3 × c-myc epitope tag containing 50 bp homology arms to the target gene with the gene-specific 'tag fwd' and 'tag rvs' primers together with a c-myc gBlock (IDT) as template (electronic supplementary material, table S1).We co-transfected the sgRNA-expressing vector and 3 × c-myc homology template into rTgSdhB or rTgMPODD parasites, selected and cloned Cas9-GFP expressing parasite by flow cytometry 2 or 3 days after transfection and then screened for successful integration of the c-myc tag using the gene-specific '3′ scrn fwd' and '3′ scrn rvs' primers (electronic supplementary material, table S1).
To generate a parasite strain in which the TgSdhB open reading frame was disrupted (i.e.where the TgSdhB was functionally knocked out), we first modified the pSAG1::Cas9-U6::sgUPRT vector by Q5 mutagenesis to express a sgRNA targeting the open reading frame in the second exon of the TgSdhB locus.We performed the Q5 mutagenesis using the SdhB KO CRISPR fwd and universal CRISPR rvs primers (electronic supplementary material, table S1), as per the manufacturer's instructions.We also PCR amplified a phleomycin selectable marker (BLE) containing 50 bp homology arms to the TgSdhB locus on either side of the sgRNA-targeting site with the primers SdhB KO fwd and rvs (electronic supplementary material, table S1) and the vector pUBTTy as template [63].We co-transfected the sgRNA-expressing vector and BLE-containing homology template into the Cas9-CAT parasite line (a kind gift from Clare Harding, U. Glasgow [64]), selected and cloned Cas9-GFP-expressing parasite by flow cytometry 2 days after transfection and then screened for successful integration of DNA into the TgSdhB open reading frame using the SdhB KO scrn fwd and rvs primers (electronic supplementary material, table S1).

Plaque assays
Plaque assays were performed as described previously [44], with 500 parasites added per flask.Flasks were incubated for 7 days before staining with crystal violet (Con and Hucker's formula, Fronine Laboratory Supplies or Gram's Crystal Violet solution, Sigma).Flasks were left to dry and scanned using a CanoScan 9000F scanner (Canon).Plaque areas were determined using the freehand area selection and measure tools in ImageJ (v.1.53 k), with plaques that had merged with neighbouring plaques excluded from the analysis.Statistical comparisons of the mean plaque areas for each parasite line and condition was performed using a one-way ANOVA with Tukey's multiple comparisons test in GraphPad Prism (v.10).

Sodium carbonate extraction assays
To test whether proteins were integral membrane proteins, we solubilized parasite fractions in alkaline sodium carbonate (Na 2 CO 3 , pH 12), as described previously [34,66].Briefly, egressed parasites were resuspended in phosphate-buffered saline (PBS) and split into three equal volumes and pelleted at 12 000 × g for 1 min.One pellet was resuspended in 1× reducing LDS sample buffer to yield the 'Total' fraction.The second pellet was resuspended in TX-100 lysis buffer (1% v/v Triton X-100, 150 mM NaCl, 2 mM EDTA and 50 mM Tris (pH 7.4), with added complete protease inhibitors (Sigma)), incubated on ice for 30 min, then centrifuged at 21 000 × g for 20 min at 4°C.The resulting pellet, containing TX-100 insoluble protein, was resuspended in 1× reducing LDS sample buffer.The TX-100 soluble proteins in the supernatant were precipitated via trichloroacetic acid precipitation, as described previously [34], before being resuspended in 1× reducing LDS sample buffer.The third parasite fraction was resuspended in 100 mM of alkaline Na 2 CO 3 (pH 12), incubated on ice for >60 min and then ultracentrifuged at 189 000 × g for 30 min at 4°C.The resulting pellet, containing integral membrane proteins, was resuspended in 1× reducing LDS sample buffer.The Na 2 CO 3 supernatant, containing soluble proteins and peripheral membrane proteins, was TCA precipitated and resuspended in 1× reducing LDS sample buffer.Samples were separated by SDS-PAGE and proteins were detected by western blotting, as described above.

Complex II and MQO enzymatic activity assays
Enzymatic activity assays were adapted from Spinazzi et al. [67].In these assays, parasite extracts were incubated in assay buffer containing the substrate succinate (to measure complex II activity) or malate (to measure MQO activity) and the redox dye 2,6-dichlorophenolindophenol (DCPIP).When oxidized, DCPIP is blue with a maximal absorbance at 600 nm; when reduced, DCPIP is colourless.The reaction is started by adding the ubiquinone analogue decylubiquinone (DUB).The parasite complex II or MQO, if functional, will oxidize succinate/malate and reduce DUB to DUBH 2 .DUBH 2 then reduces the DCPIP, rendering it colourless.Enzyme activity will, therefore, result in a decrease in absorbance of the sample at 600 nm over time.
Enzyme assay buffer containing 20 mM succinate or malate, 25 mM KH 2 PO 4 , 0.002175% w/v DCPIP, 1 µM atovaquone and 2.5 mg ml −1 fatty-acid-free BSA was prepared and aliquoted into the wells of a 24-well plate.A baseline reading was taken by measuring the absorbance at 600 nm every 30 s for 2 min using a TECAN Infinite 200 PRO plate reader warmed to 37°C.Parasite lysate (an equivalent of 6.25 × 10 7 parasites per ml) was then added to the wells and a further baseline reading was taken every 30 s for 15 min.To start the reaction, DUB was added to a final concentration of 5 µM, and DCPIP absorbance at 600 nm was measured every 30 s for 45 min.
To calculate enzymatic activity, absorbance was plotted as a function of time.The initial rate was estimated from the first 5 min after adding DUB (when the reaction was in linear phase) and divided by the extinction coefficient for DCPIP (19.1 mM cm −1 ) according to the Beer-Lambert law.For each condition, the background (initial rate in the absence of parasite lysate) was subtracted from the observed value to yield the calculated activity.As a statistical test for differences between conditions, we performed ANOVAs followed by multiple pairwise comparison testing or unpaired, two-tailed t-tests.Statistical analyses of the data were carried out in RStudio and GraphPad Prism.Graphs were built using GraphPad Prism.

Seahorse XFe96 extracellular flux analysis
Experiments measuring the mitochondrial oxygen consumption rate (mOCR) of intact extracellular parasites using a Seahorse XFe96 extracellular flux analyser were conducted as described previously [35].Data from the Seahorse flux analysis were exported from the Seahorse Wave Desktop software (Agilent Technologies).A linear mixed-effects model was applied to the data as described previously [35], setting the error between plates (between experiments) and wells (within experiments) as random effects, and the mOCR values between cell lines and days on drug (ATc) as fixed effects.Analysis of the least square means of the values was performed in the R software environment.Statistical differences between these values were tested through ANOVA (linear mixed effects), with a post hoc Tukey test.

Figure 1 .
Figure 1.TgSdhB is essential for SDH activity.(a) Western blot of proteins extracted from TgSdhB-HA parasites separated by SDS-PAGE and detected with anti-HA antibodies.(b) Immunofluorescence assay of an eight-cell vacuole of TgSdhB-HA parasites, probed with anti-HA antibodies to detect TgSdhB-HA (green) and anti-TgTom40 antibodies to detect the mitochondrion (red), with the overlap shown in the merged image.DIC, differential interference contrast transmission image.Scale bar is 5 µm.(c) Western blot of proteins extracted from TgSdhB-HA parasites in 1% (w/v) digitonin, separated by BN-PAGE and detected with anti-HA antibodies.Complex masses indicated with different coloured arrows were determined using relative migration distance (Rf) plots.(d) Western blot of proteins extracted from rTgSdhB-HA parasites, WT parasites (TgSdhB-HA) and the complemented line (cTgSdhB-Ty1/rTgSdhB-HA), cultured in the absence of ATc or the presence of ATc for the times indicated, separated by SDS-PAGE and detected using anti-HA (top), anti-Ty1 (middle) and anti-TgTom40 (bottom) antibodies.(e) SDH enzyme activity assays performed with WT parasites (grey), rTgSdhB parasites (orange) and cTgSdhB-Ty1/rTgSdhB-HA parasites (blue) cultured in the absence or presence of ATc for 3 days.Column graphs show the mean SDH activity obtained from five independent experiments (individual points shown), with error bars representing standard deviation (s.d.).ANOVA followed by Tukey's multiple pairwise comparisons test was performed, with relevant p-values shown.(f) Malate:quinone oxidoreductase (MQO) enzyme activity assays performed with rTgSdhB parasites cultured in the absence or presence of ATc for 3 days.Column graphs show the mean MQO activity obtained from five independent experiments (individual points shown) taken from the same parasite extracts as (e), with error bars representing s.d.An unpaired two-tailed t-test was performed comparing the minus and plus ATc conditions, with the p-value shown.

Figure 2 .
Figure 2. SDH is important for T. gondii proliferation and mitochondrial oxygen consumption.(a) Plaque assays with WT parasites (the TgSdhB-HA parental line), rTgSdhB-HA parasites, complemented cTgSdhB-Ty1/rTgSdhB-HA parasites and regulatable complex III mutant parasites (rTgQCR11-FLAG) cultured in the absence (top) or presence (bottom) of ATc for 7 days.Images are from a single experiment and are representative of three independent experiments.(b) Quantification of plaque sizes from the experiment shown in (a).Black lines depict the median plaque size (in arbitrary units of area) for each parasite line and condition.A one-way ANOVA followed by Tukey's multiple pairwise comparisons test was performed, with relevant p values shown.(c) Plaque assays comparing proliferation of ∆sdhB parasites to the corresponding WT line (Cas9-expressing RH strain parasites).Images are from a single experiment and are representative of two independent experiments.(d) Simplified schematic of the ETC of T. gondii parasites, showing the reactions relevant for this study.Electrons from the catabolism of carbon substrates by a range of mitochondrial dehydrogenases (including succinate dehydrogenase (SDH) and malate:quinone oxidoreductase (MQO)) results in the transfer of electrons (e − ) to the mitochondrial inner membrane (IMM) electron carrier coenzyme Q (Q).Electrons are then donated via complex III and cytochrome c (CytC) to ultimately reduce O 2 at complex IV.IMS, intermembrane space; OAA, oxaloacetate.(e) Basal mitochondrial oxygen consumption rate (mOCR) of WT (grey), rTgSdhB-HA (orange), cTgSdhB-Ty1/rTgSdhB-HA (blue) and rTgQCR11-FLAG (green) parasites cultured in the absence of ATc or presence of ATc for 1-3 days.A linear mixed-effects model was fitted to the data, with values depicting the least squares mean ± 95% confidence interval (CI) from three independent experiments.ANOVA followed by Tukey's multiple pairwise comparisons test was performed, with relevant p-values shown.

Figure 4 .
Figure 4.The myzozoan-specific protein TgMPODD is an integral membrane component of the T. gondii SDH complex.(a) Western blots of proteins extracted from TgMPODD-FLAG/TgSdhB-HA parasites, separated by SDS-PAGE and probed using anti-FLAG antibodies.(b) Immunofluorescence assay of an eight-cell vacuole of TgMPODD-FLAG/TgSdhB-HA parasites, probed with anti-FLAG antibodies to detect TgMPODD-FLAG (green) and anti-TgTom40 antibodies to detect the mitochondrion (red), with the overlap shown in the merged image.DIC, differential interference contrast transmission image.Scale bar is 2 µm.(c) Western blot of proteins extracted from TgMPODD-FLAG/TgSdhB-HA parasites in 1% (w/v) digitonin, separated by BN-PAGE, and detected with anti-FLAG antibodies.Masses of the observed complexes indicated with different coloured arrows next to bands were determined using relative migration distance (Rf) plots, taking the means of two independent experiments.(d) Immunoprecipitation (IP) of proteins extracted from TgMPODD-FLAG/TgSdhB-HA parasites in 1% (w/v) digitonin and incubated with either anti-FLAG or anti-HA agarose beads.The total fraction (T) was obtained from parasite lysates prior to immunoprecipitation.The unbound fractions (UB) contain proteins that were not retained on the agarose beads, and the bound fractions (B) contain proteins that were retained on the agarose beads.Fractions were separated by SDS-PAGE and probed with anti-HA antibodies to detect TgSdhB-HA (top), anti-FLAG antibodies to detect TgMPODD-FLAG (middle) or anti-TgTom40 antibodies (bottom) as a control for proteins that are not part of the SDH complex.Western blots are from a single experiment and are representative of two independent experiments.(e) TX-100 and alkaline sodium carbonate (Na 2 CO 3 ) extractions of TgMPODD-FLAG/TgSdhB-HA parasites.Total fractions (T) were from proteins harvested before solubilization, and pellet (P) and soluble (S) fractions were harvested after each extraction.Proteins were separated by SDS-PAGE and probed with antibodies to detect TgMPODD-FLAG (top), TgSdhB-HA (second from top), the β subunit of ATP synthase, a known peripheral mitochondrial membrane protein (TgAtpB; third from top) and TgTom40, a known integral membrane mitochondrial protein (bottom).Western blots are from a single experiment and are representative of two independent experiments.

Figure 5 .
Figure 5. TgMPODD is critical for SDH activity and important for T. gondii proliferation.(a) Western blot of proteins extracted from rTgMPODD-FLAG/TgSdhB-HA parasites grown in the absence of ATc, or in the presence of ATc for 1-3 days, separated by SDS-PAGE and probed with anti-FLAG antibodies (to detect TgMPODD-FLAG; top) and anti-TgTom40 antibodies (loading control; bottom).(b) SDH enzyme activity in WT parasites (grey, TgMPODD-FLAG/TgSdhB-HA parental line) and rTgMPODD parasites (orange), cultured for 0 or 3 days in ATc.Columns represent mean ± s.d. of three independent experiments (individual points shown).ANOVA followed by Tukey's multiple pairwise comparisons test was performed, with relevant p-values shown.(c) MQO enzyme activity in rTgMPODD parasites cultured for 0 or 3 days in ATc.Columns represent mean ± s.d. of three independent experiments (individual points shown).An unpaired two-tailed t-test was performed comparing the two conditions, with the p-value shown.(d) Plaque assays with WT parasites (parental line, TgMPODD-FLAG/TgSdhB-HA), rTgMPODD parasites, and complemented cTgMPODD/rTgMPODD parasites, cultured in the absence (top) or presence (bottom) of ATc for 7 days.Images are from a single experiment and representative of three independent experiments.(e) Quantification of plaque sizes from the experiment shown in (d).Black lines depict the median plaque size (in arbitrary units of area) for each parasite line and condition.A one-way ANOVA followed by Tukey's multiple pairwise comparisons test was performed, with relevant p-values shown.

Figure 6 .
Figure 6.Depletion of TgSdhB enhances the formation of a ~430 kDa subcomplex containing the candidate membrane-anchoring subunits of SDH.(a,b) Western blots of proteins extracted from TgMPODD-c-myc, TgSdh11-c-myc, TgSdh15-c-myc or TgSdh31-c-myc expressing parasites, all tagged in an rTgSdhB-HA/TgSdh18-FLAG background line, cultured in the absence of ATc or presence of ATc for 3 days.Proteins were separated by (a) SDS-PAGE or (b) BN-PAGE and detected with anti-c-myc, anti-FLAG, anti-HA or anti-TgTom40 antibodies.Western blots are from a single experiment and are representative of at least two independent experiments.For the BN-PAGE experiment, proteins were extracted in 1% (w/v) digitonin.The position of the ~660 kDa intact SDH complex and the putative ~430 kDa membrane anchoring subcomplex are indicated by magenta and orange arrows, respectively.Note that the day 0 and day 3 conditions on the TgSdh18-FLAG western blots are from the same exposure on the same membrane in an experiment where we also obtained data for parasites cultured for 1 and 2 days on ATc (electronic supplementary material, figureS6).

Figure 7 .
Figure 7. Loss of TgMPODD causes dissociation of the matrix-localized catalytic subunits from the membrane anchor of the SDH complex.(a,b) Western blot of proteins extracted from rTgMPODD-FLAG/TgSdhB-HA parasites grown in the absence or presence of ATc for 3 days, separated by (a) SDS-PAGE or (b) BN-PAGE and detected using anti-FLAG, anti-HA or anti-TgTom40 antibodies as indicated.For BN-PAGE, proteins were extracted in 1% (w/v) digitonin buffer.Western blots are from a single experiment and are representative of two independent experiments.Masses indicated with different coloured arrows next to bands were determined via relative migration distance (Rf) plots, taking the means of two independent experiments.(c) Western blot of proteins extracted from TgSdh18-c-myc, TgSdh11-c-myc, TgSdh15-c-myc, TgSdh31-c-myc and TgSdhA-c-myc expressing parasites, all tagged in a rTgMPODD-FLAG/TgSdhB-HA background, cultured in the absence or presence of ATc for 3 days.Proteins were extracted in 1% (w/v) digitonin, separated by BN-PAGE and detected with anti-c-myc or anti-FLAG antibodies as indicated.Western blots are from a single experiment and are representative of at least two independent experiments.The position of the ~660 kDa SDH complex is indicated with a magenta arrow.

Figure 8 .
Figure8.A model for the architecture of the SDH complex in T. gondii.(a) The functional T. gondii SDH complex comprises a ~660 kDa complex that contains all proposed subunits of the complex and may exist as a trimer (magenta arrow).The complex also exists in a higher-order structure of >720 kDa (blue arrow).Smaller complexes observed represent the membrane anchor lacking TgSdhA and with the TgSdhB protein tethered (~530 kDa complex; green arrow) or not-tethered (~430 kDa complex; orange arrow).(b) Depletion of TgSdhB leads to the formation of a ~430 kDa complex that contains the membrane-bound subunits of the complex.We propose that this represents the membrane anchor of the SDH complex in T. gondii.(c) Depletion of TgMPODD leads to the loss of TgSdhA and TgSdhB from the complex (grey and light brown arrows), with remnant complexes of ~400 and ~500 kDa containing some of the membrane anchoring subunits of the complex (dark brown arrows).These data suggest a role for TgMPODD in tethering the matrix components TgSdhB and TgSdhA to the membrane anchoring subcomplex.BN-PAGE data in the figure are derived from experiments depicted in figures 1, 3, 5 and 6.