Toxoplasma F-box protein 1 is required for daughter cell scaffold function during parasite replication

By binding to the adaptor protein SKP1 and serving as substrate receptors for the SKP1 Cullin, F-box E3 ubiquitin ligase complex, F-box proteins regulate critical cellular processes including cell cycle progression and membrane trafficking. While F-box proteins are conserved throughout eukaryotes and are well studied in yeast, plants, and animals, studies in parasitic protozoa are lagging. We have identified eighteen putative F-box proteins in the Toxoplasma genome of which four have predicted homologs in Plasmodium. Two of the conserved F-box proteins were demonstrated to be important for Toxoplasma fitness and here we focus on an F-box protein, named TgFBXO1, because it is the most highly expressed by replicative tachyzoites and was also identified in an interactome screen as a Toxoplasma SKP1 binding protein. TgFBXO1 interacts with Toxoplasma SKP1 confirming it as a bona fide F-box protein. In interphase parasites, TgFBXO1 is a component of the Inner Membrane Complex (IMC), which is an organelle that underlies the plasma membrane. Early during replication, TgFBXO1 localizes to the developing daughter cell scaffold, which is the site where the daughter cell IMC and microtubules form and extend from. TgFBXO1 localization to the daughter cell scaffold required centrosome duplication but before kinetochore separation was completed. Daughter cell scaffold localization required TgFBXO1 N-myristoylation and was dependent on the small molecular weight GTPase, TgRab11b. Finally, we demonstrate that TgFBXO1 is required for parasite growth due to its function as a daughter cell scaffold effector. TgFBXO1 is the first F-box protein to be studied in apicomplexan parasites and represents the first protein demonstrated to be important for daughter cell scaffold function.

Introduction Toxoplasma gondii is an intracellular apicomplexan parasite responsible for one of the most common parasitic infections in humans and animals [1,2]. Infections are initiated by digesting either bradyzoite-containing tissue cysts or sporozoite-laden oocysts that are disrupted in the stomach [3]. Released parasites then infect the small intestine and convert into tachyzoites, which triggers the recruitment of inflammatory cells, which are in turn infected and used to disseminate throughout the host where they convert into long-lived tissue cysts [4,5]. Occasionally, cysts reactivate and the released parasites will revert to tachyzoites, which replicate by a unique process termed endodyogeny where the two daughter parasites develop within the mother [6]. Since parasite replication underlies onset of disease in toxoplasmosis patients, it is critical to study the mechanisms regulating endodyogeny.
Formation of the inner membrane complex (IMC) is a critical step in endodyogeny. The IMC is a unique organelle comprised of an intermediate-filament like cytoskeletal network and membranous sacs whose outer leaflet anchors the actin-myosin gliding machinery [7][8][9] whereas the cytoplasmic leaflet is associated with subpellicular microtubules that extend from the apical end of the parasite for most of its length [10,11]. During daughter cell development, the IMC emerges from a structure named the daughter cell scaffold (DCS) that is located apically to the parasite's outer core of its bipartite centrosome [12,13]. The DCS serves as the docking site for Rab11b-regulated vesicles containing the material needed to form the emerging IMC [14]. Despite its central role in endodyogeny, it is largely unknown what proteins localize to the DCS nor is it understood how the DCS regulates IMC assembly.
F-box proteins (FBPs) are a family of proteins defined by the presence of a F-box domain and can be divided into three classes, FBXW, FBXL and FBXO [15] according to the presence or absence of canonical substrate recognition domains [15,16]. F-box proteins are best known as subunits of the SKP1/Cullin/F-box containing E3 ubiquitin ligase (SCF-E3) complex [17] in which the F-box domain binds SKP1 to recruit itself and/or other substrates for polyubiquitination [17,18]. Although well studied in fungi, plants, and metazoans, the function of Fbox proteins in apicomplexan parasites remains unexplored. Previously, we reported that an O 2 -regulated Toxoplasma prolyl hydroxylase modifies a proline residue in TgSKP1, which can then be modified by a series of glycosyltransferases [19][20][21]. TgSKP1 prolyl hydroxylation/glycosylation is predicted to alter F-box protein binding leading to changes in substrate recognition [22].
Here, we identify 18 putative F-box proteins in the Toxoplasma genome that, relative to other organisms, represents a significantly streamlined repertoire of F-box proteins [23,24]. We focused on Toxoplasma F-box protein 1 (TgFBXO1) and found that it localizes to the DCS early during endodyogeny and is required for DCS function. Thus, TgFBXO1 represents the first protein identified to be an executor of DCS function.

Identification of TgFBXO1 as an apicomplexan conserved F-box protein
F-box proteins contain an F-box domain of~50 amino acids that are organized into a triple αhelical bundle that mediates binding to SKP1 [25]. We employed a series of reciprocal BLASTp and hidden Markov searches seeded with 43 F-box sequences from known F-box proteins including 7 from crystal structures (S1A Fig) and 37 from two well studied yeast species (S1B Fig), to assemble a list of 18 predicted Toxoplasma gondii F-box proteins (S1 Table;  In a parallel study, we sought to identify candidate F-box proteins by investigating the TgSKP1 interactome. To maximize coverage, TgSKP1 was captured i) from lysates of extracellular parasites (80% extracellular) prepared at two different detergent concentrations using bead-bound affinity purified polyclonal antiserum (UOK75) generated against TgSKP1 [19], or ii) by using bead-bound anti-FLAG M2 monoclonal antibody to recover TgSKP1 from detergent lysates of extracellular or intracellular parasites of a strain named TgSKP1 SF in which the TgSKP1 locus was modified by the addition of a C-terminal Strep/Flag (SF) tag. To differentiate specific from adventitious interactors, non-immune rabbit IgG was used as a control for the UOK75 immunoprecipitation, and the parental RHΔhxgprtΔku80 strain was used as a control for mAb M2 immunoprecipitation. The immunoprecipitates were subjected to nLC-MS/MS analysis and proteins were assigned from peptide identifications at 1% FDR using Sequest, and quantified based on Abundance in Proteome Discoverer 2.2. After filtering to remove non-specific interactions and major organellar proteins, and relaxing the proteinlevel FDR for predicted FBPs from the bioinformatics search (S1 Fig), twenty-one proteins were identified as candidate SKP1-interacting proteins. Of these, five were present in both the TgSKP1 SF and UOK75 immunoprecipitations (Fig 1A, S2 Fig and S2 Table). These included the expected SCF complex proteins TgSKP1, CUL-1 and RBX1, a Skp1 glycosyltransferase (GAT1), and TgFBXO1 (TgGT1_310930). Three additional FBPs predicted from the bioinformatics approach were detected in the TgSKP1 SF immunoprecipitates, but two were assigned only with medium confidence.
We focused efforts on TgFBXO1 because it is conserved in Plasmodium, is important for Toxoplasma fitness according to a genome-wide CRISPR screen [26], and is the most highly expressed Toxoplasma F-box protein throughout the tachyzoite cell cycle https://toxodb.org/ toxo/app/record/gene/TGME49_310930. To confirm that TgFBXO1 interacts with TgSKP1, we cloned a C-terminal 3XHA epitope tag onto TgFBXO1 (TgFBXO1 HA ) via homologous recombination [27]. Western blotting lysates from TgFBXO1 HA expressing parasites but not the parental RHΔhxgprtΔKu80 strain with anti-HA antibodies revealed a single immunoreactive band at~100 kDa, which is the approximate expected molecular weight of TgFBXO1 HA (Fig 1B; Input 21% O 2 ). Next, TgSKP1 was detected by Western blotting TgFBXO1 HA immunoprecipitates with anti-TgSKP1 antisera (Fig 1B; IP 21% O 2 ). Taken together, the FBP and interactome discovery pipelines led to the identification of TgFBXO1 as an evolutionarily conserved F-box protein that stably associates with TgSKP1, and is potentially a major component of the SCF-E3 complex in Toxoplasma. RHΔhxgprtΔKu80 and TgSKP1 SF tachyzoites were solubilized and immunoprecipitated with affinity purified anti-TgSKP1 (UOK75) or anti-FLAG (M2), respectively. The immunoprecipitates were trypsin digested and subjected to proteomic analysis. Protein candidates were selected from a list of protein hits that were detected in multiple replicates and minimally recovered with irrelevant antibodies (for UOK75) or the untagged strain (for M2), and are reported as found in either both (red bars) or only individual immunoprecipitations with either high or medium confidence. Proteins known or predicted to interact with TgSKP1 are underlined. See S2 Fig and S2  Previously, we reported that 154 Pro of TgSKP1 is hydroxylated and glycosylated and that these post-translational modifications are predicted to change the conformation of the Fbox protein-binding region of TgSKP1 as they do in the social amoeba Dictyostelium discoideum when a conserved proline residue in DdSKP1 is similarly modified [22]. An O 2 -regulated prolyl hydroxylase modifies 154 Pro and under low O 2 conditions can be observed by Western blotting as a decrease in the apparent molecular weight of TgSKP1 [19,20,28

TgFBXO1 is important for parasite growth
Because TgFBXO1 is important for parasite fitness [26], we used a conditional knockdown system where the TgFBXO1 native promoter was replaced with an anhydrotetracycline (ATC)repressible promoter and an amino-terminal 3XHA epitope tag was cloned in frame to generate HA(ATC) TgFBXO1 (Fig 2A) [29]. HA(ATC) TgFBXO1 protein was undetectable after 24 h treatment with 1 μg/ml ATC ( Fig 2B). Next, HA(ATC) TgFBXO1 growth was assessed by plaque assay and found that relative to the parental strain growing with or without ATC that a~60% decrease in numbers of plaques were present when HA(ATC) TgFBXO1 expression was decreased ( Fig 2C and S3 Fig). Moreover, the size of the HA(ATC) TgFBXO1 plaques that did form were~70% smaller than plaques formed by HA(ATC) TgFBXO1 expressing parasites ( Fig  2D and S2 Fig). Although TgFBXO1/TgSKP1 interactions were enhanced at 0.5% O 2 , this growth phenotype was not enhanced at low O 2 levels. Therefore the rest of this study's experiments were performed under normoxic (21% O 2 ) conditions. Taken together, these data indicate that TgFBXO1 is important for optimal Toxoplasma growth.

TgFBXO1 forms an apical structure early during endodyogeny
Although detectable by Western blotting, HA(ATC) TgFBXO1 could not be visualized by immunofluorescence staining suggesting inaccessibility of the N-terminal tag in fixed parasites. We therefore assessed TgFBXO1 localization using the C-terminal-tagged TgFBXO1 HA strain and found that in interphase parasites TgFBXO1 localized to the apical end of the parasite and to the periphery where it appeared to colocalize with the IMC protein IMC3, which is a cytoskeletal IMC protein (Fig 3A; G1) [30]. We next examined TgFBXO1 localization during endodyogeny and found that early during endodyogeny TgFBXO1 localized to a perinuclear location before IMC3 associated with daughter IMCs (Fig 3A; S-phase). Later during M-phase (note lobe-shaped nuclei that are migrating into nascent daughter cells), TgFBXO1 remained primarily at the apical end of the growing daughter parasites while IMC3 extended posteriorly (Fig 3A; M-phase).
ISP1 is another IMC protein that is inserted into the membranous alveoli via fatty acylation. Before IMC3 associates with the daughter IMC, ISP1 forms an apical cap structure in the daughter parasites that remains throughout endodyogeny and into interphase [31]. Comparing ISP1 and TgFBXO1 localization revealed that TgFBXO1 attained its perinuclear localization before ISP1 did ( Fig 3B). As parasites progressed through endodyogeny, the ISP1 cap was found near TgFBXO1 although TgFBXO1 was distinct and apical to ISP1 suggesting that the two proteins were in distinct structures. In interphase parasites, apical TgFBXO1 staining did not colocalize with RNG1, which is a component of the parasite's microtubule organizing center, which is apical to the ISP1 cap (S4 Fig) [32], indicating that these two proteins also reside in distinct structures. We next used biochemical extraction to assess how TgFBXO1 associates with the IMC [33]. Parasites were mechanically lysed under hypo-osmotic conditions and cytoskeletal fractions were collected by low speed centrifugation. The pellets were resuspended in isotonic buffer and IMC embedded proteins were differentially extracted with either high salt (to disrupt electrostatic interactions), Triton X-100 (to extract membrane-associated proteins) or SDS (to extract cytoskeletal proteins). TgFBXO1 was extracted with both Triton and SDS indicating that it is most likely associated with the IMC as a membrane-associated protein in a manner similar to ISP1 (Fig 3C). However, we noted that in contrast to ISP1 that was exclusively associated with the IMC that a significant proportion of TgFBXO1 was soluble.
The first step in endodyogeny is duplication of the parasite's novel bipartite centrosome, followed by kinetochore duplication, assembly of subpellicular microtubules, formation of the IMC, and finally daughter cell budding and emergence [6,12,34]. To more definitively determine when during endodyogeny the TgFBXO1 apical structure forms, TgFBXO1 HA parasites were fixed and stained with anti-TgCentrin-1 antibodies to detect the outer centrosome core. We were able to detect parasites where TgFBXO1 remained associated with the mother IMC but containing duplicated outer centrosome cores (Fig 4A; S) indicating that centrosome duplication precedes TgFBXO1 recruitment to the forming daughter parasite. As endodyogeny progressed, TgFBXO1 were positioned apically to the duplicated outer centrosome cores Toxoplasma genome duplication and daughter cell development are coordinated by the parasite MAP kinase, TgMAPK-L1, and in the absence of TgMAPK-L1 signaling centrosome duplication continues although IMC development cannot [29]. We therefore tested whether the apical TgFBXO1 structures formed independently of centrosome duplication by treating TgFBXO1 HA parasites with either DMSO (vehicle control) or the TgMAPK-L1 inhibitor, SB505124 [35]. In contrast to control parasites, TgFBXO1 HA staining remained peripheral in SB505124-treated parasites containing �2 outer centrosome cores and numbers of apical TgFBXO1 structures in daughter parasites were significantly reduced (Fig 4B & 4C). Although it is unclear how TgMAPKL1 signaling is linked to TgFBXO1 apical structure formation, these data indicate that formation of this structure is dependent on the parasite signaling that genome duplication is completed.
The Toxoplasma kinetochore duplicates and separates after centrosome duplication but before ISP and IMC proteins form the daughter cell bud [29,36]. We therefore compared timing of TgFBXO1 apical structure formation and kinetochore duplication/migration by staining TgFBXO1 HA cells with antibodies against the kinetochore protein TgNuf2 [37]. We identified parasites with distinct TgFBXO1 apical structures that contained duplicated kinetochores that had yet to complete separation suggesting that formation of the TgFBXO1 apical structures Subpellicular microtubules, which provide positional and structural cues for daughter cell IMC development and organization [11], form after centrosome/kinetochore duplication and separation [38,39]. To assess the relationship between the TgFBXO1 apical structures and daughter cell microtubules, TgFBXO1 HA cells were stained with antibodies against acetylated tubulin, which denote stable microtubule filaments [40]. Parasites were identified that contained TgFBXO1 apical staining but before acetylated tubulin could be detected in the daughter cells ( Fig 5A). As cell cycle progressed, acetylated tubulin colocalized with apical TgFBXO1 although F-box protein staining was consistently more apical.
To test whether TgFBXO1 apical localization was dependent on microtubule assembly, TgFBXO1 HA parasites were treated with either DMSO or oryzalin (2.5 μM), which blocks polymerization of parasite but not host microtubules [39]. Oryzalin treatment, which caused the expected defects in microtubule assembly and cytokinesis [41], led to an accumulation of apical TgFBXO1 staining with most parasites possessing 8 structures (Fig 5B & 5C), indicating that TgFBXO1 apical localization was independent of microtubules.

TgFBXO1 is a component of the daughter cell scaffold
Collectively, the localization data indicate that TgFBXO1 is recruited to a structure that lies apically to the outer centrosome core as well as the emerging IMC. The DCS, which serves as a . Parasites were treated with 3 μM SB505124 (or DMSO as a vehicle control) for 24 h, fixed and then stained to detect TgCentrin1, TgFBXO1, and DNA. Note the appearance of SB505124-treated parasites with >2 centrin1 + foci and disorganized TgFBXO1 staining. Bars = 5 μm. (C). Quantification of numbers of parasites with apical TgFBXO1 structures. Data represents averages and standard deviations of three independent experiments with at least 50 parasites examined/experiment (D). TgFBXO1 HA parasites were stained to detect TgFBXO1 and the kinetochore marker TgNuf2. Note TgFBXO1 HA apical structures form after kinetochore duplication but before separation is completed as evidenced by the lobed TgNuf2 staining (Insert). Bars: 2μm. https://doi.org/10.1371/journal.ppat.1007946.g004 Toxoplasma F-box protein functions in Endodyogeny nucleation site for the daughter IMC and microtubules to emerge from, has such a localization [34,42]. The small molecular weight GTPase TgRab11b localizes early during endodyogeny to the DCS and regulates IMC protein trafficking to the nascent IMC [14]. To compare TgRab11b and TgFBXO1 HA localization, TgFBXO1 HA parasites were transfected with a Rab11b expression construct which contains a N-terminal myc-tagged ddFKBP domain to control protein levels by adding a synthetic ligand, Shield1 [14]. Parasites grown in the presence of low Shield1 concentrations (0.1 μM) express low levels of TgRab11b with no apparent effect on growth [14]. Under these conditions, examination of >50 parasites in early S-phase revealed thatTgFBXO1 and TgRab11b co-localized indicating that TgFBXO1 is a DCS component (Fig 6A; S). On the other hand, the two proteins displayed limited colocalization in interphase parasites (Fig 6A; G1).
To test whether DCS targeting of TgFBXO1 is TgRab11b dependent, TgRab11b-transfected TgFBXO1 HA parasites were grown with 1 μM Shield1, which leads to TgRab11b dysfunction due to its overexpression [14]. Under these conditions, overexpressed TgRab11b was mislocalized from the DCS and Golgi to the periphery (Fig 6A 1 bottom panels), which is consistent with [14]. TgFBXO1 was also mislocalized but unlike TgRab11b appeared as cytoplasmic punctae.
The appearance of TgFBXO1 cytoplasmic punctae in TgRab11b overexpressing parasites suggests that TgFBXO1 traffics to the DCS via a membrane trafficking pathway. In addition, TgFBXO1 detergent extractability from the IMC (Fig 3C) suggested that it is membrane-associated. The amino acid sequence of TgFBXO1 predicts neither a transmembrane domain nor a signal sequence but does have a consensus N-myristoylation site at 2 Gly and this site is conserved amongst the coccidian TgFBXO1 homologs (S5 and S6 Figs). To test whether this potential N-myristoylation site is important for DCS and IMC targeting, RHΔhxgprtΔKu80 parasites were transfected to express either wild-type TgFBXO1 (ptubTgFBXO1 HA ) or a mutant where 2 Gly is mutated to an alanine (ptubTgFBXO1 HAG2A ). In both interphase and mitotic parasites, ptubTgFBXO1 HA was properly localized to the periphery (arrow) and DCS (arrow heads), respectively ( Fig 6B). In contrast, ptubTgFBXO1 HAG2A localization appeared to be largely cytoplasmic and was not localized to either the IMC or DCS.

TgFBXO1 is important for DCS function
To test the importance of TgFBXO1 in DCS function, HA(ATC) TgFBXO1 (or parental TATiΔKu80) parasites were grown in the absence or presence of ATC, fixed 1-5 days later and TgCentrin-1/ISP1 were detected by immunofluorescence staining. In TgFBXO1-depleted daughter parasites, ISP1 was localized near TgCentrin-1 but it was disorganized and the typical apical caps observed in untreated parasites were lacking (Fig 7A and S7A Fig). Development of this phenotype was time dependent and by 5 days post-infection, was evident in~76% parasites. Similarly, the ISP1 cap normally found at the mother cell's apical end was absent in the TgFBXO1-depleted parasites.
Next, we assessed IMC3 localization and observed a time-dependent increase in IMC3-containing daughter cell buds that were malformed relative to TgCentrin-1. This phenotype was displayed as either: i) IMC3-labelled daughter buds associated with nuclei but centrin-1 staining was absent (Fig 7B (arrows) and S7B Fig). ii) Dividing parasites with only a single IMC (Fig 7B (triangles) and S7B Fig). iii) Nuclei devoid of both TgCentrin1 and IMC3 (Fig 7B  (asterisks) and S7B Fig). We quantified numbers of HA(ATC) TgFBXO1-depleted parasites with IMC3 defects and found that, like ISP1, they progressively increased to~55% of the parasites counted. Finally, we also noted interphase parasites containing IMC3 whirls (Fig 7B (X) and S7B Fig;), which was reminiscent of the effect of TgRab11b overexpression on IMC3 localization [14].
Besides IMC development, the DCS also promotes assembly of the daughter cell subpellicular microtubules. To examine whether microtubule assembly is affected in TgFBXO1-depleted parasites, HA(ATC) TgFBXO1 parasites were grown in the absence or presence of ATC and then newly assembled microtubules were imaged using anti-acetylated tubulin antibodies. In Toxoplasma F-box protein functions in Endodyogeny contrast to parasites grown in the absence of ATC, microtubules in HA(ATC) TgFBXO1 parasites with ATC were often missing or not properly organized by five days after adding ATC (Fig 7C  and S7C Fig). However, these cytoskeletal defects took longer to develop than either the ISP1 or IMC3 phenotypes suggesting that TgFBXO1 likely regulates IMC development and that the microtubule phenotypes are a secondary consequence.

Identification of DCS microdomains
Having established TgFBXO1 as a DCS effector protein, we next used super resolution microscopy to gain insight into the organization and structure of this complex. First, we sought to define the spatial organization between the DCS and the developing IMC by staining dividing TgFBXO1 HA parasites with anti-IMC3 and anti-HA antibodies. At an early stage during endodyogeny (after IMC3 begins to accumulate at the daughter cell bud but before it is cleared from the mother cell IMC), TgFBXO1 was found to accumulate as a distinct cap-like structure that was positioned apically from IMC3 structure (Fig 8A; S1 Video). However, we also noted that TgFBXO1 extended posteriorly from the cap. In addition, TgFBXO1 was not homogenously Toxoplasma F-box protein functions in Endodyogeny distributed throughout the DCS (Fig 8A). TgFBXO1 HA expressing parasites were similarly stained to detect TgFBXO1 and ISP1. In contrast to IMC3 whose localization was distinct from TgFBXO1, ISP1 was more proximal to TgFBXO1 although TgFBXO1 was positioned more apical than ISP1 (Fig 8B and S2 Video). As noted earlier, TgRab11b is another DCS-localized protein. Thus, we assessed TgRab11b and TgFBXO1 localization in dividing parasites using super resolution microscopy. Remarkably, Toxoplasma F-box protein functions in Endodyogeny TgFBXO1 and TgRab11b localization did not overlap in the DCS but instead appeared to localize to distinct regions and in some cases appeared to be interdigitated although we cannot exclude the possibility that antigen accessibility limits detection of either protein (Fig 8C and S3  Video). However, we discount this possibility since examination of > 50 parasites revealed only interdigitation but no colocalization between TgFBXO1 and TgRab11b. Taken together, these data indicate the DCS is likely composed of at least different microdomains and that TgFBXO1 and TgRab11b are distinct markers for each.

Discussion
Comprising a ubiquitous eukaryotic protein family, F-box proteins play important roles in diverse biological functions [15,16]. Although extensively studied in metazoans, plants, and fungi, little is known about their function(s) in protists and in particular apicomplexans. Here, we identified 18 putative Toxoplasma F-box proteins, which approximates the number found in two well studied yeast systems but is considerably less than the number in Dictyostelium, humans and other organisms [23,24]. TgFBXO1 was reported to be important for parasite fitness [26]; a finding we corroborate using a conditional expression system. We also demonstrated by reciprocal co-immunoprecipitation assays that TgSKP1 and TgFBXO1 interact with each other supporting TgFBXO1 designation as a F-box protein. We further found that TgFBXO1 is a DCS component; a conclusion supported by multiple lines of data. First, it localizes to a structure positioned apically to the daughter cell's outer centrosome core and developing IMC. Second, daughter cell TgFBXO1 apical structures form before ISP1 and IMC3 recruitment to developing daughter parasites. Third, TgFBXO1 becomes localized to the DCS after centrosome duplication but before the kinetochore completely separates. Finally, TgFBXO1 colocalized with TgRab11b at the DCS, albeit in distinct microdomains, and TgFBXO1 recruitment to the DCS was TgRab11b-dependent.
Sequence homology-based studies suggest that proteins related to TgFBXO1 consist of at least 3 domains (S5 Fig). The C-terminal domain and features of the middle predicted Fbox domain are conserved in proteins throughout the alveolate clade of protists (S5 Fig), which includes ciliates and dinoflagellates as well as apicomplexans. The protein acquired a 330-amino acid N-terminal domain only in Toxoplasma and closely related cyst-forming coccidia (no trace of this sequence exists in other extant proteins). Thus, a core function of the TgFBXO1 clade may have emerged early in evolution to serve a coccidia specific function, such as organization of the membranous villi that lie underneath the plasma membrane, which is the IMC in apicomplexans [43][44][45].
Decreased TgFBXO1 expression had distinct effects on two types of IMC proteins. The IMC alveoli meshwork embedded protein IMC3 formed daughter cell IMCs that were not properly aligned with the outer centrosome core. In contrast, ISP1, which is component of a distinct IMC subcompartment, was unable to form its apical cap structure during endodyogeny and this phenotype was maintained in interphase parasites. It should be noted that ISP1 is not an essential gene thus indicating that it is likely that the observed growth defects are not due to ISP1 [31] but rather ISP1 is serving as a marker for the ISP compartment. While these data are consistent with an emerging model in which IMC components are recruited to the daughter cell through distinct mechanisms [31], they also reveal that the DCS is required for both to occur. It is also possible that only ISP1 or IMC3 require TgFBXO1 and that defects observed with the other protein is a downstream effect. Discrimination between these two possibilities awaits characterization of TgFBXO1-interacting proteins.
IMC protein trafficking to the DCS and emerging bud is a poorly understood process. The small molecular weight GTPase, TgRab11b, is required for this process although it remains unclear where it executes its function and what are its effectors [14]. TgRab11b and TgFBXO1 both localize to the DCS but their effects on IMC proteins differed-defects in IMC formation vs misaligned daughter buds, respectively. These differences could be due to different approaches to achieving target protein loss of function (overexpression of an inactive TgRab11b mutant and TgFBXO1 knockdown) or could represent distinct roles for TgRab11b and TgFBXO1 in IMC biogenesis and organization. But given that TgFBXO1 requires TgRab11b for its DCS localization, we favor a model where TgRab11b functions to recruit diverse types of cargo and membrane vesicles to the DCS and that TgFBXO1 is important for executing DCS function in the development of the new IMC. This model is consistent with our finding that the DCS is composed of distinct microdomains and that each likely has distinct functions.
Fatty acylation is emerging as a critical regulator of targeting of IMC and IMC-related proteins to an emerging apicomplexan daughter cells although different proteins utilize different combinations of fatty acid moieties [31, [46][47][48][49]. Here, we report that N-myristoylation is required for TgFBXO1 targeting to the DCS. Since N-myristoylation is thought to mediate protein-membrane interactions and is a co-translational process [50,51], it was surprising that 50% of total TgFBXO1 protein was soluble. Explanations for this could include a TgFBXO1interacting protein that acts to sequester the N-terminal myristoylated glycine or that N-myristoylation of TgFBXO1 occurs post-translationally. While our future work will address these possible mechanisms, these data suggest that TgFBXO1 fatty acylation is a dynamic process that regulates the protein's trafficking and targeting.
Using a conditional expression system, TgFBXO1 was found to be required for optimal parasite growth most likely due to its function in regulating IMC development and organization. TgFBXO1 expression using this system leads to undetectable levels of the tagged protein within 24 h after adding ATC. Yet, the IMC defects were only minimally evident at this time point but rather accumulated with time over the 120 h time course that we performed. One reason for such a delay is that low but undetectable levels of TgFBXO1 are sufficient to sustain replication, a finding previously reported for Toxoplasma actin [52]. Alternatively, a second Fbox protein can compensate for the loss of TgFBXO1 albeit not as efficiently.
We are currently unable to differentiate whether TgFBXO1 acts in a SCF-E3 ubiquitin ligase dependent or independent manner [15,53,54]. It is possible that TgFBXO1 functions in IMC protein ubiquitylation to regulate their turnover and/or targeting during endodyogeny. Such a scenario is supported by the finding that many ubiquitylated Toxoplasma proteins are cell cycle regulated including ISP1 and a large number of other IMC proteins [55]. It is also possible that TgFBXO1 is itself an SCF-E3 substrate and its ubiquitylation is required during endodyogeny for either trafficking of TgFBXO1 from the IMC to the DCS or for its turnover from the mature IMC early during endodyogeny. It is noteworthy that a parasite deubiquitinase, TgOTUD3 [15,53,54,56,57] was recently reported to coordinate mitosis and cytokinesis by properly matching the correct numbers of centrosomes with numbers of daughter buds. However, we do not believe that TgFBXO1 functions in this manner because in contrast to TgOTUD3, loss of TgFBXO1 leads to unmatched numbers of outer centrosome cores and daughter buds without an increase in parasite ploidy.
Alternatively, TgFBXO1 may function in a SCF-E3 ubiquitin ligase independent manner [58][59][60]. Such a mechanism could occur by the TgSKP1/TgFBXO1 heterodimer, which according to precedent is expected to have high affinity [61][62][63], being incorporated into a supramolecular complex where TgSKP1 is shielded from interacting with cullin-1 to form a functional SCF complex. Recent cryo-EM studies of the yeast kinetochore provide an example of how the Skp1/Ctf13 (a yeast F-box protein) complex contributes to kinetochore structure in a way where Skp1 is inaccessible to Cul-1 interaction [59,60]. Certainly, it is conceivable that SCF-E3 dependent or independent mechanisms need not be mutually exclusive, because association with the SCF-E3 could be a mechanism to turn over excess TgFBXO1 that is not associated with the DCS or IMC.
In summary, this work represents the first investigation of F-box proteins in apicomplexan parasites and reveals the first DCS-localized protein that is required for DCS function. Identification of TgFBXO1-interacting proteins will allow us to determine how TgFBXO1 executes this function. In addition, it remains to be determined whether TgFBXO1/TgSKP1 interactions are regulated by O 2 -dependent hydroxylation and glycosylation of TgSKP1 and whether these impact DCS function or another process.

F-box homology search methods
Previously, 4 putative F-box-like motifs were identified in the Toxoplasma genome based on a hidden Markov algorithm [64]. Further BLASTp [65] and hidden Markov approaches using Fbox-like motifs originating from known F-box domains from 7 crystal structures (S1A Fig) and the entire collection of budding and fission yeast F-box proteins (S1B Fig) identified 14 additional putative F-box proteins in Toxoplasma (S1C Fig). Databases searched included EupathD (https://eupathdb.org/eupathdb/) (Toxoplasma, Neospora caninum, Plasmodium spp, and Eimeria spp.), Saccharomyces Genome Database (https://www.yeastgenome.org), and PomBase (https://www.pombase.org) (Schizosaccharomyces). Each candidate was used to re-search the genome until a comprehensive candidate list was created, which was trimmed of sequences that exceeded the range of variation at individual positions of known F-box sequences (see S1A and S1B Fig) with greatest emphasis given to the pattern of hydrophobic residues. Other domain identities were searched using Conserved Domains and Protein Classification search (http:// www.ncbi.nlm.nih.gov/Structure/cdd/cdd.shtml) based on database CDDV3.05-42589 PSSMs with E<0.1 [66]. Searches were also performed at Uniprot (www.uniprot.org/).
To identify TgFBXO1 homologues, the NCBI non-redundant database (November 2018) was searched using the standard BLASTp algorithm at e<100, initially with N-, F-, and C-domains (S2 Fig) of TgFBXO1 as query sequences. Except for the F-domain, only sequences from alveolates were retrieved. Repeat searches with non-apicomplexan alveolates failed to identify additional sequences outside of alveolates. Top-scoring sequences representing the diversity of known alveolate genomes were initially aligned in COBALT and manually curated for optimal alignment of hydrophobic residues and minimal indels except adjacent to Gly and Pro residues.
4 to 8 μl of the reconstituted peptides were loaded onto an Acclaim PepMap C18 trap column (300 μm, 100 Å) in 2% ACN, 0.05% TFA at 5 μl/min, eluted onto an Acclaim PepMap C18 column (75 μm × 150 mm, 2 μm, 100 Å), and eluted with a linear gradient consisting of 4-90% solvent B (solvent A: 0.1% FA; solvent B: 90% ACN, 0.08% FA) over 180 min at a flow rate of 300 nl/min on an Ultimate 3000 RSLCnano UHPLC system, and eluted into the ion source of an Orbitrap QE+ mass spectrometer (Thermo Fisher). The spray voltage was set to 1.9 kV and the temperature of the heated capillary was 280˚C. Full MS scans were acquired from m/z 350 to 2000 at 70k resolution, and MS 2 scans following higher energy collisioninduced dissociation (HCD, 30) were collected for the Top10 most intense ions, with a 30" dynamic exclusion. Intervening blank runs ensured absence of carryover between replicates. The acquired raw spectra were analyzed using Sequest HT (Proteome Discoverer 2.2, Thermo Fisher) with a full MS peptide tolerance of 10 ppm and MS 2 peptide fragment tolerance of 0.02 Da, and filtered to generate a 1% target decoy peptide-spectrum match (PSM) false discovery rate for protein assignments, which were filtered at 1% FDR or relaxed at the protein level to 5% FDR (as indicated). Variable modifications were Met oxidation, Q and N deamidation, and N-terminal acetylation; static modification was Cys carbamidomethylation.

Plaque assays
One hundred parasites were added to each well of a 6 well plate containing confluent HFFs in the presence or absence of 1μg/ml ATC. After 5 days, monolayers were methanol-fixed, stained with 0.1% crystal violet, and counted as described [71]. Briefly, plaques imaged using an Olympus SZ61 stereomicroscope equipped with a video camera. Plaque areas were measured using ImageJ software (https://imagej.nih.gov/ij/).

Immunoprecipitation
TgFBXO1 HA tachyzoites grown at 21% or 0.5% O 2 in a Baker Hypoxia Chamber (Sanford, ME) were harvested by syringe lysis and washed in ice-cold PBS. Parasites (1x10 8 ) were resuspended in 50 mM Tris-HCl pH 7.4 with 1% Triton X-100, 100 mM NaCl, 1 mM NaF, 0.5 mM EDTA, 0.2 mM Na 3 VO 4 , 1X protease inhibitor cocktail (Thermo Fisher Scientific), incubated on ice for 30 min and then subjected to sonication. Lysates were clarified by centrifugation at 16,000 xg, incubated with mouse α-anti-HA clone 12CA5 conjugated with protein G beads (Sigma-Aldrich) for 16 hours at 4˚C. Immune complexes were separated with SDS-PAGE and then Western blotted with antibodies against rat anti-HA clone 3F10 (Roche) or rabbit anti-TgSKP1 UOK75 [19].

IMC fractionation
IMC fractionation was done essentially as described in [33]. Briefly, freshly harvested tachyzoites were resuspended in hypotonic lysis buffer (10 mM Tris HCl pH 7.8, 5 mM NaCl) and lysed by repeated freeze thawing followed by centrifugation at 1,000 xg to generate fractions S1 (supernatant) and P1 (IMC). S1 was collected and centrifuged at 100,000 xg and fractions S2 (cytosol) and P2 (non-IMC membrane) were collected. Fraction P1 was resuspended in isotonic buffer (10 mM Tris HCl pH 7.8, 150 mM NaCl). SDS or Triton X-100 in isotonic buffer were added to P1 to final concentration of 1% and then centrifuged at 16,000 xg. Alternatively, P1 fraction in the presence of 500 mM NaCl were centrifuged at 100,000 xg for 1 h to isolate proteins that interact with the IMC via electrostatic interactions. All manipulations were performed at 4˚C and all buffers were supplemented with 1X Protease Inhibitor cocktail.

Immunofluorescence microscopy
Toxoplasma-infected HFFs grown on coverslips were fixed either with 4% w/v paraformaldehyde in phosphate buffered saline (PBS) for 20 min at room temperature or by ice-cold methanol for 15 min. Cells were permeabilized with 0.1% Triton X-100 in PBS for 10 min, blocked in 5% w/v bovine serum albumin (BSA) in PBS for 60 min, incubated overnight with primary antibodies (S2 Table) at 4˚C. Coverslips were washed with PBS and then incubated for 60 'with Alexa Fluor 488-or Alexa Fluor 594-conjugated secondary antibodies (1:2000, Thermo Fisher Scientific). DNA was stained either by incubation with 1 μg/ml DAPI (Thermo Fisher Scientific) for 5 minutes followed by mounting in ProLong Glass Antifade Mountant (Thermo Fisher Scientific) or with DAPI-containing VECTASHIELD mounting medium (Vector Labs; Burlingame, CA). Images were acquired using a 100X Plan Apo oil immersion 1.46 numerical aperture lens on a motorized Zeiss Axioimager M2 microscope equipped with an Orca ER charge-coupled-device (CCD) camera (Hamamatsu, Bridgewater, NJ). Images were collected as a 0.2-μm z-increment serial image stacks, processed using Volocity (version 6.1, Acquisition Module (Improvision Inc., Lexington, MA)). Images were deconvolved by a constrained iterative algorithm, pseudocolored, and merged using the Volocity Restoration Module.
For super resolution microscopy, 20 images were acquired as 0.15 μM z-increment serial images using Leica Hyvolution Acquisition Software (Leica; Buffalo Grove, IL) with a Leica TCS SP8 confocal microscope equipped with a 100x/1.47 TIRF oil immersion objective lens and both white light laser (470nm-670nm) and 405 nm diode laser. Image datasets were then deconvolved using Huygens deconvolution software (Hilversum, Netherlands) and 3D volumes generated Leica visualization software. Images from same experiments were processed using identical settings. Unless otherwise noted, data were quantified from at least 50 randomly selected images for each condition from three independently performed experiments.

Statistical analyses
When appropriate, one-way ANOVA with Tukey's post hoc test or Student's t test was performed with GraphPad Prism (GraphPad, La Jolla, CA). All proteins that were identified based on 2 or more peptides at 1% FDR, and were not predicted to reside within organelles, and were enriched �8-fold in TgSKP1 SF extracts relative to RHΔΔ extracts, are shown. Additional proteins that were seen when criteria were reduced to one peptide and the protein FDR increased to 5%, are included if they were predicted to be in the TgSKP1 interactome based on F-box sequence motifs or other information. These are labeled in green, and in parentheses if seen only with the relaxed criteria. Proteins whose interaction were supported by detection with a different antibody in panel D are underlined in bold. (B) Same as in panel A, from a 4 μl injection. Assignments with higher confidence and at greater abundance were confirmed. (C) Comparison of protein abundance, as detected in panel A (8 μl injections), in SF-tagged and untagged extracts. At the right are data from arbitrarily selected examples that were present in both samples, indicating non-specific interactions. The average total Abundance was calculated for this protein set for each sample, and used to generate a normalization factor relative to the TgSKP1 SF IC values, which was applied to each set of values in Panels A-C. (D, E) Similar analysis in which RHΔΔ tachyzoites were lysed in 1% or 0.2% NP-40 and incubated with affinity purified anti-Skp1 (UOK75) or nonspecific IgG on magnetic beads. (D) Abundance values are reported for all 4 conditions for proteins whose enrichment in UOK75 relative to non-specific IgG was �4. (E) Values for non-specifically bound proteins. No normalization was performed for this set. See S2 Table for the values and data on the identification criteria. Sequences homologous to TgFBXO1 were aligned in COBALT and subject to manual refinement. To facilitate visualization of relatedness, acidic residues are in blue, basic in dark red, small or prolines in red, and hydrophobic in green. Positions of similarity in one group are highlighted in yellow (hydrophobic), gray (acidic), green (basic), or teal (small). Underlines separate phylogenetic groups: the top two groups consist of different coccidian subclades, the next two groups represent other apicomplexans, the next group contains a chromerid (apicomplexan predecessor), and the bottom group contains non-apicomplexan alveolates including ciliates and dinoflagellates.  Fig 7. (A, B, C). Correspond to panels A, B, and C in Fig 7. (EPS) S1