Essential role of the conserved oligomeric Golgi complex in Toxoplasma gondii

ABSTRACT Survival of the apicomplexan parasite Toxoplasma gondii depends on the proper functioning of many glycosylated proteins. Glycosylation is performed in the major membranous organelles ER and Golgi apparatus that constitute a significant portion of the intracellular secretory system. The secretory pathway is bidirectional: cargo is delivered to target organelles in the anterograde direction, while retrograde flow maintains the membrane balance and proper localization of glycosylation machinery. Despite the vital role of the Golgi in parasite infectivity, little is known about its biogenesis in apicomplexan parasites. In this study, we examined the T. gondii conserved oligomeric Golgi (COG) complex and determined that contrary to predictions, T. gondii expresses the entire eight-subunit complex and that each complex subunit is essential for tachyzoite growth. Deprivation of the COG complex induces a pronounced effect on Golgi and ER membranes, which suggests that the T. gondii COG complex has a wider role in intracellular membrane trafficking. We demonstrated that besides its conservative role in retrograde intra-Golgi trafficking, the COG complex also interacted with anterograde and novel transport machinery. Furthermore, we identified coccidian-specific components of the Golgi transport system: TgUlp1 and TgGlp1. Protein structure and phylogenetic analyses revealed that TgUlp1 is an adaptation of the conservative Golgi tethering factor Uso1/p115. TgUlp1 and together with Golgi-localized TgGlp1 showed dominant interactions with the trafficking machinery that was predicted to operate endosome-to-Golgi recycling. Together, our study showed that T. gondii has expanded the function of the conservative Golgi tethering COG complex and evolved additional regulators of transport that are likely to serve parasite-specific secretory organelles. IMPORTANCE The Golgi is an essential eukaryotic organelle and a major place for protein sorting and glycosylation. Among apicomplexan parasites, Toxoplasma gondii retains the most developed Golgi structure and produces many glycosylated factors necessary for parasite survival. Despite its importance, Golgi function received little attention in the past. In the current study, we identified and characterized the conserved oligomeric Golgi complex and its novel partners critical for protein transport in T. gondii tachyzoites. Our results suggest that T. gondii broadened the role of the conserved elements and reinvented the missing components of the trafficking machinery to accommodate the specific needs of the opportunistic parasite T. gondii.

several parasite-specific organelles.Novel organelles, including rhoptries, micronemes, dense granules, and the inner membrane complex (IMC), play central roles in parasite egress, invasion, and virulence, making them attractive targets for research (1,2).Previous studies have established a close link between the biogenesis of these organelles and Golgi function (1,3).The Golgi is a conserved eukaryotic organelle that processes and modifies secreted proteins.It has been shown that T. gondii expresses a pleura of secreted proteins that contain carbohydrate modifications, which would require passing through the Golgi (4)(5)(6)(7).Major virulence factors TgMIC2, TgAMA1, ROP18, and TgCST1 are among these glycan-modified proteins.While such importance and abundance of glycosylation explains why coccidian parasites retained a nearly complete Golgi system, most apicomplexan parasites reduced this compartment (8,9).Despite its significance, the role of the Golgi in apicomplexan secretory pathways remains an enigma.
Transport between membranous organelles is mediated by vesicular trafficking, which is executed in several steps.In the donor compartment, cargo is selected and packed into coated membrane vesicles that are recognized, docked, and fused with the acceptor compartment upon arrival.The process is highly specific for types of vesicles, donor and target membranes, and involves massive molecular machinery (10).The major groups of regulators include cargo receptors, vesicle coat and its assembly factors, tethering machinery, membrane fusion SNARE receptors, and SNARE complex resolution factors.Selective capture, docking, and fusion of the arriving vesicles are mediated by a versatile group of vesicle tethering factors and specific SNARE molecules (11).Bioinformatic analyses of apicomplexan genomes identified orthologs of several multi-subunit tethering complexes but could not detect long coiled-coil tethers (12,13).HOPS, COVET, and TRAPP tethering factors in T. gondii are localized in post-Golgi compartments, the trans-Golgi network, and the endosome-like compartment (ELC), demonstrating their role in the biogenesis of secretory organelles (14)(15)(16).However, Golgi tethering complexes are insufficiently studied.
A previous bioinformatic search predicted four out of eight subunits of the conserved oligomeric Golgi (COG) complex in T. gondii, which is consistent with the current view of a reduction in trafficking machinery across apicomplexan parasites (12).The COG complex is a major Golgi multi-subunit tethering complex that has been extensively studied in model eukaryotes.The COG complex is organized into two subcomplexes, lobe A and lobe B, and facilitates the tethering and fusion of Golgi recycling vesicles, which are responsible for the maintenance of Golgi glycosylation enzymes and other Golgi resident proteins (17)(18)(19).COG complex subunits interact with a selective set of trafficking machinery, including COPI coat proteins, Golgi SNAREs, Rab effectors, and coiled-coil tethers (20)(21)(22)(23)(24)(25)(26)(27).Acute depletion of individual or multiple COG complex subunits in mammalian cells led to massive accumulations of COG complex-dependent (CCD) vesicles (28)(29)(30), while their complete knockout also changed the morphology of the Golgi and endosomal systems (31,32).It has been shown that malfunction in the COG complex alters N-and O-glycosylation of cellular proteins and causes several COG-related congenital disorders of glycosylation in humans (17,(33)(34)(35).
In the current study, we present experimental evidence that T. gondii expresses a complete oligomeric COG complex that is essential for tachyzoite growth, and we demonstrate its wider roles in parasites.We have found that the T. gondii COG complex interacts with a novel Golgi coiled-coil protein and identified a novel Golgi transport factor involved in retrograde transport in the late Golgi compartment.

Toxoplasma COG complex is composed of eight subunits
Phylogenetic analysis of the tethering complexes in metazoans identified four putative subunits of the octameric Golgi tethering COG complex in T. gondii, which could indicate that either the COG complex is reduced or the missing components were too dissimilar to be captured (Fig. 1) (12,36).To determine the exact composition of the Toxoplasma COG complex, we examined proteins associated with TgCog8, whose counterpart in other eukaryotes functions as a bridge between lobe A and lobe B of the COG complex (Fig. 1B) (19,37,38).The endogenous TgCog8 protein was fused with an auxin-induced degron (AID) and a 3×HA epitope tag (Fig. 1A) (39).Co-staining with the Golgi marker TgGRASP55 RFP confirmed that TgCog8 AID-HA was localized at the perinuclear Golgi (Fig. 1A).TgCog8 AID-HA protein complexes were isolated from dividing tachyzoites, and mass spectrometry followed by computational analysis detected 46 proteins with Significance Analysis of INTeractome (SAINT) scores of 0.5 or higher (Fig. 1C and D; Table S4).Eight factors were particularly abundant: alongside TgCog8, we detected three predicted COG complex subunits, TgCog2, TgCog3, and TgCog4, and five proteins with unknown functions (Fig. 1D).We discovered that four of these proteins were the missing COG complex subunits TgCog1, TgCog5, TgCog6, and TgCog7 (Fig. 1E; Fig. S1).TGME49_224150 contained a conserved Cog6 domain, suggesting that it is a Toxoplasma Cog6 subunit.Phylogenetic analyses of TGME49_290310, TGME49_251730, and TGME49_242030 verified that these proteins were Toxoplasma counterparts of COG1, COG5, and COG7, respectively (Fig. S1).Further examination of amino acid sequences showed that TgCog5 contains a CATCHR (complexes associated with tethering containing helical rods) fold (Fig. S2A) (40).Prior work in other eukaryotic models has shown that the binary COG5-COG7 subcomplex is stabilized by the antiparallel interactions of the α1-helix of the COG5 CATCHR fold and the N-terminal α-helixes of COG7 (41).Coincidently, the sequence alignment and folding predictions revealed three α-helixes in the N-terminus of the TGME49_242030 protein, as well as amino acid residues that are critical for COG5-COG7 interaction, confirming the TGME49_242030 identity as a TgCog7 subunit (Fig. S2B).Together with phylogenetic evidence and the nearly equimolar levels of COG subunits detected in the TgCog8 proteome, our findings

Research Article mBio
confirmed that Toxoplasma expresses a full complement of the eight subunits constitut ing the Golgi tethering COG complex.

Toxoplasma COG complex is essential for tachyzoite growth
To examine the function of the COG complex in T. gondii, we generated conditional knockdown lines for each subunit.Subunits TgCog2 through TgCog8 were analyzed in the auxin-induced degradation model, and TgCog1 was studied in the tet-OFF model (39,42).All the subunits of the Toxoplasma COG complex were abundantly expressed in the Golgi (Fig. 2).We established that treatment with either auxin for 30 min (TgCog2-TgCog8) or anhydrotetracycline (ATc) for 16 h (TgCog1) both led to robust depletion of the target proteins (Fig. 2B).Contrary to their counterparts in host cells, each subunit of the Toxoplasma COG complex was found to be essential for tachyzoite survival.Persistent downregulation of any subunit was sufficient in preventing lytic plaque formation (Fig. 2C).However, downregulating individual subunits had various effects on parasite replication.While depleting one of six subunits (TgCog2, TgCog4, TgCog5, TgCog6, TgCog7, and TgCog8) gradually arrested tachyzoite growth over multiple division cycles, depleting either TgCog1 or TgCog3 immediately affected parasite division (Fig. 2D).TgCog1-and TgCog3-deficient tachyzoites could not complete a second intravacuolar division cycle (an average of two parasites per vacuole).These results demonstrate the vital role of the COG complex in the survival of T. gondii tachyzoites, specifically lobe A subunits TgCog1 and TgCog3.

Rearrangement of the Toxoplasma Golgi during cell division
To better understand the COG complex's function in T. gondii, we first reconstructed Golgi dynamics in dividing tachyzoites (9).In eukaryotes, the Golgi expands and duplicates in synchrony with cell division to ensure that it is inherited by both daugh ter cells (43).We monitored Golgi morphology in tachyzoites by co-staining the COG complex subunit TgCog2 HA with budding stage marker TgIMC1 to visualize the assembly of internal daughters and centrosome marker TgCentrin1 to help identify parasites in the G 1 cell cycle phase (Fig. 3).Our results showed that the Toxoplasma Golgi undergoes significant rearrangement as the parasite progresses through endodyogenic cell division.The Toxoplasma Golgi exists as a single compact organelle during G 1 , duplicates in the S phase, breaks apart, and then restores its compact morphology during the internal budding process (Fig. 3A).We confirmed that the Toxoplasma Golgi loses connection with the centrosome prior to centrosome duplication at the G 1 /S transition, where the centrosome translocates to the posterior end of the parasite nucleus (Fig. 3B, G 1 and S panels) (44,45).The duplicated Golgi continues to grow and reconnects with the duplicated centrosomes, and each Golgi-centrosome unit is then delivered into the newly formed internal buds (Fig. 3A and B, budding progression).Golgi fragmenta tion is transient and characteristic of early budding (Fig. 3A, third panel) and is fully reassembled by mid-budding as a perinuclear rod-like structure at the apical end of the growing bud (Fig. 3A, bottom panel).Interestingly, TgCog2 was Golgi-associated in all cells, indicating that the entire COG complex remained membrane-bound during cell division.We also examined the internal membranes of tachyzoites by transmission electron microscopy (TEM) and confirmed the compound T. gondii Golgi organization (Fig. 3C).In tachyzoites, the Golgi system is composed of 3-5 developed cisterns and a number of Golgi-associated vesicles.TEM analyses further verified that the Golgi is associated with the apicoplast rather than the centrosome during the G 1 /S transition.T. gondii would need a repertoire of sophisticated regulatory machinery to maintain such complex organization of this essential and dynamic organelle, including the eight-subu nit Golgi tethering COG complex.

Toxoplasma COG complex deficiency severely affects the Golgi structure
Studies of the COG complex in mammalian cells revealed that siRNA and degron-assis ted depletion of either COG3, COG4, or COG7 subunits dramatically changes Golgi  S3. morphology (28,30,46).Recent TEM analyses of COG-deficient hTERT-RPE1 and HeLa cells showed accumulation of uncoated CCD vesicles followed by Golgi inflation, while prolonged COG3 deficiency caused severe Golgi fragmentation (28)(29)(30).To evaluate the effects of COG complex deficiency in T. gondii, we examined TgCog3 AID (lobe A subunit) and TgCog7 AID (lobe B subunit) tachyzoites by TEM.In untreated tachyzoite controls, the Golgi was organized as a multi-cisterna stack.However, this defined Golgi structure was lost after a short 30-min auxin treatment of TgCog3 AID parasites (Fig. 4A and B).The Golgi region was packed with free-floating large vesicles, and the Golgi cisternae appeared disorganized and fragmented (Fig. 4B and D).Prolonged TgCog3 deprivation resulted in the complete restructuring of internal membranes.In addition to vesiculation, the largest membranous structure of the cell, the ER, inflated and engulfed large portions of the cytoplasm (Fig. 4C).Depleting the lobe B subunit TgCog7 had a partial effect on the Golgi system.Both short (30 min) and long (8 h) auxin treatments caused Golgi vesicles to accumulate but did not significantly affect the Golgi cisternae (Fig. 4D through  S3. (E) Images of TgCog7-depleted parasites (30 min, IAA) show persistent vesiculation in the Golgi region.
(F) The images depict Golgi vesiculation in the RH TgCog7 AID-HA tachyzoite after 8 h of IAA treatment.
F).We also did not detect the inflation of internal membranes, even after depleting TgCog7 for 8 h (Fig. 4F).Our results suggest that the two lobes of the Toxoplasma COG complex, or certain subunits within the complex, provide different functions.This is consistent with findings in yeast, flies, and selected mammalian cell lines.The pheno types we observed in COG complex-deficient Toxoplasma share a signature of COG deficiency with human cells: accumulation of the CCD vesicles in the Golgi vicinity upon the acute depletion of COG.Like in yeast and HeLa cells, knocking down TgCog3 has a stronger effect on the Golgi structure than does downregulating COG7 (28).However, contrary to the studied eukaryotes, prolonged TgCog3 deprivation produced the inflation of an upstream secretory pathway compartment, the ER, rather than the resident compartment, the Golgi, suggesting significant differences in the pathways regulated by the Toxoplasma COG and COG complex of host cells.

Toxoplasma COG complex deficiency alters protein O-glycosylation
Protein glycosylation is one of the most common post-translational modifications of secreted proteins and is carried out by an extensive array of enzymatic machinery in the ER and Golgi to facilitate covalent attachment of oligosaccharides to specific amino acid residues.The elaborate process of glycosylation is executed in a precise order by specialized glycosyltransferases and glycosidases.Co-translational N-glycosylation of asparagine is carried out by ER enzymes, while O-glycosylation of serine or threo nine side chains is catalyzed by Golgi-localized N-acetylgalactosamine transferases (47).
Intracellular protein transport plays a central role in protein glycosylation, particularly the retrograde branch that restores the supply of Golgi glycosylation enzymes depleted by the continuous flow of anterograde transport.T. gondii encodes 68 glycogenes capable of assembling various types of glycans (5).Studies using metabolic labeling alone or in combination with genome editing have identified proteins that contain simple but abundant N-and O-glycosylations that play important roles in parasite invasion, virulence, and environmental sensing (7,(48)(49)(50).
We determined that excessive vesiculation and fragmentation of the Golgi was the primary defect of downregulating TgCog3 and TgCog7.To find out whether these COG-induced morphological changes also affected Golgi function in tachyzoites, we used lectin staining to examine global protein glycosylation.Total lysates of tachyzoites expressing the complete COG complex (−IAA) or those depleted of TgCog3 or TgCog7 subunits for 30 min or 8 h (+IAA) were probed with three types of glycan-binding lectins (Fig. S3A and B).Concanavalin A preferentially binds to the α-mannose core chain of oligosaccharides present in N-glycans, while HPA (Helix pomatia agglutinin) and Jacalin favor the immature Tn antigen of O-modified glycans (51)(52)(53).The selec tive binding of these lectins allowed us to distinguish between potential alterations in ER (N-glycosylation)-and Golgi (O-glycosylation)-dependent modifications.Lectin staining revealed that tachyzoites lacking TgCog3 or TgCog7 had defects in Golgi-based O-glycosylation (several species of HPA-and Jacalin-positive bands of glycoproteins were altered in COG-depleted samples), while ER-dependent N-glycosylation remained unaffected.Interestingly, although prolonged deprivation of TgCog3 leads to inflation of internal membranes, including the ER, this does not affect the function of ER enzymes.This suggests that ER-based glycosylation machinery is not regulated by the COG complex.Consistent with our TEM studies, depleting TgCog3 has a more prominent effect on O-glycosylation than does downregulating TgCog7.Our glycosylation analyses corroborate the primary role of the Toxoplasma COG complex in retrograde transport, particularly in recycling Golgi resident enzymes.

Toxoplasma COG complex interacts with the coat proteins of retrograde and anterograde vesicles
The conventional role of the COG complex in eukaryotes is in tethering retrograde intra-Golgi vesicles (17,33).Two types of vesicles are used in ER-Golgi and intra-Golgi transport: anterograde COPII vesicles that deliver cargo from the ER to the Golgi are coated with four-subunit COPII complexes composed of Sec24, Sec25, Sec13, and Sec31 proteins, while retrograde vesicles that restore the balance of Golgi enzymes and ER-Golgi trafficking machinery are covered with seven-subunit complexes containing α, β, β′, γ, δ, ε, and ζ COPI proteins (54,55).The Toxoplasma orthologs of COPI and COPII complexes were previously identified in silico, but neither anterograde nor retrograde Golgi transport had been studied in T. gondii (Fig. 5C and D) (56).We and the others had previously detected interactions between COG and COPI complexes in the other model systems, but surprisingly, mass spectrometry analysis of the TgCog8 pulldown identified equally enriched components of COPI and COPII coatomer complexes (Fig. 1D and 8; Table S4) (20,21,25,26,29).To verify the interactions of the Toxoplasma COG complex with both types of transport vesicles, we engineered transgenic lines co-expressing lobe A TgCog3 AID-HA or lobe B TgCog7 AID-HA proteins and subunits of COPI (TgCOPI-δ) or COPII (TgSec31) vesicles detected in TgCog8 complexes.Immunofluorescence micro scopy analyses confirmed the Golgi localization of epitope-tagged Toxoplasma coatomer proteins (Fig. S4A).We then examined the relationship between COG complex subunits TgCog3 and TgCog7 and coat proteins by high-resolution microscopy and 3D recon struction (Fig. 5A).Although both COPI and COPII vesicle coats had a moderate to high probability of colocalizing with the COG complex, the Pearson coefficient values suggested that the Toxoplasma COG complex associates with retrograde COPI vesicles (TgCOPI-δ) more closely than with anterograde COPII vesicles (Fig. 5B) (57).The strength of colocalization was comparable to that between COG complex subunits TgCog3 and TgCog7.
To verify the COG complex preference for retrograde transport, we performed a co-immunoprecipitation analysis of vesicle coat proteins and COG complex subunits (Fig. 5E and F).TgCOPI-δ myc and TgSec31 myc complexes were affinity purified and probed for TgCog3 AID-HA or TgCog7 AID-HA proteins.Both COG complex subunits were detected in the TgCOPI-δ myc pulldown, and only TgCog7 was detected in the TgSec31 myc pulldown, verifying the COG complex interaction with retrograde COPI-coated Golgi vesicles.This finding also suggests a possible interaction between the lobe B COG subcomplex and anterograde COPII-coated vesicles.However, the relative abundance of co-precipitated COG subunits suggests that the Toxoplasma COG complex preferentially interacts with retrograde transport machinery.
We then tested how COG complex deficiency affects the localization and expression of vesicle coat proteins.In line with co-immunoprecipitation results, we found that the retrograde coatomer TgCOPI-δ myc and not the anterograde coat subunit TgSec31 myc was substantially affected by TgCog3 AID-HA degradation (Fig. 5G; Fig. S4G).TgCOPI-δ myc began to relocate to the cytoplasm after brief TgCog3 AID-HA absence (+IAA, 40 min), and by 8 h, 100% of TgCog3-deficient tachyzoites had a pronounced loss of Golgi-associated TgCOPI-δ myc (Fig. 5H).In addition, extended TgCog3 AID-HA deprivation (+IAA, 8 h) led to a drastic decline in TgδCOPI myc expression (Fig. 5I).Downregulating the lobe B subunit TgCog7 AID-HA had a lesser effect on TgCOPI-δ myc that dissociated from fragmented Golgi membranes in 50% of TgCog7-deprived tachyzoites (Fig. 5F, H, and I; Fig. S4G).These results further confirmed that the Toxoplasma COG complex predominantly interacts with retrograde transport machinery and also indicated COG involvement in ER-Golgi anterograde transport.

Toxoplasma COG complex interacts with a coccidian Uso1-like Golgi tethering factor
Like the other COG complex subunits, TgCog8 also strongly interacted with a coc cidian-specific protein of unknown function, TGME49_289120 (Fig. 1D).To confirm TGME49_289120 expression and interaction with the Toxoplasma COG complex, we generated the TGME49_289120 AID conditional expression model in RHΔKu80Δhxgprt AtTIR1 parasites.Immunofluorescence microscopy analysis revealed its localization at the Golgi, and a plaque assay showed that TGME49_289120 was essential for tachyzoite  S3. (I) Western blot analysis of TgδCOPI myc stability in TgCog3-and TgCog7-deficient parasites (α-myc probe).The equal loading and the downregulation of TgCog3 and TgCog7 were confirmed with α-GRA7 and α-HA antibodies, respectively.growth (Fig. 6A and B).We also created TgCog3 and TgCog7 AID models expressing epitope-tagged TGME49_289120 myc and verified TGME49_289120 colocalization with the COG complex (Fig. 6C; Fig. S4B).To test the interaction with the COG complex, we examined TGME49_289120 myc complexes by mass spectrometry (Fig. 6E; Table S4).Corroborating proteomic studies of TgCog8, TGME49_289120 protein complexes contained a complete set of COG complex subunits, and we confirmed the presence of TgCog3 AID-HA and TgCog7 AID-HA by western blot analysis (Fig. 6D and 8).Finding a complete set of COG subunits indicated that TGME49_289120 interacts with the entire COG complex.The relative abundance of COG subunits suggested that TGME49_289120 interacted with the COG complex via TgCog5, TgCog2, and TgCog4 subunits.
Detailed analyses of interactions strongly indicated that TGME49_289120 is involved in retrograde transport, further supported by its preferential interaction with adaptor protein complex 5 (AP-5), derived from late endosomes and Golgi COPI coat proteins (Fig. 8; Table S4).Both types of vesicles are implicated in retrograde protein transport (58,59).We detected four components of the AP-5/SPG11 complex and six COPI subunits with relative enrichment of outer layer COPI components (α and β′ subunits).In addition, orthologs of the major Golgi (TgStx5, TgGOSR2, TgSec22, TgYkt6, and TgStx6) and endosome (TgStx12, TgVamp7, and TgStx8) SNARE complexes were among the strongest TGME49_289120 interactors (14,60,61).Finally, a substantial presence of cis-SNARE complex resolution factors, TgNFS, TgSNAP-α, and TgSNAP-γ, pointed toward the TGME49_289120 role in vesicle tethering and fusion rather than in assembly and budding (10).The enrichment of late endosomal and Golgi factors in Tg289120 complexes suggested that the role of this novel coccidian-specific factor lies in late endosome-to-Golgi and intra-Golgi transport.
Searching for clues of TGME49_289120 identity, we performed an in silico analysis of protein folding.We found that TGME49_289120 protein organization was like that of the eukaryotic tethering factor p115/Uso1 (Fig. 6F).Uso1/p115 is a long coiled-coil protein that promotes the capture, docking, and fusion of vesicles traveling between the ER and the cis-Golgi (62).Previous bioinformatic analyses suggested that apicomplexans lack conserved coiled-coil Golgi tethers, including Uso1 orthologs (12,36).Our examination of the TGME49_289120 structure revealed that, while vastly different at the amino acid sequence level, TGME49_289120 contained several multi-helical armadillo (ARM) repeats in the N-terminus predicted to fold in a structure reminiscent of the globular head domain of p115/Uso1 (Fig. 6G) (63).Interestingly, this region has a substantial degree of amino acid residue similarity (21%) with human p115 and includes two ARM repeats that were previously shown to interact with the COG complex via the COG2 subunit (24).Coincidently, TgCog2 was among the Toxoplasma COG subunits that preferentially interacted with TGME49_289120 (Fig. 8).Despite the lack of similarity, TGME49_289120 was also predicted to have a coiled-coil region in its C-terminus like Uso1/p115 (1,696-1,830) (Fig. 6F and G).Corroborating the prediction that a true ortholog was lost in evolution, our phylogenetic analyses determined that TGME49_289120 was not related to the eukaryotic Uso1/p115 tethering factors and only had an ortholog in the closely related coccidian parasite Hammondia hammondia (Fig. 6H).The similarities in structure, localization at the Golgi, and predicted roles in tethering vesicles between TGME_289120 and Uso1/p115 suggested that coccidian parasites either reinvented or severely modified the essential Golgi tethering factor, and we named it TgUlp1 (Uso1-like protein 1).

Toxoplasma expresses a novel Golgi transport factor
Another coccidian-specific protein with unknown function, TGME49_258080, was detected in the TgCog8 interactome (Table S4).To gather clues on its function, we built the TGME49_258080 AID model and found that TGME49_258080 was not essential for parasite replication, and similar to the COG complex, TGME49_258080 was expressed on the tachyzoite Golgi (Fig. 7A and B).High-resolution microscopy analyses of TGME49_258080 myc co-expressed with either TgCog3 AID-HA or TgCog7 AID-HA confirmed that the proteins are highly likely to colocalize (Fig. S4B and C).Examination of the TGME49_258080 protein putative 3D structure revealed two coiled-coil regions in the C-terminus with reasonable similarity to those in proteins that are involved in intracel lular protein transport in various eukaryotic models, including the conserved guanine nucleotide exchange factor Sec2 and intraflagellar transport protein from Chlamydomo nas reinhardtii IFTB1 (Fig. 7C; Fig. S4F) (64,65).We decided to name this novel Golgi-local ized TGME49_258080 protein TgGlp1 (Golgi protein 1).
Mass spectrometry analysis of isolated TgGlp1 myc complexes did not show Toxoplasma COG complex subunits, suggesting that interactions between TgGlp1 and the COG complex are indirect (Fig. 7D and 8; Table S4; Fig. S4D).Nevertheless, the TgGlp1 myc proteome had several hits in common with the TgCog8 interactome, indicating that the COPI coat was the best candidate for the link with the COG complex.Despite the lack of direct interaction, we found that the COG complex significantly affected TgGlp1 intracellular localization.Downregulation of TgCog3 and, to a lesser extent, TgCog7 led to TgGlp1 relocation from its resident compartment to dispersed associated punctata (Fig. 7E; Fig. S4E).Since TgGlp1 did not directly interact with the COG complex, the observed changes were likely due to TgGlp1 associating with the Golgi region that is highly responsive to acute degradation of TgCog3 or TgCog7, as demonstrated in our TEM experiments (Fig. 4).Interestingly, the TgGlp1 proteome had substantial overlaps with the TgUlp1 proteome, implying that these two novel coccidian-specific

DISCUSSION
The Golgi is a complex membranous organelle that functions as a transit station.It accepts vesicles from the ER (anterograde transport) and late endosomes (retrograde flow of the endosomal system) and sends vesicles to late secretory compartments, such as the TGN, late endosomes, and the plasma membrane (anterograde), and then back to the ER (retrograde transport).In addition, the Golgi constantly recycles its resident proteins in a trans-to-cis fashion (retrograde transport).Therefore, according to the current cisterna maturation model, the Golgi is equilibrated by the anterograde and retrograde flows of transport intermediates.These two directions of protein transport have different goals.Anterograde transport delivers ER-and Golgi-processed cargo to its destination, while retrograde transport recycles components of trafficking machinery and modifies enzymes to their places of function.How the Golgi segregates these two flows is still under investigation.However, it has been shown that model eukaryotes employ distinctive molecular machinery; anterograde transport is mostly mediated by COPII-, clathrin-, and AP-4-coated vesicles, while retrograde transport largely uses COPIand AP-1-, 2-, 3-, and 5-coated vesicles (66).
Most apicomplexan parasites have rudimentary Golgi systems, which can explain the lack of research attention to this organelle.Previously published data, and our study, demonstrate that T. gondii tachyzoites have a well-developed Golgi system composed of multiple cisterns and associated vesicles, corroborating retention of Golgi ribbon formation factors GRASP65/GRASP55 in the parasite genome (9).This observation sets coccidian parasites apart from the phylum and suggests an important role of Golgidependent processes in parasite survival.It also explains why T. gondii has preserved substantial transport machinery, while most apicomplexan parasites have lost many of these components (12,36).Bioinformatic and empirical analyses identified T. gondii orthologs of all known vesicle coats and fusion factors, including SNAREs, Rabs, and cis-SNARE resolution proteins (1,12,14,36,56,60).However, these studies predicted variable degrees of preservation of vesicle tethering machinery, with a lack of coiled-coil tethers and a structurally reduced COG complex (12,36).
Our study showed that T. gondii preserved a full set of subunits of the Golgi tether ing COG complex.We demonstrated that the T. gondii COG complex has a conserved function and displays several novel features.Like its higher eukaryote counterpart, the T. gondii COG preferentially interacts with the COPI vesicular coat and affects protein glycosylation, which confirms the conventional COG role in retrograde intra-Golgi transport (17,33).However, unlike in its eukaryotic counterparts, each complex subunit was essential for tachyzoite growth, raising the importance of Golgi tethering factors and the contribution of the Golgi to parasite survival.Although the immediate effect of T. gondii COG complex downregulation resembled the massive accumulation of Golgi-associated CCD vesicles observed in mammalian cells, the long-term effect was very different (Fig. 9).The extended cellular function of TgCog is likely related to its physical interaction with the anterograde ER-Golgi COPII trafficking machinery.COG complex deprivation in HeLa, HEK293T, and RPE1 human cells led to the inflation and severe fragmentation of the Golgi, but it was still capable of supporting anterograde transport and cell growth, while tachyzoites deficient in lobe A subunit TgCog3 had pronounced bloating in the ER that presented as a severely detached nuclear envelope (31,67).The ER inflation caused by prolonged COG complex deprivation in tachyzoites suggests that it is a downstream effect of the malfunction of the Golgi-to-ER segment of the retrograde transport pathway.The Golgi-to-ER block likely prevents the retrieval of ER anterograde trafficking machinery, leading to a block in ER-to-Golgi anterograde transport and ER cargo accumulation.Furthermore, without the COG tether, the Golgi may not accept the incoming transport; thereby, the TGN-and late endosome-derived retrograde vesicles would fuse with the downstream accepting compartment, the ER.A resulting disbalance of the transport system would eventually drain the anterograde (late Golgi, TGN, and endosomes) and flood the retrograde accepting stations (ER), reminis cent of the deficiencies observed in long TgCog3-deprived tachyzoites.Another possible explanation for the ER inflation caused by COG insufficiency is that the T. gondii Golgi would fail to accept anterograde ER-derived vesicles, blocking secretion.The presence of COPII coats of the anterograde vesicles supports the T. gondii COG complex involvement in anterograde transport.In human cells, COG depletion is partially compensated by a large group of COG-independent coiled-coil tethers (V.L. personal communication), while T. gondii lacks this compensatory mechanism.The normal growth of COG-deprived HEK293T cells indicates that the COG complex does not regulate ER-Golgi anterograde transport in mammalian cells.On the contrary, studies of temperature-sensitive mutants in yeast implied the yeast COG complex involvement in ER-Golgi anterograde trafficking (68,69).This suggests some similarities between COG function in yeast and T. gondii.The broad role of the COG complex in intracellular trafficking may explain the unique essentiality of this complex in T. gondii.Identification of the interactors of individual COG complex subunits and deciphering the composition of COG complex-dependent vesicles will shed light on their origin and expand our understanding of the COG complex function in T. gondii.
We made an intriguing discovery that suggests that T. gondii has either reinven ted or retained and severely altered several missing components of protein transport machinery.Two novel coccidian-specific factors detected in the TgCog8 pulldown appear to be a new addition to retrograde Golgi transport.The COG complex interactor TgUlp1 displays structural features of the missing coiled-coil Golgi tether Uso1/p115 (63).Despite that, the role of TgUlp1 is substantially different from that of Uso1, which supports the hypothesis that this factor was adapted rather than inherited.The function of conventional Uso1/p115 is limited to the ER-Golgi region where it interacts with COPII and COPI vesicles, whereas our analyses placed the T. gondii Uso1-like factor in the late Golgi compartment (trans-Golgi or TGN) where it receives the recycling AP-5 vesicles arriving from the endosomal system and, to some extent, the recycling intra-Golgi COPI vesicles (Fig. 9).Like TgUlp1, another factor identified in this study, TgGlp1, is a coccidian-specific protein with no distinctive structural features besides the presence of short, coiled-coil regions in its C-terminus.Although TgGlp1 and TgUlp1 do not interact with one another, their interactomes are nearly identical, indicating that TgUlp1 and TgGlp1 are involved in the same pathway.TgGlp1 and TgUlp1 are long proteins with coiled-coil domains that are often present in molecular tethers.Both factors favor interactions with retrograde transport machinery (AP-5 and COPI vesicles) and have an identical set of Golgi and endosomal SNAREs.However, we detected a significant difference that places TgGlp1 and TgUlp1 in different Golgi regions.Surprisingly, downregulating the COG complex affected the localization of TgGlp1 and not the direct COG complex partner TgUlp1.Upon downregulation of the lobe A subunit TgCog3, TgGlp1 was quickly redistributed to Golgi-associated vesicles, while TgUlp1 remained associated with the Golgi.This is reminiscent of the differential reaction of Golgi coiled-coil tethers to COG dysfunction in human cells.While Uso1/p115 demonstrated COG-independent Golgi localization, other tethers, like GOLGB1/giantin and GOLGA5/ golgin-84, were quickly redistributed to CCD vesicles (29).One plausible explanation is that TgGlp1 associates with the trans-Golgi or TGN that donates recycling vesicles to the preceding Golgi compartments, which are regulated by the COG complex.Thus, continuously budding TGN retrograde vesicles with attached TgGlp1 cannot fuse with acceptor Golgi membranes that lack the major Golgi tethering COG complex.We could not confirm direct TgGlp1-COG interaction, making the alternative less likely, where the COG complex would control TgGlp1 association with the Golgi.Nevertheless, the substantial overlap in COG complex localization and function with TgUlp1 and TgGlp1 suggests that they are part of the same trafficking machinery.
Why has T. gondii evolved novel transport factors?We found a rational answer in the interactomes of novel factors that were enriched in over 20 IMC proteins and suture factors, suggesting TgGlp1 and TgUlp1 involvement in the biogenesis of the parasitespecific IMC compartment.Why are these factors present only in coccidian parasites?Perhaps coccidian parasites replicate by internal budding, which requires recycling of IMC material and, in turn, requires expanding and adapting retrograde transport.In support of our hypothesis, TgGlp1 and TgUlp1 interactomes also contained Rab11b, which was previously shown to regulate IMC trafficking (70).Although we detected some rhoptries, micronemes, and dense granule factors in TgGlp1 and TgUlp1 pulldowns, their number and abundance were not comparable to IMC factors.Thus, the important observation of our study is that T. gondii supplemented and broadened the function of the conservative transport machinery, such as the COG complex, to accommodate the specific needs of the parasite.The novel components mimicking conventional machinery (Uso1/p115-like factor) may have evolved to integrate the service of novel organelles into a conserved membrane trafficking system.

Parasite cell culture
T. gondii RHΔKu80Δhxgprt AtTIR1 and RHTatiΔKu80 strains were maintained in human foreskin fibroblast (HFF) (ATCC, SCRC-1041) cells in DMEM (Millipore Sigma).A Myco plasma Detection (MP Biomedicals) PCR Kit was used to confirm that parasite strains and HFF cells were free of mycoplasma.The transgenic lines used in this study are listed in Table S1.

Phylogenetic analysis
Protein sequences were downloaded from the UniProt database and analyzed in NGPhylogeny (https://ngphylogeny.fr/) (71,72).The custom workflow included the MUSCLE algorithm and BMGE to align and curate sequences (73,74).Maximum likelihood-based inference with Smart Model Selection was implemented using the PhyML+SMS algorithm (75).We applied 1,000 bootstraps to test the optimization of each edge of the tree.The resulting phylogenetic trees were created using Newick utilities (76).Sequences of the analyzed portions are included in Table S2.

Construction of transgenic strains
Table S1 lists the transgenic lines and primers used.Targeting constructs were verified by sequencing, and PCR using gene-and epitope tag-specific primers was used to ensure that the tags were incorporated properly in their genomic locus.

Endogenous C-terminal tagging and conditional expression
To create AID conditional expression models for TgCog2, TgCog4, TgCog5, TgCog6, TgCog7, and TgCog8, genomic fragments of the 3′-end of the gene of interest were amplified by PCR and cloned into the pLIC-mAID_3xHA_HXGPRT vector digested with PacI endonuclease by the Gibson assembly method.The resulting constructs were linearized within the cloned gene fragment and transfected into the RHΔKu80Δhxgprt AtTIR1 parent.A similar approach was used to build 3×myc-epitope-tagged TgUlp1, TgGlp1, TgCOPI-δ, TgSec31, and TgCog3.Genomic fragments were cloned into the pLIC-3xmyc_DHFR-TS vector prior to linearization and transfection.

Endogenous N-terminal tagging and conditional expression
To build tet-OFF mutants of TgCog1 and TgCog2, the 5′-ends of the genes were amplified (Table S1 lists the primers used), digested with BglII/NotI (TgCog2) or BamHI/NotI (TgCog1), and ligated into the promoter replacement vector pTetO7sag4-3xHA_DHFR-TS (44).The resulting TgCog1 and TgCog2 tet-OFF constructs were linearized and transfec ted into the RHTatiΔKu80 strain.

Endogenous C-terminal tagging by CRISPR and conditional expression
To introduce the mAID-3×HA epitope into the C-terminus of TgCog3, we used the CRISPR/Cas9 approach as previously described (39).We created the pSAG1:CAS9-GFP U6: gsTgCog3 plasmid by site-specific mutagenesis of the pSAG1:CAS9-GFP U6: gsUPRT plasmid (generously provided by Dr. David Sibley) to introduce a double DNA break in the 3′UTR of the TgCog3 locus (77).A tagging cassette containing the selection marker HXGPRT flanked by 40-bp gene-specific sequences amplified by PCR was co-transfec ted with the pSAG1:CAS9-GFP U6: gsTgCog3 plasmid into the RHΔKu80Δhxgprt AtTIR1 parent.

Parasite transfection and selection
Parent strain tachyzoites were grown in HFF monolayers.Freshly lysing parasites were collected and mixed with 50 µg DNA in 100 µL Cytomix buffer supplemented with 20 mM ATP and 50 mM reduced glutathione.Electroporated parasites (Amaxa, Lonza) underwent a 24-h recovery in regular growth medium prior to drug selection.Drugresistant polyclonal populations were cloned by limiting dilution.The resulting clonal population was screened using immunofluorescence microscopy with antibodies against the tagged protein of interest.Recombination at the target locus was confirmed by PCR amplification.The expression of tagged proteins was verified by western blot analysis.

Plaque assay
Plaque assays were performed in six-well plates.Confluent HFF monolayers were infected with 50 parasites per well and treated with 500 µM indole-3-acetic acid (IAA, auxin), which prompts the degradation of mAID-tagged proteins.The viability of tet-OFF parasites was evaluated by treating infected monolayers with 1 µg/mL ATc.Plaques developed for 7 days at 37°C and then stained with crystal violet and counted.Three biological replicates of each assay were performed.

Vacuole size
The number of parasites per vacuole was determined after 16 h of growth at 37°C.Quantifications were done on 50-100 randomly selected vacuoles.

Immunofluorescence microscopy analysis
HFF-infected parasites were grown on glass coverslips, fixed, permeabilized, and then incubated with the desired antibody.The following antibodies were used: rabbit α-Myc (clone 71D10; Cell Signaling Technology), rat α-HA (clone 3F10; Roche Applied Scien ces), mouse α-centrin (clone 20H5; Millipore Sigma), and rabbit α-IMC1 kindly provided by Dr. Gary Ward (University of Vermont, VT, USA).A dilution of 1:500 was used for Alexa conjugated secondary antibodies (Thermo Fisher Scientific).Nuclei were stained using 4′,6-diamidino-2-phenylindole (DAPI, Sigma).Glass coverslips were mounted in ImmunoMount (Thermo Fisher Scientific) and analyzed under a Zeiss Axiovert micro scope with an Apotome optical slicer.Images were processed in Zen Blue 2.0 and Adobe Photoshop 2023.Single slice images were also collected on a confocal Zeiss LSM 880 Airy Scan 2 microscope at less than 50% saturation.The Pearson colocalization coefficient was calculated based on 3D reconstructions generated from confocal image Z-stacks using the "colocalization module" (Zen Blue 2.6) and the program-determined threshold.

Transmission electron microscopy
The TEM samples were processed according to published protocols with some modifications (78,79).Briefly, infected HFF monolayers were rinsed and fixed with 1% paraformal dehyde (PFA) (EMS) and 2.5% glutaraldehyde (GA) (EMS) in phosphate-buffered saline (PBS) for 2 h at room temperature and stored in 1% PFA in PBS at 4°C and finally fixed with 2.5% GA and 0.05% malachite green (EMS) in 0.1 M sodium cacodylate buffer, pH 6.8 (20 min, ice).Then, cells were washed with 0.1 M sodium cacodylate buffer, post-fixed in 0.5% osmium tetroxide and 0.8% potassium ferricyanide in 0.1 M sodium cacodylate buffer (30 min, room temperature), washed and incubated in 1% tannic acid (20 min, ice), and incubated in 1% uranyl acetate (1 h, room temperature).Specimens were gradually dehydrated in increasing ethanol concentrations, washed with propylene oxide (EMS), and incubated in a 50% PO/resin mixture before embedding in Araldite 502/Embed 812 resins (EMS).Ultrathin 50-nm sections were post-stained with aqueous uranyl acetate (EMS) and Reynold's lead citrate (EMS).Images were acquired on an FEI Technai G2 TF20 intermediate-voltage transmission electron microscope at 80 keV (FEI Co.) equipped with an FEI Eagle 4k CCD camera and processed with FEI software.

Immunoprecipitation assay
Protein samples were prepared from at least 2 × 10 9 parasites collected by filtration and centrifugation.Total proteins were extracted in phosphate-buffered saline solution with 263 mM NaCl, 1% Triton X-100, and Halt protease and phosphatase inhibitor cocktail (Thermo Scientific).Proteins were isolated on α-HA or α-Myc magnetic beads (MBL Life Science).Pulldown efficiency was verified by western blot analysis.To eliminate non-specific interactions, pulldowns from parental strains were performed using matching magnetic beads.

Western blot analysis
Protein samples made from isolated parasites or protein extracts were mixed with Laemmli sample buffer, heated at 95°C for 10 min, and sonicated.Proteins were separated via SDS-PAGE and transferred onto nitrocellulose membranes.Membranes were probed with the following primary antibodies: rat α-HA (clone 3F10; Roche Applied Sciences), rabbit α-Myc (Cell Signaling Technology), mouse α-GRA7 (kindly provided by Dr. Peter Bradley, UCLA, CA, USA), and mouse α-tubulin A (12G10; kindly provided by Dr. Jacek Gaertig, University of Georgia, Athens, GA, USA).To visualize antigens, membranes were incubated with α-rat, α-mouse, or α-rabbit secondary antibodies conjugated with horseradish peroxidase (Jackson ImmunoResearch).Proteins were visualized using an enhanced chemiluminescent detection kit (Millipore).To detect changes in protein glycosylation, membranes with separated proteins were blocked in Bio-Rad blocking buffer and probed with lectins as previously described (29).The following lectins were selected: concanavalin A that recognizes mannose and α-(1-6) polymannose (52), Helix pomatia agglutinin (HPA) that recognizes the Tn antigen (51), and Jacalin that binds to T and Tn antigens (53).Briefly, HPA-conjugated Alexa-647 (Thermo Fisher Scientific) or biotinylated concanavalin A and Jacalin (Vector) were diluted 1:1,000 in Bio-Rad blocking buffer and incubated with membranes for 1 h at room temperature.Biotinyla ted proteins were detected with 30 min of additional incubation with streptavidin-Alexa 647 (Thermo Fisher Scientific).Membranes were washed four times in PBS and imaged using the Odyssey Imaging System.

Data analysis
SAINT was used to identify TgCog8, TgUlp1, and TgGlp1 interactors with the default parameter setting in the function "SAINT express version." We incorporated proteomes of the parental strains RHΔKu80Δhxgprt AtTIR1 and RHΔKu80 to account for non-specific binding.

Structural analysis of hypothetical proteins
To identify homology models, amino acid sequences of the target proteins were examined in the SWISS-MODEL suite (82).The resulting models were further analyzed in PyMol (https://pymol.org/2/).Available images of the folded proteins were downloaded from AlphaFold2.Alignment of the protein sequences downloaded from the UniProt (https://www.uniprot.org/)database was performed using the MUSCLE algorithm and exported to Jalview (83).The domain of interest within each protein was selected and displayed using the Clustal color scheme.

FIG 1
FIG 1 Identification of the Toxoplasma COG complex subunits.(A) Endogenously tagged TgCog8 AID-HA (green) colocalizes with Golgi marker GRASP55 RFP (red).The blue DAPI stain marks the tachyzoite nucleus.(B) The schematic shows the two-lobe composition of the COG complex.Previously identified subunits are shown with bright colors.(C) Equal amounts of the input fraction (S, soluble), proteins not retained on the αHA beads (UB, unbound), and proteins retained on αHA beads (B, bound) were examined by western blot analysis to confirm the efficiency of the TgCog8 AID-HA pulldown.The tubulin A probe confirms the specificity of the TgCog8 AID-HA pulldown.(D) The log2 values of the protein spectra detected by mass spectrometry analysis of the TgCog8 AID-HA complexes and non-specific interactions detected in the parental strain are plotted on the graph.Proteins maximally enriched in the TgCog8 AID-HA complexes are encircled and listed in the table on the right.(E) The diagram shows the protein organization of the Toxoplasma COG complex subunits.

FIG 2
FIG 2 Eight essential subunits of the Toxoplasma COG complex (A) Immunofluorescence microscopy shows the Golgi expression of the HA TgCog1 tet-OFF model and AID models of TgCog2 AID-HA -TgCog8 AID-HA subunits using α-HA (red) and DAPI (blue) staining.The gray line represents parasite staining with antibodies against surface marker TgIMC1.(B) Western blot analysis confirmed the expression and downregulation of HA TgCog after 16 h of treatment with 2 µM anhydrotetracycline (ATc) and TgCog2 AID-HA -TgCog8 AID-HA after 30 min of treatment with 500 µM auxin (indole-3-acetic acid, IAA).Western blots were probed with α-HA to detect the COG complex subunits, with either α-tubulin A or α-GRA7 to confirm equal loading of the total lysates.(C) Images of host cell monolayers infected with HA TgCog1 tet-OFF or TgCog2 AID-HA -TgCog8 AID-HA AID tachyzoites and grown with or without the indicated treatment for 7 days.Representative images of three independent experiments are shown.(D) The average number of parasites per vacuole after 16 h of growth in the presence or absence of 2 µM ATc (TgCog1) or 500 µM IAA (TgCog2-TgCog8) was quantified in three independent experiments.A hundred random vacuoles were evaluated.Mean values −/+ SD are plotted on the graph.The raw counts and the t-test values are included in TableS3.

FIG 3
FIG 3 Cell cycle dynamics of the Toxoplasma Golgi.(A) Major cell cycle phases were determined based on the parasite cytoskeletal morphology (α-IMC1, internal budding) and the shape of the nucleus (DAPI).Various Golgi states and transitions, including organelle duplication, segregation, fragmentation, and reassembly, were visualized using the TgCog2 HA marker.(B) The Golgi-centrosome relationship was examined by co-staining TgCog2 HA and centrosome marker centrin 1 (α-Centrin1).(C) TEM images of dividing tachyzoites.Major organelles and structures are labeled.

FIG 4
FIG 4 The COG complex is required for proper Golgi morphology and function.(A) A TEM microphotograph of the tachyzoite expressing the entire COG complex (−IAA).The enlarged image on the right depicts four cisterns of the Golgi apparatus.(B) The image shows changes in the Golgi region of the RH TgCog3 AID-HA tachyzoite after 30 min of IAA treatment.Note the accumulation of large Golgi vesicles.(C) The prolonged TgCog3 deprivation on the tachyzoite internal membranes leads to inflation of the internal membranes, including the ER (red arrowheads).(D) Quantification of Golgi-derived vesicles in the tachyzoites after 30 min of TgCog3 and TgCog7 deficiency.Ten TEM images were evaluated.The raw counts and t-tests are included in TableS3.(E) Images of TgCog7-depleted parasites (30 min, IAA) show persistent vesiculation in the Golgi region.

FIG 5
FIG 5 Toxoplasma COG complex interacts with COPI and COPII coatomer proteins.(A) Images of the 3D reconstruction of the Golgi-localized TgCog3 AID-HA (red), TgCog7 AID-HA (red), TgδCOPI myc (green), and TgSec31 myc (green) in the perinuclear region (nucleus, DAPI, blue).(B) The Pearson coefficient was determined based on the colocalization analysis of a minimum of 10 tachyzoites.(C and D) The diagrams of the COPI (C) and COPII (D) coatomer complexes and the tables of T. gondii orthologs of individual subunits.(E and F) Immunoisolation of the TgδCOPI myc (E) and TgSec31 myc (F) complexes from parasites co-expressing endogenous TgCog3 AID-HA or TgCog7 AID-HA .The insoluble (P, pellet), soluble (S), and depleted soluble fractions (nB, not bound) and the beads with precipitated complexes (B) (10 times more than the other fractions) were probed with α-myc and α-HA antibodies to detect potential interactions (the co-IP panels) and to confirm the efficient pulldown of the bait protein (the IP panel).(G) Immunofluorescence microscopy analysis of TgδCOPI myc in parasites expressing (−IAA) or lacking TgCog3 AID-HA or TgCog7 AID-HA (+IAA).The coat protein was visualized with α-myc antibodies and co-stained with DAPI (blue) and α-TgIMC1 (traced with a gray line).(H) Quantification of the TgδCOPI myc Golgi association in parasites treated or not treated with IAA for 8 h.The mean and SD values of the vacuole counts are plotted on the graph.The raw counts and t-test values are shown in TableS3.(I) Western blot analysis of TgδCOPI myc stability in TgCog3-and TgCog7-deficient parasites (α-myc

FIG 6
FIG 6 Toxoplasma COG complex interacts with a novel Uso1-like factor.(A) Localization of TGME49_289120 AID-HA protein was determined by co-staining of the factor (α-HA), nuclear stain DAPI, and parasite surface marker TgIMC1 (traced with a gray line).A 30-min treatment with 500 μM IAA resulted in robust TGME49_289120 AID-HA downregulation.(B) Images of the host cell monolayers infected with Tg289120 AID-HA tachyzoites and grown with or without 500 μM IAA for 7 days.(C) Immunofluorescence analysis of the parasites expressing Tg289120 myc in the RH TgCog3 AID-HA or TgCog7 AID-HA mutants.Parasites were co-stained with α-myc (green), α-HA (red), and DAPI (blue).Inset shows that the proteins overlap in the Golgi region.(D) Western blot analysis of immunoprecipitated Tg289120 myc complexes from parasites co-expressing endogenous TgCog3 AID-HA or TgCog7 AID-HA .The insoluble (P, pellet), soluble (S), and depleted soluble fractions (nB, not bound) and the beads with precipitated complexes (B) (10 times more than the other fractions) were probed with α-myc and α-HA antibodies to detect protein interactions (the co-IP panels) and to confirm the efficient Tg289120 myc pulldown (the IP panel).(E) The log2 values of the protein spectra detected by mass spectrometry analysis of the Tg289120 myc complexes and proteins detected in the pulldown from the parental strain are plotted on the graph.The differently colored dots represent categories of transport proteins.(F) Schematic of Homo sapiens p115 and T. gondii Ulp1 protein organization.The signature domains and regions of similarity are shown.(G) Folding prediction for Mus musculus p115 and selected regions of T. gondii Ulp1 are shown (AlphaFold2, PyMol, and SwissProt).Note that the image of the coiled-coil region of T. gondii Ulp1 is a compilation of five different models listed in the legend below.(H) Phylogenetic tree of TgUlp1 and various Uso1/p115 orthologs.

FIG 7
FIG 7 TgGlp1 is a novel T. gondii Golgi transport factor.(A) Localization of TGME49_258080 AID-HA protein was determined by co-staining of the factor (α-HA), nuclear stain DAPI, and parasite surface marker TgIMC1 (traced with a gray line).A 30-min treatment with 500 μM IAA resulted in robust TGME49_258080 AID-HA downregulation.(B) Images of the host cell monolayers infected with Tg258080 AID-HA tachyzoites and grown with or without 500 µM IAA for 7 days.(C) Schematic of TgGlp1 protein organization.The identified domains are shown.The regions of similarity are listed below.(D) The log2 values of the protein spectra detected by mass spectrometry analysis of the TgGlp myc complexes and proteins detected in the pulldown from the parental strain are plotted on the graph.The differently colored dots represent categories of the selected transport proteins.(E) Images of tachyzoites expressing or lacking TgCog3 or TgCog7 for the indicated time.The associated changes in TgGlp1 myc expression were visualized by co-staining parasites with α-myc and DAPI (blue).The insets are overexposed images of the selected Golgi regions.

FIG 8
FIG 8 Summary of TgCog8, TgUlp1, and TgGlp1 protein interactions.The table lists the major groups of TgCog8, TgUlp1, and TgGlp1 interactors identified in the proteomic studies.The FC column shows the fold change calculated as a ratio of spectra detected in the indicated protein IP and in the IP from the parental strain.FC also accounts for protein size.The ortholog column contains the abbreviated name of the human ortholog protein.The localization column indicates the known localization of the factor in the studied model organisms.NP, not present.

FIG 9
FIG 9 The model of the COG complex function in T. gondii.The drawing depicts a portion of the secretory pathway (from bottom to top): endoplasmic reticulum (ER); cis-, mid-, and trans-Golgi; trans-Golgi network (TGN); and endosome-like compartment (ELC).The organelles are colored according to the predicted localization of the COG complex (red), TgUlp1 (orange), and TgGlp1 (yellow).The arrows show the direction of vesicular transport and the type of transported vesicles.The lobe A subunit TgCog3 is predicted to localize to the acceptor Golgi membrane.The immediate effect of acute TgCog3 degradation (30 min) is the block of the Golgi receiving function.In the absence of the Golgi tether, the anterograde and retrograde vesicles cannot dock or fuse with the Golgi and accumulate in the cytoplasm.The large membranous compartment with functional tethers, the ER, likely accepts rogue vesicles.Continued TgCog3 deprivation (8 h) leads to ER inflation, and impaired anterograde transport contributes to the phenotype.The created disbalance of the anterograde and retrograde flows results in Golgi reduction and fragmentation of the late Golgi compartments.