Actin and an unconventional myosin motor, TgMyoF control the organization and dynamics of the endomembrane network in Toxoplasma gondii

Toxoplasma gondii is an obligate intracellular parasite that relies on three distinct secretory organelles, the micronemes, rhoptries and dense granules, for parasite survival and disease pathogenesis. Secretory proteins destined for these organelles are synthesized in the endoplasmic reticulum (ER) and sequentially trafficked through a highly polarized endomembrane network that consists of the Golgi and multiple post-Golgi compartments. Currently, little is known about how the parasite cytoskeleton controls the positioning of the organelles in this pathway, or how vesicular cargo is trafficked between organelles. Here we show that F-actin and an unconventional myosin motor, TgMyoF, control the dynamics and organization of the organelles in the secretory pathway, specifically ER tubule movement, apical positioning of the Golgi and post-Golgi compartments, apical positioning of the rhoptries and finally, the directed transport of Rab6-positive and Rop1-positive vesicles. Thus, this study identifies TgMyoF and actin as the key cytoskeletal components that organize the endomembrane system in T. gondii. Author Summary Endomembrane trafficking is a vital cellular process in all eukaryotic cells. In most cases the molecular motors myosin, kinesin and dynein transport cargo including vesicles, organelles and transcripts along actin and microtubule filaments in a manner analogous to a train moving on its tracks. For the unicellular eukaryote Toxoplasma gondii, the accurate trafficking of proteins through the endomembrane system is vital for parasite survival and pathogenicity. However, the mechanisms of cargo transport in this parasite are poorly understood. In this study, we fluorescently labeled multiple endomembrane organelles and imaged their movements using live cell microscopy. We demonstrate that filamentous actin and an unconventional myosin motor named TgMyoF control both the positioning of organelles in this pathway and the movement of transport vesicles throughout the parasite cytosol. This data provides new insight into the mechanisms of cargo transport in this important pathogen and expands are understanding of the biological roles of actin in the intracellular phase of the parasite’s growth cycle.


Abstract 17
Toxoplasma gondii is an obligate intracellular parasite that relies on three distinct secretory 18 organelles, the micronemes, rhoptries and dense granules, for parasite survival and disease 19 pathogenesis. Secretory proteins destined for these organelles are synthesized in the 20 endoplasmic reticulum (ER) and sequentially trafficked through a highly polarized 21 endomembrane network that consists of the Golgi and multiple post-Golgi compartments. 22 Currently, little is known about how the parasite cytoskeleton controls the positioning of the 23 organelles in this pathway, or how vesicular cargo is trafficked between organelles. Here we 24 show that F-actin and an unconventional myosin motor, TgMyoF, control the dynamics and 25 organization of the organelles in the secretory pathway, specifically ER tubule movement, apical 26 positioning of the Golgi and post-Golgi compartments, apical positioning of the rhoptries and 27 finally, the directed transport of Rab6-positive and Rop1-positive vesicles. Thus, this study 28 identifies TgMyoF and actin as the key cytoskeletal components that organize the 29 endomembrane system in T. gondii. 30 31

32
Endomembrane trafficking is a vital cellular process in all eukaryotic cells. In most cases the 33 molecular motors myosin, kinesin and dynein transport cargo including vesicles, organelles and 34 transcripts along actin and microtubule filaments in a manner analogous to a train moving on its 35 tracks. For the unicellular eukaryote Toxoplasma gondii, the accurate trafficking of proteins 36 through the endomembrane system is vital for parasite survival and pathogenicity. However, the 37 mechanisms of cargo transport in this parasite are poorly understood. In this study, we 38 fluorescently labeled multiple endomembrane organelles and imaged their movements using 39 live cell microscopy. We demonstrate that filamentous actin and an unconventional myosin 40 motor named TgMyoF control both the positioning of organelles in this pathway and the 41 movement of transport vesicles throughout the parasite cytosol. This data provides new insight 42 into the mechanisms of cargo transport in this important pathogen and expands are 43 understanding of the biological roles of actin in the intracellular phase of the parasite's growth 44 cycle. 45

Introduction 46
Toxoplasma gondii is a member of the phylum Apicomplexa, which contains over 5000 47 species of parasites that cause substantial morbidity and mortality worldwide (1,2). T. gondii can 48 cause life-threatening disease in immunocompromised individuals and when infection occurs in 49 utero (3-5). Additionally, T. gondii is estimated to cause persistent life-long infection in 10-70% 50 of the world's population depending on geographic location (6). 51 T. gondii is an obligate intracellular parasite, and thus parasite survival and disease 52 pathogenesis rely on the parasite's lytic cycle involving host cell invasion, parasite replication 53 within a specialized vacuole termed the parasitophorous vacuole (PV), and host cell egress that 54 results in destruction of the infected cells (reviewed by (7)). To complete this lytic cycle, the 55 parasite relies on three specialized secretory organelles, the micronemes, rhoptries, and dense 56 granules. Micronemes are small vesicles that are localized predominately at the parasite's 57 apical end (8) and are important for parasite motility, attachment and initiating invasion 58 (Reviewed by (9)). Rhoptries are larger club shaped organelles that contain two sub-sets of 59 proteins (rhoptry bulb proteins (ROPs) and rhoptry neck proteins (RONs)) categorized based on 60 their functions and location within the rhoptry (10). After initial attachment to the host cell, RONs 61 contribute to the formation of the moving junction, a ring structure that aids in the propulsion of 62 the parasite into the host cell (11-15). Once invasion is initiated ROPs and dense granule 63 proteins (GRAs) are secreted into the host cell where they control the organization and structure 64 of the PV, and modulate host gene expression and immune response pathways (10,16-18). 65 Secretory proteins destined for these distinct organelles are synthesized in the 66 endoplasmic reticulum (ER) and must sequentially traverse multiple intermediate compartments 67 within T. gondii's highly polarized endomembrane system before ultimately arriving at their final 68 destination. Newly synthesized proteins are first trafficked to the Golgi which is located adjacent 69 to the nucleus at the parasite's apical end. Dense granules are formed from post-Golgi vesicles 70 (19,20) while proteins destined for the micronemes and rhoptries are trafficked through one or 71 more post-Golgi compartments (PGCs), although the exact route taken by each secretory 72 protein has not been elucidated and the function of each PGC has not been fully defined (Fig.  73 1a). Two of these compartments are marked by Rab5a and Rab7 and are referred to as the 74 endosome-like compartments, as these proteins are markers of the early and late endosomes in 75 higher eukaryotes (21 The goal of this study was to investigate the role of actin and TgMyoF in regulating 119 vesicle trafficking and organelle positioning within the endomembrane pathway. Our data 120 demonstrates that both of these proteins are required for the apical positioning and morphology 121 of the Golgi and post-Golgi compartments, ER tubule movement, and transport of Rab6-positive 122 and Rop1-positive vesicles. These results indicate that this acto-myosin system is vital for 123 controlling the organization of the endomembrane system in T. gondii and uncovers new 124 biological roles of actin in the intracellular phase of the parasite's growth cycle. 125

127
To investigate how the morphology and dynamics of the PGCs is controlled by the T. gondii 128 cytoskeleton, we fluorescently labeled the Rab5a and Rab6 PGCs by expressing 129 NeonGreenFP-Rab5a (referred to subsequently as Neon-Rab5a) and EmeraldFP-Rab6 130 (EmGFP-Rab6) along with Grasp55-mCherryFP, a marker of the cis-Golgi (23,24,51,52). As 131 expected, Neon-Rab5 did not co-localize with Grasp55-mCherryFP and was found in a distinct 132 compartment adjacent to the Golgi (23) ( Fig.1B; upper panel). Surprisingly, Rab6 also did not 133 localize to the Golgi as previously reported (24) but also localized to a compartment apical to 134 the Golgi (  vesicles revealed velocities and run-lengths of 0.92±0.01µm/s and 1.6±0.03µm respectively 162 (Table S1). 163 Given the similarities between Rab6(+) vesicle movement and actin dependent dense granule 164 movement (39) we sought to determine if actin was required for Rab6(+) vesicle transport. We 165 treated intracellular parasites expressing EmGFP-Rab6 with cytochalasin D (CD) for 60 minutes 166 to depolymerize F-actin before commencement of imaging (Video 2). We observed two 167 phenotypes associated with the loss of F-actin: First, the main Rab6(+) compartment lost its 168 apical localization and became fragmented and distributed throughout the parasite cytosol ( Fig.  169 3B). Second, the dynamic tubular morphology of the EmGFP-Rab6 compartment is lost and the 170 compartment remained static throughout the 60-second imaging period. Vesicle formation from 171 the Rab6 compartment was perturbed in CD treated parasites ( Fig. 3B; right panel). The 172 number of Rab6(+) vesicles in the parasite cytosol decreases from an average of 8±0.5 in 173 control parasites (RH parasites treated with DMSO) to 3±0.3 after CD treatment (Fig. 3C).

174
Similarly, the number of directed runs exhibited by Rab6(+) vesicles decreased from 6±0.45 in 175 control to less than 1 after CD treatment (Fig. 3D). 176 Creation of conditional TgMyoF-knockdown parasite line.

177
Since dense granule transport is dependent on F-actin and TgMyoF (39), we investigated if 178 TgMyoF was also required for Rab6(+) vesicle transport or compartment dynamics. We 179 previously used an inducible Cre-LoxP system to create a parasite line deficient in functional 180 TgMyoF, however depletion of TgMyoF protein levels after TgMyoF gene excision took ~48 181 hours (39,55). Thus, we created an inducible TgMyoF knockdown (KD) parasite line using the 182 auxin-inducible degradation system where TgMyoF protein levels could be rapidly degraded 183 (56,57) (Fig. 4A). The endogenous TgMyoF gene was C-terminally tagged with an AID-HA 184 epitope to create a TgMyoF-mAID-HA parasite line (referred to subsequently as TgMyoF-AID).

185
PCR of genomic DNA and western blot was used to confirm the accurate integration of this 186 construct ( Fig. 4B and Fig. S1). Treatment of TgMyoF-AID parasites with indole-3-acetic acid 187 (IAA) for four hours resulted in depletion of TgMyoF to undetectable levels ( Golgi, yet we have no mechanistic insight into how the position of these organelles is 196 maintained. To determine if TgMyoF is required for the apical positioning of the PGCs, we 197 ectopically expressed markers for these compartments, specifically EmGFP-Rab6, Neon-198 Rab5a, Neon-Rab7, Syn6-GFP and DrpB-GFP in TgMyoF-AID parasites treated with ethanol 199 (EtOH; control) or IAA (to deplete TgMyoF) for 18 hours before fixation. As expected in control 200 parasites, these compartments were positioned at the apical end of the parasite ( Fig. 5A left 201 panels). After TgMyoF depletion however, the Rab5a, Rab6, DrpB and Syn6 compartments 202 became fragmented and were found throughout the cytosol ( Fig. 5A; right panels). For each 203 protein, we quantified the number of compartments per parasite and found a statistically 204 significant increase after TgMyoF depletion in each case (Fig. 5B). In the case of Rab7, this 205 protein had a diffuse localization in the cytosol after TgMyoF depletion (Fig. 5A). Since Rab6 206 and Syn6 are localized to the same compartment in control parasites, we sought to determine if 207 these proteins remained colocalized after compartment fragmentation. In TgMyoF deficient 208 parasites expressing AppleFP-Rab6 and Syn6-GFP, we found that the co-localization between 209 these proteins remained after compartment fragmentation (Fig. 5C). In addition, 70% of 210 parasites also contained Rab6+/Syn6-vesicles (Fig. 5C, lower panel, white arrows). 211

TgMyoF plays a role in Rab6 vesicle transport 212
To assess the role of TgMyoF in the Rab6 compartment and Rab6(+) vesicle dynamics, 213 TgMyoF-AID parasites expressing EmGFP-Rab6 were treated with either EtOH or IAA for 18 214 hours before live cell imaging. Similar to what was observed after actin depolymerization, loss of 215 TgMyoF resulted in fragmentation of the Rab6 compartment ( Although the number of directed runs was significantly reduced, vesicle velocities were the 220 same in the absence of TgMyoF compared with controls ( Fig. 6D) (Table S1). 221 These data combined with published work demonstrate that TgMyoF and actin are required for 222 dense granule and Rab6+ vesicle transport (39), apical positioning of the post-golgi 223 compartments, and inheritance of the apicoplast (44). All of these organelles are part of the 224 endomembrane network in T. gondii. Therefore, we wanted to determine if other organelles in 225 this pathway, namely the ER, the Golgi, the micronemes and the rhoptries, relied on this acto-226 myosin system for their dynamics and/or morphology. 227

228
The ER is a large membrane bound organelle that has three distinct functional domains, the 229 nuclear envelope and peripheral tubules and peripheral cisternae which form an extensive 230 and continuous network in the parasite cytosol (52,58). Live cell imaging of parasites with a 231 fluorescently labeled ER, achieved by expression eGFP-SAG1∆GPI-HDEL (53) (referred to 232 subsequently as GFP-HDEL) reveals that the ER tubules are highly dynamic and undergo 233 continuous reorganization (Video 4). ER tubule rearrangements are clearly evident when the 234 first frame of the movie was overlaid with images taken after 5, 10 and 15 seconds of imaging 235 (Fig 7A;  contained a single Golgi, 52% contained two Golgi and 21% contained three Golgi ( Fig. 8B and  250 8D). Similarly, actin depolymerization with cytochalasin D also resulted in an increased number 251 of Golgi per parasite ( Fig. 8C and 8D). While the majority of Golgi remain closely associated 252 with the nucleus after TgMyoF-knockdown or CD treatment, the apical positioning of the Golgi 253 was lost. After TgMyoF depletion or actin depolymerization, 52% and 40% of parasites 254 respectively contained Golgi in both the apical and basal ends of the parasites compared to just 255 5% of control parasites (Fig 8E). 256 Since the Golgi in T. gondii divides by binary fission during cell division (51,63), the increased 257 number of Golgi observed after the loss of F-actin and TgMyoF could be due to uncoupling of 258 the Golgi division cycle from the cell cycle. To determine if this was the case, we first 259 determined the number of Golgi per parasite after 4 hours of IAA treatment, the time at which 260 TgMyoF is completely depleted and a length of time shorter than one parasite division cycle. 261 After 4 hours and 15 hours of IAA treatment, the number of Golgi per parasite is 262 indistinguishable indicating that Golgi fragmentation occurs quickly upon TgMyoF depletion but 263 does not continue to fragment with extended IAA treatment times ( Fig 8D). Next, we 264 investigated if loss of TgMyoF affected the number of centrosomes per parasite as it had 265 previously been demonstrated that centrosome duplication and Golgi fission are the first events 266 to take place at the start of parasite division (52) (Fig. 8A). TgMyoF-AID parasites were 267 transfected with Grasp55-mCherry and centrin1-GFP (64) to label the Golgi and centrosomes 268 respectively, and then treated with EtOH or IAA for 15 hours. In control parasites, 73% of 269 parasites contained one Golgi and one centrin (1G/1C) while the remaining ~30% of parasites 270 were at various stages of division and contained either one Golgi and twp centrin (1G/2C) or two 271 Golgi and two centrin (2G/2C) ( Fig. 7F and 7G). In contrast, only 20% of TgMyoF-KD parasites 272 contained one Golgi and one centrin (1G/1C), while 41% contained two Golgi and one centrin 273 (2G/1C), compared to just 3% of controls. 7% and 16% of IAA treated parasites contained one 274 centrin and three Golgi (1C/3G) or two centrin and three Golgi (2C/3G) respectively, which were 275 phenotypes that were never observed in the control parasites ( Fig. 8F and 8G). The number of 276 centrosomes per parasite remained unchanged in TgMyoF-KD parasites compared to controls 277 (Fig. 8H). Thus, we conclude that TgMyoF and actin are important for controlling both Golgi (visualized using an anti-AMA1 antibody) compared to 5% of controls ( Fig. 9A and 9B). Despite 289 the accumulation of these organelles in the residual body, the apical positioning of the 290 micronemes was not affected in TgMyoF-knockdown parasites when assessed by IFA (Fig. 9A).

291
In contrast, there appeared to be an increased Rop1 fluorescence throughout the parasite 292 cytosol ( Fig. 9A; magenta arrow). To further investigate the effects of TgMyoF depletion on 293 rhoptry dynamics, we expressed Rop1-NeonFP in TgMyoF-AID parasites treated for 18 hours 294 with either EtOH or IAA and imaged the parasites using live cell microscopy. In control 295 parasite's, the rhoptries were localized as expected at the apical end. The rhoptries were 296 surprisingly dynamic, and like the Rab6 compartment, were constantly rearranged (Fig. 9C,  297 inset; Video 5). In addition, Rop1 vesicles were observed throughout the parasite and exhibited 298 directed, motor-driven motion (Video 6). Upon TgMyoF knockdown we observed a large 299 decrease in the number of directed runs exhibited by Rop1-NeonFP vesicles from 11±1.1 in 300 control parasites to 1.4±0.2 after IAA treatment, even though the total number of Rop1 vesicles 301 was not statistically different between control and TgMyoF depleted parasites (6.25±0.4 and 302 9.1±0.6 in EtOH and IAA treated cells respectively) ( Fig. 9E and 9F). To further investigate the 303 effect of TgMyoF knockdown on the apical positioning of the rhoptries, we compared Rop1-304 NeonFP fluorescence intensity at the apical and basal ends in control and TgMyoF depleted 305 parasites. In control parasites the apical:basal ratio was 5.8±0.7, indicating a strong enrichment 306 of Rop1-NeonFP at the parasite's apical end. By comparison the apical:basal ratio in IAA 307 treated parasites was only 2.1±0.15. While Rop1 is still enriched at the parasites apical end, 308 there is an increase in Rop1-NeonFP fluorescence at the basal end of TgMyoF-KD parasites 309 compared to controls (Fig. 9G). 310 number, many Golgi failed to maintain their position at the parasite's apical end and were 342 observed associated with the lateral of basal sides of the nucleus (Fig. 8B and 8E). Collectively, 343

Discussion
our data demonstrate that actin and TgMyoF control both Golgi number and positioning. 344 After exiting the Golgi, proteins destined for the micronemes and rhoptries are trafficked to one 345 or more post-Golgi compartments. We have identified Rab6 as a new marker of the syntaxin 6 346 compartment, a protein that plays a role in retrograde trafficking from the Rab5a/Rab7 347 compartments to the Golgi (25). Our data builds upon previously published results indicating 348 that Rab6 localizes to cytosolic vesicles and a compartment at the apical end of the nucleus, 349 thought previously to be the Golgi since TgRab6 localizes to the Golgi when heterologously 350 expressed in mammalian cells (25). However, we find no evidence that Rab6 is found in the cis-351 or mid-Golgi. Additionally, while we also observe Rab6(+) vesicles throughout the parasite, 352 Rab6 does not colocalize with a marker for the dense granules, indicating these are distinct 353 vesicle types. This data is consistent with a recent report demonstrating that Rab11a is found on 354 the surface of dense granules and required for their secretion (71). The discrepancy between 355 our results and previously published data, is likely due to the unavailability of organelle markers 356 at the time of the previous publication (25). 357 The Rab6 compartments is dynamic and has tubular morphology that undergoes continuous 358 rearrangement, we observed new tubules growing from the compartment while others retract. 359 Vesicles were observed to bud from the tip of tubules and subsequently exhibited directed actin 360 and TgMyoF dependent movement. This dynamic morphology closely mirrors the dynamics of cytoskeleton, or that the physical connections between these compartments, that maintain the 372 compartments in close proximity at the apical end, are formed in an actin dependent manner. 373 Future studies are needed to further identify the molecular players that control the associations 374 between these compartments. 375 cytosol. In the absence of TgMyoF, there was a significant decrease in the number of directed 386 runs exhibited by Rop1-NeonFP vesicles. Currently, we do not know the function or subcellular 387 destination of these rhoptry derived vesicles and further work will be required to elucidate their 388 biological role. These results suggest that although the rhoptries are formed once per cell cycle, 389 acto-myosin dependent trafficking of proteins to or from the rhoptries occurs continuously. 390 There is an incomplete understanding of the mechanisms by which the mature rhoptries 391 are anchored to the apical end of the parasite. In the absence of TgARO1, a membrane 392 associated rhoptry protein, the apical positioning of the rhoptries was lost completely (78). This 393 contrasts with the TgMyoF knockdown phenotypes where there is increased accumulation of 394 rhoptries throughout the parasite cytosol, shown by an increase in Rop1 fluorescence at the 395 parasite's basal end (Fig. 9G), even though most parasites retain at least some intact rhoptries 396 at the apical end. Thus, the TgARO1 knockdown parasites have a more severe rhoptry 397 localization defect than TgMyoF knockdown parasites. Although TgMyoF was shown to interact 398 indirectly with TgARO1 (78), the differences in the severity of these phenotypes suggest that 399 TgMyoF is not required for TgARO1 anchoring activity. Our data suggests that TgMyoF is 400 required for movement of immature rhoptries to the apical tip but is not required for TgARO1-401 dependent anchoring once the organelles have reached their destination. 402 This study demonstrates that TgMyoF controls the dynamics, positioning, and movement of a 403 wide array of organelles in the endomembrane pathway in T. gondii. Future studies will be 404 important to elucidate the mechanism by which this single molecular motor controls the 405 movement of such a wide array of membranous cargos. Given the structural similarity between 406 TgMyoF and the well characterized cargo transporter myosin V, we previously hypothesized 407 that TgMyoF bound dense granules via its C-terminal WD40 domain and transported cargo by 408 moving processively on filamentous actin (39). The large number of membrane-bound 409 organelles whose movement is dependent on TgMyoF makes elucidating TgMyoFs mechanism 410 of action even more pertinent.

426
To determine the effect of actin depolymerization on endomembrane organization and 427 dynamics, transfected parasites were grown for 15-18 hours in confluent HFF monolayers, 428 treated for 60 minutes with either 2µM cytochalasin D or equivalent volume of DMSO before live 429 cell imaging as described below. To deplete TgMyoF, TgMyoF-AID parasites were treated with 430 a final concentration of 500µM IAA, diluted 1:1000 from a 500mM stock made in 100% EtOH. 431 For live cell imaging experiments treated time ranged from 15-18 hours. For western blot, 432 treatment time was varied as indicated in figure 3. 433

Construction of expression plasmids 434
A list of plasmids, primers and gene accession numbers used in this study can be found in 435  Tables S2, S3 and Table S5 respectively. 436

Creation of pTKOII-MyoF-mAID-HA:
437 pTKOII-MyoF-EmeraldGFP (EmGFP) (39) was digested with BglII and AflII to remove the 438 EmGFP coding sequence. AID-HA was amplified by PCR using the AID-HA ultramer as a 439 template and primer pairs AID-HA F and AID-HA R. Plasmid backbone and the PCR product 440 was gel purified and ligated via Gibson assembly using NEBuilder HiFi DNA assembly master 441 mix as per manufacturer's instructions (New England BioLabs; Ipswich, MA). Plasmids were 442 transfected into NEB5α bacteria and positive clones screened by PCR and verified by Sanger 443 sequencing. 444 Creation of pmin-eGFP-Rab6-Ble: 445 To create parental plasmid pmin-eGFP-mCherry-Ble, pmin-eGFP-mCherry was digested with 446 KpnI and XbaI. pGra1-Ble-SAG1-3'UTR plasmid (80) was digested with KpnI and XhoI to 447 remove the ble expression cassette. A fill-in reaction was performed to produce blunt ends by 448 incubating plasmids with 100µM dNTPs and T4 DNA polymerase at 12°C for 15 minutes. 449 Digested plasmids were gel purified and ligated together using T4 DNA ligase (New England 450 Biolabs). Plasmids were transfected into NEB5α bacteria and positive clones screened by PCR 451 and verified by Sanger sequencing. To create pmin-EmGFP-Rab6-ble, pmin-eGFP-mCherry-Ble 452 was digested with NheI and AflII to remove eGFP-mCherry sequence. EmGFP was amplified by 453 PCR with EmGFP-R6F and EmGFP-R6R primer pairs using pTKOII-MyoF-EmGFP as a 454 template. Rab6 coding sequence was amplified by PCR using RH cDNA as a template and 455 Rab6F and Rab6R primers. Digested plasmid backbone, EmGFP and Rab6 PCR products were 456 gel purified and ligated using NEBuilder HiFi DNA assembly master mix as per manufacturer's 457 instructions (New England BioLabs). Plasmids were transfected into NEB5α bacteria and 458 positive clones screened by PCR and verified by Sanger sequencing. 459

468
Rab5a and Rab7 coding sequences were amplified by PCR using primers sets Rab5F/R, 469 Rab7F/R and pTg-HARab5a and pTg-HARab7 (27,81) as templates. NeonGreen was amplified 470 by PCR using Neon-R5F/Neon-R5R, Neon-R7F/Neon-R7R primer sets and Ty1-471 NeonGreenPave as a template. pmin-EmGFP-Rab6 plasmid was digested with NheI and AflII 472 to remove EmGfP-Rab6 coding sequence. Plasmid backbones and PCR products were gel 473 purified and annealed using Gibson assembly with NEBuilder HiFi DNA assembly master mix as 474 per manufacturer's instructions. Plasmids were transfected into NEB5α bacteria and positive 475 clones screened by colony PCR and verified by Sanger sequencing. 476 Creation of ptub-Rop1-NeonGreenFP. 477 ptub-SAG1-∆GPI-GFP plasmid was digested with NheI and AflII to remove SAG1-GFP coding 478 sequence. Rop1-GFP coding sequence was amplified using Rop1 F/R primer pairs and RH 479 cDNA as a template. NeonGreen was amplified by PCR using NeonRop1F and NeonRop1R 480 primer pairs. Plasmid backbones and PCR products were gel purified and annealed using 481 Gibson assembly with NEBuilder HiFi DNA assembly master mix as per manufacturer's 482 instructions. Plasmids were transfected into NEB5α bacteria and positive clones screened by 483 PCR and verified by Sanger sequencing. 484 Creation of TgMyoF-AID parasite line.

485
The pTKO2_MyoF_mAID-HA plasmid was linearized with SphI and 25µg was transfected into 486 1x10 7 ∆Ku80:∆HXGPRT:Flag-Tir1 parental parasites (a gift from Dr. David Sibley, Washington 487 University (56)). Parasites were selected using mycophenolic acid (MPA) (25 µg/ml) and 488 xanthine (50 µg/ml) until approximately 70% of the parasites were HA positive. Clonal parasite 489 lines were obtained by limited dilution into a 96 well plate. After 7 days of growth, wells 490 containing a single plaque were selected for further analysis. All HA positive clones were 491 amplified in a 6 well plate and genomic DNA isolated using Qiagen DNAeasy blood and tissue 492 kit as per manufactures instructions (Qiagen, Germantown, MD) (Cat #69504). Genomic DNA 493 was analyzed for correct insertion of the pTKO2_MyoF_mAID HA plasmid into the TgMyoF 494 genomic locus by PCR using primers listed in Table S3 as outlined in Fig. S1. 495

518
Growth media was replaced with Fluorobrite DMEM (ThermoFisher; Cat# A19867) containing 519 1% FBS and 1x antimycotic/antibiotic pre-warmed to 37˚C. Images were acquired on a GE 520 Healthcare DeltaVision Elite microscope system built on an Olympus base with a 100x 1.39 NA 521 objective in an environmental chamber heated to 37˚C. This system is equipped with a scientific 522 CMOS camera and DV Insight solid state illumination module. Image acquisition speeds were 523 determined on a case-by-case basis as noted in the video legends. 524

525
Parasites were fixed with freshly made 4% paraformaldehyde (Electron microscopy sciences, 526 Hatfield, PA; Cat# 15714) in 1xPBS (ThermoFisher; Cat# 18912-014) for 15 minutes at RT. 527 Cells were washed three times in 1xPBS and permeabilized in 0.25% TX-100 diluted in 1xPBS 528 for 10 minutes at room temperature before washing three times in 1xPBS. Cells were blocked in 529 2% BSA-1XPBS for 15 minutes before antibody incubations. All antibodies were diluted in 0.5% 530 BSA-1xPBS at the concentrations indicated in Table S4. DNA was stained with 10µM DAPI 531 diluted in 1xPBS for 10 minutes and then washed three times in 1xPBS. Cells in mattek dishes 532 were either imaged immediately or stored in 1xPBS at 4˚C. Coverslips were mounted onto 533 slides using Prolong Gold anti-fade reagent (ThermoFisher; Cat # P36930) and allowed to dry 534 overnight before imaging. 535

536
Vesicle tracking and counting.

537
Vesicle tracking was performed using MtrackJ plug-in in Fiji (National Institutes of Health) as 538 previously described (82). Line scan analysis was performed using the "plot profile" tool in Fiji.

539
Kymographs were made using Fiji plugin KymographBuilder. Fiji plugin cell counter was used to 540 quantify the number of vesicles/parasites. Statistical significance was determined using students 541 t-test. 542

543
To count number of PGC "objects" in each parasite in control and TgMyoF knockdown 544 parasites, transfected parasites were fixed and stained with the anti-IMC1 antibody and DAPI 545 and z-stack images were acquired using Deltavision elite imaging system. For each PGC 546 image, maximum intensity projection of Z-stacked images were converted to binary images 547 using Fiji and particles counted using the "analyze particles" tool. IMC1 staining was used to 548 determine number of parasites/vacuole. Statistical significance was determined using students 549 t-test. 550 Quantification of Golgi and centrosome number.

551
To quantify the number of Golgi and centrosome per parasite, TgMyoF-AID parasites were 552 transfected with Grasp55-GFP plasmid alone or Grasp55-mCherry and Centrin1-GFP plasmids, 553 grown in confluent HFF monolayers overnight and treated with either EtOH or IAA for the final 4-554 or 15-hours of growth before fixation and processing for immunofluorescence. Number of Golgi 555 per parasite and number of centrosomes per parasites were counted manually using the cell 556 count tool in Fiji. Statistical significance was determined using students t-test. 557

558
To quantify the number of parasites with microneme and rhoptry proteins in the residual body, 559 TgMyoF parasites were transfected with Rop1-NeonFP plasmid, and were grown for 15 hours in 560 either EtOH or IAA before fixation and immunocytochemistry with an anti-AMA1 antibody (Table  561 S4). The number of vacuoles containing rhoptries or micronemes in the residual body were 562 manually counted. N = 57 vacuoles for AMA-1 quantitation, N=77 vacuoles for Rop-1 563 quantitation from two independent experiments. Statistical significance was determined using 564 students t-test. 565 Loss of rhoptry positioning at the parasites' apical end. 566 To quantify Rop1 localization at the apical or basal ends of the parasite, Rop1-NeonFP was 567 transiently expressed in TgMyoF-AID parasites treated with EtOH or IAA for 15 hours. Parasites 568 were imaged live as described above. Using the first frame of each movie, the apical half and 569 the basal half of each parasite was outlined manually and the mean fluorescent intensity of 570 Rop1 in the apical and basal ends of the parasites were calculated using Fiji. The apical to 571 basal mean fluorescence intensity was calculated for each parasite. A ratio of 1 indicates that 572 Rop1 is evenly distributed in the apical and basal ends. A ratio >1 indicates more rop1 is 573 localized in the apical end than the basal end and a ratio <1 indicates more rop1 in the basal 574 end compared to the apical end. Rop1-NeonFP fluorescence in the residual body was excluded 575 from this analysis. Statistical significance was determined using students t-test. 576

581
This work was funded by National Institutes of Health R21AI121885 awarded to ATH and the 582 University of Connecticut Research excellence program awarded to ATH. The funders had no 583 role in study design, data collection and interpretation, the decision to submit the work for 584 publication or manuscript preparation. The authors declare that no competing interests exist. 585 586 Acknowledgements. 587 We thank members of the Heaslip Lab, Dr. Ken Campellone and members of the Campellone 588 lab (University of Connecticut) for helpful discussion during the course of these experiments. 589 We thank our colleagues for sharing reagents: Dr. Gary Ward (University of Vermont) for the Grasp55-mcherry (Golgi marker; yellow) and NeonFP-Rab5 (magenta; upper panel) or 832 (EmGFP-Rab6 (magenta; lower panel) were fixed and stained with DAPI (cyan). (C) RH 833 parasites expressing AppleFP-Rab6 (magenta; all panels) with markers of the endomembrane 834 system as indicated (yellow). Nuclei were stained with DAPI (cyan). Panels i-iii are deconvolved 835 images of a single focal plane from fixed intracellular parasites. Panels iv-v are a single focal 836 plane from live intracellular parasites expressing AppleFP-Rab6 (magenta) and SAG1∆GPI-837 GFP (marker for the dense granules) or Syn6-GFP respectively. Panel vi is a single extracellular 838 parasite expressing AppleFP-Rab6 that fixed and stained with an anti-CPL antibody and DAPI. 839 Scale bar = 5µm. Inset scale bar in panel iv is 1µm. 840