Phosphorylation of a Toxoplasma gondii tyrosine transporter by calcium-dependent kinase 3 is important for parasite fitness

Toxoplasma gondii parasites rapidly exit their host cell when exposed to calcium ionophores. The calcium-dependent protein kinase 3 (TgCDPK3) was previously identified as a key mediator in this process, as TgCDPK3 knockout (Δcdpk3) parasites fail to egress in a timely manner. Phosphoproteomic analysis comparing WT with Δcdpk3 parasites revealed changes in the TgCDPK3-dependent phosphoproteome that included proteins important for regulating motility, but also metabolic enzymes, indicating that TgCDPK3 controls processes beyond egress. Here we have investigated a predicted direct target of TgCDPK3, a putative transporter of the major facilitator superfamily (MFS) and show that it is rapidly phosphorylated after induction of calcium signalling. Conditional knockout (KO) of the transporter reveals an essential role in the lytic cycle during intracellular growth with a transcriptome signature of amino acid-starved parasites. Using a combination of metabolomics and heterologous expression, we confirmed a primary role in tyrosine import. Complementation with phosphorylation site mutants shows that phosphorylation of serine 56 (S56) by TgCDPK3 gives the parasites a growth benefit in competition assays. Collectively, these findings validate an important, albeit non-essential role for TgCDPK3 in the regulation of metabolic processes, in addition to motility. Author summary Toxoplasma gondii is an obligate intracellular parasite. To survive and spread throughout the host it must repeatedly infect, replicate within and exit, host cells. These recurring cycles of infection and egress rely on signalling pathways that allow the parasites to sense and respond rapidly to their environment. While some key kinases and secondary messengers within these pathways have been identified, functional analysis of non-kinases has been very limited. This is especially true for candidates that are not predicted to play a role in active motility or are not known to function in established signalling pathways. Here we have followed up on an unexpected target of the T. gondii calcium-dependent kinase 3 (TgCDPK3), a plant-like calcium dependent kinase, that was previously shown to play an important role in calcium-mediated exit from the host cell. We show that, in addition to controlling motility of the parasite (as previously shown), TgCDPK3 phosphorylates an essential tyrosine transporter in the plasma membrane. Mutational analysis of the phosphorylation sites demonstrates an important role in maintaining parasite fitness, thus demonstrating that TgCDPK3 plays a pleiotropic role in controlling both egress and metabolism.


Abstract 22
Toxoplasma gondii parasites rapidly exit their host cell when exposed to calcium 23 ionophores. The calcium-dependent protein kinase 3 (TgCDPK3) was previously 24 identified as a key mediator in this process, as TgCDPK3 knockout (∆cdpk3) 25 parasites fail to egress in a timely manner. Phosphoproteomic analysis comparing 26 WT with ∆cdpk3 parasites revealed changes in the TgCDPK3-dependent 27 phosphoproteome that included proteins important for regulating motility, but 28 also metabolic enzymes, indicating that TgCDPK3 controls processes beyond 29 egress. Here we have investigated a predicted direct target of TgCDPK3, a putative 30 transporter of the major facilitator superfamily (MFS) and show that it is rapidly 31 phosphorylated after induction of calcium signalling. Conditional knockout (KO) 32 of the transporter reveals an essential role in the lytic cycle during intracellular 33 growth with a transcriptome signature of amino acid-starved parasites. Using a 34 combination of metabolomics and heterologous expression, we confirmed a 35 primary role in tyrosine import. Complementation with phosphorylation site 36 mutants shows that phosphorylation of serine 56 (S56) by TgCDPK3 gives the 37 parasites a growth benefit in competition assays. Collectively, these findings 38 validate an important, albeit non-essential role for TgCDPK3 in the regulation of 39 metabolic processes, in addition to motility. The fast growing tachyzoite stage of the protozoan parasite Toxoplasma gondii requires 61 cycles of host cell invasion, replication, and lysis for its successful proliferation within 62 the host. Each step of this lytic cycle involves tightly regulated signalling pathways, the 63 intricacies of which remain largely unknown. Paramount to parasite survival is the 64 ability to sense and respond to changes in the environment for which the divalent 65 calcium ion (Ca 2+ ) acts as an important secondary messenger (1). Changes in free 66 intracellular [Ca 2+ ] i levels, via release of Ca 2+ from organellar Ca 2+ stores, can be 67 induced by the addition of Ca 2+ ionophores, such as A23187 or phosphodiesterase 68 inhibitors (2,3). Ca 2+ flux regulates key processes including secretion of micronemes 69 prior to host cell entry (4), parasite motility (5), and host cell egress (6) and invasion 70 (7). Inversely, these processes can all be inhibited by Ca 2+ immobilisers or chelators, 71 such as BAPTA-AM (5,8-10). Ca 2+ release leads to the activation of Ca 2+ binding 72 proteins such as calmodulins, calcineurin B-like kinases and calcium-dependent protein 73 kinases (CDPKs). T. gondii calcium-dependent protein kinase 3 (TgCDPK3), for 74 example, has been implicated in the regulation of ionophore induced egress, IIE (i.e. 75 the rapid exit of tachyzoites from a host cell on addition of ionophore) and ionophore 76 induced death, IID (i.e. the loss of infectivity of EC parasites after prolonged exposure 77 to ionophore) (9). TgCDPK3 KO (∆cdpk3) (11), mutants (9), and chemically inhibited 78  TgCDPK3-mediated phosphorylation. We show that ApiAT5-3 is rapidly 106 phosphorylated at serine 56 (S56) during the first minute of induced egress. Using a 107 conditional KO approach, we show that ApiAT5-3 is essential, and that deletion leads 108 to a delayed death phenotype that is accompanied by a transcriptional response relating 109 to translational stress. In growth competition assays performed with parasite lines that 110 rely on phosphomutants or phosphomimetics, we show that phosphorylation of S56 111 appears to be important, but not essential, for parasite fitness. Finally, using a 112 combination of metabolic analysis and heterologous expression in Xenopus laevis 113 oocytes we confirm that ApiAT5-3 transports tyrosine but has only limited capacity to 114 transport BCAAs. This data confirms that TgCDPK3 phosphorylates several targets in 115 its vicinity, controlling diverse processes at the plasma membrane and thus contributing 116 to a range of biological processes in the parasite. 117

119
ApiAT5-3 is located at the parasite periphery and phosphorylated during ionophore 120 induced egress in a TgCDPK3-dependent manner. 121 ApiAT5-3 was previously identified to be phosphorylated at Serine 56 in a 122 potentially allowing for direct interaction with TgCDPK3, which also localises to 132 the plasma membrane (Fig. 1A). ApiAT5-3 contains several phosphorylation sites 133 at its N-terminus, of which S56 was the only one previously identified as being 134 TgCDPK3-dependent (Fig. 1B, upper panel). It is entirely plausible, however, that 135 kinases other than TgCDPK3 act during egress to phosphorylate additional 136 residues on the ApiAT5-3 N-terminus. To investigate this, we queried a dataset 137 recently generated in our laboratory in which we have quantified, using tandem-138 mass-tag technology (21), phosphorylation site abundance on T. gondii proteins 139 across 4 time points (0, 15, 30 and 60 s) following ionophore-treatment (Caia 140 Dominicus, in preparation). From the ~850 phosphorylation sites that are 141 phosphorylated or de-phosphorylated during egress, we identified S56 of ApiAT5-142 3 as increasingly phosphorylated over time (Fig. 1B, lower panel). We also 143 identified several proteins already known to be more phosphorylated in response 144 to Ca 2+ signalling including MyoA, Myosin F and DrpB (17,22,23). None of the other 145 phosphorylation sites on the ApiAT5-3 N-terminus increased in phosphorylation 146 state prior to, or during egress. However, S14 of ApiAT5-3 was dephosphorylated 147 during ionophore-treatment. Collectively these data indicate that S56 is 148 phosphorylated in a TgCDPK3-dependent manner upon Ca 2+ stimulation, and that 149 a phosphatase is acting on S14 during the same period, while the other 150 phosphorylation sites appear unaffected. ApiAT5-3 depletion is predicted to have a high fitness cost (Toxo DB 7.1 (24)). 165 Accordingly, we generated a conditional KO using the dimerisable cre 166 recombinase (DiCre) strategy. We replaced the endogenous copy of apiAT5-3 with 167 a recodonised version in RH ∆ku80 DiCre parasites, by double homologous 168 recombination, using a double-guide strategy (25) ( Fig. 2A). We initially placed a 169 loxP site adjacent to the Kozac sequence of ApiAT5-3 but were unable to obtain 170 correct integration. We hypothesised that the loxP sequence might be interfering 171 with promotor elements and moved it 100 and 200 bp upstream of the predicted 172 start ATG, respectively. Both of these constructs correctly integrated into the 173 genome. Subsequent analyses were performed with the resulting ApiAT5-3_loxP, 174 with the loxP at ATG -100bp. Integration was confirmed by PCR amplification ( WT parasites, tdTomato expressing RH parasites (RH Tom) were spiked into the 205 imaging wells at a 1:1 ratio. These analyses revealed that apiAT5-3 KO does not 206 lead to early egress or an inability to invade, but rather a lack of replication, often 207 with the ability to undergo a first division, but failing to go beyond 2 208 parasites/vacuole (Fig. 2F, Movies S1A and B). 209 As we showed that ApiAT5-3 is phosphorylated directly after ionophore-210 treatment ( Fig. 1B), we postulated that it may be required for ionophore induced 211 egress. To assess this, we performed egress assays of the DMSO-and RAP-treated 212 lines in the presence of 8 µM Ca 2+ ionophore. However, there was no significant 213 difference between the KO and WT ( Collectively these data show that ApiAT5-3 is an essential protein that is required 242 for intracellular replication. Its depletion leads to a complete arrest in growth 243 which is not accompanied by a substantial stress response, but rather modest 244 signs of translational stress. 245 246 Mutation of S56 to alanine, but not a phosphomimetic leads to a reduction in fitness. 247 Having established that ApiAT5-3 is essential for the lytic cycle, we next sought to 248 examine the role of TgCDPK3-mediated phosphorylation in ApiAT5-3 function. To 249 do this, we complemented ApiAT5-3_loxP parasites with either WT ApiAT5-3, or 250 variants where S56 is mutated to alanine (S56A) or to aspartic acid (S56D). To 251 prevent possible differences in growth between the parasite lines due to 252 differential expression of the complementation constructs, we inserted each into 253 the uprt locus by double homologous recombination, under control of the 254 endogenous promoter (Fig. 4A). Complementation into the uprt locus was verified 255 by PCR (Fig. 4B). The complementation constructs also carried a C-terminal HA 256 epitope tag to verify correct trafficking to the plasma membrane. 257 Immunofluorescence displayed correct trafficking in all variants (Fig. 4C). 258 To compare fitness between the WT, the phosphomimetic (S56D), and the 259 phosphomutant (S56A) complemented lines in the absence of apiAT5-3, we 260 deleted the endogenous copy using RAP-treatment. This results in parasite strains 261 that solely rely on the complemented copy of the gene. We confirmed correct 262 excision of apiAT5-3 by virtue of YFP expression post RAP-treatment, and PCR 263 analysis (Fig. 4C, Fig. S2A). RAP-treated parasite lines were viable and allowed us 264 to isolate clones by limiting dilution, all of which restored growth in plaque assays 265 (Fig. S2B). This shows that i) complementation of apiAT5-3 by expression at the 266 uprt locus fully restores ApiAT5-3 function and ii) that neither the introduction of 267 phosphomimetics nor phosphomutants of S56 are lethal to parasite growth. 268 This is not surprising as deletion of TgCDPK3, the kinase putatively responsible 269 for ApiAT5-3 phosphorylation during egress, does not lead to a severe growth 270 phenotype. Accordingly, phosphomutants would not be expected to display 271 drastic differences in growth. We therefore performed competition assays in 272 which we compared growth of YFP expressing complementation lines that fully 273 rely on the complementation variant for growth (∆apiAT5-3 ApiAT5-3/S56A/S56D ) 274 mixed in a 1:1 ratio with their non-excised, colourless counterpart (ApiAT5-275 3 ApiAT5-3/S56A/S56D ). Using the ratio of 4',6-diamidino-2-phenylindole (DAPI) stained 276 parasites (DAPI labels the DNA of all parasites) and YFP expressing parasites (YFP 277 is expressed only in the complementation lines) we followed growth over 14 days 278 in biological triplicates. While ∆apiAT5-3 ApiAT5-3 parasites showed no difference in 279 growth compared to their WT control, ∆apiAT5-3 ApiAT5-3_S56A was reduced by 280 84.0% after 14 days (Fig. 4D). Strikingly, ∆apiAT5-3 ApiAT5-3_S56D was not 281 outcompeted and grew at similar levels to the WT control. 282 Collectively these data indicate that phosphorylation of S56 while not essential, is 283 important for intracellular growth. 284 285

ApiAT5-3 is a primary transporter of tyrosine, but not branched chain amino acids 286
The predicted homology of ApiAT5-3 to a BCAA transporter and the profound up-287 regulation of the BCKDH complex in ∆cdpk3 parasites suggested a direct role for 288 ApiAT5-3 in BCAA transport. To test this, we expressed apiAT5-3 in the 289 heterologous expression system, X. laevis oocytes. Concurrently with our study, 290 data were presented that ApiAT5-3 may be a tyrosine transporter (Giel van 291 Dooren, personal communication and pre-published in BioRx (19)). We therefore 292 tested BCAA import and replicated the tyrosine uptake capacity of ApiAT5-3 in 293 oocytes expressing WT ApiAT5-3 (Fig. 5A). Measuring unidirectional influx, we 294 observed a significant (4.0-fold) increase in the uptake of 14 C-tyrosine into 295 ApiAT5-3-expressing oocytes compared to either water-injected or uninjected 296 control oocytes under the conditions tested, consistent with results from (19). We 297 also observed moderate ApiAT5-3-dependent phenylalanine influx, but not for the 298 BCAA valine (Fig. S3A), suggesting that, while ApiAT5-3 is capable of tyrosine 299 transport, it is unlikely to be a major BCAA transporter. 300 To verify the role of ApiAT5-3 in tyrosine transport in our conditional KO 301 parasites, we measured intracellular 13 C-tyrosine levels in RAP-treated ∆apiAT5-302 3 ApiAT5-3 (WT) and ApiAT5-3_loxP (KO) parasites (74 hrs post excision), after 1 hr 303 in the presence of growth media containing 13 C-tyrosine. In an analogous manner, 304 we also measured 13 C-isoleucine uptake in order to verify if ApiAT5-3 is also a 305 BCAA transporter. ∆apiAT5-3 ApiAT5-3 was used instead of DMSO-treated ApiAT5-306 3_loxP to control for any potential effects of rapamycin on parasite metabolism. 307 This analysis verified a reduction of 13 C-labelled tyrosine uptake (40.5% 308 compared to ∆apiAT5-3 ApiAT5-3 ), but not isoleucine uptake (4.3% compared to 309 ∆apiAT5-3 ApiAT5-3 ) (Fig. 5B). We also measured the intracellular abundance of all 310 detectable amino acids when labelling with 13 C-tyrosine. We observed a reduction 311 of intracellular tyrosine abundance (63.2%) in the ∆apiAT5-3 ApiAT5-3 cells (as 312 expected), but not phenylalanine which was slightly increased in relative 313 abundance (17.45%), suggesting that while ApiAT5-3 is able to transport 314 phenylalanine in oocytes, it is not the major phenylalanine transporter in T. gondii 315 (Fig. S3B). It is important to note that our metabolome analysis was performed at 316 the end of cycle 2 after RAP-treatment, when ∆apiAT5-3 ApiAT5-3 parasites are still 317 viable but start to display a reduction of growth (Fig. 2e). Therefore, we predict 318 that low levels of ApiAT5-3 ApiAT5-3 present at this stage are responsible for the 319 residual transport of tyrosine. Interestingly, we also observed a reduction in 320 intracellular aspartate (38.5%) and glycine (28.3%) in ∆apiAT5-3 ApiAT5-3 cells (Fig.  321   S3B). Since T. gondii is not known to be auxotrophic for these amino acids we 322 reasoned that the observed death phenotype is unlikely caused by a defect in 323 glycine or aspartate import, and instead focussed our subsequent analysis on 324 tyrosine. We also observed an increase in the abundance of glutamine, valine, 325 isoleucine and proline, indicating potential wider metabolic effects. 326

327
To test whether exogenous tyrosine can complement the loss of ApiAT5-3 we 328 grew parasites in media with 5× the normal amount of tyrosine (2 mM). Despite 329 various attempts to restore normal growth or generate ∆apiAT5-3 clonal lines, we 330 could not obtain viable parasites in high tyrosine (Fig. 5C). In other organisms, 331 phenylalanine can be converted into tyrosine. Therefore, we tested whether 332 phenylalanine supplementation (2 mM) can rescue the growth phenotype of 333 apiAT5-3 KO parasites. No growth rescue could be observed (data not shown). 334 Together, these results suggest that ApiAT5-3 is the only transporter of tyrosine 335 in T. gondii and that phenylalanine cannot be readily converted into tyrosine in T. 336 gondii parasites. 337

338
To identify whether differences in phosphorylation state of ApiAT5-3 affected 339 tyrosine transport, we performed isotopic tyrosine labelling of extracellular 340 parasites using the ∆apiAT5-3 ApiAT5-3_S56A and ∆apiAT5-3 ApiAT5-3_S56D 341 phosphomutant strains. No significant differences could be observed in these 342 assays indicating that, if tyrosine import is affected, differences may be obstructed 343 by the intrinsic variability of the assay (data not shown). 344 The fitness phenotype of mutating S56 to alanine is modest in a single lytic cycle 345 and thus not predicted to have substantial impact on the import of tyrosine. 346 Because we could not measure differences in tyrosine import in our 347 phosphorylation site mutants we turned to heterologous assays where we 348 expressed ApiAT5-3 WT, S56A and S56D in X. laevis oocytes and tested tyrosine 349 uptake. Although there was a trend towards a reduction in tyrosine uptake in the 350 S56A expressing oocytes (average 19.5% reduction in S56A relative to WT 351 ApiAT5-3 expressing oocytes), this difference was not statistically significant (Fig.  352   5D). The S56D expressing oocytes display a marginal increase in tyrosine uptake 353 of 14.0%, that again was not statistically significant. 354 Collectively these data show that conditional deletion of ApiAT5-3 causes a lethal 355 reduction in tyrosine import that cannot be compensated for by amino acid 356 supplementation. We also show that while phosphorylation of ApiAT5-3 at S56 is 357 not required for tyrosine transport, mutating that site to alanine does lead to a 358 subtle decrease in tyrosine transport in a heterologous assay. 359 360

361
TgCDPK3 has previously been implicated in controlling distinct biological 362 processes such as gliding motility and metabolism. How these are linked, however, 363 has been unclear. Mutants that display only IIE and IID phenotypes have been 364 identified (9,26), arguing that TgCDPK3 may be an upstream regulator of both 365 processes. Here we show that upon activation by the Ca 2+ ionophore A23187, 366 TgCDPK3 leads to an increase in ApiAT5-3 phosphorylation at S56. This occurs at 367 the same time, and to a similar intensity, as other previously identified targets of 368 TgCDPK3 (e.g. serine 21/22 of MyoA) and other kinases involved in signalling (e.g. 369 TgCDPK1). MyoA and ApiAT5-3 are both located at, or close to, the plasma 370 membrane. It is conceivable that, upon activation, TgCDPK3 phosphorylates 371 proteins at the plasma membrane, some of which are important for motility and 372 others (e.g. transporters) that prepare the parasite for the extracellular milieu. In 373 this study we have identified a function for the phosphorylation of S56, which 374 becomes rapidly phosphorylated prior to egress in a TgCDPK3-dependent 375 manner. Interestingly, ApiAT5-3 possesses several other phosphorylation sites in 376 its N-terminus, aside from S56, that either do not change in phosphorylation state 377 or, in the case of S14, appear dephosphorylated during induced Ca 2+ signalling. 378 How S14 dephosphorylation and S56 phosphorylation are controlling ApiAT5-3 379 function requires further investigation, however it is evident that mutating S56 to 380 a non-phosphorylatable residue substantially reduces parasite fitness. 381 Phosphorylation of transporters has been shown to regulate affinity, specificity 382 and flux of cargo (27-32). Accordingly, the observed fitness cost in S56A mutants 383 could be indicative of a reduction in tyrosine import, for which the parasite is 384 auxotrophic (33). 385 While we observed a clear role for ApiAT5-3 in tyrosine transport, expression of 386 the ApiAT5_S56A phosphomutant only led to a modest reduction in tyrosine 387 import in oocytes, which was not statistically significant. One explanation could 388 be that the effect on tyrosine import is beyond the limit of detection in our assays. 389 We observed an 84.0% reduction in growth of ∆apiAT5-3 ApiAT5-3_S56A mutants 390 compared to WT parasites over 14 days. This translates into a reduced replication 391 rate of ~6.74% per 24 hrs. Tyrosine uptake assays in oocytes, as well as in 392 extracellular parasites are, for reasons of cell viability, performed in <1 hr, and 393 may therefore not pick up the subtle differences predicted to occur if tyrosine is 394 the only limiting factor in these assays. Furthermore, as indicated above, ApiAT5-395 3 shows a complex phosphorylation pattern at its N-terminus and other 396 phosphorylation sites could contribute to ApiAT5-3 regulation. Mimicking these 397 conditions in a heterologous assay, or in parasites, without prior knowledge of the 398 abundance of the phosphorylation on each of these residues becomes hard to 399 interpret and should be the subject of further studies. 400 Apart from regulating amino acid transport, phosphorylation of transporters has 401 also been shown to regulate trafficking to the surface (34-36). However, a role for 402 S56 in trafficking is less likely for two reasons: i) we did not observe any obvious 403 defects in surface translocation of the transporter in parasites and ii) TgCDPK3 404 phosphorylates S56 shortly before, or during egress, at which state the 405 transporter is already on the surface. If S56 phosphorylation was important for 406 surface translocation, we would expect this to occur at an earlier stage. However, 407 we cannot exclude the possibility that minor differences in trafficking capacity 408 impact tyrosine transport, resulting in the growth phenotype. 409 Whatever the molecular explanation for the phenotype, it is evident that 410 TgCDPK3-mediated phosphorylation of ApiAT5-3 is important for parasite fitness. appears to be imported in our apiAT5-3 KO line (Fig. 5B), this is likely due to the 427 presence of residual ApiAT5-3 protein in the plasma membrane after RAP-428 treatment. Along with our inability to rescue growth upon tyrosine 429 supplementation, we conclude that this residual tyrosine import is unlikely due to 430 an alternative transporter. Further to this, our transcriptomic analysis argues 431 against a rapid sensing and transcriptional compensation for the lack of tyrosine 432 import, so if upregulation of alternative transporters occurs, it will be a slow 433 process. Another explanation may be that slight differences in the genetic 434 background or passage history, and potential epigenetic changes in the parental 435 strains, leads to a difference in capacity for amino acid transport. There is some 436 indication that this may be the case as, in our metabolomics experiments, the For Western blot analysis, intracellular parasites were lysed in Laemmli buffer (60 511 mM Tris-HCl pH6.8, 1% SDS, 5% glycerol, 5% b-mercaptoethanol, 0.01% 512 bromophenol blue) and heated to 37°C for 30 mins prior to separation on a 10% 513 sodium dodecyl-polyacrylamide gel. Proteins were transferred onto a 514 nitrocellulose membrane prior to blocking in 3% milk, 0.1% Tween-20 PBS. HA-515 tagged ApiAT5-3 was detected using rat anti-HA (1:500), followed by goat anti-rat 516 horseradish peroxidase-conjugated secondary antibody (1:2500). 517 IFA's were performed on intracellular parasites grown in HFFs on glass coverslips. 518 1×10 5 parasites were seeded 24 hrs prior to fixation with 3% formaldehyde (FA). 519 PBS 0.1% Triton X-100 was added to the fixed cells for 10 mins prior to blocking 520 with 3% bovine serum albumin in PBS for 1 hr. ApiAT5-3::HA was visualised using 521 rat anti-HA (1:500) followed by addition of Alexa488 conjugated donkey anti-rat 522 secondary antibody (1:2000) and DAPI, 5 µg/ml. 523

Plaque assay and amino acid complementation 524
For plaque assay analysis, 150 parasites were seeded on confluent HFF 525 monolayers, grown in 24-well plates, and left undisturbed for 5 days, before fixing 526 with chilled methanol and staining with 0.1% crystal violet. To assess growth in 527 excess tyrosine, plaque assays were repeated either at normal tyrosine levels (400 528 µM L-tyrosine disodium salt; as per Gibco manufacturer) or in DMEM 529 supplemented with 2mM L-tyrosine disodium salt (dissolved for 1 hour at 50 °C). 530 To ensure tyrosine had successfully dissolved samples of the media were analysed 531 by GC-MS as previously described (18). 532 were incubated with 8 µM Ca 2+ ionophore A23187 for 0, 0.5, 1, 1.5, 2, 2.5, 3, 3.5, 4 557 and 5 mins prior to the addition of 16% FA to a final concentration of 3% for 15 558 mins. Wells were subsequently washed with PBS and stained with 5 µg/ml DAPI. 559 Automated image acquisition of 25 fields per well was performed on a Cellomics 560 Array Scan VTI HCS reader (Thermo Scientific) using a 20× objective. Image 561 analysis was performed using the Compartmental Analysis BioApplication on HCS 562 Studio (Thermo Scientific). 563 564 Competition assays and flow cytometry 565 5×10 6 ApiAT5-3 ApiAT5-3 , ApiAT5-3 ApiAT5-3_S56A and ApiAT5-3 ApiAT5-3_S56D parasites 566 were mixed in a 1:1 ratio with ∆apiAT5-3 ApiAT5-3 , ∆apiAT5-3 ApiAT5-3_S56A and 567 ∆apiAT5-3 ApiAT5-3_S56D respectively. 5×10 4 parasites were added to fresh HFF 568 monolayers before spinning the rest of the sample at 72 × g to remove host cell 569 debris for 1 min. The supernatant was spun at 2049 × g for 5 mins. The pellet was 570 fixed for 10 mins in 3% FA, washed in PBS and stained with 5 µg/ml DAPI. The 571 sample was washed and resuspended in PBS before running on a flow cytometer. 572 All parasites were gated on DAPI fluorescence to prevent results being skewed by 573 remaining unstained host cell debris. The proportion of DAPI (+); YFP (+) 574 (representing ∆apiAT5-3 ApiAT5-3/S56A/S56D ) compared to DAPI (+); YFP (-) 575 (representing ApiAT5-3 ApiAT5-3/S56A/S56D ) was calculated. The process was 576 repeated 14 days later for comparison to day 0. 577 578 Oocyte maintenance and radiotracer uptake assays 579 ApiAT5-3, ApiAT5-3_S56A and ApiAT5-3_S56D were PCR amplified from 580 ∆apiAT5-3 ApiAT5-3 , ∆apiAT5-3 ApiAT5-3_S56A and ∆apiAT5-3 ApiAT5-3_S56D cDNA, 581 respectively, using primers 30 and 31 to add a region of homology to the XkbN 582 plasmid at the 5' end and a HA tag to the 3' end of each gene. These fragments were 583 then amplified with primers 32 and 33 to add a 3' XkbN homology overhang. These 584 were aligned using Bowtie 2 (46) to Ensembl Protist's release 35 of T. gondii 618 (ToxoDB-7.1). They were then quantified using RSEM before being processed 619 using Bioconducor (47). We used DESeq2 (48) to account for gene length and 620 library size, and to test for the interaction between treatment and time point to 621 generate the differential genelist. We corrected for multiple testing using the 622 Benjamini-Hochberg procedure for false discovery rates. To validate the 623 recodonised transcript, we both re-aligned to a custom genome rebuilt to include 624 the novel sequence, and also used a pseudo-alignment approach to quantify purely 625 the reads associated with the novel sequence (49). 626