The HCF101 protein is an important component of the cytosolic iron–sulfur synthesis pathway in Toxoplasma gondii

HCF101 is a putative P-loop containing nucleoside triphosphate hydrolase that belongs to the Mrp/NBP35 family (also called ApbC) of dimeric iron–sulfur carrier proteins that are found ubiquitously in all domains of life and thus likely appeared early during evolution [27,28]. Compared with NBP35, in addition to the nucleotide hydrolase domain HCF101 also contains a N-terminal DUF59/MIP18-family domain [29], as well as a C-terminal domain of unknown function (DUF971). We performed homology searches in the database to retrieve members of the Mrp/NPB35 family encoded by the T. gondii genome. Besides TgNBP35 that has an unusual mitochondrial localization in T. gondii [20,24], in line with previous studies [24,30], we identified TGGT1_318590 as a potential homolog of HCF101. A broader phylogenetic analysis that included members of the Mrp/NPB35 family from a wide variety of eukaryotes (S1 Table), revealed that this putative T. gondii homolog clearly segregates with members of the HCF101 group, clearly distinct from NBP35 homologs (Fig 1A). Moreover, its primary sequence analysis confirmed a 3 domain-organization typical of HCF101 proteins, while NBP35 homologs mostly consist of the central ATPase domain (Fig 1B). Of note, our analysis, in accordance with a previous study [24], confirmed the absence of a CFD1 homolog in Apicomplexa, contrarily to what has been suggested by a recent report [31]. In conclusion, TGGT1_318590 may be named TgHCF101.

Fig 1. TgHCF101 is a cytosolic protein.

(A) Evolutionary relationship of proteins of the MRP family. Eukaryotic sequences from HCF101 homologs were aligned and submitted to phylogenetic analysis with the maximum likelihood method. Scale bar represents 1 residue substitution per site. Bacterial P-loop ATPase sequences were used as an outgroup. HCF101 from A. thaliana is indicated by a green arrowhead, TgHCF101 by a yellow arrowhead and the 2 homologs present in C. velia are indicated by red arrowheads. NBP35 homologs from A. thaliana, T. gondii, and C. velia are indicated by purple, blue, and brown arrowheads, respectively. (B) Schematic representation of homologs for HCF101 in A. thaliana (AtHCF101), in T. gondii (TGGT1_318590) and in C. velia (Cvel_23131, Cvel_17212), and homologs for NBP35 in A. thaliana (AtNBP35), in T. gondii (TGGT1_280730), and in C. velia (Cvel_19821). Main domains are highlighted on the sequences; TP: predicted transit peptide, MIP18-like domain (or Domain of unknown function 59, DUF59), the ATPase domain and DUF971. (C) Strategy for generating the inducible knockdown of TgHCF101 by promoter replacement and simultaneous N-terminal tagging of the TgHCF101 protein in the TATi ΔKu80 cell line. (D) Diagnostic PCR for checking correct integration using the primers mentioned in (B), on genomic DNAs of a transgenic parasite clone and of the parental strain. (E) Immunoblot analysis of the cKD-TgHCF101 mutant and parental line showing efficient tagging and down-regulation of TgHCF101 starting at 24 h of treatment with ATc. Actin was used as a loading control. (F) Immunofluorescence assay showing a cytosolic signal for TgHCF101 protein (labeled with an anti-HA antibody), with no particular co-localization with the apicoplast (Apico, labeled with anti-PDH-E2 antibody), and total depletion of the protein after 48 h of ATc treatment. DNA was stained with 4′,6-diamidino-2-phenylindole (DAPI). Scale bar = 5 μm. The data underlying this figure can be found in S1 Table and in https://doi.org/10.6084/m9.figshare.28238195.

https://doi.org/10.1371/journal.pbio.3003028.g001

In A. thaliana, the HCF101 mutant was found to be impaired in photosynthesis [32]: interfering with HCF101 function has an impact on the maturation of both photosystem I (PSI) and ferredoxin-thioredoxin reductases, which contain [4Fe-4S] clusters [25,26]. The apicoplast harbored by apicomplexan parasites has lost the ability to perform photosynthesis [33], and some members of the Apicomplexa phylum, like Cryptosporidium, have even completely lost the plastid [34] while it has retained a HCF101 homolog (Figs 1A and S1). Thus, a function linked to photosystem seems unlikely for apicomplexan parasites. Sequence analysis confirmed that, in contrast to plant homologs, TgHCF101 lacks a predicted N-terminal transit peptide, which is typically needed for plastid import (Fig 1B). Interestingly, Chromera velia, a close non-parasitic relative of Apicomplexa whose plastid has retained its photosynthetic capacity [35], has 2 HCF101 isoforms: one with a predicted transit peptide (thus potentially addressed to the plastid), and one without (Figs 1 and S1). Specific analysis of selected eukaryotic HCF101 homologs showed that Apicomplexa proteins, like ciliates and dinoflagellates, form distinct groups that are clearly separate from the Archaeplastida group of plants, as well as algae bearing plastid-associated HCF101 isoforms (S1 Fig). Interestingly, while the transit peptide-bearing C. velia isoform segregates with homologs of plastid-containing eukaryotes, the second isoform is closer to the Apicomplexa homologs (S1 Fig). A recent study suggests that the plastid-located HCF101 likely evolved from an ancestral cytosolic version of the protein [24]. This led us to hypothesize that apicomplexan HCF101 would be involved in a cytosolic-related, rather than in a plastid-associated, Fe-S cluster transport function.

To assess experimentally the localization of TgHCF101, we first tagged an ectopic copy with a C-terminal GFP and observed a cytosolic localization (S2A Fig), in line with what had been described previously when an additional copy of TgHCF101 bearing a C-terminal Human influenza hemagglutinin (HA) tag was expressed [24]. However, a stable TgHCF101-GFP cell line could not be established, suggesting the potential importance of C-terminal accessibility for proper protein function. We then next generated a transgenic cell line in which we modified the 5′ region of the TgHCF101 gene by homologous recombination to replace the endogenous promoter by an inducible-Tet07SAG4 promoter, and at the same time adding a sequence coding for a N-terminal HA tag (Fig 1C and 1D). In this conditional knock-down (cKD) HA-TgHCF101 cell line, the addition of anhydrotetracycline (ATc) can repress transcription of TgHCF101 through a Tet-Off system [36]. Two independent transgenic clones were obtained and found to behave similarly in the initial phenotypic assays we performed, so only one will be described in details here.

Immunoblot analysis showed an N-terminal HA-tagged protein at around 70 kDa (the predicted molecular mass for TgHCF101) and depletion of the protein was efficient upon addition of ATc (Fig 1E). Immunofluorescence assay (IFA) showed a punctate cytosolic signal for TgHCF101, clearly distinct from the apicoplast, and efficiently depleted by the addition of ATc (Fig 1F). Disruption of the apicoplast-located SUF pathway has a strong impact on the lipoylation of the E2 subunit of the pyruvate dehydrogenase (PDH-E2) by Fe-S-containing lipoate synthase LipA [19,22]. However, depletion of TgHCF101 did not seem to have an impact on the lipoylation profile of PDH-E2 (S2B Fig). Disruption of the SUF pathway also typically impacts the apicoplast-located isoprenoid synthesis, which involves 2 Fe-S oxidoreductases (IspG and IspH) and has been shown to have downstream consequences on the glycosylation of proteins involved in parasite motility [22]. We thus assessed the gliding motility of the parasites after TgHCF101 depletion and found that it was only affected after 4 days of ATc treatment, which might be reflecting long-term impact of TgHCF101 depletion on parasite fitness rather than a direct consequence of apicoplast-related Fe-S cluster synthesis (S3A and S3B Fig).

We next assessed the impact of TgHCF101 depletion on the parasite’s lytic cycle: We evaluated the capacity of TgHCF101-depleted tachyzoites (a highly replicative and highly invasive form responsible for the acute form of the disease caused by T. gondii) to generate lysis plaques. For this, we infected a monolayer of host cells in absence or continuous presence of ATc for 7 days (Fig 2A and 2B). Depletion of TgHCF101 completely prevented plaque formation. While absence of plaques highlights a fitness problem for the parasites, it does not necessarily implies their death: for instance, in response to stress, tachyzoites can convert into a slow-growing and cyst-enclosed persisting stage called bradyzoites (associated with the chronic form of the disease), that can reactivate when they encounter more favorable conditions [37]. We thus performed a similar experiment, in which the ATc was washed out at the end of the 7-day incubation, and then incubated the parasites for an extra 7 days in the absence or presence of ATc before evaluating plaque formation (Fig 2C). We did not observe any plaques after ATc removal, suggesting that the parasites were not able to recover after 7 days of TgHCF101 depletion, likely because they were dead.

Fig 2. TgHCF101 is essential for parasite growth and survival.

(A) Plaque assays were carried out by infecting a monolayer of HFFs with TATi ΔKu80 or cKD-TgHCF101 cell lines for 7 days in the presence or absence of ATc. Scale bar = 2 mm. (B) Quantification of plaque area observed in (A). Results are expressed as percentage of lysed area relative to control (TATi ΔKu80 -ATc, set as 100% for reference). Values represented are mean ± SD from n = 3 independent biological replicates, ** p-value ≤0.01, Student t test. (C) Plaque assays were carried out with cKD-TgHCF101 parasites as described in (A) but for the ‘7d+7d-‘ condition, ATc was washed-out after 7 days and parasites were allowed to grow for another 7 days without ATc, while in the ‘7d+7d+’ condition, ATc treatment was maintained for 7 more days. The ‘7d –‘control was kept without ATc for 7 days of growth. Scale bar = 2mm. (D) Replication assay of parental (TATi ΔKu80) and transgenic cell lines (cKD-TgHCF101): parasites were pre-cultured for 48 h in the presence or absence of ATc and allowed to invade HFF-coated coverslips for another 24 h in the presence or absence of ATc, for a total of up to 72 h of treatment with ATc. Number of parasites per vacuole was quantified for each condition and expressed as a percentage, 200 vacuoles were counted for each condition. Values represented are mean ± SD of 3 independent biological replicates, **** p-value ≤0.0001 by two-way ANOVA with Dunnett’s multiple comparison test, showing a significant difference when comparing the TATi ΔKu80 control and cKD-TgHCF101+ATc parasites for percentage of vacuoles containing 1 or 8 parasites. (E) Immunofluorescence assay of cKD-TgHCF101 parasites showing asynchronous division of parasites growing in the same vacuole (outlined with a yellow dotted line) upon TgHCF101 depletion (+ATc condition: parasites pre-incubated for 48 h with ATc and allowed to invade coverslips for another 48 h in the presence of ATc). Individual parasites and budding daughter cells are outlined by anti-IMC3 antibody staining (magenta), DNA was stained with DAPI. Scale bar = 5 μm. (F) Electron microscopy of cKD-TgHCF101 parasites pre-incubated with ATc for 24 h before being released from their host cell and allowed to reinvade for 24 h in the presence (+ATc) of ATc, or grown in absence of ATc (-ATc). Inset shows magnification of a selected region. D: daughter bud, R: rhoptry, A: apicoplast, M: mitochondrion, CC: cytoplasmic cleft. The asterisk denotes unincorporated nuclear material. Scale bar = 2 μm (0.5 μm for inset magnification). (G) DNA content analysis by flow cytometry on TATi ΔKu80 and cKD-TgHCF101 parasites treated or not with ATc up to 4 days and stained with propidium iodide. 1N, 2N, 3N represent the ploidy, with S1 Data and http://flowrepository.org/id/FR-FCM-Z8SR.

https://doi.org/10.1371/journal.pbio.3003028.g002

To assess whether this defect in the lytic cycle is due to a replication problem, cKD HA-TgHCF101 parasites were preincubated in ATc for 48 h and then released mechanically, before infecting new host cells for an additional 24 h in ATc prior to parasite counting. We noted that incubation with ATc led to an accumulation of vacuoles with fewer parasites (Fig 2D). To get more precise insights into the impact of TgHCF101 depletion on parasite division, we labeled dividing parasites with inner membrane complex (IMC) protein IMC3 [38] to detect growing daughter cells (Fig 2E). T. gondii tachyzoites develop inside a mother cell by a process called endodyogeny [39] and this division is usually highly synchronous within the same parasitophorous vacuole, yet after 2 days of ATc treatment vacuoles showed a marked lack of synchronicity for daughter cell budding (Fig 2E). Investigation of the consequences of TgHCF101 depletion on developing parasites by electron microscopy highlighted important defects on cytokinesis, including incomplete daughter cell budding or defaults in organellar segregation (Figs 2F and S4). We also observed occasional vacuoles and cytoplasmic clefts (Figs 2F and S4). However, the overall appearance of organelles seemed essentially normal (Figs 2F and S4). To get a more general overview, we performed IFA using organelle-specific markers, which confirmed that TgHCF101 depletion leads to defaults in the replication or segregation of organelles, some daughter cells appearing devoid of apicoplast, mitochondrial or nuclear material (S5 Fig). Finally, through propidium iodide labeling and flow cytometer-based analysis of DNA content, we observed that long-term depletion of TgHCF101 led to the emergence of an increasing subpopulation of parasites with sub-1N DNA content, thus smaller than the typical haploid 1N DNA content of the parasites (Fig 2G), suggesting an impairment in DNA synthesis.

Through electron microscopy analysis, we also noticed that the depletion of TgHCF101 induced the appearance of structures resembling lipid droplets (LDs) in several parasites (Figs 3A and S4). In order to properly quantify this, we used Nile red, a selective fluorescent stain for neutral lipids, and microscopic imaging (Fig 3B). We noticed a strong increase in both the number and size of LD in cKD HA-TgHCF101 parasites incubated for 2 days with ATc (Fig 3C and 3D). To see if this was a common feature induced by abolishing Fe-S synthesis, we quantified LDs in mutants of the mitochondrial (TgISU1) and apicoplast (TgSUFC) pathways [19]. Depletion of these proteins did not lead to an increase in LD formation (S6A and S6B Fig), indicating that LD induction may be linked to a pathway distinct from the mitochondrial or plastid-related Fe-S cluster assembly machinery. We then also tested a recently characterized mutant for TgABCB7L (a mitochondrial transporter of a sulfur precursor upstream of the CIA pathway [40]) and a mutant of the CIA shuttle protein TgNAR1 (www.ToxoDB.org entry TGGT1_242580) that we generated, and which is important for parasite fitness (S7 Fig). Interestingly, both these CIA-related mutants induced an increase in LD numbers (although more modest for the TgABCB7L mutant), while only the TgNAR1 led to a statistically significant increase in LD size (Fig 3C and 3D).

Fig 3. TgHCF101 depletion induces lipid droplet accumulation.

(A) Electron microscopy of cKD-TgHCF101 parasites pre-incubated with ATc for 48 h and allowed to reinvade for 24 h in the presence of ATc. Scale bar = 1 μm. The panel on the right corresponds to a magnification of the selection on the left panel, highlighting lipid droplets. Scale bar = 500 nm. (B) Fluorescent imaging of parasites from the cKD-TgHCF101 cell line treated for 72 h with ATc or in grown the absence of ATc, shows an accumulation of lipid droplets upon TgHCF101 depletion. LDs were detected with Nile red (orange), parasites are outlined with an anti-IMC3 antibody (green) and DNA is stained with DAPI. DIC = Differential interference contrast. Scale bar = 5 μm. (C, D) Correspond to the quantification of lipid droplet area and number, respectively, and 100 parasites were analyzed per condition. The parental (TATi ΔKu80) and transgenic parasites (cKD HA-TgHCF101, cKD TgABCB7, cKD TgNAR1) were grown in absence of ATc, or in presence of ATc for 72 h. Values are represented as the mean ± SD of n = 3 independent biological replicates (different symbols represent different series); ns, not significant (p-value >0.05), * p-value ≤0.05, ** p-value ≤0.01, *** p-value ≤0.001, **** p-value ≤0.0001, Student’s t test. The data underlying this figure can be found in S1 Data. cKD, conditional knock-down; LD, lipid droplet; SD, standard deviation.

https://doi.org/10.1371/journal.pbio.3003028.g003

As LD droplet induction seems to be a feature common to CIA pathway mutants and to the TgHCF101 mutant, we thought that TgHCF101 may play a role in this pathway. Additional supporting evidence comes from the characterization of a C-terminal targeting complex recognition (TCR) signal ending with an aromatic residue ([LIM]-[DES]-[WF]) in the clients proteins of the cytosolic Fe-S cluster machinery or their adaptors [41], which can be clearly identified in the apicomplexan HCF101 homologs (Fig 4A). Of note, this motif is also present in the C. velia isoform that has no transit peptide and thus which is likely to have a plastid-independent function (Fig 4A). Importantly, the last 3 amino acids of TgHCF101 (LEW, Fig 4A) constitute a canonical TCR motif. Thus, from the cKD HA-TgHCF101 mutant, we generated cell lines in which we expressed either a wild-type copy or one deleted for this putative TCR signal (Fig 4B and 4C). As expected, expression of these additional copies was unaltered by ATc treatment (Fig 4D), and the corresponding proteins localized to the cytoplasm (Fig 4E). Upon depletion of the ATc-regulated copy, expressing the extra wild-type copy was efficient in restoring parasite growth (Fig 4F and 4G) and preventing LD accumulation (Fig 4H and 4I). Yet, in sharp contrast, expressing the TCR-deleted extra copy did not improve parasite fitness (Fig 4F and 4G) and did not prevent induction of LD formation (Fig 4H and 4I).

Fig 4. A C-terminal TCR-like motif is necessary for TgHCF101 function.

(A) Alignment of the C-terminal region of HCF101 homologs from different eukaryotes, including the plant A. thaliana, the 2 isoforms of Apicomplexa-relative photosynthetic algae C. velia, as well as several Apicomplexa species (T. gondii, Eimeria tenella, P. falciparum, and apicoplast-less Cryptosporidium parvum). The tryptophan-containing TCR signal is highlighted in yellow. Consensus tree was obtained with protein alignment and bootstrap values (500 replicates) are indicated at the base of the nodes. (B) Schematic representation of the strategy for generating cell lines expressing myc-tagged wild-type (WT) or TCR motif-deleted (-LEW) copies of TgHCF101 by integrating an extra copy of the gene of interest by double homologous recombination at the Uracil Phosphoribosyltransferase (UPRT) locus. Negative selection with 5-fluorodeoxyuridine (FUDR) was used to select transgenic parasites based on their absence of UPRT expression. (C) Diagnostic PCR for verifying integration at the UPRT locus thanks to the primers described in B. (D) Immunoblot analysis showing expression of the additional myc-tagged copies upon depletion of the ATc-regulated HA-tagged copy. (E) Immunofluorescence with anti-myc antibody confirms the cytoplasmic localization of the extra copies. Parasite shape is outlined. DNA was stained with DAPI. Scale bar = 5 μm. (F) Plaque assay showing restoration of growth by the additional WT copy upon depletion of the ATc-regulated copy, contrarily to the TCR mutant. (G) Representative images of vacuoles (outlined) containing parasites grown continuously in the presence of ATc for 72 h and stained by Nile red (orange) for lipid droplets. DIC: differential interference contrast. DNA was stained with DAPI. Scale bar = 10 μm. (H, I) Correspond to the quantification of lipid droplet area and number, respectively; 100 parasites were analyzed per condition. WT or TCR-depleted (-LEW) parasites were grown in presence of ATc for 72 h. Values are represented as the mean ± SD of n = 3 independent biological replicates (different symbols represent different series); ns, not significant (p-value >0.05), ** p-value ≤0.01, **** p-value ≤0.0001, Student’s t test. The data underlying this figure can be found in S1 Data. HA, hemagglutinin; SD, standard deviation; TCR, targeting complex recognition.

https://doi.org/10.1371/journal.pbio.3003028.g004

To assess if the induction of LD formation could play a role in the demise of the parasites, we used T863, a pharmaceutical inhibitor of the acyl coenzyme A:diacylglycerol acyltransferase 1 (DGAT1) enzyme, which is important for storing neutral lipids in cytoplasmic LD [42]. While treatment with T863 efficiently decreased the number of LDs (S6C, S6D, and S6E Fig), long-term incubation during plaque assays did not allow any recovery in parasite fitness (S6F Fig). This suggests that LDs are not detrimental to TgHCF101-depleted parasites (or at least not solely responsible for their demise), but instead could be part of an integrated stress response, as they have been shown in other eukaryotes to be up-regulated in response to cellular injuries including nutrient-related and oxidative stresses [43].

Disrupting proteins involved in Fe-S cluster assembly may have a direct effect on the stability and expression levels of local Fe-S proteins. Thus, to get insights into the effect of TgHCF101 depletion at the molecular level, we performed global label-free quantitative proteomic analyses. Of course, this may also affect downstream cellular pathways or functions, and other pathways may also be up-regulated in compensation. cKD HA-TgHCF101 or TATi ΔKu80 parental control parasites were treated for 3 days with ATc prior to a global proteomic analysis and compared for protein expression. We selected candidates with a log₂(fold change) ≤-0.55 or ≥0.55 (corresponding to a ~1.47-fold change in decreased or increased expression) and a p-value n = 4 biological replicates) and we completed this data set by selecting some candidates that were consistently and specifically absent from the mutant cell lines or only expressed in these (S2 and S3 Tables). Many proteins with higher abundance were bradyzoite stage-specific and more particularly components of the cyst wall or their cell surface [44,45] (Fig 5A and 5B and S2 Table). Tachyzoites can convert to the persistent bradyzoite form upon stress, so to verify if TgHCF101 depletion was inducing stage conversion we used a lectin from the plant Dolichos biflorus, which recognizes the SRS44/CST1 cyst wall glycoprotein in differentiating cysts [46] (Fig 5C and 5D). We could see that TgHCF101 depletion induced the appearance of up to 10% of lectin-labeled vacuoles during the first 48 h of intracellular growth in the presence of ATc. However, when parasites were kept for up to a week in the presence of ATc, the proportion of lectin-positive vacuoles plateaued at 14%, and only a very limited number of vacuoles looked like bona fide mature cysts (less than 2% were both lectin-positive and containing more than 2 parasites) (Fig 5E).

Fig 5. TgHCF101-depleted parasites express bradyzoite-specific markers but are unable to complete their conversion.

(A) Volcano plot showing differential expression of proteins impacted by TgHCF101 depletion after 72 h of ATc treatment analyzed by label-free quantitative proteomic. X-axis corresponds to the log₂ of the fold-change (FC) and the Y-axis corresponds to the -log₁₀ of the p-value, when comparing cKD-TgHCF101 expression values to the TATi ΔKu80 parental cell line. Statistical analyses were performed with ANOVA on 4 independent biological replicates. Cut-offs were set at ≤1.5- or ≥1.5-FC and p-value ≤0.05. Significant hits corresponding to stage-specific protein are highlighted in green on the graph. (B) Clustering of bradyzoite (Bz) or tachyzoite (Tz)-specific proteins of the SRS family shows specific enrichment of bradyzoite proteins upon TgHCF101 depletion. (C) Immunofluorescence assay of cKD-TgHCF101 treated for 24 h or 48 h with ATc, cyst wall is labeled with DBL, parasites periphery is outlined with anti-IMC3 antibody and DNA is stained with DAPI. Scale bar = 10μm. (D) Corresponds to the quantification of the percentage of vacuoles presenting a DBL positive signal as shown in (C). Values are represented as the mean ± SD of n = 3 independent biological replicates, * p-value ≤0.05, Student’s t test. (E) Immunofluorescence assay of cKD-TgHCF101 treated for 7 days in the presence of ATc, cyst wall is labeled with DBL, parasites are outlined with anti-IMC3 antibody and DNA is stained with DAPI. The percentage of DBL positive vacuoles of corresponding size (more than 2 parasites per vacuole, top; or 2 parasites per vacuole or less, bottom) is specified as mean ± SD of n = 3 independent biological replicates. DIC = Differential interference contrast. Scale bar = 10 μm. The data underlying this figure can be found in S2 and S3 Tables and in S1 Data. cKD, conditional knock-down; SD, standard deviation.

https://doi.org/10.1371/journal.pbio.3003028.g005

The label-free quantitative proteomic analyses also highlighted proteins which were less expressed in absence of TgHCF101. Among them were a surprisingly large number of rhoptry bulb proteins (S3 Table and S8A Fig). Rhoptries are club-shaped organelles that comprise a narrow tubular neck opening at the anterior pole of the parasite, and a most posterior bulbous part; the proteins secreted from these different sub-compartments are either involved in invasion and parasitophorous vacuole formation, or in the modulation of host cell defenses, respectively [47]. We assessed if there was any major impact of TgHCF101 depletion on rhoptry morphology or function. We performed IFA with anti-armadillo repeats only protein (TgARO), a protein homogenously anchored to the surface of the rhoptry membrane [48], and found no particular defect in rhoptry morphology or positioning (S8B Fig), in accordance with our initial electron microscopy analysis (Fig 2F). Quantification of intracellular evacuoles (rhoptry-secreted vesicular clusters) did not indicate any particular problem in the secretory capacity of the organelles upon TgHCF101 depletion (S8C Fig). Finally, we assessed by immunoblot the expression profile of one of the potentially less-expressed rhoptry bulb protein but found only a slight alteration in the expression profile (a modest increase in the non-mature from of the protein), but not in the overall amount of protein (S8D Fig). We conclude that although TgHCF101 depletion may lead to a secondary impact on rhoptry content, it does not extensively impact organelle morphology or secretory function.

Given the potential implication of TgHCF101 in the biogenesis of Fe-S proteins, we also specifically searched for putative Fe-S proteins in the less expressed proteins highlighted by the label-free quantitative proteomic analysis. We have previously estimated the Fe-S proteome using a computational predicting metal-binding sites in protein sequences [49], which, coupled with the data from global mapping of T. gondii proteins subcellular location by HyperLOPIT spatial proteomics [50], allows predicting client proteins present in the mitochondrion, apicoplast, or cytosol [19]. We found 3 Fe-S proteins that were less expressed upon TgHCF101 depletion (Fig 6A). Interestingly, they were all connected to the CIA pathway as they were NAR1, an Fe-S carrier to the CIA core machinery [15], and 2 Fe-S proteins depending on the CIA machinery for their maturation [51]: ABCE1 (Rli1 in yeast), a cytosolic ribosomal recycling factor and NTHL1, a nuclear DNA base-excision repair enzyme. In order to verify the proteomics data, we tagged TgABCE1 with a myc epitope in the context of the cKD HA-TgHCF101 mutant (S9A and S9B Fig). IFA showed a decrease in the cytosolic TgABCE1 signal (Fig 6B), which was supported by quantitative immunoblot analysis indicating a statistically significant decrease in TgABCE1 upon treatment of cKD HA-TgHCF101 mutant with ATc (Fig 6C and 6D).

Fig 6. TgHCF101 is involved in Fe-S cluster biogenesis of the CIA pathway.

(A) Volcano plot showing differential expression of proteins impacted by TgHCF101 depletion after 72 h of ATc treatment analyzed by label-free quantitative proteomic. Cut-offs were set at ≤1.5- or ≥1.5-fold change (FC) and p-value ≤0.05. Significant hits corresponding to predicted Fe-S cluster proteins are highlighted in red on the graph. (B) Immunofluorescence assay for the detection of myc-tagged TgABCE1 in the cKD TgHCF101 genetic background. Parasites were pre-incubated for 48 h in the presence of ATc and allowed to invade HFF-coated coverslips for another 48 h in the presence of ATc. The control (no ATc) was infected 24 h prior to fixation. TgABCE1 was detected with an anti-myc antibody and DNA was stained with DAPI. Parasites periphery is outlined by white dotted lines. Scale bar = 5 μm. (C) Immunoblot analysis of TgABCE1 abundance shows decrease upon TgHCF101 depletion following up to 4 days of treatment with ATc of the cKD HA-TgHCF101 TgABCE1-myc cell line. TgABCE1 was detected with anti-myc antibody and anti-actin antibody was used as a loading control. (D) Decrease of TgABCE1 expression upon TgHCF101 depletion was quantified by band densitometry analysis and normalized on the loading control of each respective lane. The relative abundance of TgABCE1 is presented as a percentage relative to the untreated control, set as 100%, for each biological replicate. Values are represented as the mean ± SD from n = 4 independent biological replicates, ** p-value ≤0.01, **** p-value ≤0.0001; ns, not significant (p-value ≥0.05), Student’s t test. (E) Immunofluorescence assay for the detection of HA-tagged TgPOLD1 in the cKD TgHCF101 genetic background. Parasites were pre-incubated for up to 4 days in presence of ATc on HFF-coated coverslips. The control (no ATc) was infected 24 h prior to fixation. TgPOLD1 was detected with an anti-HA antibody and DNA was stained with DAPI. Parasites periphery is outlined by white dotted lines. Scale bar = 5 μm. (F) Immunoblot analysis of TgPOLD1 abundance in conditions of TgHCF101 depletion following up to 4 days of treatment with ATc of the cKD TgHCF101 TgPOLD1-HA cell line. TgPOLD1 expression was detected with anti-HA antibody and anti-actin antibody was used as a loading control. (G) Changes in TgPOLD1 expression upon TgHCF101 depletion was quantified by band densitometry analysis and normalized on the loading control of each respective lane. The relative abundance of TgPOLD1 is presented as a percentage relative to the untreated control, set as 100%, for each biological replicate. Values are represented as the mean ± SD from n = 5 independent biological replicates, * p-value ≤0.05; ** p-value ≤0.01; ns, not significant (p-value ≥0.05), Student’s t test. The data underlying this figure can be found in S2 and S3 Tables and in S1 Data. cKD, conditional knock-down; CIA, cytosolic iron–sulfur cluster assembly; HA, hemagglutinin; HFF, human foreskin fibroblast.

https://doi.org/10.1371/journal.pbio.3003028.g006

We did not manage to tag TgNTHL1 to verify if its abundance was effectively affected by TgHCF101 depletion. The catalytic subunit of DNA polymerase delta (TgPOLD1), which is another nucleus-associated Fe-S protein, was found to be decreasing in abundance in the CIA-related TgABCB7L mutant [40]. Our quantitative proteomic analysis did not highlight a particular decrease in TgPOLD1 abundance upon TgHCF101 depletion: it was found to be stable, or even slightly more abundant (by a 1.4-fold), after 2 days of TgHCF101 depletion. Yet, we knew it could be tagged and decided to use this to investigate more thoroughly a potential global impact TgHCF101 depletion on nuclear and cytoplasmic Fe-S proteins. We thus created a conditional knockdown line of TgHCF101 in the POLD1-HA background [40] and verified that it displayed a similar phenotype to the original cKD HA-TgHCF101 cell line (S10 Fig). In the TgABCB7L mutant, TgPOLD1 abundance was found to be sharply decreasing early, before any noticeable impact on TgABCE1 [40]. In contrast, upon TgHCF101 depletion we observed a limited and delayed decrease in TgPOLD1 abundance (Fig 6E–6G) and intriguingly, TgPOLD1 levels even seemed to increase in the first 2 days of TgHCF101 depletion (Fig 6F and 6G). This is in sharp contrast with the more rapid and complete impact on TgABCE1 (Fig 6B–6D). Of note, IFAs highlighted that although TgHCF101 depletion did not lead to a marked loss of TgPOLD1, in some parasites the protein remained in the cytoplasm instead of being translocated to the nucleus (Fig 6E), which in fact corresponds to a stage where we have shown that DNA content of the parasite is affected (Fig 2G).

To get further insights on the role played by TgHCF101 in the CIA machinery, we next performed co-immunoprecipitations (co-IPs) and mass spectrometry identification of associated proteins, comparing lysates of cKD HA-TgHCF101 parasites grown in the presence of ATc or not (S4 Table). Strikingly, we pulled-down all 3 T. gondii homologs of the core components of the CTC: MET18, CIA1, and AE7 (an MIP18 family protein) (Fig 7A). We managed to tag the TgCIA1 protein with a myc epitope tag in the context of the HA-tagged TgHCF101 cell line (S9C and S9D Fig) and performed a reverse IP by which we managed to specifically pull-down TgHCF101, confirming their interaction (Fig 7B). It should be noted that this result was only obtained when this experiment was performed after chemical crosslinking, suggesting a possibly indirect or transient interaction between the proteins.

Fig 7. TgHCF101 is associated to the CIA targeting complex and specifically interacts with TgABCE1.

(A) Volcano plot showing differential expression of TgHCF101 and co-immunoprecipitated proteins in the cKD HA-TgHCF101 cell line after TgHCF101 depletion (with 72 h of ATc treatment) or not, as analyzed by quantitative proteomic. Cut-offs were set at ≤1.5- or ≥1.5-fold change and p-value ≤0.05. Up-regulated proteins are highlighted in blue, significant hits corresponding to predicted proteins of the CIA targeting complex and target protein TgABCE1 were annotated on the graph. (B) Immunoblot analysis of a reverse co-immunoprecipitation assay of the myc-tagged TgCIA1 protein in the cKD HA-TgHCF101 background shows specific co-immunoprecipitation of TgHCF101, which is absent upon depletion by ATc. The anti-SAG1 antibody was used as a control for unspecifically bound proteins. (C) Immunoblot analysis of a reverse co-immunoprecipitation assay of the myc-tagged TgABCE1 protein showing TgHCF101 is co-immunoprecipitating. The anti-SAG1 antibody was used as a control for unspecifically bound proteins. (D) TgHCF101 interacts with TgABCE1 in a Gal4-based yeast two-hybrid assay. YRG2 cells co-transformed with AD- and BD-fusion proteins were grown to stationary phase, then serially diluted to OD₆₀₀ values ranging from 1 to 5 × 10⁻³ before being spotted onto a control plate (+His, upper panel) to assess cell viability, and a Y2H test plate (-His, middle panel) to assess interaction. Plates were incubated at 30°C, and yeast growth was recorded after 5 days. As shown in the middle panel, TgHCF101 exhibited a strong interaction with TgABCE1, independent of cloning orientation. The strongest interaction was observed between AD-TgHCF101 and BD-TgABCE1, with co-transformed cells growing at the lowest dilution tested in the presence of 10 mM 3-aminotriazole (3AT) as a competitive inhibitor. Neither TgHCF101 nor TgABCE1 alone showed HIS3 transactivation in the presence of 3AT. These results are representative of 3 independent experiments. (E) Immunoblot analysis of puromycin incorporation in the parental (TATi ΔKu80) and cKD TgHCF101 cell lines untreated or treated with ATc for 72 h. TATi ΔKu80 treated with cycloheximide (CHX) was used as a control for translation inhibition. The puromycin signal was detected with anti-puromycin antibody, and total protein content was visualized by stain-Free imaging technology. Anti-actin antibody was also used as a loading control. (F) Variation of puromycin incorporation in the different conditions was quantified by band densitometry and normalized on the total protein content of each respective lane. Puromycin labeling is presented as a percentage relative to the untreated control, set as 100% for each biological replicate. Values are represented as the mean and SD of 4 independent biological replicates; * p-value ≤0.05; ** p-value ≤0.01; ns, not significant (p-value ≥0.05), Student’s t test. The data underlying this figure can be found in S4 Table and in S1 Data. cKD, conditional knock-down; CIA, cytosolic iron–sulfur cluster assembly; HA, hemagglutinin.

https://doi.org/10.1371/journal.pbio.3003028.g007

Another interesting candidate that was found co-immunoprecipitated with TgHCF101 was TgABCE1, strengthening the possibility of an important relationship between the 2 proteins. We next performed a reverse IP and could recover small amounts of TgHCF101 co-eluting specifically with TgABCE1 (Fig 7C). To further investigate this potential interaction, we performed yeast two-hybrid assay, based on interaction-dependent transactivation of the HIS3 reporter gene (Fig 7D). This experiment confirmed that there is a direct interaction between TgHCF101 and TgABCE1, and highlighted a particularly strong interaction, as it was retained when using 3-aminotriazole, a competitive inhibitor of the HIS3 gene product (Fig 7D). While this confirms that TgABCE1 is likely a client protein of TgHCF101, we could not confirm a direct interaction between TgHCF101 and TgNTHL1 using the same method (S11 Fig). Given the known implication of ABCE1 in translation, we next evaluated the impact of TgHCF101 depletion on translation by using a puromycin-based assay (Fig 7E). Puromycin can mimic aminoacyl-tRNA and becomes covalently attached to nascent peptides, thereby allowing the evaluation of protein elongation rates with an anti-puromycin antibody [52]. Using this method, we could show that TgHCF101 depletion leads to a modest but a statistically significant decrease in the overall translation rate (Fig 7E and 7F).

Transgenic cell lines were generated in the TATi ΔKu80 genetic background line lacking the Ku80 gene and expressing the TATi transactivator required for the TetOff system [64]. For positive selection of transgenic parasites bearing resistance cassettes for expressing dihydrofolate reductase thymidylate synthase (DHFR-TS) or chloramphenicol acetyltransferase (CAT) were grown with 1 μm pyrimethamine or 20 μm chloramphenicol (Sigma-Aldrich, SML3579, C0378), respectively. Conditional depletion in the TetOff conditional knockdown lines was achieved by incubation with 0,5 μg/ml anhydrotetracycline (ATc, Fluka 37919) for the indicated duration of the assay.

cDNA corresponding to the TgHCF101 gene (TGGT1_318590) was amplified by PCR with primers ML4715 and ML4716 (all primers used in the present study are listed in S5 Table) and sub-cloned using XhoI and KpnI into the pEZS-NL vector (D. Ehrhardt, https://deepgreen.dpb.carnegiescience.edu/cell%20imaging%20site%20/html/vectors.html) for a C-terminal GFP fusion. The TgHCF101-GFP cassette was them amplified by PCR using the ML4817 and ML4818 primers and cloned using BclI and EcoRV in a BglII/EcorV-digested pTUB-IMC1-TdT vector [72] to drive the expression from a tubulin promoter. Tachyzoites were transfected with 100 μg of plasmid and observed by fluorescence microsocopy.

To generate the construct for the tetracycline-regulated conditional depletion of TgHCF101 and add an N-terminal HA tag, we digested the DHFR-TetO7Sag4 plasmid [72] by BglII and inserted a fragment coding for a single HA tag generated through hybridization of the ML4924 and ML4925 oligonucleotides. Then, a 1,066 bp fragment corresponding to the 5′ coding part of TgHCF101 was amplified by PCR with primers ML4926 and ML4927 and inserted after digesting with BglII and NotI, to yield the DHFR-TetO7Sag4-HA-TgHCF101 plasmid. The TATi ΔKu80 cell line was transfected with 80 μg of the BsiWI-linearized plasmid. Transgenic parasites, named cKD HA-TgHCF101, were selected with pyrimethamine and cloned by limiting dilution. Positive clones were verified by PCR with primers ML1771 and ML6043.

The cKD TgHCF101-HA cell line was complemented by adding an extra copy of the TgHCF101 gene under the control of a tubulin promoter at the UPRT locus. The entire TgHCF101 cDNA sequence (1,935 bp) was amplified by PCR using primers ML6409/ML6410 from the plasmid pGBKT7-myc-HCF101 to express an N-terminal myc-tagged HCF101 copy. This sequence was then cloned downstream of the tubulin promoter in the pUPRT-TUB vector [64], resulting in the pUPRT-myc-TgHCF101 plasmid. The plasmid was subsequently linearized prior to transfecting the mutant cell line, together with a plasmid expressing Cas9 and a UPRT-specific guide RNA under the control of a U6 promoter [73]. Transgenic parasites were selected using 5 μm 5-fluorodeoxyuridine (Sigma-Aldrich) and cloned by serial limiting dilution to obtain the cKD TgHCF101 WT complemented cell line.

A CRISPR-based strategy was used to C-terminally tag proteins of interest in the cKD HA-TgHCF101 background. Guide RNAs (gRNAs) targeting the C-terminal end of the gene of interest were selected using CHOP-CHOP tool (https://chopchop.cbu.uib.no/). gRNAs were cloned into the pU6-Cas9 Universal Plasmid (Addgene, 52694) using BsaI. Donor DNA was amplified by PCR using the high fidelity KOD DNA polymerase (Novagen) to amplify a fragment containing the tag and the resistance cassette, using the pLIC-myc-CAT plasmid as a template, adding at both ends 30-nucleotide overhangs homologous to the C-terminal region of the gene of interest for homologous recombination. TgABCE1 (TGGT1_216790) tagging was performed by using primers ML5126 and ML5127 for the gRNA and ML5128 and ML5129 to amplify donor DNA, yielding the cKD HA-TgHCF101 TgABCE1-myc cell line. Integration of the construct was controlled by PCR using GoTaq DNA polymerase (Promega) with primers ML6115 and ML4310. TgCIA1 (TGGT1_313280) tagging was performed by using primers ML5932 and ML5933 for the gRNA and ML5934 and ML5935 to amplify donor DNA, yielding the cKD HA-TgHCF101 TgCIA1-myc cell line. Integration of the construct was controlled by PCR using GoTaq DNA polymerase (Promega) with primers ML5936 and ML4310.

The POLD1-HA cell line [40] was used as a background to generate a tetracycline-regulated conditional TgHCF101 mutant. Primers ML4926 and ML4927 were used to amplify a 1,066 bp 5′ fragment of the TgHCF101 coding sequence and cloned using BglII/NotI in the DHFR-TetO7Sag4 plasmid [72] to yield the DHFR-TetO7Sag4-TgHCF101 plasmid. The POLD1-HA cell line was transfected with 80 μg of this plasmid linearized by NsiI. Transgenic parasites were selected with pyrimethamine and cloned by limiting dilution. Positive clones were verified by PCR with primers ML2456 and ML6043.

Semi-quantitative RT-PCR was performed as described previously [74]. Briefly, 1 μg of RNA was used as a template for each RT-PCR reaction and specific primers for TgHCF101 (ML5112/ML5113), TgNAR1 (ML6467/ML6468), or TUB2 (Tubulin β chain) (ML841/ML842) were used. Twenty-four cycles of PCR were performed.

For IFAs, coverslips seeded with HFFs were infected by T. gondii tachyzoites. Intracellular parasites and host cell monolayer were then fixed with 4% (w/v) paraformaldehyde (PFA, diluted in phosphate-buffered saline, PBS) for 20 min. After fixation, cells were permeabilized with 0.3% (v/v) Triton X-100 (diluted in PBS) for 10 min. Coverslips were blocked with 2% (w/v) bovine serum albumin (BSA) for 1 h prior to immunolabeling with primary antibody for 1 h. After 3 washes in PBS, corresponding secondary antibody was incubated for 1 h. Coverslips were finally incubated with 1 μg/ml 4,6-diamidino-2-phenylindole (DAPI) for 5 min before 3 washes in PBS and, lastly, mounted using Immu-Mount (Thermo Fisher) onto microscope slides. Primary antibodies used were prepared in 2% BSA (diluted in PBS) and used at the following concentrations: rat monoclonal anti-HA (1:1,000, 3F10 Roche), mouse monoclonal anti-myc (1:100, 9E10 Sigma), rabbit anti-IMC3 (1:1,000) [75], mouse anti-SAG1 (T41E5, 1:1,000) [76], rabbit anti-armadillo repeats only (ARO, 1:1,000) [48], rabbit anti-PDH-E2 (1:500) [22], mouse anti-F1β ATPAse (1:1,000, gift of P. Bradley). Highly cross-adsorbed Alexa Fluor 488-conjugated or Alexa Fluor 594-conjugated anti rat-, rabbit-, or mouse-IgG secondary antibodies were all from Thermo Fisher and were diluted at 1:4,000. Cysts were stained with biotin labeled Dolichos Biflorus lectin (1:300, Sigma L-6533) and detected with FITC-conjugated streptavidin (1,300, Invitrogen, SNN1008). Lipid droplets were detected with Nile Red (1 μg/ml, Sigma 72485), using the CY3/DsRed BP550/25 FT570 BP605/70 filter cube (Zeiss) on an epifluorescence microscope.

The lytic cycle of tachyzoites was assessed by plaque assay as described previously [59]. Briefly, tachyzoites from the cKD HA-TgHCF101 transgenic cell line or parental strain (TATi ΔKu80) were allowed to invade monolayers of HFFs in the presence or absence of ATc. Parasites were cultivated for 7 days at 37°C and 5% CO₂ and fixed with 4% (w/v) PFA (diluted in PBS) for 20 min. Cells were stained with 0.1% crystal violet solution (V5265, Sigma-Aldrich), washed, and air-dried before imaging on an Olympus MVX10 microscope. For the reversibility assay, drug washout was carefully performed after 7 days of ATc pretreatment and cultures were kept for another 7 days of growth, non-treated control conditions were also infected at this time point.

Protein extracts were prepared from 10⁷ extracellular parasites resuspended in Laemmli buffer at a final concentration of 10⁶ parasites/μl. Extracts were treated with benzonase to remove DNA from samples and resolved by SDS-PAGE before being transferred on nitrocellulose membrane for subsequent protein detection. Primary antibodies used for immunodetection were resuspended at their respective working concentration in 5% (w/v) milk in TNT buffer (0.1 M Tris-HCl (pH 7.6), 0.15 M NaCl, 0.05%, Tween 20). Antibodies used in this study for immunoblot detection were rat anti-HA (1:1,000, Roche), mouse anti-myc (1:100, Sigma), mouse anti-SAG1 (1:50, hybridoma), mouse anti-actin (1:25, hybridoma) [77], mouse anti-ROP7 (1:1,000, T43H1) [78], mouse anti-puromycin (1:1,000, 12D10, MABE943 Sigma), and rabbit anti-lipoic acid (1:500, ab58724 Abcam) [79]. Alkaline phosphatase-conjugated anti mouse- or anti rabbit-IgG were from Promega and used at 1:2,500. Alkaline phosphatase-conjugated anti rat-IgG secondary antibodies were from Sigma and used at 1:2,500. Horseradish peroxidase-conjugated anti rat-, rabbit-, or mouse-IgG secondary antibodies were all from Thermo Fisher and were diluted at 1:10,000.

Parasites from the TATi ΔKu80 and cKD HA-TgHCF101 cell lines were treated for 72 h in HFF seeded in T75 cm² flasks. After suitable incubation, parasites were released from host cells with a cell scraper and passed through a 25G needle before filtration on a fiber glass wool column. Parasites were pelleted and washed in Hank’s balanced salt solution (HBSS, Gibco). Parasites were resuspended in lysis buffer (1% SDS, 50 mM Tris HCL (pH 8), 10 mM EDTA (pH 8)) and protein quantification was determined with the bicinchoninic acid kit (Abcam). For each condition, 20 μg of protein resuspended in Laemmli buffer were resolved on a 10% SDS-PAGE for 35 min at 100V. Proteins were fixed with a combination of acetic acid and ethanol and stained in PageBlue Protein Staining Solution (Thermo Scientific). Each lane was cut in 3 identical pieces which were digested with trypsin and peptide extraction was done as previously described [80].

LC-MS/MS experiments were performed using an Ultimate 3000 RSLC nano system (Thermo Fisher) interfaced online with a nano easy ion source and an Exploris 240 Plus Orbitrap mass spectrometer (Thermo Fisher). The.raw files were analyzed with MaxQuant version 2.0.3.0 using default settings (PMID: 19029910). The minimal peptide length was set to 6. The files were searched against the T. gondii proteome (March 2020, https://www.uniprot.org/proteomes/UP000005641-8450). Identified proteins were filtered according to the following criteria: at least 2 different trypsin peptides with at least 1 unique peptide, an E value below 0.01 and a protein E value smaller than 0.01 were required. Using the above criteria, the rate of false peptide sequence assignment and false protein identification were lower than 1%. Proteins were quantified by label-free method with MaxQuant software using unique and razor peptides intensities [81]. Statistical analyses were carried out using RStudio package software. The protein intensity ratio (protein intensity in mutant/protein intensity in parent) and statistical tests were applied to identify the significant differences in the protein abundance. Hits were retained if they were quantified in at least 3 of the 4 replicates in at least 1 experiment. Proteins with a statistically significant (p https://maxquant.net/perseus/). Perseus was also used for hierarchical clustering of bradyzoite- and tachyzoite-specific surface antigens of the SRS family using RNAseq data of Hehl and colleagues [82], available on www.Toxodb.org.

Parasites of the cKD HA-TgHCF101 transgenic cell line were treated for 3 days in the presence or absence of ATc in T175 cm² seeded with Vero cells. After treatment, intracellular parasites were released by scraping of the host cells and 3 passages through a 26G needle. To eliminate cell host debris, parasites were filtered through fiber glass wool and harvested by centrifugation at 650 g for 5 min and washed 3 times in DPBS (Gibco). Parasites were resuspended in lysis buffer (1%NP40, 50 mM Tris-HCl (pH 8), 150 mM NaCl, 4 mM EDTA, supplemented with cOmplete Mini protease inhibitors mix (Roche)) and incubated overnight at 4°C on a rotating wheel. Centrifugation of insoluble material was performed at 13,500 g for 30 min at 4°C. The supernatant was transferred to a tube containing 50 μl of anti-HA magnetic beads (Thermo Fisher, 88836) for 4 h at 4°C on a rotating wheel. The depleted fraction was then removed and beads were washed 5 times with lysis buffer. For the elution of immunoprecipitated proteins, beads were incubated in 50 μl of HA peptide solution at 2 mg/ml (Thermo Fisher, 26184) for 1 h at 37°C on a rotating wheel. The eluted proteins were mixed with Laemmli buffer and resolved on a 10% SDS-Page for 35 min at 100V. In-gel proteins were fixed with a combination of acetic acid and ethanol and stained with PageBlue Protein Staining Solution (Thermo Scientific). Each lane was cut in 3 identical pieces which were digested with trypsin, peptide extraction was done as previously described [80] and mass spectrometry identification was performed as described for label-free quantitative proteomic.

Proteomic data were analyzed on R following the publicly available script of DEP analysis package (https://github.com/arnesmits/DEP, v.1.7.1). The following parameters were used for differential expression analysis: data set was normalized by variance-stabilizing transformation and missing values were imputed via MNAR (missing not at random) method, the fold change cutoff was set at 1.5 and p-value at 0.05. Statistical analyses were performed by a differential enrichment test based on protein-wise linear models and empirical Bayes statistics [83], and p-values were adjusted by the Benjamini–Hochberg correction.

RNA was extracted from parasites using NuceloSpin RNA kit (Macherey-Nagel). cDNA was obtained using SuperScript III First Strand Synthesis SuperMix for RT-qPCR (Invitrogen) using oligodT and following the manufacturer’s instructions. Specific cDNAs were subsequently amplified using KOD DNA polymerase (Novagen). Cloning was performed with In-Fusion HD cloning kit (Takara) for each candidate into pGADT7 and pGBKT7 vectors (Clontech, Takara Bio) to allow expression of AD-(Gal4 activation domain) and BD-(Gal4 DNA binding domain) fusion proteins, respectively. Primers were designed using the InFusion Cloning Primer Design Tool (https://takarabio.com/) and are listed in S5 Table.

All experiments were performed in the Gal4-based yeast two-hybrid (Y2H) reporter strain YRG2 (MATα, ura3-52, his3-200, ade2-101, lys2-801, leu2-3, 112, trp1-901, gal4-542, gal80-538, lys2::UAS_GAL1-TATA_GAL1-HIS3, URA3::UAS_GAL4,_17MERS(x3)-TATA_CYC1-LacZ) (Stratagene, Agilent). Yeast cells were co-transformed by pGAD/pGBK construct pairs and selected on plates containing the minimal YNB medium (0.7% Yeast Nitrogen Base medium without amino acids, 2% glucose, 2% Difco agar) supplemented with histidine (H), adenine (A), Lysine (K), and uracil (U) according to Gietz and Woods [84]. For all Y2H interaction assays, individual co-transformed colonies were cultured on liquid YNB+HUKA, adjusted to an OD₆₀₀ of 0.05 and dotted on YNB+UKA plates containing (+His plates) or missing (-His plates) histidine (7 μl per dot). Plates were incubated at 30°C and cell growth was recorded after 5 days. Constructs pairs including empty vectors (AD-fusion + empty BD or empty AD + BD-fusion) were tested first to evaluate the capacity of each protein of interest to transactivate the HIS3 reporter gene in the absence of interactor and to determine the appropriate 3-aminotriazole (3AT) concentration required to abolish false interactions. True positive interactions were visualized as cells growing on YNB+UKA in the absence of histidine (-his plates) and eventually in the presence of 3AT if required. Increasing 3AT concentrations were also used to challenge the strength of positive interactions. For all assays, at least 3 independent transformations per construct pair were performed.

DNA content analysis by flow cytometry was performed as described previously [85]. Parasites from the cKD HA-TgHCF101 strain and from the background line TATi ΔKu80 were treated for up to 4 days in the presence or absence of ATc. Extracellular parasites were removed by washing with HBSS and intracellular parasites were released from host cell by scraping with a cell scraper (VWR, 734–2602), passages through 26G needles and finally filtered on a fiber glass wool column. Parasites were then fixed overnight in a solution of 70% ethanol and 30% PBS at 4°C. After fixation, parasites were washed in PBS and stained with a 30 μm propidium iodide solution for 30 min. Parasites DNA content was then analyzed by flow cytometry with an Aurora cytometer (Cytek) from the MRI facility. Raw FCS files and gating strategy for cytometry experiments are available at http://flowrepository.org/id/FR-FCM-Z8SR.

Parasites from the parental TATi ΔKu80 and the cKD HA-TgHCF101 transgenic cell lines were grown in the presence or absence of ATc for up to 4 days, and 1.10⁶ freshly egressed parasites were harvested and resuspended in 400 μl motility buffer (155 mM NaCl, 3 mM KCl, 2 mM CaCl₂, 1 mM MgCl₂, 3 mM NaH₂PO₄, 10 mM HEPES, 10 mM glucose) and 100 μl of the suspension was immediately placed on poly-L-lysine coated slides and incubated at 37°C for 15 min. Unattached parasites were removed by performing 3 washes with PBS. Finally, attached parasites were fixed with 4% (w/v) PFA and stained with anti-SAG1 antibody. Trails length (at least 100 per condition) were captured on random fields with a 63× objective on a Zeiss AXIO Imager Z2 epifluorescence microscope and were quantified using the NeuronJ plug-in on the ImageJ software as described previously [86].

Rhoptry secretion was quantified by performing an evacuole assay [87]. Parasites from the cKD HA-TgHCF101 strain and cKD TgARO strain were grown in the presence or absence of tetracycline for up to 72 h. Freshly egressed parasites were then pretreated for 10 min in DMEM supplemented with 1 μm cytochalasin D (cytD), an inhibitor of actin polymerization. Parasites were allowed to secrete their rhoptry content as they were put in contact with HFF cells for 15 min in the presence of cytD at 37°C. Parasites and HFF cells were then immediately fixed with 8% (w/v, in PBS) PFA, permeabilized with 0.1% Triton X-100 and stained with anti-ROP1 and anti-SAG1 to detect secreted rhoptry material and parasites, respectively. For each technical replicate, 20 random fields were quantified and 3 technical replicates were performed for each of the 3 biological replicates.

Statistical analyses were generally performed with the Prism 8.3 software (Graphpad). For proteomics experiments, statistics were performed with version 4.2.1 of the R package (2022-06-23, https://www.R-project.org/) using the Differential Enrichment analysis of Proteomics Data (DEP, v.1.23.0) package. Unless specified, values are expressed as means ± standard deviation (SD).