RNG2 is prone to proteolysis and forms oligomers in vitro

RNG2 codes for a 290 kDa protein that encompasses a central part annotated as tropomyosin domain (Tropo) within a long coil-coiled region flanked by disordered N- and C-termini (Fig 2A). To decipher the biochemical properties of RNG2, recombinant full-length and truncated RNG2 variants with an N-terminal His-Tag and a C-terminal TwinStrep-Tag were expressed in Sf9 insect cells (Fig 2A). All RNG2 constructs showed robust expression; however, only the ΔN+CTer variant exhibited good solubility (Fig 2B and 2C). Full-length RNG2 was completely insoluble, whereas the ΔNTer, ΔTropo, and ΔCTer variants showed partial solubility (Fig 2B and 2C). Western blot analysis with anti-His and anti-Strep antibodies further highlighted the presence of both N- and C-terminally truncated fragments for almost all the variants (Fig 2C). The ΔNTer and ΔN+CTer variants did not display any dominant C-terminally truncated fragment, with a single most-intense band detected with anti-His antibody corresponding to the predicted size of the full-length proteins (Fig 2C).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. Biochemical investigation of wild-type and processed RNG2.

(A) Representation of the RNG2 constructs expressed in Sf9 insect cells. (B) Insect cell protein expression and solubility of the RNG2 truncation variants, as assessed by Coomassie blue staining. (C) Protein expression and solubility of RNG2 truncation variants were assessed via western blotting. (D) Size-exclusion chromatography (SEC) chromatogram of the affinity-purified RNG2-ΔN+CTer variant obtained via Strep-tactin purification. A Coomassie-stained gel of the protein-containing fractions of the SEC is shown on the right. The data underlying this figure can be found in S2 Data. (E) Co-purification assays using metal-affinity chromatography of cells co-infected with RNG2 fragments assessed by western blot analysis and Coomassie blue staining. All fragments have a C-terminal STREP-tag.

https://doi.org/10.1371/journal.pbio.3003506.g002

Purification via the N-terminal His-Tag was not achievable for any of the constructs. In contrast, purification by pulling on the C-terminal TwinStrep-Tag proved successful for RNG2 constructs where the C-terminal 310 residues were truncated (ΔCTer and ΔN+CTer). However, purification required to use salt concentrations largely exceeding physiological conditions (800 mM NaCl). Affinity-purified proteins were subjected to size exclusion chromatography (SEC), and the fractions were analyzed by SDS‒PAGE, mass spectrometry (MS), and western blotting (WB). For both C-terminally truncated constructs (ΔCTer and ΔN+CTer), purification led to the isolation of the same three most-abundant products with sizes corresponding to 155, 90, and 60 kDa (Figs 2D and S1A). All the fragments coeluted in the void volume of the SEC, corresponding to molecule sizes greater than 600 kDa, indicating that all soluble fragments purified formed large oligomers (Figs 2D and S1A). WB analysis confirmed that the purified products had lost the N-terminal His-Tag (S1B Fig). This provides an explanation for the failure to isolate RNG2 fragments by using the N-terminal His-tag and strongly suggests that these protein constructs are susceptible to proteolytic cleavage. MS analysis confirmed that the purified material corresponded to RNG2 fragments and allowed mapping of the fragment boundaries (Figs 2D and S1C).

To isolate homogeneous RNG2 truncated proteins, each of the 155, 90, and 60 kDa fragments (encompassing residues 985–2290, 1529–2290, and 1793–2290, respectively) were individually expressed and purified via the same workflow as described earlier. Like the larger constructs, each fragment eluted in the void volume by SEC, indicating oligomerization (S1D Fig). To test whether fragments of different sizes could form hetero-oligomers, pull-down experiments were performed on cells co-infected with pairs of fragments of different sizes, where one of the fragments was affinity-purified via a unique N-terminal His-tag (Fig 2E). Purification of either the 155 kDa or the 60 kDa fragment did not result in co-elution with the 90 kDa fragment, indicating that the oligomers observed are likely homo-oligomers. Since homo-oligomerization is thermodynamically favorable [25], the fragments that co-eluted during the SEC of purified ΔCTer and ΔN+CTer constructs likely did not assemble into hetero-oligomers (Fig 2D and 2E).

Collectively, heterologous expression of RNG2 provided evidence that the protein is highly unstable and susceptible to rapid proteolytic degradation. Only a mixture of truncated RNG2 products of the same size could be purified, which were associated with large homo-oligomers containing fragments of 155, 90, and 60 kDa.

RNG2 decorates 22 tethers between the conoid and the APR

Double epitope tagging at both extremities of RNG2 previously revealed that the N-terminal tag stains the conoid, whereas the C-terminal tag labels the APR [23]. To resolve this conundrum and scrutinize the biochemical properties of RNG2 in T. gondii, we introduced epitope tags at the N-terminus (Myc), center (Ty), and C-terminus (HA), along with a C-terminal mAID cassette (Fig 3A). Prediction of the maximum length of RNG2 suggested that its coiled-coil regions could span approximately 215 nm, consistent with the distance between the conoid and the APR (Fig 3A). The endogenous triple epitope-tagged RNG2 clonal line was confirmed by genomic PCR (S2A Fig) and WB analysis (Fig 3B). Indirect immunofluorescence assay (IFA) using the three antibodies (Myc, Ty, HA) confirmed the apical localization of the RNG2 as well as tight regulation under auxin (indole-3-acetic acid [IAA]) treatment (S2B Fig). Several RNG2 truncated products of high molecular weight were detected with anti-HA and anti-Ty antibodies. The specificity of those bands was confirmed by their disappearance in parasites treated with IAA. However, the anti-Myc antibody detected only the full-length version at a high molecular weight with additional smaller fragments around 100 kDa, which is consistent with N-terminal processing events (Fig 3B). In addition, the use of RNG2 rabbit polyclonal antibodies raised against the recombinant central fragments of RNG2 (155, 90, 60 kDa) (Fig 2D), revealed the same processing events (Fig 3B). By comparing the WB obtained from insect cells expressed variants and the ones using all the epitopes-tags in T. gondii, using the sizes of the processed products the map we could determine that RNG2 displays five main cleavage sites that are conserved between insect cells and the parasite (S3A Fig). Moreover, in T. gondii neither the full length nor processed products could be observed soluble using the three epitope tags (Myc-Ty-Ha) (S3B Fig). Interestingly, the extensive proteolytic cleavage of RNG2 observed in the parasite did not differ between intracellular and extracellular states, indicating that it is constitutive and not dependent on conoid dynamics (Fig 3C). To refine the dynamic behavior of RNG2, the position of the different tags was assessed via U-ExM on extracellular tachyzoites on extruded or retracted conoids (Fig 3D). The N-terminus of RNG2 (Myc) was found at the base of the conoid both in extruded and retracted states, whereas the C-terminus (HA) was always localized on top of the SPMTs at the APR level, confirming the previous observation [23]. The central Ty-tag localized between the conoid and the SPMTs, as did the anti-RNG2 antibodies raised against the central region of RNG2. Colocalization of the Ty- and anti-RNG2 tags with HA- and Myc-tags confirmed that RNG2 is oriented with its N-terminal region at the conoid, its central region spanning the space between the conoid and the APR, and its C-terminal region at the APR (S3C Fig). Notably, Ty-Tag colocalized with the N-terminal Myc-Tag only when the conoid was retracted, suggesting a shortening of the protein length in the retracted state (S3D Fig).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 3. RNG2 spans between the APR and the conoid, forming 22 tethers.

(A) Illustration of endogenous tagging of RNG2, including three epitope tags and a minimal auxin inducible (mAID) motif. The estimation of the protein length corresponding to the distance coverage of the central part of the protein in linear conformation. (B) Protein depletion after 48 h of auxin (IAA) treatment assessed by western blot. (C) Western blot of parasite lysate collected in intracellular or extracellular state (D) U-ExM image of the locations of RNG2-mAID depending on the tag insertion and RNG2 polyclonal antibody (recognizing the central region). Images were taken with extruded and retracted conoids. Scale bar = 3 μm. (E) Iterative expansion microscopy (iU-ExM) of the N-terminal (Myc), middle (Ty), and C-terminal (HA) tags of RNG2-mAID. Scale bar = 3 μm. Right panel: Distribution of the numbers of dots counted for the anti-HA staining across 13 images. The data underlying this figure can be found in S2 Data. (F) Left panel: Electron microscopy images of the Toxoplasma gondii apical complex as an ultrastructure control. Right panel: Immunoelectron microscopy images obtained with the α-GAC and α-RNG2 antibody. Red arrows = gold particles. Scale bar = 100 nm.

https://doi.org/10.1371/journal.pbio.3003506.g003

To achieve higher resolution, we employed iterative ultrastructure expansion microscopy (iU-ExM), which enables a 16-fold expansion compared with the 4-fold expansion achieved by conventional U-ExM [26]. With increased resolution analysis based on a dozen of images, we observed that the RNG2 C-terminus exhibited approximately 22 puncta aligning with the tips of the SPMTs (Fig 3E). Additionally, around 22 puncta could sometimes be detected at the RNG2 N-terminus on the conoid, although these were less distinct. (Fig 3E).

A study based on electron microscopy (EM) ultrastructural analysis revealed the existence of BFs organized in a cartwheel-like conformation and linking the conoid to the SPMTs [24]. To test whether RNG2 could be a component of BFs, we performed IEM using rabbit anti-RNG2 antibodies (Fig 3F). Although BFs were readily detected in the control samples (Fig 3F), when processed for IEM, many of the BFs disappeared, likely due to the numerous washes and incubations with the antibodies. However, a large accumulation of gold labels was found in the space between the conoid and APR and on the BFs still visible in the preparation (Fig 3F). As a control, we used rabbit polyclonal anti-GAC antibodies, which showed accumulation of gold particles at the PCRs where GAC is localized [11] (Fig 3F). Taken together, the iU-ExM and IEM data showing the spread of RNG2 between the APR and the conoid support its identification as a component of the 22 BFs.

RNG2 is essential for maintaining conoid attachment to the APR during retraction

Given biochemical and imaging evidence for a role of RNG2 in conoid-APR attachment, we revisited its function, taking advantage of the fast depletion of the protein with the auxin degron system. RNG2 depletion by IAA, as well as deletion of the gene Δrng2, were confirmed to be fitness-conferring by plaque assays (S4A Fig). In intracellular parasites, no obvious cytoskeletal difference was observed in RNG2-depleted parasites under U-ExM (Fig 4A). Conoid extrusion can be triggered by artificial elevation of cGMP levels with BIPPO, a phosphodiesterase inhibitor [27], and unambiguously observed by U-ExM and EM [11,27,28]. Strikingly, in BIPPO-stimulated extracellular RNG2-depleted parasites, the conoid detached from the cell apex and was observed floating in the cytosol by both U-ExM and EM (Fig 4B and 4C). Importantly, conoid detachment was prominently observed in more than 80% of BIPPO-stimulated extracellular parasites. To determine whether actomyosin-driven conoid extrusion triggers conoid detachment, we performed a conoid extrusion assay using parasites pretreated with cytochalasin D (CD), an inhibitor of actin polymerization [11]. Under normal conditions, BIPPO stimulation resulted in more than 84% of parasites exhibiting conoid extrusion, whereas treatment with CD nearly abolished extrusion, with 95% of parasites retaining a retracted conoid (Fig 4D). In contrast, only 9% of BIPPO-stimulated RNG2-depleted parasites showed conoid extrusion, while 77% exhibited conoid detachment. Notably, CD treatment of RNG2-depleted parasites preserved a retracted conoid phenotype, similar to the parental strain (Fig 4D). These results indicate that conoid detachment caused by RNG2 depletion occurs only upon stimulation. EM imaging revealed that the detached conoid remained intact, with PCRs and ICMTs still associated with the tubulin cone (Fig 4C). Additionally, the electron-dense structure capping the 22 SPMTs known as the APR remained unaffected under EM imaging (Fig 4C).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 4. RNG2 depletion leads to conoid detachment from the apical pole of the parasite.

(A) U-ExM images of the tubulin cytoskeleton of RNG2-mAID parasites treated for 30 h with auxin showing no cytoskeletal defects. White arrows: conoids (B) U-ExM images showing extruded conoid and detached conoid in RNG2-mAID-depleted parasites and RNG2 knockout parasites (Δrng2). Scale bar = 3 μm. (C) Electron microscopy images of negatively stained tachyzoites with extruded conoids and detached conoids in RNG2-mAID-depleted parasites and Δrng2 parasites. Scale bar = 500 nm. (D) Quantification of the conoid state in three categories: extruded, retracted, and detached. Under all conditions, extracellular tachyzoites were triggered for conoid extrusion via BIPPO in the absence or presence of auxin (IAA) and with or without cytochalasin D (CD). N = 100 parasites in 3 biological replicates. The data underlying this figure can be found in S2 Data. (E): Cryo-electron tomography performed on RNG2-mAID parasites with detached conoid. Upper panel shows two parasites with completely detached conoids. Lower panel shows a conoid detached and falling towards the cytoplasm. (F): Quantification of the presence of the apical vesicle (AV) in the tomograms of parasites depleted or not for RNG2. n = 23. The data underlying this figure can be found in S2 Data.

https://doi.org/10.1371/journal.pbio.3003506.g004

To gain further insight into the consequences of RNG2 depletion, cryo-electron tomography (Cryo-ET) was performed on parasites with detached conoids (Fig 4E). While all components of the apical complex were correctly positioned in untreated parasites, the apical vesicle (AV), a plasma membrane-embedded structure essential for rhoptry discharge and invasion, was the only structure that remained correctly positioned in conoid-less parasites (Figs 4E, 4F, S5A and S5B and S1–S3 Movies). Interestingly, we captured the apical end of a parasite with a falling conoid, showing that conoid-associated components such as ICMTs, MVs, and rhoptry necks moved with the conoid while remaining correctly positioned within it (Fig 4E and S3 Movie). In addition, consistent with the positioning of the AV, no difference in localization was observed for the Nd6 protein previously localized at the rosette [29] (S5C and S5D Fig). Taken together, these findings demonstrate that RNG2 is essential for retaining the conoid at the apical pole after extrusion.

Conoid detachment in the absence of RNG2 disrupts motility and rhoptry positioning

To investigate quantitatively and at the molecular level the consequences of the conoid detachment on the structures of the apical complex, markers of the conoid (CPH1), the PCRs (Pcr4), and the APR (APR1) were epitope-tagged in the background of the RNG2-mAID strain [11,19,30] (Fig 5A). In parasites detaching their conoid, the three components remained correctly positioned, suggesting no alteration of the structures but simply a disconnection (Fig 5A).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 5. Phenotypic consequences of the conoid detachment.

(A) U-ExM images of CPH1 (conoid marker) and APR1 (APR marker) with the conoid detached from the APR. Scale bar = 3 μm. (B) U-ExM of secretory organelles, with MIC2 used as a microneme marker and RON9 used as a rhoptry neck marker. Images taken with the conoid attached or detached from the APR. Scale bar = 3 μm (C) Western blot analysis of ethanol-triggered microneme secretion. ESA, Excreted secreted antigens. (D). Representative images of BIPPO-induced egress. Scale bar = 20 μm. (E) Left panel: representative images of STAT6-P-positive cells with a Tir1 strain. Right panel: Quantification of rhoptry secretion was performed to assess STAT6 phosphorylation. At least 200 cell nuclei were counted for data plotting, n = 3. ATc = anhydrotetracycline. The data underlying this figure can be found in S2 Data.

https://doi.org/10.1371/journal.pbio.3003506.g005

T. gondii tachyzoites possess eight to 12 rhoptries, with two of them having their necks docked inside the conoid and primed for discharge [13]. Micronemes similarly accumulate at the apical pole of the parasite [31]. Since the conoid is assumed to serve as funnel for the secretion of micronemes and rhoptries, we assessed the impact of conoid detachment on these processes. U-ExM revealed no striking disorganization of the micronemes as they could accumulate apically; however some micronemes were still associated to the detached conoid, pointing to the existence of physical interactions between the conoid and these organelles (Fig 5B). More strikingly, the rhoptries were considerably disorganized in RNG2-depleted parasites (Fig 5B). Consistent with these observations, induced microneme secretion showed only a partial reduction of exocytosis following conditional depletion of RNG2 (Figs 5C and S4B). The level of micronemes secreted allowed RNG2-depleted parasites to successfully egress, lysing both the parasitophorous vacuole membrane (PVM) and host cell plasma membrane upon BIPPO stimulation (Figs 5D and S4C). However, RNG2-depleted parasites remained clustered at the site of host cell lysis, indicating an impairment in motility. This was confirmed by a severe defect in the gliding trails assay (S4D Fig). Because proper rhoptry positioning is crucial for discharge, we assessed rhoptry secretion in RNG2-deficient parasites. Rhoptry secretion was measured indirectly, by monitoring the phosphorylation of host cell nuclear STAT6, a marker for rhoptry discharge [32]. In aspartyl protease 3-depleted parasites, used here as a control, rhoptry secretion is completely blocked [33], whereas RNG2-depleted parasites show a ~50% reduction in rhoptry discharge, comparable to that observed in the Δrng2 mutant (Fig 5E). Consequently, depletion of RNG2 resulted in a reduction of invasion efficiency to less than half of that seen in the parental strain (S4E Fig).

To conclude, the disconnection of the conoid from the APR led to detachment of rhoptries and a small portion of microneme from the apical tip, resulting in severe defects of rhoptry discharge as well as a modest defect of microneme secretion. While the invasion capacity of parasite is adversely affected by the failure of rhoptry discharge, the minor effect of microneme secretion does not impair the rupture of the PVM or the host cell. In RNG2-depleted parasites, conoid detachment, including the PCRs, disrupted apical F-actin polymerization and mislocalized GAC, preventing glideosome assembly and causing a profound motility defect.

Domain mapping of RNG2 interfaces with the conoid and APR

To investigate the regions of RNG2 involved in conoid tethering, the Δrng2 parasite line was complemented with truncated versions of RNG2 expressed as a second copy under the control of the tubulin promoter (Fig 6A). WB analysis confirmed that each variant was expressed at the expected size and revealed processed forms similar to those observed for the endogenous protein (Fig 6B). IFA and U-ExM analyses of all truncated proteins confirmed their localization to the APR, with an additional cytoplasmic signal likely resulting from overexpression (Figs 6C and S6A). Plaque assays assessing the fitness of the complemented lines showed that the WT, ΔNTer, and ΔTropo variants fully restored parasite fitness after deletion of endogenous RNG2 (Figs 6D and S6A). In contrast, the ΔCTer and ΔN+CTer variants only partially restored parasite fitness, indicating that the C-terminal region of RNG2 is critical for its physiological function, whereas the N-terminal region and the central tropomyosin-like domain appear dispensable (Figs 6D and S6A). Although the ΔCTer- and ΔN+CTer- complemented strains displayed a mild invasion defect, the WT and ΔTropo strains fully rescued the invasion phenotype. The ΔNTer variant, however, showed a slight but significant reduction in invasion (~10%) (Fig 6E).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 6. Investigation of RNG2 subdomain functions.

(A) Representation of the RNG2 variants expressed in the Δrng2 background. (B) Western blot analysis of the expression of RNG2 variants expressed in the Δrng2 background. (C) U-ExM images of the localization of the RNG2 variants. Scale bar = 1 μm. (D) Quantification of the plaque sizes created by the different RNG2 strains via a plaque assay. Ten plaques were measured per biological replicate. Three biological replicates were evaluated. The data underlying this figure can be found in S2 Data. (E) Quantification of the percentage of invasion by the different RNG2 strains in the invasion assay. One hundred parasites were measured per biological replicate. Three biological replicates were evaluated. The data underlying this figure can be found in S2 Data. (F) Quantification of the percentages of conoid retraction, extrusion, and detachment by the different RNG2 strains in the conoid detachment assay. One hundred parasites were measured per biological replicate. Three biological replicates were evaluated. The data underlying this figure can be found in S2 Data. (G) U-ExM images of the localization of the RNG2 variants under conditions in which the conoid detached from the apical pole. Scale bar = 1 μm. (H) Representation of the second round of RNG2 variants expressed in the Δrng2 background. (I) Western blot analysis of the expression of the second round of RNG2 variants expressed in the Δrng2 background. (J) Quantification of the plaque sizes created by second-round RNG2 variants via a plaque assay. Ten plaques were measured per biological replicate. Three biological replicates were evaluated. The data underlying this figure can be found in S2 Data. (K) Quantification of the percentage of invading cells in response to the second round of RNG2 overexpression via an invasion assay. One hundred parasites were measured per biological replicate. Three biological replicates were evaluated. The data underlying this figure can be found in S2 Data. (F) Quantification of the percentage of conoid retraction/extruded/detaching by the second round of the RNG2 variant in the conoid detachment assay. One hundred parasites were measured per biological replicate. Three biological replicates were evaluated. (L) U-ExM images of the localization of the second round of RNG2 variants under conditions in which the conoid detached from the apical pole. Scale bar = 1 μm. The data underlying this figure can be found in S2 Data.

https://doi.org/10.1371/journal.pbio.3003506.g006

To assess the ability of each construct to correctly maintain the conoid at the apical pole, conoid detachment was assessed by U-ExM after activation with BIPPO. The wild-type and ΔTropo variants successfully complemented the depletion of endogenous RNG2, resulting in very few detached conoids. In contrast, 50% of conoid detachment was scored in ΔCTer and ΔN+CTer variant-complemented strains (Fig 6F). Although the ΔNTer strain did not exhibit a significant fitness defect, approximately 20% of parasites displayed a detached conoid, correlating with the mildly impaired invasion phenotype (Fig 5F). Following conoid extrusion, the protein variants either remained tethered to the APR, as observed with the ΔNTer variant, or were associated with the floating conoid, as seen in the ΔCTer and ΔN+CTer strains. These observations confirm that the N-terminal region of RNG2 associates with the conoid, while the C-terminal region mediates interaction with the APR (Fig 6G).

Two additional RNG2 variants were engineered to investigate the internal coiled-coil region: one with an extended central region containing six tropomyosin domain repeats (RNG2-6Tropo), and another shorter variant in which most of the internal sequence was deleted (RNG2-ΔCenter) (Fig 6H). Both variants were dually tagged with Myc-tags and Ty-tags at their N- and C-terminal ends, respectively. WB analysis confirmed that both variants were expressed correctly and exhibited several processed forms, particularly in the extended RNG2 protein (Fig 6I). Comparison of the blots indicated the presence of several fragments of identical size between the two variants, suggesting that a major N-terminal fragment of approximately 110 kDa and two key C-terminal fragments of 60 and 40 kDa were produced (Fig 6I). Both variants accumulated at the apical pole, as shown by IFA (S6B Fig). While the 6Tropo variant fully complemented the Δrng2 strain, the ΔCenter strain failed to do so, leading to conoid detachment (Figs 6J–6L and S6B). The 6Tropo variant dually localized to the conoid and APR (Fig 6M). In contrast, RNG2-ΔCenter localized to both the conoid and APR when detected with antibodies against either the N- or C-terminal regions, suggesting that the shortened protein is too short to span the gap between the two structures (Fig 6M). These results indicate that while the N- and C-terminal extremities of RNG2 are sufficient for binding the conoid and APR, respectively, the central region is essential for tethering the conoid at the apical pole.

Noting that three main truncation products of RNG2 during purification from insect cells, we asked whether these isoforms might be physiologically relevant. The three truncated variants (155, 90, and 60 kD) were expressed in Δrng2 strain to assess their function (Fig 7A). All three variants bear the epitope tags both at N-terminal (Myc) and C-terminal (Ty), and WB analysis confirmed the expression of three variants (Fig 7B), located at apical region of the parasites. While the 155 kD variant exhibited dotted accumulated signal in the cytosol similar to the wild-type, expression of 90 and 60 kD smeared within the cytosol (S6C Fig). U-ExM revealed that all three truncated variants localized to the APR, with their N- and C-termini colocalized, unlike the dual localization observed for endogenous RNG2 (Fig 7C). Functional assessment by plaque assay (Figs 7D and S6C), invasion assay (Fig 7E), and conoid anchoring (Fig 7F) showed that none of the three isoforms could complement the loss of endogenous RNG2.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 7. RNG2 is functional only as intact, full-length protein.

(A) Representation of the RNG2 three processed products originally identified in insect cells (155kDa, 90kDa, 60kDa) expressed in the Δrng2 background. (B) Western blot analysis of the expression of RNG2 variants expressed in the Δrng2 background. (C) U-ExM images of the localization of the RNG2 variants. Scale bar = 3 μm. (D) Quantification of the plaque sizes created by the different RNG2 strains via a plaque assay. Ten plaques were measured per biological replicate. Three biological replicates were evaluated. The data underlying this figure can be found in S2 Data. (E) Quantification of the percentage of invasion by the different RNG2 strains in the invasion assay. One hundred parasites were measured per biological replicate. Three biological replicates were evaluated. The data underlying this figure can be found in S2 Data. (F) Quantification of the percentages of conoid retraction, extrusion, and detachment by the different RNG2 strains in the conoid detachment assay. One hundred parasites were measured per biological replicate. Three biological replicates were evaluated. The data underlying this figure can be found in S2 Data. (G) Design of the experiment to test if different fragments of RNG2 can cooperate to form a functional tether. (H) Conoid detachment assessed by U-ExM in transiently transfected parasites treated for 48h under auxin to down-regulate the endogenous RNG2-mAID. The data underlying this figure can be found in S1 Data. (I) Representation of the skip peptide (P2A) variants expressed in the RNG2-mAID-HA strain. (J) Western blot analysis of the expression and skipping pattern verifying the successful skipping of ribosome on the P2A sequence. (K) U-ExM images of auxin treated parasite to deplete the endogenous RNG2, showing the localization of both P2A variants and their ability to maintain the conoid at the apical pole. (L) Quantification of the plaque sizes created by RNG2 P2A variants via a plaque assay. Ten plaques were measured per biological replicate. Three biological replicates were evaluated. The data underlying this figure can be found in S2 Data.

https://doi.org/10.1371/journal.pbio.3003506.g007

RNG2 needs to be intact to tether the conoid at the apical pole

Given the extensive proteolytic processing of RNG2, both in a heterologous system and in the parasite, and its tendency to oligomerize, we hypothesized that the processed products might form oligomers in the parasite and provide flexible length to accommodate rapid distance changes during conoid extrusion and retraction. To test whether different fragments of RNG2 can interact and thereby restore function, RNG2-mAID-HA parasites stably expressing the RNG2-ΔCenter variant were transiently transfected with the 155 kDa fragment that spans the deleted region (Fig 7G). After 48 h of IAA treatment to deplete endogenous RNG2-mAID-HA, conoid extrusion was induced using BIPPO, and the proportion of conoids correctly positioned at the apical pole was assessed. As previously shown, ~80% of conoids were detached in the RNG2-mAID-HA strain, while only ~20% were detached when a full-length copy of RNG2 was transiently expressed (Fig 7H). As observed earlier (Fig 7F), transient expression of the 155 kDa fragment alone did not rescue conoid anchoring, with ~80% of conoids remaining detached. In contrast, transient transfection of full-length RNG2 into the RNG2-mAID-HA strain stably expressing RNG2-ΔCenter drastically reduced the number of detached conoids. These results indicate that the fragments do not functionally cooperate or physically interact in a way that restores RNG2 function, similar to what was observed in vitro.

As a second approach to assess whether RNG2 proteolytic processing is functionally important, we designed a construct in which a P2A “skip” peptide [34,35] was inserted within the tropomyosin domain of RNG2. This insertion causes ribosomal skipping during translation, leading to expression of two separate RNG2 fragments (RNG2-P2A WT). In parallel, we generated a control construct in which key residues of the P2A sequence were mutated to prevent skipping [34], thereby preserving RNG2 as a single polypeptide (RNG2-P2A MUT). Both constructs were stably expressed in the RNG2-mAID-HA line to assess whether the co-expressed fragments could functionally complement RNG2 depletion (Fig 7I). WB analysis confirmed co-expression of both N- and C-terminal fragments from the RNG2-P2A WT construct, while the RNG2-P2A-MUT remained unsplit and displayed the same proteolytic processing as the endogenous RNG2 (Fig 7J). Upon BIPPO-induced conoid extrusion, the strain expressing RNG2 P2A-WT exhibited detachment of the conoid from the apical pole. The N-terminal fragment of RNG2 localized to the displaced conoid, while the C-terminal fragment remained associated with the APR (Fig 7K). In contrast, in the strain expressing the RNG2 P2A-MUT, the protein localized between the conoid and APR and rescued the phenotype of conoid detachment (Fig 7K). In line with this observation, the RNG2-P2A-WT failed to complement by plaque assay, while the RNG2-P2A-MUT fully complemented the depletion of the endogenous protein (Fig 7L). Collectively, these results demonstrate that full-length RNG2 is required to tether the conoid at the apical pole.

Parasite maintenance

T. gondii tachyzoites were amplified from HFFs (ATCC) in Dulbecco’s modified Eagle’s medium (DMEM, Gibco) supplemented with 5% fetal calf serum (FCS, Gibco), 2 mM glutamine, and 25 µg/ml gentamicin (Gibco). Parasites and HFFs were maintained at 37 °C with 5% CO₂.

Generation of transgenic cell lines

Baculovirus-infected Insect cell expression and purification

All recombinant RNG2 proteins were expressed in Sf9 insect cells and purified following a similar protocol. All the plasmids encoding RNG2 possessed an N-terminal His10-tag followed by a TEV recognition sequence and an HRV-3C recognition sequence, followed by a TwinSTREP tag. Baculoviruses were generated following standard procedures, and the cells were infected for 48 h at 27 °C in a shaking incubator. Purifications were performed via a combination of STREP-tactin affinity purification and size-exclusion chromatography. Briefly, a pellet from 600 ml of cells was resuspended in 120 ml of lysis buffer (PBS supplemented with 800 mM NaCl, 1 mM EDTA, 0.1% Triton X-100, and 3 mM beta-mercaptoethanol) and lysed by shear force using a LM-20 microfluidizer set at 20,000 psi. The lysate was centrifuged at 35,000g for 35 min at 4 °C, and the supernatant was applied to a 5 ml STREP-TACTIN XT 4Flow column (IBA) at 0.5 ml/min. The column was washed with 50 ml of lysis buffer and eluted with 15 ml of lysis buffer supplemented with 50 mM biotin. The eluted protein was concentrated to 1 ml via a 30 MWCO Amicon concentrator and loaded on a Superdex 200 10/300 GL column at 22 °C previously equilibrated in PBS supplemented with 3 mM DTT. Fractions were analyzed via SDS‒PAGE, pooled, and flash frozen in liquid nitrogen before being stored at −80 °C.

Expansion microcopy (U-ExM)

For this study, the expansion microscopy protocol applied to T. gondii tachyzoites was followed as previously described [28]. Briefly, parasites resuspended in warm PBS containing 10 μM BIPPO were seeded on 12 mm coverslips precoated with poly-D-lysine (Gibco) for 10 min. After excess PBS was removed, the coverslips were incubated in PBS containing 0.7% formaldehyde and 1% acrylamide for 3 h at 37 °C. Polymerization of the expansion gel was performed on ice and contained a monomer mixture (19% sodium acrylate/10% acrylamide/0.1% bis-acrylamide), 0.5% ammonium persulfate, and 0.5% tetramethylenediamine. Fully polymerized gels were denatured at 95 °C for 90 min in denaturation buffer (200 mM SDS, 200 mM NaCl, 50 mM Tris, pH = 9) and expanded in pure H₂O overnight.

The next day, the expansion ratio of the fully expanded gels was determined by measuring the diameter of the gels. The well-expanded gels were shrunk in PBS and stained with primary and secondary antibodies diluted in freshly prepared 2% PBS/BSA at 37 °C for 2 h. Three washes with PBS/0.1% Tween for 10 min were performed after primary and secondary antibody staining. The stained gels were expanded again in pure H₂O overnight for further imaging. All U-ExM images used in this study were acquired via Leica TCS SP8 microscope with the lens HC PL Apo 100×/1.40 Oil CS2. Images were taken with Z-stacks and deconvolved with the built-in setting of Leica LAS X. Final images were processed with ImageJ, and the maximum projected images are presented in this study.

Iterative Ultrastructure Expansion Microscopy (iU-ExM)

Iterative ultrastructure expansion microscopy was applied as described [26]. Briefly, parasites resuspended in warm PBS containing 10 μM BIPPO were seeded on 12 mm coverslips precoated with poly-D-lysine (Gibco) for 10 min. After excess PBS was removed, the coverslips were incubated in PBS containing 0.7% formaldehyde and 1% acrylamide for 3 h at 37 °C. The first monomer solution was added to fill in the space of the gelation chamber. Following an incubation of 15 min on ice, an incubation at 37 °C for 45 min was completed. The formed gels were transferred into the denaturation buffer and incubated at 85 °C for 90 min. The gels were placed in water to expand 5–6 times their size and were shrunk with PBS to apply primary and secondary antibodies. The gels were bathed in water again and were cut into pieces that were incubated 3 times for 10 min on an agitating platform with activated neutral gel. The gels were dried by sliding it onto the microscope slide. A coverslip was placed on top, and the chamber was incubated at 37 °C for 1 h. The gel was bathed in the combination of acrylamide and formaldehyde for overnight at 37 °C under agitation. The following day, the gel was washed twice with 1× PBS for 30 min before applying the 3rd monomer solution to achieve an expansion factor of approximately 16×. The gel was washed 3 times for 10 min under agitation on ice and was dried on the microscope slide. A coverslip was placed on top within a chamber and incubated for 1 h at 37 °C. The gel was bathed in 200 mM NaOH solution for 1 h under agitation at RT and was washed in 1× PBS for 20 min until the pH reaches 7. The last round of expansion was carried out with water, and cells were imaged using Leica TCS SP8 microscopy with the lens HC PL Apo 100×/1.40 Oil CS2. Images were taken with Z-stack and deconvolved with the built-in setting of Leica LAS X or with Huygens software. Final images were processed with ImageJ, and the maximum projected images were presented in this study.

Electron microscopy

Extracellular tachyzoites were prepared in the same manner as previously described [28]. Briefly, the extracellular parasites were pelleted in PBS. Conoid extrusion was induced by incubation with 40 µl of BIPPO in PBS for 5 min at 37 °C. A 4 µl sample was applied to a glow-discharged 200-mesh Cu electron microscopy grid for 10 min. The excess sample was removed by blotting with filter paper and immediately washed three times in double-distilled water. Finally, the sample was negatively stained with a 0.5% aqueous solution of phosphotungstic acid for 20 s and air-dried. Electron micrographs of parasite apical poles were collected with a Tecnai 20 transmission electron microscope (FEI, the Netherlands) operated at an acceleration voltage of 80 kV and equipped with a side-mounted CCD camera (MegaView III, Olympus Imaging Systems) controlled by iTEM software (Olympus Imaging Systems).

Immunoelectron microscopy

The isolation of the cytoskeleton from tachyzoites of T. gondii was performed as previously reported [24]. Briefly, a suspension of tachyzoites (1 × 10⁶/ml) suspended in PHEM solution (10 mM HEPES, 10 mM EGTA, 1 mM MgCl₂, 50 μg/ml N-Tosyl-L-phenylalanine chloromethyl ketone, 50 μg/ml Nα-p-Tosyl-L-lysine chloromethyl ketone, and 17.4 μg/ml phenylmethylsulfonyl fluoride, pH 6.9) was exposed to 0.1% Triton X-100 in PHEM for 1 min at RT. The cytoskeleton fraction was centrifuged at 200,000g for 15 min at 4 °C. Under such conditions, most of the cytoskeletons were integrated into the typical crescent shape of tachyzoites. To isolate the cytoskeletons in a cartwheel-like conformation, tachyzoites were exposed to Triton X-100 solution for 3 min. For IEM, cytoskeletons were deposited on Ni grids covered with a formvar film (Polysciences, Warrington, PA, USA), fixed with 1% PAF in PHEM, blocked with 0.1% BSA, and then incubated for 2 h with anti-RNG2 (1:100 dilution) diluted in PHEM inside a humid chamber at RT. After washing, the samples were incubated for 2 h with IgG goat anti-rabbit IgG antibodies (ZYMED Laboratories, San Francisco, CA) coupled to 10 nm colloidal gold particles diluted in PHEM, washed with PHEM, counterstained with uranyl acetate, and examined via TEM. As a negative control, the cytoskeleton was incubated with preimmune serum and then with a secondary antibody coupled to gold particles.

Cryo-electron tomography

Freshly egressed parasites pretreated 48 h ±IAA were harvested and washed with PBS. Conoid extrusion was induced by incubating the parasites in warm 10 µM of BIPPO in PBS for 10 min. Formaldehyde 4% was then used to fix the samples. Fixation was quenched with PBS-Glycine, washed with PBS, and kept on ice. 3 µL of sample (~4 × 10⁶ tachyzoites) were loaded on EM grid for plunge freezing into a liquid ethane on an EM GP2 automatic plunger (Leica Microsystems, Wetzlar, Germany) after 4–5 sec front blotting. Images were collected on a Thermo Fisher 300 kV Titan Krios G4 Cryo-TEM paired with the Thermo Scientific Falcon 4 Direct Electron Detector. Tilt-series were collected with a dose-symmetric scheme ranging from −55° to +55° with 2° increments at a magnification of 53000X, a pixel size of 2.42 Å, and a defocus range of −6 to −8 µm. The cumulative dose of each tilt-series was 140 e⁻/Ų. Tilt series were motion corrected and CTF estimated using Warp, and tomograms aligned and reconstructed with AreTomo software [49,50]. Segmentation was manually generated in IMOD software, and videos were created with UCSF ChimeraX [51,52]. Quantifications were done in IMOD. In total, 23 tomograms were analyzed for each condition.

Plaque assay

HFF monolayers were infected with a serial dilution of T. gondii tachyzoites and grown for seven days at 37 °C. The cells were fixed with paraformaldehyde-glutaraldehyde for 10 min, followed by neutralization with 0.1 M PBS/glycine. The fixed monolayer was then stained with crystal violet for 2 h and then washed three times with PBS.

Conoid extrusion assay

To induce conoid extrusion, freshly egressed tachyzoites were incubated with 10 µM BIPPO for 10 min at room temperature. The parasites were then transferred to poly-D-lysine-coated coverslips, followed by the ultrastructure expansion microscopy protocol as described previously. The conoid status of a minimum of 100 parasites was determined for quantification. The data presented in this study were acquired from three biological replicates.

Invasion assay

Coverslips covered with the HFF monolayer were infected with T. gondii tachyzoites, centrifuged at 1,200g for 1 min, and incubated at 37 °C for 30 min. Infected HFFs were fixed with paraformaldehyde-glutaraldehyde and neutralized with PBS/0.1 M glycine. The fixed samples were blocked with 5% PBS-BSA for 20 min at room temperature, followed by 1 h of staining with anti-SAG1 antibody and three washes in PBS. The stained samples were fixed again with 1% formaldehyde for 7 min and permeabilized with PBS/0.2% Triton X-100 for 20 min. The samples were then stained with anti-GAP45 antibody and secondary antibodies. At least 100 parasites per condition were counted to determine the invasion ratio. The data presented in this study were from three biological replicates.

Microneme secretion assay

Freshly egressed parasites (syringed out in the case of parasites with severe egress defects) were washed twice in warm DMEM, pelleted, and resuspended in media containing 10 μM BIPPO. The suspensions were incubated at 37 °C for 15 min and centrifuged at 2,000g to allow separation of the pellet/supernatant (ESA) fraction. The isolated ESA fractions were further centrifuged to remove any remaining cell debris. All the samples were subjected to western blot analysis with anti-MIC2, anti-MIC6, anti-AMA1, anti-Catalase, and anti-GRA1 antibodies.

Egress assay

Fresh tachyzoites were used to infect HFF on coverslips, which were subsequently cultured for 30 h at 37 °C. Infected cultures were incubated with DMEM containing BIPPO (10 µM) for 10 min at 37 °C, followed by PFA/Glu fixation and neutralization with 0.1 M PBS/glycine. The coverslips were stained as described previously with the anti-GAP45 antibody and anti-GRA3. At least 100 vacuoles per condition were counted. The data presented in this study were acquired from three biological replicates.

Gliding trail assay

Freshly egressed tachyzoites were resuspended in warm DMEM containing 10 μM BIPPO, seeded on 12 mm coverslips coated with poly-L-lysine, and centrifuged at 1,200g for 2 min. The plates with coverslips were incubated for 30 min at 37 °C and fixed/stained as described. The anti-SAG1 antibody was resuspended in 2% PBS/BSA, and the corresponding fluorescent secondary antibody was used for visualization by imaging.

STAT6-P rhoptry secretion assay

A total of 5 × 10⁶ freshly egressed parasites were harvested and resuspended in 250 μl of DMEM for infection of one coverslip seeded with an HFF monolayer. After 30 s of centrifugation at 1,000g, the coverslips were incubated at 37 °C for 20 min, followed by 8 min of fixation in ice-cold methanol (MeOH). After 30 min of blocking with 5% PBS/BSA, the coverslips were incubated overnight at 4 °C with a STAT6-P antibody. The next day, the coverslips were visualized with a fluorescent secondary antibody and DAPI. The rhoptry discharge efficiency was determined by the ratio of the number of STAT6-P-positive cell nuclei to the total number of nuclei (visualized by DAPI). At least 200 cell nuclei were counted for quantification. The data presented in this study were acquired from three biological replicates.

Western blot analysis for monitoring auxin-induced protein degradation

HFF monolayers seeded in 6 cm petri dishes were infected with freshly egressed parasites and grown for 48 h in the absence or presence of IAA (500 µM). The parasites were pelleted at 1,200 rpm, resuspended in protein loading buffer containing 2% SDS, and boiled for 15 min at 95 °C. To monitor the depletion over various periods, the HFF monolayer was infected with freshly egressed parasites for 1 h at 37 °C. The extracellular parasites were then washed with DMEM supplemented with 500 μM IAA for 1, 2, 4, 8, or 12 h. After 12 h of depletion, all the samples were pelleted and resuspended in SDS. Protein depletion was assessed by WB using an anti-HA antibody against the protein of interest and anti-MIC2 as a loading control.