Dual apical methyltransferases orchestrate motility initiation in apicomplexan parasites

Abstract

Apicomplexan parasites such as Toxoplasma gondii initiate motility through rapid, spatially confined cytoskeletal activation at their apical end. While calcium-, lipid-, and kinase-based signalling pathways have been partially elucidated, how these cues are translated into mechanical force remains unclear. Here, we uncover a dual methyltransferase mechanism that orchestrates this process. We characterise TgPCKMT, a PreConoidal ring–associated lysine (K) MethylTransferase, as an essential upstream regulator of motility. TgPCKMT anchors the actin nucleator Formin-1 (TgFRM1) at the conoid, enabling conoid protrusion and F-actin assembly. Loss of TgPCKMT abolishes TgFRM1 recruitment, blocks conoid extrusion, and arrests invasion and egress despite preserved conoid structure. In contrast, the apical methyltransferase TgAKMT, previously linked to motility through recruitment of the glideosome-associated connector (TgGAC), acts downstream, disengaging from the conoid upon activation and likely promoting TgGAC-dependent force transmission. TgPCKMT depletion prevents TgAKMT translocation, revealing that actin assembly and lysine methylation are mechanistically coupled. Together, these findings define a two-step methylation regulatory module that coordinates actin nucleation with force propagation, uncovering methylation as a central regulatory axis for motility initiation in apicomplexan parasites.

This study identifies the Apicomplexa-specific methyltransferase PCKMT and reveals how it works with the previously known AKMT in a two-step pathway that initiates parasite motility and invasion.

Introduction

Apicomplexan parasites such as Toxoplasma gondii rely on tightly coordinated cytoskeletal changes to invade and exit host cells. At the apex, the conoid complex and its preconoidal rings (PCRs) have emerged as regulatory hubs for motility initiation. Once enigmatic, these rings are now recognised as platforms that concentrate regulatory and structural effectors essential for invasion and egress¹. For example, the Conoid Gliding Protein (TgCGP) localises to the PCRs and is required for their integrity; parasites lacking TgCGP lose the rings and become non-motile^2,3. Thus, the PCRs act as architectural and regulatory hubs where motility factors assemble to ensure that gliding begins at the right time and place⁴.

Motility initiation is also governed by intracellular signalling. In T. gondii, the cGMP–TgPKG cascade elevates Ca²⁺ and triggers microneme secretion^5,6. This is followed by conoid extrusion and F-actin nucleation at the apex by Formin-1 (TgFRM1), an actin regulator essential for invasion^7,8. Newly formed actin filaments (F-actin) are then linked to surface adhesins by the glideosome-associated connector (TgGAC), thereby transmitting force for directed motion^9,10. Despite this well-defined kinase and actin network, the contribution of post-translational modifications, particularly lysine methylation, to motility initiation remains largely unexplored.

Although direct functional studies are limited, there is strong precedent that methylation modulates cytoskeletal dynamics and motility in other systems. For example, SETD2-mediated methylation of actin affects cell migration¹¹. In addition, independent studies have identified methylated proteins in motile cilia and flagella^12,13, suggesting a broader link between methylation and cytoskeletal dynamics.

In apicomplexan parasites, the Apical lysine (K) MethylTransferase TgAKMT represents one of the few methylation enzymes directly implicated in motility control. It localises to the conoid and is essential for parasite egress and invasion; upon motility induction, TgAKMT rapidly translocates from the conoid to the cytosol^14,15. This dynamic behaviour suggests a regulatory role at the apex, where methylation may act as a reversible switch coupling conoid extrusion to actin assembly.

Here, we characterised a second SET-domain methyltransferase, TgPCKMT (PreConoidal Ring-associated Lysine (K) MethylTransferase), that localises to the preconoidal rings (PCRs) and acts upstream of TgAKMT to trigger motility initiation. Using conditional knockouts, high-resolution imaging, and proteomics, we demonstrate that TgPCKMT recruits the actin nucleator Formin-1 (TgFRM1) to the conoid to enable conoid protrusion and actin filament assembly, and productive gliding motility. Loss of TgPCKMT abolishes TgFRM1 recruitment, blocks conoid protrusion, and prevents the actin-dependent disengagement of TgAKMT from the apex. Conversely, a catalytically inactive TgPCKMT mutant supports partial TgFRM1 recruitment but fails to restore motility, demonstrating that methyltransferase activity is essential. Together, these findings reveal a dual-methyltransferase regulatory module at the conoid: TgPCKMT licenses actin assembly by anchoring TgFRM1, thereby enabling TgAKMT-driven remodelling and TgGAC recruitment. This two-step methylation logic synchronises actin nucleation with force transmission to initiate motility.

Results

PCKMT is a conserved apical methyltransferase associated with the preconoidal rings

In previous studies, we identified several apical complex components essential for parasite egress². Among them, TgCGP was shown to be critical for maintaining the integrity of the PCRs in mature T. gondii tachyzoites³. TgCGP depletion led to loss of the PCRs and impaired retention of key motility factors at mature conoid complexes, including the actin nucleator TgFRM1 (Fig. 1a). In the same study, a proximity-labelling assay identified a previously uncharacterised SET-domain methyltransferase (TGME49_292170) as a potential TgCGP interactor. This protein is localised exclusively to the PCRs, and its stable association depended on TgCGP. Immunofluorescence and high-resolution imaging confirmed that this methyltransferase, hereafter referred to as TgPCKMT resides at the preconoidal rings, positioned immediately above the conoid body motor myosin TgMyoH¹⁶ (Fig. 1b). Structural prediction using AlphaFold3 revealed that TgPCKMT (~175 kDa) adopts a multidomain architecture composed of conserved TPR, ankyrin, and SET domains, followed by a tandem proline-rich region at the C-terminus (Supplementary Fig. 1a, b).

Fig. 1. TgPCKMT localises to the preconoidal rings, anchors TgFRM1, and is essential for motility in T. gondii.

a Schematic overview of the preconoidal ring (PCR) network. Structural components (TgCGP, TgPcr4, and TgPcr5) form the PCR scaffold, while TgCGP supports the recruitment of a subset of regulatory proteins (TgFRM1, TgICAP16, TgDAP1, and TgPCKMT—shown in magenta-) at the preconoidal rings. b STED super-resolution image of a T. gondii tachyzoite showing that TgPCKMT (magenta) forms a discrete ring positioned apical to TgMyoH (yellow), consistent with preconoidal ring localisation. Scale bar, 1 µm; representative of n = 3. c Localisation of Tg PCKMT at the conoid (cyan) of mature parasites (cyan arrow) and developing daughter cells (white arrow) in non-induced parasites. The shape of the parasites is shown with α-GAP45 antibody (magenta). Upon rapamycin induction, TgPCKMT signal is lost. Scale bar, 2 µm; representative of n = 3. Endogenous tagging of TgFRM1 (d), TgCGP (e), and TgPcr4 (f) (in yellow) in the conditional loxPpckmt-Halo (cyan) strain shows that TgFRM1 signal is markedly reduced at the conoid upon PCKMT depletion, whereas TgCGP and TgPcr4 remain unaffected, indicating a unidirectional dependency. Cyan arrow: mature conoids; white arrows: daughter cell conoids. Scale bar: 5 µm and 1 µm in magnified images. Images representative of n = 3 replicates. g Tomographic slices and corresponding segmentations from cryo-electron tomograms of wild-type (WT) and TgPCKMT-depleted parasites (PCKMT iKO). Segmented cellular structures are colour-coded: bright pink, plasma membrane; dark pink, preconoidal rings (PCRs); yellow, conoid; khaki, apical polar ring (APR); green, microtubules (MTs); orange, rhoptries; turquoise, micronemes; beige, inner membrane complex (IMC); dark red, intraconoidal microtubules (ICMTs); grey, microtubule-associated vesicles; light green, apical vesicle (AV); dark green, dense granule (observed only in the WT tomogram). The images shown are representative of the tomograms suitable for analysis (PCKMT iKO, n = 60 of 89 acquired tomograms; WT, n = 25 of 46 acquired tomograms). Scale bar, 200 nm. For full tomographic reconstructions and segmentations, see Supplementary Movies 1–3. h Representative plaque assay images of the DiCreΔKu80 (parental) and loxPpckmt-Halo strains after 7 days ± 50 nM rapamicin (n = 3). Cyan boxes indicate regions shown at higher magnification. Scale bars: 3 mm in overview images and 1 mm in magnified images. No plaques were detectable following TgPCKMT depletion under these assay conditions. i Quantification of invasion efficiency after 72 h ± 50 nM rapamicin. Data are normalised to the parental control (DiCreΔKu80). Bars show mean ± SD; white dots represent means of each biological replicate (n = 3). Statistical significance was determined by two-tailed Student’s t test. Source data are provided as a Source Data file. j Quantification of gliding speed. Box plots show the median as the centre line and the 25th and 75th percentiles as the lower and upper bounds of the box. Whiskers extend to the smallest and largest values within 1.5 times the interquartile range, and points beyond the whiskers represent outliers. Dots represent individual parasites pooled across three biological replicates. Statistical significance was determined by two-tailed Student’s t test. Source data are provided as a Source Data file.

To characterise the role of TgPCKMT in the lytic cycle, we introduced loxP sites flanking the pckmt locus in a strain expressing DiCre recombinase (Supplementary Fig. 1c–e)^17,18. Time-course quantification revealed that by 48 h post-induction, over 80% of the population had lost detectable TgPCKMT signal, reaching near-complete 100% knockout by 72 h (Fig. 1c, Supplementary Fig. 1f).

Given its identification as a TgCGP-associated factor, we next asked whether TgPCKMT depletion disrupts the localisation of apical proteins that depend on the PCRs for recruitment. Strikingly, loss of TgPCKMT resulted in a complete loss of FRM1 from the PCRs (Fig. 1d), establishing TgPCKMT as essential for anchoring this key actin nucleator at the apex. Conversely, TgFRM1 depletion did not affect TgPCKMT localisation, indicating a unidirectional dependency (Supplementary Fig. 1g). TgDAP1, another conoid-associated factor that is dispensable for parasite survival³, also exhibited partial mislocalisation upon TgPCKMT depletion (Supplementary Fig. 1h, i), whereas TgICAP16 remained correctly positioned in both mature and budding daughter cells (Supplementary Fig. 2a). Notably, TgICAP16 was previously reported to contribute moderately to host cell invasion¹⁹. These findings indicate that TgPCKMT selectively stabilises a subset of apical complex proteins, with TgFRM1 being its most prominent effector.

To exclude potential conoid structural defects similar to those reported in TgCGP-depleted parasites³, we next examined whether loss of TgPCKMT affected the organisation of other preconoidal components. TgCGP localisation was not affected by TgPCKMT depletion, and likewise, TgPcr4 (PreConoidal Ring protein 4) remained correctly positioned at the preconoidal rings (Fig. 1e, f), indicating that TgPCKMT is not required for maintaining overall preconoidal integrity. To further rule out subtle structural perturbations, we performed cryo-electron tomography (cryo-ET), which revealed a morphologically intact conoid and associated apical complex in mature PCKMT-KO parasites (Fig. 1g; Supplementary Fig. 2b–e; Supplementary Movies 1–3).

Having established that loss of TgPCKMT selectively affects specific apical factors without grossly altering conoid architecture, we next assessed its functional relevance during the lytic cycle. Plaque assays revealed a complete loss of parasite growth after one week of induction with 50 nM rapamycin (Fig. 1h). Replication assays confirmed that intracellular division proceeds normally in the absence of TgPCKMT (Supplementary Fig. 3a), indicating that the loss of TgPCKMT primarily affects motility-associated processes such as invasion, egress, or gliding. Indeed, parasites lacking TgPCKMT exhibited a severe invasion defect, comparable to knockouts of core motility factors, including Tg-actin or TgMyoA²⁰ (Fig. 1i). Gliding motility was also profoundly impaired: only a small fraction of parasites initiated motility, and those that did moved at reduced speed (Fig. 1j; Supplementary Fig. 3b; Supplementary Movie 4).

Consistent with these defects, TgPCKMT-depleted parasites also failed to egress from host cells even when stimulated with calcium ionophore (Fig. 2a). To dissect this phenotype in more detail, we analysed actin dynamics during induced egress using parasites expressing chromobody-emerald (CbEm) to visualise F-actin and TgSAG1ΔGPI-dsRed marker² to visualise the parasitophorous vacuole space. In non-induced parasites, actin filaments depolymerised upon stimulation, the vacuole ruptured, and F-actin relocalised to the parasite basal end, enabling effective egress (Fig. 2b; Supplementary Movie 5). In contrast, TgPCKMT-depleted parasites exhibited a pronounced egress block: although the parasitophorous vacuole ruptured, F-actin failed to accumulate at the posterior pole, indicating a failure in motility initiation reminiscent of TgCGP-deficient parasites² (Fig. 2c; Supplementary Movie 5).

Fig. 2. Conserved apical methyltransferase function: TgPCKMT and PfSET9 are required for parasite propagation.

a Quantification of T. gondii egress efficiency using a fixed-egress assay. Parasites were pre-induced with or without rapamycin (Rapa) for 72 h, and egress was triggered with 4 µM calcium ionophore A23187 (CI) for 5 min. Egress efficiency was normalised to the DiCreΔKu80 control. Bars show mean ± SD; white dots indicate biological replicates (n = 3). Statistical significance was assessed by two-tailed Student’s t test. b, c Representative stills from live egress imaging (Supplementary Movie 4). Parasites expressing Chromobody-emerald (CbEm; yellow, F-actin) and SAG1ΔGPI-dsRed (magenta, parasitophorous vacuole) were stimulated with 4 µM CI. b Non-induced parasites show actin remodelling and egress. (c) TgPCKMT-depleted parasites fail to initiate motility despite vacuole rupture. Scale bar: 5 µm. d Expansion microscopy of conoid extrusion following ~10 min CI stimulation. White arrows mark extruded conoids. The yellow arrow indicates improperly extruded conoid. and cyan arrows indicate non-extruded conoids. Scale bar: 5 µm. e Quantification of conoid extrusion. ≥100 parasites were counted per condition. Bars show mean ± SD; dots represent biological replicates (n = 3). Two-tailed Student’s t test; n.s., non-significant. f Immunofluorescence of PfSET9-depleted schizonts (48 h post-induction with 100 nM Rapa) show depletion of PfSET9 (magenta) and weaker PfFRM1 (yellow) signals. Scale bar: 5 µm. g Quantification of parasite growth over 6 days reveals a pronounced growth defect following PfSET9 depletion. The points represent the mean across biological replicates (n = 3). The points show mean ± SD across 3 biological replicates. h Quantification of the intraerythrocytic stage distribution across the different induction groups (non-induced (-Rapa), induced at ring ( + Rapa R), trophozoite ( + Rapa T), or schizont stages ( + Rapa S)) over a period of 6 days. Stacked bars show the distribution of parasites across the indicated blood stages, normalized to the corresponding non-induced population on the same day. Numbers within the bars indicate the total number of parasites assigned to each stage across three independent biological replicates. Data are presented as mean ± SD across biological replicates (n = 3). A minimum of 2100 red blood cells were counted to obtain a minimum of 100 parasites in the non-induced condition across each biological replicate at Day 0. Approximately the same number of red blood cells ( ≥ 2100) were counted across all conditions for all time points. Statistical significance was assessed separately for each day using one-way ANOVA followed by a two-sided Dunnett’s multiple-comparisons test comparing each induced group ( + Rapa R, +Rapa T and +Rapa S) with the matched non-induced control ( − Rapa). Dunnett-adjusted P values are shown in the panel. Very small tail probabilities were evaluated on the logarithmic scale in R and converted back to standard P values for reporting; the displayed values are not log-transformed P values. Source data are provided as a Source Data file.

Since motility initiation is tightly linked to conoid protrusion²¹, we used expansion microscopy to compare wild-type, Tgpckmt-induced knockout (iKO), and Tgfrm1-iKO parasites. In both TgPCKMT- and TgFRM1-depleted parasites, conoid extrusion was significantly impaired compared to controls (Fig. 2d, e), suggesting that TgPCKMT acts upstream of F-actin polymerisation.

To explore whether the function of TgPCKMT might be conserved across apicomplexan parasites, we next examined its Plasmodium falciparum orthologue, annotated as set9. This gene was previously detected in transcriptional signatures of sexually committed parasites²², and available transcriptomic data indicate that PfSET9 is expressed across multiple life stages, including the asexual blood stages. However, its localisation and function have not been described so far. We generated a conditional mutant by inserting a loxP intron into the Pfset9 locus in a DiCre background^23,24 (Supplementary Fig. 3c–f). Immunofluorescence analysis revealed that PfSET9 localises as a distinct apical dot, in close proximity with PfFRM1, in segmented merozoites during the asexual blood stage (Fig. 2f, Supplementary Fig. 3e). Upon PfSET9 depletion, PfFRM1 signal appeared reduced or partially mislocalised compared to controls, suggesting that proper PfFRM1 positioning depends, at least in part, on SET9 activity and indicating a conserved functional link between TgPCKMT/PfSET9 and FRM1-like factors in regulating apical assembly and invasion (Fig. 2f). Following rapamycin-induced excision at different developmental stages (Supplementary Fig. 3g), parasites failed to transition to ring stages and exhibited a pronounced growth arrest (Fig. 2g, Supplementary Fig. 3h), consistent with an invasion defect reminiscent of the phenotype observed in T. gondii PCKMT knockouts. We next quantified the parasite stages at each time point post-induction and observed a marked reduction in rings, despite schizont numbers remaining comparable to those in non-induced parasites, ultimately leading to a severe reduction in parasites in the subsequent cycle (Fig. 2h). These findings indicate a failure in invasion, and potentially egress, in parasites depleted of PfSET9. Intriguingly, early gametocytes were observed in induced parasites (Fig. 2h; Supplementary Fig. 3i), raising the possibility of an association with gametocyte formation.

Together, these data identify TgPCKMT as an apical lysine methyltransferase that is essential for conoid-dependent motility in T. gondii and point to an evolutionarily conserved role for its Plasmodium orthologue, PfSET9, in invasion-associated processes across apicomplexans. Our data support an apical localisation of PfSET9 in proximity to PfFRM1 and reveal an invasion defect reminiscent of that observed in T. gondii. However, a detailed mechanistic investigation of PfSET9 function, comparable to that performed for TgPCKMT/TgFRM1, remains the subject of ongoing work.

Catalytic activity of PCKMT is required for anchoring FRM1 and triggering parasite motility

To explore how TgPCKMT exerts its function, we tested whether its methyltransferase activity, mediated by a conserved SET domain, is required for TgFRM1 recruitment and motility initiation.

The catalytic histidine (H542) lies within the highly conserved HxC motif, a hallmark of active SET-domain methyltransferases. Sequence alignment of TgPCKMT with TgAKMT, other apicomplexan SET-domain enzymes, and the human cytoskeletal methyltransferase SMYD2²⁵ confirmed conservation of key catalytic residues across species (Supplementary Fig. 4a), supporting its likely enzymatic activity. This residue corresponds to H447 in TgAKMT, which was previously shown to be essential for methyltransferase activity when mutated to valine¹⁵. Based on this conservation, we introduced a point mutation substituting the catalytic histidine with alanine (H542A) in the TgPCKMT SET domain to disrupt catalytic activity.

To evaluate the functional impact of this substitution, we generated complementation lines expressing a second copy of either wild-type TgPCKMT (PCKMT^com) or the mutant version (PCKMT^H542A) in the inducible knockout background (Supplementary Fig. 4b, 5).

Complementation of the TgPCKMT-inducible knockout (iKO) strain with PCKMT^com fully restored parasite growth in plaque assays, while the PCKMT^H542A mutant failed to rescue growth (Supplementary Fig. 6a). Further analysis revealed that PCKMT^com rescued all steps of the lytic cycle, including invasion, gliding, egress, and conoid extrusion (Fig. 3a–d; Supplementary Fig. 6b–d). In contrast, the PCKMT^H542A line exhibited a partial rescue: invasion and gliding were moderately improved compared to the knockout, but egress remained severely impaired (Fig. 3a–d; Supplementary Fig. 6b; Supplementary Movie 6). Conoid extrusion was also defective in the mutant line, indicating that catalytic activity is required for this critical early step in motility initiation (Fig. 3c, Supplementary Fig. 6c, d).

Fig. 3. TgPCKMT catalytic activity is essential for TgFRM1 anchoring and motility initiation.

a Quantification of invasion efficiency in the indicated strains post induction ± Rapa (rapamycin) for 72 h. Bars show mean ± SD; white dots represent biological replicates (n = 3). b Quantification of gliding motility speed in PCKMT^com (complemented strain) and PCKMT^H542A (catalytically inactive mutant) parasites ± Rapa. Box plots show the median as the centre line and the 25th and 75th percentiles as the lower and upper bounds of the box. Whiskers extend to the smallest and largest values within 1.5 times the interquartile range, and points beyond the whiskers represent outliers. Dots represent individual parasites pooled from three biological replicates (n = 3). c Quantification of conoid extrusion in PCKMT^com and PCKMT^H542A parasites ± Rapa after CI stimulation for ~10 min. ≥100 parasites were analysed per condition. Bars show mean ± SD; white dots represent biological replicates (n = 3). d Quantification of egress efficiency in PCKMT^com and PCKMT^H542A parasites pre-induced with or without Rapa. Egress was triggered with 4 µM calcium ionophore A23187 (CI) for 5 min. ≥100 vacuoles were scored per condition. Bars show mean ± SD; white dots represent biological replicates (n = 3). e Immunofluorescence showing that PCKMT^com (cyan) expression is maintained after rapamycin-induced depletion of endogenous PCKMT (magenta), with FRM1(yellow) retaining normal localisation at the conoid. Cyan arrows indicate mature parasites; white arrows mark daughter cells. Scale bar: 5 µm in overview images and 1 µm in magnified images. Image shown is a representative of n = 3. f Representative image showing reduced TgFRM1 (yellow) localisation in PCKMT^H542A (cyan) parasites after rapamycin depletion of endogenous PCKMT (magenta). The white arrow indicates conoidal PCKMT^H542A signal. Three biological replicates were performed. Scale bar: 5 µm in overview images and 1 µm in magnified images. Image shown is a representative of n = 3. g Quantification of relative TgFRM1 intensity in the indicated strains. Each dot represents the mean TgFRM1 intensity per vacuole comprising at least 15 vacuoles per biological replicate (n = 3). Data are presented as mean ± SD. Statistical significance for all quantifications in this figure was assessed by a two-tailed Student’s t test; n.s.: non-significant. Source data are provided as a Source Data file.

We next examined how the catalytic mutation in TgPCKMT affects TgFRM1 recruitment to the conoid. To do this, we performed immunofluorescence assays in parasites expressing second-copy complementation lines. Unexpectedly, expression of the second copy, either PCKMT^WT or PCKMT^H542A, was only robustly detected upon depletion of the endogenous copy, rather than in parallel with it as anticipated (Fig. 3e, f). This may reflect integration site effects at the uprt locus or a very tight regulation of TgPCKMT expression.

In the complemented strain, TgFRM1 remained stably anchored at the conoid, whereas in the PCKMT^H542A mutant, TgFRM1 was largely absent from mature tachyzoites (Fig. 3f) but still detectable in developing daughter buds, consistent with previous observations that TgFRM1 is recruited early to nascent conoids during cell division³ (white arrows). This pattern suggests that the retention of TgFRM1, and potentially other conoidal proteins, within the mature conoid requires a post-budding maturation step. Quantification of TgFRM1 fluorescence intensity confirmed a strong reduction at mature conoids in the mutant strain, nearly comparable to the full TgPCKMT knockout (Fig. 3g). Closer inspection revealed heterogeneous TgFRM1 expression within vacuoles: in the catalytic-inactive mutant, approximately 64% of vacuoles contained a mixed population of parasites with or without TgFRM1 signal, whereas only 25% displayed uniform TgFRM1 localisation at mature conoids (Fig. 4a). These results indicate that the catalytic activity of TgPCKMT is required for stable TgFRM1 recruitment and retention at the mature conoid, whereas its initial incorporation during daughter conoid formation can occur independently of catalysis, consistent with previous observations that several PCR components transiently associate with budding conoids during division before disengaging upon maturation³.

Fig. 4. Quantitative conoid localisation of TgPCKMT^H542A and TgFRM1, reciprocal independence of TgAKMT and PCKMT, and STED mapping of apical methylation.

a Quantification of TgPCKMT^H542A and TgFRM1 expression at the conoid. Parasite vacuoles were classified into three categories: (1) no detectable expression within the vacuole, (2) partial expression, where only a subset of parasites displayed detectable conoidal signal, and (3) uniform expression, where all parasites within the vacuole showed conoidal localisation. Proteins with even faint signals were scored as expressed regardless of intensity. At least 100 vacuoles were analysed per condition across three biological replicates (n = 3). Bars show mean ± SD. b Quantification of relative TgPCKMT^H542A intensity at the conoid in the loxPpckmt-Halo/pckmt^H542A-SYFP2 strain. Each dot represents the mean conoid intensity per vacuole comprising at least 15 vacuoles per biological replicate (n = 3). Data are presented as mean ± SD. c TgAKMT (yellow) localisation remains unaltered upon TgPCKMT (cyan) depletion. Cyan arrow indicates AKMT at the mature conoid. Scale bar: 5 µm. d TgPCKMT (cyan) localisation remains unaltered upon TgAKMT (yellow) depletion. Cyan arrow indicates PCKMT at the mature conoid. White arrow: daughter cell conoids Scale bar, 5 µm. Zoom in scale bar, 2 µm. e STED microscopy of extracellular T. gondii tachyzoites. (i) Maximum-intensity z-projection showing overall methylation pattern. (ii) Merge of three optical sections highlighting apical structures, with strong methylation signals at the preconoidal rings (pink arrow) and apical polar ring (blue arrow). (iii) Methylation puncta align along cortical microtubules. Scale bar, 1 µm. α-AcTub: anti-acetylated tubulin antibody (magenta); α-H4K20me3: anti-trimethylation in lysine 20 of the histone 4 antibody (yellow). Image shown is a representative of n = 3. Statistical significance (a, b) was assessed by a two-tailed Student’s t test. n.s.: non-significant. Source data are provided as a Source Data file.

Interestingly, not all vacuoles expressed detectable levels of the second-copy PCKMT^H542A protein, raising the possibility that an active SET domain may be required for the proper recruitment or stabilisation of TgPCKMT itself at the conoid (Fig. 4a, b). In fact, AKMT^H447V also failed to localise fully at the conoid¹⁵. This suggests a potential feedback mechanism in which enzymatic activity reinforces localisation or retention of these methyltransferases at their site.

Taken together, these results demonstrate that TgPCKMT’s catalytic activity is essential for anchoring TgFRM1 at the conoid and initiating parasite motility. The inability of the H542A mutant to restore growth, conoid extrusion, or egress confirms that enzymatic activity, rather than mere structural presence, is required for TgPCKMT function. The reduced localisation of both TgFRM1 and the inactive TgPCKMT variant further suggests a feedback mechanism in which catalytic activity stabilises TgPCKMT and its partners at the conoid.

PCKMT and AKMT localise independently at the apical complex

PCKMT is not the only methyltransferase localised at the conoid. To investigate potential functional interplay with the apical methyltransferase TgAKMT, we examined TgAKMT dynamics and the reciprocal localisation of both enzymes in their respective knockout backgrounds.

TgAKMT is a SET-domain methyltransferase shown to regulate gliding motility¹⁵ by promoting the recruitment of TgGAC, which links F-actin to the parasite surface adhesins⁹. Immunofluorescence analysis revealed that TgPCKMT depletion did not affect TgAKMT localisation at the conoid in intracellular parasites (Fig. 4c), nor did it impact TgGAC recruitment (Supplementary Fig. 6e). Conversely, TgAKMT knockout did not alter the localisation of TgPCKMT, which remained at the preconoidal rings (Fig. 4d). These results indicate that TgPCKMT and TgAKMT localise independently and may function in parallel to prepare the conoid and associated cytoskeletal components for motility activation.

Distinct Methylation Networks Governed by PCKMT and AKMT

Next, to further dissect the functional distinction between TgPCKMT and TgAKMT, we analysed global methylation patterns using an antibody specific for trimethylated lysine 20 of histone H4 (α-H4K20me3)²⁶. Interestingly, this antibody revealed a broad methylation profile concentrated at the apical complex and extending along the subpellicular microtubules (Fig. 4e, Supplementary Fig. 7a).

In intracellular TgPCKMT-depleted parasites, we observed no detectable change in H4K20me3 signal compared to non-induced parasites (Fig. 5a), suggesting that TgPCKMT does not contribute to global methylation patterns under these conditions. In extracellular parasites, the signal was more elevated than in intracellular parasites, suggesting an increase in methylation after motility initiation, but again unchanged in the absence of TgPCKMT (Fig. 5b).

Fig. 5. Apical lysine methyltransferases establish a conoid methylation network controlling TgFRM1 localisation in T. gondii.

a–c Immunofluorescence analysis of methylation using anti-H4K20me3 antibody in parasites induced ± rapamycin (Rapa). α-H4K20me3, magenta; DAPI, cyan. a Intracellular loxPpckmt-Halo parasites. b Extracellular loxPpckmt-Halo parasites. c Extracellular loxPakmt-YFP parasites. Comparable methylation patterns were observed with α-H3K36me3, α-Kme1/2, and α-Pan-Kme3 antibodies (Supplementary Fig. 7c,d). Scale bars, 5 µm in overview and scale bar 1 µm in magnified images. d Volcano plots showing differential methyl-protein enrichment between induced and non-induced parasites for loxPpckmt-Halo and loxPakmt-YFP strains. Grey points indicate proteins that did not meet the significance threshold; cyan and red points indicate significant negative and positive log₂ fold changes, respectively. e Representative images of PCKMT–TurboID parasites stained with streptavidin (yellow) and DAPI (cyan). In the absence of biotin, a strong streptavidin signal is detected at the apicoplast (magenta arrows). After 6 h induction with 150 µM biotin, additional puncta appear at the conoid (white arrows). Insets show conoidal localisation. Scale bars: 5 µm in overview and scale bar 1 µm in magnified images. f Volcano plots from TurboID proximity-labelling experiments for TgPCKMT and TgAKMT after 6 h biotin incubation. Grey points indicate proteins that did not meet the significance threshold; cyan and red points indicate significant negative and positive log2 fold changes, respectively. g Representative images of TgAKMT–TurboID parasites stained with streptavidin (yellow) and DAPI (cyan). Following 6 h biotin induction, signal is detected both apically (white arrows) and at the basal complex (green arrows) besides the expected localisation at the apicoplast (magenta arrow). Insets show apical puncta with mature (cyan arrow) and daughter cells (white arrows). Scale bars: 5 µm in overview and scale bar 1 µm in magnified images. h Venn diagram summarising overlap between methyl-enrichment and TurboID datasets, highlighting TgPCKMT and TgFRM1 as the only proteins consistently identified across all analyses (see source data). Yellow, TgAKMT–TurboID; grey, TgPCKMT–TurboID; red, TgPCKMT methyl-enrichment; cyan, TgAKMT methyl-enrichment. The immunofluorescence experiments shown in panels (a–c, e and g) were independently repeated three times with similar results. Source data are provided as a Source Data file.

In contrast, depletion of TgAKMT resulted in a striking loss of methylation signal in extracellular parasites (Fig. 5c). This confirms TgAKMT as the dominant contributor to methylation detectable by these antibodies during the lytic cycle as previously described⁹. To address the possibility that TgPCKMT acts on a restricted subset of targets not detectable by population-level immunofluorescence, we generated a double knockout strain lacking both TgAKMT and TgPCKMT, hereafter referred to as the Double Apical MethylTransferases inducible KnockOut (DAMTiKO). However, this mutant displayed methylation levels indistinguishable from the TgAKMT single knockout (Supplementary Fig. 7b), further supporting a limited role for TgPCKMT in bulk lysine methylation.

Comparable methylation patterns were observed using alternative antibodies, including α-H3K36me3, α-Kme1/2, and α-Pan-Kme3, in both intracellular and extracellular parasites (Supplementary Fig. 7c, d). Although these antibodies might recognise partially overlapping methylation marks and can display sequence bias²⁷, their consistent apical labelling across different reagents strongly supports that the observed methylation pattern reflects genuine, biologically relevant modifications rather than antibody cross-reactivity. Consistent with this, Western blot analysis confirmed a corresponding reduction in methylation levels upon enzyme depletion (Supplementary Fig. 8a).

To explore whether PCKMT targets a restricted subset of substrates, we performed α-H4K20me3–mediated methylated protein enrichment (Fig. 5d; Supplementary Fig. 8b, c) and TurboID-based proximity labelling (Fig. 5e, f; Supplementary Fig. 9) in both wild-type and KO backgrounds. Comparative analysis of these proteomes revealed largely distinct sets of interactors for TgPCKMT and TgAKMT, with only TgFRM1 and TgPCKMT emerging as consistent hits in all datasets (Fig. 5h). To further characterise these datasets, we performed Gene Ontology (GO) enrichment analysis of the proteins reduced upon TgAKMT or TgPCKMT depletion (Supplementary Fig. 8c). TgAKMT-associated proteins were significantly enriched for cytoskeletal organisation and motility-related processes, including the myosin complex, actin cytoskeleton, and vesicle organisation, consistent with its broad role in apical and pellicular remodelling. By contrast, the limited number of proteins affected in the TgPCKMT methylome yielded no meaningful GO enrichment, in line with its narrow substrate specificity. Similar trends were observed in the TurboID datasets (Supplementary Fig. 9f, g), reinforcing the functional partitioning between TgAKMT and TgPCKMT revealed by the proteomic analyses.

Among the apical proteins detected in these datasets, TgMyoL was of particular interest because recent atomic models of the conoid PCR-P2 ring place TgPCKMT, referred to as AKMT2 in that study, in close proximity to TgMyoL and its associated calmodulin-like light chains²⁸. This structural organisation suggests that TgPCKMT and TgMyoL occupy the same specialised region of the apical complex. Consistent with this, TgMyoL was detected in the TgPCKMT proximity-labelling dataset and was also enriched in the TgAKMT methylome dataset. We therefore examined whether TgMyoL contributes to TgPCKMT localisation. Upon TgMyoL depletion, TgPCKMT-Halo remained concentrated at the conoid despite loss of MyoL-YFP signal, indicating that TgMyoL is not required for TgPCKMT recruitment or stable positioning at the PCR-P2 region (Supplementary Fig. 10a).

To further validate candidate proteins emerging from these datasets, we endogenously tagged several additional candidates that appeared in both the TgAKMT and TgPCKMT TurboID datasets and localised to the conoid of mature or developing parasites, including the recently described apical polar ring scaffold TgASAF1^3,29,30. None of these proteins showed altered localisation following TgPCKMT depletion (Supplementary Fig. 10b–f).

Together, these findings demonstrate that TgPCKMT and TgAKMT define distinct yet complementary methylation networks at the apical complex. TgPCKMT acts on a narrow subset of substrates, most notably anchoring TgFRM1 to the preconoidal rings, whereas TgAKMT methylates a broader range of cytoskeletal targets at the conoid and pellicle. This division of labour explains the contrasting methylation patterns observed by immunofluorescence and underscores the robustness of our combined proteomic and imaging-based analyses.

PCKMT links actin assembly to AKMT mobilisation at the conoid

In the current model of motility initiation, signalling cues trigger conoid protrusion^21,31, followed closely by the disengagement of TgAKMT¹⁵ from the apex and the translocation of TgGAC⁹ together with F-actin to the parasite’s basal end to generate the force needed for gliding. To position TgPCKMT within this cascade, we analysed TgAKMT dynamics during egress induction. To capture the onset of TgAKMT translocation, parasites were stimulated with a calcium ionophore before fixation, and lysed parasitophorous vacuoles were stained with anti-TgSAG1 under non-permeabilising conditions to identify those that had initiated egress, including in the induced knockout lines that are severely impaired in motility. In wild-type parasites, calcium ionophore stimulation triggers TgAKMT disengagement from the conoid and relocalisation to the posterior end. In contrast, TgPCKMT-depleted parasites retained TgAKMT at the conoid in most cells, with quantification revealing a near-complete block in translocation (Fig. 6a; Supplementary Fig. 10g).

Fig. 6. Dynamic relocalisation of TgAKMT requires TgPCKMT and actin polymerisation.

Immunofluorescence analysis of TgAKMT (yellow) dynamics during calcium-triggered egress in loxPpckmt-Halo (magenta) (a) and loxPfrm1-Halo (magenta) (b) parasites induced ± rapamycin (Rapa) for 54 h. Egress was stimulated with 4 µM calcium ionophore A23187 (CI) for 5 min. Anti-SAG1 (cyan, non-permeabilising) marks parasites that have ruptured the parasitophorous vacuole membrane (PVM) but not completed egress. Cyan arrows indicate localization of the conoid; scale bars: 2 µm (merged) and 1 µm (zoom). c TgAKMT (yellow) translocation occurs independently of TgGAC (magenta). loxPgac-Halo parasites were induced ± Rapa for ≥54 h and stimulated with 4 µM CI for 5 min. Anti-SAG1 (cyan) marks lysed vacuoles. Cyan arrows highlight the absence of TgAKMT at the conoid. Scale bars: 2 µm (merged) and 1 µm (zoom). d TgAKMT (yellow) relocalisation depends on actin polymerization (magenta). Parasites expressing CbHalo-labelled F-actin were treated with 2 µM cytochalasin D (CytD) for 1 h before egress induction with 4 µM CI for 5 min. In CytD-treated parasites, TgAKMT remains apical (cyan arrows). Scale bars: 2 µm (merged) and 1 µm (zoom). e Quantification of TgAKMT translocation from panels (a–d). In the +Rapa condition, only parasites lacking the corresponding floxed protein (TgPCKMT, TgFRM1, or TgGAC) were considered. Bars show mean ± SD; dots represent the mean of each biological replicate (n = 3). Statistical significance was assessed by a two-tailed Student’s t test (n.s. = not significant). (f) Quantification of the conoid-to-body fluorescence intensity ratio of TgAKMT translocation shown in panels (a–d). Under the +Rapa condition, only parasites in which the corresponding floxed protein (TgPCKMT, TgFRM1, or TgGAC) was successfully depleted were included in the analysis. Box plots show the median as the centre line and the 25th and 75th percentiles as the lower and upper bounds of the box. Whiskers extend to the smallest and largest values within 1.5 times the interquartile range, and points beyond the whiskers represent outliers. Data were pooled from three independent biological replicates (n = 3). At least 10 vacuoles were quantified per condition, and at least three parasites were analyzed per vacuole. Statistical significance was determined using a two-tailed Student’s t test. g Fluorescence-recovery-after-photobleaching (FRAP) analysis of apical proteins. Parasites expressing the indicated fluorescently tagged constructs were imaged for 10 s before bleaching at the apical tip, followed by 60 s recovery. The dotted vertical line marks the bleaching time point. Normalised fluorescence intensity (mean ± SD) is plotted over time (n = 3 biological replicates). Recovery rates (τ) were obtained by exponential fitting. A two-way repeated-measures ANOVA compared recovery dynamics to the non-recovering control TgIMC1-Halo; p values are colour-coded as in the legend. h Montage of selected frames from Supplementary Movie 6 illustrating the rapid fluorescence recovery quantified in (f). Orange arrows denote the bleached region. Source data are provided as a Source Data file.

We next examined whether the conoid-associated formin TgFRM1⁸ acts upstream of TgAKMT translocation. TgFRM1 depletion phenocopied the TgPCKMT knockout, with TgAKMT failing to relocalise upon ionophore stimulation (Fig. 6b; Supplementary Fig. 10h). These data position TgFRM1 as a key factor upstream of TgAKMT mobilisation and suggest that TgPCKMT may promote local actin polymerisation by recruiting or activating TgFRM1 at the conoid. Finally, to refine pathway order, we analysed TgAKMT localisation in parasites lacking TgGAC. Contrary to earlier reports⁹, TgAKMT readily disengaged from the conoid in TgGAC knockout parasites, although it appeared more diffusely distributed within the cytosol (Fig. 6c; Supplementary Fig. 10i).

To test whether actin polymerisation is required for TgAKMT relocalisation, we treated wild-type parasites with Cytochalasin D (CytD). Following ionophore induction, TgAKMT remained concentrated at the conoid (Fig. 6d; Supplementary Fig. 10j), consistent with previous reports^9,15. Notably, the cytosolic TgAKMT subpopulation observed in the absence of TgPCKMT or TgFRM1, was frequently diminished or absent under CytD treatment, suggesting that actin dynamics facilitate its redistribution within the cytosol, even when conoid disengagement is blocked. The sporadic presence of diffuse TgAKMT signal further implies that additional F-actin nucleation centres, such as those formed by Formin 2 (TgFRM2) near the Golgi–apicoplast region¹⁷, might contribute to this residual mobility.

We quantified these dynamics using two complementary approaches. First, we scored the percentage of vacuoles showing TgAKMT translocation across all conditions, including only parasites deficient for the floxed genes analysed above or treated with CytD to abolish actin polymerisation (Fig. 6e). Second, to capture the mixed apical and cytosolic TgAKMT signal observed in some conditions, we measured the relative TgAKMT intensity at the conoid versus the parasite body/cytosol (Fig. 6f). This ratio-based analysis confirmed that TgPCKMT- and TgFRM1-depleted parasites, as well as CytD-treated parasites, retain high conoid-to-cytosol TgAKMT ratios, consistent with impaired mobilisation. By contrast, TgGAC-depleted parasites showed reduced ratios, consistent with TgAKMT disengagement from the conoid despite altered cytosolic distribution. Together, these results indicate that TgAKMT redistribution depends on the TgPCKMT–TgFRM1–actin axis but occurs independently of TgGAC.

Apical methyltransferase forms a dynamic complex regulated by PCKMT

The residual redistribution of TgAKMT under CytD treatment suggested that a component of its apical dynamics may occur independently of full motility activation. To examine this dynamic behaviour directly, we performed fluorescence recovery after photobleaching (FRAP) experiments targeting the apical pools of TgAKMT and TgGAC. Both proteins recovered fluorescence after bleaching, indicating they are not stably anchored to the conoid but instead dynamically associate with the apical complex (Fig. 6g, h; Supplementary Movie 7). Quantitative analysis showed that TgGAC-YFP recovered rapidly (with a half-time of fluorescence recovery (τ½) of 8.9 s and a maximum recovery level (Imax) of 78.1%) and TgAKMT-YFP recovered more slowly but to a similar extent (τ½ = 29.4 s, Imax = 73.4%). As expected for a stably integrated structural protein, Tg IMC1-Halo showed negligible recovery. (Fig. 6f, g)

Attempts to perform FRAP on TgPCKMT-YFP and TgFRM1-YFP were unsuccessful due to the protein’s low expression levels: the YFP signal was barely detectable and photobleached before acquisition. Nevertheless, live imaging and fixed-cell analysis revealed no evidence of relocalisation following stimulation, suggesting that TgPCKMT is a stably anchored conoid component (Fig. 6a, ( − Rapa); Supplementary Fig. 10f).

Together, these results indicate that apical proteins such as TgAKMT and TgGAC form a dynamic, reversible complex whose localisation depends on TgPCKMT-mediated actin polymerisation and underpins motility initiation.

Discussion

Our findings demonstrate that two conoid-anchored, Apicomplexa-specific methyltransferases, TgPCKMT and TgAKMT, play distinct yet complementary roles in motility initiation. TgPCKMT is required for the recruitment and anchoring of TgFRM1, enabling actin filament formation and conoid extrusion, while TgAKMT¹⁵ ensures the recruitment of TgGAC, linking F-actin to surface adhesins for force transmission⁹. Together, these enzymes define a dual methylation system that coordinates cytoskeletal activation and mechanical coupling at the parasite apical complex, ensuring that motility is initiated only once both systems are properly engaged.

Strikingly, TgPCKMT operates with exceptional specificity: proximity labelling and methyl-enrichment identified only a limited number of candidate substrates, with TgFRM1 emerging as the most prominent effector linked to motility. Consistent with this focused role, the TgPCKMT-associated datasets were not broadly enriched for cytoskeletal motors or vesicle-organisation factors, distinguishing TgPCKMT from TgAKMT and supporting its function as a regulatory licensing factor rather than a motor-associated remodelling component. Loss of TgPCKMT disrupts TgFRM1 recruitment to mature conoids, and mutation of the catalytic residue H542 mirrors this phenotype, impairing conoid extrusion and gliding. These findings indicate that TgPCKMT’s enzymatic activity, and to a certain extent its structural presence, are both required to license TgFRM1-dependent motility initiation. At present, whether distinct methylation events differentially regulate TgFRM1 positioning at the conoid versus its activation state remains unresolved. Defining the precise methylation site(s), direct substrate(s), and temporal regulation of these modifications during conoid assembly versus acute motility activation will therefore be an important future direction, but lies beyond the scope of the present study. This dual role as both scaffold and methyltransferase echoes that of large SET-domain proteins like KMT2D/MLL4, which stabilise interacting partners while methylating specific chromatin targets³². By analogy, TgPCKMT functions as a minimalist apical scaffold whose catalytic activity is channelled almost exclusively towards TgFRM1, creating a tightly focused methylation regulatory module at the onset of actin assembly.

In contrast to this narrow specificity, TgAKMT targets a broader set of substrates, as reflected by its wider distribution in microscopy and proteomic analyses. Notably, this broader methylome was enriched for several myosins and myosin-associated proteins, including TgMyoL, TgMyoA, TgMyoH and the myosin light chain TgMLC3, which localise to the apical complex. At present, it remains unclear whether these proteins represent direct methylation substrates or reflect spatial proximity within densely packed apical assemblies. Consistent with previous work, TgAKMT relocalises from the conoid to the cytosol upon motility activation¹⁵, and our data show that this translocation requires both TgPCKMT and TgFRM1. Surprisingly, TgGAC was absent from our methylome enrichment assays, suggesting that it is not directly methylated but instead recruited via one or several methylated intermediates. Further work is needed to elucidate the mechanism of TgGAC recruitment to the apical complex. What is clear, however, is that both TgAKMT and TgGAC are recruited to the apical complex in immotile parasites and translocate towards the basal pole immediately after motility is initiated via actin dynamics. Whether actin itself, or an actin-associated protein, is among TgAKMT’s direct substrates linked to this translocation remains an open question. Together, these differences in localisation, interactome composition, and dynamics support a hierarchical model in which TgPCKMT acts as a stable upstream licensing factor at the PCRs, while TgAKMT functions downstream as a dynamic effector that remodels the pellicle upon actin-dependent activation.

In metazoan cells, several methyltransferases directly modify cytoskeletal proteins to regulate filament dynamics and cell motility. SETD3 methylates β-actin at His73, a modification that stabilises filaments and promotes motility^33,34, while SETD2 and SET8 have been reported to methylate α- and β-tubulin as well as β-actin, influencing microtubule organisation and actin polymerisation^11,35,36. Other enzymes, such as PRMT5 and EZH2, have also been implicated in actin methylation and cytoskeletal control^37–39. In contrast, our data suggest that neither TgAKMT nor TgPCKMT directly methylates actin in T. gondii. Actin was not enriched in either our methyl-enrichment or TurboID assays, and no specific actin methylation signals were detected in the presence or absence of these enzymes. This implies that any impact of these apical methyltransferases on actin dynamics is likely indirect, possibly mediated through their primary effectors, such as TgFRM1 and other apical complex components, or that any direct actin methylation is transient or occurs at levels below detection in our current experimental setup. This may allow tight spatial control of actin assembly without altering core actin properties.

Interestingly, α- and β-tubulin appeared enriched in the wild-type but not in the TgAKMT-KO methyl-enrichment dataset. Tubulin methylation has been reported in other eukaryotes, and the lysine 40 (K40) residue of α-tubulin is a particularly well-conserved methylation site known to influence microtubule function in mammalian cells³⁶. Consistent with this, previous proteomic analyses in T. gondii identified methylated peptides in both α- and β-tubulin isoforms⁴⁰, supporting the possibility of direct methylation by TgAKMT. However, our recent mutagenesis of K40 in α-tubulin, designed to mimic (K40M) or prevent methylation (K40A), did not alter microtubule stability⁴¹. This suggests that the functional impact of α-tubulin methylation in T. gondii may not involve structural stabilisation but could instead fine-tune microtubule dynamics or interactions through more subtle regulatory mechanisms.

Together, these data support a model in which the core motility machinery is pre-assembled at the apical complex of non-motile parasites, poised for rapid activation upon stimulation (Fig. 7). Consistent with current models¹, in resting parasites, TgPCKMT anchors TgFRM1 at the preconoidal rings, preparing the site for actin nucleation (Fig. 7; step 1). Upon stimulation, TgFRM1 promotes conoid extrusion through local actin polymerisation, thereby facilitating engagement with TgGAC and the glideosome (Fig. 7; step 2). Concurrently, TgAKMT disengages from the conoid in an actin-dependent manner, likely methylating downstream targets that complete the transition to productive motility. In the absence of TgPCKMT, TgFRM1 is not recruited to the conoid, preventing local actin nucleation. Nevertheless, TgAKMT and TgGAC are still independently recruited, but neither protein can translocate from the apical complex, resulting in a complete failure to initiate motility (Fig. 7; step 3).

Fig. 7. Coordinated methylation logic driving motility initiation in apicomplexan parasites.

(1) In non-motile parasites, TgPCKMT, TgAKMT, TgGAC, and TgFRM1 are pre-assembled at the apical complex, poised for rapid activation. (2) Upon stimulation, TgPCKMT enables TgFRM1-dependent actin nucleation at the conoid, promoting conoid protrusion and the apical-to-basal translocation of TgAKMT and TgGAC, which couple F-actin to surface adhesins to drive motility. The exact route of TgAKMT movement (question mark) remains unresolved—whether it traverses the sub-pellicular space between the plasma membrane (PM) and the inner membrane complex (IMC) or diffuses through the cytosol is unknown. (3) In the absence of TgPCKMT, TgFRM1 fails to localise to the conoid, actin assembly does not occur, and TgAKMT/TgGAC translocation is blocked, preventing motility initiation. (4) The lower panel summarises the temporal sequence of activation, TgPCKMT → TgFRM1 → actin → TgGAC/TgAKMT, illustrating the cascade that links methylation to cytoskeletal activation and force transmission.

This model is further supported by FRAP analyses, which revealed that TgAKMT and TgGAC are not stably anchored but instead exchange dynamically between apical and cytosolic pools. This dynamic exchange may maintain the system in a primed state that can be rapidly biased toward translocation by TgPCKMT-licensed actin polymerisation. How such upstream signals are sensed at the apical complex and coupled to TgPCKMT-dependent licensing has remained unclear. Recent structural and functional studies provide a coherent framework for understanding how motility initiation is regulated at the apical complex. Cryo-electron tomography of the PCR-P2 ring has revealed a modular architecture in which TgPCKMT, TgFRM1, and the myosin-like protein TgMyoL are positioned in close proximity, together with calmodulin and myosin light chain subunits, suggesting a localized platform for calcium-responsive regulation^28,42. TgMyoL has been identified as a non-motor, calmodulin-associated component that mechanically tethers the preconoidal rings to the conoid and is important for conoid extrusion and gliding motility, while being dispensable for microneme secretion⁴³. Together, these findings suggest that the TgMyoL–calmodulin complex forms a calcium-responsive structural element at the PCRs. However, TgPCKMT remains apically localised upon TgMyoL depletion, and published TgMyoL-depletion phenotypes are less severe than the near-complete block in motility initiation caused by TgPCKMT loss. These observations suggest that TgMyoL does not primarily regulate motility initiation through controlling TgPCKMT recruitment to the PCRs, and instead point to a coordinated, but not strictly hierarchical, relationship between these components within the PCR–conoid regulatory module.

In P. falciparum, the TgPCKMT orthologue PfSET9 has so far been implicated only in gametocyte differentiation²². Our data now reveal that PfSET9 localises apically in segmented merozoites and that its depletion causes a strong invasion defect during the asexual blood stages, mirroring the phenotype observed in TgPCKMT knockouts. We also observed a possible increase in gametocyte formation, consistent with the previously reported developmental role of PfSET9. Given the reduced conoid complex of Plasmodium parasites⁴⁴, these findings raise questions about how PfSET9 is positioned within this simplified apical architecture and whether it contributes to motility across the life cycle. In particular, ookinetes, which harbour a more prominent conoid and rely heavily on motility, represent attractive systems to further explore PfSET9 function in future studies.

In summary, our data reveal a tightly ordered methylation–actin cascade that coordinates motility initiation (Fig. 7; step 4). TgPCKMT first licenses TgFRM1-mediated actin nucleation, which in turn enables TgAKMT translocation and TgGAC recruitment to transmit force. This stepwise sequence places lysine methylation at the core of motility control, providing a unifying framework for how apical signalling, cytoskeletal dynamics, and mechanical output are synchronised in T. gondii and likely conserved across apicomplexans.

Methods

Culture conditions

T. gondii culture

Human foreskin fibroblasts (HFF; ATCC SCRC-1041) were cultured in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% foetal bovine serum, 25 mg/mL gentamicin, and 2 mM L-glutamine until they reached 100% confluency. Parasites were then allowed to infect the HFF monolayer and maintained at 37 °C with 5 % CO₂ in an incubator.

P. falciparum culture

Asexual parasites were maintained in RPMI 1640 with AlbuMAX (Gibco, Thermo Fisher Scientific) and supplemented with 2.5mM L-glutamine and 25 µg/mL Gentamicin. Parasite cultures were incubated at 37 °C in a 94% N₂, 5% CO₂, and 1% O₂ gas mixture. Blood smears were routinely made and stained with Hemacolor® Rapid staining of blood smear (Sigma-Aldrich) to determine the parasitemia of the Plasmodium cultures.

Generation of plasmids

T. gondii plasmid generation

Guide RNAs targeting the gene of interest were designed using EuPaGDT⁴⁵ and are listed in Supplementary Data 1 (Table 1). The sgRNAs were ligated into a vector coding for Cas9-YFP expression as previously described⁴⁶. Briefly, gRNA sequences were ordered as single-stranded oligomers (ThermoFisher) with overhangs to allow for ligation to the Cas9-YFP vector⁴⁶. Primers were diluted at equimolar concentrations in annealing buffer (10 mM Tris, pH 7.5–8.0, 50 mM NaCl, 1 mM EDTA), heated at 95 °C for 5 min and left to cool down for at least 2 h. Cas9-YFP vector was linearised using BsaI (NEB). Annealed primers and digested vector were ligated using T4 ligase (NEB) overnight at room temperature. The ligated vector was transformed into DH5α bacteria, and colonies were screened for positive gRNA-containing vectors.

To generate the vectors containing the second copy of pckmt fragments containing promoter region, TPR domain, SET domain, Ankyrin domain, proline-rich motif, and 3´UTR region were amplified by PCR with overhangs to allow for insertion with Gibson assembly kit (NEB) as described in Supplementary Fig. 5. Insertion of the different fragments was verified by PCR.

P. falciparum plasmid generation

To generate the conditional Pfset9 mutant, a synthetic DNA fragment corresponding to Pfset9 (PlasmoDB gene ID: PF3D7_0508100) was synthesised by GenArt (Thermo Fisher Scientific) and cloned into an SLI-based vector containing a loxP-flanked intron as previously described²³ using the restriction enzymes BglII and SalI. Briefly, a 1 kb homology region corresponding to the endogenous PfSET9 locus was retained upstream of the synthetic loxP-containing intron, which was inserted immediately upstream of the SET domain. The sequence downstream of the loxP-containing intron through to the end of the coding sequence was recodonised, and the endogenous stop codon was removed to place the recodonised sequence in frame with a C-terminal HA tag positioned upstream of the SalI restriction site. A schematic representation of the resulting vector is shown in Supplementary Fig. S3c.

Generation of transgenic parasites

T. gondii parasites

The CRISPR/Cas9 system was used to generate transgenic parasites. Repair templates for integrating a tag and a loxP sequence were generated following the method described by Stortz et al.^2,17,47. Briefly, for tagging, the repair templates were PCR-amplified from vectors carrying tags such as 3xHA, SYFP2, Halo, and SNAP. Primers used for this process were designed to bind the vectors and included 50 bp of homology to the gene. For the upstream loxP, oligonucleotides containing the loxP sequence, flanked by 33 bp of homology on both sides of the gene, were ordered as single-stranded DNA from ThermoFisher.

The HaloTag system was employed to label PCKMT due to its versatility in conjugating different fluorophores and its ability to provide high signal-to-noise ratios and specificity when combined with Janelia dyes (Promega). This approach was particularly advantageous for imaging this low-expressing protein. Other proteins, such as FRM1 or AKMT, were tagged with YFP derivative (SYFP2), and GAC was tagged with mCherry or YFP.

Parasite transfection, sorting, and screening for positive clones were performed as described by Stortz et al.¹⁷. In short, repair templates and Cas9-YFP expression vectors were co-transfected into freshly lysed or mechanically lysed RHDiCre Δku80⁴⁸ tachyzoite parasites. At 24–48 h after transfection, parasites were sorted using a FACSAria III cell sorter (BD Biosciences). Parasites were first identified on the basis of their forward- and side-scatter properties. Doublets and aggregates were excluded by singlet gating based on forward-scatter pulse geometry, after which Cas9–YFP-positive parasites were selected according to their green fluorescence. Individual YFP-positive parasites were deposited into separate wells of 96-well plates. FACS was used solely as a preparative step to isolate transfected parasites and was not used to generate quantitative data presented in the manuscript. Clonal lines were subsequently validated by genomic PCR and, where required, sequencing.

When indicated in the figure legend, TLAP1 (TrxL1-associating proteins)⁴⁹ endogenously tagged was used to show the parasite’s shape.

CBem and second copies of pckmt were integrated into the uprt (uracil phosphoribosyltransferase) locus by CRISPR/Cas9-mediated targeting. The corresponding amplicons, flanked by 50-nucleotide homology arms to the 5′ and 3′ untranslated regions (UTRs) of uprt, were inserted as previously described². Parasites lacking UPRT activity were selected using 1 µM 5-fluoro-2-deoxyuridine (FUDR; ThermoFisher)⁵⁰.

Primers are listed in Supplementary Data 1 (Table 2), and the transgenic parasite lines generated/used in this study are listed in Supplementary Data 1 (Table 3).

To induce KO, 50 nM rapamycin was added for one to two hours, followed by a wash with fresh DMEM media, and parasites were allowed to grow until fixation.

P. falciparum parasites

10 µg of plasmid DNA was transfected into mature schizonts of DiCre-expressing 3D7 parasite line (B11DiCre)⁵¹, isolated using 60% Percoll (Sigma-Aldrich), using the Amaxa P3 Primary cell 4D-Nucleofector® X Kit L (Lonza). 24 h post-transfection, parasites maintaining the plasmid episomally were selected using 2.5 nM WR99210 (Jacobus Pharmaceutical Company). The parasites were then subjected to 400 µg/ml G418 (Sigma-Aldrich) to select for genomic integration of the construct. The media and drugs were replenished every 2 days. Genomic DNA was isolated from 200 µL of infected red blood cells (iRBCs) using the Qiagen Blood and Tissue kit. An insertion PCR was conducted to verify the genomic integration using the appropriate primers (Supplementary Data 1, Table 2). Synchronous cultures were obtained by isolating mature schizonts using 60% Percoll (Sigma-Aldrich), following which they were allowed to infect fresh RBCs for 5 h. Subsequently, 0–5-hour-old rings were obtained using 5% D-Sorbitol solution (Sigma-Aldrich). To understand the role of PfSET9 in the parasites, KO was induced at the various asexual blood stages of the synchronised parasite culture- at rings (0 h), trophozoites (24 h), and schizonts (42 h), using 100 nM Rapamycin (Sigma-Aldrich) while maintaining a non-induced, synchronised parasite population as a control.

Guide RNA sequences, primer sequences, plasmid details and parasite strains used or generated in this study are listed in Supplementary Data 1 (Tables 1–3). Newly generated plasmids and parasite lines are available from the corresponding author upon request.

Immunofluorescence assay (IFA)

In T. gondii

HFFs were cultured on sterile coverslips in 24-well plates to establish a monolayer for parasite infection. The parasites were then allowed to infect the HFF monolayer for 24–48 h. The coverslips were fixed with 4% Paraformaldehyde (PFA) for 20 min, followed by block-permeabilisation using 3% BSA and 0.1% Triton-100 in PBS for 20 min. Primary and secondary antibodies of required dilutions were prepared in the block-permeabilising solution (See antibody list in Supplementary Data 1, Table 4). The coverslips were mounted onto slides using ProLong™ Gold Antifade Mountant with DNA Stain DAPI (Thermo Scientific, USA) and imaged using a Leica DMi8 inverted live cell widefield microscope or an Abberior STED microscope.

To visualise Halo-tagged proteins, parasites were incubated with 4-12 nM HaloTag Janelia 646 (Promega, GA112A) overnight prior to fixation. For colocalization analysis, 50 nM HaloTag Janelia 646 was used. After dye application, parasites were washed three times with PBS and incubated in media for 10 min prior to fixation.

In P. falciparum

Immunofluorescence analysis was performed as described in previous studies^52,53. Thin blood smears were created on microscopic slides and fixed with 4% paraformaldehyde (PFA) in PBS for 10 min. The fixed smears were washed three times with 1x PBS. The parasites were permeabilised with 0.1% Triton X-100 in PBS for 10 min, followed by 3 washes with 1x PBS. The parasites were blocked overnight with 3% BSA in PBS at 4 °C. The parasites were incubated with anti-HA (1:500, Rat, Roche) and anti-PfFRM1 (1:500, Baum Lab) diluted in blocking solution for 1 h at room temperature. The unbound primary antibody was removed with three 5-min washes with PBS. The parasites were then incubated with appropriate secondary antibodies (Supplementary Data 1, Table 4) diluted in blocking solution for 45 min at room temperature. The washing steps were repeated as described above. All the steps described were performed in a humid chamber. A couple of drops of ProLong™ Gold Antifade mounting with DAPI (Invitrogen) were added before covering the slides with 24 x 60 mm coverslips.

Expansion microscopy in T. gondii samples

ExM was performed as described by Dos Santos et al.,⁵⁴ with some modifications. Briefly, intracellular parasites on 12 mm coverslips were fixed with 4% PFA, followed by three washes with PBS, and stored at 4 °C. The fixed coverslips were then treated with a 1.4% formaldehyde and 2% acrylamide mix at 37 °C for 5 h, followed by a gelation step. A 35 µL drop of Monomer Solution (19% sodium acrylate, 10% acrylamide, 0.1% bis-acrylamide) supplemented with 0.5% Tetramethylethylenediamine (TEMED) and 0.5% Ammonium persulfate (APS) per coverslip was used for gelation on ice in a humid chamber for 5 min, then at 37 °C for 1 h. After gelation, the gel was denatured with denaturation buffer (200 mM SDS, 200 mM NaCl, 50 mM Tris in ultrapure water, pH 9) at 80/85 °C for 1.5 h. Gels were then expanded overnight in deionized water (dH₂O) at 4 °C. A small piece of the gel was cut and subjected to immunostaining. For immunostaining, antibodies were diluted in freshly prepared PBS with 2% BSA. The gel was incubated with primary antibodies for 3 h at 37 °C, followed by four washes with PBS containing 0.2% Triton X-100 (PBS-Tx100). The gel was then incubated with secondary antibodies for 2.5 h at 37 °C or overnight at 4 °C, followed by an additional 2 h at 37 °C, and then washed four times with PBS-Tx100. Antibody concentrations are listed in Supplementary Data 1 (Table 4). The gel was fully expanded in dH₂O for a few hours or overnight. To stain the nuclei, 0.8 µM Hoechst 33342 (diluted in dH₂O) was then used for 3 min, followed by one wash. After determining the orientation of the gel, it was mounted on a glass-bottom dish pre-coated with poly-L-lysine. Widefield images were acquired in Z-stacks with 0.3 µm increments using a Leica DMi8 wide-field microscope.

Sample preparation for cryo-ET

Human foreskin fibroblasts infected with T. gondii from the loxPpckmt-Halo strain were treated with 50 nM rapamycin for 72 h to induce the knockout. The uninduced (wild-type) control sample originated from an unpublished dataset prepared for a previously published manuscript³, using the loxPcgp-Halo/pcr4-Halo strain that was, apart from the induction, prepared in the same way as iKO. On the day of harvesting, parasite-infected fibroblasts were stained with 0.04 µM JaneliaFluor 646 Halo dye (Promega, GA1120) for 1 h at 37 °C. Cells with parasites were scratched from the dishes, centrifuged for 5 min at 800 × g and resuspended in fresh DMEM media. Samples were then syringed three times and centrifuged for 5 min at 600 × g. Parasites were resuspended in 0.5 mL PBS and stained with NucBlue (Hoechst 33342, Thermo Fisher Scientific) for 15 min at 37 °C. Parasites were washed with PBS and treated with ±2 µM calcium ionophore (Sigma-Aldrich A23187) for 5 min at RT and washed with PBS. Carbon lacey cryo-EM grids with mesh 300 (Quantifoil) were plasma-cleaned for 30 s with a 90:10 argon: oxygen mixture using plasma cleaner (Fischione 1070) set to 100% power and flow of 30 SCCM. Subsequently, 4 µL of parasites were applied on the grid mounted on GP1 plunge freezer (Leica), backside blotted for 2–3 s and plunge frozen into liquid ethane. The environment in the plunger was set to 99% humidity and 22 °C. Vitrified grids were clipped into Autogrids in liquid nitrogen vapours and stored in liquid nitrogen.

Cryo-electron tomography

Grids with vitrified parasites were loaded onto a Titan Krios G4 electron microscope (Thermo Fisher Scientific) operated at 300 kV, equipped with Falcon4i camera and Selectris X energy filter. Low magnification images were taken to assess ice thickness and parasite distribution over different grid squares. Then, medium magnification montages of the grid squares were acquired to identify apical sites of the parasites. Tilt series were acquired using SerialEM⁵⁵ with a dose symmetric tilting scheme⁵⁶, range of ±60° with 2° increment, at pixel size of 3.033 Å, exposure time of around 1 s with exposure dose of ~2.3 e⁻/Ų that accumulated to around 140 e⁻/Ų throughout the entire tilt series.

Tomogram reconstruction

Frames acquired in the EER format were converted to TIFF, followed by motion correction, CTF estimation and tilt selection using Relion-5⁵⁷. Tilt stacks were aligned using the AreTomo2 wrapper implemented in Relion-5. Tomograms were reconstructed with stand-alone AreTomo2⁵⁸. Reconstructed tomograms were then subjected to CTF-deconvolution from IsoNet⁵⁹ to improve visualization.

Segmentation and visualization

Reconstructed tomograms were segmented using MemBrain-seg⁶⁰ and Dragonfly (Comet Technologies Canada Inc. Dragonfly 3D World (v. 2024.1), Comet Technologies Canada Inc., https://dragonfly.comet.tech/). Firstly, membrane segmentation was predicted using MemBrain-seg (using model MemBrain_seg_v10_beta.ckpt) and imported to Dragonfly. Then, non-membranous structures were segmented manually using either full or Otsu thresholding during segmentation. Classes of segmented organelles were exported as TIFF files and a 3D Gaussian filter from FIJI⁶¹ was applied before importing to ChimeraX⁶² for visualization together with the reconstructed tomograms. Supplementary Movies were assembled from the reconstructed tomograms and segmentations using Python 3.13.1 scripts solely for visualisation; these scripts were not used for image processing, segmentation or quantitative analysis.

Plaque assay

Parasites were released by passing them through a 26-G needle. 1000 parasites per condition were allowed to infect the HFF monolayer cultured on a 6-well plate. The parasites were left undisturbed at 37 °C and 5% CO₂ for 6–7 days to form plaques. The cells are then washed with PBS and fixed with ice-cold 100 % Methanol (−20 °C) for 20 min and later stained with Giemsa stain to visualise the plaques. The plaques were imaged using a Leica DMi8 (Leica, Germany) inverted live cell widefield microscope.

Invasion assay

Parasites were pre-induced ±50 nM Rapa for 72 h. 5 × 10⁶ freshly egressed tachyzoites were resuspended in 2 mL DMEM and kept on ice for 10 min before inoculation. Parasites were then allowed to invade confluent HFF monolayers for 1 h at 37 °C, then fixed with 4% PFA for 10 min and washed three times with PBS. To distinguish extracellular and intracellular parasites, sequential immunofluorescence staining was performed. For extracellular labelling, samples were blocked without permeabilisation and incubated with anti-SAG1 for 1 h, followed by secondary antibody. After PBS washes, intracellular parasites were visualised by blocking buffer with 2% BSA, followed by staining with anti-IMC1(plus secondary) to label all parasites. Samples were imaged by a Leica DMi8 widefield microscope. A minimum of 100 parasites were scored per condition to calculate invasion efficiency, expressed as the percentage of intracellular (dual-labelled) versus extracellular (single-labelled) parasites.

Replication assay

Parasites were allowed to infect HFF monolayer-seeded glass coverslips for one hour. The non-invaded parasites were then washed off with PBS to synchronise the replication of parasites within the host cells. DMEM media (Sigma) was added to the coverslips, and the parasites were incubated at 37 °C for 24 h to allow for replication within the host cells. Induced KO for loxP-pckmt-halo were pre-incubated with 50 nM rapamycin for 72 h prior to the infection of coverslips.

The coverslips were then fixed, and the standard IFA protocol was followed as described in the immunofluorescence assay section, using anti-GAP45 (1:5000), to visualise the parasites as the primary antibody. The parasites were imaged using the Leica DMi8 inverted live cell widefield microscope (Leica, Germany), and the number of vacuoles in the various cell stages of the parasites was determined for each mutant strain. To improve the statistical significance of the observations, three biological replicates were performed.

Live gliding assay

Pre-warmed FBS was added onto the surface of a Cellvis 29 mm glass-bottom dish (Cellvis, USA) and incubated for 1–2 h and then aspirated to provide a surface to facilitate parasite gliding. Parasites were harvested and passed through a 3 µm filter to remove debris. The parasites were pelleted at 1500 × g for 5 min and washed with a pre-warmed gliding buffer to remove media. They were then resuspended in the gliding buffer (1 mM EGTA and 100 mM HEPES in HBSS solution), ensuring a concentration of 4 × 10⁶ parasites/mL. The parasites were added to the FBS-coated glass-bottom dishes. Gliding efficiency was recorded using the Leica DMi8 inverted live cell widefield microscope with x100 oil objective by capturing the gliding of the parasite every second for a total of 10 min in the DIC channel. In the case of iKO parasites, a stack image was performed after each video to determine the efficiency of our protein of interest (POI) depletion. Only parasites lacking the analysed POI were considered for the quantification of gliding motility for iKO parasites; only parasites with movement were measured for speed and distance.

The distance travelled by the parasites ( ≥ 9 parasites per replicate) as well as the speed at which the gliding motion occurred were assessed using the ManualTracking Plugin in Image J across three independent biological replicates.

Egress assays

To investigate PCKMT behaviour during egress, loxPpckmt-Halo/CbEm parasites induced with ±50 nM Rapa for 54 h before triggering egress, transiently transfected with the vector pTub- pTub-sag1ΔGPI-dsRed and grown in HFFs on a glass-bottom dish for approximately 30 h before induction of egress with 4 µM Calcium ionophore A23187(CI) (Sigma-Aldrich, C7522-1mg). Images were taken in auto-focus mode with a 100x oil objective lens. All videos were recorded in triplicate per condition at a minimum. Live microscopy was performed using a Leica DMi8 widefield microscope equipped with a DF C9000 GTC camera in a heated chamber with 5% CO₂.

For the fixed egress assay, parasites were pre-induced ±50 nM Rapa for 54 h, and Halo dye was added to live cultures 2–4 h before fixation. Egress was triggered with 4 µM CI for 5 min, and then fixed with PFA. Without permeabilisation, samples were blocked and stained with anti-SAG1 to label the extracellular parasites, followed by the corresponding secondary antibody. After permeabilisation with Triton X-100, an IMC antibody (plus secondary) was applied to label all parasites; an auto 488 antibody was used to anti-SYFP2. To block F-actin polymerisation, 2 µM Cytochalasin D(CytD) was added for 1 h prior to egress induction. Antibody details are provided in Supplementary Data 1 (Table 4).

For AKMT fluorescence intensity quantification, fixed egress samples were imaged under identical acquisition settings. Only parasites lacking visible daughter cells were selected for analysis. Mean fluorescence intensities of the conoid and parasite body were measured in Fiji/ImageJ, and the relative conoid-to-body intensity ratio was calculated. At least three parasites per vacuole and at least 10 vacuoles per biological replicate were analyzed (n = 3 independent experiments).

Protrusion assay

LoxPpckmt-Halo, loxPpckmt-Halo/pckmt^com-SYFP2, loxPpckmt-Halo/pckmt^H452A-SYFP2 and loxPfrm1-Halo. Parasites were pretreated ±50 nM rapamycin for 72 h prior to the assay. Parasites were mechanically released, filtered, pelleted, and resuspended in pre-warmed DMEM (without FBS) containing 4 µM CI. They were then added to 24-well plates containing poly-L-lysine-coated coverslips and subjected to brief centrifugation at 1000 x g for 1 min. After incubation at 37 °C for 8–10 min to promote extrusion, parasites were fixed with PFA. The parasites were then processed for ExM and stained with anti-acetylated tubulin and Halo dye. Three biological replicates were performed, with at least 100 parasites counted per condition and three biological replicate were performed. The Halo antibody did not work effectively to detect the direct absence of PCKMT and FRM1 from the conoid.

Fluorescence recovery after photobleaching (FRAP)

Intracellular Toxoplasma gondii parasites expressing YFP-tagged proteins (GAC-YFP, AKMT-YFP, MyoH-YFP) were seeded in µ-dishes (Ibidi) and cultured for 24 h prior to imaging. Before FRAP, the medium was replaced with FluoroBrite™ DMEM (ThermoFisher) supplemented with 10% foetal bovine serum (FBS), L-glutamate, and gentamycin. Imaging was performed using a Leica DMi8 widefield microscope equipped with a FRAP module and a 488 nm laser.

Fluorescence was recorded every second for 10 seconds before photobleaching. Bleaching was performed at the apical tip of single parasites using a high-intensity pulse of the 488 nm laser, and fluorescence recovery was monitored every second for 60 seconds. Experiments were performed in biological triplicates, and at least six vacuoles were analysed in total across the replicates.

Image analysis was conducted using Fiji (ImageJ). Intensities were measured in the bleached area, a non-bleached parasite in the same field of view, and the background. Signals were corrected by subtracting background fluorescence and normalised using the non-bleached parasite to account for overall photobleaching. Recovery curves were generated from the normalised values. Data were then transformed so that the pre-bleach average intensity corresponded to 100%, ensuring comparability across samples with slightly different bleaching depths. To determine the characteristic recovery time (Tau, τ), the normalized post-bleach data were fitted to a single exponential function of the form:

$$I\left(t\right)=A\cdot \left(1-e^{-\frac{t}{\tau }}\right)$$

Where I(t) is the normalized intensity at time t, A is the plateau (maximum recovery) and τ is the time constant representing the time required to reach ~63% of the final recovery level. From this fit, the half-time of recovery (t₁/₂) was also calculated using the relationship:

$$t_{1/2}=\tau \cdot \ln \left(2\right)$$

which corresponds to the time at which fluorescence intensity reaches 50% of the recovery plateau.

Enrichment of methylated proteins and pull-down

Roughly 10⁸ parasites of the strains loxPpckmt-Halo, loxPakmt-syfp2 and double AKMT/PCKMT inducible KO ± rapamycin for 72 h were collected per technical (n = 2) and biological replicates (n = 3; total samples = 6 per strain/condition). Parasites were released mechanically and filtered through a 3 µm membrane (Millipore). Parasites were washed with ice-cold PBS. Pelleted parasites were resuspended in 1 ml RIPA buffer (0.5% sodium deoxycholate, 150 mM NaCl, 1 mM EDTA, 0.1% SDS, 50 mM Tris-HCl (pH 8.0), 1% Triton TX-100) with Pierce™ protease inhibitors and incubated at 4 ^oC for 1 h.

During this time, 100 µl of Protein G Dynabeads™ (Thermofisher) per sample were each conjugated with 9.4 µg of αH4K20m3 antibody (Abcam) diluted in 200 µL PBS with 0.02% Tween™20 (Sigma) for 10 min at room temperature, followed by 3 washes in PBS with 0.02% Tween™ and a final wash in RIPA buffer.

Lysed parasites were incubated for 1 h with the conjugated magnetic beads. After that, the beads were then washed five times with 1 mL of RIPA buffer without Triton TX-100 but with protease inhibitor, followed by three washes with 50 mM Tris-HCl (pH 8). Washing steps were performed on ice. 10 µl samples of lysed parasites prior to incubation with beads, output after incubation and last wash together with 10% of the washed beads were preserved for Western blot analysis; the rest of the beads were pelleted and stored at −80 °C before being sent for mass spectrometry.

Proximity labelling assay

10⁸ parasites of the strains pckmt-TurboID, akmt-TurboID and RHΔKu80DiCre (wild-type parasite) pre-treated for 2 or 6 h with or without 150 μM biotin were collected following the protocol described above in biological replicates (n = 4). Additionally, 10⁷ parasites treated with ±150 μM biotin were collected for biotinylation analysis via Western blot. Pelleted parasites were also lysed with RIPA buffer with protease inhibitors for 1 h. The supernatant from the lysis was incubated with 100 μL of beads per sample (Dynabeads™ MyOne™ Streptavidin T1, Invitrogen), prewashed with PBS, for 30 min at room temperature while being gently rotated. The beads were then washed five times with 1 mL of RIPA buffer without Triton TX-100 but with protease inhibitor, followed by three washes with 50 mM Tris-HCl (pH 8). Washing steps were performed on ice. While 10% of the beads were preserved for Western blot analysis, the rest of the beads were pelleted and stored at −80 °C before being sent for mass spectrometry.

Protein quantification by LC-MS

Immunoprecipitated proteins were quantified by LC-MS essentially as described previously, with minor modifications⁶³. Briefly, beads were incubated with 10 ng/µL trypsin in 1 M urea and 50 mM ammonium bicarbonate (NH₄HCO₃) for 30 min, followed by a wash with 50 mM NH₄HCO₃. The supernatant was then digested overnight in the presence of 1 mM DTT. After digestion, peptides were alkylated and desalted prior to LC-MS analysis. For LC-MS/MS, peptides were loaded onto an Ultimate 3000 RSLCnano system and separated on a 25 cm analytical column (75 µm inner diameter, 1.6 µm C18, Aurora-IonOpticks) using a 50-min gradient from 2 to 35% acetonitrile in 0.1% formic acid.

The effluent from the HPLC was directly electrosprayed into an Orbitrap Exploris 480 (Thermo) operated in data-dependent mode to automatically switch between full scan MS and MS/MS acquisition. Survey full scan MS spectra (from m/z 350-1200) were acquired with resolution R = 60,000 at m/z 400 (AGC target of 3 × 106). The 20 most intense peptide ions with charge states between 2 and 6 were sequentially isolated to a target value of 1×105, and fragmented at 30% normalised collision energy. Typical mass spectrometric conditions were: spray voltage, 1.5 kV; no sheath and auxiliary gas flow; heated capillary temperature, 275 °C; intensity selection threshold, 3×10⁵.

Protein identification and quantification were performed with MaxQuant v2.4.1.0 using iBAQ-based quantification. Searches were carried out against the Toxoplasma gondii protein database (Uniprot_UP00000152_Toxoplasmagondii_Me49_20221024-08.04.16.64.fasta) with the following parameters: precursor mass tolerance 10 ppm, fragment mass tolerance 20 ppm, peptide FDR 0.1, protein FDR 0.01, minimum peptide length 7, fixed modification: carbamidomethylation (C), variable modification: oxidation (M). Razor and unique peptides were used for quantification, with a minimum of one peptide per protein and at least two ratio counts.

Quantified protein values (iBAQ, Z-score normalised) were compared using the adjusted t-test implemented in Perseus (v.2.1.0)⁶⁴. Missing values were imputed from a normal distribution (width = 0.3, downshift = 4). Differential enrichment analysis was performed using a permutation-based false discovery rate (FDR) approach, with thresholds set at 0.05 for AKMT and DAMT-iKO, and 0.1 for PCKMT in the methylation enrichment assay. For the proximity labelling experiment, an FDR threshold of 0.08 was applied for PCKMT. In all cases, a constant s₀ parameter of 0.1 was used to stabilise variance estimation.

Volcano plots depicting the different abundance of proteins detected were constructed using R (v. 4.5.1) and RStudio (2025.09.1 Build 401).

Data analysis and statistics

Two-tailed Student t test and one-way ANOVA were used to determine the statistical significance between groups.

Software and online applications

All sequences were obtained from the ToxoDB database^65,66. Protein structure predictions and models were generated using ColabFold v1.5.2⁶⁷, Alphafold3⁶⁸, and ChimeraX (v. 1.9⁶⁹). sgRNA designing was aided by the Eukaryotic Pathogen sgRNA Design Tool (EuPaGDT)⁴⁵. In silico cloning was conducted using ApE (v. 3.1.3⁷⁰) and Benchling (https://benchling.com). BioEdit (v.7.0.5.3⁷¹) was used in the analysis of sequencing results. Image acquisition was carried out using Leica LAS X software (v. 3.10.0.28982; RRID: SCR_013673) for the Leica DMi8 inverted live cell widefield microscope and Imspector (v.16.3.19714-w2408) for the Abberior STED microscope. All image processing and analysis were done using ImageJ (v. 2.16.0/1.5p)^61,72. Cryo-ET data were acquired using SerialEM (v.4.1.1)⁵⁵. Relion-5 (v. 5.0.0)⁵⁷ and AreTomo2 (v. 1.1.2)⁵⁸ were used for tomogram reconstruction and IsoNet (v. 0.2.1)⁵⁹ for CTF-deconvolution. Tomogram segmentation was performed using MemBrain-seg (v. 0.0.5)⁶⁰ and Dragonfly (Comet Technologies Canada Inc. Dragonfly 3D World (v. 2024.1), Comet Technologies Canada Inc., https://dragonfly.comet.tech/). Visualisation of tomograms together with segmentations was prepared using FIJI/ImageJ (v. 2.16.0/1.54p)⁶¹, ChimeraX (v.1.8)⁶² and custom Python (v. 3.13.1) scripts. Inkscape v1.4 was employed to generate the figures for this manuscript (www.inkscape.org).

RStudio (v. 4.4.1 and 4.5.1) was utilised to perform statistical analysis and data visualisations.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

Author contributions

P.Q. performed the characterisation of PCKMT, generation of mutants and analysed the data. T.K.: performed all experiments on P. falciparum. O.K. performed cryo-ET, analysed the cryo-ET data and wrote the cryo-ET results section. W.L. created some of the strains used in this publication and performed some imaging assays. I.F. performed the LC-MS and conducted the relevant analysis. S.M. analysed the cryo-ET data and contributed resources. E.J-R. designed and coordinated the project and experiments, analysed the data, wrote the paper and contributed with resources.

Peer review

Peer review information

Nature Communications thanks Nishith Gupta, and Sabrina Absalon for their contribution to the peer review of this work. A peer review file is available.

Funding

P.Q. was supported by a China Scholarship Council fellowship (grant number 202106300007). E.J.-R. received support for this work from the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation) through the Priority Programme SPP 2225 “EXIT Strategies of Intracellular Pathogens” (project JI 463/2-2) and a DFG equipment grant (INST 86/1831-1). Open Access funding enabled and organized by Projekt DEAL.

Data availability

The mass-spectrometry proteomics data generated in this study have been deposited in the ProteomeXchange Consortium via the PRIDE⁷³ partner repository under accession code PXD067879. The cryo-electron tomography data generated in this study have been deposited in the figshare repository under 10.6084/m9.figshare.30674381⁷⁴ [https://figshare.com/account/articles/30674381]. The numerical data underlying the graphs and statistical analyses generated in this study are provided in the Source Data file. Additional data, including uncropped gel and blot images, are provided in the Supplementary Information. No data generated in this study are subject to restricted access or data-privacy limitations. Newly generated parasite strains, plasmids and related reagents are available from the corresponding author upon reasonable request, subject to applicable material-transfer and biosafety requirements.

Competing interests

The authors declare no competing interests.

Supplementary information

The online version contains supplementary material available at 10.1038/s41467-026-77078-y.