Calmodulin-like proteins localized to the conoid regulate motility and cell invasion by Toxoplasma gondii

Abstract

Toxoplasma gondii contains an expanded number of calmodulin (CaM)-like proteins whose functions are poorly understood. Using a combination of CRISPR/Cas9-mediated gene editing and a plant-like auxin-induced degron (AID) system, we examined the roles of three apically localized CaMs. CaM1 and CaM2 were individually dispensable, but loss of both resulted in a synthetic lethal phenotype. CaM3 was refractory to deletion, suggesting it is essential. Consistent with this prediction auxin-induced degradation of CaM3 blocked growth. Phenotypic analysis revealed that all three CaMs contribute to parasite motility, invasion, and egress from host cells, and that they act downstream of microneme and rhoptry secretion. Super-resolution microscopy localized all three CaMs to the conoid where they overlap with myosin H (MyoH), a motor protein that is required for invasion. Biotinylation using BirA fusions with the CaMs labeled a number of apical proteins including MyoH and its light chain MLC7, suggesting they may interact. Consistent with this hypothesis, disruption of MyoH led to degradation of CaM3, or redistribution of CaM1 and CaM2. Collectively, our findings suggest these CaMs may interact with MyoH to control motility and cell invasion.

Fig 1. Endogenous tagging and generation of knockouts in T. gondii.

A. Schematic of the CRISPR/Cas9 tagging system. Tagging plasmids were generated with various tags (green box) flanked by common ends (red and black boxes) and including a common stop codon (gray box) followed by the HXGPRT 3’ UTR (yellow box) and the selectable marker HXGPRT. Amplification of this central region with primers that contained short homology regions HR1 (purple box) and HR2 (blue box) together with the common flanks (red and black boxes) generated products for gene-specific tagging. Co-transfection of these amplicons with a CRISPR/Cas9 plasmid bearing the gene-specific single guide RNA (sgRNA3’) was used to add an epitope tag (green box) at the C-terminus of the endogenous locus. See S1 Fig for more details. B. Localization of CaM1, CaM2 and CaM3 containing C-terminal 6HA tags. Detected with mouse anti-HA (green) and rabbit anti-GAP45 (red). Scale bar, 2 μM. C. Schematic of the double CRISPR/Cas9 gRNA system used for generation of clean knockouts using two sgRNAs matching the 5’ and 3’ ends of the coding sequence. The entire coding sequence was replaced by the DHFR marker flanked by short homology regions (HR3, red; HR2, blue). Primers (p) used for diagnostic PCR. D. Diagnostic PCR of knockouts compared to the parental ku80^KO line. CDPK1, PCR control. E. Plaque numbers formed by the knockouts compared to the parental ku80^KO line. ns, not significant, analyzed by one-way ANOVA.

Fig 2. Development of the auxin inducible degron (AID) system in T. gondii.

A. Structure of auxin indole-3-acetic acid (IAA). B. Plaque formation of the ku80^KO line grown in 500 μM IAA or ethanol control (0.1%) (mock) for 6 days. Scale bar = 2 mm. C. Plaque formation of the ku80^KO line grown in IAA or ethanol control (0.1%) mock) for 6 days. Mean ±S.D. from three independent experiments with triplicates for each (n = 9). ns, not significant, analyzed by one-way ANOVA. D. Heterologous co-expression of Oryza savita auxin receptor TIR1 (TIR1-3xFLAG)(parental line) (red) and an N-terminal YFP fusion with the Arabidopsis thaliana IAA17 (AID) (YFP-AID-3xHA) (green) in the TIR1 background. Scale bar, 2 μm. E. Schematic illustration of conditional degradation of AID-tagged proteins in T. gondii. F. Dose-response and G. time course of YFP-AID-3HA knockdown following IAA treatment by Western blotting with antibodies to YFP-AID-3HA (mouse anti-HA) or Aldolase (as a loading control).

Fig 3. Generation of AID tagged lines in the TIR parental line of T. gondii.

A. Western blot analysis using antibodies to detect CaM1-AID or CaM3-AID (mouse anti-HA to the AID-3HA tag), TIR1-3Flag (rat anti-Flag) and aldolase (rabbit anti-aldolase, ALD). B and C. Degradation of AID tagged proteins in cam2^KO/CaM1-AID (B) and CaM3-AID (C) lines after addition of auxin (500 μM IAA) for different time periods. Mock indicates parasites grown with 0.1% ethanol for 36 hr. CaM1-AID or CaM3-AID proteins were detected with mouse anti-HA and rabbit anti-aldolase (ALD) antibodies served as a loading control. Band intensities were analyzed by ImageJ, and ratios of anti-HA vs. anti-ALD signal were calculated (HA/ALD) and expressed as a percentage of the mock treatment (i.e. 100%). D and E. Degradation of AID tagged proteins in cam2^KO/CaM1-AID (D) and CaM3-AID (E) parasites after 24 hr incubation with 500 μM IAA (+IAA) or ethanol vehicle 0.1% (-IAA). CaM1-AID or CaM3-AID proteins were detected with mouse anti-HA (green) and rabbit GAP45 (red) antibodies served as a control to label the parasite. Scale bar, 2 μM. F. Plaque formation by parasites grown on HFF monolayers. Scale bar, 0.5 cm. Insert images in the CaM3-AID line, scale bar (red) = 1 mm. G. Measurement of plaque numbers and sizes for the CaM3-AID line treated with and without auxin. N≥ 25, ***, P < 0.0001. Mann Whitney non-parametric test.

Fig 4. Analysis of parasite replication, conoid protrusion, apical organelle distribution and secretion in parental and mutant lines.

A. Parasite replication after 24 hr incubation with ± IAA (500 μM vs 0.1% ethanol). ns, not significant. B. Proportion of parasites with extruded conoid. Parasites grown for 2 days ± IAA (500 μM vs 0.1% ethanol), stimulated with 3 μM A23187 or DMSO vehicle control for 10 min. ns, not significant. C and D. Distribution of MIC2 (mouse anti-MIC2 (green) and ROP5 (rabbit annti-ROP5 (green) upon depletion of AID fusion proteins. Parasites grown ± IAA (500 μM vs 0.1% ethanol) for 24 hr in HFF monolayers and stained for IFA. Parasites were counterstained with mouse anti-IMC1 (red) or rabbit anti-GAP45 antibodies (green). Scale bar, 2 μM. E. Quantification of micronemal secretion using MIC2-GLuc-myc reporter lines. Parasites were grown for 2 days ±IAA (500 μM vs 0.1% ethanol), stimulated with 1% ethanol—1% BSA and secretion was monitored by releases of luciferase (see methods). Relative Luminescence Unit (RLU). ns, not significant. F and G. Detection of rhoptry secretion by ROP1 staining. Parasites were grown for 2 days ± IAA (500 μM vs 0.1% ethanol), harvested and used to detect formation of evacuoles (arrows) on fresh monolayers of HFF cells in the presence of cytochalasin. Parasites were counted from triplicate samples on three separate experiments and ratios of parasites associated with evacuoles in were plotted. Scale bar, 5 μm. Panels A, B, E, F, G mean ± S.D. from three independent experiments with triplicates for each (n = 9). One-way ANOVA with Tukey’s multiple comparison test for B and E and two-way ANOVA with Tukey’s multiple comparison test for pair-wise multiple comparisons across each vacuole size for A, Man-Whitney non-parametric test for F and G.

Fig 5. Analysis of egress, invasion, and motility in parental and mutant lines.

A. Parasites grown for 30 hr ± IAA (500 μM vs 0.1% ethanol) were stimulated with 3 μM A23187 to simulate egress. Rabbit anti-GRA7 (red) and mouse anti-IMC1 (green) antibodies were used to distinguish intact vs. egressed vacuoles. *** P ≤ 0.0001, significant for the time points of 2, 5, 10 and 15 min, but not significant for 0 and 20 min. Scale bar, 5 μM. B. Quantitative analysis of invasion by parasites grown for 2 days ± IAA (500 μM vs 0.1% ethanol) and used to challenge fresh HFF monolayers on coverslips for 20 min. Extracellular parasites (invaded) were distinguished from those that remained extracellular (attached) by differential IFA staining (see methods). *** P ≤ 0.0001. C. Evaluation of cell entry past the moving junction. Parasites grown for 2 days ± IAA (500 μM vs 0.1% ethanol) were used to challenge fresh HFF monolayers on coverslips for 3 min, fixed and stained with rabbit anti-RON4 (green) and mouse anti-SAG1 (red) without permeabilization. Parasites with RON4 dots were considered to be apically attached (red column), and parasites with RON4 positive rings were classified as partially invaded (green column). *** P ≤ 0.0001. Scale bar, 2 μM. D. Parasite motility as monitored by video microscopy. Parasites grown for 2 days ± IAA (500 μM vs 0.1% ethanol) were allowed to glide on serum-coated coverslips. Time-lapse video microscopy was used to score different motile behaviors. *** P ≤ 0.0001, the cam2^KOCaM1-AID line showed significant decrease in twirling and increase non-productive movement when grown in +IAA vs. -IAA or the TIR1 parental line, **, P ≤ 0.0001, the CaM3-AID line showed a significant decrease in twirling and increase in circling when grown in +IAA vs. -IAA or the TIR1 parental line. Panels A, B, C, D represent means ± S.D. from three independent experiments with triplicates for each (n = 9). Two-way ANOVA with Tukey’s multiple comparison test for A, C and D, and one-way ANOVA with Tukey’s multiple comparison test for B.

Fig 6. Assessment of the roles of EF hand domains in CaM1 and CaM2.

A. Calcium-dependent solubility as detected by cell fractionation and Western blotting. Tagged parasites were lysed in 1% Triton X-100 in the presence of either 5 mM EDTA or 5 mM CaCl₂ and fractionated by centrifugation. CaM1, CaM2, or CaM3 were detected with mouse anti-HA (green), while mouse anti-IMC1 was used as a control for the pellet (p) and rabbit anti-aldolase (ALD) was used as a control for the supernatant (s). B. Diagram of wild type (WT) and CaM1 and CaM2 mutants showing the residues in conserved or degenerated EF hands (predicted by ScanProsite). Names of the proteins are shown to the right (i.e. D38A represents an Asp residue at D38 that was mutated to Ala). C and D. Structural modeling of TgCaM1 (C), TgCaM2 (D) highlighting their EF hand domains. C, Top: Structure of TgCaM1 is shown with conserved Asp residues (red) chelating a calcium iron (green ball). Bottom: enlargement of the TgCaM1 EF1 domain showing the wild type (left) and the triple EF1m mutant (right). D, Top: The structure of TgCaM2 is shown with conserved Asp residues (red) in the EF1 but not in EF2. Bottom: enlargement of TgCaM2-EF1 showing intact EF1 domain that chelate calcium (right) and the degenerate EF2 domain (left). E. Western blot detection of CaM mutants grown for 2 days ± IAA (500 μM vs 0.1% ethanol). Cell pellets were resolved by SDS-PAGE and Western blotted using mouse anti-HA to detect CaM-AID fusions, mouse anti-Ty to detect complementing alleles, and rabbit anti-aldolase (ALD) antibodies as a loading control. F. Evaluation of complementation by plaque formation. Scale bar, 0.5 cm.

Fig 7. Interaction of CaM1, CaM2 and CaM3 with MyoH at the conoid.

A. Localization of Ty-tagged CaMs (green) in a MyoH-3HA strain (red) by AiryScan super-resolution microscopy. Conoid extrusion was stimulated prior to fixation and IFA staining, MLC1 (blue) used to detect the parasite. Scale bar, 0.5 μm. B. Biotin labeling of CaM lines labeled with BirA and analyzed by MS/MS. Normalized Spectral Abundance Factors (NSAF) were plotted as average values from the combination of two independent experiments. Further data are found in S4 Table. C. Localization of CaMs upon depletion of MyoH. Parasites were grown for 8 hr ± IAA (500 μM vs 0.1% ethanol and evaluated by IFA using mouse anti-Ty (green) and rabbit anti-aldolase (red) antibodies. SAS6SL was tagged with Ty and served as a control. D. Western blot detection of CaMs (anti-Ty) in the MyoH-AID (anti-HA) strain. Rabbit aldolase (ALD) as a loading control. See S4 Fig for more details.