The apical annuli of Toxoplasma gondii are composed of coiled-coil and signalling proteins embedded in the inner membrane complex sutures

Abstract

The apical annuli are among the most intriguing and understudied structures in the cytoskeleton of the apicomplexan parasite Toxoplasma gondii. We mapped the proteome of the annuli in Toxoplasma by reciprocal proximity biotinylation (BioID), and validated five apical annuli proteins (AAP1-5), Centrin2, and an apical annuli methyltransferase. Moreover, inner membrane complex (IMC) suture proteins connecting the alveolar vesicles were also detected and support annuli residence within the sutures. Super-resolution microscopy identified a concentric organisation comprising four rings with diameters ranging from 200 to 400 nm. The high prevalence of domain signatures shared with centrosomal proteins in the AAPs together with Centrin2 suggests that the annuli are related and/or derived from the centrosomes. Phylogenetic analysis revealed that the AAPs are conserved narrowly in coccidian, apicomplexan parasites that multiply by an internal budding mechanism. This suggests a role in replication, for example, to provide pores in the mother IMC permitting exchange of building blocks and waste products. However, presence of multiple signalling domains and proteins are suggestive of additional functions. Knockout of AAP4, the most conserved compound forming the largest ring-like structure, modestly decreased parasite fitness in vitro but had no significant impact on acute virulence in vivo. In conclusion, the apical annuli are composed of coiled-coil and signalling proteins assembled in a pore-like structure crossing the IMC barrier maintained during internal budding.

Keywords: BioID; Toxoplasma gondii; apicomplexa; cytoskeleton; proteomics.

Figure 1.. Protein-protein interaction (PPI) network analysis of the apical annuli.

A. Episomal expressed Ty-BioID2-Centrin2 localizes predominantly to the apical annuli (arrows) but also to the preconoidal ring, the centrosome and the basal complex. Application of 150 μM biotin overnight facilitates biotinylation as detected with streptavidin-A594 (biotinylated proteins, red). Endogenous biotinylation of apicoplast proteins is always detected with streptavidin (asterisks). B. Endogenously triple Myc-tagged (Myc₃) TGGT1_230340 localizes towards the apical end of the parasites in 5-6 puncta. Parasites were co-stained with beta-tubulin (red) to highlight the parasite’s periphery. C. Airyscan imaging of TGGT1_230340 Myc₃-tagged parasites transiently expressing Ty-Centrin2 (red). TGGT1_230340 only colocalizes with Centrin2 at the apical annuli and is not present in other subcellular Centrin2 localizations. Due to its localization we named TGGT1_230340 apical annuli protein 4 (AAP4). D. AAP4 was endogenously tagged with Ty-BioID2 at the N-terminus (see Fig. S1E). Eipon application of 150 μM biotin, increased biotinylation of the apical annuli can be detected by streptavidin staining (biotinylated proteins, red). Endogenously biotinylated proteins of the apicoplast are always detected (asterisks). Blue: DAPI stain. E. PPI networks were modeled by calculating probabilistic bait-prey interactions using SAINTexpress (Lambert et al., 2015) and plotted with Cytoscape (Saito et al., 2012, Shannon et al., 2003). Preys with an AvgP (average individual probability for SAINTexpress analysis) ≥ 0.5 are shown for each bait. F. The statistical support for the core protein set of the apical annuli is shown as a dot plot generated with ProHits-viz (Knight et al., 2017). Note that the relative abundance compares between the two samples and not within an individual sample (e.g. AAP4 was one of the most abundant proteins identified in Centrin2 BioID data, but compared to the number of spectra in the AAP4 data set it deceivingly appears to be relatively low abundant in the Centrin2 data set). Also note that AAP2 is present in the AAP4 BioID data set and TSC1 and TSC3 in the Centrin2 BioID data set, but only with low AvgP scores (< 0.2). AvgP: average individual probability for SAINTexpress analysis, SpecSum: sum of all spectra for an individual protein. G. The data set was further assembled into a prey-prey correlation map showing the apical annuli core set. The distance between individual preys is expressed by color, Black indicates correlating preys, non-correlating preys are shown in white. See Table S1 for the complete data.

Figure 2.. Localization and dynamics of the apical annuli proteins (AAPs).

A. AAP genes were endogenously tagged with a triple Myc-epitope tag (Myc₃) and co-localized with a specific antisera that recognizes AAP4 (red). Images were acquired on the Airyscan system. Yellow arrows indicate cross-reactive signal seen with the AAP4 antiserum close to the nucleus (see also Fig. 6A and Fig. S5B). B. Protein dynamics of identified AAPs were further followed along the intracellular division cycle of the tachyzoite. The periphery of the parasite is visualized with Tg-β-tubulin antiserum staining (red). Arrowheads in the middle plane indicate annuli presence in early daughter buds. All AAPs are present in late daughter buds when the mother’s cytoskeleton is being degraded. Blue: DAPI stain. C. AAPs localize to the apical annuli in extracellular parasites, although AAP5 only exhibits a weak signal.

Figure 3.. Annotation and sequence analysis of the AAPs.

A. AAP annotation and functional data available through ToxoDB.org (Gajria et al., 2008). Fitness scores defined by genome-wide CRISPR screen for fitness across three lytic cycles (Sidik et al., 2016). *In GT1 annotation AAP1 comprises two genes: TGGT1_242790A (fitness score 0.02, with an extended N-terminus) and TGGT1_242790B (fitness score −1.01). B. Domain annotations of AAP1-5 made through searches on ToxoDB, SMART, and PFam databases, NCBI Nr PBLAST searches and coiled-coil predictions (coils window size of 28; 100% probability predictions are shown). The long, largely α-helical domain in AAP5 is between brackets as this feature is not displayed for AAP1-4, but it is the only distinguishable feature in AAP5. Phosphorylation sites detected in the tachyzoite phosphoproteome are marked with vertical ticks (Treeck et al., 2011). The number at the C-terminus indicates the number of amino acids. C. Coils prediction by the coils server using a window size of 21 overlaid with the PSIPRED predicted α-helical repeats (ToxoDB) identify 11 repeated regions (y-axis represents 0-100% probability range of α-helix or coiled-coil), whose consensus repeat sequence is provided in the logo plot (Crooks et al., 2004) at the bottom (see also Fig. S3). D. Conservation of the AAP proteins across the Apicomplexa accessed through BLASTP of the Toxoplasma AAPs against EuPathDB (Aurrecoechea et al., 2013) and OrthoMCL (Chen et al., 2006). Colors represent likelihood of functional conservation based on manual assessment of the quality and length of sequence alignments and genomic synteny (green: robust ortholog; yellow: putative ortholog; red: no ortholog), cyst: cyst forming; monox: monoxenic n.h.: no homology; endo: asexual division by endodyogeny and/or endopolygeny; schizo; asexual division by schizogony. The absence of tissue cysts in C. suis is likely a secondary loss in this lineage as it is phylogenetically is more closely related to Toxoplasma than to Sarcocystis (Carreno et al., 1998).

Figure 4.. The methyltransferase AAMT localizes to the annuli in intracellular parasites.

A. Endogenously triple-Myc tagged (Myc₃) AAMT displays apical annuli localization in mature parasites (top panels, boxed area magnified 2-fold in inset). We further detected AAMT signal association with the conoid and in forming daughter buds (lower panels, marked by arrowheads). B. Airy scan imaging of AAMT-Myc₃ parasites co-localized with AAP4 antiserum (red). White arrows indicate the apical annuli, as highlighted by AAMT and AAP4 staining. Note that the AAMT localization to the annuli is not as complete as detected for other AAPs. Yellow arrows indicate cross-reactive signal seen with the AAP4 antiserum close to the nucleus (see also Fig. 6A and Fig. S5B). C. In extracellular parasites AAMT redistributes to the (transverse) IMC sutures but remains associated with the conoid (asterisks). Tg-β-tubulin serum (red) (Morrissette & Sibley, 2002) highlights the cortical cytoskeleton. Blue: DAPI stain.

Figure 5.. The apical annuli display a concentric ring-like architecture.

A. SR-SIM analysis of endogenously Myc-tagged AAP and Centrin2 parasite lines co-stained with Tg-β-tubulin antiserum (Morrissette & Sibley, 2002). The lower 50% of the imaged stack was combined into a Z projection for each image. See Supplementary Movies S1 and S2 for a 3D-reconstruction of the entire AAP4 image stack. B. Quantification of AAP donuts. Diameters of at least 40 individual annuli were measured for each cell line. Error bars represent standard deviation. Statistics: paired two-tailed t-test analysis indicated that all signals except AAP5 and Centrin2 (p=0.485) are significantly different from each other (p<0.0001). C. SR-SIM analysis of C-terminally triple-Myc tagged AAP2 co-stained with a specific antiserum recognizing AAP4. AAP2 signals, shown to exhibit the smallest average annuli diameter, were observed as donuts and are embedded by the AAP4 signal. Image is cropped to the apical section of a tachyzoite. D. SR-SIM analysis of C-terminally triple-Myc tagged AAP4 co-stained with a specific ISC2 antiserum. The annuli, highlighted by AAP4, localize to the apical end of the IMC sutures (ISC2 signal). See Supplementary Movie S3 for a 3D-reconstruction of the entire image stack. E. Schematic presentation of the apical annuli architecture incorporating microscopy and PPI data. PI-PLC interaction is gleaned from (Hortua Triana et al., 2018).

Figure 6.. Knockout analysis of AAP4 reveals decreased fitness in vitro and reduced virulence in vivo.

A. IFA using specific α-AAP4 serum (green). White arrowheads indicate AAP4 signal at the annuli, yellow arrows indicate a cross-reactive signal seen close to the nucleus (see also Fig S5B). Parasite periphery is shown by α-IMC3 serum (red). Blue: DAPI stain. B. Representative plaque assays of AAP4-KO parasites after seven days of growth. AAP4-KO parasites form significantly smaller plaques compared to control (RHΔKu80) or AAP4-KO complement parasites. C. Quantification of plaque assays. Three biological replicates are shown; 40-80 independent plaques per condition and experiment were quantified; error bars represent standard deviation; For statistical analysis an one-way analysis of variance (ANOVA) test was applied and significance determined with a post-hoc Tukey’s honestly significant difference (HSD) test. D. C57BL/J6 mice infected with 1000 AAP4-KO parasites survive one day longer than mice infected with equal numbers of control (RHΔKu80) or AAP4-KO complemented parasites. Weight changes relative to the starting day (day 0) are shown for each of the four mice per group (round symbols). Horizontal bars represent the group average. Weight change patterns did not show significant differences between the groups, as tested by one-way ANOVA test. An additional infection experiment with a 100-parasite inoculum is present in Fig. S6.