A Novel Toxoplasma Inner Membrane Complex Suture-Associated Protein Regulates Suture Protein Targeting and Colocalizes with Membrane Trafficking Machinery

Abstract

The cytoskeleton of Toxoplasma gondii is composed of the inner membrane complex (IMC) and an array of underlying microtubules that provide support at the periphery of the parasite. Specific subregions of the IMC carry out distinct roles in replication, motility, and host cell invasion. Building on our previous in vivo biotinylation (BioID) experiments of the IMC, we identified here a novel protein that localizes to discrete puncta that are embedded in the parasite's cytoskeleton along the IMC sutures. Gene knockout analysis showed that loss of the protein results in defects in cytoskeletal suture protein targeting, cytoskeletal integrity, parasite morphology, and host cell invasion. We then used deletion analyses to identify a domain in the N terminus of the protein that is critical for both localization and function. Finally, we used the protein as bait for in vivo biotinylation, which identified several other proteins that colocalize in similar spot-like patterns. These putative interactors include several proteins that are implicated in membrane trafficking and are also associated with the cytoskeleton. Together, these data reveal an unexpected link between the IMC sutures and membrane trafficking elements of the parasite and suggest that the suture puncta are likely a portal for trafficking cargo across the IMC. IMPORTANCE The inner membrane complex (IMC) is a peripheral membrane and cytoskeletal system that is organized into intriguing rectangular plates at the periphery of the parasite. The IMC plates are delimited by an array of IMC suture proteins that are tethered to both the membrane and the cytoskeleton and are thought to provide structure to the organelle. Here, we identified a protein that forms discrete puncta that are embedded in the IMC sutures, and we show that it is important for the proper sorting of a group of IMC suture proteins as well as maintaining parasite shape and IMC cytoskeletal integrity. Intriguingly, proximity labeling experiments identified several proteins that are involved in membrane trafficking or endocytosis, suggesting that the IMC puncta provide a gateway for transporting molecules across the structure.

Keywords: Apicomplexa; BioID; Toxoplasma gondii; dynamin-related protein; inner membrane complex.

FIG 1

TGGT1_202220 localizes to cytoplasmic spots along the IMC sutures and is associated with the cytoskeleton. (A) Diagram of TGGT1_202220 showing two predicted coiled-coil (CC) domains. (B) IFA of endogenously 3×HA-tagged TGGT1_202220 showing several faint puncta in the cytoplasm (arrows). Magenta, mouse anti-HA; green, rabbit anti-IMC6. (C) IFA showing that the cytoplasmic puncta appear near the apical ends of the developing daughter buds during endodyogeny (arrow). Magenta, mouse anti-HA; green: rabbit anti-IMC6. (D) IFA showing endogenous tagging of TGGT1_202220 with spaghetti monster HA (smHA) enables better detection of the cytoplasmic puncta (arrow). Magenta, mouse anti-HA; green, rabbit anti-IMC6. (E) TGGT1_202220 tagged with smHA also shows enhanced staining near the developing daughter buds (arrow). Magenta, mouse anti-HA; green, rabbit anti-IMC6. (F) IFA showing that smOLLAS-tagged TGGT1_202220 puncta colocalize with the IMC sutures (ISC6-3×HA). The arrow in the top panel shows a punctum on a longitudinal suture. The arrow in the bottom panel points to a punctum on a transverse suture. Insets shows magnifications of the boxed regions highlighting suture colocalization. Magenta, rat anti-OLLAS; green, mouse anti-HA. (G) Western blot analysis of TX-100 detergent fractionation shows that 3×HA-tagged TGGT1_202220 partitions to the cytoskeletal pellet with the alveolin IMC6 and is not released like the membrane-associated IMC protein ISP3. T, total; S, detergent-soluble supernatant; P, detergent-insoluble cytoskeletal pellet. Bars = 2 μm.

FIG 2

Gene knockout of ISAP1. (A) IFA showing lack of HA staining in Δisap1 parasites. Magenta, mouse anti-HA; green, rabbit anti-IMC6. (B) PCR and diagram showing that the Δisap1 strain contains the correct amplicon for the replacement of ISAP1 with the selectable marker hypoxanthine-xanthine-guanine phosphoribosyl transferase (HPT) and lacks the ISAP1-coding amplicon. Primer positions are shown with arrows, and amplicons agree with the anticipated sizes for the knockout using wild-type genomic DNA as a control. (C) Complementation with the ISAP1 coding sequence driven from the ISC6 promoter restores the spot-like pattern similar to the wild-type protein. Magenta, mouse anti-HA; green, rabbit anti-IMC6. (D) Western blot analysis of HA-tagged Δisap1 and complemented strains shows that complementation restores levels of the ISAP1 protein similar to that seen for the HA-tagged strain. (E) Quantification of plaque assays showing disruption of ISAP1 results in a 69% decrease in plaque size (****, P < 0.0001). The defect is rescued by complementation (ISAP1^comp). The ISAP1 phosphomutant (Fig. S3) also rescues the plaque defect. (F) Graph showing an 80% reduction in plaque efficiency of Δisap1 parasites (****, P < 0.0001), which is mostly rescued upon complementation. (G) Ionophore induced egress is not significantly affected in Δisap1 parasites. (H) Host cell invasion is reduced by 42% in Δisap1 parasites (*, P > 0.05). (I) IFA at 24 h postinfection showing that loss of ISAP1 results in swollen parasites that have dysregulated endodyogeny and morphological defects. The arrow points to a swollen parasite. Magenta, mouse anti-ISP1; green, rabbit anti-IMC6. (J) IFA at 32 h postinfection showing more severe defects in morphology and daughter cell formation. The arrow points to four daughter buds in a maternal parasite. Magenta, mouse anti-ISP1; green, rabbit anti-IMC6. (K) Quantification of vacuoles with misshapen parasites and/or dysregulated endodyogeny in Δisap1 parasites (*, P < 0.05). Bars = 2 μm.

FIG 3

Loss of ISAP1 affects the integrity of the IMC cytoskeleton and results in mistargeting of a subgroup of cytoskeletal IMC suture proteins. (A) IFA of IMC1 and IMC6 in Δisap1 parasites showing gaps in the IMC cytoskeleton (arrows). The more internal IMC1/IMC6 staining pattern throughout the figure is due to the planes imaged for visualizing the gaps or sutures at the periphery of the parasite. Green, mouse anti-IMC1 and rabbit anti-IMC6. (B) IFA with the cytoskeletal transverse sutures protein TSC2 also shows the gaps in the Δisap1 cytoskeleton (arrows). TSC2 correctly targets the transverse sutures. Magenta, mouse anti-HA; green, rabbit anti-IMC6. (C) Quantification of the IMC6 cytoskeletal gaps showing 94.7% of Δisap1 vacuoles have parasites with gaps (***, P < 0.001). (D) IFA of Δisap1 parasites showing that HA-tagged ISC3, ISC5, and ISC6 correctly target the sutures. Magenta, mouse anti-HA; green, rabbit anti-IMC6. (E) IFA of Δisap1 parasites showing that HA-tagged TSC3 and -4, Ty-tagged TSC5, and Myc-tagged TSC6 also correctly target the sutures. Magenta, mouse anti-HA, mouse anti-Ty, mouse anti-Myc; green, rabbit anti-IMC6. (F and G) IFA showing that ISC1 and ISC2 are mislocalized to cytoplasmic spots in Δisap1 parasites. Magenta, rat anti-ISC1, rat anti-ISC2; green, rabbit anti-IMC6. (H) Disruption of ISAP1 in an ISAP1-smOLLAS- and ISC4-HA-tagged background results in the loss of ISC4. Magenta, mouse anti-HA (ISC4); green, rabbit anti-IMC6. (I) Western blot analysis of TX-100 detergent fractionation shows that ISC2 is no longer associated with the cytoskeleton in Δisap1 parasites. T, total; S, detergent-soluble supernatant; P, detergent-insoluble cytoskeletal pellet. Bars = 2 μm.

FIG 4

The CC domains of ISAP1 are not required for function. (A) Diagram showing deletions of the CC domains individually or together. (B) ISAP1 CC domains predicted by the COILS server (43). (C) IFA showing that the 3×HA-tagged CC deletions localize to puncta similarly to wild-type ISAP1. Magenta, mouse anti-HA; green, rabbit anti-IMC6. Bars = 2 μm. (D) Western blot analysis comparing expression levels of 3×HA-tagged ISAP1^comp to the CC deletions. (E) Plaque assays showing that ISAP1 CC deletion constructs can fully rescue the plaque defect of Δisap1 parasites (**, P < 0.01).

FIG 5

The N-terminal region of ISAP1 is important for punctum trafficking and function. (A) Diagram showing the N- and C-terminal deletion constructs. (B) Plaque assays showing that the conserved C-terminal region of ISAP1 can be deleted but the N-terminal 121 amino acids of the protein is important for function (**, P < 0.01; ***, P < 0.001). (C) IFA showing that the full-length smMyc-tagged ISAP1 complement (comp) targets puncta similarly to the endogenously smHA-tagged protein (wt) (arrows). Loss of the N-terminal 121 or 250 amino acids of ISAP1 results in relocalization to the periphery of the parasite, consistent with IMC localization. Magenta, rabbit anti-HA; green, mouse anti-Myc. Bars = 2 μm. (D) Detergent fractionation showing that 3×HA-tagged Δ2–121 and Δ2–250 constructs still associate with the cytoskeleton. ISP3 and IMC6 were used as controls for the membrane and cytoskeletal fractions, respectively. T, total; S, detergent-soluble supernatant; P, detergent-insoluble cytoskeletal pellet. (E) Western blot analysis comparing expression levels of 3×HA-tagged ISAP1^comp to those of proteins with N- and C-terminal deletions.

FIG 6

In vivo biotinylation using ISAP1 as bait identifies candidate interacting proteins. (A) Diagram showing the ISAP1-BioID2 fusion generated by endogenous gene tagging. A 3×HA tag is included for detection of the fusion protein. (B) IFA showing that the ISAP1-BioID2 fusion targets puncta similarly to the wild-type protein and is active, as assessed by faint streptavidin staining upon addition of biotin to the medium (arrows). Asterisks indicate the endogenously biotinylated signal in the apicoplast. Magenta, mouse anti-HA; green, streptavidin 488. Bars = 2 μm. (C) Table showing the top 25 hits from streptavidin purification of ISAP1-BioID2 following detergent fractionation of the cytoskeleton. Two replicates were performed (replicates are labeled A and B), and untagged parasites plus biotin were used as the control. Spectral counts are shown for each sample. GWCS, phenotype score assigned in a genome-wide CRISPR/Cas9 screen (22).

FIG 7

Endogenous tagging of candidates reveals proteins that colocalize with ISAP1. (A) IFA showing the EPS15/intersectin-1-like protein TGGT1_227800-smOLLAS localizes to discrete spots that colocalize with ISAP1-smHA. Magenta, rabbit anti-HA; green, rat anti-OLLAS. (B) IFA showing that TGGT1_297520-3×HA also colocalizes with the ISAP1-smOLLAS puncta. Magenta, rat anti-OLLAS; green, mouse anti-HA. (C) The AP2 adaptor complex member TGGT1_272600-3×HA also colocalizes with ISAP1-smOLLAS. Magenta, rat anti-OLLAS; green, mouse anti-HA. (D and E) Detergent fractionation shows that TGGT1_227800 and TGGT1_297520 are tethered to the cytoskeleton, although both of the proteins are labile and suffered substantial reproducible breakdown during fractionation. ISP3 and IMC6 were used as controls as described above. (F) IFA showing that 3×HA-tagged TGGT1_227800, TGGT1_297520, and TGGT1_272600 all retain their suture punctum localization in Δisap1 parasites. Green, mouse anti-HA; magenta, rabbit anti-Myc detecting 3×Myc-tagged ISC6. Bars = 2 μm.

FIG 8

The membrane trafficking protein DrpC colocalizes with ISAP1 and associates with the cytoskeleton. (A) IFA showing that 3×HA-tagged DrpC mostly colocalizes with ISAP1-smOLLAS (white arrows), though unique DrpC spots are also present (green arrows). Magenta, rat anti-OLLAS; green, rabbit anti-HA. (B) Detergent fractionation shows that a substantial portion of DrpC is surprisingly tethered to the cytoskeleton. ISP3 and IMC6 were used as controls as described above. (C) The related dynamin-related protein DrpB is readily solubilized by detergent extraction. (D) IFA showing that 3×HA-tagged DrpC retains its cytoplasmic punctum localization in Δisap1 parasites. Magenta, mouse anti-HA; green, rabbit anti-IMC6. Bars = 2 μm.