Rapid invasion of host cells by Toxoplasma requires secretion of the MIC2-M2AP adhesive protein complex

Abstract

Vertebrate cells are highly susceptible to infection by obligate intracellular parasites such as Toxoplasma gondii, yet the mechanism by which these microbes breach the confines of their target cell is poorly understood. While it is thought that Toxoplasma actively invades by secreting adhesive proteins from internal organelles called micronemes, no genetic evidence is available to support this contention. Here, we report successful disruption of M2AP, a microneme protein tightly associated with an adhesive protein called MIC2. M2AP knockout parasites were >80% impaired in host cell entry. This invasion defect was likely due to defective expression of MIC2, which partially accumulated in the parasite endoplasmic reticulum and Golgi. M2AP knockout parasites were also unable to rapidly secrete MIC2, an event that normally accompanies parasite attachment to a target cell. These findings indicate a critical role for the MIC2-M2AP protein complex in parasite invasion.

Fig. 1. Targeted disruption and genetic complementation of M2AP. (A) The MIC2–M2AP complex. MIC2 is a type I membrane protein that anchors the complex to the microneme/plasma membrane. MIC2 consists of a C-terminal cytoplasmic (C) domain, a motif (M) domain comprising six tandem segments similar to the type I repeat of thrombospondin and an integrin-like I-domain (I). M2AP has an N-terminal propeptide, which is removed (small arrow) in the parasite Golgi, a central domain consisting of β-sheets (β) and a C-terminal domain comprised of random coil sequence (coil). An anonymous parasite proteinase, MPP1, cleaves (large arrow) MIC2 near or within the transmembrane anchor, releasing the complex from the parasite surface. (B) Schematic depiction of the M2AP knockout and complementation strategy. The knockout plasmid, pm2apKO, contains an HXGPRT expression cassette for drug selection and targeting sequences designed to mediate double-crossover homologous recombination at the M2AP locus of the parental strain, ΔHX, thereby replacing M2AP with HXGPRT in the knockout clone, m2apKO. For complementation, m2apKO was co-transfected with a 1:10 ratio of pCAT and pM2AP and, after selection, parasites expressing CAT alone (2G11) or CAT and M2AP (1C4) were cloned and analyzed. (C) Southern blot analysis of m2apKO and associated clones. Genomic DNA was digested with NcoI and NaeI and hybridized with probes for HXGPRT, M2AP and CAT, respectively. The HXGPRT probe (left panel) hybridized to the endogenous HXGPRT locus disrupted by an insertional mutation in ΔHX (Donald and Roos, 1998). An additional band corresponding to incorporated HXGPRT at the M2AP locus is seen in m2apKO, 2G11 and 1C4. The M2AP probe (middle panel) hybridized to two bands of the expected size from the M2AP locus in ΔHX, which was deleted in m2apKO and 2G11 and restored by incorporation of pM2AP (expressing the M2AP cDNA) in 1C4. The faint bands in the upper part of the middle panel are due to incomplete stripping of the HXGPRT probe. Hybridization with a CAT probe (right panel) revealed incorporation of pCAT in 2G11 and 1C4. (D) Western blot probed with anti-actin and anti-M2AP antibodies. M2AP expression was undetectable in m2apKO and 2G11 but restored in 1C4. (E) Phase contrast and indirect immunofluorescence images of intracellular tachyzoites showing M2AP in the micronemes of ΔHX, loss of expression in m2apKO and 2G11, and re-expression of micronemal M2AP in 1C4. Scale bar, 5 µm. (F) Growth rate of m2apKO was similar to ΔHX, 2G11 and 1C4, based on the number of parasites in vacuoles determined 26 h after inoculation in HFF. Data represent three independent experiments, each from counting six fields/clone at 600× total magnification.

Fig. 2. Disruption of M2AP causes MIC2 retention in the ER/Golgi. (A) Photomicrographs of intracellular tachyzoites showing co-localization of MIC2 and AMA1 in ΔHX micronemes, but retention of MIC2 in the Golgi (closed arrowhead) or ER (open arrowhead) of m2apKO and 2G11. M2AP re-expression in 1C4 restored MIC2 localization to the micronemes (bottom row). Scale bar, 5 µm. (B) Illustration of a tachyzoite indicating the approximate positions of sections shown in (C) including a longitudinal section of Golgi (box with broken line) and a cross-section through the apical region (broken line). (C) Colloidal gold immunolabeling of ultrathin sections revealed normal localization of MIC2 in the apical micronemes of ΔHX, but an abnormal abundance of MIC2 in the Golgi of m2apKO with correspondingly less MIC2 in the micronemes. Re-expression of M2AP in 1C4 restored normal localization of MIC2 to the micronemes. Scale bar, 0.2 µm. (D) Quantification of immunogold labeling confirmed significantly enhanced (P = 0.0005, Mann–Whitney test) labeling of MIC2 in the Golgi of m2apKO compared with ΔHX and 1C4. Twelve randomly chosen sections were quantified. (E) Quantification of 10 randomly chosen apical sections revealed significantly less (P = 0.0007, Mann–Whitney test) MIC2 in the micronemes of m2apKO compared with ΔHX and 1C4. Bars represent mean values in (D) and (E).

Fig. 3. MIC2 requires M2AP for proper expression and rapid secretion. (A) Western blot showing equal expression of actin, GRA1, AMA1 and MIC4 in all clones, but reduced expression of MIC2 in m2apKO and 2G11 compared with ΔHX and 1C4. (B) Short exposure (left panel) of a western blot loaded with 1.0, 0.75 and 0.5 relative amounts of ΔHX tachyzoite lysate compared with a 1.0 equivalent (normalized to actin) of m2apKO lysate. A longer exposure (right panel) of the same blot shows an additional ∼95 kDa MIC2 band (arrow) exclusively in m2apKO. Relative molecular masses are indicated in kDa. (C) Western blots of parasite lysates probed with an I-domain-specific mAb 6E9 (left panel) or a C-domain-specific antibody MαC-dom. (right panel, intentionally overexposed). Absence of reactivity with MαC-dom. suggests that the ∼95 kDa species is a degradation product resulting from proteolytic removal of the C domain. (D) Western blots of culture supernatants from constitutively secreting (Const./Basal Sec’n) tachyzoites or from tachyzoites induced for microneme secretion by ethanol (EtOH-Ind. Sec’n) or A23187 (A23187-Ind. Sec’n). While constitutive/basal secretion (top left panel) of MIC2 was not substantially affected by disruption of M2AP, induced secretion was severely impaired (top center and top right panels). Lower panels show the same blots probed with antibodies to GRA1 as a loading control.

Fig. 4. Disruption of M2AP sharply reduces T.gondii attachment to host cells. (A) Representative fluorescence images of fibroblast monolayers infected with ΔHX and m2apKO parasites using a red/green invasion assay (see Materials and methods). External tachyzoites are stained red, internal tachyzoites are stained green and host nuclei are stained blue. (B) Quantification of invasion using the red/green assay. Red bars represent external, attached parasites while green bars represent internal, penetrated parasites. Asterisks indicate that attachment or penetration was significantly lower (P < 0.05, two-tailed Student’s t-test) than ΔHX. Tachyzoites treated with BAPTA-AM (B-AM) or cytochalasin D (CytD) were included as positive controls for defects in attachment or penetration, respectively. Data are mean values ± SEM of four independent experiments, counting six randomly selected fields for each sample. (C) Tachyzoite attachment to glutaraldehyde-fixed fibroblasts. A single asterisk indicates that attachment was significantly lower (P < 0.05, two-tailed Student’s t-test) than ΔHX and a double asterisk indicates that attachment was significantly higher than m2apKO (P < 0.05, two-tailed Student’s t-test). Data are mean values ± SEM of four independent experiments, counting six randomly selected fields for each sample.