Attachment of Toxoplasma gondii to a Specific Membrane Fraction of CHO Cells

Abstract

We have observed previously that attachment of Toxoplasma gondii to synchronized host cells is considerably increased at the mid-S phase (4 h postrelease). Synchronized CHO host cells at the mid-S phase were fractionated by molecular weight, and the antigens were used to produce a panel of polyclonal mouse antisera. The polyclonal antisera raised against fraction 4 with molecular mass ranging approximately from 18 to 40 kDa significantly reduced attachment to mid-S-phase host cells. Immunofluorescence assays demonstrated strong reactivity to mid-S-phase host cells and identified a number of potential receptors on Western blots. These data indicate that there is a specific host membrane receptor for parasite attachment that is upregulated during the mid-S phase of the host cell cycle.

Toxoplasma gondii is an obligate intracellular parasite that causes one of the most common parasitic infections of humans and other mammals. Parasite invasion of the host cells is essential in order for the parasite to undergo a replicative phase. First, recognition and attachment of the parasites to the host cells must occur. Following the attachment process, the parasite initiates the invasion through the plasma membrane. During penetration through the cell plasma membrane, a constricted site, which moves toward the posterior end of the parasite, forms. Once the parasite is inside, a small cytoplasmic projection closes the host cell plasma membrane, leaving the parasite within a parasitophorous vacuole.

Several candidate ligands have been evaluated in the attachment process. Both surface antigens, in particular SAG1 (p30) (2, 3, 4, 5, 9, 14), and microneme proteins (such as MIC2) have been implicated as important in the process of attachment (5). Alteration of these ligands results in decreased infectivity of the parasite. Recent data would suggest that at least some of the microneme proteins are principally involved in the process of host cell penetration (1, 2).

That T. gondii can infect a wide range of host cells would imply the presence of a common receptor(s) (6, 7). Competition assays have demonstrated that parasite attachment can be saturated (14). Moreover, attachment by the parasite to the host cells may be specific in that a receptor-mediated interaction occurs with higher frequency during the mid-S phase of the host cell cycle (8). Murine polyclonal antibody raised against host cells harvested at mid-S phase showed more significant inhibition of attachment of host cells by toxoplasmas than did polyclonal antibody raised against host cells harvested at the beginning of the S phase. This observation indicates that the potential receptor for attachment to the host cell by toxoplasmas is abundant during the mid-S phase. In this report, we characterize and partially identify the host cell receptor for parasite attachment that is upregulated during the mid-S phase of the cell cycle.

Murine polyclonal antibody against a specific host cell membrane fraction inhibits parasitic attachment.

The parasites (RH strain) were maintained by serial passage in a human foreskin fibroblast confluent monolayer (11, 15) cultured in modified Eagle's medium with 10% heat-inactivated fetal calf serum and antibiotic as described previously (10). Chinese hamster ovary cells (CHO-pro 5; ATCC CRL-1781) were used as host cells in all experiments. These cells were maintained in alpha minimum Eagle medium supplemented with 10% heat-inactivated fetal calf serum and antibiotics.

CHO cells were synchronized by a combination of serum starvation and hydroxyurea block (12, 16, 17). Cells were seeded at low density with medium containing 10% fetal calf serum for 24 h at 37°C. The monolayers were serum starved for 48 h and then treated with 1 mM hydroxyurea, 10% fetal calf serum, and antibiotics for 12 h. Cells were washed with warm medium and allowed to progress in synchrony to the S phase by incubation in medium containing 10% serum. Cell synchronization and passages through the S phase were monitored by fluorescence-activated cell sorter analysis. At the appropriate time points, cells were fixed in 95% ethanol overnight. The ethanol was washed off with phosphate-buffered saline (PBS). Cells were resuspended in 400 μl of PBS, and 50 μl of RNase A (5 mg/ml) and 500 μl of propidium iodide (50 μl/ml in 50 mM sodium citrate) were added. Data were analyzed by the Modfit program.

The synchronized cells were harvested 4 h postrelease (H-4 antigen) by scraping the cells off the flask and washing twice with ice-cold PBS. They were incubated with Dounce buffer with protease inhibitor (10 mM Tris-Cl [pH 7.6], 0.5 mM MgCl₂, 10 μg of leupeptin/ml, 10 μg of aprotinin/ml, 1 mM phenylmethylsulfonyl fluoride in 100% ethanol, 1.8 mg of iodoacetamide/ml) for 10 min at 4°C. After the cells were pressurized in a nitrogen cavitation device to 100 lb/in² for 10 min, they were homogenized with tonicity restoration buffer (10 mM Tris-Cl [pH 7.6], 0.5 mM MgCl₂, 0.6 M NaCl, and protease inhibitors) and nuclear fractions were removed by low-speed centrifugation. The supernatant was centrifuged with 0.5 M EDTA at a final concentration of 5 mM for 45 min in a Beckman Ti 70-I rotor at 100,000 to 150,000 × g at 4°C. The membrane pellet was dissolved in Triton X-100 lysis buffer (300 mM NaCl, 50 mM Tris-Cl, 0.5% Triton X-100, and protease inhibitors) with repeated vortexing for 30 to 45 min. Insoluble materials were pelleted for 15 min at 10,000 × g, at 4°C, and the supernatant was saved.

Approximately 100 to 150 μg of whole membrane fraction was solubilized in sample buffer (nonreducing) for 5 min at 100°C and electrophoresed on 10% gels without lanes. The bands of gels were cut into horizontal strips by comparing the relative mobility of the protein sample to the distance traveled by a known molecular-weight-marker protein into different segments. The different fractions were named according to molecular weight as follows: fraction 1, 250,000 to 140,000; fraction 2, 140,000 to 70,000; fraction 3, 70,000 to 40,000; fraction 4, 40,000 to 18,000; and fraction 5, 18,000 to 0.

Electro-Eluter Model 422 (Bio-Rad) was used to elute the protein from the gel. BALB/c mice were immunized with different fractions (1 to 5) of whole-cell membrane antigen by using Freund's complete adjuvant for the first injection and Freund's incomplete adjuvant for the subsequent injections. The animals received 10 total injections over a 12-week period (6 to 10 μg of protein/injection). Serum was prepared and stored at −20°C. The immunoglobulin G (IgG) titer was checked by enzyme-linked immunosorbent assay before use of an attachment assay.

Mice that received either fraction 2 (140,000 to 70,000) or fraction 3 (70,000 to 40,000) died within 7 to 21 days postvaccination. The sera obtained from mice immunized with fraction 1 (250,000 to 140,000), fraction 4 (40,000 to 18,000), or fraction 5 (18,000 to 0) were evaluated for their ability to inhibit parasite attachment to a population of mid-S-phase cells. The asynchronized CHO cells were preincubated in antisera diluted 1:100 for 30 min prior to the addition of the tachyzoites.

Parasites (3 × 10⁶) were added to the culture and incubated for 30 min at 37°C. The coverslips were then washed to remove extracellular parasites by immersing them ten times in warm medium. Host cells' adherent and intracellular parasites were then fixed for 20 min with Bouin's fixative and were dehydrated and Giemsa stained. Infection was quantitated by counting the number of intracellular and attached parasites per 200 host cells on at least three replicate coverslips under a microscope.

In this assay, antisera raised against fractions 1, 4, and 5 inhibited tachyzoite attachment to both live and fixed asynchronous host cells. When compared with controls, the antisera raised against fractions 1 and 5 blocked parasite invasion by 12.5 and 10.5% in the fixed assay and 3.5 and 13.7% in the live assay, respectively. However, the inhibition observed for the cells preincubated with antisera against fraction 4 was significant in both cases, being 53.2 and 52.8% in the fixed and live assays, respectively (Fig. 1A and B).

FIG. 1.

Effects of polyclonal mouse antisera raised against whole-cell-fraction and fraction-1, -4, and -5 MDBK cells on live and fixed attachment assays. Host cells were incubated with medium only or preimmune (normal mouse serum [NMS]) or immune antisera for 30 min prior to the addition of the parasites. Results are expressed as the number of parasites attached to 200 host cells for three replicate cultures.

Rabbit polyclonal antibody against fraction 4 inhibits parasite attachment.

New Zealand White rabbits were immunized only with eluted fraction 4 by using Freund's complete adjuvant for the first injection and Freund's incomplete adjuvant for the subsequent injections. The animals received 10 total injections over a 12-week period (40 to 50 μg of protein/injection). They were terminally bled on the fifth day after the 10th immunization. High-titered anti-fraction 4 IgG (1:6,400) was tested for efficacy by enzyme-linked immunosorbent and immunofluorescence assays (IFA). To be certain that the inhibitory activity was specific to the antibody, the IgG for fraction 4 was purified by affinity chromatography on protein G coupled to Sepharose (Pharmacia, Piscataway, N.J.) and used at various concentrations in a blocking assay with asynchronous, fixed CHO cells (Fig. 2). The antisera against fraction 4 were then evaluated by using the attachment assay with glutaraldehyde-fixed asynchronous or synchronous CHO cells. Floating coverslips of nonconfluent, synchronous host cells were washed twice in medium and fixed in 2% glutaraldehyde (grade 1; Sigma) for 5 min at 4°C. The fixed cells were then washed three times in PBS and incubated overnight in 0.16 M ethanolamine (pH 8.3). The fixed cells were then incubated with minimum Eagle medium–0.2% bovine serum albumin containing either anti-fraction 4 antibodies diluted 1:100 or 25 μg of purified anti-fraction 4 IgG (determined from Fig. 2) per ml for 30 min. Then parasites (3 × 10⁶) were added to the culture and incubated for 30 min at 37°C. The coverslips were then washed to remove extracellular parasites by immersing them ten times in warm medium. Host cells' adherent and intracellular parasites were then fixed for 20 min with Bouin's fixative and were dehydrated and Giemsa stained. Infection was quantitated by counting the number of intracellular and attached parasites per 200 host cells on at least three replicate coverslips. When compared to the asynchronous control culture, antisera against fraction 4 blocked parasite attachment significantly (72% with anti-fraction 4 and 80% with column-purified anti-fraction 4 IgG) (Fig. 3). Anti-fraction 4 IgG also showed significant sequential inhibition of parasitic attachment in synchronized CHO cells when compared with that in the control synchronized culture (Fig. 4A and B).

FIG. 2.

Titration of rabbit anti-fraction 4. Shown is the effect of different concentrations of anti-fraction 4 IgG on the attachment of the RH strain of Toxoplasma tachyzoites to fixed, asynchronous CHO cells over a 30-min period. Host cells were preincubated in various concentrations of purified IgG for 30 min prior to the addition of parasites. Results are expressed as the number of parasites attached to 200 host cells.

FIG. 3.

Effect of anti-fraction 4 polyclonal antisera or anti-fraction 4 purified IgG on the attachment of the RH strain of Toxoplasma tachyzoites to asynchronous (live) CHO-pro cells over a 30-min period. Host cells were used untreated (control) or treated with preimmune (normal rabbit serum [NRS or NRIgG]) or immune anti-fraction 4 (F4) serum or purified IgG (F4IgG). Results are expressed as the number of parasites attached to 200 host cells for three replicate cultures.

FIG. 4.

CHO cells were synchronized with serum starvation and hydroxyurea block, resulting in a halt in cell cycle progression through the S phase at the G₁/S boundary. Synchrony was then monitored by fluorescence-activated cell sorter analysis; data are represented as the percentage of cells undergoing the S phase (A). (B) Shown are the kinetics of the attachment of T. gondii to synchronous CHO cells fixed at various time points in the S phase in the presence of polyclonal anti-fraction 4 IgG antibody or normal rabbit IgG antibody or medium alone. Host cells were preincubated with the antibodies (25 μg/ml) for 30 min prior to the addition of parasites. Results are expressed as the number of parasites (mean ± standard deviation) attached to 200 host cells (n = 3).

Indirect fluorescence antibody labeling.

Purified IgGs, both normal and anti-fraction 4, were labeled with fluorescein isothiocyanate (Fluorolink antibody labeling kit; Amersham Pharmacia Biotech). Synchronous and asynchronous CHO cells were fixed in 50% methanol and 50% acetone for 2 min. Cells were washed with PBS and blocked with 5% bovine serum albumin in PBS for 45 m. They were then incubated with either normal rabbit IgG or anti-fraction 4 IgG at a concentration of 10 μg/ml in PBS for 90 min at room temperature. Cells were washed in PBS 3 times and mounted in slow-fade light antifade buffer (Molecular Probes, Eugene, Oreg.). The cells were examined with a Zeiss Axiophot epifluorescence microscope (Fig. 5). It is clearly evident that these sera showed greater reactivity to the cells which were at 4 h postrelease than to the cells which were at either 0 or 8 h following release of CHO cells.

FIG. 5.

Localization of rabbit anti-fraction 4 antibodies on glutaraldehyde-fixed asynchronous and synchronous CHO cells by confocal microscopy. Cells are asynchronous CHO and normal rabbit serum (A), asynchronous CHO and anti-fraction 4 (B), synchronous 0 h and anti-fraction 4 (C), synchronous 2 h and anti-fraction 4 (D), synchronous 4 h and anti-fraction 4 (E), synchronous 8 h and anti-fraction 4 (F). Both antibodies were used at 10 μg/ml.

Western blotting with rabbit polyclonal anti-fraction 4 antibody.

Synchronized CHO cells were harvested at 4 h following release from hydroxyurea block, and prepared membrane fractions were separated by sodium dodecyl sulfate–10% polyacrylamide gel electrophoresis (SDS–10% PAGE). The proteins were transferred onto a nitrocellulose membrane. The membrane was incubated with primary antibodies (25 μg of rabbit polyclonal purified anti-fraction 4 IgG/ml) overnight at 4°C and for 1 h with secondary antibodies (alkaline phosphatase-conjugated anti-rabbit IgG). The membrane was developed with alkaline phosphatase-conjugated substrate (Bio-Rad) until color developed at room temperature.

A series of studies were performed in an attempt to identify the specific antigen(s) recognized by the anti-fraction 4 rabbit polyclonal IgG. The most valuable information arose from Western blot analysis, in which this antiserum was found to identify four to six bands with apparent molecular masses of 115, 100, 68, 30, and 25 kDa (Fig. 6). There were no obvious bands that provided a differential between the control and test antibody (data not shown). Further studies included biotin labeling of the 4-h-postrelease host cell surface antigens followed by immunoprecipitation (data not shown). None of these assays provided additional information regarding the specific character of the receptor for the parasite ligand.

FIG. 6.

Western blot analysis. Proteins from synchronized-CHO-cell (at 4 h) membrane were separated by mini-SDS–10% PAGE, transferred to nitrocellulose, and probed with rabbit anti-fraction 4 antibodies diluted 1:50. Molecular weights are indicated in thousands.

In this report, we have observed that synchronized host cells are most favorable to attachment when cells are at the peak of DNA synthesis (4 h postrelease). Attachment can be blocked by inhibitors of protein synthesis and partially affected by inhibition of glycolysis (unpublished data). This would suggest that the ubiquitous receptor for parasite attachment is a complex molecule or perhaps a number of molecules. The polyclonal antisera generated against the 4-h-postrelease cells inhibit attachment to both live and fixed cells. Antisera raised against MDBK cells can block parasite attachment to CHO cells, indicating that the receptor(s) is well conserved (8).

Several different approaches were utilized to identify the host cell receptor for parasite attachment. Silver staining of synchronized MDBK/CHO cell membrane prepared at different time points after hydroxyurea release and separated by SDS-PAGE failed to demonstrate a difference in protein expression (data not shown). A purified IgG fraction directed at proteins from fraction 4 (18,000 to 40,000) was reactive by IFA to host cells and demonstrated considerable interference with parasite attachment at 4 h postrelease. The IFA studies suggested that cells at 4 h postrelease were more reactive to the antibody than cells assayed at the time of release (0 h). We next attempted to identify expression of novel membrane antigens by Western blotting using the anti-fraction 4 sera. For 4-h membrane preparations, multiple bands were noted on Western blots. Although the mice were immunized with antigens from fraction 4, the antisera recognized a number of bands representing higher masses, including 115, 100, and 68 kDa. Unfortunately, there was no clearly demonstrable expression of unique antigens at the time point of interest.

Our findings further indicate that the receptor for parasite attachment is a complex molecule that has several structural properties. These data suggest that this receptor(s) has biochemical qualities consistent with being a protein as well as a glycosylated product. Since none of the approaches utilized, including protein or glycosylation inhibition, antibody blocking, or use of host cells devoid of proteoglycans, provided a distinct loss of the ability for the parasite to attach, it is possible that there is more than a single molecule that serves as the receptor for the parasite. Just as there is a family of parasite surface antigen-related products (the SRS family) (13), it is not inconceivable that there are different receptors expressed with different frequencies on a number of host cells. This could perhaps account for the ubiquity of the parasite in our environment and its ability to infect all mammalian cells. It further allows for successful attachment once the host has mounted an immune response against one surface ligand. This would be a survival method that would provide the parasite the opportunity to infect many different cell lines in the infected host.