Conformationally Correct Expression of Membrane-Anchored Toxoplasma gondii SAG1 in the Primitive Protozoan Giardia duodenalis

Abstract

To explore the possibility of expressing membrane-anchored exodomains of heterologous surface antigens in Giardia, a chimeric construct containing the Toxoplasma gondii SAG1 gene was made. The Giardia system is shown here to provide a means of generating correctly folded chimeric surface proteins in a native and unmodified form.

SAG1 is the immunodominant surface antigen of Toxoplasma gondii tachyzoites and has been shown to elicit a protective immune response against lethal challenge with parasites of the virulent RH strain in mice and other model animals. During previous development of SAG1 as a vaccine candidate against toxoplasmosis and a diagnostic reagent, recombinant forms of SAG1 were produced in a variety of bacterial or eukaryotic expression systems. Native SAG1 adopts a specific tertiary structure defined by intramolecular cysteine bridges that give rise to immunologically relevant conformational epitopes (4). Thus, production of correctly folded recombinant protein in bacterial expression systems has been inherently difficult. Similarly, illegitimate posttranslational modifications in the molecule were common when SAG1 was expressed in eukaryotic systems. The primitive eukaryote Giardia duodenalis (syn. G. lamblia and G. intestinalis) has an efficient protein export pathway to keep its surface provided with membrane-anchored variant-specific protein (VSP). Like SAGs and related proteins, Giardia VSPs contain many cysteine residues, which are likely to be important for attaining their functional conformation (1). Although initial studies have shown that VSPs are posttranslationally modified (7, 13), recent data suggest that at least in the case of VSP-H7, these protein-associated glycans may not be covalently bound (A. Hülsmeier and P. Köhler, unpublished data), indicating that posttranslational modification of surface antigens may be more sparingly used in Giardia than previously believed. In the present study, we wanted to test (i) whether an important T. gondii antigen, SAG1, can be expressed as both a correctly folded and unmodified membrane protein and (ii) whether this recombinant protein reacts with antibodies in human patient sera and could be used as a diagnostic reagent.

A chimeric gene containing targeting sequences for VSP surface expression fused to the SAG1 exodomain (SAG1-VSPct) was constructed for expression in Giardia under the control of a stage-specific inducible promoter. The chimeric SAG1-VSPct cassette was assembled as follows. The cyst wall protein 1 (CWP1) promoter region, including the transcription start site and a hydrophobic leader sequence, was amplified from genomic DNA from Giardia strain WBC6 (ATCC Nr 50803) (16) with primers CWP1-XbaI-s (sense; ATTCTAGACTAGCCACGCATGGGCTGT) and CWP1-nsiI-as (antisense; GTATGCATGACGAGCACCTCCCTCTGA). The SAG1 exodomain was amplified from the plasmid SAG1-GPI (15) with primers SAG1-nsi1-s (sense; GTatgcatCTGAGTAGCCGGGCTATGA) and SAG1-BglII-as (antisense; CATAGATCTAGCCCGGCAAACTCCAGT). The VSPH7 transmembrane domain plus the cytoplasmic pentapeptide was amplified from genomic DNA from Giardia strain H7 (12) (ATCC Nr 50581) with primers H7-BglII-s (sense; CATagatctAATAGCACCGGCGGCGATAGTG) and H7-PacI-as (antisense; GCTTAATTAATCACGCCTTCCCGCGGCAGACGAAC). The complete SAG1-VSPct gene was ligated into the green fluorescent protein (GFP) expression vector C1-GFP (6). The C-terminal portion of SAG1-VSPct consists of the hydrophobic VSP transmembrane anchor and a short cytoplasmic pentapeptide, CRGKA, replacing the original glycosylphosphatidyl inositol anchor signal of SAG1. Parasites of the WB strain were stably transformed with the SAG1-VSPct plasmid, and production of the chimeric protein was induced by encystation as described previously (6). Expression of SAG1-VSPct protein was determined by Western blot analysis of total lysate and indirect immunofluorescence of fixed but intact parasites. For detection of the SAG1 exodomain, a monoclonal antibody (MAb), DG52 (3), and human patient sera, which had been titrated against total membrane proteins of T. gondii tachyzoites of strain RH, were used (F. Grimm, unpublished results). For Western blots, total cell lysates (from 5 × 10⁵ cells/lane; 15 h postencystation) were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis under nonreducing conditions unless stated otherwise and transferred onto a nitrocellulose membrane. Western blotting of total protein from encysting parasites transformed with the SAG1-VSPct construct and probed with DG52 showed a single band of ∼28 kDa (Fig. 1). In contrast, total protein from uninduced trophozoites, the parental WBC6 strain, and encysting WB-SAG1-VSPct parasites separated in the presence of 100 mM dithiothreitol did not react with DG52. Reactivity of MAb DG52 with SAG1-VSPct could be increased >10-fold by treatment of cells with Triton X-100 (data not shown). T. gondii total protein probed with DG52 revealed, in addition to SAG1, several minor bands representing cross-reactions with other members of the SAG family.

FIG. 1.

Western analysis of WB-SAG1-VSPct transgene (Ct) or T. gondii lysate (Toxo). Wild-type Giardia (WB) or induced (ind) transgenes were probed under reducing (R) or nonreducing (NR) conditions. Lysates were probed either with human sera (patients 1 and 2) or with the anti-SAG1 MAb DG52. MAb DG52-reactive bands are designated by asterisks.

To test whether SAG1-VSPct would also be recognized by naturally occurring antibodies of human toxoplasmosis patients, we performed Western blot analysis with lysates from transformed Giardia. Sera from 12 toxoplasmosis patients with acute (high immunoglobulin M [IgM], low IgG; e.g., patient 1 in Fig. 1) or chronic (low IgM, high IgG; e.g., patient 2 in Fig. 1) infections and two negative control sera were used to examine the extent to which transgenic parasites were able to present native and conformation-dependent SAG1 epitopes in addition to the epitope recognized by MAb DG52. The sera reacted specifically with the nonreduced SAG1-VSPct product, but, interestingly, failed to recognize any linear SAG1 epitope in the reduced forms. As shown for two representative examples in Fig. 1, the serum from acute infection with high levels of anti-Toxoplasma IgM and low levels of IgG showed a weaker signal when detected with anti-IgG secondary antibodies. Although there was a possibility of interference with detection of anti-SAG1 antibodies, the polyclonal human sera investigated here did not evoke additional, strongly cross-reacting bands when incubated with lysate of transformed Giardia. In contrast, they reacted with several higher- and lower-molecular-weight proteins of T. gondii tachyzoites in addition to SAG1. Controls with sera from uninfected human subjects showed no reaction with either transgenic Giardia or T. gondii lysates.

To confirm that the recombinant SAG1-VSPct was not posttranslationally modified by glycosylation, specifically at the single Asn-X-Ser site in SAG1, we performed ECL (enhanced chemiluminescence) glycoprotein detection experiments (Amersham Pharmacia UK, Ltd.), as described previously (7), with lysates from induced and noninduced transgenic Giardia on Western blots. Endogenous VSP bands and a putative GPI-anchored protein (GP49) were detected, but no specific signal at the position of SAG1-VSPct could be seen in lanes with parasite lysates (Fig. 2).

FIG. 2.

Recombinant SAG1-VSPct is not glycosylated. Glycans in separated lysates of transformed Giardia trophozoites (T [uninduced]), encysting cells (E [induced]), and a control protein (C [transferrin]) were detected as previously described (7). The position of SAG1-VSPct from induced cells is shown in a separate lane by reaction with antibodies to SAG1 (S). The migration of the size markers is indicated (in kilodaltons).

By indirect immunofluorescence with MAb DG52 or the human patient sera described above, SAG1-VSPct was found associated with the plasma membrane and flagella (Fig. 3). In parasites induced to express SAG1-VSPct, the observed staining pattern was in accordance with the distribution of the endogenous VSP-H7 in Giardia strain H7 in immunoelectron microscopy studies (14) and immunofluorescence (A. B. Hehl, unpublished observation), as well as that of VSP TSA417 in strain WB (5). Surface exposition of SAG1-VSPct was not found on trophozoites. These results demonstrate that SAG1-VSPct expressed in Giardia retains its tight tertiary structure, exposing immunologically relevant conformational epitopes.

FIG. 3.

Immunofluorescence microscopy (A, B, D, and E) and corresponding differential interference contrast images (C and F) of induced WB-SAG1-VSPct transgenes. Cells in panels A and B were labeled with DG52, and those in panels D and E were labeled with human serum from patient 2. Single planes (A and D) and three-dimensional-reconstructed image stacks (B and E) are shown.

The variability between commercially used serological assays to detect T. gondii-specific Igs is largely due to varying conditions for tachyzoite cultivation (mouse or tissue culture) and a lack of standard methods for preparing tachyzoite antigen (2). In addition, the pattern of T. gondii antigen immunoreactivity with human sera varies with the Ig class and the stage of infection. The GPI-anchored SAG1 is highly abundant and the most immunogenic surface protein of the T. gondii tachyzoites present during the acute phase of the disease, but it is not expressed in bradyzoites during the chronic stage of infection (9). Therefore, the SAG1 antigen has been the focus of intensive research with the aim of developing a subunit vaccine against T. gondii infection or as a diagnostic tool. Expression of recombinant SAG1 in different prokaryotic systems, as well as in eukaryotic cells, such as insect cells (8), Pichia pastoris (11), and CHO cells (10), remained unsatisfactory due to improper folding and/or fortuitous glycosylation events. These deviations from the native SAG1 structure lead to a strong decrease in protective responses or the production of antibodies against natural epitopes in animals. In contrast, SAG1 expressed in Giardia retains its correct tertiary structure, and most importantly, SAG1-VSPct appears also not to be glycosylated. Simple axenic cultivation in a semidefined medium and well-established techniques for stable maintenance of transfected DNA make Giardia an extremely useful system for small- and medium-scale expression of membrane-bound recombinant surface antigens. Finally, the expression of heterologous surface antigens such as SAG1, or specific subdomains thereof, in a native conformation on the Giardia surface provides a powerful tool for the investigation of host-parasite interactions.