Simple and Efficient Method for Measuring Anti-Toxoplasma Immunoglobulin Antibodies in Human Sera Using Complement-Mediated Lysis of Transgenic Tachyzoites Expressing β-Galactosidase

Abstract

A simple and efficient method using transgenic Toxoplasma gondii tachyzoites expressing β-galactosidase was developed for detection of specific antibodies against the parasite in sera of patients. The titers obtained with the new test were similar to those obtained with the Sabin-Feldman dye test run in parallel. Although significant changes in endpoint titers were not observed when sera drawn sequentially at 2- to 3-week intervals were tested with both procedures, apparent differences in antibody affinity were observed with the new test which were not perceptible with the Sabin-Feldman dye test. Like the Sabin-Feldman dye test, the new test is based on complement lysis of tachyzoites, but it is much easier to perform and the reaction is read colorimetrically instead of visually.

Infection with Toxoplasma gondii is routinely diagnosed by demonstration of the presence of specific antibodies against the parasite in serum using a number of serologic assays, including the Sabin-Feldman dye test (9). The dye test is a complement-lysis-based assay and is the international “gold standard” for diagnosis of toxoplasmosis (8, 9). The dye test measures principally immunoglobulin G (IgG) antibodies and is both sensitive and specific. Since IgG antibodies persist in the dormant stage of the infection, detection of these antibodies in a single sample does not provide sufficient information regarding the timing of the initial infection or disease manifestation. A more accurate diagnosis could be made if sera were obtained at regular intervals. Sequential serum samples could then be tested in parallel to determine if the IgG titers have changed over time. A significant rise in the IgG titer is suggestive of an evolving recently acquired infection. Conversely, a significant decrease in the titer suggests that the infection is moving toward the chronic stage. The dye test is considered by many diagnostic laboratories to be more reliable than commercially available enzyme-linked immunosorbent assay (ELISA) kits for demonstration of IgG antibodies and can often show a change in titer between sequential samples. Unfortunately, it is time consuming and cumbersome because it requires that live parasites treated with each serum dilution be analyzed under the microscope. Consequently, the dye test is presently employed by relatively few diagnostic laboratories (8).

For the present study, we sought to improve and simplify the dye test by using tachyzoites of T. gondii in which the gene for the bacterial enzyme β-galactosidase (β-Gal) was introduced. This procedure allowed the development of a microtiter assay with the accuracy of a complement-based assay that can be read colorimetrically and avoid many of the pitfalls of the dye test.

MATERIALS AND METHODS

Construction of β-Gal plasmid.

The entire open reading frame of the Escherichia coli β-Gal gene was amplified from λgt11 using primers LACNSI (5′-GGGATGCATATTACGGATTCACTGG-3′) and LACPAC (5′-GGGTTAATT AATTATTTTTGACACCAGAC-3′) carrying flanking sequences for the restriction sites NsiI and PacI, respectively, and treated with both restriction enzymes. The DNA fragment corresponding to the chloramphenicol acetyltransferase open reading frame was excised from the T. gondii SAG1 promoter construct (10) with the restriction endonucleases NsiI and PacI (Promega Corp., Madison, Wisc.) and replaced with the β-Gal cassette. The resulting plasmid was designated SAG1/1 β-GAL.

Transfection.

Parasites were transfected using restriction enzyme-mediated integration as described previously (1). Briefly, 20 μg of SAG1/1 β-GAL DNA was linearized with the restriction endonuclease NotI (Promega Corp.) and phenol extracted to eliminate residual enzymatic activity. Following ethanol precipitation, the DNA was resuspended in cytomix buffer (10). Immediately prior to electroporation, 100 U of NotI was added to the cuvette containing the parasites and DNA. Following electroporation, parasites were inoculated into T25 flasks containing human foreskin fibroblast (HFF) cells and placed under 20 μM chloramphenicol selection. After three passages, the parasites were cloned by limiting dilution in 96-well microtiter plates containing HFF cells. Cloned, stable transformants expressing β-Gal were identified in 96-well cultures grown in Dulbecco's modified Eagle's medium lacking phenol red (Life Technologies, Rockville, Md.) but containing 100 μM chlorophenol red–β-d-galactopyranoside (CPRG) (Boehringer Mannheim, Indianapolis, Ind.) as previously described (3, 7).

Parasites.

Wild-type and transgenic T. gondii (RH strain) cells were maintained in HFF monolayers cultured in Dulbecco's modified Eagle's medium containing 10% fetal bovine serum (Hyclone, Logan, Utah), 25 mM HEPES, 50 U of penicillin ml⁻¹ and 50 μg of streptomycin ml⁻¹ incubated at 37°C in 5% CO₂. The organisms were also maintained by repeated passage in Swiss Webster mice by intraperitoneal injection with tachyzoites. For intraperitoneal passage of the β-Gal transgenic parasites, organisms were suspended in phosphate-buffered saline (PBS) containing fresh 100 μM chloramphenicol prior to injection. Peritoneal fluids containing tachyzoites were collected at 3 days postinfection.

Serum samples.

Serum samples were provided by the Toxoplasma Serology Laboratory of the Research Institute, Palo Alto Medical Foundation, and by P. Thulliez from the Pasteur Institute in Paris, France. All samples were examined in the Sabin-Feldman dye test, the IgM-ELISA, the IgA-ELISA, and differential agglutination (AC/HS) test (2, 5, 6, 9, 11, 12). These tests comprise the toxoplasma serologic profile (TSP) (5), and their results in combination with the individual's clinical history were used to classify the samples into three groups. Group 1 sera were from 8 individuals not infected with T. gondii (seronegative), group 2 sera were from 11 individuals with a TSP suggestive of a recently acquired infection (acute TSP) (5), and group 3 sera were from 11 individuals with a TSP suggestive of an infection acquired in the distant past (chronic TSP) (5). An additional group, group 4, was comprised of samples from four pregnant women who had shown seroconversion during pregnancy. Sera from these women were collected periodically for as long as 1 year after seroconversion. Thirteen serum samples, eight from group 1, two from group 2, two from group 3, and one from group 4, were used to standardize the assay with transgenic tachyzoites and as controls.

Complement-mediated colorimetric microtiter test (β-Gal assay).

Twenty-five microliters of serum from patients was diluted 1:4 with Hanks balanced salt solution (HBSS; Life Technologies, Rockville, Md.) and heated at 56°C for 30 min to inactivate preexisting complement. Twofold dilutions of the serum were then made in 96-well microtiter plates. Transgenic RH tachyzoites (β-Gal clone) were harvested from the peritoneal cavities of mice and pelleted by centrifugation at 120 × g, washed twice with PBS containing 1% heparin, passed four times through a 22-gauge needle to disperse parasite aggregates, and resuspended at a concentration of 5 × 10⁷ parasites per ml in PBS. An equal volume of the parasite suspension was mixed with an equal volume of “accessory factor.” For the present study, accessory factor was human serum from a single, seronegative donor that provided the source of complement for both the dye test and the current β-Gal assay. Fifty microliters of the tachyzoite-accessory factor mixture was added to each well containing 75 μl of each serum dilution. One set of control wells contained patient serum to be diluted 1:16 without addition of the tachyzoite-accessory factor mixture. A second set of control wells contained only the tachyzoite-accessory factor mixture without addition of patient serum. One hundred ten microliters of each mixture was transferred to 0.2-ml thin-walled PCR tubes (Perkin-Elmer, Emeryville, Calif.) and incubated at 37°C for 15 min. Thereafter, intact parasites were pelleted by centrifugation at 1,000 × g for 5 min and 30 μl of the supernatant was transferred to a microtiter plate. β-Gal activity was assayed in duplicate on 10 μl of each supernatant using CPRG as the colorimetric indicator (3, 7). The absorbance at 570 nm was read on a Dynatech MR5000 microplate reader (Dynatech, Chantilly, Va.). An additional total lysis control was included by mixing 100 μl of tachyzoite-complement mixture with 110 μl of HBSS buffer and 40 μl of 5× reporter lysis buffer (Promega Corp.) followed by incubation at 50°C for 10 min. Although all specimens had been previously tested in the Sabin-Feldman dye test (9) in the Toxoplasma Serology Laboratory, serum samples analyzed by the β-Gal assay were also tested in parallel using the standard dye test.

RESULTS

Transgenic parasites.

Of the 20 transgenic clones positive for β-Gal expression in the 96-well cultures, 2 clones had apparently higher levels of β-Gal expression than the other 18. During passage and propagation in cell culture these two clones converted the CPRG substrate in the culture medium to red much sooner than the other clones. These two clones were expanded, and their genomic arrangements were analyzed by restriction digest and Southern blot. The Southern blot results suggested that a single copy of the β-Gal transgene had integrated in the genome of both clones (data not shown). In addition, the Southern blot patterns were identical, suggesting that these clones were siblings.

Endpoint titration.

Transgenic parasites were incubated with serum dilutions plus complement in an assay that was analogous to the conventional dye test. However, instead of direct microscopic analysis of the treated parasites as in the dye test, the supernatant of each reaction was assayed for the release of β-Gal from the parasites. When serum samples from patients that were previously shown to possess IgG antibodies against T. gondii by the conventional dye test were tested in the β-Gal assay, a dilution was reached at which the patient serum no longer stimulated a significant release of β-Gal above the nonspecific release observed with the saline control (Fig. 1). Although it is an approximation, this dilution is similar to the endpoint in the dye test, which is the first dilution at which less than 50% of the parasites appear unstained. Furthermore, the endpoint dilution in the β-Gal assay was within fourfold of the endpoint obtained in a parallel dye test titration (Table 1) (for example, patient 1 had a dye test endpoint dilution of 1:256 and a β-Gal endpoint dilution of 1:256, patient 2 had a dye test endpoint dilution of 1:512 and a β-Gal endpoint dilution of 1:256, while patient 3 had a dye test endpoint dilution of 1:4,096 and a β-Gal endpoint dilution of 1:2,048 [Fig. 1]) A similar correlation was found between the β-Gal assay endpoint dilution and the dye test endpoint dilution in each of the additional 20 positive sera tested (data not shown). In the β-Gal assay, a relationship between the serum concentration and the amount of β-Gal released from the parasites was observed at numerous dilutions preceding the endpoints. At the more concentrated serum dilutions, more β-Gal was released. Although a relationship between the serum concentration and the percentage of unstained versus stained tachyzoites was observed in the dye test, this relationship was easily quantified only at two or three dilutions immediately preceding the endpoint. At the more concentrated dilutions, 100% of the tachyzoites appeared to be unstained, making it difficult to discern differences between adjacent dilutions.

FIG. 1.

Endpoint titrations of serum samples from three individuals with serologic profiles consistent with chronic (patients 1 and 2) or acute (patient 3) T. gondii infection. Patients 1 and 2 were classified as having chronic TSP profiles (6) with dye test titers of 1:256 and 1:512, respectively, negative IgM ELISA (5) titers (1.3 and 0.9, respectively) and equivocal patterns in the differential agglutination AC/HS (2) test (50/100 and 50/400, respectively). Patient 3 had an acute TSP profile (5) and a dye test titer of 1:4,096, positive IgM titer of 4.1, positive IgA ELISA (11) titer of 2.1, and an acute AC/HS pattern (400/3,200). Absorbance readings at 570 nm are shown. Sera from patients 1 and 2 were not tested at the 1:2,048 or 1:4,096 dilutions.

TABLE 1.

Comparison of β-Gal assay with the Sabin-Feldman dye test

Patient no.	Specimen datea	Endpoint dilution in:		TSP
		Dye test	β-Gal assay
1	12/12/97	1:256	1:256	Chronic
2	11/05/97	1:512	1:256	Chronic
3	02/06/97	1:4,096	1:2,048	Acute
4	09/17/97	1:256	1:256	Chronic
	10/07/97	1:64	1:256	Chronic
5	12/18/96	1:512	1:2,048	Acute
	01/02/97	1:1,024	1:2,048	Acute

^aDate sample was taken (month/day/year). 

When parasites were incubated with sera from eight patients that tested negative for T. gondii IgG antibodies in the dye test, no significant release of β-Gal was detected even at a 1:4 dilution of the sera (data not shown).

Sequential serum samples.

In our initial development of the assay, we observed that sera from various patients that had similar endpoint dilutions in the β-Gal assay released very different amounts of β-Gal at the more concentrated dilutions (data not shown), suggesting that the relationship between serum dilution and β-Gal release might be exploited to discern changes in IgG titers or quality from sequentially drawn sera which did not have obvious differences in dye test titers. We tested sera drawn sequentially from several seroconverters whose dye test titers did not change significantly over a 2- to 3-week period (Fig. 2). Based on the absence of IgM antibodies and the lack of a substantial change in the dye test titers (1:256 on 17 September 1997 and 1:64 on 7 October 1997), patient 4 was diagnosed as having an infection acquired in the more distant past. Using the β-Gal assay, similar endpoints were found (1:512) for both sera from patient 4 and no significant difference was observed in the pattern of the readings over the dilution series (Fig. 2A). However, with sera from patient 5 a very different pattern was observed (Fig. 2B), in which the later serum sample (2 January 1997) had significantly higher readings at the more concentrated dilutions than the sample obtained 2 weeks earlier (18 December 1996), although the endpoints for the two sera were the same (1:2,048) in the β-Gal assay and were within twofold of each other when run in parallel in the dye test. Although patient 5 showed only a small increase in dye test titers over the 2-week interval between the first and second samples (1:512 to 1:2,048), this patient was diagnosed as having a recently acquired infection based on the short interval between a previous dye-test-negative serum sample (19 August 1996) and the first IgG-positive serum (18 December 1996). Similar patterns, as in Fig. 2B, were observed in this study with four other sequentially drawn serum samples from patients with recently acquired infection (data not shown). When sequential samples with similar endpoints (in the dye test and β-Gal test) were compared, significant differences in β-Gal readings were found at parallel dilutions. This phenomenon was apparently restricted to samples obtained during recently acquired infections, since, as observed with patient 4 (Fig. 2A), sequential serum samples from an additional four chronically infected patients did not display differences in β-Gal readings at parallel dilutions, although these sera had similar endpoints by both the dye test and the β-Gal assay (data not shown).

FIG. 2.

Parallel endpoint titrations of sequential serum samples obtained over 2- to 3-week periods from individuals with serologic profiles consistent with chronic (A) or acute (B) T. gondii infection. From patient 4 (A), two serum samples were obtained 1 year after seroconversion, with dye test titers of 1:256 (17 September 1997) and 1:64 (7 October 1997) and negative IgM titers (0.7 and 0.8, respectively). Patient 5 (B) seroconverted between a dye-test-negative sample (19 August 1996) and the first dye-test-positive sample (18 December 1996). Two sequential samples from patient 5 had dye test titers of 1:512 (18 December 1996) and 1:1,024 (2 January 1997) and IgM-ISAGA titers of 12. Absorbance readings at 570 nm are shown.

DISCUSSION

Titers of anti- T. gondii IgG in the dye test are defined from the endpoint dilution, which is the greatest serum dilution where the complement-mediated lysis of tachyzoites returns to the baseline lysis that occurs from the complement accessory factor alone. In the present β-Gal assay as well, endpoint dilutions were used to estimate T. gondii IgG titers. However, the β-Gal assay revealed differences in lysis intensity between parallel dilutions of serum specimens that had similar endpoints in both the β-Gal assay and the dye test. At the 1:64 dilutions, serum from patient 2 had significantly higher readings than serum from patient 1 (Fig. 1), but both had endpoints of 1:256. Similarly, differences were observed between sequential specimens from individuals that had shown seroconversion in the recent past. For patient 5, the amount of β-Gal released from tachyzoites during complement fixation (reflecting the intensity of lysis) increased during a 2-week period despite the lack of change in the endpoint titers during this period. These differences in intensity may reflect differences in the avidity of the IgG antibodies as they mature. The maturing antibodies from the later phase of the infection are known to have a higher avidity (4) and may provide a more efficient substrate for the complement system. The rate of complement lysis is dependent on the binding affinities of all components in the reaction, a characteristic that is similar to the kinetic properties of enzymes, where the rate of product production is dependent on the affinity of the enzyme for the substrate. The complement cascade is initiated by the binding of the antibodies to the parasite, and the initial rate of lysis is dependent on the affinities of these antibodies. Although the absolute quantity of antibodies in two serum samples may be the same, resulting in similar titration endpoints, the higher-avidity antibodies in combination with complement will increase the complement lesions in the tachyzoites, resulting in an increased release of the 116-kDa β-Gal protein into the medium. The conventional dye test makes use of the same lesions but employs a small vital dye (methylene blue) to monitor creation of these lesions, while in the β-Gal assay lesions are monitored as the release of a very large macromolecule. The theory behind staining in the conventional dye test is that complement-lysed cells are unable to retain the dye and are not stained, while live cells retain the dye and are stained. Small complement lesions or small numbers of lesions in the tachyzoites may be sufficient to prevent accumulation of the dye. Therefore, at the more concentrated serum antibody dilutions, all tachyzoites are unstained regardless of the number or size of the lesions. However, in the present assay it is likely that as the number of lesions per parasite or the size of the lesions increases, more of the large β-Gal molecules are released.

In much the same way as differences in ELISA avidities are currently being used (4), the differences in β-Gal lysis intensity might be exploited as a way to follow evolving infections and to estimate the timing of the onset of T. gondii infection in pregnancy, where timing is critical to determination of the risks of congenital transmission. However, the nature of these differences in β-Gal lysis would need to be explored in more detail. Furthermore, numerous sequential sera would need to be tested to establish a correlation between lysis intensity and the timing of infection.

For the purposes of this study, the endpoint values were approximated as the dilution point where the assay readings were similar to the background release of β-Gal observed in a parallel saline control. However, there were day-to-day and batch-to-batch variations in the background release of β-Gal that were most likely due to differences in the viability of the parasites in each batch. These differences will have an effect on the endpoint titers, especially if the background release is high. Indeed, we observed that the same sera would give endpoint titers that varied by twofold when tested on different days (data not shown). However, this twofold variance is also observed in the dye test and is the reason sequential specimens are tested in parallel on the same day to more accurately determine changes in titers. Although this is a preliminary study with only a small number of specimens tested and very few modifications were made to the method to improve standardization, the results are promising and warrant further exploration of this method as a routine assay. Toward that end, extensive validation and standardization will be needed. However, as an easy method for accurately measuring IgG titers, the β-Gal assay may be preferable to the conventional dye test.