JOURNAL OF CLINICAL MICROBIOLOGY, Sept. 1987. p. 1597-1600 Vol. 25, No.

9
0095-1137/87/091597-04$02.00/0
Copyright C 1987, American Society for Microbiology

Comparative Study of Tissue Culture and Mouse Inoculation

Methods for Demonstration of Toxoplasma gondii
FRANCIS DPEROUIN,1* MARIE CHRISTINE MAZERON,2 AND YVES J. F. GARIN'
Laboratoires de Parasitologie-Mycologie' et de Bactériologie-Virologie,2 Hôpital Saint-Louis,
75475 Paris Cedex 10, France
Received 3 February 1987/Accepted 22 May 1987

Two methods for the isolation of Toxoplasma gondii were analyzed and compared. Bradyzoites or tachyzoites
of three strains of T. gondii were injected into mice and introduced in parallel onto MRC5 fibroblasts cultured

Downloaded from http://jcm.asm.org/ on March 20, 2019 by guest

on cover slips. In the cultures, the parasites were more readily identified by an indirect immunofluorescence
assay than by examination of unstained or Giemsa-stained cultures. With the RH strain, the tachyzoites
replicated actively, and large foci of parasites were observed in 24 h. The bradyzoites or tachyzoites of the other
strains could also be cultivated, but grew rather slowly; 2 days after inoculation, early stages of multiplication
could be observed: from day +4, Toxoplasma clusters or foci were easily identified at a 100 magnification.
X

The course of infection in mice was greatly dependent on the virulence of the strain and on the parasitic stage
inoculated. In the chronically infected mice, evidence of Toxoplasma infection was only detected 45 days after
inoculation through the demonstration of cysts in the brain or the presence of specific antibodies in the serum.
The mean ratio of infected mice and positive cultures was compared in relation to the inoculum size. The tissue
culture method was found to be at least as sensitive as mouse inoculation. Since Toxoplasma organisms may be
isolated within a few days in tissue culture, it is proposed that this method should be used when early isolation
of the parasite is crucial for the diagnosis of toxoplasmosis.

Toxoplasma gondii is usually isolated from susceptible maintained through passage at 6-month intervals by intra-
laboratory animals. Mice are most commonly used, since peritoneal inoculation of brain cysts. To obtain tachyzoites,
they can be readily infected by intraperitoneal injections of cysts from chronically infected mice were inoculated intra-
trophozoites or bradyzoites. Depending on the virulence of peritoneally into mice treated with cortisone acetate (0.2
the strain, mice develop either an acute infection with mg/day subcutaneously from the day of inoculation); perito-
parasite-rich ascites or a chronic infection characterized by neal exudate containing tachyzoites was harvested 5 to 7
the presence of cysts in the brain. days later.
Inoculation of blood, body fluids, or tissue extracts into For the collection of tachyzoites, the peritoneal cavity of
mice may be used for diagnosis (1, 12), but the result is often infected mice was washed with 5 ml of sterile phosphate
delayed because the encysted parasites may not be identified buffer solution (PBS; 0.01 M, pH 7.2). Contaminating cells
before 30 days after inoculation. were eliminated by gel filtration through Trisacryl GFO5
Several reports have demonstrated that tissue culture (IBF-France); 10 ml of gel was packed into a 20-ml syringe
methods could be applied to the rapid isolation of Toxo- barrel and washed with 40 ml of PBS. An equal volume of
plasma organisms from blood (7, 17) or infected tissues (2, peritoneal wash collected from infected mice was allowed to
5), and could serve for diagnosis when serological tests are pass through the column; the filtrate contained tachyzoites
inconclusive. However, the sensitivity of these methods has with less than 0.01% contaminating peritoneal cells.
not been evaluated. This was the main objective of this Bradyzoites (C and H strains) were obtained from chron-
study, in which three strains of T. gondii were cultivated on ically infected mice. Cysts were isolated from brain tissues
human fibroblasts (MRC5 strain) and inoculated into mice. on a Percoll gradient (Pharmacia, Sweden) (3) and then
disrupted by trypsin digestion (trypsin 1/250 [Difco Labora-
MATERIALS AND METHODS tories]; 0.5% in PBS, 5 min at 37°C). Bradyzoites were
centrifuged and washed with PBS.
Mice. Adult male and female Swiss albino mice were used. Viable tachyzoites and bradyzoites (brightly refringent by
All were negative for anti-Toxoplasma antibodies in the phase-contrast microscopy) were suspended in minimum
direct agglutination test (Bio-Mérieux, France). essential medium (MEM) (Flobio, France) supplemented
Parasites. Three strains of parasites were studied. The RH with 10% fetal calf serum (Flow Laboratories, France) and
virulent strain was maintained in mice by syringe passages of enumerated in a hemacytometer. They were inoculated into
peritoneal fluid from infected mice at 3-day intervals. The mice or cultures within 2 h after collection.
two other strains were of human origin; one (C strain) was Toxoplasma antigen. Toxoplasma antigen was prepared
obtained from a congenitally infected placenta (5), and the from T. gondii tachyzoites of the RH strain. Purified para-
other (H strain) was isolated from the blood of a bone sites were lysed with distilled water and then disrupted by
marrow transplant patient with disseminated toxoplasmosis. six successive cycles of freezing and thawing. After centrif-
These strains were of low virulence in mice and could be ugation at 10,000 x g for 1 h, the supernatant fluid was
collected and used as the antigen for immunization of
*
Corresponding author. rabbits. The protein content was 0.5 mg/ml (13).
1597
1598 DEROUIN ET AL. J. CLIN. MICROBIOL.

Toxoplasma antibodies. Antibodies to T. gondii were ob- presence of brain cysts was determined by examination of
tained by immunization of New Zealand White rabbits. six brain fragments (one frontal and two parietal fragments
Initial immunization was made with 500 Fxg of antigen mixed of each hemisphere).
with 0.5 mg of muramyl dipeptide (Choay Chimie, France) Statistical analysis. The effects of method mousee inocula-
and 1 ml of incomplete Freund adjuvant (Difco); 200 1dl was tion versus culture), strain, and inoculum dose were evalu-
injected intradermally at five sites. An identical weekly ated in a maximum likelihood logistic regression analysis of
booster was administered for 2 months until a serum agglu- the number of positive mice per the total number of inocu-
tination titer of 1:4,096 was obtained. The rabbits were lated mice or cultures (4).
exsanguinated, and immunoglobulin G (IgG) was purified by Correlations between the number of positive brain smears
ion-exchange chromatography (DEAE Trisacryl; IBF, or cultures and the number of parasites inoculated were
France) (16). studied by the nonparametric Spearman test (R) (8).
Cell cultures. Human embryonic fibroblast cell line MRC5
(Bio-Mérieux, France) was maintained in MEM containing RESULTS
10% decomplemented fetal calf serum, kanamycin (200
ptg/ml), and ampicillin (100 ,ug/ml). One milliliter of a sus- Development of toxoplasmas in culture. Tachyzoites and

Downloaded from http://jcm.asm.org/ on March 20, 2019 by guest

pension containing 105 fibroblasts was seeded onto 12-mm- bradyzoites from all strains could be grown in the fibroblast
diameter round cover slips (GIBCO Laboratories) placed cultures. Tachyzoites of the RH strain replicated in the cells
into each well of 24-well plates (Nunc, Sweden). Cultures within 6 to 9 h. By 12 h, parasites often formed rosettes or
were incubated at 37°C in a moist 5% C02-95% air atmo- dispersed into the cytoplasm of the host cells. No limiting
sphere and used within 2 weeks after preparation. membrane was observed by IF or Giemsa staining. At 48 h,
Identification of T. gondii in cell cultures. Three methods most of the host cells were infected and some of them
were used to identify T. gondii: (i) direct examination with detached from the cover slips; at 96 h, monolayers were
an inverted microscope (x250); (ii) examination of Giemsa- destroyed and large numbers of tachyzoites were found in
stained cultures (x400); and (iii) indirect immunofluores- the supernate.
cence (IF) (15). Cover slip cultures were rinsed with PBS, Tachyzoites and bradyzoites of the H and C strains grew
fixed with cold acetone (15 min at -20°C), and then incu- rather slowly; no marked difference was noticed in develop-
bated in the wells for 30 min at 37°C with antitoxoplasma ment in culture between the tWo strains. The early stages of
antibodies diluted at 1/200 in PBS. After two washes with multiplication were observed during days 1 and 2. IF staining
PBS, a fluorescent anti-rabbit IgG (Institut Pasteur, France) caused the entire cytoplasm of some infected cells to fluo-
diluted at 1/100 in PBS containing 1/25,000 Evans blue was resce. At days 4 and 8, the parasites formed compact
added. After 30 min at 37°C, the cover slips were washed and clusters limited by a thin fluorescent membrane. On day 10,
mounted onto slides for examination with an Olympus BH2 some pseudocysts (clusters) had burst, and large foci of
microscope with a reflected-light fluorescent illuminator newly infected cells were seen.
(light source, HBO 100 W; excitation filter, 490 nm; barrier Comparison of methods for identification of toxoplasmas.
filter, 515 nm). By direct examination (x 250), a cytopathic effect was de-
A comparative evaluation of the three methods was made tected after 24 to 48 h in culture with the RH strain and on
with cultures inoculated with 103 tachyzoites or bradyzoites day 4 with the other strains. It presented as foci of fibroblasts
from the different strains. In the cultures inoculated with the with a granular appearance, but Toxoplasma organisms
virulent RH strain, the medium was changed after an initial could not be identified. Thereafter, pseudocysts were ob-
contact of 3 h with the parasites, and then two cover slips served as enlarged fibroblasts containing refringent inclu-
were examined; other cultures were maintained at 37°C and sions. Large typical cysts were rare. Usually, cytopathic
examined at 6, 9, and 12 h and 1, 2, 4, and 8 days after effects could not be differentiated from nonspecific degener-
inoculation (two cover slips). In the cultures inoculated with ative changes of the monolayers.
the low-virulence C and H strains, the medium was changed Giemsa staining of the cultures allowed a better analysis of
at day + 1 in all the wells, and two cover slips were the structure of both parasites and hosts cells; however,
examined. Other examinations were done after 2, 4, 8, and scarce parasitized cells were difficult to detect in the stained
10 days. monolayers.
Comparison of culture and mouse inoculation. Suspensions By IF, parasites could be identified at the early stages of
of tachyzoites of the RH, C, and H strains and bradyzoites of multiplication. At a magnification of x 100, the infected cells
the C and H strains containing 1, 3, 10, 30,-and 100 parasites could be detected, and Toxoplasma organisms were un-
were prepared. Each suspension was inoculated simulta- equivocally identified when examined at x400. Since this
neously into four cell culture wells (1 ml per well) and into method was found to be more sensitive for identification of
each of two mice (1 ml, intraperitoneally). Four control mice Toxoplasma organisms in cultures, it was used for the
were injected with MEM alone. comparative study of the sensitivity of tissue culture versus
All assays were repeated in a separate experiment (in one mouse inoculation.
of the two assays with RH tachyzoites, only one mouse was Toxoplasma infection in mice. Mice injected with tach-
used for each inoculum dose). For each suspension, one of yzoites of the RH strain died of acute infection within 9 days
the four cover slip cultures was examined by IF as already and presented parasite-rich ascites. With the low-virulence
described on days 2, 4, 8, and 10. H strain, the course of infection was dependent on the
Inoculated mice were followed up for 45 days. Any animal parasite stage inoculated: mice infected with tachyzoites
dying during this period was examined for the presence of died of acute infection, but very few parasites were found in
toxoplasmas in peritoneal exudate or in the brain; serological the peritoneal exudate at the time of death; mice infected
tests were not performed. Surviving mice were sacrificed at with bradyzoites developed chronic infection. With the
day 45 and tested serologically by the agglutination test avirulent C strain, mice developed chronic infection whether
(Bio-Mérieux) and the IF antibody test with fluorescent infected with tachyzoites or bradyzoites. A total of 29 mice
anti-mouse immunoglobulin (Institut Pasteur, France). The were chronically infected (18 with the C strain and 11 with
VOL. 25, 1987 COMPARISON OF INOCULATION METHODS FOR T. GONDII 1599

the H strain). In 28 of 29 mice, cysts were found in at least TABLE 1. Sensitivity of mouse inoculation and tissue culture to
one brain smear; for one mouse, six brain smears were inoculation with tachyzoites and bradyzoites
negative but blind passage was positive. Brain cysts were
found either grouped or dispersed and ranged in size from 15 No. of positive No. of posi- Organisms
to 100 ,um in diameter. The number of positive smears was Parasite No.
inocu- cultures/no.
inoculated tiveb mice/no.
inoculated found
mice)(no.in: of
found to correlate poorly with inoculum size (R = 0.38, P < stage lated (mean positive (mean positive Brain
0.05) (Fig. 1). ratio)` ratio) Ascites cysts
The immunofluorescent antibody test was strongly posi-
tive (titers of >1:500) for the 29 chronically infected mice; Tachyzoites 1 1/24 (0.04) 0/10 (0) 0 0
3 4/24 (0.17) 2/11 (0.18) 1 1
the agglutination test was positive for 28 of 29. For control 10 11/24 (0.46) 4/11 (0.36) 3 1
mice, brain smears and serological tests were negative. 30 18/24 (0.75) 10/11 (0.91) 7 3'
Comparison of sensitivity of the methods. Cultures were 100 22/24 (0.92) 8/11 (0.73) 7 1
examined after 2, 4, 8, and 10 days; for each inoculum size,
results included examination of six cover slip cultures inoc- Bradyzoites 1 0/16 (0) 0/8 (0)
ulated with tachyzoites (three strains in duplicate wells) and 3 5/16 (0.31) 3/8 (0.37)
10 12/16 (0.75) 6/8 (0.75)

Downloaded from http://jcm.asm.org/ on March 20, 2019 by guest

four cultures inoculated with bradyzoites (two strains in
duplicate wells). The ratio of positive to inoculated cultures 30 14/16 (0.87) 6/8 (0.75)
was calculated for each inoculum. No significant difference 100 16/16 (1) 8/8 (1)
was found when the cultures were examined at each 2-day Mean positive ratio calculated from eight determinations (on days 2. 4, 8,
interval (P = 0.27); therefore, mean positive values were and 10. duplicate wells) for each inoculum level of three strains (tachyzoites)
calculated from the four determinations (24 cultures with or two strains (bradyzoites).
h Mice were considered positive on the basis of a positive serological test or
bradyzoites and 16 cultures with tachyzoites for each by the presence of parasites in the brain or ascites.
inoculum size) (Table 1). "
One found positive after blind passage.
One culture of the 24 inoculated with one tachyzoite was
positive; however, a significant number of positive cultures
were only observed for an inoculum containing more than the two methods according to strain or parasite stage inoc-
three bradyzoites or tachyzoites (P < 0.01). We noticed that ulated (P = 0.11).
with the low-virulence strains, the host cells examined on
day 2 contained only a few parasites; examination of the DISCUSSION
cultures from day 4 on was easier because parasites formed
clusters which could be identified at low magnification Animal inoculation is usually considered the most sensi-
(x 100). tive method for the isolation of T. gondii from tissues or
Mice developed either acute or chronic infections; thus, body fluids (1, 11). However, several studies have demon-
Toxoplasma infection was assessed by a positive serological strated that in mice, susceptibility to Toxoplasma and course
test or by the presence of parasites in brain biopsies or of infection may be affected by several factors, such as the
peritoneal exudates. Following these criteria, positive ratios route of infection and the infecting dose (6, 11, 12). In
(i.e., number of mice infected/number inoculated) were addition, the mouse virulence of the parasite may influence
calculated for mice inoculated with tachyzoites or brady- the ability to isolate Toxoplasma, since some strains infect
zoites (Table 1). mice poorly (10, 14).
The minimum infecting dose was three tachyzoites or The reliability of the mouse model requires regular fol-
bradyzoites; inoculation of 10, 30, and 100 parasites resulted low-up studies for the presence of Toxoplasma and serolog-
in an increase in the infection rate (Fig. 2). ical tests. In acutely infected mice, tachyzoites can be found
Ratios of positive mice and cultures were both signifi- in the peritoneal exsudate, but may be rare even at the time
cantly related to the inoculum dose (P < 0.03) (Fig. 2). No of death in mice infected with the H strain. In most in-
significant difference between the two methods was found (P stances, mice develop a chronic infection that requires 30 to
= 0.96), and no difference was observed in the sensitivity of 45 days. The detection of anti-Toxoplasma antibodies by IF
is a good indicator of infection because all the chronically
infected mice became seropositive while the controls re-
POSITIVE
SMEARS
6 POSITIVE

5 1o0 --- - - -_
--

4- >

3.

2. 50 1

1i

i 3 10 3Ô 160

PARASITES
INOCULATED à3 10
30 100 PARASITES
INOCULATED
FIG. 1. Relationship between the number of positive brain
smears (six examined) and the inoculum size (tachyzoites or FIG. 2. Relationship between the mean ratio (±+ 1 standard error)
bradyzoites) in 29 chronically infected mice. Symbols: M. C strain of positive culture (O) and mice (-) and the number of bradyzoites
(18 mice); A, H strain (11 mice). or tachyzoites inoculated.
1600 DEROUIN ET AL. J. CLIN. MICROBIOL.

mained negative. However, the demonstration of cysts in the 3. Cornelissen, A. W. C. A., J. P. Overdulve, and J. M. Hoen-
brain provides the best evidence of infection; at least three derboom. 1981. Separation of Isospora (Toxoplastna) gondii and
brain smears should be examined since the cysts may be rare cystozoites from mouse brain tissue by continuous density-
and their number poorly correlated to the inoculum size. gradient centrifugation. Parasitology 83:103-108.
Finally, the methods used for the demonstration of Toxo- 4. Cox, D. R. 1970. Analysis of binary data, p. 152. Chapman &
plasma infection in mice can only be applied for routine Hall, London.
5. Derouin, F., M. C. Mazeron, and Y. J. F. Garin. 1986.
diagnosis in specialized laboratories; the results are often Toxoplasmose congénitale. Diagnostic rapide par mise en évi-
delayed since mice usually develop a chronic rather than an dence de toxoplasmes dans le placenta par culture cellulaire.
acute infection. In contrast, tissue culture methods which Presse Méd. 15:1684.
are commonly used for isolation of viral pathogens are more 6. Eyles, D. E., and N. Coleman. 1956. Relationship of size of
readily available for the cultivation of T. gondii (9). inoculum to time to death in mice infected with Toxoplasma
Our results show that tachyzoites or bradyzoites from gondii. J. Parasitol. 42:272-275.
different strains of T. gondii can be grown in MRC5 fibro- 7. Hofflin, J. M., and J. S. Remington. 1985. Tissue culture
blasts cultures. With an IF technique, Toxoplasmia organ- isolation of To.vopilasma from blood of a patient with AIDS.
Arch. intern. Med. 145:925-926.
isms were easily identified in fibroblasts within 2 days after 8. Hollander, M., and D. A. Wolfe. 1973. Nonparametric statistical
inoculation; however, later examination (after 4 days) is

Downloaded from http://jcm.asm.org/ on March 20, 2019 by guest

methods, p. 503. John Wiley and Sons, Inc., New York.
preferable for easy identification of the parasites in the 9. Hughes, H. P. A., L. Hudson, and D. G. Fleck. 1986. In vitro
cultures since large foci or clusters of toxoplasmas can be culture of Toxoplasina gondii in primary and established cell
observed. lines. Int. J. Parasitol. 16:317-322.
The sensitivity of tissue culture and mouse inoculation 10. Jacobs, L., and M. L. Melton. 1954. Modifications in the
methods for the demonstration of Toxoplasmna organisms virulence of a strain of Toxoplasma gondii by passage in various
was compared by using several inocula of bradyzoites and hosts. Am. J. Trop. Med. Hyg. 3:447-457.
tachyzoites. The two methods were found to be equally 11. Johnson, A. M. 1984. Strain dependent route of challenge-
dependent murine susceptibility to toxoplasmosis. Z. Parasit-
sensitive, since the mean positive ratio of infected mice enkd. 70:303-309.
versus that of cultures was not significantly different whether 12. Jones, F. E., D. E. Eyles, N. Coleman, and C. L. Gibson. 1958.
the inoculum was 1 or 100 parasites. A comparison of methods for the isolation of Toxoplasma from
We propose the use of the tissue culture technique for the suspected hosts. Am. J. Trop. Med. Hyg. 7:531-535.
diagnosis of active toxoplasmosis when serological tests are 13. Lowry, O. H., N. J. Rosebrough, A. L. Farr, and R. J. Randalil.
inconclusive. In congenitally infected children, as well as in 1951. Protein measurement with the Folin phenol reagent. J.
immunocompromised patients, rapid demonstration of the Biol. Chem. 193:265-275.
presence of Toxoplasma organisms in blood or tissue ex- 14. Nakayama, I. 1967. Survey on toxoplasmosis in 200 patients
tracts may be crucial for therapy. suspected of toxoplasmosis with special reference to the method
of its detection. Jpn. J. Parasitol. 16:464-469.
15. Riggs, J. L. 1979. Immunofluorescence staining, p. 141-151. In
ACKNOWLEDGMENTS E. H. Lennette and N. J. Schmidt (ed.), Diagnostic procedures
for viral. rickettsial, and chlamydial infections. American Public
We thank Dr. Chavance (INSERM U169, Villejuif, France) for his Health Association. Washington, D.C.
helpful assistance with statistical analysis and Peter H. David for 16. Saint-Blancard, J., J. M. Kazin, P. Riberon, and F. Petit. 1982.
reviewing the manuscript. A simple and rapid procedure for large scale preparation of IgG
and albumin from human plasma by ion exchange and affinity
LITERATURE CITED chromatography, p. 305-312 lI T. J. C. Gribrau, J. Visser, and
R. F. J. Nivard (ed.), Affinity chromatography and related
1. Abbas, A. M. A. 1967. Comparative study of methods used for techniques, 1982. Elsevier, Amsterdam.
the isolation of Toxoplasina gondii. Bull. W.H.O. 36:344-346. 17. Shepp, D. H., R. C. Hackman, F. K. Conley, J. B. Anderson,
2. Chang, C. H., C. Stulberg, R. O. Bollinger, R. Walker, and J. and J. D. Meyers. 1985. Toxoplasma gondii reactivation identi-
Brough. 1972. Isolation of Toxoplasina gondii in tissue culture. fied by detection of parasitemia in tissue culture. Ann. Intern.
J. Pediatr. 81:790-791. Med. 103:218-221.