Immunology and Cell Biology (2005) 83, 483–489 doi:10.1111/j.1440-1711.2005.01356.

Research Article
Toxoplasma gondii regulates ICAM-1 mediated monocyte adhesion
to trophoblasts
ALEXANDER W PFAFF, SOPHIE GEORGES, AHMED ABOU-BACAR,
VALERIE LETSCHER-BRU, JEAN-PAUL KLEIN, MARC MOUSLI and
ERMANNO CANDOLFI

Cellular and Molecular Physiopathology and Infection Laboratory, Inserm UMR-S 392, Parasitology and Tropical
Pathology Institute, Faculty of Medicine, Louis Pasteur University, Strasbourg, France

Summary Materno-foetal transmission causes one of the most serious forms of infection with the intracellular
protozoan parasite Toxoplasma gondii. In the placenta, trophoblast cells constitute the barrier between maternal
circulation and foetal tissue. We looked at the factors that determine the extent of cell adhesion to human BeWo
trophoblast cells during T. gondii infection. BeWo monolayers stimulated with the supernatant of T. gondii-infected
PBMC showed a large increase in THP-1 cell adhesion and upregulation of the intercellular adhesion molecule
(ICAM)-1. Neutralization of cytokines by corresponding antibodies demonstrated that anti-IFN- γ, but not anti-TNF-α
or anti-IL-1β, led to a significant reduction of THP-1 adhesion to a BeWo monolayer. Treatment of BeWo cells with
single cytokines failed to induce upregulation of adhesion. In contrast, simultaneous treatment with IFN- γ and either
TNF-α or IL-1β mimicked strongly the effect of infected cell supernatant. The results suggest that IFN- γ plays a
pivotal role in the cell adhesion process through upregulation of ICAM-1 and in the process of congenital
transmission of T. gondii.

Key words: BeWo, cell adhesion, ICAM-1, Toxoplasma gondii, trophoblast.

Introduction IFN-γ.11 This indicates an enhancing effect of IFN- γ concern-

ing materno-foetal transmission. As direct invasion of cells by
The protozoan parasite Toxoplasma gondii has the remarkable
T. gondii does not depend on IFN-γ, this cytokine might aid
ability to infect and multiply in all nucleated cells of its
the parasite by increasing adhesion of infected blood cells to
human or animal host. Infection is generally benign but can
placental cells. Adhesion of circulating cells to the trophoblast
be life-threatening in immunocompromised persons and for
barrier remains poorly studied, particularly in the context of
foetuses in the case of materno-foetal transmission. 1,2 After
infection. In the present work, we investigate the regulation
oral ingestion, the parasite first affects the intestinal epithe-
by T. gondii of monocyte adhesion to trophoblasts, the con-
lium before passing into blood and lymphatic vessels. 3,4 The
comitant inflammatory environment, as well as the adhesion
multiplication stage, the tachyzoites, proliferate in specialized
receptors involved in this process, using the in vitro human
parasitophorous vacuoles. Upon lysis of the host cell, they
trophoblast cell line BeWo model.12,13
rapidly infect adjacent cells. Consequently, local adhesion of
infected cells could influence the dissemination of infection.
In the placenta, specialized cells of foetal origin, trophoblasts, Materials and methods
constitute a continuous barrier between maternal and foetal
Cell lines and parasites
blood circulation systems. These cells have a central position
in the control of materno-foetal passage of pathogens, 5 The human trophoblast cell line BeWo was kindly provided by Elisabeth
including T. gondii.6 Menu (Unité de Rétrovirologie, Institut Pasteur, Paris, France) and was
Infection with T. gondii elicits a Th1-type immune reac- maintained in Dulbecco’s modified Eagle medium with 4.5 g/L
tion with prominent production of IFN- γ, TNF-α and IL-1β.7,8 glucose supplemented with 20% heat-inactivated FCS, 2 mmol/L
This response limits both the initial infection of intestinal L-glutamine, 50 U/mL penicillin, 50 µg/mL streptomycin and 0.5 µg/mL

cells9 and dissemination in the host.10 However, we recently amphotericin B (Invitrogen, Cergy Pontoise, France). The cultures
observed diminished placental and foetal infection in a model were kept at 37°C in 5% CO2 and saturated humidity. The monocytic
cell line THP-1 was kept as suspension culture in RPMI 1640 medium
of pregnant BALB/c mice during acute primary T. gondii
(Invitrogen), supplemented as described above, except for FCS, which
infection following neutralization of IFN- γ, despite blood
was added at 10%. Tachyzoites of the virulent RH strain of T. gondii
parasite burdens higher than those without neutralization of
(Sabin 1941) were maintained in Swiss-Webster mice.

Correspondence: Dr Alexander Pfaff, Institut de Parasitologie et Cell adhesion assay

de Pathologie Tropicale, Faculté de Médecine, 3 rue Koeberlé, 67000
Strasbourg, France. Email: [email protected] BeWo cells (3 × 104) were cultured on collagen coated eight-well
Received 16 February 2005; accepted 17 March 2005. glass culture slides (LabTek; Nunc, Roskilde, Denmark) for 5 days

© 2005 Australasian Society for Immunology Inc.

484 AW Pfaff et al.

until they reached confluence. Before the adhesion assay, THP-1 cells with gene specific oligonucleotides. The membranes were then
were preincubated with phorbol 12-myristate, 13-acetate (Sigma, St exposed to a phosphorimager screen for 4 days. Quantification was
Quentin Fallavier, France) at 100 nmol/L for 6 h at 37°C to retain done with ImageQuant software (Amersham Biosciences). Normali-
their monocytic phenotype. BeWo monolayers and THP-1 cells were zation of mRNA expression on the two membranes was performed
washed separately and 3 × 104 THP-1 cells per well were added to with the aid of the housekeeping genes ubiquitin C, G3PDH, β-actin
BeWo monolayers and incubated for 90 min at 37°C. Unbound THP-1 and ribosomal protein L13a, spotted on the same membranes.
cells were removed by gently washing the wells three times with
200 µL medium. After air-drying overnight, THP-1 cells were stained
by butyrate esterase colouration,14 and adhering THP-1 cells were
Quantification of ICAM-1 mRNA by quantitative RT–PCR
counted by microscopy (at ×400 magnification) in at least 25 fields Total RNA was isolated using TriReagent (Molecular Research
per well. The results were calculated as bound THP-1 cells/mm2. Center, Cincinnati, OH, USA), according to the manufacturer’s
instructions. Five µg of RNA was reverse transcribed using the First-
Strand cDNA Synthesis Kit (Amersham Biosciences). The following
Stimulation of BeWo cells primers were synthesized by MWG Biotech (Courtaboeuf, France):
BeWo monolayers were stimulated for 24 h at the indicated concen- (i) ICAM-1 (product size 238 bp, annealing temperature 55°C) sense
trations of human recombinant IFN-γ, TNF-α and IL-1β cytokines (5′-TAT GGC AAC GAC TCC TTC T-3′), antisense (5′-CAT TCA
(Sigma) or by the supernatants of T. gondii-infected or uninfected GCG TCA CCT TGG-3′); and (ii) G3PDH (product size 137 bp,
PBMC cultures that were isolated from buffy coat of T. gondii annealing temperature 59°C) sense (5′-AGC AAT GCC TCC TGC
negative donors, provided by the Etablissement Français du Sang ACC ACC AAC-3′), antisense (5′-CCG GAG GGG CCA TCC ACA
(Strasbourg, France). Briefly, after centrifugation over Ficoll Histo- GTCT-3′). Quantification of ICAM-1 and G3PDH transcripts was
paque 1077 (Sigma), the PBMC were washed three times, and then performed by comparison to a DNA standard, generated by conven-
they were resuspended in complete RPMI 1640 medium at a concen- tional PCR, purification of the amplified DNA using the Qiaquick
tration of 2 × 106 cells/mL. T. gondii tachyzoites were added at a ratio PCR Purification Kit (Qiagen, Courtaboeuf, France) and analysis by
of one parasite per mononuclear cell. After incubation for 24 h at spectrophotometry at 260 nm. Quantification of specific cDNA was
37°C, the supernatants were removed, filtered through a 0.22 µm carried out on the LightCycler PCR system, using the LightCycler-
filter, pooled and immediately frozen at –80°C until use. Infection DNA Master SYBR Green I Kit (Roche Diagnostics, Meylan,
of PBMC was routinely checked by cytospin and standard May– France). After 7 min of denaturation at 95°C, the reaction was cycled
Grünwald–Giemsa staining.15 Typically, approximately 50% of the 40 times for 10 s at 95°C, 5 s or 10 s at the respective annealing
monocytes were infected, whereas there were very few infected temperature for ICAM-1 and G3PDH, respectively, and 20 s or 7 s at
lymphocytes. Standard ELISA tests were performed as described 72°C for ICAM-1 and G3PDH, respectively. Product specificity was
elsewhere11 to quantify IFN-γ, TNF-α and IL-1β in the supernatants determined by melting curve analysis.
of infected PBMC, which varied between 6 and 20 ng/mL, 5 and
10 ng/mL, and 1 and 2 ng/mL, respectively, while the concentrations Measurement of ICAM-1 by flow cytometry
of these cytokines in the supernatants of uninfected PBMC were
always lower than 150 pg/mL. When indicated, BeWo monolayers BeWo cells were harvested from the wells using trypsin-EDTA
were infected with 1.5 × 105 T. gondii tachyzoites per well 24 h solution (Invitrogen) and washed two times with PBS supplemented
before the adhesion assay, and THP-1 cells were infected at a rate of with 5% FCS. Cells (5 × 105) were stained with FITC-conjugated
three parasites per THP-1 cell, also for 24 h. The resulting infection antihuman ICAM-1 mouse IgG1 or isotype control (Immunotech,
rate was at 60–65% of THP-1 cells. Marseille, France). The cells were incubated at 4°C for 30 min, washed
with PBS-5% foetal bovine serum (FBS), and resuspended in PBS-
1.5% paraformaldehyde for analysis. Data were acquired with a flow
Cytokine neutralization by mAbs cytometer (FACScan; Becton Dickinson, Le Pont-De-Claix, France),
operated with the accompanying software (CellQuest, BD Biosciences,
Culture supernatants of infected PBMC were incubated with mAbs San Jose, CA, USA). Ten thousand events per sample were acquired.
raised against human IFN-γ (clone 25718.11; Sigma; 10 µg/mL),
human TNF-α (clone D2E7; Abbott, Rungis, France; 20 µg/mL) or
human IL1β (clone 8516.311; Sigma; 10 µg/mL) for 1 h at 37°C, and Co-culture of trophoblasts and T. gondii-infected monocytes
they were transferred to BeWo monolayers. Anti-intercellular
BeWo cells were grown in 24-well plates, stimulated and washed as
adhesion molecule (ICAM)-1 mAb (clone 6.5B5; DAKO, Trappes,
described above. THP-1-infected cells (2 × 105) per BeWo monolayer
France) or isotype control antibody (BD Biosciences, Le Pont de
well were added and incubated for 90 min at 37°C. The cultures were
Claix, France) were added at 10 µg/mL to prestimulated BeWo
gently washed three times with 400 µL medium to remove unbound
monolayers and incubated for 1 h.
monocytes, and maintained for a further 24 h at 37°C. The percentage
of infected trophoblasts was evaluated by flow cytometry. Briefly, the
cells of the co-cultures were harvested and stained with an antihuman
Gene expression array analysis
CD45-PE mAb (BD Biosciences) to exclude THP-1 cells. Cells were
Confluent BeWo monolayers in culture flasks were incubated with incubated at 4°C for 30 min, washed with PBS-5% FBS, resuspended
supernatants of uninfected or T. gondii-infected PBMC for 24 h. in 100 µL of Cytofix/Cytoperm (BD Biosciences) and incubated for
Cells were then removed from the flasks using a cell scraper, and 20 min at 4°C. Cells were washed two times with PermWash buffer
total RNA was isolated with the NucleoSpin II Kit (BD Biosciences). (BD Biosciences), treated with anti-T. gondii antibody (Neomarkers,
Five µg of RNA per sample was then subjected to gene expression Westinghouse, CA, USA) for 45 min, washed and stained with
array analysis, using the Cell Interaction Kit (BD Biosciences), antirabbit IgG-FITC (Sigma) for 45 min at 4°C. After final washes,
following the manufacturer’s recommendations. Briefly, RNA was cells were resuspended in PBS-1.5% paraformaldehyde and analysed
reverse transcribed, labelled with 32P-dATP (Amersham Biosciences, by flow cytometry as above. Twenty thousand events per sample
Freiburg, Germany) and hybridized on a nylon membrane spotted were acquired.
 Effect of T. gondii on cell adhesion 485

Figure 2 Effect of Toxoplasma gondii infection on adhesion of

THP-1 cells to BeWo monolayers. BeWo cells were treated with
the supernatants of uninfected PBMC (control) or T. gondii-
infected PBMC (stimulated). Tachyzoite infection of BeWo or
THP-1 cells was performed simultaneously during cell stimulation.
Figure 1 Involvement of cytokines in the THP-1 cell adhesion Data are means ± SD of three separate experiments performed in
process to BeWo monolayers. (A) BeWo cells were treated with duplicate and analysed using the Student’s t-test. , No infection;
supernatant (Sn) of Toxoplasma gondii-infected PBMC and , THP-1 infection; , BeWo infection.
cytokines were neutralized by anti-IFN-γ, anti-TNF-α or anti-
IL-1β antibodies. Adhesion of THP-1 cells was then performed
on pretreated BeWo monolayers. *P < 0.05, as compared to PBMC. Neutralization of all three cytokines did not further
treatment with supernatant of infected PBMC in the absence of decrease THP-1 adhesion (data not shown). We also deter-
neutralizing antibodies. (B) BeWo cells were stimulated for 24 h mined the influence of these cytokines on the adhesion
with IFN-γ (20 ng/mL), TNF-α (10 ng/mL) or IL-1β (1 ng/mL). process. We added the cytokines IFN- γ, TNF-α and IL-1β,
Data are means ± SD of three separate experiments performed in alone or in combination, to the BeWo cell culture, before the
duplicate and analysed using the Student’s t-test. *P < 0.05, as adhesion assay. As shown in Figure 1B, no cytokine was able
compared to medium group. to enhance adhesion of THP-1 cells when added alone.
However, when IFN-γ was added in combination with either
TNF-α or IL-1β, adhesion was significantly increased. No
Statistical analysis such effect was observed with the combination of TNF-α and
IL-1β, without addition of IFN-γ. These results suggest a
All experiments were performed at least three times. Data are
expressed as mean ± SD. Statistical analysis was performed using the
pivotal role of IFN-γ in the supernatant of T. gondii-infected
Student’s t-test except for co-culture analysis, which was assessed PBMC for increased adhesion of THP-1 cells to BeWo
with a χ2-test. P-values < 0.05 were considered to be statistically trophoblasts.
significant. We looked at whether T. gondii infection of either THP-1
or BeWo cells has any effect on adhesion (Fig. 2). Infection
of THP-1 cells did not change their adhesion capacity. We
Results then wanted to investigate THP-1 adhesion to an already
infected BeWo cell layer. As shown in Figure 2, infection by
Effects of cytokines on THP-1 cell adhesion to BeWo cells T. gondii did not alter the number of adhering THP-1 cells.
As T. gondii infection is associated with production of The above results were observed with both non-stimulated
inflammatory and Th1 cytokines, we studied the effect of the BeWo cells and cells stimulated with the supernatant of
supernatant that was obtained from T. gondii-infected PBMC T. gondii-infected PBMC. We conclude therefore that infec-
on THP-1 adhesion to BeWo monolayers, as well as the tion of the cells does not change adhesion.
effects of cytokines on this process. Figure 1A shows that
stimulation of BeWo cells with the supernatant resulted in a
Expression analysis of adhesion molecules on stimulated
high and significant increase of THP-1 cell adhesion to the
BeWo cells
BeWo monolayer. The possible roles of cytokines secreted in
the supernatant were assessed by treatment of the supernatant To identify the variations of adhesion molecules expressed
with neutralizing antibodies. Only the neutralization of IFN- γ, in BeWo cells following stimulation with supernatant of
but not of TNF-α or IL-1β, led to a significant reduction of uninfected and T. gondii-infected PBMC, a gene expression
THP-1 adhesion to the BeWo monolayer, which almost array analysis was performed by comparing mRNA abundance
reached the same level as with the supernatant of uninfected of various adhesion molecules. In Table 1, we show the
486 AW Pfaff et al.

Table 1 Upregulated adhesion molecule-related genes in stimulated

BeWo cells†

Antigen GenBank Fold increase‡

Accession No.
ICAM-1 (CD54) J03132 7.7
Tenascin R X98085 7.0
Integrin beta 7 M73780 6.0
Chondroitin sulphate U16306 5.8
proteoglycan core protein 2
Chondroitin sulphate M55172 5.6
proteoglycan 1
Integrin alpha 8 L36531 4.8
†BeWo cells were stimulated for 24 h with supernatants of unin-

fected or Toxoplasma gondii-infected PBMC. mRNA expression was

assessed by BD Biosciences’ Cell Interaction Gene Expression Array,
as described in the Materials and methods. ‡The number indicates the
increase of mRNA in stimulated BeWo cells versus unstimulated
BeWo cells. Only genes with a greater than fourfold increase of
expression are listed.

Figure 3 Effect of ICAM-1 neutralization on THP-1 cell adhesion

to BeWo monolayers. BeWo cells were incubated with anti-ICAM-1
expression of the few adhesion-related genes that were upreg- () or isotype control antibody () at 10 µg/mL for 1 h following
ulated after BeWo cell stimulation. The most remarkable treatment with supernatants of uninfected PBMC (control) or
molecule was ICAM-1, but upregulation of expression of two Toxoplasma gondii-infected PBMC (stimulated) for 24 h. Data
integrin subunits and chondroitin sulphate related genes was are means ± SD of three separate experiments performed in dupli-
also observed. Genes for other molecules regularly described cate and analysed using the Student’s t-test. *P < 0.05, as com-
to be involved in adhesion to vascular cells, such as all other pared to control antibody.
integrins, VCAM-1 and E-selectin, were not expressed (data
not shown).

and either TNF-α or IL-1β increased ICAM-1 expression

Regulation of ICAM-1 expression
highly and significantly. In addition, the analysis of ICAM-1
The regulation of ICAM-1 expression by IFN-γ, TNF-α and mRNA expression levels showed a similar pattern of stimula-
IL-1β in BeWo trophoblast cells was assessed by analysing tion as protein expression under the same experimental
the mRNA and protein levels. First, however, we investigated conditions (Fig. 5), only the combination of IFN-γ and either
the role of ICAM-1 on BeWo cells. BeWo monolayers were TNF-α or IL-1α being able to significantly augment ICAM-1
treated with an anti-ICAM-1 antibody after incubation with mRNA production.
supernatants of naive or infected PBMC, before the adhesion These results show that IFN-γ is crucial for upregulating
assay (Fig. 3). In unstimulated BeWo cells, no significant ICAM-1 expression on BeWo cells and as observed in the
effect on the adhesion process was observed when ICAM-1 adhesion assays, IFN-γ was also the pivotal cytokine in the
was neutralized by anti-ICAM-1 on the BeWo surfaces. This supernatant of T. gondii-infected PBMC involved in upregu-
indicates that ICAM-1 is not involved in the adhesion process lation of ICAM-1 expression.
in unstimulated BeWo cells. In contrast, the highly increased
THP-1 adhesion upon stimulation of BeWo cells by treatment
Transmission of T. gondii from infected THP-1 to BeWo cells
with the supernatant of infected PBMC was abrogated by
neutralization of ICAM-1 on BeWo cells by mAb treatment. To study the possible transmission of the parasite from infected
This result suggests that the augmentation of adhesion to monocytes to placenta, we co-cultured T. gondii-infected
BeWo cells is largely due to a substantial upregulation of THP-1 cells with an uninfected BeWo monolayer as a model
ICAM-1 expression on stimulated BeWo cells. of monocyte–trophoblast parasite transmission. Following
We then investigated the role of cytokines on the expres- 90 min of initial incubation, the nonadherent infected THP-1
sion of ICAM-1 on BeWo cells. Stimulation of BeWo cells cells were removed and co-cultures were further incubated for
with supernatant of T. gondii-infected PBMC considerably 24 h to assess the influence of cytokines involved in the
upregulated ICAM-1 expression on the cell surface (Fig. 4), previously observed adherence process. Table 2 shows that
which corroborated our results of the gene expression array BeWo cells were infected with T. gondii. The infection rate
analysis (Table 1). Blocking studies using anti-cytokine anti- was significantly increased when BeWo cells were pretreated
bodies with the supernatant of T. gondii-infected PBMC with the supernatant of T. gondii-infected PBMC. The trans-
showed that only anti-IFN-γ treatment was able to signifi- mission of T. gondii from infected THP-1 cells to BeWo cells
cantly inhibit ICAM-1 expression. Moreover, the addition of was significantly reduced by anti-ICAM-1 and anti-IFN-γ
single cytokines to BeWo cells failed to increase ICAM-1 treatment, but was not affected by anti-TNF-α or anti-IL-1β
expression significantly, whereas the combination of IFN-γ treatment. These results demonstrate strongly that adhesion of
 Effect of T. gondii on cell adhesion 487

ICAM-1/ G3PDH mRNA (x10.000)

Figure 5 Effects of cytokines on ICAM-1 mRNA abundance in

BeWo cells. Cells were stimulated with IFN-γ (20 ng/mL), TNF-α
(10 ng/mL) or IL-1β (1 ng/mL), alone or in combination, or
exposure to supernatants (Sn) of Toxoplasma gondii-infected
PBMC, for 24 h. Results are expressed as ratio of ICAM-1 and
G3PDH transcripts as determined by quantitative reverse transcrip-
tion (RT)–PCR. Data are means ± SD of three separate experiments
Figure 4 Effects of cytokines on ICAM-1 protein expression at performed in duplicate and analysed using the Student’s t-test.
the surface of BeWo cells. (A) BeWo cells were stimulated for *P < 0.05, as compared to medium group.
24 h with the supernatants (Sn) of uninfected or Toxoplasma
gondii-infected PBMC. The supernatant of infected PBMC was
neutralized by the indicated mAbs for 1 h and then added to the Table 2 Effect of neutralizing antibodies on Toxoplasma gondii
BeWo cells. ICAM-1 protein expression was determined by flow transmission to BeWo cells†
cytometry. The values represent the ratios of the fluorescence
intensities with anti-ICAM-1 specific antibody and isotype control BeWo stimulation Added Infected BeWo
antibody. *P 0.05, as compared to the non-neutralized super- antibody cells (%)
natant. (B) BeWo cells were stimulated for 24 h with IFN-γ Sn of uninfected PBMC None 1.1‡
(20 ng/mL), TNF-α (10 ng/mL) or IL-1β (1 ng/mL), alone or in Sn of infected PBMC None 4.2
combination. Data are means ± SD of three separate experiments Anti-IFN-γ 1.9‡
performed in duplicate and analysed using the Student’s t-test. Anti-TNF-α 4.1
Anti-IL-1β 4.5
*P < 0.05, as compared to medium group.
Anti-ICAM-1 2.8‡
Isotype control 4.0
infected monocytes is an important process of enhancing the
† BeWo cells were stimulated for 24 h with supernatants (Sn) of
placental infection rate, as well as the central roles of IFN-γ
and ICAM-1 in parasite transmission. uninfected or T. gondii-infected PBMC. Anti-IFN-γ, anti-TNF-α and
anti-IL-1β mAbs were added to the supernatant and incubated for 1 h
before transfer to the BeWo monolayer. Anti-ICAM-1 antibody was
Discussion added to the BeWo culture for 1 h following stimulation. T. gondii-
infected THP-1 cells were then added and incubated for 90 min. After
Transplacental passage is one of the major complications of
removing nonadherent cells, incubation was maintained for a further
T. gondii infection.1 Elucidating the mechanisms that control
24 h. Infection of BeWo cells was evaluated by flow cytometry.
this passage is of great interest for developing new therapy ‡P < 0.05, compared to BeWo cells stimulated with supernatant of
strategies. IFN-γ, along with other inflammatory cytokines, infected PBMC, without neutralizing antibody, χ2-test.
plays a major role in controlling infection in the host. 7,8
However, our observation of diminished placental and foetal
infection in T. gondii-infected mice upon neutralization of macrophages in the intervillous space. 19 Trophoblast cells
IFN-γ despite a higher blood parasite burden 11 suggests that are also able to produce chemokines and their receptors. 20,21
IFN-γ may have a yet unknown regulatory role at the The role of these processes in materno-foetal transmission of
materno-foetal interface. Its production might, at that particu- T. gondii is not clear. However, immunohistochemical studies
lar site, even be harmful for the host, as it facilitates the in a rat model of congenital infection showed T. gondii
materno-foetal passage of the parasite. Inflammatory cytokines multiplication in intact trophoblast cells. A recent study showed
were associated with severe placental damage in human that T. gondii exploits ICAM-1 expression on endothelial
pregnancies.16 In vitro studies showed that adhesion of cells for enhanced invasion.22 In our study, we show another
monocytes, which is promoted by IFN-γ and TNF-α, can lead regulation mechanism that involves inflammatory cytokines
to damage and apoptosis in placental cells. 17,18 Moreover, enhancing adhesion of infected blood cells to trophoblast cells.
enhanced production of chemokines in the placenta during Monocytes were previously suggested to be an important
an inflammatory reaction contributes to an accumulation of reservoir of T. gondii in the blood.23 Therefore, enhanced
488 AW Pfaff et al.

adhesion of monocytes to placental cells may enhance Acknowledgements

materno-foetal transmission of the parasite. Moreover, activa-
This work was supported by the Marie Curie Research Training
tion of adhesion mechanisms has been implicated in patho-
Programme of the European Commission, project MCFI-
logical processes for various infectious agents, notably
2001-01071, and by Les Hôpitaux Universitaires de Strasbourg.
Plasmodium spp. Expression of adhesion molecules by brain
We wish to thank Dr Falkenroth and the Laboratoire
endothelial cells is linked to the cerebral complication of
d’Hématologie for their valuable help with the staining of
malaria.24 Infected pregnant women have large accumulations
slides.
of infected red blood cells in the placenta. 25 Adhesion of
Plasmodium-infected red blood cells to placental cells relies
mainly on expression of chondroitin sulphate expression on References
trophoblast cells,26 which is probably different for leucocytes,
1 Wong SY, Remington J. Toxoplasmosis in pregnancy. Clin.
where adhesion relies on expression of ICAM-1, VCAM-1, Infect. Dis. 1994; 18: 853–62.
E-selectin and other receptors. In the present work, we 2 Luft B, Brooks RG, Conley FK, McCabe RE, Remington JS.
investigated in vitro the regulating mechanisms of cell adhe- Toxoplasmic encephalitis in patients with acquired immune
sion to trophoblast. deficiency syndrome. JAMA 1984; 252: 913–17.
Gene expression array, quantitative PCR and FACS analy- 3 Bertoli F, Espino M, Arosemena JR, Fishback JL, Frenkel JK. A
sis showed a significant increase of few adhesion related spectrum in the pathology of toxoplasmosis in patients with
genes in BeWo cells. One report similar to our present study acquired immunodeficiency syndrome. Arch. Pathol. Lab. Med.
demonstrates that ICAM-1 is the principal receptor for 1995; 119: 214–24.
increased monocyte adhesion to activated primary trophoblast 4 Dubey J, Speer C, Shen S, Kwok O, Blixt J. Oocyst-induced
cultures.27 In contrast, another in vitro study has demonstrated murine toxoplasmosis: life cycle, pathogenicity, and stage con-
low sensitivity of BeWo cells to IL-1β- and TNF-α-mediated version in mice fed Toxoplasma gondii oocysts. J. Parasitol.
induction of ICAM-1 expression,28 which may be related to 1997; 83: 870–82.
the low or absent expression of ICAM-1 on trophoblast cells 29 5 Burton GJ, Watson AL. The structure of the human placenta:
and in histological sections of Plasmodium-infected placentas.30 Implications for initiating and defending against virus infections.
The present data give a homogenous picture of the central Rev. Med. Virol. 1997; 7: 219–28.
role of IFN-γ in increasing the cell adhesion process and 6 Ferro EA, Bevilacqua E, Favoreto-Junior S, Silva DA, Mortara RA,
ICAM-1 expression. In fact, treatment of BeWo cells with Mineo JR. Calomys callosus (Rodentia : Cricetidae) trophoblast
cells as host cells to Toxoplasma gondii in early pregnancy.
cytokines or treatment of supernatant of T. gondii-infected
Parasitol. Res. 1999; 85: 647–54.
PBMC with anti-cytokine antibodies shows that IFN-γ is
7 Denkers EY, Gazzinelli RT. Regulation and function of T-cell-
indispensable, but not sufficient, for activation of BeWo cells.
mediated immunity during Toxoplasma gondii infection. Clin.
Either TNF-α or IL-1β is required to provide the necessary
Microbiol. Rev. 1998; 11: 569–88.
second signal for cell adhesion. The requirement for a second 8 Chang HR, Grau GE, Pechere JC. Role of TNF and IL-1 in infec-
signal to upregulate ICAM-1 expression in BeWo cells is not tions with Toxoplasma gondii. Immunology 1990; 69: 33–7.
required in other cell types.31,32 Moreover, BeWo cells support 9 Dimier IH, Bout DT. Inhibition of Toxoplasma gondii replica-
less adhesion of leucocytes than endothelial cells 28 or MRC-5 tion in IFN-gamma-activated human intestinal epithelial cells.
fibroblast cells (AW Pfaff, pers. obs., 2005). However, infec- Immunol. Cell Biol. 1997; 75: 511–14.
tion of trophoblast cells with cytomegalovirus was recently 10 Alexander J, Scharton-Kersten TM, Yap G, Roberts CW, Liew FY,
shown to upregulate ICAM-1 expression, 18 suggesting that Sher A. Mechanisms of innate resistance to Toxoplasma gondii
T. gondii infection elicits different reactions of the cells when infection. Philos. Trans. R. Soc. Lond. B. Biol. Sci. 1997; 352:
compared to viral infection. This is possibly related to the 1355–9.
formation of the unique parasitophorous vacuole during 11 Abou-Bacar A, Pfaff AW, Georges S et al. Role of NK cells and
T. gondii invasion, which effectively interferes with host cell gamma interferon in transplacental passage of Toxoplasma
activation.33 However, trophoblast cells are able to react to gondii in a mouse model of primary infection. Infect. Immun.
strong inflammatory stimuli, which are present during acute 2004; 72: 1397–401.
T. gondii infection. This stimulation is enough to significantly 12 Lagaye S, Derrien M, Menu E et al. Cell-to-cell contact results in
enhance cell adhesion of potentially infected blood cells. a selective translocation of maternal human immunodeficiency
Signalling pathways triggered by different pathogens in tropho- virus type 1 quasispecies across a trophoblastic barrier by both
blast cells should be investigated to shed light on the particu- transcytosis and infection. J. Virol. 2001; 75: 4780–91.
13 Ellinger I, Schwab M, Stefanescu A, Hunziker W, Fuchs R. IgG
lar immune response at the materno-foetal interface.
transport across trophoblast-derived BeWo cells: a model system
In conclusion, the present data suggest that the inflamma-
to study IgG transport in the placenta. Eur. J. Immunol. 1999; 29:
tory cytokines secreted during T. gondii infection enhance the
733–44.
adhesion of blood cells to the trophoblast layer in the pla-
14 Mufti GJ, Schaefer HE, Sandberg AA, Flandrin G. M Dunitz
centa, and that this process could contribute to placental and (ed.) An Atlas of Malignant Haematology. London, UK: Lippin-
ultimately foetal infection. ICAM-1 plays a major part in cell cott Williams and Wilkins, 1996.
adhesion to activated trophoblast cells. IFN-γ is the central 15 Locquin M, Langeron M. Dunod (ed.) Manuel de Microscopie.
mediator of trophoblast activation in the context of T. gondii Paris, France: Masson SA, 1978.
infection, while TNF-α and IL-1β have auxiliary functions. 16 El-Shazly S, Makhseed M, Azizieh F, Raghupathy R. Increased
Further investigation of the transmission process using the expression of pro-inflammatory cytokines in placentas of women
BeWo trophoblast model will help to increase our insight into undergoing spontaneous preterm delivery or premature rupture
the process of congenital transmission of T. gondii. of membranes. Am. J. Reprod. Immunol. 2004; 52: 45–52.
 Effect of T. gondii on cell adhesion 489

17 Garcia-Lloret MI, Winkler-Lowen B, Guilbert LJ. Monocytes 26 Rogerson S, Brown G. Chondroitin sulphate A as an adherence
adhering by LFA-1 to placental syncytiotrophoblasts induce receptor for plasmodium falciparum-infected erythrocytes. Para-
local apoptosis via release of TNF-alpha. A model for hemato- sitol. Today 1997; 13: 69–75.
genous initiation of placental inflammations. J. Leukoc. Biol. 27 Xiao J, Garcia-Lloret M, Winkler-Lowen B, Miller R, Simpson K,
2000; 68: 903–908. Guilbert LJ. ICAM-1-mediated adhesion of peripheral blood
18 Chan G, Stinski MF, Guilbert LJ. Human cytomegalovirus- monocytes to the maternal surface of placental syncytiotropho-
induced upregulation of intercellular cell adhesion molecule-1 on blasts: implications for placental villitis. Am. J. Pathol. 1997;
villous syncytiotrophoblasts. Biol. Reprod. 2004; 161: 797–803. 150: 1845–60.
19 Suguitan AL Jr, Leke RG, Fouda G et al. Changes in the levels 28 Jarousseau AC, Thibault G, Reverdiau P et al. Adhesive proper-
of chemokines and cytokines in the placentas of women with ties of choriocarcinoma cells toward lymphocytes activated or
Plasmodium falciparum malaria. J. Infect. Dis. 2003; 188: not by interleukin-2. Cell. Immunol. 1994; 157: 38–47.
1074–82. 29 Gaffuri B, Vigano P, Nozza A, Gornati G, Di Blasio AM,
20 Dolcini G, Derrien M, Chaouat G, Barre-Sinoussi F, Menu E. Vignali M. Expression of intercellular adhesion molecule-1
Cell-free HIV type 1 infection is restricted in the human tropho- messenger ribonucleic acid and protein in human term placen-
blast choriocarcinoma BeWo cell line, even with expression of tal cells and its modulation by pro-inflammatory cytokines
CD4, CXCR4 and CCR5. AIDS Res. Hum. Retroviruses 2003; (interleukin-1beta and tumor necrosis factor alpha). Biol.
19: 857–64. Reprod. 1998; 58: 1003–1008.
21 Moussa M, Mognetti B, Dubanchet S et al. Expression of beta 30 Sugiyama T, Cuevas LE, Bailey W et al. Expression of inter-
chemokines in explants and trophoblasts from early and term cellular adhesion molecule 1 (ICAM-1) in Plasmodium falciparum-
human placentae. Am. J. Reprod. Immunol. 2001; 46: 309–17. infected placenta. Placenta 2001; 22: 573–9.
22 Barragan A, Brossier F, Sibley LD. Transepithelial migration of 31 Takahashi K, Takigawa M, Takashiba S et al. Role of cytokine
Toxoplasma gondii involves an interaction of intercellular adhe- in the induction of adhesion molecules on cultured human gin-
sion molecule 1 (ICAM-1) with the parasite adhesin MIC2. Cell gival fibroblasts. J. Periodontol. 1994; 65: 230–35.
Microbiol. 2005; 7: 561–8. 32 Spoelstra FM, Postma DS, Hovenga H, Noordhoek JA,
23 Channon JY, Seguin RM, Kasper LH. Differential infectivity Kauffman HF. Interferon-gamma and interleukin-4 differentially
and division of Toxoplasma gondii in human peripheral blood regulate ICAM-1 and VCAM-1 expression on human lung fibro-
leukocytes. Infect. Immun. 2000; 68: 4822–6. blasts. Eur. Respir. J. 1999; 14: 759–66.
24 Aikawa M, Iseki M, Barnwell JW, Taylor D, Oo MM, Howard RJ. 33 Lingelbach K, Joiner KA. The parasitophorous vacuole mem-
The pathology of human cerebral malaria. Am. J. Trop. Med. brane surrounding Plasmodium and Toxoplasma: an unusual
Hyg. 1990; 43: 30–37. compartment in infected cells. J. Cell. Sci. 1998; 111: 1467–75.
25 Walter PR, Garin Y, Blot P. Placental pathologic changes in
malaria. A histologic and ultrastructural study. Am. J. Pathol.
1982; 109: 330–42.