Lower Expression of TLR2 and SOCS-3 Is Associated with
Schistosoma haematobium Infection and with Lower
Risk for Allergic Reactivity in Children Living in a Rural
Area in Ghana
Franca C. Hartgers1*, Benedicta B. Obeng1,2, Yvonne C. M. Kruize1, Marjolijn Duijvestein1, Anna de
Breij1, Abena Amoah2, Irene A. Larbi2, Ronald van Ree3, Michael D. Wilson2, Laura C. Rodrigues4,
Daniel A. Boakye2, Maria Yazdanbakhsh1
1 Department of Parasitology, Leiden University Medical Centre, Leiden, The Netherlands, 2 Noguchi Memorial Institute for Medical Research, University of Ghana, Legon,
Accra, Ghana, 3 Department of Experimental Immunology, Academic Medical Center, Amsterdam, The Netherlands, 4 Department of Infectious Diseases, London School
of Hygiene & Tropical Medicine, London, United Kingdom
Abstract
Background: Helminth infections are prevalent in rural areas of developing countries and have in some studies been
negatively associated with allergic disorders and atopy. In this context little is known of the molecular mechanisms of
modulation involved. We have characterized the innate immune responses, at the molecular level, in children according to
their helminth infection status and their atopic reactivity to allergens.
Methodology/Principal Findings: The mRNA expression of several genes of the innate immune system that have been
associated with microbial exposure and allergy was examined in 120 school children in a rural area in Ghana. Helminth
infections were common and atopy rare in the study area. The analysis of gene expression in ex vivo whole blood samples
reflected the levels of corresponding proteins. Using this approach in a population of school children in whom the presence
of Schistosoma haematobium infection was associated with protection from atopic reactivity, we found that the level of
TLR2 and SOCS-3, genes associated with atopy in the children, were significantly downregulated by presence of S.
haematobium infection.
Conclusions: S. haematobium infections modulate the expression of genes of the innate immune system (TLR2 and SOCS-3);
these are genes that are associated with increased allergic inflammatory processes, providing a molecular link between the
negative association of this infection and atopy in rural children in Ghana.
Citation: Hartgers FC, Obeng BB, Kruize YCM, Duijvestein M, de Breij A, et al. (2008) Lower Expression of TLR2 and SOCS-3 Is Associated with Schistosoma
haematobium Infection and with Lower Risk for Allergic Reactivity in Children Living in a Rural Area in Ghana. PLoS Negl Trop Dis 2(4): e227. doi:10.1371/
journal.pntd.0000227
Editor: Andrew Scott MacDonald, University of Edinburgh, United Kingdom
Received November 26, 2007; Accepted March 20, 2008; Published April 16, 2008
Copyright: 
 2008 Hartgers et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits
unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: This work was supported in part by The Netherlands Foundation for the Advancement of Tropical Research (grant WB 93-443), The Dutch Organization
for Scientific Research (NWO), ZonMW TOP program 912-03-048, GA2LEN-FOOD-CT-2004-506378, GLOFAL-FOOD-CT-2005-517812, and EUROPREVALL-FOOD-CT-
2005-514000.
Competing Interests: The authors have declared that no competing interests exist.
* E-mail: F.C.Hartgers@lumc.nl
Introduction reached between pro- and anti-inflammatory responses, such that
Th2 responses are kept under control when allergens are
In the last few decades, allergic diseases have become a major encountered.
health burden in the western world. Although these disorders In rural areas in the developing world, chronic helminth
clearly have a genetic component, their rapid change in infections are highly prevalent. These infections not only result in
prevalence points to environmental conditions that have changed skewing of the immune responses towards Th2, but also induce the
during this time frame. In the same time frame, there has been a higher production of anti-inflammatory molecules such as IL-10 to
decrease in exposure to microbial products as a result of changing prevent the elimination of helminths, which at the same time
lifestyle with, among others, improved sanitation and access to protect the host against the pathological consequences of excessive
clean water. Interestingly, in the developing world, the prevalence inflammation [1]. Such an anti-inflammatory environment
of allergies is relatively low, particularly in rural areas, where induced by chronic helminth infections might modulate immune
exposure to infectious agents is high. There is increasing evidence responses to other antigens. For example, chronic infection with
that exposure to pathogen-derived compounds influences the schistosomes or Onchocerca was shown to modulate the immune
maturation of the immune system and therefore the balance response to tetanus toxoid following vaccination [2,3].
www.plosntds.org 1 April 2008 | Volume 2 | Issue 4 | e227
mRNA Expression Profiles of Rural Ghanaian Children
Author Summary only of TLRs but also of molecules such as SOCS-1 and SOCS-3,
involved in downstream signalling, were studied in whole blood
Inflammatory diseases such as atopic disorders are a major samples of rural Ghanaian school children.
health problem in the Western world, but their prevalence The results of this study showed that high expression of TLR2
is also increasing in developing countries, especially in and SOCS-3 was associated with allergic skin reactivity, whereas
urban centres. There is increasing evidence that exposure helminth infection was associated with lower expression levels of
to a rural environment with high burden of compounds TLR2 and SOCS-3, providing a potential regulatory link between
derived from parasites and microorganisms is associated helminth infection and allergies at the molecular level.
with protection from atopic disorders. Since urbanisation is
progressing at a rapid pace, particularly in less-developed
nations, there is a need to understand the molecular Methods
processes that control the progress towards the develop- Study population
ment of allergic diseases in developing countries. In this
The study population consisted of schoolchildren between 5 and
study we have examined a population of school children
living in a rural area of Ghana, where helminth (worm) 14 years of age. Children whose parents consented by signing or
infections are prevalent and associated with protection thumb printing an informed consent form were registered to
from skin reactivity to house dust mite. Blood samples participate in a large study on allergy and parasitic infections (B.B.
were collected from these children and analysed for the Obeng et al, manuscript submitted). The Institutional Review
expression levels of several genes involved in the Board of the Noguchi Memorial Institute for Medical Research,
development of a pro allergic immune system. The results Accra, Ghana approved the study. Skin reactivity to mite was
point at a potential molecular link that might explain the negatively associated with S. haematobium infection (OR 0.5, 95%
negative association between schistosome infections and CI 0.2–1.0, p = 0.05), particularly in areas where prevalence of
allergies. schistosomiasis is high (OR 0.3, 5% CI 0.1–0.9, p = 0.04). Blood
samples from children from two rural schools with high prevalence
of S. haematobium infection were used for RNA isolation to study
Epidemiological studies have revealed both positive and gene expression. In these schools, the reactivity to house dust mite
negative associations between helminth infections and allergies was low (9%) compared to school children from Accra (15%), free
(reviewed in [4]). It is thought that severe, chronic infections are of any helminth infections, and with a relatively high socioeco-
often associated with suppression of allergic reactivity. For nomic status.
example chronic infections with intestinal helminth, such as with The study subjects were fist selected randomly, one out of three
hookworm, have been shown to suppress allergic diseases [5,6]. children from whom blood samples were available were selected
These observations have been confirmed for schistosomiasis, (107 children). In order to increase power, all skin prick test (SPT)
demonstrating lower skin reactivity to allergen in infected positive children from these schools were added to our randomly
individuals [7,8]. Additionally, removal of helminths by long-term selected subjects, along with randomly selected SPT negative
anti-helminth treatment in Venezuelan or Gabonese children children (16 children in total), resulting in a group of 123 children
resulted in increased atopic reactivity to house dust mite [9,10], (Table 1).
even though a shorter anti-helminth treatment did not show an
effect on atopy in one study [11]. Detection of helminth and malarial infection
In a population of rural Ghanaian school children a negative The participants were given specimen bottles and were asked to
association was found between infection with Schistosoma haemato- collect a fresh stool and urine sample for the detection of helminth
bium and skin reactivity to mite allergen (Obeng et al, submitted for infections. Stool examination was performed by the Kato-Katz
publication). Within this study we aimed to identify the molecular method for the detection of hookworm and trichuris, and the total
mechanisms by which schistosome infections may modify immune number of eggs was calculated per gram of faeces. Urine samples
responses and modulate inflammatory reactions such as atopy. To were used for the detection of S. haematobium by passing 10 ml of
address this we selected two groups of genes that have been urine through a filter with 10-micron pore size. A subject was
described previously to play a role in allergic diseases. Toll-like considered positive for helminth infection if eggs of any of the
receptors (TLRs) have been shown in several studies outside Africa helminth species were detected.
to change in expression levels following exposure to microorgan- Blood samples were collected from all participants for the
isms [12–15]. In a European study these molecules were linked to detection of the malaria infection by Giemsa-stained thick smear
allergy: children of farmers in Alpine regions, exposed to high (GTS) examination.
microbial burden and with a low prevalence of atopy, had altered
levels of TLR2 [16], suggesting that exposure to microorganisms Skin prick test for mite allergen
might modulate the innate immune system and thereby suppress The immediate hypersensitivity skin prick test with inhalant
the development of allergic disorders. The molecules suppressor of allergen extracts was performed by using the standard prick
cytokine signalling (SOCS)-1 and SOCS-3 have recently been method on the volar surface of the right forearm with the
described and reported to be involved in TLR signalling and standardized extracts of 6 allergens; Dermatophagoides pteronyssinus
inflammatory diseases [17–22]. Elegant studies by Kubo and co (Der P), Dermatophagoides farinae (Der F), cat, dog, peanut and grass
workers in animal models have shown SOCS-3 to be involved in mix (HAL Allergen Laboratories, The Netherlands). Histamine
regulation of immune responses in allergic disease [23,24]. dihydrochloride (1/1000) and glycerinated saline solution were
Given that helminth products have been shown to modulate used as positive and negative controls, respectively. The wheal
cells of the innate immune system and to interfere with pathways diameter was measured after 15 minutes and the result considered
that are activated via TLR stimulation [15,25], we asked whether positive when the wheal size was at least 3 mm in diameter, in the
in an area in Africa where helminth infections are highly prevalent absence of significant reactivity of the diluent negative control.
and allergic disorders are low, we can find molecular pathways None of the children were taking anti-allergic medicine that might
that may explain the relationship. To this end, gene expression not interfere with SPT or had ever been treated with specific
www.plosntds.org 2 April 2008 | Volume 2 | Issue 4 | e227
mRNA Expression Profiles of Rural Ghanaian Children
Table 1. Characteristics of study population. RNA isolation from cell subsets
From a subset of the donors, PBMC were isolated and
monocytes and T cells were separated by subsequent labelling
School children n = 123 and magnetic cell separation of cells with CD3 and CD14
Microbeads (Miltenyi Biotech, Germany). Fractionated monocytes
Age (yrs)
and T cells were mixed with Nuclisens lysis buffer. Fluorescence
median 9.0 activated cell sorting (FACS) of the isolated cells indicated that
(min-max) (5–14) these fractions were at least 90% pure. The unlabeled cell fraction
Sex depleted for monocytes and T cells was also collected and mixed
M/F (% M) 66/57 (54%) with lysisbuffer (remaining cell fraction). Samples were stored and
RNA was isolated as described for whole blood samples. The
Helminth infection
percentages of monocytes and T cells in all donors were
at least one helminth (%) 60/113 (52%)
determined by flow cytometry of the PBMC prior to the isolation
S. haematobium (%) 46/120 (38%) procedure.
hookworm (%) 29/119 (24%)
other helminth (%) 4/120 (3%) cDNA synthesis and real-time PCR
Schistosome egg load Reverse transcription of RNA was performed using moloney
GM (eggs/10 ml urine) 35 murine leukaemia virus reverse transcriptase (M-MLV RT)
(Invitrogen). Samples without RT were regularly taken along to
(min-max) (1–1112)
control for genomic DNA contamination. Gene expression was
Hookworm egg load
assessed with real-time quantitative PCR (Prism 7700, Applied
GM (eggs/gram faeces) 186 Biosytems). PCR reactions were performed in duplicate in
(min-max) (20–3540) accordance with the TaqmanTM assay instructions using Taqman
Malaria infection (%) 59/117 (50%) probes and qPCR Core kit reagents (both Eurogentec, Seraing,
Belgium). Gene expression was normalized to the housekeeping
Egg loads are represented as geometric means for infected individuals only gene 18S rRNA and calculations were performed as described
Malaria infection was examined by microscopic examination of thick blood
[28]. Analysis of the expression of 8 different housekeeping genes
smears
doi:10.1371/journal.pntd.0000227.t001 in a subset of the samples indicated that 18S rRNA was a stable
housekeeping gene in our samples. Sequences of primers and
probes have been obtained from Dr. Roger Lauener (TLR2,
immunotherapy. Since most of the positive reactions seen were TLR4, 18S rRNA, [16]) and from Dr. Masato Kubo (SOCS-1,
against mite allergen, we have focused on children having a SOCS-3, [23]). IgE mRNA levels were determined by a primer
positive skin test for mite allergen. and probe set specific for the CH1 region of IgE (forward primer
59-CAA TGCCACCTCCGTGACTC-39, reverse primer 59-
Detection of total and mite-specific IgE CGTCGCAGGACGACTGTAAG-39 and probe 59-ATCGTC-
CACAGACTGGGTCGACAACAAA-39). For each gene, after
Serum levels of total IgE were measured by the enzyme linked
normalisation for the housekeeping gene, the donor with the
immunosorbent assay (ELISA) as described before [26]. Results
were expressed as international units per ml (IU/ml). lowest expression was set to 1. For the expression levels of TLR2
and SOCS-3 in isolated cell subsets, expression of SOCS-3 or
Serum levels of house dust mite (HDM) antibodies were
TLR2 in T cells was set to 1 for each donor.
determined by radio allergosorbent test (RAST) as described
previously [27] (CLB, Amsterdam, The Netherlands). Results were
expressed as international units per ml (IU/ml). One IU is 2.4 ng Comparison of mRNA and surface expression of TLR2
IgE. Subjects were considered sensitised when concentrations of Whole blood from 5 donors was diluted 1:1 with RMPI 1640
specific IgE of more than 0.7 IU/ml were measured. medium (Gibco) and stimulated for 16 hours and 24 hours in 96-
well round bottom plates with medium, 10 mg/ml SEA (schisto-
RNA isolation from whole blood somal egg antigens), 100 ng/ml LPS (Sigma), 100 mg/ml poly I:C
Immediately after venapuncture into heparinised tubes, 0.8 ml or 5 mg/ml TNF-a (Sanquin, The Netherlands). After 16 hours
of whole blood was added to 3.6 ml of Nuclisens lysis buffer the blood was mixed with ABI Lysis buffer (Applied Biosystems)
(Biomérieux, Boxtel, The Netherlands) to stabilise the RNA. and RNA was extracted using the ABI6100 according to their
Samples were stored for a maximum of two weeks at 4uC, after protocol (Applied Biosystems). cDNA synthesis and quantitative
which they were put at 280uC for long-term storage. The PCR for TLR2 was performed as described above. 24 hours after
Nuclisens Isolation kit (Biomérieux) was used for the isolation of stimulation, cells were mixed with FACS lysing solution (BD
total nucleic acid (approximately 1 ml of blood mixed with lysis Biosciences) to lyse the erythrocytes, washed with PBS and cells
buffer per isolation) according to the manufacturer’s instructions. stained with anti-TLR2 PE (clone T2.5, eBioscience). Flow
Genomic DNA was removed by treating the samples with RNAse- cytometric analyses were performed using a Becton Dickinson
free DNAse (Invitrogen, Breda, The Netherlands) for 30 minutes FACSCalibur flow cytometer (BD Biosciences) and analysed using
at 37uC, followed by the Nuclisens isolation procedure to isolate FlowJo analysis software (Tree Star Inc.).
the purified RNA. RNA was isolated from the same samples
before and after one year of storage at 280uC, and mRNA levels Statistical analysis
of several genes of interest were compared. There was no Association of total and mite-specific IgE with skin reactivity to
difference in gene expression indicating that mRNA was stable mite was analysed by logistic regression of log-transformed values.
in this buffer for at least one year at 280uC, and that the The association of gene expression with skin reactivity or helminth
procedure was consistent. infection was analysed by logistic regression using a value of each
www.plosntds.org 3 April 2008 | Volume 2 | Issue 4 | e227
mRNA Expression Profiles of Rural Ghanaian Children
Table 2. Determinants of a positive skin prick reactivity to 1.00], p = 0.05, adjusted for age, sex, school and levels of mite IgE).
house dust mite. Additionally, the odds ratio of the association between the log of
mite IgE and atopy is clearly lower in children infected with
helminths (OR 5.8 [1.0–33.2]; p = 0.05) compared to the odds ratio
Adjusted OR [95% CI]* P in non-infected children (OR 18.7 [2.4–145.8]; p = 0.005).
The level of mite-specific IgE was a strong determinant for the
total IgE$ 1.93 [0.78–4.75] 0.16
risk of positive skin reactivity to mite (Table 2). In contrast, the
mite IgE$ 9.90 [2.57–38.26] 0.001 level of total IgE was not significantly associated with atopy.
TLR2# 2.56 [0.79–8.29] 0.12 However, total IgE was associated with atopy in the children that
TLR4# 0.93 [0,32–2,74] 0.90 were not infected with helminths (OR = 5.6, CI 95%: 1.0–30.9;
SOCS-1# 5.85 [1.52–22.42] 0.01 p,0.05).
SOCS-3# 4.44 [1.27–15.53] 0.02
Validation of ex vivo mRNA expression profiles
*adjusted for age, sex and helminth infection In order to evaluate whether the in vivo status of the immune
$log-transformed system could be evaluated by analysing the mRNA expression in
#high expression of indicated genes defined as levels above the geomean of all
whole blood, RNA was isolated from peripheral blood samples
samples tested (3.2 for TLR2; 5.0 for TLR4;3.6 for SOCS-1; 6.0 for SOCS-3)
doi:10.1371/journal.pntd.0000227.t002 collected in our study population. In agreement with IgE serum
levels, the levels of IgE mRNA were significantly higher in helminth-
gene to separate high and low gene expression, since the infected children compared to non-infected children (Figure 1A),
association between gene expression and allergen reactivity might and a high correlation was seen between the mRNA expression and
not be a linear one. This value was based on the geomean of the serum IgE levels (r = 0.58, p,0.001; Figure 1B). In addition the
relative expression data (low expression: below geomean; high mRNA levels of TLR2 were compared to surface TLR2 protein
expression: above geomean). The regression analysis was per- expression using flow cytometry in whole blood samples. There was
formed adjusting for age, sex and helminth infection (Table 2) or a strong correlation between TLR2 mRNA expression and TLR2
for age, sex and skin reactivity to mite (Table 3). The comparison surface expression (r = 0.58; p,0.001), validating the use of mRNA
of the non-adjusted means of gene expression between skin prick levels measured in whole blood samples as a reflection of the
positive and negative children and between helminth-infected and measured immunological events in vivo.
non-infected children was determined by the non-parametric
Mann-Whitney test. Correlation between mRNA and surface The association of TLR2, SOCS-1 and SOCS-3 mRNA
TLR2 levels and between mRNA and serum IgE levels was expression and skin reactivity to mite allergen
compared using the non-parametric Spearman’s correlation test. In Ghanaian school children living in an area highly endemic
for parasitic infections, there was a significantly higher expression
Results of TLR2 in subjects with positive skin reactivity to house dust mite
(Figure 2A); high expression of TLR2 doubled the risk of atopy
Characteristics of the study population in rural Ghana (OR 2.6, Table 2), whereas there was no such association for
The study population here originates from two schools selected TLR4 and skin reactivity (OR 0.9; Table 2 and Figure 2B).
from a large study on allergy and parasitic infections. The schools Children with high expression of SOCS-1 or SOCS-3 had a
were in a rural area highly endemic for helminth infections (see significantly increased risk for skin reactivity (OR 5.8 and 4.4,
Material and Methods section). Fifty-four percent of the children respectively, Table 2). Both SOCS-1 and SOCS-3 expression were
were infected with at least one helminth species (Table 1). As significantly elevated in those who were skin prick positive to house
indicated in Table 1, the most prevalent helminth species was dust mite compared to non-atopic children (Figures 2C and 2D).
Schistosoma haematobium, followed by hookworm, Ascaris lumbricoides To determine the source of TLR2 and SOCS-3 expression in
and Trichuris trichiura. In the population where mRNA expression whole blood samples, monocytes and T cells were isolated from
was analyzed (see Materials and Methods), 14 out of 74 schistosome five donors. Although the mRNA expression of TLR2 was high in
negative children had a positive skin reaction to mite (19%), monocytes (63 to 275-fold higher than in T cells, Figure 3A),
whereas 5 out of 46 schistosome-infected children were SPT correction for the percentages of monocytes and T cells in the
positive for mite (11%) resulting in a significant negative association peripheral blood mononuclear cells, indicated that monocytes, T
between infection with S. haematobium and atopy (OR 0.26 [0.07– cells and other cells contributed similarly to the TLR2 expression
Table 3. Association between helminth infection and gene expression.
TLR2# TLR4# SOCS-1# SOCS-3#
OR* P OR* P OR* P OR* P
Helminth 0.26 0.001 0,40 0.024 1.38 0.42 0.29 0.003
[0.11–0.60] [0.18–0,89] [0.63–3.01] [0.13–0.66]
Schistosomes 0.24 0.001 0.56 0.15 1,12 0.77 0.35 0.010
[0.11–0.55] [0.26–1.24] [0.51–2.46] [0.16–0.78]
*adjusted for age, sex, and skin reactivity to mite
#high expression of indicated genes defined as levels above the geomean of all samples tested (3.2 for TLR2; 5.0 for TLR4;3.6 for SOCS-1; 6.0 for SOCS-3)
doi:10.1371/journal.pntd.0000227.t003
www.plosntds.org 4 April 2008 | Volume 2 | Issue 4 | e227
mRNA Expression Profiles of Rural Ghanaian Children
Figure 1. Validation of ex vivo mRNA expression profiles. (A) Relative IgE mRNA expression in blood of helminth-infected (helminth pos) and
non-infected (helminth neg) individuals. Horizontal bars represent median values per group. (B) Scatter plot showing correlation between levels of
IgE mRNA and serum levels of IgE. (C) Scatter plot showing correlation between TLR2 mRNA expression in whole blood measured by real-time PCR
(16 hours after stimulation) and surface expression of TLR2 by flow cytometry in same samples (24 hours after stimulation).
doi:10.1371/journal.pntd.0000227.g001
measured in whole blood (Figure 3C). In contrast, the mRNA The analysis of the expression levels of SOCS-1 and SOCS-3
expression of SOCS-3 could clearly be attributed to the T cell revealed that children infected with helminths had significantly
fraction with little contribution from monocytes or other cells lower expression levels of SOCS-3, but not of SOCS-1 (Figures 4C
(Figure 3B–C). and 4D). Helminth positivity in a child was associated with low
gene expression of SOCS-3, but not of SOCS-1 (Table 3)). As for
Expression of TLR2 and SOCS-3 mRNA is lower in TLR2, low expression of SOCS-3 was associated with S.
helminth-infected children haematobium rather than hookworm infection.
TLR expression can be altered following exposure to ligands
expressed by microorganisms and parasites. Helminth parasites Discussion
carry signature molecules that can interact with TLRs and therefore
Using gene expression profiles in whole blood from children
could affect their expression. The expression of both TLR2 and
TLR4 genes was lower in children with a helminth infection; the living in a rural area in Ghana, we found that high expression of
effect being more prominent for TLR2 (Figures 4A and 4B). Indeed, TLR2, SOCS-1 and SOCS-3 mRNA was associated with positive
infection with helminths predicted low expression of TLR2 and, to a skin reactivity to house dust mite. Presence of Schistosoma
lesser extent, of TLR4 (Table 3). Two major helminth species haematobium infection, reported to decrease the risk of atopy [7]
prevalent in the study area were Schistosoma haematobium and and observed in the current study, affected the expression levels of
hookworm. The mRNA expression of TLR2 in helminth-infected TLR2 and SOCS-3, which were significantly lower in infected
children was significantly lower only in children infected with S. children.
haematobium, which was associated with low expression of TLR2 There are few studies that have looked at the association of
(Table 3), whereas no such association was found for TLR4. Malaria TLR expression and allergy, and those that have, are all in
infection was associated with higher levels of TLR2 expression (not European populations [29,30]. Of these, only one study has
shown), and adjustment for malaria infection did not change the investigated the levels of TLR expression in an age group
association between helminth infection and TLR2 expression. Thus, comparable to our study. European children living on farms and
induction of lower expression of TLR2 by infection with reported to have a lower risk of developing atopic disorders were
schistosomes seems to be specific. shown to have higher expression levels of TLR2 and CD14,
www.plosntds.org 5 April 2008 | Volume 2 | Issue 4 | e227
mRNA Expression Profiles of Rural Ghanaian Children
Figure 2. Relative gene expression levels of TLR2, TLR4, SOCS-1 and SOCS-3 in skin test positive and negative children. Relative
mRNA expression levels of TLR2 (A), TLR4 (B), SOCS-1 (C) and SOCS-3 (D) in blood of children positive (spt pos) or negative (spt neg) for skin reactivity
to house dust mite. Horizontal bars represent median values per group. * p,0.05.
doi:10.1371/journal.pntd.0000227.g002
compared to non-farmer children [16]. As farmer children would or by stimulation of TLR and limit the production of
be expected to be exposed to a high burden of environmental inflammatory cytokines [33]. In murine models, transgenic over
microorganisms, the results are in contrast to the lower expression expression of SOCS-3 has been shown to mediate and maintain
of TLR2 in our helminth infected subjects compared to uninfected allergic responses [23]. Furthermore, T cell expression of SOCS-3,
Ghanaian school children. Children living in a rural area in but not of SOCS-1, was associated with atopic disease in humans
Ghana are expected to have exposures that are higher in intensity, and increased with disease severity [23]. These results suggest that
and different in nature in terms of the sort of microorganisms and SOCS-3 is involved in Th2 skewed responses. However, although
parasites, compared to European farmer children. Moreover, the helminth infections are clearly associated with Th2 responses, we
European study did not examine the TLR2 levels in atopic and have found that SOCS-3 expression is decreased in helminth-
non-atopic individuals as we do here, showing that atopy was infected children. Cell subset analysis indicated that T cells were
associated with high expression of TLR2. the main source of SOCS-3 mRNA leading us to conclude that in
Our data support a role for suppression of atopy by current helminth infected children, with strong Th2 responses, SOCS-3
infection with a systemic helminth, Schistosoma haematobium. The expression is low in T cells. So although both allergic disorders and
finding that S. haematobium infected children show a lower helminth infections are characterized by Th2 responses, SOCS-3
expression of TLR2 gene, is supported by the results that baseline is associated with allergic disorders, but not with helminth
expression levels of TLR2 protein were also shown to be lower in infection. This would suggest that in allergic subjects, with high
individuals infected with another systemic helminth infection, the expression of SOCS-3, Th2 responses are associated with
filarial nematode, Wuchereria bancrofti [31,32]. Importantly, lower inflammation; whereas in allergic subjects with a helminth
expression of TLR correlated with a lower expression of co- infection and consequent low expression of SOCS-3, Th2 cells
stimulatory molecules such as CD80 and CD86 and lower are not associated with inflammation. Interestingly, a recent report
production of the inflammatory cytokines IL-6 and TNF-a [32]. using T cell-specific SOCS-3 conditional knockout mice indicated
Our results also indicated an association of skin reactivity to that in the absence of SOCS-3 expression, the levels of typical Th2
house dust mite with higher gene expression for both SOCS-1 and cytokines in peripheral CD4+ T cells were either unaffected (IL-5)
SOCS-3. SOCS genes are involved in the pathogenesis of several or only slightly lower (IL-4), whereas following T cell stimulation,
inflammatory diseases. They are induced upon cytokine signalling the production of the anti-inflammatory cytokines, IL-10 and
www.plosntds.org 6 April 2008 | Volume 2 | Issue 4 | e227
mRNA Expression Profiles of Rural Ghanaian Children
Figure 3. Expression of TLR2 and SOCS-3 mRNA in cell subsets. Mean of relative mRNA expression of TLR2 (A) and SOCS-3 (B) in T cells (CD3),
monocytes (CD14) and cells depleted of T cells and monocytes (rest) from 5 donors. Y error bars indicate the standard error of the mean from each
group. (C) Mean expression levels of TLR2 and SOCS-3 in T cells, monocytes and the rest of the cells corrected for the percentages of cells within the
PBMC population for each donor.
doi:10.1371/journal.pntd.0000227.g003
TGF-b1, were significantly higher. The abolition of SOCS-3 the cytokine environment might influence the expression of TLR.
expression in T cells ameliorated ovalbumin-induced airway Th1 cytokines such as IFN-c seems to increase expression levels of
hyperresponsiveness in vivo [34]. Furthermore, CD4+CD25+foxp3 TLR [41], whereas Th2 cytokines as IL-4 and IL-13, abundantly
positive regulatory T cells were shown to have low SOCS-3 present in helminth infected individuals, downregulate TLR
expression as compared to Th2 cells, indicating that low SOCS-3 expression and function [42,43]. Thus, the exposure of European
expression in T cells is associated with suppressive function [35]. farmers to Th1 inducing agents might be reflected in higher TLR2
These data raise the possibility that a decrease in SOCS-3 T cell expression, whereas in our subjects the exposure to Th2 inducing
expression by helminth infection might shift the balance towards a agents might lead to low TLR2 expression. The relationship
modified Th2 response [36] with a more anti-inflammatory between TLR2 and SOCS-3 expression might not be a direct one.
function, and thereby might suppress allergic inflammation [1]. In rural Ghana, helminth infection is associated with low TLR2 as
As malaria infection was prevalent in our study area, we looked well as low SOCS-3 expression, whereas the expression of TLR2 is
at this infection and found that it had no effect on atopy high in a European rural area. If TLR2 expression is merely the
(multivariate analysis, data not shown) and interestingly found that result of exposure to pathogens, and low SOCS-3 expression is
malaria infection was associated with a higher expression of TLR2 associated with protection from allergy, it would be of interest to
(data not shown), indicating that different pathogens might induce know whether SOCS-3 is also lower in the European farmers’
different regulation of TLR expression. Indeed, other protozoa environment despite a higher TLR2 expression.
such as Entamoeba histolytica and Trypanosoma cruzi inhibit immune In summary, ex vivo whole blood analysis of mRNA profiles in
responses by down-regulating TLR2 expression [37] or signalling children infected with helminths compared to non-infected
via TLR2 [38,39]. There are numerous studies supporting either children living in a rural area in Ghana have shown that chronic
up- or downregulation of TLR2 and TLR4, depending on the helminth infections are associated with a lower expression of
stimulus and cell type studied [40]. Both an increase and a TLR2 and SOCS-3. The difference in the expression of SOCS-3
decrease in TLR2 expression might reflect repeated TLR in helminth infected and uninfected children might result from the
stimulation, the direction as well as the downstream signaling interaction of helminth derived molecules with the immune system
pathways being dependent on the type of pathogen. Alternatively, leading to modulation of downstream signalling and induction of
www.plosntds.org 7 April 2008 | Volume 2 | Issue 4 | e227
mRNA Expression Profiles of Rural Ghanaian Children
Figure 4. Relative gene expression levels of TLR2, TLR4, SOCS-1 and SOCS-3 in helminth-infected and helminth free children.
Relative mRNA expression levels of TLR2 (A), TLR4 (B), SOCS-1 (C) and SOCS-3 (D) in blood of children positive (helminth pos) or negative (helminth
neg) for helminth infection. * p,0.05; ** p,0.01.
doi:10.1371/journal.pntd.0000227.g004
‘‘modified’’ Th2 cells. Larger epidemiological studies will be We are indebted to the children for their participation in this study. We
needed to be able to test this hypothesis directly and to confirm thank HAL Allergy (Haarlem, The Netherlands) for providing the allergens
that helminths modify the development of allergy by modulating for the skin reactivity tests.
the expression levels of TLR2 and SOCS-3.
Author Contributions
Acknowledgments Conceived and designed the experiments: FH MW LR DB MY.
Performed the experiments: FH BO YK MD AdB AA IL RvR. Analyzed
We thank all field workers from the Noguchi Memorial Institute for
the data: FH BO MD RvR LR MY. Wrote the paper: FH LR MY.
Medical Research who were involved in the recruitment and field work.
References
1. Maizels RM, Yazdanbakhsh M (2003) Immune regulation by helminth parasites: 8. Araujo MI, Lopes AA, Medeiros M, Cruz AA, Sousa-Atta L, Sole D,
cellular and molecular mechanisms. Nat Rev Immunol 3: 733–744. Carvalho EM (2000) Inverse association between skin response to aeroallergens
2. Sabin EA, Araujo MI, Carvalho EM, Pearce EJ (1996) Impairment of tetanus and Schistosoma mansoni infection. Int Arch Allergy Immunol 123: 145–148.
toxoid-specific Th1-like immune responses in humans infected with Schistosoma 9. Lynch NR, Hagel I, Perez M, Di Prisco MC, Lopez R, Alvarez N (1993) Effect
mansoni. J Infect Dis 173: 269–272. of anthelmintic treatment on the allergic reactivity of children in a tropical slum.
3. Cooper PJ, Espinel I, Paredes W, Guderian RH, Nutman TB (1998) Impaired J Allergy Clin Immunol 92: 404–411.
tetanus-specific cellular and humoral responses following tetanus vaccination in 10. van Den Biggelaar AH, Rodrigues LC, van Ree R, van der Zee JS, Hoeksma-
human onchocerciasis: a possible role for interleukin-10. J Infect Dis 178: Kruize YC, Souverijn JH, Missinou MA, Borrmann S, Kremsner PG,
1133–1138. Yazdanbakhsh M (2004) Long-term treatment of intestinal helminths
4. Smits HH, Hartgers FC, Yazdanbakhsh M (2005) Helminth infections: increases mite skin-test reactivity in Gabonese schoolchildren. J Infect Dis 189:
Protection from atopic disorders. Current Allergy and Asthma Reports 5: 42–50. 892–900.
5. Flohr C, Tuyen LN, Lewis S, Quinnell R, Minh TT, Liem HT, Campbell J, 11. Cooper PJ, Chico ME, Vaca MG, Moncayo AL, Bland JM, Mafla E, Sanchez F,
Pritchard D, Hien TT, Farrar J, Williams HC, Britton J (2006) Poor sanitation Rodrigues LC, Strachan DP, Griffin GE (2006) Effect of albendazole treatments
and helminth infection protect against skin sensitization in Vietnamese children: on the prevalence of atopy in children living in communities endemic for
A cross-sectional study. American Journal of Tropical Medicine and Hygiene 75: geohelminth parasites: a cluster-randomised trial. Lancet 367: 1598–1603.
321. 12. Braun-Fahrlander C, Riedler J, Herz U, Eder W, Waser M, Grize L, Maisch S,
6. Scrivener S, Yemaneberhan H, Zebenigus M, Tilahun D, Girma S, Ali S, Carr D, Gerlach F, Bufe A, Lauener RP, Schierl R, Renz H, Nowak D, von
McElroy P, Custovic A, Woodcock A, Pritchard D, Venn A, Britton J (2001) Mutius E (2002) Environmental exposure to endotoxin and its relation to asthma
Independent effects of intestinal parasite infection and domestic allergen in school-age children. N Engl J Med 347: 869–877.
exposure on risk of wheeze in Ethiopia: a nested case-control study. Lancet 13. Velasco G, Campo M, Manrique OJ, Bellou A, He H, Arestides RS, Schaub B,
358: 1493–1499. Perkins DL, Finn PW (2005) Toll-like receptor 4 or 2 agonists decrease allergic
7. van Den Biggelaar AH, van Ree R, Rodrigues LC, Lell B, Deelder AM, inflammation. Am J Respir Cell Mol Biol 32: 218–224.
Kremsner PG, Yazdanbakhsh M (2000) Decreased atopy in children infected 14. Eisenbarth SC, Piggott DA, Huleatt JW, Visintin I, Herrick CA, Bottomly K
with Schistosoma haematobium: a role for parasite-induced interleukin-10. (2002) Lipopolysaccharide-enhanced, Toll-like Receptor 4-dependent T Helper
Lancet 356: 1723–1727. Cell Type 2 Responses to Inhaled Antigen. J Exp Med 196: 1645–1651.
www.plosntds.org 8 April 2008 | Volume 2 | Issue 4 | e227
mRNA Expression Profiles of Rural Ghanaian Children
15. Kane CM, Cervi L, Sun J, McKee AS, Masek KS, Shapira S, Hunter CA, is there an influence of maternal allergy? J Allergy Clin Immunol 115:
Pearce EJ (2004) Helminth antigens modulate TLR-initiated dendritic cell 1304–1310.
activation. J Immunol 173: 7454–7461. 30. Krauss-Etschmann S, Hartl D, Heinrich J, Thaqi A, Prell C, Campoy C,
16. Lauener RP, Birchler T, Adamski J, Braun-Fahrlander C, Bufe A, Herz U, von Molina FS, Hector A, Decsi T, Schendel DJ, Koletzko BV (2006) Association
Mutius E, Nowak D, Riedler J, Waser M, Sennhauser FH (2002) Expression of between levels of Toll-like receptors 2 and 4 and CD14 mRNA and allergy in
CD14 and Toll-like receptor 2 in farmers’ and non-farmers’ children. Lancet pregnant women and their offspring. Clin Immunol 118: 292–299.
360: 465–466. 31. Babu S, Blauvelt CP, Kumaraswami V, Nutman TB (2006) Cutting Edge:
17. Fujimoto M, Tsutsui H, Xinshou O, Tokumoto M, Watanabe D, Shima Y, Diminished T Cell TLR Expression and Function Modulates the Immune
Yoshimoto T, Hirakata H, Kawase I, Nakanishi K, Kishimoto T, Naka T (2004) Response in Human Filarial Infection. J Immunol 176: 3885–3889.
Inadequate induction of suppressor of cytokine signaling-1 causes systemic 32. Babu S, Blauvelt CP, Kumaraswami V, Nutman TB (2005) Diminished
autoimmune diseases. Int Immunol 16: 303–314. expression and function of TLR in lymphatic filariasis: a novel mechanism of
18. Marine JC, Topham DJ, McKay C, Wang D, Parganas E, Stravopodis D, immune dysregulation. J Immunol 175: 1170–1176.
Yoshimura A, Ihle JN (1999) SOCS1 deficiency causes a lymphocyte-dependent 33. Baetz A, Frey M, Heeg K, Dalpke AH (2004) Suppressor of cytokine signaling
perinatal lethality. Cell 98: 609–616. (SOCS) proteins indirectly regulate toll-like receptor signaling in innate immune
19. Naka T, Fujimoto M, Tsutsui H, Yoshimura A (2005) Negative regulation of cells. J Biol Chem 279: 54708–54715.
cytokine and TLR signalings by SOCS and others. Adv Immunol 87: 61–122. 34. Kinjyo I, Inoue H, Hamano S, Fukuyama S, Yoshimura T, Koga K, Takaki H,
20. Mansell A, Smith R, Doyle SL, Gray P, Fenner JE, Crack PJ, Nicholson SE, Himeno K, Takaesu G, Kobayashi T, Yoshimura A (2006) Loss of SOCS3 in T
Hilton DJ, O’Neill LA, Hertzog PJ (2006) Suppressor of cytokine signaling 1 helper cells resulted in reduced immune responses and hyperproduction of
negatively regulates Toll-like receptor signaling by mediating Mal degradation. interleukin 10 and transforming growth factor-beta1. J Exp Med 203:
Nat Immunol 7: 148–155. 1021–1031.
21. Suzuki A, Hanada T, Mitsuyama K, Yoshida T, Kamizono S, Hoshino T,
35. Pillemer BBL, Xu H, Oriss TB, Qi Z, Ray A (2007) Deficient SOCS3 expression
Kubo M, Yamashita A, Okabe M, Takeda K, Akira S, Matsumoto S,
in CD4+CD25+FoxP3+ regulatory T cells and SOCS3-mediated suppression of
Toyonaga A, Sata M, Yoshimura A (2001) CIS3/SOCS3/SSI3 plays a negative
Treg function. Eur J Immunol 37: 2082–2089.
regulatory role in STAT3 activation and intestinal inflammation. J Exp Med
36. Platts-Mills T, Vaughan J, Squillace S, Woodfolk J, Sporik R (2001)
193: 471–481.
Sensitisation, asthma, and a modified Th2 response in children exposed to cat
22. Shouda T, Yoshida T, Hanada T, Wakioka T, Oishi M, Miyoshi K, Komiya S,
allergen: a population-based cross-sectional study. Lancet 357: 752–756.
Kosai K, Hanakawa Y, Hashimoto K, Nagata K, Yoshimura A (2001) Induction
37. Maldonado C, Trejo W, Ramirez A, Carrera M, Sanchez J, Lopez-Macias C,
of the cytokine signal regulator SOCS3/CIS3 as a therapeutic strategy for
Isibasi A (2000) Lipophosphopeptidoglycan of Entamoeba histolytica induces an
treating inflammatory arthritis. J Clin Invest 108: 1781–1788.
antiinflammatory innate immune response and downregulation of toll-like
23. Seki Y, Inoue H, Nagata N, Hayashi K, Fukuyama S, Matsumoto K, Komine O,
receptor 2 (TLR-2) gene expression in human monocytes. Arch Med Res 31:
Hamano S, Himeno K, Inagaki-Ohara K, Cacalano N, O’Garra A, Oshida T,
Saito H, Johnston JA, Yoshimura A, Kubo M (2003) SOCS-3 regulates onset S71–S73.
and maintenance of T(H)2-mediated allergic responses. Nat Med 9: 1047–1054. 38. Brodskyn C, Patricio J, Oliveira R, Lobo L, Arnholdt A, Mendonca-Previato L,
24. Ozaki A, Seki Y, Fukushima A, Kubo M (2005) The control of allergic Barral A, Barral-Netto M (2002) Glycoinositolphospholipids from Trypanosoma
conjunctivitis by suppressor of cytokine signaling (SOCS)3 and SOCS5 in a cruzi interfere with macrophages and dendritic cell responses. Infect Immun 70:
murine model. J Immunol 175: 5489–5497. 3736–3743.
25. Van der Kleij D, van Den Biggelaar AH, Kruize YC, Retra K, Fillie Y, 39. Ropert C, Gazzinelli RT (2004) Regulatory role of Toll-like receptor 2 during
Schmitz M, Kremsner PG, Tielens AG, Yazdanbakhsh M (2004) Responses to infection with Trypanosoma cruzi. J Endotoxin Res 10: 425–430.
Toll-like receptor ligands in children living in areas where schistosome infections 40. Fan H, Cook JA (2004) Molecular mechanisms of endotoxin tolerance.
are endemic. J Infect Dis 189: 1044–1051. J Endotoxin Res 10: 71–84.
26. van Den Biggelaar AH, Lopuhaa C, van Ree R, van der Zee JS, Jans J, Hoek A, 41. Bosisio D, Polentarutti N, Sironi M, Bernasconi S, Miyake K, Webb GR,
Migombet B, Borrmann S, Luckner D, Kremsner PG, Yazdanbakhsh M (2001) Martin MU, Mantovani A, Muzio M (2002) Stimulation of toll-like receptor 4
The prevalence of parasite infestation and house dust mite sensitization in expression in human mononuclear phagocytes by interferon-gamma: a
Gabonese schoolchildren. Int Arch Allergy Immunol 126: 231–238. molecular basis for priming and synergism with bacterial lipopolysaccharide.
27. Aalberse RC, Koshte V, Clemens JG (1981) Immunoglobulin E antibodies that Blood 99: 3427–3431.
crossreact with vegetable foods, pollen, and Hymenoptera venom. J Allergy Clin 42. Mita Y, Dobashi K, Endou K, Kawata T, Shimizu Y, Nakazawa T, Mori M
Immunol 68: 356–364. (2002) Toll-like receptor 4 surface expression on human monocytes and B cells is
28. User Bulletin #2 (2005) ABI Prism 7700 Sequence Detection System. P/N modulated by IL-2 and IL-4. Immunol Lett 81: 71–75.
4303859 Rev. A, Stock No. 777802-001. 43. Mueller T, Terada T, Rosenberg IM, Shibolet O, Podolsky DK (2006) Th2
29. Amoudruz P, Holmlund U, Malmstrom V, Trollmo C, Bremme K, Scheynius A, cytokines down-regulate TLR expression and function in human intestinal
Sverremark-Ekstrom E (2005) Neonatal immune responses to microbial stimuli: epithelial cells. J Immunol 176: 5805–5814.
www.plosntds.org 9 April 2008 | Volume 2 | Issue 4 | e227