SEARCH

SEARCH BY CITATION

Keywords:

  • birth cohort;
  • cord blood;
  • environmental exposure;
  • FOXP3;
  • regulatory T cells.

Abstract

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. Authors’ contribution
  8. Conflict of interest
  9. References
  10. Supporting Information

Background:

Regulatory T cells (Tregs) with stable FOXP3 expression are characterized by a specific demethylated region in the FOXP3 gene (Treg-specific demethylated region, TSDR). The aim of this study was to analyse the influence of prenatal factors on cord blood Treg numbers, as detected by changes in the TSDR demethylation, and the subsequent risk for allergic diseases.

Methods:

Analyses were performed within the LINA study in blood samples from pregnant women (34th gestational week) and in cord blood (n = 346 mother–child pairs). Treg numbers were detected via DNA demethylation in the FOXP3 TSDR. At age 1, total and specific IgE was measured in children's blood. In addition, maternal cytokine production (Th1/Th2/Th17) was analysed. Exposure and disease outcomes were assessed by questionnaires.

Results:

Boys had lower Treg numbers compared with girls (< 0.001). Parental atopy history, particularly maternal hay fever and paternal asthma were related to lower Treg numbers in cord blood (adj. MR = 0.81, 95% CI = 0.68–0.97; adj. MR = 0.60, 95% CI = 0.45–0.81). Maternal cytokines (IL-13, IL-17E and IFN-γ) and maternal smoking/exposure to tobacco smoke during pregnancy were also associated with decreased cord blood Treg numbers (adj. MR = 0.89, 95% CI = 0.97–1.00). Children with lower Treg numbers at birth had a higher risk to develop atopic dermatitis (adj. OR = 1.55, 95% CI = 1.00–2.41) and sensitization to food allergens (adj. OR = 1.55, 95% CI = 1.06–2.25) during the first year of life.

Conclusions:

These results indicate that both genetic and environmental factors presumably influence the development of foetal Tregs. Low cord blood Treg numbers may predict early atopic dermatitis.

As allergic diseases start to develop very early in life, it has been proposed that genetic predisposition as well as multiple prenatal factors [e.g. in utero microenvironment, maternal smoking and exposure to environmental tobacco smoke (ETS), diet] influence the in utero programming of the foetal immune system and therefore the susceptibility to develop allergic diseases [1-3]. Regulatory T cells (Tregs) play an essential role in immune regulation of atopic diseases [4]. It has been hypothesized that in utero induced impairment in these regulatory pathways may play a key role in providing the basis for an allergy skewed phenotype. However, data concerning the relevance of in utero events for the imprinting of Tregs and children's atopic predisposition is still limited and contradictory. It has been suggested that Tregs are impaired in neonates with hereditary allergy risk [5, 6]. In contrast, Rindsjö et al. [7] did not detect an effect of maternal allergic disease on Treg subsets in cord blood. These inconsistent results might be due to difficulties in characterizing Tregs in large birth cohort studies; further studies are indispensable to evaluate the relevance of in utero events for cord blood Tregs and allergic predisposition.

In humans, Tregs are best characterized by the expression of the transcription factor FOXP3, which is essential for Treg induction and stability [8]. However, FOXP3 is also expressed transiently in conventional CD4+ T cells upon activation [9-11]. Recently, it has been shown that both induction and stabilization of FOXP3 expression is under epigenetic control [12-14]. Sequence analyses revealed a highly conserved noncoding region in the FOXP3 locus, the Treg-specific demethylated region (TSDR). Demethylation of the TSDR corresponds with the stability of FOXP3 expression in CD4+CD25hi Tregs, whereas activated conventional human T cells expressing FOXP3 only transiently show a fully methylated TSDR [12]. Thus, the demethylation state of the TSDR in the FOXP3 gene has been suggested to be a valuable biomarker which can reliably detect and quantify stable CD4+CD25hi Tregs in adults as well as in cord blood in large field studies [12, 15, 16].

The objective of the present study was to identify prenatal factors influencing Treg numbers at birth and evaluate the relevance of cord blood Treg numbers for allergic outcomes of the child during the first year of life. Maternal atopy might influence the development of foetal Tregs either directly or via maternal immune status during pregnancy. Furthermore, environmental exposure during pregnancy might influence the maternal immune status or directly modulate Treg numbers at birth. We therefore hypothesize that maternal/paternal atopy, maternal immune status during pregnancy and/or environmental exposure may influence the Treg number at birth with a potential relevance for the allergy risk in infancy.

Methods

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. Authors’ contribution
  8. Conflict of interest
  9. References
  10. Supporting Information

For more detailed information, please see the methods section of Appendix S1.

Study design

Six-hundred and twenty-two mother–child pairs were recruited within the prospective birth cohort study LINA (Lifestyle and environmental factors and their Influence on Newborns Allergy risk) from March 2006 until December 2008 in Leipzig, Germany. Blood samples were obtained during pregnancy (mother, 34th week of gestation), at birth (venous umbilical cord blood) and on the child's first birthday. Data on confounding variables, prenatal exposure, life style factors and children's disease outcomes were obtained from questionnaires filled in by the parents 4 weeks before birth and on the child's first birthday. The present investigation comprises of only study participants who have received a complete blood analysis at pregnancy/birth and at the age of one (n = 346, greatest common subset, see Fig. S1 and methods/study design of Appendix S1). Participation was voluntary, and informed consent was given by the parents. This study was approved by the Ethics Committees of the University of Leipzig (file reference 046-2006).

Parental history of atopy, prenatal exposure and outcomes in the first year of life

Information on parental history of atopy and prenatal exposure was assessed by detailed questionnaires. To address atopy history, we included asthma, hay fever and atopic dermatitis (‘Did you ever suffer from asthma, allergic rhinitis, atopic dermatitis?’, ‘Did your partner ever suffer from asthma, allergic rhinitis, atopic dermatitis?’). Exposure variables such as smoking/exposure to ETS at home (‘Did you or anybody else smoke inside your dwelling during the last twelve months?’), any renovation activities (‘Did you paint, install new furniture or change your flooring either before conception, during first to 13th gestational week, 14th to 26th gestational week or 27th to 40th gestational week?’), the usage of chemicals (‘How often did you use disinfectants or air freshener during pregnancy?’), and traffic (‘How is the distance between your dwelling and a main street, lower than ten metres or above 150 metre?’) were considered.

Information on children's allergic outcomes was obtained from questionnaires. We included physician-diagnosed atopic dermatitis (‘Has a doctor diagnosed your child with atopic dermatitis in the last twelve months?’) and cradle cap (‘Has a doctor diagnosed your child with cradle cap in the last twelve months?’). Furthermore, symptoms of atopic dermatitis observed by the parents were assessed (‘Did your child suffer from any symptoms of atopic dermatitis in the last twelve months?’).

Whole blood assay

Blood samples from mother–child pairs were processed within 6 h for further analysis. For cytokine analyses, whole blood samples were stimulated ex-vivo with phytohemagglutinin (PHA, 50 μg/ml; Sigma Aldrich, Munich, Germany) or lipopolysaccharide (LPS, 1 μg/ml, E. coli 026:B6; Sigma Aldrich). After incubating for 4 h at 37°C, samples were diluted with RPMI-1640 medium without supplements in a ratio of 1 : 1 and centrifuged. Cell-free supernatants were collected and stored at −80°C until subsequent analysis.

Cytokine measurement

Concentrations of IL-4, IL-5, IL-6, IL-8, IL-10, IL-13, MCP-1 (CCL2), TNF-α and IFN-γ in the supernatants of stimulated whole blood samples were detected by flow cytometry using the BD CBA Human Soluble Flex Set system (BD Bioscience, Heidelberg, Germany) according to the manufacturer's instructions and as described previously [17]. Detection limits for each cytokine were 3 pg/ml, excepting IL-5, where the detection limit was 2 pg/ml. IL–17A and IL-17E concentrations were measured by ELISA according to the manufacturer's instructions (eBiosciences, Frankfurt, Germany and ANTIGENIX AMERICA Inc, New York, NY, USA, respectively). Detection limit was 4 pg/ml for IL-17A and 7 pg/ml for IL-17E. Cytokine concentrations below the detection limit were assigned a value that was half of the detection limit and were included in the analysis.

Immunoglobulin E measurement

The concentrations of total IgE and specific IgE against food (Fx5) and inhalant allergens (Sx1) in sera of 1-year-old children were determined by the Phadia CAP System (Phadia GmbH, Freiburg, Germany). The allergen multipanel Fx5 consists of hen's egg, cow's milk, wheat, fish, peanut and soy. Sx1 includes timothy, rye, mugwort, birch, house dust mite (Dermatophagoides pteronyssinus), cat and dog. Samples with a specific IgE concentration >0.35 kU/l and with total IgE concentration >3.8 kU/l were regarded as positive, according to published cut off values for 1-year-old children [18].

Isolation of DNA

Genomic DNA was isolated from whole blood using the DNA Blood Mini Kit (Qiagen, Hilden, Germany) according to the manufacturer's instructions. Briefly, whole blood was treated with 20 μl Proteinase K (activity: 600 mAU/ml solution; Qiagen) and incubated with Buffer AL in a ratio of 1 : 1 (10 min, 56°C, 350 rpm).

Cell lysis was stopped by adding 200 μl of ethanol. The complete sample volume was transferred onto a QIAamp Mini spin column and centrifuged (17 540 g). DNA bound to the QIAamp membrane was washed two times using buffers AW1 and AW2 (17 540 g, 1 min; 20 590 g, 5 min). DNA was eluted by adding buffer AE and stored at 4°C until subsequent analysis.

DNA bisulfite conversion and FOXP3 methylation-specific real-time PCR

Bisulfite treatment of genomic DNA was conducted using EpiTect 96 Bisulfite Kit (Qiagen), as described by the manufacturer. In brief, 1 μg of genomic DNA, 85 μl of the Bisulfite Mix and 35 μl of the DNA Protect Buffer were mixed in a 96-well EpiTect Conversion Plate, and bisulfite conversion was performed in a thermal cycler according to the manufacturers’ instructions. Converted DNA was stored at −20°C until further analysis.

Quantification of demethylation in TSDR was performed by real-time PCR by Epiontis (Berlin, Germany; for details see Appendix S1). As described previously, the number of Tregs in cord blood is presented as percentage corresponding to the measured amount of TSDR demethylation in the FOXP3 gene [15, 16].

Statistical analysis

Statistical tests were performed using Statistica for Windows Version 8.0 [StatSoft Inc. (Europe), Hamburg, Germany]. The chi-squared test for cross relationship was performed to ensure the equal distribution of parameters in the analysed subcohort and the entire LINA cohort. Values for Treg numbers and cytokine concentrations were not normally distributed, implying the usage of nonparametric tests. First, bivariate analyses for the relationship between gender, parental atopy (defined as occurrence of asthma or atopic dermatitis or hay fever), environmental exposure during pregnancy and Treg numbers at birth (Mann–Whitney U-test) as well as between maternal cytokine production during pregnancy and cord blood Treg numbers were performed. Second, these relationships were verified using linear regression models with log-transformed Treg values adjusted for every single factor found to be significantly correlated with Treg numbers (calculation of adjusted mean ratios, adj MR). Each exposure variable in pregnancy was first tested individually as an influencing variable with children's cord blood Treg numbers being the outcome variable. In addition, we also tested all factors possibly influencing cord blood Treg numbers together in one regression model. Third, adjusted odds ratios (OR) were calculated for the relationship between cord blood Treg numbers and allergic outcomes in the first year of life. According to the results of bivariate analysis, either maternal atopic dermatitis or parental history of atopy was selected to confound this relationship. Although we did not observe a bivariate relationship between smoking/ETS exposure, gender, parental educational level and allergic outcomes, we decided to include these confounding variables because of their already described impact on allergy risk. Regarding atopic outcomes, analyses were comprised of the occurrence of atopic dermatitis and cradle cap, if diagnosed by a physician during the first year of life. Furthermore, the questionnaire reported symptoms of atopic dermatitis were considered. All P values <0.05 were considered to be significant. Due to the fact that we investigated a priori hypotheses, adjustment for multiple testing (Bonferroni correction) was not performed [19, 20].

Results

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. Authors’ contribution
  8. Conflict of interest
  9. References
  10. Supporting Information

Characteristics of the study population

Complete data, including methylation analysis of the FOXP3 locus in cord blood, maternal cytokine production during pregnancy and IgE measurements in children's blood samples that were 1 year of age, were available from 346 mother–child pairs (see Fig. S1; results, Appendix S1). Characteristics of the study population are listed in Table 1. There were no differences in the distribution of considered parameters in the analysed subcohort compared with the entire LINA cohort (Table 1). It is noteworthy that 64.7% of parents had a history of atopy, and 10.4% of children suffered from atopic dermatitis during the first year of life.

Table 1. Characteristics of the analysed subcohort and the entire LINA cohort during pregnancy and at birth (n = 622 mother–child pairs, seven twins) and at the age of one (n = 606)
ParametersAnalysed subcohort Pregnancy and birthEntire LINA cohortP valuea
n (%), N = 346n (%), N = 622
  1. a

    P value from chi-squared test for cross relationship.

  2. b

    History of atopy is defined as occurrence of asthma or atopic dermatitis or hay fever.

  3. c

    Low, 9 years of schooling or less ‘Hauptschulabschluss’; intermediate, 10 years of schooling ‘Mittlere Reife’; high, 12 years of schooling or more ‘(Fach-)hochschulreife’.

  4. d

    Values were categorized according to the published cut-off values [18]. Concentrations >3.8 kU/l were regarded as enhanced.

  5. e

    According to the Pharmacia CAP System, concentrations >0.35 kU/l were regarded as positive.

Gender of the child
Male189 (55.0)327 (52.0)0.671
Female157 (45.0)302 (48.0)
Parental history of atopyb
Negative122 (35.3)210 (33.8)0.960
Single positive155 (44.8)291 (46.8)
Double positive69 (19.9)121 (19.4)
Maternal history of atopyb
No195 (56.4)330 (53.1)0.639
Yes151 (43.6)292 (46.9)
Paternal history of atopyb
No204 (58.9)381 (61.3)0.729
Yes142 (41.1)241 (38.7)
Parental educationc
Low7 (2.0)17 (2.7)0.947
Intermediate82 (23.7)144 (23.2)
High257 (74.3)461 (74.1)
Pet keeping during pregnancy
None199 (57.5)358 (88.0)0.376
Cat64 (18.5)113 (17.9)
Dog43 (12.4)70 (11.1)
Other87 (25.1)165 (26.2)
Smoking inside dwelling during pregnancy
Never303 (87.6)527 (84.7)0.949
Occasionally19 (5.5)43 (6.9)
Once per week2 (0.6)4 (0.7)
Daily22 (6.3)48 (7.7)
 At the age of 1 year 
 n (%), N = 346n (%), N = 606
Total & specific IgE at age one
Enhanced total IgE (>3.8kU/l)d274 (79.2)404 (66.6)0.988
Sensitization to food allergens (fx5)e52 (15.0)79 (13.0)
Sensitization to inhalant allergens (sx1)e13 (3.7)17 (2.8)
Atopic diseases at age one
Symptoms of atopic dermatitis38 (10.9)59 (9.73)0.970
Physician diagnosed atopic dermatitis36 (10.4)60 (9.90)
Physician diagnosed cradle cap202 (58.4)354 (58.4)

Relationship between gender, parental history of atopy and Treg numbers in cord blood

In the analysed subcohort, the median value of Treg numbers in cord blood, as detected by changes in demethylation of TSDR, was 1.18% [interquartile range (IQR) 0.73–1.71].

In our analyses, the number of Tregs varied by gender of newborns. In cord blood of female neonates, the number of Tregs was significantly higher in comparison with the male ones (< 0.001, Table 2). Furthermore, parental history of atopy was related to lower Treg numbers in cord blood (Table 2). Adjusted mean ratios for gender, maternal smoking/exposure to ETS at home and parental educational level revealed an inverse relationship between parental history of atopy and Treg numbers in cord blood (adjusted MR = 0.78, 95% CI = 0.61–0.99 double positive, Fig. 1 and Table S1). Furthermore, we found lower Treg numbers in cord blood of neonates in the case of maternal hay fever (adjusted MR = 0.81, 95% CI = 0.68–0.97) and paternal asthma (adjusted MR = 0.60, 95% CI = 0.45–0.81) (Fig. 1, Table S1).

image

Figure 1. Relationship between parental history of atopy and Treg numbers in cord blood as detected by demethylation of Treg-specific demethylated region in the FOXP3 gene. Linear regression models (shown are MR and 95% confidence intervals) were adjusted for gender, maternal smoking/environmental tobacco smoke exposure at home and parental educational level. πHistory of atopy is defined as occurrence of asthma or atopic dermatitis or hay fever.

Download figure to PowerPoint

Table 2. Relationship between gender, parental history of atopy and Treg numbers in cord blood as detected by demethylation of Treg-specific demethylated region in the FOXP3 gene
ParameterTreg numbers in cord blood (%)
NMedian (IQR)P valuea
  1. Treg numbers are presented as medians and interquartile range (IQR).

  2. a

    Analyses were performed using Mann–Whitney U-test. Significant values are in boldface (< 0.05).

  3. b

    History of atopy is defined as occurrence of asthma or atopic dermatitis or hay fever.

Gender
Male1890.99 (0.61–1.55) 
Female1571.39 (0.86–1.95)<0.001
Parental history of atopyb
Negative1221.31 (0.83–1.86) 
Single positive1551.14 (0.70–1.72)0.167
Double positive691.04 (0.58–1.50)0.049
Maternal history of atopyb
Negative1951.21 (0.97–1.80) 
Positive1511.13 (0.67–1.64)0.264
Asthma
No3111.18 (0.74–1.72) 
Yes341.25 (0.63–1.60)0.820
Hay fever
No2311.22 (0.80–1.80) 
Yes1131.04 (0.58–1.59)0.064
Atopic dermatitis
No2871.20 (0.73–1.71) 
Yes571.13 (0.80–1.72)0.943
Paternal history of atopyb
Negative2041.25 (0.77–1.83) 
Positive1421.05 (0.58–1.59)0.061
Asthma
No2991.25 (0.76–1.86) 
Yes330.85 (0.44–1.38)<0.001
Hay fever
No2111.22 (0.77–1.78) 
Yes1161.05 (0.59–1.70)0.174
Atopic dermatitis
No2961.19 (0.75–1.82) 
Yes301.00 (0.51–1.51)0.145

Relationship between environmental exposure during pregnancy and Treg numbers in cord blood

Analysis of environmental exposure included maternal smoking/ETS at home, renovation activities, the usage of chemicals like disinfectants or air fresheners, as well as traffic exposure during pregnancy. Prenatal maternal smoking/ETS exposure at home as well as the usage of disinfectants during pregnancy were found to be associated with reduced Treg numbers in cord blood (= 0.039, = 0.042, respectively, Table 3). Treg numbers in cord blood did not vary significantly following renovation activities, usage of air fresheners, or traffic exposure during pregnancy (Table 3). After adjustment for gender, parental history of atopy, parental educational level and maternal cytokine production, only the association between maternal smoking/ETS exposure at home remained significant (adjusted MR = 0.89, 95% CI = 0.79–1.00, = 0.047; Fig. 2, Table S2).

image

Figure 2. Relationship between maternal prenatal exposure and Treg numbers in cord blood as detected by demethylation of Treg-specific demethylated region in the FOXP3 gene. Linear regression models (shown are MR and 95% confidence intervals) were adjusted for gender, parental history of atopy, parental education and maternal production of IL-13, IL-17E, IFN-γ upon phytohemagglutinin stimulation. πETS, environmental tobacco smoke. ПQuestionnaire reported painting, installation of new furniture or changing the flooring either before conception or during pregnancy.

Download figure to PowerPoint

Table 3. Association between exposures during pregnancy and Treg numbers in cord blood as detected by demethylation of Treg-specific demethylated region in the FOXP3 gene
Exposure during pregnancyTreg numbers in cord blood (%)
NMedian (IQR)P valuea
  1. Treg numbers are presented as medians and interquartile range (IQR).

  2. ETS, environmental tobacco smoke.

  3. a

    Analyses were performed using Mann–Whitney U-test. Significant values are in boldface (< 0.05).

  4. b

    Questionnaire reported painting, installation of new furniture or changing the flooring either before conception or during pregnancy.

  5. c

    Values are presented in metre and refer to the distance between dwelling and main streets.

Maternal smoking/ETS exposure at home
Never3031.17 (0.73–1.76) 
Daily220.85 (0.52–1.38)0.039
Any renovationb
No761.04 (0.62–1.72) 
Yes2451.19 (0.73–1.66)0.527
Usage of disinfectants
No1211.38 (0.81–1.80) 
Yes2191.05 (0.70–1.68)0.042
Usage of air freshener
No1371.35 (0.75–1.81) 
Yes2051.05 (0.73–1.63)0.135
Trafficc
≤10m391.32 (0.84–1.65) 
>150m1191.20 (0.70–1.74)0.713

Relationship between maternal cytokine production during pregnancy and Treg numbers in cord blood

The relationship between maternal production of inflammatory markers IL-6, IL-8, IL-10, TNF–α and MCP-1 (CCL2) upon LPS stimulation as well as between PHA-induced production of maternal Th1/Th2/Th17 cytokines (IFN-γ, IL-4, IL-5, IL-13, IL-17E and IL-17A) and Treg numbers in cord blood was determined by linear regression adjusted for gender, parental history of atopy, maternal smoking/ETS at home and parental educational level (Fig. 3, Table S3). According to these models, maternal PHA-induced IL-13 (adjusted MR = 0.85, 95% CI = 0.75–0.96), IL-17E (adjusted MR = 0.92, 95% CI = 0.87–0.97) and IFN-y (adjusted MR = 0.90, 95% CI = 0.83–0.98) production were associated with significantly reduced Treg numbers in cord blood (Fig. 3, Tables S3 and S4).

image

Figure 3. Relationship between phytohemagglutinin-/lipopolysaccharide-stimulated cytokine production in maternal blood and Treg numbers in cord blood as detected by demethylation of Treg-specific demethylated region in the FOXP3 gene. Linear regression models (shown are MR and 95% confidence intervals) were adjusted for gender, parental history of atopy, maternal smoking/environmental tobacco smoke exposure at home and parental educational level.

Download figure to PowerPoint

To verify the results obtained by testing possible influencing factors in separate models, we applied an additional modelling approach in which all factors potentially influencing Treg numbers at birth were included in one regression model. Following this approach, we obtained comparable results, confirming the data that was gained from single models (see Appendix S1, results and Fig. S2).

Factors influencing maternal cytokines

We analysed the relationship between maternal history of atopy, environmental exposure and cytokine production during pregnancy. Increased maternal IL-13 production was related to maternal history of atopy, maternal hay fever and asthma, whereas increased IL-17E levels were associated with the usage of disinfectants during pregnancy (see results Appendix S1, Tables S6 and S7).

Relationship between Treg numbers in cord blood, diseases and atopic sensitization in the first year of life

Analyses of the relationship between cord blood Treg numbers and allergic outcomes included the occurrence of physician-diagnosed cradle cap, atopic dermatitis (symptoms and physician diagnosed) as well as sensitization to food and inhalant allergens in the first year of life (Table 4, Table S5, results, Appendix S1).

Table 4. Association between Treg numbers in cord blood as detected by demethylation of Treg-specific demethylated region in the FOXP3 gene and allergic diseases as well as sensitization to food and inhalant specific allergens in the first year of life
Outcome at age oneTreg numbers in cord blood (%)
NMedian (IQR)P valuea
  1. Treg numbers in cord blood are presented as medians and interquartile range (IQR).

  2. a

    Analyses were performed using Mann–Whitney U-test. Significant values are in boldface (< 0.05).

  3. b

    According to the Pharmacia CAP System, concentrations >0.35 kU/l were regarded as positive.

Atopic diseases
Atopic dermatitis
Symptoms
No3071.22 (0.75–1.74) 
Yes380.86 (0.51–1.47)0.021
Diagnosed
No3091.22 (0.75–1.72) 
Yes360.91 (0.54–1.39)0.046
Cradle cap
No1431.28 (0.74–1.72) 
Yes2021.15 (0.73–1.69)0.535
Sensitizationb
To food allergens
No2941.21 (0.76–1.76) 
Yes520.97 (0.76–1.76)0.032
To inhalant allergens
No3331.19 (0.74–1.70) 
Yes131.04 (0.55–1.71)0.549

According to our adjusted logistic regression models, children with lower Treg numbers in cord blood had a significantly higher risk of developing an atopic dermatitis in the first year of life (physician-diagnosed: adjusted OR = 1.55, 95% CI = 1.00–2.41, = 0.048, Fig. 4, Table S5). Upon analysis of specific IgE levels in sera of the 1-year-old children, we found a higher risk for sensitization against food allergens in case of lower Treg numbers at birth (adjusted OR = 1.55, 95% CI = 1.06–2.25, = 0.020, Fig. 4, Table S5).

image

Figure 4. Relationship between Treg numbers, atopic diseases and sensitization in cord blood as detected by demethylation of Treg-specific demethylated region in the FOXP3 gene. гLogistic regression models (OR and 95% confidence intervals) were adjusted for gender, maternal atopic dermatitis, maternal smoking/environmental tobacco smoke exposure (ETS) at home and parental educational level. μAdjusted odds ratios for gender, parental history of atopy and parental educational level (because of the distribution of cases maternal smoking/ETS exposure at home could not be considered). *According to the Pharmacia CAP System, concentrations >0.35 kU/l were regarded as positive.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. Authors’ contribution
  8. Conflict of interest
  9. References
  10. Supporting Information

Our results reveal differences in cord blood Treg numbers stably expressing FOXP3 depending on gender, family history of atopy as well as prenatal exposure to tobacco smoke. In our study, we found that being male or having parents with a history of atopy was related to lower Treg numbers in cord blood. Furthermore, the capacity of maternal blood cells to produce cytokines (IFN-γ, IL-13 and IL-17E) upon stimulation was inversely related to Treg numbers in cord blood. In addition, maternal smoking/exposure to ETS at home was associated with decreased Treg numbers. Finally, the results indicate that children with low Treg numbers at birth might be at a higher risk of developing atopic dermatitis or sensitization to food allergens in the first year of life (Fig. 5).

image

Figure 5. Overview on prenatal factors that might be associated with alterations in Treg numbers and thereby modulate the susceptibility to diseases at the age of one. *Maternal atopy, asthma or hay fever; **usage of disinfectants during pregnancy.

Download figure to PowerPoint

The strength of our study lies in the fact that the measurement of Treg numbers was performed by analysing the methylation state of the TSDR in the FOXP3 gene, which constitutes as an exclusive, specific and reliable marker for stable Tregs [12-15]. We analysed TSDR demethylation in relation to various factors in the prenatal environment, set in a large prospective birth cohort. Our study comprises mother–child pairs that are well characterized regarding immune parameters as well as for atopic outcomes and exposure variables. A weakness of the LINA study in general is the potential bias by high rates of participating atopic parents (64.7%). We have considered this point by including parental atopy history as confounding variable in the regression models. One other limitation of the study is the low number of exposed participants to certain variables (e.g. daily smoking/ETS at home n = 22). We are aware of the fact that testing the possible influencing factors on Treg numbers in separate models might not be the most convincing approach. To overcome this weakness, we additionally tested all potentially influencing factors in one regression model. Both approaches generated similar results. As usual in population-based studies, we also have the limitation to define the cause–effect relation in our observed associations.

In our study cohort, the median value of Treg numbers in cord blood, as detected by changes in demethylation of TSDR, was 1.18% (IQR = 0.73–1.71). This value is comparable to data from a recent publication by Schaub et al. [21] using the same method for Treg quantification. Depending on farm exposure, they found median Treg frequencies to be between 1.6–2.0 (IQR = 1.2–2.4).

Results from epidemiologic studies addressing the influence of parental history of atopy on cord blood Tregs are conflicting. Our finding that a family history of atopy is associated with reduced Treg numbers in cord blood is in line with previous studies demonstrating an impairment of Tregs at birth in neonates with a hereditary allergy risk [5, 6, 22]. In contrast, Rindsjö et al. [7] did not detect an effect of maternal allergic disease on Treg subsets in cord blood. This might be explained by differences in the experimental setup. In the study of Rindsjö et al., Tregs were analysed flow cytometrically by the expression of FoxP3 and a panel of T-cell homing markers, whereas in the other studies, the suppressive capacity of Tregs and/or the expression of Treg-related proteins including FoxP3 were determined [6, 22].

In our study, male neonates had significantly lower Treg numbers compared with females, suggesting gender-dependent differences in Treg development. By now, there is no evidence concerning the relationship between gender and Treg numbers. However, gender-associated disproportion has been described in immune development and allergic immune responses [reviewed in [23, 24]]. As Tregs play an important role in preventing allergic diseases, our results (lower Treg numbers in male cord blood) fit with the well-documented finding that atopic diseases in early childhood are predominant in boys [23, 25].

Concerning maternal intrinsic factors, it has been described previously that the maturation of the foetal immune system is determined by maternal immune status, atopic disease, as well as the capacity of adaptive and innate immune responses [1, 2]. For example, maternal sensitization to allergens was found to be associated with an elevated production of the Th2 cytokine IL-13 in human infants [26]. In a previous study, we demonstrated that total maternal IgE in pregnancy was predictive for total IgE in cord blood and in the first year of life, and that maternal TNF-α and IL-10 production during pregnancy was an indicator for high concentrations of these cytokines in children's blood at the age of one [27]. Furthermore, we showed that high blood TNF-α concentrations during pregnancy are protective against sensitization to inhalant allergens in 1-year-old children [27]. There are no data in the literature concerning the influence of maternal immune responses in pregnancy on the development of foetal Tregs. In our study Th1, Th2 and Th2-related cytokines IFN-γ, IL-13 and IL-17E (IL-25) were associated with reduced Treg numbers in cord blood. Thus, a strong maternal adaptive immune response in pregnancy might antagonize the development of foetal Tregs. Our data therefore support previous in vitro findings demonstrating an inhibitory influence of Th1/Th2 polarizing cytokines on Treg development [28]. We furthermore investigated factors influencing maternal cytokine production during pregnancy (see Appendix S1). Focusing on maternal cytokines relevant for children's Treg development, we found that increased maternal IL-13 production was related to maternal history of atopy, maternal hay fever and asthma, whereas increased IL-17E levels were associated with the usage of disinfectants during pregnancy. Thus, at least maternal atopy and usage of disinfectants might increase maternal IL-13 and IL-17E production and thereby influence the number of Treg in cord blood.

In our study, maternal smoking/exposure to ETS at home was associated with lower Treg numbers in cord blood compared with neonates without exposure. The influence of maternal smoking on Tregs at birth has not been investigated so far. However, maternal smoking has been shown to modify neonatal immune functions. This results in higher neonatal Th2 responses, as well as perturbed toll-like receptor (TLR)-mediated responses in cord blood at birth [29, 30]. Furthermore, TLR-mediated responses at birth have been linked to the expression of Treg-related genes [22].

In addition to maternal intrinsic factors and exposure, paternal genetic/epigenetic factors may also affect the maturation of the foetal immune system. Analysing the influence of genetic predisposition on Tregs as detected by changes in the demethylation of TSDR in our study depicted a significant relationship between paternal asthma and lower Treg numbers in cord blood. Thus, in case of inheritance of FOXP3 methylation patterns, the atopic father seems to play an important role, too. Considering the fact that boys had reduced Treg numbers, male gender might in general be a predisposition in terms of lower Treg numbers determined by increased methylation of TSDR in the FOXP3 gene.

Epigenetic alterations are discussed as one potential mechanism mediating the impact of the environment on gene expression [3]. Thus, environmental factors might have the potential to influence the epigenetic fixation of the Treg phenotype by TSDR demethylation or methylation. Recently, it has been shown that prenatal tobacco smoke exposure affects global and gene-specific DNA methylation [31]. The authors demonstrated that in utero exposure to tobacco smoke is related to global hypomethylation, whereas single CpG loci were hypermethylated in several genes. Whether tobacco smoke exposure during pregnancy directly influences epigenetic regulation of FOXP3 or involves other pathways remains to be elucidated.

Finally, we were interested in the predictive value of low Treg numbers in cord blood regarding the development of diseases and allergic sensitization in the first year of life. Even though Tregs in humans constitute only up to two per cent of the total peripheral CD4+ T cells [32], we were able to detect significant associations between cord blood Treg numbers and allergic outcomes. We demonstrated that atopic dermatitis and sensitization to food allergens in the first year of life were associated with lower Treg numbers in cord blood. Our findings are in line with results from a study demonstrating an association between impaired suppressive function of Tregs in cord blood at birth and egg allergy in the first year of life [33]. Thus, Treg numbers in cord blood may be considered as a predictive marker for the development of allergic diseases later in life.

In a previous investigation of the LINA study, we found that maternal cytokines during pregnancy did not directly correlate with atopic dermatitis of the child at the age of one [27]. However, we found an association between maternal and children's immune response. Deduced from the results of the present study, we hypothesized that maternal cytokines in pregnancy do not directly influence the development of atopic diseases in the offspring, but rather they might affect the maturation of the foetal immune system (e.g. Treg development) and thereby the development of allergies and allergic sensitization.

Altogether our results suggest that male sex, maternal intrinsic factors (atopy history, IL-13 and IL-17E production), parental history of atopy as well as ETS exposure during pregnancy are related to lower Treg numbers in cord blood (Fig. 5). As we found a relationship between maternal history of atopy and maternal IL-13 production, we hypothesize that the maternal atopic background affects foetal Treg development via the cytokine milieu during pregnancy. Prenatal tobacco smoke exposure seems to affect foetal Tregs directly without affecting maternal immune responses, which might be relevant for Treg development. As a consequence of lower cord blood Treg numbers, the risk for the development of allergic sensitization against food allergens and atopic dermatitis increases.

Mechanisms by which Treg affects the allergic outcome and how environmental factors act on Treg development have to be elucidated in future studies. Nevertheless, by measuring demethylation of TSDR in the FOXP3 gene, the results of our study provide evidence for prenatal modulation of Treg numbers and encourage further investigation in this direction.

Acknowledgments

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. Authors’ contribution
  8. Conflict of interest
  9. References
  10. Supporting Information

We cordially thank the participants of the LINA study as well as Beate Fink and Anne Hain for their excellent technical assistance. The authors are grateful to Neil Jones and Sonja Nakasian for the careful revision of the manuscript. This work was partially supported by the Helmholtz Impulse and Networking Fund through Helmholtz Interdisciplinary Graduate School for Environmental Research (HIGRADE) and by the German Research Foundation (KFO250).

Authors’ contribution

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. Authors’ contribution
  8. Conflict of interest
  9. References
  10. Supporting Information

Each named author substantially contributed to this paper. Gunda Herberth, Mario Bauer, Stefan Röder, Michael Borte, and Irina Lehmann were involved in the development of the study design and the field work. Denise Hinz, Gunda Herberth, Mario Bauer and Sven Olek performed the FOXP3 TSDR demethylation measurement. Stefan Röder, Denise Hinz, Gunda Herberth and Irina Lehmann contributed to the statistical analysis. Ulrich Sack contributed with IgE measurements and discussion of the IgE data. Jochen Huehn and Jan Christoph Simon contributed with the discussion of FOXP3 TSDR demethylation and allergy outcome. Gunda Herberth, Denise Hinz and Irina Lehmann wrote the paper; all authors were involved in the revision of the final text.

References

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. Authors’ contribution
  8. Conflict of interest
  9. References
  10. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. Authors’ contribution
  8. Conflict of interest
  9. References
  10. Supporting Information
FilenameFormatSizeDescription
all2767-sup-0001-figs1.docxWord document34KFigure S1. Flow chart of the LINA study population with the analysed subcohort.
all2767-sup-0002-figs2.docxWord document138KFigure S2. Regression model with all factors possibly influencing cord blood Treg numbers.
all2767-sup-0003-tables1.docxWord document17KTable S1. Relationship between parental history of atopy and Treg numbers in cord blood as detected by demethylation of TSDR in the FOXP3 gene.
all2767-sup-0004-tables2.docxWord document16KTable S2. Relationship between maternal prenatal exposure and Treg numbers in cord blood as detected by demethylation of TSDR in the FOXP3 gene.
all2767-sup-0005-tables3.docxWord document16KTable S3. Relationship between PHA-/LPS-stimulated cytokine production in maternal blood during pregnancy and number of Tregs in cord blood as detected by demethylation of TSDR in the FOXP3 gene.
all2767-sup-0006-tables4.docxWord document15KTable S4. Cytokine production during pregnancy in PHA-/LPS-stimulated blood samples. Values are presented as medians and the interquartile range.
all2767-sup-0007-tables5.docxWord document17KTable S5. Relationship between Treg numbers in cord blood as detected by demethylation of TSDR in the FOXP3 gene and the development of allergy as well as sensitisation to food or inhalant specific allergens in the first year of life.
all2767-sup-0008-tables6.docxWord document19KTable S6. (a) Association between maternal history of atopy and PHA induced maternal cytokine production in pregnancy. Cytokine concentrations are presented as medians and interquartile range (IQR). (b) Association between maternal history of atopy and PHA induced maternal cytokine production in pregnancy. Cytokine concentrations are presented as medians and interquartile range (IQR).
all2767-sup-0009-tables7.docxWord document21KTable S7. (a) Association between exposures during pregnancy and maternal cytokine production upon PHA-stimulation. Cytokine concentrations are presented as medians and interquartile range (IQR). (b) Association between exposures during pregnancy and maternal cytokine production upon PHA-stimulation. Cytokine concentrations are presented as medians and interquartile range (IQR). (c) Association between exposures during pregnancy and maternal cytokine production upon PHA-stimulation. Cytokine concentrations are presented as medians and interquartile range (IQR).
all2767-sup-0010-Suppl_Material.docWord document416KAppendix S1. Methods and Results.

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.