Susan L. Prescott School of Paediatrics and Child Health University of Western Australia PO Box D184 Princess Margaret Hospital Perth WA 6001 Australia
Background: During pregnancy, variations in maternal–foetal cellular interactions may influence immune programming. This study was carried out to determine if maternal responses to foetal alloantigens are altered by maternal allergic disease and/or previous pregnancies.
Methods: For this cohort study, peripheral blood was collected from allergic (n = 69) and nonallergic (n = 63) pregnant women at 20, 30, 36-week gestation and 6-week postpartum (pp). Cord blood was collected at delivery. Mixed lymphocyte reactions were used to measure maternal cytokine responses [interleukin-6 (IL-6), IL-10, IL-13 and (interferon-γ) IFN-γ] at each time point towards foetal mononuclear cells.
Results: Maternal cytokine responses during pregnancy (20, 30 and 36 weeks) were suppressed compared to the responses at 6-week pp. The ratio of maternal IFN-γ/IL-13 and IFN-γ/IL-10 responses were lower during pregnancy. Allergic mothers had lower IFN-γ responses at each time-point during pregnancy with the greatest difference in responses observed at 36-week gestation. When allergic and nonallergic women were further stratified by gravidity group, IFN-γ responses of allergic multigravid mothers were significantly lower than nonallergic multigravid mothers during pregnancy.
Conclusions: During normal pregnancy, peripheral T-cell cytokine responses to foetal alloantigens may be altered by both allergic status of the mother and previous pregnancies. These factors could influence the cytokine milieu experienced by the foetus and will be further explored in the development of allergic disease during early life.
Patterns of perinatal immune responses have been convincingly linked to the risk of subsequent allergic disease. The most consistent of these observations has been that the ‘normal’ T-helper cell type 1 (Th-1) immaturity of the neonatal period is more profound in newborns with allergic predisposition (1–5). This has generated enormous interest in the potential factors that could influence early immune function and predisposition to disease. An increasing number of studies have examined the effects of environmental influences during pregnancy and demonstrated that exogenous factors, such as microbial exposure (6), maternal diet (7, 8) and smoking (9) can have effects on developing immune responses. However, endogenous influences such as variations in cellular interactions at the maternal–foetal interface have not been extensively studied.
Immune interactions between mother and foetus play an important role in determining cytokine homeostasis during pregnancy. Successful pregnancy involves a relative Th-2 cytokine bias in the placenta, which occurs under the influence of hormonal changes (10–12). This is reflected by foetal responses at birth which are also relatively ‘Th-2 skewed’ (4), indicating that this shift in cytokine expression has a critical effect on subsequent immune development.
Direct interaction between mother and foetus has the capacity to influence the pattern of foetal immune development by subtle alterations of immune responses during pregnancy. A recent exploratory study by our group showed that maternal responses to foetal alloantigen were directly related to foetal Th-2 [interleukin-13 (IL-13), IL-10] cytokine responses and allergic outcomes at 6 years of age (13). Here, we explore the hypothesis that allergic mothers, who have a predisposition towards Th-2-like responses to allergens, also produce such a response to foetal alloantigen during pregnancy, modifying the Th-1/Th-2 polarity during pregnancy. Maternal atopy has been reported to have a stronger effect than paternal allergy on allergy risk (14, 15) as well as Th-1 immune dysfunction in newborns that later developed allergic symptoms (16). These observations support a direct gestational influence of maternal allergy on immune development. In this study, we examine how maternal allergic status influences maternal responses to foetal alloantigens.
Study design and recruitment criteria
This was a longitudinal study of 169 healthy pregnant women recruited from the private clinics of obstetricians in the Perth metropolitan area, Western Australia between August 2002 and March 2005. At 20-week gestation, pregnant women were recruited into study groups according to allergic status and gravidity and followed-up throughout their pregnancy and postpartum (pp). Maternal allergy was defined as a doctor diagnosed clinical history of asthma, eczema or allergic rhinitis plus a positive skin prick test (SPT) to one or more common allergens. Nonallergic mothers had no history of allergic disease and negative SPT to all tested allergens. Women were considered as ‘primigravid’ if they had no known previous pregnancies (including miscarriages and terminations). For the purpose of this study, ‘multigravid’ women were those with one or more previous pregnancies with their current partner (thereby excluding multiparous women having their first pregnancy with a new partner).
Mothers were excluded if: (i) they were aged <18 and >44 years of age; (ii) there were any maternal or foetal complications (including pre-eclampsia or major congenital anomalies); (iii) parental smoking in pregnancy or during the last 2 years; (iv) the delivery was preterm (<36 weeks gestation) or (v) they used immunomodulatory medications (including glucocorticoids for threatened preterm labour).
Collection and cryopreservation of mononuclear cells
Maternal peripheral blood samples were collected from mothers at 20, 30, 36-week gestation and 6-week pp. Thirty millilitres of blood was collected via routine venipuncture of the cubital fossa vein, into an equivolume of heparinized (preservative-free heparin) Roswell Park Memorial Institute (RPMI, Gibco; Life Technology, Paisley, UK) tissue culture medium. At the time of delivery cord blood (CB) samples were collected from the placental vessels via venipuncture (19G needle) into heparinized (preservative-free) RPMI tissue culture medium as previously described (4). Mononuclear cells (MNC) from all collections were isolated by Ficoll-Hypaque gradient centrifugation and cryopreserved (in 7.5% DSMO) using established techniques (4) which have been shown not to significantly alter cell function.
Mixed Lymphocyte reactions
Vials of cryopreserved MNC were thawed at 37°C then transferred to a 10 ml tube to which cooled RPMI was added drop wise. The cells were pelleted via centrifugation at 500 g for 5 min and then resuspended to 1 ml in AIM-V, serum free tissue culture medium (Gibco) supplemented with 2-mercaptoethanol (2ME; 4 × 10−5M final concentration, Sigma, Castle Hill, Australia). Cell viability and concentration was determined by staining with trypan blue (4%).
Maternal ‘responder’ populations (125 μl at a concentration of 0.5 × 106 cells/mL) were added to culture wells. Cord blood MNC ‘stimulator’ populations were CD3-depleted using CD-3 Dynabeads (Dynal Biotech ASA, Oslo, Norway, as per manufacturers instructions) and irradiated at 3000 rad (Gammacell 3000 Elan; MDS Nordion, Ontario, Vic., Canada) to inhibit proliferation by this population. Stimulator cells were then added to culture wells (125 μl at 0.5 × 106 cells/ml) containing responders. All cell populations were diluted in AIM-V (Gibco) supplemented with 2ME (4 × 10−5M final concentration; Sigma) plus 5% heat inactivated AB serum (Sigma-Aldrich Inc, St Louis, MO, USA). Maternal responder MNC were also cultured with stimulator autologous MNC or MNC pooled from 16 unrelated donors (URD) as a negative or positive control respectively. Mixed lymphocyte reaction (MLR) cultures were incubated at 37°C in 5% CO2 for 72 h for cytokine protein measurement or 5 days for lymphoproliferation. Responder and stimulator populations were also stimulated separately (250 μl at 1 × 106/ml) with recombinant IL-2 to ensure that the CD3 depletion and irradiation process was successful and in PHA to ensure cells were viable.
Cytokine protein detection
Cytokines [IL-6, IL-10, IL-13, interferon-γ (IFN-γ)] in cell culture supernatants were quantified using a ‘sandwich-type’ time-resolved fluorometry as previously described (8). Matched antibody pairs, (consisting of unlabelled capture antibody and biotinylated detection antibody), and recombinant protein standards (Pharmingen; BD Biosciences, San Jose, CA, USA) were used at an optimal concentration established for our laboratory. Data were expressed as concentration in pg/ml. The detection limit of the assay was 5 pg/ml for all cytokines. Samples with concentrations below this value were not included in the statistical analysis. IL-4 and IL-5 were not measured in this study as these cytokines were generally low level (i.e. close to or below the detection limit of the assay), in the majority of samples from the pilot study.
After 5-day incubation, 0.5 μCi of tritiated thymidine [(methyl-3H) thymidine 5 mCi] (Amersham Biosciences UK Ltd, Aylesbury, UK) per well was added for the last 16–18 h of incubation as a measure of DNA synthesis. Disintegrations per minute (dpm) were measured on a beta counter (Packard Matrix 9600; Packard Instrument Co., Meriden CT, USA). The median of each triplicate was used as a measure of T-cell proliferation. A positive response was defined as a delta dpm >1000 (dpm of test minus dpm of unstimulated control).
All statistical analysis was performed using Statistical Package for Social Sciences (spss version 13.0 for Macintosh; SPSS, Inc., Chicago, IL, USA). The majority of the cytokine data did not have a normal distribution. Where this could be normalized by natural logarithmic transformations, the results were displayed as geometric mean ± 95% CI. Group comparisons of paired and unpaired data were analysed by student’s t-test. Linear regression analyses were used to control for possible confounders (e.g. breast feeding) on the relationship between two continuous variables. P-values < 0.05 were considered significant.
Ethical approval was granted by Princess Margaret Hospital for Children, King Edward Memorial Hospital, St John of God Hospital and Mercy Hospital ethics committees and all women gave informed written consent.
Characteristics of the cohort
Of the 169 pregnant women initially recruited into the study, 18 were excluded because of missed CB collections and three who withdrew before delivery. Additionally one mother was excluded because of diagnosis of pre-eclampsia and two for premature delivery. A further 13 mothers–infant pairs were not included as the concentration of maternal or cord MNC collected was not high enough for the MLR cultures. The remaining 132 females included 69 allergic and 63 nonallergic women who were matched for gravidity status. Within the allergic group, the predominant condition was allergic rhinitis which affected 48/69 (70%) mothers, while asthma and eczema both affected 22/69 (32%). There was no statistical difference in maternal age, mode of delivery, infant gender, gestational age or appearance, grimace, activity, and respiration score in the allergic group compared to the nonallergic group. Neonates of allergic women tended to have lower birth measurements, however, this only reached statistical significance for head circumference (P = 0.044, Table 1).
Table 1. Characteristics of the cohort
Nonallergic n = 63 (%)
Allergic n = 69 (%)
Data shown as n (%) or mean (SD) for normally distributed data.
†Partial or exclusive breast feeding at 6-week postpartum.
Age at delivery (years)
32.8 (SD 4.0)
32.2 (SD 4.0)
21.4 (SD 2.3)
20.7 (SD 2.0)
30.4 (SD 0.9)
30.1 (SD 1.7)
36.2 (SD 0.8)
36.3 (SD 0.5)
6.8 (SD 1.4)
6.5 (SD 1.8)
Mode of delivery
Birth gestation (weeks)
39.2 (SD 1.0)
39.3 (SD 1.1)
3.47 (SD 0.43)
3.42 (SD 0.41)
50.2 (SD 2.0)
49.8 (SD 2.9)
35.2 (SD 1.6)
34.6 (SD 1.3)*
APGAR (5 mins)
Maternal responses to foetal alloantigens
Mixed lymphocyte reactions were employed to measure the maternal response to CB MNCs and thus the contribution of paternally inherited antigens expressed by the foetus. As expected, negative control cultures resulted in low levels of lymphoproliferation (Fig. 1A) and low cytokine production (data not shown) at all time-points. Of note, maternal proliferative responses to MNCs pooled from URD were significantly higher than maternal responses to foetal antigens (Fig. 1A, P < 0.001). For the combined study population (Fig. 1) there were no significant changes in maternal cytokine or proliferative responses to CB MNC from 20 to 36-week gestation and this was evident regardless of maternal allergic status. All cytokine and proliferative responses were significantly suppressed during gestation compared to 6 weeks pp (Fig. 1; proliferations: P < 0.001 IL-6 P < 0.001; IL-10 P = 0.002; IL-13 P < 0.001; IFN-γP < 0.001 for 6-week pp comparison to 36-week response). Notably maternal responses to URD were similar regardless of the pregnancy status (Fig. 1A, data shown for proliferative responses only).
Next, we examined for changes in the patterns of cytokine production. The ratio of IFN-γ/IL-13 (Fig. 2A) and IFN-γ/IL-10 (Fig. 2B) cytokine production was determined at each time-point. Both ratios were significantly lower during pregnancy compared to 6-week pp. There was no change in the IFN-γ/IL-13 and IFN-γ/IL-10 ratios over the course of pregnancy (data not shown).
Effect of allergy and gravidity on maternal responses to foetal alloantigens
The cytokine responses of allergic and nonallergic women towards CB MNC were compared (Table 2). There was a trend for allergic mothers to have lower IFN-γ responses at each time-point during pregnancy with the greatest difference in responses observed at 36-week gestation (P = 0.024). When allergic and nonallergic women were further stratified by gravidity group (Fig. 3, data shown for 36-week gestation only) IFN-γ responses of allergic multigravid mothers were significantly lower than nonallergic multigravid mothers (P = 0.017 at 36 weeks; P = 0.040 at 30 weeks; P = 0.008 at 20 weeks). There was a nonsignificant trend towards higher IFN-γ responses during pregnancy in allergic primigravid mothers compared to nonallergic primigravid mothers (P = 0.322 at 36 weeks; P = 0.933 at 30 weeks; P = 0.315 at 20 weeks). Allergic mothers tended to have higher IL-13 and higher proliferative responses (though not significant) to CB MNC both during and after pregnancy which was still observed after stratifying for gravidity status (Fig. 3B, IL-13 responses shown at 36-weeks gestation only: P = 0.674 for primigravids, P = 173 for multigravids). No trends were observed for IL-6 and IL-10 when allergy and gravidity groups were compared.
Table 2. Maternal responses to cord mononuclear cells during and after pregnancy. Comparison between allergic and nonallergic mothers.
Response to foetus mean (95% CI)
pp, post-partum; dpm, disintegrations per minute.
Effect of other factors
Some mothers did not have a detectable cytokine response (<5 pg/ml) to their foetus and the percentage of ‘nonresponsive’ mothers differed according to the cytokine measured (% of nonresponders: IL-6 = 49%; IL-10 = 54%; IL-13 = 25%; IFN-γ = 40%) while 94% of the mothers demonstrated a positive proliferative response (delta dpm >1000). All mothers exhibited cytokine and proliferative responses to URD suggesting that nonresponsiveness was related to degree of human leukocyte antigen (HLA) mismatch. When mothers were grouped as a responder or nonresponders, there was no significant difference between allergic vs nonallergic or primigravid vs multigravid mothers for all cytokines.
Lactational status at the 6-week pp visit was of particular interest as there is evidence that breastfeeding may influence maternal immunity after delivery (17). Maternal cytokine and proliferative responses before pregnancy remained significantly lower than the nonpregnant state after controlling for partial or exclusive breast feeding. Specifically using partial correlation the adjusted regression coefficients ‘r’ and P-values are as follows for the relationship between 36 and 6-weeks pp responses: IL-6 adjusted r = 0.709, P < 0.001; IL-10 adjusted r = 0.432, P = 0.017; IL-13 adjusted r = 0.737, P < 0.001; IFN-γ adjusted r = 0.268, P = 0.003; lymphoproliferation adjusted r = 0.679, P < 0.001).
The mode of delivery (caesarean section vs vaginal delivery), maternal age, infant gender, infant birth measurements (weight, length and head circumference), and birth gestation had no effect on maternal responses.
Many factors during pregnancy, both maternal and environmental, may play a role in the development of the immune system. The cellular interaction between mother and foetus could be the most direct influence on foetal immune responses. To our knowledge, this is the first study to measure maternal peripheral responses to the foetus over the course of pregnancy and in the context of allergic disease.
In this study, maternal cytokine responses during pregnancy were suppressed compared to the nonpregnant state. Consideration of the effect of lactational status on the pp responses was also of concern as some research has observed increased cytokine production of both Th-1 and Th-2 cytokines from lymphocytes of breastfeeding mother in ex vivo cultures compared to formula feeders and controls (17–19). However, in this study cytokine responses were still higher pp after adjusting for breastfeeding status. Alternatively, the suppression of maternal alloantigen responses during pregnancy could be caused by actions of pregnancy hormones such as progesterone (20) and/or development of maternal tolerance specifically to paternal alloantigen (21). There is general agreement that the placental interface is not completely impermeable during human pregnancy as there is evidence for two way cell trafficking across this barrier (22, 23) and antibodies directed towards paternal HLA expressed on foetal cells are commonly detected in the peripheral blood of pregnant women. This indicates that maternal immune cells are likely to encounter and respond to foetal antigens throughout the course of pregnancy, highlighting the importance of these mechanisms in preventing foetal immune rejection.
Regulatory T cells (TReg) including the naturally occurring CD4+ CD25+ TReg cells expand during normal human pregnancy (24). The consequences of a deficient activity of TReg cells, namely immunological spontaneous abortion, has been demonstrated in a murine model (25) suggesting these cells may be important for preventing maternal immune rejection. The cytokine IL-10 is produced by some TReg subsets, however, our results did not show an increase in maternal IL-10 responses during pregnancy, rather they were significantly dampened compared to the nonpregnant state. In other studies, IL-10 levels were not upregulated in abortion-prone mice who received TReg cells from normal pregnant mice (26), and IL-4/IL-10 knockout in mice resulted in normal pregnancies (27). Therefore, IL-10 may not be crucial in pregnancy for promoting maternal tolerance to paternal alloantigens. Instead TReg cells may promote maternal–foetal tolerance by activating alternative pathways, such as production of other immunosuppressive molecules i.e. TGF-β (28, 29).
There are a number of factors which may have influenced the alloimmune responses observed in this study. Previous studies have suggested that maternal allergy can affect T-cell responses during pregnancy as circulating IFN-γ and IL-13 producing T lymphocytes were increased in asthmatic pregnant women (30). The tendency for allergic mothers in this study to have higher IL-13 but reduced IFN-γ responses to their foetus could consequentially lead to a gestational environment where there is potentially reduced signals for foetal Th-1 maturation that may contribute to atopic risk of the infant (13). Birth order has also been reported to influence allergy risk with studies showing negative associations between sib-ship size and allergic manifestations (reviewed in 31). It is still not known, however, whether the reduced risk of children born into larger families is because of more frequent exposure to pathogens during early childhood or to increasing parity influencing the maternal interaction with her offspring (32). If birth order does play an important role in allergy risk, it would be expected that subsequent pregnancies might lead to lower Th-2 and/or increased Th-1 type responses to the foetus. However, the results of this study are not consistent with the birth order effect. Whether a lack of difference is caused by sample size, or requires study of other Th-2 cytokines such as IL-4 or IL-5 is not certain. Unexpectedly, multigravid mothers with allergic disease showed the lowest IFN-γ responses compared to nonallergic mothers and both allergic and nonallergic primigravids. It is possible that the maternal Th-1 immune response to alloantigen may be less aggressive in this particular subset of women. Further follow-up study of the infants born to these women is required to determine if these subtle interactions between allergic and gravidity status of the mother have any clinical significance after birth. Human leukocyte antigen mismatch between mother and foetus is also likely to account for some of the variations in immune responses in this study (13) and HLA compatibility may also explain why there was a subset of mothers who showed nondetectable cytokine responses. This laboratory will be typing mothers and infants for future studies to confirm these speculations.
In conclusion, during normal pregnancy, peripheral T-cell cytokine responses to foetal alloantigens may be altered by both allergic status of the mother and previous pregnancies. Follow-up studies of the infants are underway to determine the role of these responses in the development of allergic disease.
This project was funded by the National Health and Medical Research Council (NHMRC) of Australia. Janet Dunstan is supported by the Child Health Research Foundation of Western Australia. Thank you to Elaine Pascoe who assisted with statistical analyses, Amira Wahden for her contribution to volunteer recruitment as well as Heidi Lehmann and Jenefer Wiltschut who assisted with the laboratory experiments. We would also like to thank volunteers, obstetricians and midwives who assisted with the study.