|Home | About | Journals | Submit | Contact Us | Français|
Fetal exposures have come under investigation as risk factors of early life allergic disease. In this study we aimed to examine the relationships between dog or cat exposure and naturally occurring regulatory T cells (Treg cells), thought to play an important role in immune tolerance, in pregnant women. A cross-sectional analysis was conducted among 204 pregnant women who were queried regarding dog and cat exposure. Treg cells (CD4+CD25+Foxp3+ lymphocytes) and allergen-specific IgE were measured in venous blood samples. Atopy was defined as allergen-specific IgE ≥ 0.35 kU/L reactive with common allergens including dust mite, dog, cat, Timothy grass, ragweed, Alternaria alternata, egg white or cockroach. Nonparametric Wilcoxon rank sum tests and linear regression models of log transformed Treg cell levels were used in analyses. Among women sensitized to dog, those who had a dog or cat in the home had lower Treg cell levels compared with those who had no dog or cat. However, among women not sensitized to dog, those with a dog or cat in the home had higher Treg cell levels compared with those who did not. Among women sensitized to cat, those who had a dog or cat in the home had lower Treg cell levels compared with those who had no dog or cat. Gestational age at blood draw did not affect the associations. We conclude that Treg cell levels during pregnancy vary in association with both dog and cat exposure and atopic status.
A recent cross-sectional study reported that almost half (45.6%) of the children in the United States ages 6–9 years are atopic (positive skin test to at least one of ten allergens) (Arbes SJ, et al. 2005). Research on childhood allergic disease has largely focused on early childhood exposures, and causal mechanisms remain poorly understood. Over the last several years, fetal exposures, especially to pets and livestock, have come under investigation as potentially critical factors in the development of allergic diseases. However, studies of maternal exposure to pets and livestock during pregnancy have provided mixed evidence about whether prenatal exposure to pets or livestock might reduce the risk of allergic disease in early life (Aichbhaumik N, et al. 2008; Kerkhof M, et al. 2005; Schönberger HJ, et al. 2005; Ege MJ, et al. 2006; Ege MJ, et al. 2008). Additionally, mechanistic investigations regarding the role of prenatal pet exposure in subsequent allergic disease risk are lacking.
To investigate potential mechanisms, we studied the relationship between regulatory T cells and pet exposure in a cohort of pregnant women. Although there is no consensus, regulatory T cells are thought to play an important role in immune tolerance. The subset of CD25+ regulatory T cells may be key in allergic disease mechanisms because they may suppress Th2 and Th1 responses (Shi HZ, et al. 2005; Seroogy CM and Gern JE 2005; Bacchetta R, et al. 2007; Bellinghausen I, et al. 2005; Lee JH, et al. 2007; Maggi L, et al. 2007).
For this current investigation, we examined the associations between self-reported prenatal exposure to dogs and cats and regulatory T cell percentages (Treg cells were defined as CD4+CD25+Foxp3+ lymphocytes) measured in the blood of 204 pregnant women in the Detroit, Michigan area.
The Wayne County Health, Environment, Allergy and Asthma Longitudinal Study (WHEALS) is an ongoing birth cohort study in southeastern Michigan designed to examine the relationships between early life exposures such as pets, infections or endotoxin and allergic diseases in early childhood. We recruited 21–49 year old women from a predefined geographic area who were in their second trimester or later of pregnancy, were seeing a Henry Ford Health System (HFHS) medical group obstetrician at one of five clinics, and planned to stay in the Detroit area for at least two years after delivery. Women completed an interview about their health histories at the time of recruitment and provided a venous blood sample. Women were required to speak English well enough to provide written informed consent. This research was approved by the HFHS IRB.
Recruitment began in September 2003 and was completed in November 2007. However, maternal blood samples were analyzed for Treg cells only in women recruited after November 28, 2006 because that is when funding for Treg cell analyses in a subset of the mothers became available. We have analyzed all 204 prenatal blood samples collected from the women recruited after that date. These maternal blood samples were drawn between 1 and 102 days prior to delivery.
Details on pet exposure were collected during the prenatal interview. Women were asked whether they had owned or cared for any pets in their home for more than 1 week since they learned they were pregnant. Pet keepers were then asked to identify the type of pet, the length of time the pet had been in their home and the usual number of hours the pet was inside the home each day. For this report, we define a woman as exposed to a pet if she reported owning or caring for dogs or cats that spent 12 or more hours each day in the home for more than 1 week.
Additional covariates considered in the analyses were maternal self-reported race, medications, and pregnancy history. Women were asked to report their race based on U.S. Census 2000 categories. They were also asked to report all medications taken during pregnancy and whether they were currently taking them at the time of the blood draw. Medications were then grouped into any medication for asthma (currently using pill or inhalant taken daily or as needed), any medication for allergies (currently using pill or inhalant taken daily or as needed) and antibiotics (any during pregnancy). Maternal report of the timing (month and year ended), duration (weeks or months) and outcome of each of their prior pregnancies also was recorded.
Details on maternal smoking and secondhand smoke exposure were also collected during the prenatal interview. Women were asked if they were currently smoking at least 1 tobacco product per day and whether they had smoked in the first three months of their pregnancy (at least 1 tobacco product per day). Women were also asked whether other residents of their home smoked. For the analyses, each woman was assigned an indicator variable for maternal smoking during pregnancy and another indicator for whether she lived with smokers. We did not collect information on smokeless tobacco use.
Aliquots (0.5 ml) of whole blood collected in heparinized tubes were incubated for 6 hours at 37°C. Brefeldin A was added after two hours of incubation. The samples then were incubated with BD FastImmune™ EDTA Solution (BD Biosciences, San Jose, CA) for 15 minutes followed by BD FACS Lysing Solution (BD Biosciences) for 10 minutes to lyse red blood cells and fix white blood cells.
Samples were stored at − 80°C until staining and flow cytometry analysis. At the time of staining, samples were thawed and wash buffer (phosphate buffered saline (PBS) plus 0.5% bovine serum albumin) was added. Samples then were centrifuged, the supernatant decanted and the cells resuspended in wash buffer. Each 0.5 ml blood sample was subdivided into 100 μl aliquots for staining.
Fluorochrome-conjugated antibodies against the T cell surface marker CD25 (anti-CD25 FITC, BD Biosciences) along with anti-CD4 PerCP-Cy5.5 (BD Biosciences) for detection of CD4+ T cells were added to each 100 μl staining aliquot. Following incubation with cell surface marker antibodies, cells were then permeabilized by addition of Fixation/Permeabilization buffer (eBioscience Inc., San Diego, CA) and incubated with fluorochrome-conjugated antibody against Foxp3 (anti-Foxp3 PE, eBioscience), a transcription factor used as an intracellular marker of Treg cells. Cells were incubated at room temperature in the dark and fixed with 1% paraformaldelhyde in PBS. Flow cytometry was performed within 24 hours of staining.
Flow cytometry was performed with an LSR flow cytometer (BD Biosciences). The data were collected and analyzed using CellQuest Pro software (BD Biosciences). Matched isotype controls were included in each assay to adjust for nonspecific staining. Cells were gated on lymphocytes using forward and side light scatter properties and a minimum of 10,000 lymphoctye-gated events were acquired. CD4+ T lymphocytes were analyzed with bivariate dot plots of CD25 versus Foxp3. Values were obtained for the proportion of CD4+ lymphocytes that are CD25+ and express Foxp3. Treg cells are expressed as the percentage of CD4+ lymphocytes that are CD25+Foxp3+.
We performed reproducibility tests for flow cytometry analysis of duplicate samples within the same run and of duplicate samples run on different days. The coefficients of variation (CV) for these tests were within 5% for duplicate samples within the same flow cytometry run and within 15% for duplicate samples run on different days. A single technician performed all staining and flow cytometry procedures. All Treg cells analyses were reviewed by a single immunologist (KRB).
Venous blood was also collected for assessment of allergen-specific IgE. Plasma was isolated and stored at − 80°C until assayed. Measurements of allergen-specific IgE were performed following the standard manufacturer’s protocols using the Pharmacia UniCAP system (Pharmacia-Upjohn Diagnostic Division, Kalamazoo, MI). Allergen-specific IgE was analyzed for dust mite, dog, cat, Timothy grass, ragweed, Alternaria alternata, egg white and cockroach. One percent of all assays were repeated in a different assay run on a different day to provide estimates of inter-assay reliability. The geometric mean coefficient of inter-assay variation was 5.9% for all eight allergens. Sensitization was defined as a positive allergen-specific IgE result of ≥ 0.35 kU/L. Atopy was defined as having at least one allergen-specific sensitization.
We used robust descriptive statistics (geometric means and 95% confidence intervals) to describe Treg cell levels during pregnancy for all women and for different subgroups. Firstborn status, first pregnancy, medication use, current asthma, sensitization to any of eight allergens, sensitization to dog, sensitization to cat, tobacco smoke exposure and self-reported African-American race were evaluated as effect modifiers and then as confounders through stratified analyses and change in effects criteria (20%), respectively. These factors were chosen because they have been identified as having potential impact on allergic risk in prior studies. Linear regression models with interaction terms were also used to evaluate effect modification and confounding of associations with log transformed Treg cell levels. Using the blood draw date and the expected delivery date from the interview, and confirmed in the medical record, we calculated the gestational age at the time of blood draw. Gestational age at draw was considered as a factor potentially affecting the associations between pet exposure and Treg cell levels.
The majority of the 204 women in our sample were African American (67.2%), and had a prior pregnancy (74.0%) and a prior live birth (58.8%) (Table 1). The average age was 29.4 years (standard deviation, 5.4 years), and some women smoked during pregnancy (10.8%) or had current asthma (12.3%). Almost a quarter of the women (23.0%) lived with at least one smoker during pregnancy. Most women were atopic (59.9%), and 28.4% had a dog or cat in the home 12 or more hours per day during pregnancy. All but one pet was in the home for at least 1 month prior to the interview.
The geometric mean for the percentage of Treg cells (% of CD4+ lymphocytes that were CD25+Foxp3+) for all 204 women was 0.83% (95% CI = 0.69%, 1.01%). The levels of Treg cells did not vary by pregnancy history (Table 2), race, baby sex, maternal allergic sensitization, maternal smoke exposure, dog or cat in the home, medications or current asthma status (Table 3), even after adjusting for gestational age at time of blood draw.
In order to assess potential relationships between pet exposure and Treg cell levels we stratified results by whether a pet, either a cat or dog, was present in the home during pregnancy (Table 4). The geometric means of Treg cell levels for the pet exposed and pet unexposed women are presented for different subgroups of women. No association between Treg cells and pets were found among the entire group nor among any of the subgroups evaluated including status by parity, gravidity, ethnicity, antibiotic or allergy medication use, atopy or asthma history.
We also wished to assess whether chronic exposure to pets among pet allergic women was associated with altered Treg cell levels. Among women sensitized to dog, those who had a dog or cat in the home had lower Treg cell levels compared with those who had no dog or cat (geometric mean = 0.32%, 95% CI = 0.11%, 0.93% versus 0.89%, 95%CI = 0.56%, 1.42%; p value = 0.04) (Table 4). However, among women not sensitized to dog, those with a dog or cat in the home had higher Treg cell levels compared with those who did not (1.15%, 95% CI = 0.83%, 1.60% versus 0.75%, 95% CI = 0.57%, 0.99%; p value = 0.09) (Table 4). In a linear regression model of log transformed Treg cell levels, the interaction term between being sensitized to dog and dog or cat exposure was statistically significant (p value = 0.007). These associations were not affected by time of blood draw.
Similar to those sensitized to dog, among women sensitized to cat, those with a pet in the home had lower Treg cell levels compared with those who had no dog or cat (geometric mean = 0.42%, 95% CI = 0.12%, 1.39% versus 1.10%, 95% CI = 0.73%, 1.66%; p value = 0.04) (Table 4). Among women not sensitized to cat, those with a dog or cat in the home had similar Treg cell levels compared with those who did not (1.03%, 95% CI = 0.72%, 1.47% versus 0.71%, 95% CI = 0.54%, 0.94%; p value = 0.15) (Table 4). In a linear regression model of log transformed Treg cell levels, the interaction term between being sensitized to cat and dog/cat exposure was statistically significant (p value = 0.02). Again, these associations were not affected by time of blood draw.
To further describe these observed associations and to evaluate potential confounders, we built final models, including an interaction term between dog or cat in the home, whether the mother was sensitized to dog, and whether the mother was sensitized to cat. There were no confounding covariates (Table 5). The final model indicates a pattern in which Treg cell levels were lowest in those women who were sensitized to either a dog or cat and kept at least one dog or cat in their home during pregnancy.
Of the 42 women with dog sensitization, 9 had at least 1 dog, 2 had at least 1 cat with no dogs and 31 had no dogs or cats. Among the 37 women with cat sensitization, 4 had at least 1 cat, 5 had at least 1 dog with no cats and 28 had no dogs or cats. Limited sample size did not allow us to investigate whether sensitization and keeping of that same type of animal affected levels of Treg cell levels differently than our “unmatched” analyses.
Overall, there was a pattern of lower Treg cell levels in pregnant women who were sensitized to dog or cat and had a dog or cat in their home. Since the data are cross-sectional, we could not assess whether a cause and effect relationship exists between sensitization and lower Treg cell levels among indoor dog or cat owners. Theoretically, low Treg cell levels may increase the risk of sensitization. Alternatively, chronic exposure to allergen among pet-sensitized women may result in lower Treg cell levels. Further speculation based on our data would be premature. The data call to mind the statement from Maggi et al. (2007): “[C]aution should be advised before considering allergic disorders as the simple consequence of defective T-cell regulatory mechanisms in the response to common environmental allergens.” There is complexity in the relationships between exposures and Treg cells that we have only begun to explore.
Our data are unique and not directly comparable to other studies. We used Foxp3 to identify Treg cells in a non-experimental setting and in unstimulated blood samples from women who were exposed or not exposed to dogs or cats in their usual environment. This approach may better reflect actual environmental exposure compared with experimental analyses in which ex-vivo cell stimulation with specific antigens may not directly translate to usual human exposure. Also, we did not directly measure the suppressive function of the Treg cells following allergen-specific stimulation, a methodologic approach that could have been informative.
Though our study population included a large proportion of African American women (67.2%), the data do not indicate that there are racial differences in the associations between pets and Treg cells. Thus we do not think that our results limit comparison to other study populations.
As with any study, this work had limitations. A larger sample size would have improved precision of these estimated associations of Treg cell levels to the likelihood of sensitization among those who lived with dogs or cats. For example, if non-sensitized women with chronic exposure to pets had high Treg cell levels, this could indicate that the presence of a pet may provoke a functional immune response to increase Treg cell levels in the presence of chronic exposure to a potential allergen. In contrast, those who were sensitized may fail to increase Treg cell levels resulting in progression to the production of allergen-specific IgE and immediate hypersensitivity.
An additional limitation is that we did not validate maternal report of pet exposure. However, it is unlikely that accuracy of pet reporting varied by Treg cell levels, which would have been unknown to the women. Thus, biased results due to inaccurate pet reporting are unlikely.
Although the mechanisms are not known, the role of prenatal exposure to pets on the risk of allergic diseases in children has come under study. Since the role of Treg cells on allergic disease risk has been the recent focus of much discussion (Shi HZ, et al. 2005; Seroogy CM and Gern JE 2005; Bacchetta R, et al. 2007; Bellinghausen I, et al. 2005; Lee JH, et al. 2007; Maggi L, et al;. 2007), we chose to study Treg cells in pregnant women to further investigate the potential mechanisms underlying the role of prenatal exposures in early life allergic disease risk. We view this presented work as an early step in further understanding the relationship between maternal environmental exposures, maternal immune function and the immune function and allergic disease risk of her newborn. The results seen in these pregnant women may facilitate the study of risk factors in their children’s health, but may not be the same in women who are not pregnant. Future analyses will further study the relationship between maternal immunity during pregnancy and the child’s early life health and immune parameters, including cytokines and IgE levels.
This work was funded by NIAID (AI070606 and AI069271) and the Fund for Henry Ford Hospital.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.