|Home | About | Journals | Submit | Contact Us | Français|
Bisphenol A (BPA) is a high–production volume chemical classified as an environmental estrogen and used primarily in the plastics industry. BPA’s increased usage correlates with rising BPA levels in people and a corresponding increase in the incidence of asthma. Due to limited studies, the contribution of maternal BPA exposure to allergic asthma pathogenesis is unclear. Using two established mouse models of allergic asthma, we examined whether developmental exposure to BPA alters hallmarks of allergic lung inflammation in adult offspring. Pregnant C57BL/6 dams were gavaged with 0, 0.5, 5, 50, or 500 μg BPA/kg/day from gestational day 6 until postnatal day 21. To induce allergic inflammation, adult offspring were mucosally sensitized with inhaled ovalbumin containing low-dose lipopolysaccharide or ip sensitized using ovalbumin with alum followed by ovalbumin aerosol challenge. In the mucosal sensitization model, female offspring that were maternally exposed to ≥ 50 μg BPA/kg/day displayed enhanced airway lymphocytic and lung inflammation, compared with offspring of control dams. Peritoneally sensitized, female offspring exposed to ≤ 50 μg BPA/kg/day presented dampened lung eosinophilia, compared with vehicle controls. Male offspring did not exhibit these differences in either sensitization model. Our data demonstrate that maternal exposure to BPA has subtle and qualitatively different effects on allergic inflammation, which are critically dependent upon route of allergen sensitization and sex. However, these subtle, yet persistent changes due to developmental exposure to BPA did not lead to significant differences in overall airway responsiveness, suggesting that early life exposure to BPA does not exacerbate allergic inflammation into adulthood.
Asthma is a global disease with no definitive cause or cure. This complex, chronic condition is described by variable symptoms, including reversible airway obstruction, wheezing, chest tightness, and shortness of breath. In a 2010 assessment, asthma affected over 300 million people (GINA, 2011), and in 2009, it claimed 250,000 lives worldwide (GINA, 2009), making its study a global priority. Allergic asthma is the most common form of asthma and describes about 60% of all asthmatics (Milgrom, 2003). Allergic asthma is triggered by the response to allergens, such as pollens or house dust mites. Asthma prevalence has increased since the 1970s, a period that also saw an increased use of numerous chemicals in manufacturing processes, with increased exposure of the general population (Kwak et al., 2009; Vollmer et al., 1998). This association has raised concern about the potential for a cause-and-effect relationship between the widespread use of some chemicals, particularly those that are not only used in manufacturing and industrial processes but are detected in a large proportion of the population, with the increased incidence or severity of diseases such as asthma. One chemical that has garnered considerable attention of late is bisphenol A (4,4′-(propane-2,2-diyl) diphenol or BPA) (Braun and Hauser, 2011; Hengstler et al., 2011; Vandenberg et al., 2007).
BPA is composed of two functional phenol groups, giving it properties that are useful in many manufacturing processes and commonly used consumer products. BPA’s primary use is to make polycarbonate plastics. Other applications include epoxy resins lining food and beverage containers, dental sealants, carbonless copy and thermal papers, and as a precursor in flame retardants (Debenest et al., 2010; Liao and Kannan, 2011; Lopez-Cervantes and Paseiro-Losada, 2003; Suzuki et al., 2000). Concerns about BPA stem from several factors, including the knowledge that it is one of the highest production volume chemicals in the world, and products that contain BPA often break down, releasing monomers (Cannon et al., 2000). Biomonitoring surveys by the U.S. Centers for Disease Control and Prevention have found BPA in blood and urine samples from 95% of the U.S. population (Calafat et al., 2005). Additional concern arises because BPA is an endocrine disruptor. Specifically, BPA binds to estrogen receptor-α (ER-α) and ER-β. Although BPA’s affinity for ER-α and ER-β is ≥ 10,000-fold lower than 17β-estradiol (Kuiper et al., 1998), studies suggest that BPA may initiate biological effects through both ER-dependent and ER-independent signaling pathways (Miyakoshi et al., 2009; Singleton et al., 2004). Multiple reports, in a variety of animal models, paint a concerning, albeit inconsistent, picture of the potential for BPA to contribute to negative health effects, including chromosomal and reproductive system abnormalities, impaired brain, neurological, and cardiovascular functions, as well as possibly contributing to diabetes and obesity (Allard and Colaiacovo, 2010; Larocca et al., 2011; Leranth et al., 2008; Melzer et al., 2010; Sakurai et al., 2004). The differential effects of estrogen and the modification of ER signaling on allergic asthma have also been reported although the mechanisms by which estrogens influence this disease remain to be determined (de Oliveira et al., 2010; Shah et al., 2010). These reports highlight the potential for BPA to perturb human health and contribute to diseases, such as allergic lung inflammation and asthma, although the underlying mechanisms by which BPA could cause these effects remain an area of active research.
Few epidemiological studies have explicitly examined the relationship between BPA exposure and the developing human immune system although contact dermatitis and asthma have been reported in industrial workers exposed to BPA and BPA conjugates, suggesting a possible link between chronic BPA exposure and immune-mediated diseases (Hannu et al., 2009; Jensen and Andersen, 2003). Several in vitro and animal studies have examined the consequences of BPA exposure on mast cell degranulation, lymphocyte proliferation, antibody responses, and regulatory T cells, with data suggesting that the developing immune system may be a particularly sensitive target of BPA (Midoro-Horiuti et al., 2010; Nakajima et al., 2012; Shin-ichiro et al., 2007; Yan et al., 2008; Yoshino et al., 2004). However, there are still gaps in our knowledge concerning BPA’s effects in the context of allergic asthma. This study examines, for the first time, whether maternal exposure permanently affects metrics of allergic inflammation development/severity into adulthood and whether gender influences the potential proallergic effects of BPA (Melgert et al., 2005; Shah et al., 2010). Furthermore, we examined whether the route of allergen sensitization affects any BPA-related differences.
Animals. C57BL/6 mice (6–8 weeks; NCI, Bethesda, MD) were housed in prewashed polysulfone microisolator cages under pathogen-free conditions. Mice received soy-free, AIN76-semi-PD1RR chow (Test Diet, Richmond, IN) to avoid phytoestrogenic effects that might mimic possible estrogenic effects of BPA. Glass water bottles and reverse osmosis-purified water were used. Paired mice were checked daily for vaginal plugs, indicating pregnancy. Pregnant females were housed singly and initially treated by oral gavage with 50 or 500 μg BPA/kg/day (d) (Sigma-Aldrich, St Louis, MO) or peanut oil vehicle control, beginning on gestational day 6 and continuing daily through postnatal day (PND) 21. An estrogen control was not used as BPA may act via other receptors in addition to ERs; thus it is not clear that BPA and natural estrogens will have the same effects (Chung et al., 2011). The results of these initial experiments and new information in the literature prompted the subsequent examination of maternal exposure to lower BPA doses (0.5 and 5 μg BPA/kg/day). Dosing to the dam was initiated on the 6th day of pregnancy to eliminate any possible effects of BPA on implantation, which occurs in the mouse between the 4th and 5th day of gestation, and all pups were weaned on PND 21, at which time all maternal transfer of BPA ceased. Offspring were not directly given any BPA. The selected dose range spans the current tolerable daily intake of 50 μg BPA/kg/day (USEPA, 2010). Maternal treatment with these doses of BPA did not affect offspring bodyweight, which was monitored over the life span of the offspring, pregnancy duration, litter size, or the sex ratio of litters (Roy et al., 2012).
Using a subset of pregnant dams, we determined the amount of circulating BPA detected following this oral dose of BPA. BPA was measured at the Laboratory of Analytical Chemistry, New York State Department of Health, using high performance liquid chromatography coupled with electrospray triple-quadrupole mass spectrometry (Padmanabhan et al., 2008). Four hours after gavage of pregnant mice, free and total BPA were measured (7 mice per treatment group). The mean circulating levels of total BPA in pregnant mice given 50 µg/kg/day were 2.941ng/ml, whereas mean circulating BPA levels in vehicle control–treated dams was 0.434ng/ml (p = 0.0173). Interestingly, in the BPA-treated group, none of the samples tested were below the limit of quantitation (LOQ), whereas in the samples from the group dosed with peanut oil vehicle, 4 out of 7 samples were below the LOQ.
On PND 21, offspring were weaned and housed in same-sex groups under the BPA-controlled conditions described above until maturity (6–10 weeks). Over the course of this research, several separate cohorts of maternally exposed mice were generated over a period of many months. Vehicle-treated dams were included in all pregnancy cohorts, so that BPA exposed offspring and vehicle-exposed offspring were age matched and derived from a single shipment of nulliparous mice (i.e., each batch of mice that we used as dams). Our experimental strategy utilized the available offspring in the various exposure groups as they were available. For all presented data, the offspring of vehicle- and BPA-treated dams were impregnated and treated in the same time frame. For experiments, age and sex-matched, maternally exposed, mature offspring were derived from separate vehicle- or BPA-treated dams. For example, a group with an N = 5 contained five mice, each from a different exposed dam of the same vehicle or BPA dose. Animal work was done in accordance with protocols approved by the University of Rochester Animal Care and Use Committee.
Mucosal sensitization model. Mice were anesthetized with avertin ((2,2,2-tribromoethanol); Aldrich, Milwaukee, WI) and intratracheally (i.t.) sensitized on days 0, 1, and 2 with 100 μg of endotoxin-depleted ovalbumin (OVA), with, or without, 100ng of Escherichia coli lipopolysaccharide (LPS) (O55:B5; Sigma-Aldrich) in PBS (50 μl/mouse). OVA (Grade V, Sigma-Aldrich) was endotoxin-depleted using Endotoxin Detoxi-Gel (Pierce, Rockford, IL). This protocol results in TH2-type allergic airway inflammation following aerosol OVA challenge (Eisenbarth et al., 2002). Mice were challenged twice per day on days 14, 15, and 16 with aerosolized 1% OVA in PBS (1h), using a jet nebulizer (DeVilbiss, Somerset, PA) and constant airflow rate. Forty-eight hours after the final aerosol challenge, mice were sacrificed by avertin overdose. We chose the 48-h time point based on prior studies and to better compare outcomes with mice in the peritoneal sensitization groups (Hogaboam et al., 2008; Lawrence et al., 2008).
Peritoneal sensitization model. Mice were ip sensitized on days 0 and 4 with PBS or 100 μg OVA in PBS with 100 μl Imject Alum (Pierce; 200 μl/mouse). Although this is an artificial route of allergen encounter, this widely used model relies on the robust TH2-promoting properties of injected alum. Mice were challenged once (1h), on day 11, with aerosolized 1% OVA in PBS, using a challenge chamber, jet nebulizer, and constant airflow rate. Forty-eight hours postaerosol challenge, mice were sacrificed using avertin overdose. In previous studies, inflammatory cells, cytokines, and OVA-specific IgE were detected 48h after single aerosol challenge (Hogaboam et al., 2008; Lawrence et al., 2008). Based on this work, we elected to use the time point for the current study.
Histological analysis. Lungs were fixed in situ with 10% neutral-buffered formalin, embedded in paraffin, and 5-μm sections were stained with hematoxylin and eosin (H&E). Coded samples were scored for tissue inflammation by at least two different scientists using the following scale: 0, no airway involvement, few/no inflammatory cells in tissue; 1, limited inflammation, small number of airways with inflammatory cell clusters, most airways have no inflammation; 2, modest inflammation, several airways with small inflammatory cell clusters, no areas of intense infiltrate; 3, moderate-to-intense inflammation, many airways with inflammatory cells and occasional areas of intense infiltrate; 4, significant airway inflammation with multiple intense foci, massive perivascular/periairway inflammation, most/all airways involved. The combined scores of five lobes resulted in an overall number. The maximum possible inflammation score for each whole lung was 20.
Bronchoalveolar lavage. Lungs were lavaged as previously described (Warren et al., 2000). The supernatant from the first 1ml wash was frozen at −80°C for cytokine analysis, whereas recovered cells from all washes were pooled. A TC-10 (Bio-Rad, Hercules, CA) was used to calculate the total number of cells recovered. Bronchoalveolar lavage (BAL) cells (5×104) were spun onto glass microscope slides using a cytospin (Shandon Scientific, Waltham, MA), fixed, and stained with H&E.
Flow cytometric analysis of lung-derived immune cells. The pulmonary vasculature was perfused (PBS) via the right ventricle, and immune cells were obtained from collagenase-digested whole lung as described previously (Neff-LaFord et al., 2007). Leukocytes were filtered, washed, and counted, and nonspecific binding was blocked using anti-CD16/32 and rat IgG. Subsequently, cells were stained with fluorochrome-conjugated antibodies against cell surface CD4, CD8, CD11b, CD11c, CD25, CD45.2, MHCII, Siglec-F, or the intracellular protein, FoxP3 (eBioscience, San Diego, CA; BD Pharmingen, San Jose, CA). For the latter, cells were permeabilized with saponin postfixation and incubated with an anti-FoxP3 antibody. A live/dead stain (Invitrogen, Carlsbad, CA) was used to selectively gate on live cells during analyses. Events were acquired using an LSR II cytometer (BD Biosciences, Franklin Lakes, NJ) with subsequent analysis using FlowJo 8.8.7 (Tree Star, Ashland, OR).
OVA-specific antibody measurements. Circulating OVA-specific IgE levels were analyzed by stacking ELISA (Hogaboam et al., 2008). Serial dilutions of OVA-specific IgE (GeneTex, Irvine, CA) and plasma from unexposed/unsensitized/unchallenged C57BL/6 mice were used as positive and negative controls, respectively. Plasma from exposed mice was serially diluted, and biotinylated anti-mouse IgE subclass antibody (eBioscience) was used, in combination with avidin-peroxidase and ABTS substrate (2,2′-azino-bis(3-ethylbenzothiazoline-6-sulphonic acid, Sigma-Aldrich) to detect OVA-specific IgE in the samples.
Measurement of airway hyperresponsiveness. Airway responsiveness to methacholine was assessed using the FlexiVent, invasive, airway plethysmography system (Scireq, Chandler, AZ) and analyzed using Lab Chart Version 6 software (ADInstruments, Inc., Colorado Springs, CO). Mice were paralyzed with succinylcholine, anesthetized with avertin, tracheostomized, and intubated with a cannula connected to a ventilator. The system was calibrated for airflow and pressure. Pressure measurements, postbaseline, were obtained in response to airway challenge with saline and increasing concentrations of aerosolized methacholine in saline (0.1, 0.3, 1, 3, 10, 30, and 100mg/ml) using a jet nebulizer. Resistance and elastance measurements to describe overall airway responsiveness were derived from the pressure measurements. Resistance is a measurement describing a reduction in airflow due to factors such as bronchoconstriction, whereas elastance describes the “stretching/recoil” potential of the airway.
Statistical analysis. Given that the dams, not offspring, were directly exposed, all offspring from a litter are considered as a single statistical unit. Statistical analyses were performed on all measured endpoints using JMP software (Version 9, SAS, Cary, NC). Mean differences between independent variables were compared among treatment groups using two-way ANOVA, followed by a post hoc test (Tukey’s-Kramer HSD [Honestly Significant Difference]). Also, nonparametric analyses of data were performed using a Kruskal-Wallis test. Differences between two groups at a single point in time were evaluated using a Student’s t-test or Tukey’s-Kramer HSD test. Differences were considered significant when p values were less than 0.05. All statistical analyses were conducted within animals of the same sex.
In light of the disparity among allergic inflammation mouse models currently employed in research and our aim to better understand the potential long-term impact of maternal exposure to BPA on allergic inflammation, we utilized two established mouse models of allergic inflammation. These models differ in the route of antigen sensitization, but share a similar model allergen (OVA) and strategy for allergen challenge (OVA aerosolization). In one model, sensitization is via the airway, and in the other model, OVA is administered into the peritoneum. Airway sensitization is a physiological route of mucosal allergen exposure, and low-dose LPS is thought to act as a TH2-promoting adjuvant, activating airway epithelial cells and lung-resident dendritic cells (Eisenbarth et al., 2002; Tan et al., 2010). In contrast, inhalation of endotoxin-depleted OVA alone usually results in tolerance to the antigen. In the more widely used peritoneal sensitization model, OVA is suspended in the adjuvant alum. The precise mechanisms by which alum and LPS act as adjuvants in these models remain to be fully determined; however, it is thought that they cue innate and inflammatory responses differently and may even influence aspects of antigen processing and presentation (Eisenbarth et al., 2002; Kool et al., 2008a,b). In both models, OVA sensitization and challenge were performed in immunologically mature offspring of both sexes.
In the mucosal sensitization model, mice inhaled either endotoxin-depleted OVA alone, or with low-dose LPS as adjuvant, followed 2 weeks later by OVA aerosol challenge. We observed that adult female offspring, but not adult male offspring, that were maternally exposed to 50 and 500 μg BPA/kg/day developed slightly enhanced lymphocyte airway infiltration when compared with offspring of vehicle control–treated dams (Fig. 1). Furthermore, when we examined lung histopathology, only adult female offspring demonstrated consistently enhanced whole-lung inflammation compared with adult offspring of vehicle control–treated dams (Fig. 2). This enhanced airway and whole-lung inflammation that occurred in adult female offspring sensitized using OVA alone suggests that the threshold for mounting a response to inhaled adjuvant-free OVA, as a model allergen, was lower in these mice. In contrast, the effects of maternal exposure to BPA on lung inflammation in the male offspring were less pronounced, with some statistically significant differences in mean score of the OVA-alone groups. However, the overall level of lung inflammation was similar among the groups of male offspring, regardless of maternal exposure and whether sensitization was with OVA or OVA + LPS. Regardless of BPA dose, neither female nor male adult offspring presented differences in airway eosinophilia or neutrophilia when sensitized with OVA + LPS. Of note, the statistically significant reduction in airway neutrophilia observed in adult male offspring (Fig. 1E) is unexplained, but it was consistent among experiments.
We next further characterized the lymphocytic lung inflammation. Given that maternal BPA exposure has been shown to influence helper (Yoshino et al., 2004) and regulatory CD4+T cell populations (Yan et al., 2008), we examined common lung T-cell populations by flow cytometry. There were no differences in the number (or percentage) of CD4+, CD8+, or CD4+CD25+FoxP3+ T lymphocyte subpopulations in adult offspring that were maternally exposed to 50 or 500 μg BPA/kg/day, when compared with the adult offspring of corresponding vehicle-treated dams (Fig. 3). Furthermore, although we noted differences in some metrics of allergic inflammation that could be attributed to maternal BPA exposure, we did not observe such differences in disease-relevant lavage fluid cytokines, such as IL-5 and IL-17, or circulating OVA-specific IgE levels (data not shown).
U.S. Federal guidelines currently put the daily upper limit of safe human exposure to BPA at 50 μg/kg/day (USEPA, 2010). Hence, we selected this daily dose and an order of magnitude higher for our aforementioned initial investigations. However, several animal studies have demonstrated that acute or maternal exposure to BPA at levels < 50 μg/kg/day can have deleterious effects in various disease models (Al-Hiyasat et al. 2002; Chitra et al. 2003; Maffini et al. 2006; Sakaue et al. 2001). In light of these studies, we investigated the consequences of maternal exposure to 10 and 100 times lower levels of BPA (0.5 or 5 μg BPA/kg/day) in adult offspring, using the mucosal sensitization model described above. When comparing like-sensitization/challenge groups, there were no significant differences in total airway eosinophils, neutrophils, or lymphocytes in adult offspring (Fig. 4). Furthermore, the overall magnitude of the response, as measured by the number of airway eosinophils, neutrophils, and lymphocytes, generally appeared to be less in adult male than in adult female offspring. As previously observed with higher BPA doses, an investigation of the CD4+, CD8+, and CD4+CD25+FoxP3+ lymphocyte populations in the lung revealed few differences between like-sensitization/challenge groups in adult offspring that were previously maternally exposed to 0, 0.5, or 5 μg BPA/kg/day (Fig. 5). Moreover, similar to the higher maternal BPA doses tested, although we noted subtle, insignificant differences in some metrics of allergic inflammation that could be attributed to maternal BPA exposure, we found no differences in cytokines, such as IL-5 and IL-17, in lung lavage fluid, or circulating OVA-specific IgE levels (data not shown). Previous allergic asthma-modeling studies in mice (Melgert et al., 2005) have shown that female mice are more prone to display characteristics of asthma than male mice. Despite the proposed link between estrogens and asthma, these data show that maternal BPA exposure, at the doses used, does not alter this sex-specific characteristic and that maternal exposure to BPA does not promote allergic inflammation in adult male offspring.
In the commonly used peritoneal sensitization model, mice were ip sensitized with OVA + alum followed by OVA aerosol challenge. This leads to OVA uptake by spleen antigen-presenting cells and is believed to promote robust TH2 immune responses in both inflammasome-dependent and -independent manners (Eisenbarth et al., 2008; Kool et al., 2008a). Interestingly, and in contrast to developmentally exposed offspring that were sensitized with inhaled OVA plus low-dose LPS, we found that adult female offspring that were maternally exposed to 0.5, 5, or 50 μg BPA/kg/day displayed significantly dampened airway eosinophilia, compared with adult female offspring exposed to either vehicle control or 500 μg BPA/kg/day (Fig. 6A). There were no significant, BPA-related differences in airway neutrophil or lymphocyte frequency in these sensitized/challenged mice (Figs. 6B & C). The attenuated airway inflammation observed was similarly reflected in the total number of lung-derived eosinophils, wherein adult female offspring that were maternally exposed to 0.5–50 μg BPA/kg/day had significantly fewer CD11b+CD11c−Siglec-F+cells (eosinophils) compared with adult female offspring of dams that received vehicle or 500 μg BPA/kg/day (Fig. 6D and Supplementary fig. 1). Although mean differences did not reach statistical significance for any comparison, the total number of CD4+ T cells in the lung displayed a similar reduced trend for the same maternal BPA exposure groups (Fig. 6E). Furthermore, circulating OVA-specific IgE levels were significantly lower in adult female offspring that were maternally exposed to 0.5, 5, or 500 μg BPA/kg/day (Fig. 6F), compared with vehicle-exposed offspring.
Given that we observed statistically significant differences in multiple endpoints of allergic inflammation using the peritoneal sensitization model, we wanted to determine whether these differences affected overall airway mechanics. Airway hyperresponsiveness was measured using invasive plethysmography. This measurement relates to a clinically relevant human endpoint, and because multiple pathways contribute to changes in this measurement, the results may reveal differences that are not apparent in the examination of individual phenotypic or cellular endpoints. As shown in Figure 7, there were no significant alterations in airway hyperresponsiveness in sensitized/challenged adult female offspring that were maternally exposed to BPA, compared with vehicle control. This suggests that although maternal exposure to BPA caused subtle changes in some metrics of allergic airway inflammation, the net effect of developmental exposure to BPA on overall airway mechanics during antigen challenge remain unaffected.
In light of increasing concern about human exposure to BPA from a variety of sources and the reported correlation of this exposure with increasing rates of allergic asthma, we investigated the hypothesis that maternal exposure to BPA affects the development and severity of allergic inflammation into adulthood using two distinct mouse models of allergic asthma. In addition to using the conventional and commonly used OVA + alum, peritoneal sensitization model, we felt it was critical to investigate the consequences of maternal BPA exposure on allergic inflammation in the context of a model that uses mucosal sensitization, a more physiological route of antigen encounter. Importantly, although antigen sensitization differs in the models, both models used aerosol challenges with antigen to induce allergic inflammation. Furthermore, we wanted to determine whether any impact that gestational exposure to BPA may have on allergic inflammation would persist into immunologically mature, adult animals.
The major findings of this study were that the effects of maternal exposure to a range of doses of BPA on allergic inflammation in adult offspring were subtle. Many immune endpoints were unaffected by developmental exposure. For the endpoints that were altered by developmental exposure to BPA, the effects observed were qualitatively different and critically dependent on the route of antigen sensitization and the sex of the offspring. Furthermore, when tested, the subtle effects on allergic inflammation caused by maternal BPA exposure did not have a significant impact on airway function, compared with vehicle control–exposed mice. Interestingly, in the more widely used peritoneal sensitization model of allergen immunization, maternal exposure to 0.5, 5, and 50 μg BPA/kg/day attenuated lung eosinophilia in adult female offspring following OVA challenge. In contrast, when antigen sensitization occurred via the mucosal route, maternal exposure to BPA resulted in enhanced airway lymphocytic inflammation in adult female offspring, which occurred even in the absence of inhaled low-dose LPS adjuvant. The differences observed between these two mouse models of allergic inflammation suggest that maternal BPA exposure differentially influences immune responses into adulthood, but the nature of this effect depends on the site and context of initial antigen encounter. Additionally, in mice that were maternally exposed to BPA, we have shown that these influences did not impact the airway response in mice that underwent the peritoneal model of allergic inflammation.
The mucosal sensitization model involves toll-like receptor (TLR)-4, CD14, and LPS-binding protein in response to LPS, thereby facilitating the OVA-specific response. It has been shown that the amount of LPS influences the type of inflammatory response generated (pro-TH1 or pro-TH2), thus influencing the development and severity of allergic asthma (Eisenbarth et al., 2002; Tan et al., 2010). In contrast, insoluble aluminum salts, like those used in the peritoneal sensitization model, are the principle adjuvants used in clinical vaccine preparations and have been shown to elicit TH2-type-specific immune responses to allergens (Grun and Maurer, 1989). The mechanism by which alum acts as an adjuvant remains controversial. However, multiple studies indicate that aluminum salts elicit their actions through TLRs. One such study showed that MyD88-deficient mice produced normal amounts of IgG1 but excessive amounts of IgE in response to antigen immunization in aluminum salts (Schnare et al., 2001). Another study showed enhanced antibody responses in MyD88/TRIF double-deficient mice, where no TLR signaling could occur (Gavin et al., 2006). There is controversy, some showing dependence (Kool et al., 2008b) and others showing independence (McKee et al., 2009) of the inflammasome in this process. Thus, it is possible that maternal BPA exposure influences these different mechanisms of allergen uptake and presentation, which offers an explanation for differential effects observed when we used these distinct models for antigen sensitization.
An unexpected and important finding of our studies using the peritoneal model was that maternal BPA exposure dampened multiple features of adult allergic inflammation. Other recent studies (Midoro-Horiuti et al., 2010; Nakajima et al., 2012) utilized a similar OVA + alum peritoneal sensitization model but reported enhanced airway hyperresponsiveness, eosinophilia, and circulating OVA-specific IgE in neonatal offspring that were maternally exposed to 10 μg/ml BPA in the dams’ drinking water. Several distinctions in study design, such as the maternal dose of BPA, mouse strain, and no segregation of male and female offspring for neonatal sensitization, may contribute to the differences in our findings and those reported by this other group. However, the age difference of the mice at the time of antigen sensitization and challenge offers a likely and very important distinction between these studies and ours. The development of the fetal immune system is under the integrated control of multiple systems, including those involving hormones, and one possibility is that chemical exposure is more detrimental during this period. In fact, there is evidence that the developing immune system is considerably more sensitive to perturbation than the fully mature immune system (Hogaboam et al., 2008; Luebke et al., 2004). It would appear that although all studies of maternal exposure to BPA demonstrated modulatory effects on allergic inflammation in the offspring, the exacerbation of overall airway responsiveness observed in neonatally sensitized mice appears to result in a different outcome than when sensitization occurs later in life. When taken together, it may be that gestational exposure to BPA leads to transient early life changes in allergic lung inflammation, but that this effect wanes with age and may not be permanent. Furthermore, the metabolic capacity of the exposed dam may also determine exposure to the fetus (Doerge et al., 2011).
Several other pieces of evidence support the idea that there may be a developmental window in which BPA exposure affects allergic sensitization. A recent study showed that Ugt2b1 expression, an isoform of UDP glucuronosyltransferase that is related to clearance of BPA in rats, is undetectable in mouse fetuses and newborn pups but approaches adult levels by PND 25 (Nakajima et al., 2012). This suggests that BPA conjugation mechanisms change with developmental state, which may influence the potential for BPA to modulate biological systems during different windows of development. Further support for a developmental window for BPA to affect the immune system comes from separate studies, in which we examined whether daily, acute, oral exposure of adult mice to BPA alters allergic lung inflammation in either the mucosal and peritoneal sensitization models. Our data revealed no effect of BPA treatment on allergic inflammation using either model of allergic inflammation (unpublished observations). Consistent with these findings, a recent study found no correlation between urinary BPA levels and allergy/hay fever (Rees Clayton et al., 2011). Taking all of the above into consideration suggests that the BPA-related subtle effects we have observed in maternally exposed adult mice occurred during development; however, these changes did not confer a permanent detrimental impact on allergic inflammation, nor was airway responsiveness adversely affected.
Allergic lung inflammation is just one of many facets of immune system responsiveness. Interestingly, in parallel animal studies, we recently reported that maternal exposure to BPA exposure did not compromise virus-specific adaptive immunity against primary and recall influenza A virus infection in adult offspring. However, we observed that infected adult mice that were developmentally exposed to BPA had a transient reduction in the extent of infection-associated pulmonary inflammation and anti-viral gene expression in lung tissue (Roy et al., 2012). These observations are consistent with the idea that early life exposure to BPA affects aspects of tissue inflammation, even in the absence of modulating adaptive immune responses, and suggest that early life exposure to BPA may affect epithelial cells resident in peripheral tissues, such as the lung.
In summary, we present here a novel evaluation of the effects of maternal exposure to four different doses of BPA and report the effect of this exposure on several measures of allergic lung inflammation in adult offspring. In contrast to reports that maternal, and even gestational only, exposure to BPA alters aspects of lung inflammation and airway responsiveness when antigen sensitization and challenge occur shortly after birth, using multiple doses of BPA to the dam, our data reveal that maternal exposure to BPA has only subtle effects on allergic inflammation in adult offspring, with some endpoints unaffected at any dose. Moreover, when observed, the direction of change and magnitude of the effect of developmental exposure to BPA are critically dependent upon route of antigen sensitization and sex of the adult offspring. Finally, it is important to consider that these subtle effects, although persistent, did not lead to significant differences in overall airway responsiveness, suggesting that early life exposure to BPA may not exacerbate allergic inflammation into adulthood.
National Institutes of Health/National Institute of Environmental Health Sciences (RC2-ES018750 to BPL and SG, R01-ES017250 to BPL, R01-HL097141 to BPL, P30ES01247 to BPL, T32-HL066988 to SMB, T32-ES07026 to AR).
We thank Jill Gresens for her assistance with mouse colony breeding and BPA dosing of dams and Dr Kannan Kurunthachalam (Wadsworth Center, New York State Department of Health) for measuring BPA levels in mouse plasma samples.