Maternal Bisphenol A Exposure Promotes the Development of Experimental Asthma in Mouse Pups

Background We recently reported that various environmental estrogens induce mast cell degranulation and enhance IgE-mediated release of allergic mediators in vitro. Objectives We hypothesized that environmental estrogens would enhance allergic sensitization as well as bronchial inflammation and responsiveness. To test this hypothesis, we exposed fetal and neonatal mice to the common environmental estrogen bisphenol A (BPA) via maternal loading and assessed the pups’ response to allergic sensitization and bronchial challenge. Methods Female BALB/c mice received 10 μg/mL BPA in their drinking water from 1 week before impregnation to the end of the study. Neonatal mice were given a single 5 μg intraperitoneal dose of ovalbumin (OVA) with aluminum hydroxide on postnatal day 4 and 3% OVA by nebulization for 10 min on days 13, 14, and 15. Forty-eight hours after the last nebulization, we assessed serum IgE antibodies to OVA by enzyme-linked immunosorbent assay (ELISA) and airway inflammation and hyperresponsiveness by enumerating eosinophils in bronchoalveolar lavage fluid, whole-body barometric plethysmography, and a forced oscillation technique. Results Neonates from BPA-exposed mothers responded to this “suboptimal” sensitization with higher serum IgE anti-OVA concentrations compared with those from unexposed mothers (p < 0.05), and eosinophilic inflammation in their airways was significantly greater. Airway responsiveness of the OVA-sensitized neonates from BPA-treated mothers was enhanced compared with those from unexposed mothers (p < 0.05). Conclusions Perinatal exposure to BPA enhances allergic sensitization and bronchial inflammation and responsiveness in a susceptible animal model of asthma.


Research
The dramatic increase in the prevalence of childhood asthma in developed coun tries over the last two to three decades sug gests that environmental changes related to industrializa tion may be involved in this epidemic (McCoy et al. 2005). One public health approach to understanding the asthma epidemic would be to identify environ mental exposures that increased during the period when asthma prevalence began to increase in industrialized countries and to test their effects in experimental animals. Those compounds that enhance various features of asthma might be candidates for promoting the development or morbidity of asthma in humans. However, even the major reductions in some air pol lutants accomplished in many areas have yet to have major impacts on the prevalence or even the morbidity of asthma. This experi ence suggests that the multi factorial nature of the disease makes it difficult to see the effects of incremental improvements in our environ ment when exposures to other types of toxi cants persist or even increase.
We recently found that exposing cultured mast cells to estradiol (E 2 ) strongly potentiates the synthesis and release of allergic mediators, acting through a membrane form of estrogen receptor α (ERα) (Zaitsu et al. 2007). We also examined the effects of environ mental estrogens, alone and in combination with physiologic concentrations of E 2 , on the acti vation of a human mast cell line and primary cultures of murine mast cells . Like E 2 , low concentrations of envi ronmental estrogens caused a rapid, partial degranulation of mast cells. Exposing HMC1 cells (a human mast cell line) to a combination of suboptimal concentrations of E 2 and an environmental estrogen had an additive effect on degranulation. Environmental estrogens also enhanced the release of βhexosaminidase induced by allergen crosslinking of IgE (immunoglobulin E) on the surface of these cells. Bonemarrow-derived mast cells defi cient in ERα expression had significantly reduced responses to some concentrations of environmental estrogens, suggesting that the degranulating activity of environmental estro gens on mast cells is mediated, at least in part, through ERα and that the dose response is non monotonic, as previously described for non genomic responses to endogenous and environmental estrogens (Bulayeva and Watson 2004). Recent studies suggest that mast cells, in addition to their role as effects of allergic reactions, also contribute to the process of allergic sensitization (Nakano et al. 2009;Sayed and Brown 2007). We therefore exam ined whether environmental estrogens might enhance allergic sensitization and inflamma tion, thereby promoting the develop ment of asthma and other allergic diseases.
To better understand the potential for environmental estrogens to promote the develop ment of childhood asthma, in the pres ent study we tested the hypothesis that prenatal and/or neonatal exposures to one of the most ubiquitous environmental estrogens, BPA, enhance the development of allergic asthma in a susceptible animal model. BPA is a substrate of polycarbonate plastics and has been pro duced in increasingly large quantities in indus trialized countries since the 1950s. BPA is used to form plastic bottles, as a lining for food and beverage cans, and as a flame retardant and is a component of dental fillings. Interestingly, the increase of asthma prevalence among children started in the 1970s (Vollmer et al. 1998), 20 years (approximately one generation) after largescale BPA production started in several industrialized regions of the world.

Materials and Methods
Animals. Female BALB/c mice were obtained from Harlan (Houston, TX) and were housed in pathogenfree conditions in the animal research facility of the University of Texas Medical Branch (UTMB; Galveston, TX) in accordance with the National Institutes of Health and UTMB institutional guidelines for animal care. Each mother and her litter were housed separately and fed a caseinbased diet (Research Diet, New Brunswick, NJ) to elimi nate estrogenic effects in the typical soybased mouse diet from 1 week before BPA load ing until the end of the study (Curran et al. 2004). We used animal cages made of poly sulfone (Tecniplast, Buguggiate, Italy). We tested these cages for contamination with BPA by adding water to a cage and maintaining it Background: We recently reported that various environmental estrogens induce mast cell degranulation and enhance IgE-mediated release of allergic mediators in vitro. oBjectives: We hypothesized that environmental estrogens would enhance allergic sensitization as well as bronchial inflammation and responsiveness. To test this hypothesis, we exposed fetal and neonatal mice to the common environmental estrogen bisphenol A (BPA) via maternal loading and assessed the pups' response to allergic sensitization and bronchial challenge. Methods: Female BALB/c mice received 10 µg/mL BPA in their drinking water from 1 week before impregnation to the end of the study. Neonatal mice were given a single 5 µg intraperitoneal dose of ovalbumin (OVA) with aluminum hydroxide on postnatal day 4 and 3% OVA by nebulization for 10 min on days 13, 14, and 15. Forty-eight hours after the last nebulization, we assessed serum IgE anti bodies to OVA by enzyme-linked immunosorbent assay (ELISA) and airway inflammation and hyperresponsiveness by enumerating eosinophils in bronchoalveolar lavage fluid, wholebody barometric plethysmography, and a forced oscillation technique. results: Neonates from BPA-exposed mothers responded to this "sub optimal" sensitization with higher serum IgE anti-OVA concentrations compared with those from unexposed mothers (p < 0.05), and eosinophilic inflammation in their airways was significantly greater. Airway responsiveness of the OVA-sensitized neonates from BPA-treated mothers was enhanced compared with those from unexposed mothers (p < 0.05). conclusions: Perinatal exposure to BPA enhances allergic sensitization and bronchial inflammation and responsiveness in a susceptible animal model of asthma. for 1 week at room temperature. The concen tration of BPA in the water was assayed with a highly sensitive gas chromatography-mass spectrometry (GCMS) method (detection limit, 0.01 pg/mL). We found no detectable levels of BPA in the water. All experimenta tion was conducted under a protocol approved by the UTMB Institutional Review Board. The animals were treated humanely and with regard for alleviating suffering.
BPA loading of female mice and pups. To simulate the relevant sources of BPA exposure in the human fetus and infant, we fed BPA (Sigma, St. Louis, MO) to the female mice before pregnancy. Female BALB/c mice were given 10 µg/mL BPA in 1% ethanol solution in their drinking water for 1 week before mating and throughout their pregnancy and lactation. We chose the concentration of BPA to feed the female mice based on a previous study (Kabuto et al. 2004) in which the body burdens of free BPA in the mothers and pups were similar to those described in human tissues and fluids. In that study, 5 or 10 µg/mL BPA in the mother's drinking water resulted in 10 -8 to 10 -7 M con centrations in neo natal tissues, which is in the range reported from environmentally exposed humans (Table 1). Control female mice were given 1% ethanol solution in their drinking water from 1 week before mating until the end of the study. All of the outcome measures were performed only on pups.
Allergic sensitization and exposure. To detect effects of BPA on allergic sensitization, we gave suboptimal doses of OVA to the pups, as described previously (Fedulov et al. 2008;Hamada et al. 2007). A single intraperitoneal (ip) injection of 5 µg OVA with 1 mg alum (Sigma) as an adjuvant was given on postnatal day 4 (PND4). These mice were subsequently exposed to aerosolized 3% OVA solution in phosphatebuffered saline (PBS) for 10 min on PNDs 13, 14, and 15, using a jet nebu lizer (Pari II; Pari Industries, Richmond, VA). Markers of the development of asthma were assessed on PND17.
Pulmonary function testing. To assess airway hyper responsiveness (AHR), we used wholebody barometric plethysmography (Buxco Electronics, Sharon, CT) 48 hr after the OVA challenge on PND17 (Leme et al. 2006) as previously described (Castro et al. 2006). Measurements of airway responses were performed on individual, unrestrained, non anesthetized mice in a fourchamber plethysmograph. AHR was expressed as an enhanced pause (Penh), a dimensionless parameter used to measure pulmonary resis tance, calculated from changes in the pattern of chamber pressure induced by methacholine challenge. After a brief acclimatization in the chamber, the mice received an initial baseline challenge of saline, followed by increasing doses of nebulized methacholine (1, 10, 25, and 50 mg/mL). Recordings were taken for 3 min after each nebulization. The respira tory rate in breaths per minute was extrapo lated from readings of every 10 breaths. The box pressure waveforms generated from the respiratory cycle were used to calculate peak expiratory pressure (PEP), peak inspiratory pressure (PIP), and the time of expiration. Penh was then calculated using the formula Penh = pause × PEP/PIP. Penh values were averaged and reported as a percentage of base line nebulized saline inhalation.
Some of the pups from each treatment group were analyzed using the forced oscilla tion technique (flexiVent; SCIREQ, Montreal, Quebec, Canada) to validate the results of the wholebody barometric plethysmography, as described previously by Shore et al. (2003). Briefly, we treated these animals similarly to those described above, except we gave them aerosolized 3% OVA on PNDs 18-20. On PND22 (48 hr after the last aerosol chal lenge), animals were anesthetized with xyla zine (7 mg/kg) and pento barbital sodium (50 mg/kg). After performing a tracheostomy, we used a tubing adaptor (19 gauge; Becton Dickinson, Franklin Lakes, NJ) to cannu late the trachea and then ventilated the mice at 150 Hz with a tidal volume of 0.3 mL. Pressure at the airway opening was meas ured by the flexiVent system. We applied a posi tive endexpiratory pressure of 3 cm H 2 O. We used a 2.5Hz sinusoidal forcing function to meas ure dynamic pulmonary resistance (RL) by the forced oscillation technique. We obtained dose-response curves to nebulized methacholine (0.1, 1, 10, 30 and 50 mg/mL); the five highest values of RL obtained after each dose were averaged to obtain the final values for each dose.
Quantifying the inflammatory cells in bronchoalveolar lavage (BAL) fluid. On PND17, mice were sacrificed, and total leuko cytes and eosinophils in BAL fluid were counted. Immediately after sacrifice, cells in the lungs were recovered by flushing the isolated trachea with 0.5 mL PBS twice. Total leuko cytes were counted using a hemo cytometer. Eosinophil  counts were calculated from differential cell counts in 150 µL fluid deposited onto glass slides using a Cytospin 3 centrifuge (400 × g for 4 min; Shandon Lipshaw, Pittsburgh, PA) and stained with hematoxylin and eosin (Anderson and Poulsen 2003). The results were expressed as the absolute number of total cells and eosinophils. Quantification of OVA-specific anti bodies in mouse sera. To evaluate the effects of the environmental estrogen BPA on in vivo sensi tization, we measured allergenspecific IgE and IgG1 in the sera of the mouse pups at the time of sacrifice. Individual measurements of IgE and IgG1 were performed by ELISA (enzyme linked immunosorbent assay). We used OVA specific monoclonal IgE and IgG1 anti bodies (Gene Tex, Inc., San Antonio, TX) as stan dards, and biotinylated antimouse IgE (R&D Systems, Minneapolis, MN) and horseradish peroxidase antimouse IgG (H&L; Zymed, San Francisco, CA) for detection. The detec tion limit for both IgE antiOVA and IgG antiOVA was 5 ng/mL. Values below the detection limit are shown on the baseline.
Statistical analysis. Results are expressed as the mean ± SE. Statistical analysis was performed using oneway analysis of vari ance. Where differences between groups were present, they were further analyzed by the Student's ttest. A pvalue of < 0.05 was defined as statistically significant.

Results
The effect of BPA exposure on pregnancy. We found no significant differences in the BPA exposed and unexposed mothers in time to impregnation, litter size, birth weights, or sex ratio of the pups (data not shown). Figure 1 shows the effect of BPA exposure on the development of AHR as assessed by whole body barometric plethysmograph and the forced oscillation technique. Airway respon siveness to methacholine was significantly increased in BPAOVA pups compared with all other groups using both methods of analy sis. These differences were present in response to both 25 and 50 mg/mL methacholine when meas ured by wholebody barometric plethysmograph and 30 mg/mL by the forced oscillation technique. The patterns of Penh and lung resistance responses to methacholine were quite similar. This result is consistent with a previous report by Adler et al. (2004) in BALB/c mice that showed a signifi cant cor relation between Penh and lung resistance.

Effect of BPA on allergen-induced AHR.
Effect of BPA on pulmonary inflammation. To determine whether BPA alters allergenin duced pulmonary inflammation, we quantified total and differential cell counts in BAL fluid. We observed a significant increase in eosinophils in BAL fluid from BPAOVA pups ( Figure 2B) compared with all other groups. We found no significant difference in the total cell number between the groups (Figure 2A). We derived the data for each group from 6-7 mothers and 12-16 pups. Pups from each litter were distributed in both the OVAsensitized and non sensitized groups, so the unit of analysis was the individual pups. Statistical analysis was performed with and without the inclusion of the BPAOVA outlier shown in Figure 3. Both analyses showed p < 0.05 for eosinophil con centration in BAL fluid from BPAOVA pups compared with all other groups.
Effect of BPA on allergen-specific antibody production. The concentration of IgE anti OVA antibodies in the sera from BPAOVA pups was significantly higher than that for the other three groups (p < 0.05; Figure 3). IgG1 antiOVA concentrations in sera from these four groups of pups did not differ. Statistical analysis performed with and without the out lier (BPAOVA pup) showed statistical sig nificance (p < 0.05) for IgE antiOVA in the sera from BPAOVA pups compared with all other groups, but no significance for IgG antiOVA concentrations. The outlier in Figure 3 was the same BPAOVA pup with the high eosinophil number in the BAL fluid ( Figure 2B).

Discussion
To begin testing the hypothesis that expo sure of genetically susceptible animals to an environ mental estrogen during critical peri ods of immune development promotes aller gic sensitization and enhances subsequent inflammatory reactions, we chose to analyze the responses of neo natal mice to BPA expo sure during the peri natal period. We found that AHR, eosinophilic inflammation, and allergenspecific IgE were all significantly increased in the allergensensitized/challenged, BPAexposed pups compared with those that were not exposed to BPA or not sensitized to OVA. These findings are quite consistent with our hypothesis. The BALB/c strain of mice is considered to be susceptible to allergic sensitization and has been used extensively as a model of allergic asthma after sensitization with OVA. However, most studies of OVA induced asthma in BALB/c mice have started the sensitization process later in life and have used multiple and larger ip doses of OVA to prime the allergic response and larger inhaled doses to complete the sensitization and to induce airway inflammation and hyperreactiv ity. We chose an intentionally "sub optimum" sensitization protocol that was designed and used extensively by Kobzik and colleagues (Fedulov et al. 2008;Hamada et al. 2007) to examine maternal influences on development of the asthma phenotype in neo natal BALB/c mice. This approach allowed us to demon strate in our model that early life exposure to BPA enhanced the development of asthma under conditions that otherwise would not induce this phenotype. However, the present study does not exclude the possibility that exposure to BPA in utero only, during early infancy only, or even in adulthood may have similar effects on the development or manifes tations of asthma.
Our finding that BPA exposure enhanced allergic sensitization (serum IgE antiOVA anti bodies), eosinophilic airway inflammation, and bronchial hyperreactivity does not allow us to distinguish primary from secondary or tertiary effects of BPA on the development of these components of the asthma pheno type. However, because exposure to both BPA and OVA was required to induce AHR in our model, it seems very likely that allergic sensiti zation is required for BPA to promote asthma development, although sensitization alone may not be sufficient to induce the asthma pheno type. This proposition is in keeping with the study by Ohshima et al. (2007) that found that BPA exposure in mice also promotes the manifestation of food allergy in orally sensi tized mice. Our finding that our BPA expo sure/OVA sensitization protocol did not cause an increase in mouse IgG1 antibodies to OVA [another Thelper cell (T H 2)driven isotype] may be explained by the recent finding that IgG1 is a switch intermediate between IgM and IgE expression (Sudowe et al. 1997). Because BPA has been shown to induce production of a large amount of interleukin4 (IL4) (Johnson et al. 1996;King et al. 1998;Siegrist 2001), this may drive the IgG1expressing B cells on to IgE expression. However, this finding may also be a manifestation of the immaturity of the Bcell system at this age (Johnson et al. 1996;King et al. 1998;Siegrist 2001).
The key cell types and mechanisms by which BPA promotes allergic asthma in mouse pups remain to be elucidated. However, a number of sex steroid effects on immune sys tem functions have been described (Watson and Gametchu 2001), although relatively few have been explored mechanistically. Estrogens (10 -7 M E 2 , nonylphenol, and octylphenol) promote a T H 2 response associated with increased IL4 and decreased interferonγ (IFNγ) production from CD4 + CD8 + thy mocytes and naive CD4 + T cells isolated from C57BL/6 mice (Iwata et al. 2004). Lambert et al. (2005), using cells from an ERα knock out mouse on a C57BL/6 background, found that 10 -9 M E 2 acts through ERα to increase IL4 and GATA3 expression and essential T H 2 cytokine and transcription factors in CD4 + cells and to reduce IFNγ production from macrophages. E 2 (10 -8 to 10 -6 M) also has effects on major histocompatibility com plex class II expression on spleen dendritic cells isolated from BALB/c mice (Yang et al. 2006). High concentrations of BPA (10 -5 to 10 -4 M) and nonyl phenol (10 -7 to 10 -6 M) have been shown to induce IL4 and IgE pro duction from CD4 + T cells (Lee et al. 2003).
Environmental toxins are encountered to varying extents in both industrialized and developing nations and rural and urban environ ments. Some of these compounds are byproducts of industrial plants or agricul ture, such as pesticides. They often contami nate water supplies, exposing fish and other animals, including humans. Although the World Health Organization strongly supports breastfeeding, breastmilk monitoring stud ies suggest that environmental chemicals that may affect children's health are transmitted through breastfeeding (Solomon and Weiss 2002;Wang et al. 2004). This is especially true for environmental lipidsoluble pollut ants such as poly halogenated compounds, because these chemicals tend to degrade slowly in the environment, bioaccumulate and bioconcentrate in the food chain, and have long halflives in humans. Because the fat content of breast milk is relatively high, the concentration of some of these pollutants is 100 times higher in maternal milk than in plasma (Dewailly et al. 1993). As the final consumers in the food chain, human infants may consume the highest concentrations of lipidsoluble environ mental pollutants, which might enhance their risk of developing asthma or other allergic diseases due to the actions of environ mental estrogens.
Newborns generally have a T H 2skewed pattern of immunity, partially due to low pro duction of IL12 and the propensity of T H 1 cells to undergo apoptosis after antigen expo sure (Allam et al. 2005;Prescott et al. 1998). Subsequently, "immunematuring" infec tions are thought to promote a shift toward T H 1 responses in most children, whereas some remain prone to develop T H 2 responses (Prescott et al. 1999) and their disease con sequences. In keeping with this progres sion, most cases of asthma develop in early childhood. Thus, environmental exposures in utero and during first months of life are likely to be very important (Holt and Jones 2000). Further, Gern et al. (1999) found that asthma, at any age, is likely to originate in childhood or earlier. Therefore, immunologic events in children and experimental models of airway injury and repair in early life may provide important clues to the inception and pathogenesis of asthma.
We observed that our adult mice drank about 5 mL of 10µg/mL BPA in their water and therefore estimate that they consumed about 2 mg/kg/day of BPA. Lee et al. (2003) reported that injection of keyhole limpet hemocyanin (100 µg) into the footpad fol lowed by BPA (25 mg/kg/day) or nonyl phenol (5 mg/kg/day) by ip injection induced production of both IL4 and IgE in adult mice. It is possible that the mice in their study had higher tissue levels of BPA than did ours, because we found that the range of envi ronmental estrogens that induced mast cell degranulation in vitro was somewhat broad, 10 -11 to 10 -9 M , sug gesting that a broad range of BPA may also promote allergic sensitization in vivo.
The results described here indicate that we must give due consideration to the possible impact of environmental estrogens on normal immune development and on the develop ment and morbidity of immunologic diseases, such as asthma. More extensive studies are required to analyze the cellular and molec ular mechanisms that underlie these effects during specific developmental windows and thus identify approaches to prevent exposures or remediate effects of the xeno estrogens. Understanding the implications of our study for human asthma will require epidemiologic studies that examine the effect of BPA burden of mothers and their children on the risk of developing childhood and adult asthma in large populations.