Chronic exposure to perfluorinated compounds: Impact on airway hyperresponsiveness and inflammation

Abstract

Emerging epidemiological evidence reveals a link between lung disease and exposure to indoor pollutants such as perfluorinated compounds (PFCs). PFC exposure during critical developmental stages may increase asthma susceptibility. Thus, in a murine model, we tested the hypothesis that early life and continued exposure to two ubiquitous household PFCs, perfluorooctanoic acid (PFOA) and perflurooctanesulfonic acid (PFOS), can induce lung dysfunction that exacerbates allergen-induced airway hyperresponsiveness (AHR) and inflammation. Balb/c mice were exposed to PFOA or PFOS (4 mg/kg chow) from gestation day 2 to 12 wk of age by feeding pregnant and nursing dams, and weaned pups. Some pups were also sensitized and challenged with ovalbumin (OVA). We assessed lung function and inflammatory cell and cytokine expression in the lung and examined bronchial goblet cell number. PFOA, but not PFOS, without the OVA sensitization/challenge induced AHR concomitant with a 25-fold increase of lung macrophages. PFOA exposure did not affect OVA-induced lung inflammatory cell number. In contrast, PFOS exposure inhibited OVA-induced lung inflammation, decreasing total cell number in lung lavage by 68.7%. Interferon-γ mRNA in the lung was elevated in all PFC-exposed groups. Despite these effects, neither PFOA nor PFOS affected OVA-induced AHR. Our data do not reveal PFOA or PFOS exposure as a risk factor for more severe allergic asthma-like symptoms, but PFOA alone can induce airway inflammation and alter airway function.

scientific evaluation of xenobiotic exposure on lung health is critical in understanding the etiology of chronic lung diseases such as asthma, as a wide range of environmental factors influence one's susceptibility to acquire lung diseases. In particular, the onset of allergic asthma may be influenced by chemical, infectious, and allergic exposures in utero and in infancy. Thus studies evaluating the impact of xenobiotic insults, starting at early lifetime, on the onset and severity of lung disease in adulthood are of great interest. Here we evaluated if exposure to perfluorinated compounds (PFCs) spanning over critical stages of development impacted the lung's susceptibility for asthma-related symptoms using a murine model.

PFCs are persistent organic pollutants widely used in consumer products such as nonstick cookware, stain-resistant carpets, flame retardants, adhesives, and pesticides (34). Health concerns about PFCs stems from several factors, including: they are ubiquitous in the environment and can accumulate in the body (2, 27, 32). Perfluorooctanoic acid (PFOA) and perflurooctanesulfonic acid (PFOS) are detected in drinking water (25) and house dust (14, 32), and their volatile chemical precursors are detectable in indoor air (32, 34, 35). Exposure through ingestion of contaminated food is a major source of PFOA and PFOS intake in humans (8, 36). Both chemicals have been detected in blood samples collected from the general public, with human half-lives for PFOS and PFOA of 3.8 and 5.4 yr, respectively (28). They are readily absorbed, are not metabolized in humans, and cross the placental barrier to accumulate in the developing fetus (2, 17, 20). Newborns are exposed to PFCs through breastfeeding (21). Additional concerns arise because PFCs can disrupt endocrine pathways and have immune modulator effects (reviewed in Ref. 34). Specifically, PFOA and PFOS bind to peroxisome proliferator-activated receptors (PPARs), a key nuclear hormone receptor with roles in lipid metabolism and inflammation (reviewed in Ref. 6).

Multiple reports, in a variety of animal models, paint a concerning picture of the potential for PFOA and PFOS to contribute to negative health effects, including developmental toxicity (23) and pulmonary complications resulting in neonatal mortality (12, 13). In humans, there is evidence of higher (self-reported) shortness of breath and asthma among people who had been exposed to PFOA via drinking water (1). A more recent study from Taiwan involving ∼450 children revealed a positive association between PFC exposure and immunological biomarkers of juvenile asthma (7).

In experimental models, exposure to PFOA and PFOS had been demonstrated to alter inflammatory responses and cytokine production in allergic airways. For instance, a short (4-day) dermal exposure to high-dose PFOA (25 mg/kg body wt) was shown to augment allergen-induced IgE production and airway hyperreactivity in mice, suggesting PFOA exposure may increase the risk of allergy and asthma (10). PFOA has been shown ex vivo to induce excessive release of human mast cell-derived proinflammatory cytokines that can directly modulate airway contractility (33).

We hypothesized that exposure to PFOA and PFOS, beginning in utero, would predispose offspring for reduced lung function and altered allergic response in the lungs. We used a mouse model of long-term exposure to ingested PFOA and PFOS to 1) measure changes in lung function and airway hyperresponsiveness (AHR), 2) determine if PFC exposure aggravates AHR induced by ovalbumin (OVA) sensitization and challenge, and 3) determine if PFC exposure modulates the allergic inflammatory response to OVA in the lungs.

MATERIALS AND METHODS

Animals and chemical exposure protocol.

All animal procedures were performed according to protocols and guidelines approved by the University of Manitoba Animal Ethics Committee (Winnipeg, MB, Canada). Timed-pregnant dams were fed PFOA- or PFOS-contaminated diet ad libitum (∼4–6 g/day) beginning on gestation day (GD) 2. The first day with vaginal plug was implemented as GD 0 in our protocol. The chemical-contaminated diet was prepared by mixing in 4 mg of PFOA or PFOS (dissolved in methanol)/1 kg of LabDiet 5001 Rodent powder diet (LabDiet, Brentwood, MO). Perfluoro-n-octanoic acid (PFOA, ≥98%) was purchased from Wellington Laboratories (Guelph, ON, Canada). Heptadecafluorooctanesulfonic acid potassium salt (PFOS, ≥98%) was purchased from Sigma-Aldrich (St. Louis, MO).

The pregnant dams were fed the contaminated diet through the pregnancy and lactation periods. The pups, once weaned, were housed based on gender in groups of four to five and fed the same diet as their dams. Equal numbers of male and female offspring (4:4 or 5:5/group) were randomly selected from a larger pool of animals at ∼3 wk after the birth to be assessed for lung function at 12 wk of age. To assess impact of sustained chronic PFC exposure on allergic inflammatory and airway hyperreactive response to allergen, a subgroup of chemical-exposed cohorts underwent an OVA protocol at 8–9 wk of age (Fig. 1). During the protocol, mice were sensitized by intraperitoneal injection of 2 μg of OVA (Albumin Chicken Egg Grade V; Sigma-Aldrich) adsorbed to 2 mg Inject Alum (Thermo Scientific, Waltham, MA) in 0.5 ml saline. Fifteen days after the first OVA sensitization, a second intraperitoneal OVA was administered. Mice were challenged with intranasal OVA (50 μg OVA in 50 μl saline, applied directly to nostrils) on day 15, and intranasal OVA challenge was repeated two times in the following two days. Experimental assessments were made 48 h postfinal OVA challenge. Body weight and wet liver weights were also measured at this experimental endpoint.

Fig. 1.

Schematic diagram showing perfluorooctanoic acid (PFOA) and perflurooctanesulfonic acid (PFOS) exposure timeline and ovalbumin (OVA) protocol. A: timed-pregnant dams were fed PFOA- or PFOS-enriched diet ad libitum beginning on gestation day (GD) 2. The PFC-enriched diet was prepared by mixing in 4 mg of PFOA or PFOS (dissolved in methanol)/1 kg of diet. Dams were fed PFC-enriched diet throughout pregnancy and the lactation period. Postweaning, the pups were maintained on the PFC-enriched diet until 12 wk of age. B: subgroups of PFC-exposed mice were subjected to OVA sensitization and challenge: beginning at 8–9 wk of age, mice were sensitized by ip (OVA IP) injection (2 μg OVA adsorbed to 2 mg of alum in 0.5 ml saline). Fifteen days later (day 15) a second OVA IP was administered along with an initial intranasal OVA challenge (IN OVA): 50 μg OVA in 50 μl saline applied directly to nostrils. IN OVA was repeated for two more consecutive days (days 16 and 17 of the OVA protocol). On day 19, 48 h after the final OVA IN challenge, lung function was assessed using a flexiVENT; thereafter, we collected bronchoalveolar lavage fluid (BALF), blood, and lung specimens.

Liquid chromatography-tandem mass spectrometry analysis of PFOA in serum.

At the experimental endpoint, whole blood was collected, left at room temperature to clot, and centrifuged at 10,000 rpm for 10 min at 4°C. The supernatant (blood serum) was isolated and stored at −80°C until extraction and analysis. The PFOA extraction procedure was adopted from a recent study (11, 15). Briefly, 20 μl of 2.5 pg/μl internal standard (perfluoro-n-[1,2,3,4-¹³C₄]octanoic acid, >98% purity, >94% isotopic purity) (Wellington Laboratories), 50 μl of methanol, and 50 μl of serum were mixed. The solutions were vortex mixed and then centrifuged at 14,000 rpm for 10 min. The supernatant was isolated, vortex mixed with 60 μl of acetonitrile, and chilled at −20°C for 1 h. The mixture was centrifuged at 14,000 rpm for 10 min. Supernatant (110 μl) was combined with 110 μl of 0.1% formic acid in a polypropylene autosampler vial. The solution was vortexed and then subjected to liquid chromatography-tandem mass spectrometry (LC-MS/MS) analysis using online solid-phase extraction.

Concentrations of PFOA were quantified by LC-MS/MS analysis using online solid-phase extraction, using a HTC PAL autosampler (CTC Analysis, Zwingen, Switzerland), two Agilent LC pumps (Agilent Technologies, Mississauga, ON, Canada), and an Agilent 6410 triple quadrupole MS/MS. Quantitation was performed by isotope dilution, using dynamic multiple-reaction monitoring utilizing electrospray ionization in negative mode. The limit of detection and limit of quantification in bovine plasma were determined as the concentrations with signal-to-noise ratios of 3 and 10, respectively, and were estimated to be 0.014 and 0.048 ng/ml, respectively. Percent recoveries of PFOA were determined to be 103%. These results are similar to what has been seen for recoveries in other studies (5, 27, 29).

Respiratory mechanics and bronchoalveolar lavage collection.

To measure lung mechanics, mice were anesthetized with pentobarbital sodium (90 mg/kg ip injection) and tracheotomized with a 20-gauge polyethylene catheter that was further connected to a flexiVent small animal ventilator (Scireq Montreal, PQ, Canada). Mice were mechanically ventilated with a tidal volume of 10 ml/kg body wt, 150 times/min. Lung mechanics were measured using the forced-oscillation technique, and a positive end-expiratory pressure of 3 cmH₂O was used for all studies. Mice were subjected to an increased dose of nebulized methacholine (MCh) challenge protocol to assess concentration response characteristics of respiratory mechanics. Saline (30 μl) containing 0, 3, 6, 12, 25, and 50 mg/ml MCh was delivered over 10 s using an inline nebulizer at the beginning of each dose point. Low-frequency forced-oscillation technique was used to assess the effects of MCh challenge on the respiratory mechanics. During the low-frequency forced oscillation, mechanical ventilation was interrupted, and then a volume perturbation signal was applied using a preset flexiVent Prime-8 protocol. By fitting respiratory system input impedance to the constant-phase model (19, 37), flexiVent software calculated Newtonian resistance (Rn), peripheral tissue damping (G), and tissue elastance or stiffness (H); each parameter was normalized according to body weight. Values for each parameter were calculated as the mean of all 20 perturbation cycles performed after each MCh challenge.

Following lung mechanic measurement, lungs were lavaged with 1 ml of saline, two times, for a total of 2 ml. Bronchoalveolar lavage fluid (BALF) cell count was estimated using a hemocytometer. For differential counts, cells were stained with a modified Wright-Giemsa stain (HEMA 3 STAT PACK) (Fisher Scientific, Waltham, MA), and cell distributions were analyzed by manually identifying and counting eosinophils, lymphocytes, macrophages, and neutrophils in six randomly chosen fields of view examined with a light microscope at ×400 magnification. Cell counts were totaled across the six fields of view, and the ratio was used to calculate the absolute number cells per milliliter of original BALF.

Quantitative PCR.

Right lungs were stabilized and stored in RNAlater (Qiagen, Mississauga, ON, Canada) at −20°C. RNA was extracted from the tissue using a Qiagen RNeasy Plus Mini Kit as per the manufacturer's instructions. RNA concentration and purity were assessed using a spectrophotometer (NanoDrop 2000; Thermo Scientific). The qScript cDNA superMix kit (Quanta BioSciences, Gaithersburg, MD) was used for the reverse transcription of RNA (1 μg) to cDNA. Quantitative PCR was carried out with the 7500 Real-Time PCR System (Applied Biosystems, Foster City, CA) using primer pairs for eotaxin, ribosomal 18S RNA (18S), interleukin-13 (IL-13), interferon-γ (INF-γ), and tumor necrosis factor-α (TNF-α) (Table 1). Each quantitative PCR reaction was comprised of 12.5 μl SYBR Green (Applied Biosystems, Warrington, UK), 1 μl of forward and 1 μl of reverse primer (10 μM primer stock), 1 μl of cDNA and built up to 25 μl with water before being sealed, briefly centrifuged, and run on the thermocycler using the recommended cycling protocol for the SYBR reagent (Life Technologies, Waltham, MA) with a primer annealing temperature of 58°C. Product specificity was determined by dissociation curve analysis using the supplied 7500 Sequence Detection software version 1.4 (Applied Biosystems). Ribosomal 18S was used as the internal standard with relative gene expression calculated using the ΔΔC_t method as previously described (24).

Table 1.

Summary of primer sequences used in quantitative PCR

Target	NCBI Accession No.	Primer Sequence	PCR Product Size, bp
Eotaxin-1	NM_008176.3	F 5′-TTCTATTCCTGCTGCTCACGGTCA	164
		R 5′-GGCTTTCAGGGTGCATCTGTTGTT
18S	NR_003278	F 5′-CGCCGCTAGAGGTGAAATTC	62
		R 5′-TTGGCAAATGCTTTCGCTC
IL-13	NM_008355.3	F 5′-GATCTGTGTCTCTCCCTCTGA	109
		R 5′-AGGTCCACACTCCATACCA
INF-γ	NM_008337.3	F 5′-GGCCATCAGCAACAACATAAG	112
		R 5′-GTTGACCTCAAACTTGGCAATAC
TNF-α	NM_013693.3	F 5′-GCCTCCCTCTCATCAGTTCTAT	104
		R 5′-CACTTGGTGGTTTGCTACGA

Primers were designed using Primer Quest (IDT.com) and initially verified in silico using Primer Blast (NCBI), UNA Fold (IDT) and in silico PCR (UCSC) before being confirmed with standard PCR and one-dimensional electrophoresis. All primer sets were also tested using PCR.

18S, ribosomal 18S RNA; IL-13, interleukin 13; INF-γ, interferon-γ; TNF-α, tumor necrosis factor-α; F, forward; R, reverse.

Histological assessment of goblet cells in airways.

Histological assessment of airways was performed using the left lungs. Lungs were inflated with 1 ml of 10% buffered formalin (pH 7.4) to a pressure of 25 cmH₂O. Paraffin-embedded lung specimens were sectioned (6 μm thick) and stained with Periodic acid-Schiff stain (Sigma-Aldrich).

Medium-sized airways (4,000 μm > basement membrane length > 1,000 μm) were assessed for goblet cell metaplasia and mucus production. Based on the ratio of goblet cell area to whole cross-sectional epithelial area, the severity of goblet cell metaplasia and mucus hypersecretion was scored as previously described (3). Airways were scored from zero to three (3 being most severe) by two blinded independent observers: a score of zero for no goblet cells, a score of one (mild hyperplasia and mucus secretion) for occupation of <1/3 of the epithelial area by goblet cells with mild mucus secretion, a score of two (moderate hyperplasia and mucus secretion) for occupation of ≥1/3 to ≤2/3 of the epithelial area and moderate mucus secretion, and a score of three (severe hyperplasia) for occupation of >2/3 of the epithelial area and copious mucous secretion. For each group, 12 airways from 4 mice were scored two times to obtain an average score.

Statistical analysis of data.

Data are expressed as means ± SE. One- or two-way ANOVA, with Tukey-Kramer or Bonferroni post hoc tests, respectively, were used for comparisons of means from multiple groups. The Kruskal-Wallis test, followed by Dunn's Multiple Comparison, was used to compare the mucus score. Statistical analysis was carried out using IBM SPSS 20 (IBM, Armonk, NY) software. A P value <0.05 was considered statistically significant.

RESULTS

Exposure to PFOA and PFOS induces weight gain and liver enlargement.

To confirm that in-body PFC accumulation developed as a result of our ingestion protocols, we used a validated method to measure PFOA serum concentrations at 12 wk of age. PFOA serum levels in exposed mice reached 4,800 ± 1,100 ng/ml. Exposure to dietary PFOA and PFOS both correlated with significant body weight gain and liver enlargement at 12 wk of age. Compared with naïve controls, PFOA- and PFOS-exposed animals had mean body weights that were 13 and 9% greater (ANOVA, P < 0.05), respectively, and their liver weight-to-body weight ratio was 80 and 44% greater (ANOVA, P < 0.05), respectively.

PFOA, but not PFOS, exposure induces AHR.

Exposure to ingested PFOA, but not PFOS, resulted in a left upward shift of the MCh dose-response curves of airway resistance (Rn), indicating AHR (Fig. 2A). Central conducting airways in PFOA-exposed mice were hyperreactive to MCh challenge, marked by a 64.7% greater maximum Rn after inhalation of 50 mg/ml MCh. Indeed, Rn measured in the PFOA-exposed mice were significantly higher than that of chemical-naïve mice at every MCh dose >12 mg/ml (Fig. 2A). Airway sensitivity, as reflected by MCh provocative concentration 100 (PC100 = the dose of nebulized MCh required to elicit a 100% increase in baseline Rn), revealed PFOA-exposed mice were hypersensitive to MCh; there was 68.7% decrease in PC100 in the PFOA-group compared with chemical-naïve mice (Fig. 2B).

Fig. 2.

Early life and continued exposure of mice to PFOA alone induces airway hyperresponsiveness. Lung function was assessed in mice that were exposed to PFOA or PFOS. Airway mechanics were assessed after inhalation of nebulized saline and increasing concentrations of methacholine (MCh, 3–50 mg/ml). MCh dose-response curve for Newtonian resistance (Rn) (A), tissue damping (G, C), and tissue elastance (H, D) are plotted for each group. B: central or conducting airway sensitivity to MCh was measured by calculating provocative concentration of MCh needed to elicit 100% increase in Rn (PC100). Significant difference from the chemical-naïve was determined by 2-way ANOVA, and statistical difference compared with the chemical-naïve control was denoted P < 0.05 (*), P < 0.01(**), and P < 0.001(***). Error bars represent means ± SE. The no. of animals studied in each group was 8–10.

In PFOA-exposed mice, maximum G, which reflects peripheral tissue resistance, was 83.4% greater than for chemical-naïve mice (Fig. 2C). Similarly, maximum H in the PFOA-exposed mice was 85.6% greater than that of the naïve (Fig. 2D), indicating PFOA-only exposed lungs were stiffer and required greater work of breathing.

As was the case with PFOA, PFOS exposure resulted in greater sensitivity of airways to MCh as indicated by a significant decrease in PC100 (Fig. 2B). PFOS exposure, however, did not increase maximum Rn, nor did it affect the changes in tissue G or H caused by methacholine challenge (Fig. 2, A, C, and D).

PFOA and PFOS exposure does not alter allergic AHR.

We next determined whether altered lung mechanics resulting from lifelong exposure to PFOA or PFOS is a risk factor for the development of a greater degree of AHR after OVA sensitization and challenge. OVA-alone challenge consistently induced AHR to MCh, as indicated by a 93% increase in maximum Rn and 82% decrease in MCh PC100 (Fig. 3, A and B). Mice exposed to PFOA and OVA, or PFOS and OVA, exhibited AHR that was similar in magnitude to that induced by OVA alone (Fig. 3, A and B). Thus, low-dose chronic exposure to PFOA and PFOS did not alter the magnitude and severity of AHR induced by allergic airway inflammation.

Fig. 3.

Exposure of mice to dietary PFOA or PFOS does not affect allergen-induced airway hyperresponsiveness. Lung function was assessed in mice that were sensitized and challenged with OVA in addition to PFOA or PFOS exposure. Airway mechanics were measured after inhalation of nebulized saline and increasing concentrations of MCh (3–50 mg/ml). MCh dose-response curves for Rn (A), G (C), and H (D) are plotted for each group. B: conducting airway sensitivity to MCh was measured by calculating PC100. OVA-alone challenge consistently induced AHR to MCh, as indicated by a 93% increase in maximum Rn and 82% decrease in PC100. Statistical comparisons between OVA-alone and chemical-naïve control were compared by ANOVA, and statistical difference between OVA-alone and the chemical-naïve control was denoted P < 0.05 (*), P < 0.01(**), and P < 0.001(***). No statistical difference was found between PFOA + OVA, PFOS + OVA, and OVA-alone group (ANOVA, P > 0.05). Error bars represent means ± SE. The no. of animals studied in each group was 8–10.

Lung inflammation induced by PFOA or PFOS exposure.

Mice exposed to PFOA alone exhibited a 25-fold higher total cell count than chemical-naïve mice (Fig. 4A). The increase in total cell number in PFOA-exposed mice was associated with a dramatic accumulation of macrophages, which comprised 97% of the total cell number (Fig. 4B). Increased macrophage number was associated with increased abundance of IFN-γ mRNA, but not eotaxin-1, TNF-α, or IL-13, in the lung (Fig. 5). We also examined whether goblet cell abundance in the airway epithelium increased with PFCs exposure and found no evidence of difference in goblet cell mucus score between PFOA-exposed and chemical-naïve mice (Fig. 6).

Fig. 4.

Inflammatory cell counts in BALF are altered by exposure to PFOA or PFOS. BALF was collected after mice were anesthetized and mechanically ventilated to assess respiratory mechanics. A: total cell counts in BALF were estimated using a hemocytometer. Values are expressed as no. of cells/ml of BALF. One-way ANOVA was used to compare the total cell counts. Cell distribution was analyzed by manual counting of macrophages (B), eosinophils (C), and neutrophils (D) in 200 μl of BALF after cytospin, fixation, and modified Wright-Giemsa staining. Cell counts were obtained in 6 random fields examined from a light microscope (×40 magnification). Statistical difference in the differential counts was determined by ANOVA. Means were compared with that of the naïve and are denoted with P < 0.01(**) and P < 0.001(***) for significant difference. #Significant difference compared with the OVA-only group (ANOVA, P < 0.05). As expected, OVA-alone challenge induced a significant increase in total, macrophage, and eosinophil count. In OVA-challenged mice, PFOS decreased the total leukocyte count in BALF by 68.7% compared with OVA alone. This decrease in the total cell count was associated with a blunting of macrophage nos. by 64.8% compared with OVA-only exposed mice. In contrast, PFOA had no detectable impact on the allergen-induced leukocyte infiltration observed. Data points represent counts from 6 to 8 mice in each group (n = 4 for PFOS). Mean values for each group are indicated with a horizontal bar, and error bars for means ± SE are shown.

Fig. 5.

Relative mRNA abundance of inflammatory cytokines in mouse lungs is altered by exposure to PFOA and PFOS. mRNA abundance was measured using quantitative RT-PCR, with normalization to an internal standard, ribosomal subunit 18S. Relative mRNA abundances for eotaxin (A), tumor necrosis factor-α (TNF-α, B), interferon-γ (INF-γ, C), and interleukin-13 (IL-13, D) are presented. Relative expression between groups was compared based on ΔΔC_t values for PFOA and PFOS exposure alone, OVA sensitization and challenge, or OVA sensitization and challenge in combination with PFOA or PFOS exposure. All data are plotted relative to the abundance of mRNA for each target that was detected in the OVA-only exposed mice. Statistical comparison of mean ΔΔC_t was carried out using ANOVA included in SPSS 20, and significant differences from the chemical-naïve mice are denoted with * when P < 0.05. Error bars represent the means ± SE. Data are from at least 2 replicates from 3–4 mice in each group. ND, not detected (C_t >35).

Fig. 6.

Semiquantitative analysis of goblet cell and mucus abundance in airways of mice exposed to PFOA or PFOS revealed no impact on the OVA-induced goblet cell hyperplasia and mucus production by PFOA or PFOS exposure. Histological images of murine airways from chemical-naïve and no OVA challenge mice (A), PFOA-only exposed mice (B), PFOS-only exposed mice (C), chemical-naïve and OVA-sensitized/challenged mice (D), PFOA-exposed and OVA-sensitized/challenged mice (E), and PFOA-exposed and OVA-sensitized/challenged mice (F) are presented. Arrows in each panel indicate areas of positive staining for mucous and goblet cells in the airway epithelium. G: scores from semiquantitative assessment of goblet cell and mucus scores are plotted. To obtain a mean score, 12 medium-sized airways (cross-sectional epithelial basement membrane length >1,000 μm) from 4 independent animals were scored two times by 2 blinded observers. Goblet cell hyperplasia and mucus abundance were scored from 0 to 3; 0 = no presence of goblet cell; 1 = mild hyperplasia with occupation of <1/3 of the epithelial area by goblet cells with mild mucus, 2 = moderate hyperplasia with occupation of ≥1/3 to ≤2/3 of the epithelial area with mucus and goblet cells, and 3 = severe goblet cell hyperplasia with >2/3 of the airway epithelial with mucus and goblet cells. Kruskal-Wallis test along with Dunn's Multiple Comparison were used to compare the means. ***Significant statistical difference compared with the naïve mice (P < 0.001).

In contrast to PFOA, there was no change in total or differential cell count in BALF from mice exposed to PFOS alone (Fig. 4). However, we did observe a selective increase in lung IFN-γ mRNA (Fig. 5C) as with PFOA-exposed mice. Similar to histological observations made in PFOA-exposed mice, there was no obvious impact on airway goblet cell abundance in PFOS-exposed mice (Fig. 6).

Impact of PFC exposure on OVA-induced lung infiltration by leukocytes.

As expected, OVA exposure resulted in dramatic lung infiltration of leukocytes as reflected by increased numbers of macrophages and eosinophils in BALF (Fig. 4). This correlated with an increase in both eotaxin-1, TNF-α, and IL-13 mRNA in the lung (Fig. 5). OVA sensitization and challenge induced significant goblet cell hyperplasia and mucus staining in conducting airways (Fig. 6).

Surprisingly, we found that leukocyte infiltration of the lungs that was induced by allergic sensitization and challenge was blunted by dietary PFOS exposure but was not affected by PFOA exposure (Fig. 4). PFOS decreased the total leukocyte count in BALF by 68.7% compared with the OVA alone (Fig. 4A). This decrease in the total cell count was associated with a blunting of macrophage numbers by 64.8% compared with OVA-only exposed mice (Fig. 4B). In contrast, PFOA had no detectable impact on the allergen-induced leukocyte infiltration observed (Fig. 4). Similar to mice that were exposed to chemical alone, IFN-γ mRNA was significantly greater in the lungs of mice that were coexposed with PFOA and OVA (Fig. 5). Notably, in line with lung inflammatory cell numbers, in PFOS- and OVA-exposed mice the magnitude of expression of IL-13, eotaxin-1, and TNF-α, but not IFN-γ, appeared to be lower than that seen in PFOA- and OVA-exposed mice. Semiquantitative histological assessment of conducting airway revealed that neither PFOA nor PFOS affected OVA exposure-induced accumulation of goblet cells and mucous (Fig. 6).

DISCUSSION

To our knowledge, this is the first study to investigate effects of chronic exposure of PFOA and PFOS on respiratory mechanics and airway inflammation. Our protocol design ensured PFC exposure span over all critical stages of development (both in utero and early postnatal) and adulthood; and, therefore, it offers a unique platform to gauge the potential adverse effects of early life and continued exposure to PFC. Here, we show that PFOA exposure, but not PFOS, led to offspring presenting a phenotype characterized by AHR in association with elevated lung macrophages and IFN-γ in adulthood. Neither PFOA nor PFOS exposure increased the risk for developing excessive allergic inflammation or AHR. Rather, PFOS had a suppressive effect on allergen-induced lung inflammation without impacting the degree of OVA-induced AHR. Our data show that prolonged PFC exposure is associated with aberrant expression of the macrophage-activating family member IFN-γ. Although this may have potential to modulate local immunity, it does not appear to compound allergen-driven lung dysfunction.

One of the intriguing findings of this study was that, despite similarities in structure and function of PFOA and PFOS, the effects of long-term exposure to airway inflammation and lung function have several distinctive features. PFOA exposure alone, but not PFOS alone, was associated with significant AHR. PFOA exposure was linked with increased numbers of lung macrophages, whereas PFOS has no such effect. However, PFOS and PFOA exposures were both associated with elevated expression of the immune regulator IFN-γ in the lungs. This suggested that PFC exposure might underpin differential impact of allergen-induced inflammatory responses in PFOA- and/or PFOS-exposed mice. Indeed, PFOS exposure was associated with suppressed OVA-induced leukocyte infiltration to the lungs, whereas PFOA exposure had no detectable impact. Two compounds' distinctive impacts on the lungs likely arise from differential affinity and activation of PPAR by PFOA and PFOS. PFOS have higher affinity to PPARγ than PFOA, and PFOA have higher affinity and activation of PPARα than PFOS (16). PPARγ agonist are known to have anti-inflammatory properties and regulate cytokine secretion by airway smooth muscle (4, 18). In this regard, future study using a murine PPAR knockout model is highly desirable to elucidate the mechanism of AHR observed in this study.

Another interesting finding of this study is that exposure to PFOA and PFOS alone associates with increased abundance of INF-γ mRNA in the lungs. This effect is retained in mice that are also challenged with OVA. Notably, we observed a concomitant enhancing effect of PFOA alone on IFN-γ and macrophage number in the lungs. This is of interest because the accumulation of macrophages in several disorders, for example, in adipose tissue and at sites of atherosclerotic plaque formation, is associated with local inflammation that includes IFN-γ production as part of a local immune network involving resident and infiltrating immune cells that interact with the macrophage (30, 39). IFN-γ modulates innate and adaptive immunity against allergic insult and viral and intracellular bacterial infections, and it is an important activator of macrophages (31). Although our study was not designed to delineate mechanisms that link PFC exposure to increased IFN-γ expression, our findings do support future work in this area. Moreover, our work uncovers areas for further investigation and suggests that the impact of chronic PFC exposure on response to infectious agents and resulting changes in lung inflammation and function should be of interest.

Our observation that neither PFOA nor PFOS had any impact on OVA-induced lung dysfunction, which is manifest as AHR, is in disagreement with findings from an acute dermal PFOA exposure protocol that showed augmented airway hyperreactivity in mice exposed to PFOA and OVA (10). At first glance, the difference may be attributed to the difference in the concentration of PFC exposure; our exposure level approximated a gavage dose of 1 mg·kg body wt⁻¹·day⁻¹, which was below the dermal exposure level (2.5 mg PFOA·kg body wt⁻¹·day⁻¹) that showed enhanced OVA response (10). However, our exposure protocol includes maternal exposure during gestation and nursing (42 days at 1 mg/kg) as well as 9 wk (63 days at 1 mg/kg) of PFOA or PFOS ingestion postweaning. Thus, in total, the mice in our study received ∼105 mg/kg (42 days + 63 days at 1 mg/kg). In contrast, the acute dermal exposure study delivered 2.5 mg/kg for 4 days, for a cumulative exposure of only 10 mg·kg⁻¹·animal⁻¹. Thus, although daily exposure of animals in our study was lower, cumulative exposure was at least 10 times greater than that used for the dermal study, and, as such, the lack of impact on lung function would not appear to be linked to any deficit in the magnitude of PFC exposure. Of note, the previous acute study used whole body plethysmography instead of the force oscillation method we used to reveal the impact on airway hyperreactivity. Whole body plethysmography provides nonspecific lung mechanics, whereas the force oscillation technique provides a much more specific measure of conducting airways mechanics (19). Therefore, enhanced OVA-induced responsiveness with higher-dose PFOA and plethysmography measure could reflect more global changes in lung physiology, whereas the lack of effect on Rn using the force oscillation technique likely is more reflective of the airway response per se.

One may argue that the OVA sensitization and challenge protocol presented here was sufficient to induce maximum detectable changes in lung function, thereby precluding any potential additive impact that PFOA or PFOS could have. However, this possibility is unlikely to be the case because the Rn values we measured across groups ranged as high as 2.60 cmH₂O·s⁻¹·ml⁻¹, with the mean Rn in OVA-exposed mice being only 1.97 cmH₂O·s⁻¹·ml⁻¹. This indicates that we can detect and induce central airway resistance at much higher levels than the mean that we measured in chemical-only and chemical-OVA exposed animals. Hence, it is unlikely that any real augmentation of Rn by PFC exposure remained undetected due to technical limitations of using a high concentration of OVA or limits in measurement range of the small animal ventilator. Nonetheless, in future studies, experiments that employ a submaximal allergen sensitization arm may be revealing, as is the case for work examining bisphenol A exposure and its capacity to enhance airway hyperreactivity after OVA challenge (26).

In addition, because our study was designed to assess AHR when it was at its peak in OVA-only animals (48 h postallergen), this time point may not be ideal to track peak cytokine expression in lung tissue. It seems likely that cytokine levels would attain peak values before maximum AHR is attained; thus, our cytokine mRNA data may not necessarily reflect causal mechanisms. Notably, mRNA for the cytokines we measured was generally relatively low at the time point that we studied, being in the order of <1% of ribosomal 18S (data not shown). Therefore, future study assessing the impact on cytokine expression as a causal relationship with PFOS exposure and immune suppression should use an earlier time point.

The different pharmacokinetics and elimination kinetics in mice compared with humans may limit extrapolation of our findings to humans. For instance, we discovered that serum concentrations of PFOA measured in 12-wk-old mice (4,800 ± 1,100 ng/ml) were two to three orders of magnitude higher than that detected in the general public (0.5–20 ng/ml) (22, 29). Nevertheless, there are significantly higher levels of PFOA serum concentrations reported in occupationally exposed persons (422–999 ng/ml) and individuals living in areas of high environmental exposure (155–556 ng/ml) (9). Indeed, in retired workers who had been exposed for extended periods of time during fluorochemical production, serum PFOA concentrations have been reported to be between 72 and 5,100 ng/ml, with a mean of 691 ng/ml (28). Combined, our model most likely mimics the exposure level of highly exposed occupational workers and individuals living in areas of high environmental concentrations. Moreover, the elimination half-life of PFOA and PFOS is 17–19 days in mice, which is much longer than that for other experimental animals, such as rat (2 h-6 days with gender variance), rabbit (5–7 h), and chicken (4.6 days) (34). This difference makes the murine model we developed more suitable for assessing the impact of chronic, low-dose exposure to PFCs.

In conclusion, in utero-through-adulthood exposure of allergen-naïve mice to PFOA alone induced lung inflammation characterized by macrophage accumulation and IFN-γ expression, as well as AHR, a hallmark symptom of asthma. Our findings are important because they strongly support the current production phased out by eight major manufacturer's through a voluntary stewardship agreement with the U.S. Environmental Protection Agency in an effort to eliminate global PFOA production by 2015 (38). Moreover, continued monitoring of theses chemicals is necessary since these chemicals persist in the environment and in the human population because of their nondegradable nature and extremely long biological half-lives (28, 29). Taken into consideration with epidemiological studies that link PFOA exposures to juvenile asthma, our findings support the currently held public opinion that minimization of exposure to PFCs is in the best interest of children and expectant mothers.

GRANTS

Funding for this study was provided by the Natural Sciences and Engineering Research Council (Collaborative Health Research Project to A. B. Becker, A. J. Halayko, and C. S. Wong and Discovery Grant to C. S. Wong), the Manitoba Health Research Council (graduate studentship to M. H. Ryu and A. Jha and postdoctoral fellowship to M. Loewen and O. O. Ojo), and the Canada Research Chairs Program (to A. J. Halayko and C. S. Wong).

DISCLOSURES

Authors have no relevant conflict of interest to disclose related to research presented here.

AUTHOR CONTRIBUTIONS

Author contributions: M.H.R., A.J., O.O.O., T.H.M., S.B., K.A.D., N.N., and M.L. performed experiments; M.H.R., O.O.O., T.H.M., S.B., N.N., and A.J.H. analyzed data; M.H.R., A.J., and A.J.H. interpreted results of experiments; M.H.R., O.O.O., and A.J.H. prepared figures; M.H.R., A.J., K.A.D., and A.J.H. drafted manuscript; M.H.R., A.J., O.O.O., T.H.M., K.A.D., C.S.W., M.L., A.B.B., and A.J.H. edited and revised manuscript; M.H.R., A.J., O.O.O., T.H.M., S.B., K.A.D., N.N., C.S.W., M.L., A.B.B., and A.J.H. approved final version of manuscript; C.S.W., M.L., A.B.B., and A.J.H. conception and design of research.