Asthma-protective agents in dust from traditional farm environments

Abstract

Background: Growing up on traditional European or US Amish dairy farms in close contact with cows and hay protects children against asthma, and airway administration of extracts from dust collected from cowsheds of those farms prevents allergic asthma in mice.

Objectives: This study sought to begin identifying farm-derived asthma-protective agents.

Methods: Our work unfolded along 2 unbiased and independent but complementary discovery paths. Dust extracts (DEs) from protective and nonprotective farms (European and Amish cowsheds vs European sheep sheds) were analyzed by comparative nuclear magnetic resonance profiling and differential proteomics. Bioactivity-guided size fractionation focused on protective Amish cowshed DEs. Multiple in vitro and in vivo functional assays were used in both paths. Some of the proteins thus identified were characterized by in-solution and in-gel sodium dodecyl sulfate-polyacrylamide gel electrophoresis enzymatic digestion/peptide mapping followed by liquid chromatography/mass spectrometry. The cargo carried by these proteins was analyzed by untargeted liquid chromatography-high-resolution mass spectrometry.

Results: Twelve carrier proteins of animal and plant origin, including the bovine lipocalins Bos d 2 and odorant binding protein, were enriched in DEs from protective European cowsheds. A potent asthma-protective fraction of Amish cowshed DEs (≈0.5% of the total carbon content of unfractionated extracts) contained 7 animal and plant proteins, including Bos d 2 and odorant binding protein loaded with fatty acid metabolites from plants, bacteria, and fungi.

Conclusions: Animals and plants from traditional farms produce proteins that transport hydrophobic microbial and plant metabolites. When delivered to mucosal surfaces, these agents might regulate airway responses.

Keywords: Asthma; asthma protection; farm effect; microbial metabolites.

Figure 1.. DEs from European cow but not sheep sheds protect against experimental asthma

A. Balb/c mice were immunized with OVA + Alum i.p. (day 0 and 7) and challenged i.n. with OVA (day 15 and 17). DEs (5 mg of dust equivalent) were administered i.n. 8 times. BAL eosinophilia was assessed at day 19. B: BAL eosinophilia in mice receiving OVA (n=14), OVA + AM cowshed DE (n=13) or OVA + EUC-01 DE (n=5). Autoclaved (auto) samples were also tested (n=5 mice/group for OVA + AM cowshed DE and n=4 for OVA + EUC-01 DE). C: BAL eosinophilia in mice receiving OVA (n=4), OVA + unfractionated (0/0) EUC-01 DE (n=5) or OVA + EUC-01 DE fraction >10 kDa (n= 5). D. BAL eosinophilia in mice receiving OVA (n=10), OVA + autoclaved >10 kDa DE fractions EUC-02 (n=4), EUC-03 (n=3), or EUC-04 (n=3), or OVA + EUS-01 (n=4). Data (mean ± SEM) are from three independent experiments. E-G: lung Il5 mRNA levels in all samples tested in panels B-D. Data were analyzed using an unpaired, two-tailed t test (B, E-G) or a Wilcoxon two-sample test (C and D) after assessing the normality of value distributions with the Shapiro-Wilk test.

Figure 2.. NMR spectra of European and Amish cowshed DEs reveal the presence of random coil peptides/proteins and carbohydrates

Area-normalized ¹H NMR spectra (800 MHz, D₂O) show the section of aliphatic peptide side chains, OCH, and CCH units of carbohydrates. Eight European cowshed DEs (EUC-01, EUC-01.2, EUC-01.3, EUC-02/06) and one Amish (AM) cowshed DE (yellow line) are shown individually. The average ¹H NMR spectrum of the eight European cowshed DEs is shown by the dotted black line. All cowshed DEs show closely related curvature of ¹H NMR spectra, corroborating rather congruent structural main features that primarily originate from amino acid side chains (δ_H ~ 0.5–4.5 ppm for aliphatic, blue shade “a”) and subordinate, from OCH and HOCH₂ units in carbohydrates (δ_H ~ 3.2–4.5 ppm, red shade “b”). The entire ¹H NMR spectrum, also showing aromatic side chains (δH ~ 6.5–8.2 ppm), is shown in Fig. S2.

Figure 3.. Comparative NMR spectra of sheep shed versus cowshed DEs

Area-normalized ¹H NMR spectra (800 MHz, D₂O) show the section of aliphatic peptide side chains, and OCH and HOCH₂ units of carbohydrates, OCH, and CCH units. Two sheep shed DEs (EUS-01 and EUS-02) are compared with the computed average of eight European cowshed DEs. DEs derived from sheep shed show higher relative proportions of carbohydrates, comprising OCH and HOCH₂ units (δ_H ~ 3.3–4.3; red shade “b”) together with lower proportions of peptide CONHCαH units and lower proportions of peptides/proteins (δ_H ~ 0.5–4.5 ppm (Csp3H), blue shade “a”) than cowshed DEs. The entire ¹H NMR spectrum, also showing aromatic side chains (δ_H ~ 6.5–8.2 ppm) is shown in Fig. S4.

Figure 4.. The protective activity of Amish cowshed DEs resides within the 28–64 kDa range

A. Aqueous Amish cowshed DEs were fractionated by SEC collecting fractions (named as above the graph) every 1.5 minutes. Red and black lines represent spectrophotometric readings at 220 and 280 nm, respectively. B: Fractions were pooled into four consecutive groups (F-I, 6–16.5 min; A-E, 16.5–24 min; J-N, 24–31.5 min; O-S, 31.5–40 min) and mass-adjusted to the original concentration (100 mg/ml of dust equivalents). Fraction bioactivity was assessed by comparing percentages of BAL eosinophils in mice treated with OVA or OVA + fraction as in Fig. 1A. Shown are mean percentages of BAL eosinophils ± SE (8–9 mice/group from 2 independent experiments for each fraction pool). C: Effects of fractions A-E (each mass-adjusted to 100 mg/ml of dust equivalent) on OVA-induced BAL eosinophilia. Shown are mean percentages of BAL eosinophils ± SE (8–9 mice/group from 2 independent experiments for each fraction). Unfractionated Amish (AM) cowshed DE served as a positive control. Differences in percent BAL eosinophilia between mice treated with OVA and OVA + fraction were assessed by an unpaired, two-tailed t test after evaluating the normality of value distribution with the Shapiro-Wilk test.

Figure 5.. Fraction B inhibits OVA-induced AHR and selectively boosts human airway epithelial barrier function

A. Balb/c mice were sensitized with OVA (50 μg) + Alum (6 mg) i.p. (day 0 and 7) and treated i.n. with Amish cowshed DE fractions A-E (5 mg dust equivalent/treatment) 8 times over 14 days (Fig. 1A). Airway resistance was measured at day 19 after i.n. OVA challenge (100 µg, day 15–17) and methacholine (0–30 mg/ml) nebulization. Data (mean ± SEM) are from one of two experiments, 4–5 mice/treatment group. AHR differences between OVA- and OVA+DE fraction-treated mice were assessed by an unpaired, two-tailed t test after assessing the normality of sample value distribution with the Shapiro-Wilk test. B. TEER was measured in 16HBE14o- human bronchial epithelial cells cultured for 48 hours with fractions A-E or unfractionated Amish cowshed DE (2 mg dust equivalent/well) and was expressed as fold-increase in sample-treated over serum-starved (FCS^-) cells. TEER differences between cultures were assessed by an unpaired, two-tailed t test after evaluating the normality of sample value distribution with the Shapiro-Wilk test. C. Silver-staining of SDS-PAGE (4–12% reducing gel) of unfractionated (UF) Amish cowshed DE and fractions A-E (10 µg dust equivalent/lane). M: MW markers.