Pre- and Neonatal Imprinting on Immunological Homeostasis and Epithelial Barrier Integrity by Escherichia coli Nissle 1917 Prevents Allergic Poly-Sensitization in Mice

Abstract

A steady rise in the number of poly-sensitized patients has increased the demand for effective prophylactic strategies against multi-sensitivities. Probiotic bacteria have been successfully used in clinics and experimental models to prevent allergic mono-sensitization. In the present study, we have investigated whether probiotic bacteria could prevent poly-sensitization by imprinting on the immune system early in life. We used two recombinant variants of probiotic Escherichia coli Nissle 1917 (EcN): i) EcN expressing birch and grass pollen, poly-allergen chimera construct (EcN-Chim), and ii) an "empty" EcN without allergen expression (EcN-Ctrl). Conventional mice (CV) were treated with either EcN-Chim or EcN-Ctrl in the last week of the gestation and lactation period. Gnotobiotic mice received one oral dose of either EcN-Chim or EcN-Ctrl before mating. The offspring from both models underwent systemic allergic poly-sensitization and intranasal challenge with recombinant birch and grass pollen allergens (rBet v 1, rPhl p 1, and rPhl p 5). In the CV setting, the colonization of offspring via treatment of mothers reduced allergic airway inflammation (AAI) in offspring compared to poly-sensitized controls. Similarly, in a gnotobiotic model, AAI was reduced in EcN-Chim and EcN-Ctrl mono-colonized offspring. However, allergy prevention was more pronounced in the EcN-Ctrl mono-colonized offspring as compared to EcN-Chim. Mono-colonization with EcN-Ctrl was associated with a shift toward mixed Th1/Treg immune responses, increased expression of TLR2 and TLR4 in the lung, and maintained levels of zonulin-1 in lung epithelial cells as compared to GF poly-sensitized and EcN-Chim mono-colonized mice. This study is the first one to establish the model of allergic poly-sensitization in gnotobiotic mice. Using two different settings, gnotobiotic and conventional mice, we demonstrated that an early life intervention with the EcN without expressing an allergen is a powerful strategy to prevent poly-sensitization later in life.

Keywords: BALB/c; Escherichia coli Nissle 1917; allergic poly-sensitization; germ-free; hygiene hypothesis; mouse model; mucosal tolerance.

Figure 1

Perinatal application of Escherichia coli Nissle 1917 (EcN)-Chim and EcN-Ctrl in conventional mice reduced airway inflammation. (A) Schematic representation of the perinatal application of recombinant EcN expressing birch-grass pollen chimera (EcN-Chim) and EcN expressing empty plasmid (EcN-Ctrl) in a conventional mouse model of poly-sensitization. Female BALB/c mice were given either EcN-Chim or EcN-Ctrl orally during the gestation and lactation. On day 21, female offspring derived from these mothers were separated and were then sensitized and challenged with 5 µg Bet v 1, 5 µg Phl p 1, and 5 µg Phl p 5 as described in poly-sensitization model. Mice were sacrificed 72 h after the last challenge. (B) Absolute numbers of macrophages, lymphocytes, eosinophils, and neutrophils in bronchoalveolar lavage (BAL) and (C) Representative cytospins of BAL of one mouse per group stained with hematoxylin and eosin (H&E; 100x magnification). (D) IL-5 and (E) IL-13 cytokines in BAL. (F) Representative lung tissue sections of one mouse per group stained with Periodic Acid Schiff (PAS) (Red; 10 x magnification; scale bars 100 µm). (G) Levels of Bet v 1-specific serum IgE. (B, D, E, G) represents the mean ± SEM from two experiments (total n = 6 to 9 mice per group). Error bars show mean ± SEM. *P < 0.05, **P < 0.01, ****P < 0.0001 by the One-way ANOVA followed by the Bonferroni’s Multiple Comparison Test.

Figure 2

Perinatal and neonatal mono-colonization of Escherichia coli Nissle 1917 (EcN)-Chim and EcN-Ctrl in germ-free (GF) mice reduced airway inflammation. (A) Schematic representation of the neonatal colonization of GF mice with recombinant EcN expressing EcN-Chim and EcN-Ctrl in a gnotobiotic mouse model of poly-sensitization. Eight-week-old GF mice were colonized with a single dose of either EcN-Chim or EcN-Ctrl by intragastric gavage and mated 10 days later. On day 21, all mono-colonized female offspring were separated from their mothers and divided into two groups. One group was sensitized and challenged with 5 µg rBet v 1, 5 µg rPhl p 1, and 5 µg rPhl p 5 and other not. Mice were sacrificed 72 h after the last challenge (B) Absolute numbers of macrophages, lymphocytes, eosinophils, and neutrophils in bronchoalveolar lavage (BAL). (C) Representative cytospins of BAL of one mouse per group stained with H&E 100x magnification). (D) IL-5 and (E) IL-13 cytokines in BAL. (F) Quantification of mucus-producing goblet cells. (G) Representative lung tissue sections of one mouse per group stained with Periodic Acid Schiff (PAS) (Red; 10x magnification; scale bars 100 µm); arrows indicate cell infiltration. (B, D–F) represents mean ± SEM from two experiments (total n = 7 to 10 mice per group). Error bars show mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001 by the One-way ANOVA followed by the Bonferroni’s Multiple Comparison Test.

Figure 3

Perinatal and neonatal mono-colonization with Escherichia coli Nissle 1917 (EcN)-Chim and EcN-Ctrl inhibits the development of Th2-type allergen-specific immune response. Mice were treated as indicated in Figure 2A. Serum samples were obtained from mice on sacrifice day. Allergen-specific antibody levels in mouse sera were determined by ELISA and RBL. The figure represents levels of Bet v 1, Phl p 1, Phl p 5-specific serum (A) IgG1 (B) IgG2a, and (C) IgE. Data represent mean ± SEM from two experiments (total n = 7 to 10 mice per group). Error bars show mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001 by the One-way ANOVA followed by the Bonferroni’s Multiple Comparison Test.

Figure 4

Perinatal and neonatal mono-colonization with Escherichia coli Nissle 1917 (EcN)-Ctrl led to increased IL-10, Foxp3, and IFNγ mRNA expression in the lung. Mice were treated as indicated in Figure 2A. After sacrifice, lung samples were collected and expression of IL-10, Foxp3 and IFNγ mRNA was measured by real-time (RT)-PCR. The figure represents the ratio of the target genes: (A) IL-10, (B) Foxp3, and (C) IFNγ to the GAPDH reference gene and (D) IL-10, (E) Foxp3, and (F) IFNγ to the β-actin reference gene. Data represents mean ± SEM from two experiments (total n = 7 to 10 mice per group) analyzed by the One-way ANOVA followed by the Bonferroni’s Multiple Comparison Test. Error bars show mean ± SEM. Data represents mean ± SEM from two experiments (total n = 7 to 10 mice per group). Error bars show mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001 by the One-way ANOVA followed by the Bonferroni’s Multiple Comparison Test.

Figure 5

Perinatal and neonatal mono-colonization of either EcN-Chim or EcN-Ctrl activated both TLR2 and TLR4 expression and maintained ZO-1 levels in the lung. Mice were treated as indicated in Figure 2A. For immunohistochemistry analysis, lung sections were processed and stained as described in Material and Methods using antibodies against TLR2, TLR4, and ZO-1 followed by 3,3′-Diaminobenzidine (DAB) staining. Samples were analyzed under a light microscope with a 40x objective. Immunohistochemistry staining for TLR2 (A), TLR4 (B), and ZO-1 (E) is represented by a brown color (indicated by arrows). Quantification of TLR2 (C), TLR4 (D), and ZO-1 (F) expression was performed by optical density analysis (OD) of bronchial epithelial cell layer using ImageJ software. (C, D, F) represents mean ± SEM from two experiments (five mice per group were tested). Error bars show mean ± SEM. *P < 0.05, **P < 0.01, ****P < 0.0001 by the One-way ANOVA followed by the Bonferroni’s Multiple Comparison Test.