Perinatal Maternal Administration of Lactobacillus paracasei NCC 2461 Prevents Allergic Inflammation in a Mouse Model of Birch Pollen Allergy

Background The hygiene hypothesis implies that microbial agents including probiotic bacteria may modulate foetal/neonatal immune programming and hence offer effective strategies for primary allergy prevention; however their mechanisms of action are poorly understood. We investigated whether oral administration of Lactobacillus paracasei NCC 2461 to mothers during gestation/lactation can protect against airway inflammation in offspring in a mouse model of birch pollen allergy, and examined the immune mechanisms involved. Methods BALB/c mice were treated daily with L. paracasei in drinking water or drinking water alone in the last week of gestation and during lactation. Their offspring were sensitized with recombinant Bet v 1, followed by aerosol challenge with birch pollen extract. Results Maternal exposure to L. paracasei prevented the development of airway inflammation in offspring, as demonstrated by attenuation of eosinophil influx in the lungs; reduction of IL-5 levels in bronchoalveolar lavage, and in lung and mediastinal lymph node cell cultures; and reduced peribronchial inflammatory infiltrate and mucus hypersecretion. While allergen-specific IgE and IgG antibody levels remained unchanged by the treatment, IL-4 and IL-5 production in spleen cell cultures were significantly reduced upon allergen stimulation in offspring of L. paracasei treated mice. Offspring of L. paracasei supplemented mothers had significantly reduced Bet v 1-specific as well as Concanavalin A-induced responses in spleen and mesenteric lymph node cell cultures, suggesting the modulation of both antigen-specific and mitogen-induced immune responses in offspring. These effects were associated with increased Foxp3 mRNA expression in the lungs and increased TGF-beta in serum. Conclusion Our data show that in a mouse model of birch pollen allergy, perinatal administration of L. paracasei NCC 2461 to pregnant/lactating mothers protects against the development of airway inflammation in offspring by activating regulatory pathways, likely through TLR2/4 signalling.


Introduction
The prevalence of allergic diseases in industrialized countries has increased rapidly over the past decades. The reasons for the steady increase are not yet fully understood. While genetic factors contribute to increased disease risk [1,2], genetic predisposition and heritable factors cannot solely explain the rise in allergies. Rather, environmental factors such as reduced microbial stimulation of the immune system in infancy as a consequence of improved hygiene seem to play important roles. However, not only has the extrinsic microbial environment changed, but so has the composition of intestinal microbiota possibly due to differences in diet or over-use of antibiotics [3,4]. Several studies have shown differences in the composition of gut microbiota between allergic and non-allergic children [5,6,7,8] and reduced diversity of the infant's intestinal flora was associated with increased risk of allergic sensitization [9]. In a prospective study, children who later developed allergic disease were less often colonized with bifidobacteria during the first year of life [10]. Moreover, atopic sensitization was correlated with a reduced ratio of faecal bifidobacteria to clostridia in the early perinatal period [6]. These findings provide a rationale for use of probiotic bacteria to prevent allergic disease.
There is evidence that events occurring in the first year of life and even before delivery have the potential to program persisting immunological phenotypes that determine the subsequent risk of allergic disease [11,12]. Thus, pre-, peri-, and/or postnatal interventions offer a promising approach to modulate immune responses and promote a non-allergic status. Moreover, interventions during pregnancy/lactation might have considerable advantages in terms of convenience and compliance compared to childdirected interventions. Indeed, preclinical and clinical studies have shown that perinatal interventions with specific probiotic bacteria can mediate protection against infant allergic diseases [13].
In animal studies, we have shown previously that probiotic bacteria might regulate the allergic phenotype through a number of different pathways including: (i) induction of Th1-type immunity [14]; (ii) generation of regulatory responses [15,16]; (iii) and production of IgA [14]. Using a mouse model of allergic poly-sensitization, we recently demonstrated that Lactobacillus paracasei NCC 2461, a strain with the capacity to induce Th1 and regulatory responses in vitro [17], applied at the time of sensitization and challenge significantly suppressed airway inflammation and down-regulated allergen-specific immune responses [15]. In the present study we show that maternal supplementation with L. paracasei NCC 2461 protects against the development of allergic airway inflammation in offspring, likely mediated by induction of regulatory responses.

Animals
Pregnant BALB/c mice (day 14 of pregnancy) were purchased from Charles River (Sulzfeld, Germany) and maintained under conventional housing conditions. Female TLR2-and TLR4deficient and wild type mice with C57BL/6 genetic background were obtained from M. Müller (Vienna, Austria).

Probiotic Bacteria
Probiotic strain L. paracasei NCC 2461 (CNCM I-2116, hereafter L. paracasei) was provided by Nestlé Research Center (Lausanne, Switzerland). For perinatal oral administration, spraydried bacterial preparation was diluted in drinking water at 5610 8 CFU/ml. For in vitro experiments, L. paracasei preparation was inactivated with 1% formaldehyde-PBS for 3 h at room temperature, washed twice with sterile PBS, and stored at -20uC.

Perinatal L. paracasei Exposure
Based on preliminary studies of dosage and timing of L. paracasei application (data not shown), pregnant BALB/c mice (mothers; n = 5/group; two independent experiments were performed) were treated with an average daily dose of 2610 9 CFU L. paracasei/ mouse, starting from the third trimester of gestation (day -7) and continuing throughout the lactation period until the day of the weaning (day 21). Age-matched control animals received water only (Experimental design; Fig 1).

Bronchoalveolar Lavage and Differential Cell Counts
Offspring from L. paracaseior control-treated mother mice were terminally anesthetized and bronchoalveolar lavage (BAL) was performed two times using 2 ml PBS containing protease inhibitor cocktail (Roche, Mannheim, Germany). Total leukocytes were counted and cytospins (Shadon CytospinH, Shadon Southern Instruments, USA) were stained with hematoxylin and eosin (H&E; HemacolorH, Merck, Darmstadt, Germany) for differential cell counts (200 cells were counted per cytospin). Cell-free supernatants were stored at -20uC for further analysis.

Lung Histology
Whole lungs were fixed with 10% formaldehyde-PBS and paraffin-embedded. Tissue sections were stained with H&E. Airway mucus occlusion was analyzed on periodic acid-Schiffstained (PAS, Sigma-Aldrich) sections.

IgG1, IgE and IgA Antibody Levels in BAL, Serum and Gut Samples
Blood samples were taken by tail bleeding on the day of sacrifice and sera were stored at -20uC. Small intestine was excised and faeces removed by flushing the lumen with 2 ml PBS. The intestine was cut open lengthwise and frozen in 1 ml of protease inhibitor cocktail. After thawing, samples were incubated in 20% saponine solution (Sigma-Aldrich) overnight to permeabilize cell membranes. Supernatants were collected after centrifugation (2000 g; 10 min) and stored at -20uC. Levels of serum anti-Bet v 1 IgG1 and IgE and total levels of gut IgA were measured by ELISA. For the detection of antigen-specific antibodies, microtitre plates (Nunc, Wiesbaden, Germany) were coated with Bet v 1 (2 mg/ml). Serum samples were diluted 1/1000 for IgG1 and 1/10 for IgE. For the detection of total IgA, plates were coated with anti-IgA (2 mg/ml; Pharmingen, San Diego, CA). BAL fluid was applied neat and gut lavages were diluted 1/2500. Rat anti-mouse IgG1, IgE and IgA antibodies (1/500; Pharmingen, San Diego, CA) were applied and peroxidase-conjugated mouse anti-rat IgG antibodies (1/1000; Jackson, Immuno Labs., West Grove) were used for the detection. Data represent mean values 6 SEM of optical density (OD) values from duplicate wells.

Ex vivo Stimulation of Cell Cultures
Spleens, lungs, mediastinal lymph nodes, and mesenteric lymph nodes (MLN) were collected on sacrifice and single-cell suspensions were prepared from each organ. Where indicated, cells (2.5610 6 / ml) were stimulated with Bet v 1 (10 mg/ml), Concanavalin A (ConA) (1 mg/ml) or media alone in 96-well plates at 37uC for 72 h in culture medium (RPMI 1640 supplemented with 10% heatinactivated FCS, 2 mM L-glutamine, 100 U/ml penicillin, 100 mg/ ml streptomycin). Proliferative responses of spleen and MLN cell cultures were determined by scintillation counting after addition of 3 [H]-thymidine (0.5 mCi/well; Amersham, Buchter, Braunschweig, Germany) for the last 18 h. In addition, spleen cells from offspring of sham-treated mothers were incubated with L. paracasei (10 7 CFU/ ml) admixed to Bet v 1 (10 mg/ml) for 48 h.

Cytokine Measurements
Levels of IL-4, IL-5, IL-10, and TGF-b were measured by ELISA (Endogen, Cambridge, MA) according to the manufacturer's instructions. Levels of IFN-c were measured as previously described [19].

Quantification of Foxp3 mRNA Expression in Offspring Lung by Real-time RT-PCR
Total RNA was extracted from lung tissue of offspring from L. paracaseior control-treated mothers as described previously [18]. The housekeeping gene 5-aminolevulinic acid synthase 1 (ALAS1) (Universal ProbeLibrary probe #64; Roche) was used to standardize the amount of sample cDNA. Data are presented as the relative ratio of the target gene to ALAS1.

Generation and Stimulation of Bone Marrow Derived Dendritic Cells
Bone marrow derived dendritic cells (BM-DC) were generated as previously described [20]. Briefly, the bone marrow precursors were isolated from femurs and tibias of BALB/c mice. Cells were cultured at 2610 5 /ml in bacteriological Petri dishes in 10 ml culture medium with GM-CSF (20 ng/ml; Sigma-Aldrich). Fresh medium was added at days 3 and 6 and BM-DC were used on day 10 of culture. BM-DC were stimulated with L. paracasei at a final concentration of 10 6 or 10 7 CFU/ml for 24 h, after which supernatants were tested by ELISA. LPS (1 mg/ml) was used as a positive control.

Statistical Analysis
All data are shown as mean 6 SEM. Significance was analyzed using the non-parametric Mann-Whitney U-test (Graph Prism; Graph Pad Software, Inc, San Diego, CA). Significant differences were considered at p,0.05.

Maternal Perinatal L. paracasei Application does not Influence Sensitization to Bet v 1 in Offspring
In a mouse model of birch pollen allergy (Fig 1), sensitization and challenge of offspring induced allergen-specific antibody production. Maternal application of L. paracasei did not hinder this process, and anti-Bet v 1 IgG1 and IgE antibody titres in serum did not differ between offspring of L. paracasei-treated and control mice (IgG1: 1.1260.16 vs. 1.6560.17, p = 0.76; IgE: 0.5860.12 vs. 0.6160.14, p = 0.89; n $8 per group; data from 3 separate experiments which each yielded similar results).

Maternal Perinatal L. paracasei Supplementation Protects
Offspring from Development of Allergic Airway Inflammation L. paracasei was detected by PCR analysis using strain specific primers exclusively in faecal samples from L. paracasei-treated, and not from control mothers (data not shown). Importantly, L. paracasei was not detected in offspring of either L. paracasei-treated or control mothers (data not shown).
Perinatal administration of L. paracasei to mothers significantly inhibited the development of airway inflammation in their progeny. Numbers of BAL eosinophils were significantly reduced in sensitized/challenged offspring from L paracasei-treated mothers as compared to sensitized/challenged offspring from sham-treated control mothers (Fig 2A). Offspring of L. paracasei-exposed mothers showed considerably reduced peribronchial inflammation with only rare mononuclear cell infiltrates (Fig 3B) as compared to offspring of control mice that showed massive cellular infiltration around bronchi and blood vessels (Fig 3A). In addition, goblet cell hyperplasia and mucus hypersecretion were strongly reduced in offspring of L. paracasei-treated mothers as compared to the control group (Fig 3C-D). Furthermore, levels of the IL-5 in BAL fluid (Fig 2B), as well as lung tissue ( Fig 2C) and mediastinal lymph node (Fig 2D) cell culture supernatants from offspring of L. paracaseiexposed mothers were significantly reduced compared with that from offspring of sham-treated controls.

Maternal L. paracasei Supplementation Results in Upregulation of Foxp3 in Offspring Lung Tissues
Since we observed marked suppression of airway inflammation and IL-5 responses in the lungs of offspring from L. paracaseitreated mothers, we examined levels of Foxp3 mRNA expression in this organ. Offspring of L. paracasei-treated mothers exhibited increased expression of Foxp3 mRNA in comparison to controls (Fig 4A). Additionally, offspring of mothers exposed to L. paracasei had markedly increased serum levels of the regulatory cytokine TGF-b as compared to offspring from control mothers ( Fig 4B). As TGF-b is an important switch factor for B cells to produce IgA antibodies [21] and a potential regulatory role of IgA against allergic disease has been suggested [22] we examined IgA production in offspring. We found that maternal L. paracasei treatment had no effect on total IgA levels in BAL from offspring (0.960.2 OD vs. 1.060.2 OD, p = 0.678; offspring from perinatal L. paracasei vs. offspring from perinatal sham, respectively; n$5 per group; data from 2 separate experiments which each yielded similar results). However, offspring from L. paracasei treated mothers had increased levels of total gut IgA when compared to control mice (2.160.0 OD vs. 1.960.0 OD, p = 0.003; offspring from perinatal L. paracasei vs. offspring from perinatal sham, respectively; n$5 per group; data from 2 separate experiments which each yielded similar results).

Maternal L. paracasei Supplementation Inhibits Systemic Allergen-specific Recall Responses in Offspring
To test whether the effects on airway inflammation in offspring were associated with changes in distal immune responses, allergenspecific responses were examined in offspring splenocyte and MLN cultures. Production of IL-4 and IL-5 in Bet v 1-stimulated splenocyte cultures was significantly lower in the offspring of L. paracasei-supplemented mothers compared with offspring of control mothers (Fig 5A-B). Interestingly, production of the regulatory cytokine IL-10 ( Fig 5C) and the Th1 cytokine IFN-c (Fig 5D) was also markedly lower in offspring of L. paracasei-treated mothers as compared to offspring of controls. Furthermore, IL-4 and IFN-c secretion by Bet v 1-stimulated MLN cells from offspring of L. paracasei treated mothers were significantly lower than for offspring of control mothers (IL-4: 1.860.5 pg/ml vs. 6.062.0 pg/ml, p = 0.02; and IFN-c: 73.5619.0 pg/ml vs. 220.7664.8 pg/ml, p = 0.03; perinatal L. paracasei vs. perinatal sham, respectively; n $4 per group; data from 2 separate experiments which each yielded similar results). We further assessed whether L. paracasei might have similar effects on allergen-specific cytokine production in vitro. Splenocytes from offspring of sham-treated mothers were cultured with Bet v 1 in the presence or absence of L. paracasei. Bet v 1-induced IL-4 and IL-5 production was significantly reduced in the presence of L. paracasei (Fig 5E-F).

paracasei Supplementation Inhibits Mitogeninduced Cytokine Responses in Offspring
Having determined that perinatal L. paracasei supplementation reduced allergen-specific immune responses in offspring, we aimed to assess whether mitogen-induced immune responses were influenced as well. Indeed, ConA-stimulated splenocytes from offspring of L. paracasei-supplemented mothers had significantly reduced secretion of IL-5, IL-10, and IFN-c in comparison to splenocytes from offspring of control mice (Fig 6B-D). Similarly, we observed a trend towards reduced secretion of Th1-and Th2-related cytokines by ConA-stimulated MLN cells in offspring of L. paracaseitreated mothers in comparison to offspring of sham-treated controls (IL-4: 6.361.1 pg/ml vs. 10.262.0 pg/ml, p = 0.23; and IFN-c: 281.6693.5 pg/ml vs. 451.86133.5 pg/ml, p = 0.35; perinatal L. paracasei vs. perinatal sham, respectively; n $3 per group; data from 2 separate experiments which each yielded similar results).

paracasei Supplementation Inhibits Allergenand Mitogen-induced Proliferative Responses of Spleen and MLN Cells in Offspring
Remarkably, maternal L. paracasei-supplementation led to reduced allergen-induced and mitogen-induced proliferative responses of spleen and MLN cells from offspring ex vivo (Fig 6E-F).

L. paracasei Acts through TLR2 and TLR4 to Induce Production of Regulatory Cytokines by Splenocytes
We have shown previously that intranasal application of L. paracasei led to increased mRNA expression of TLR2 and TLR4 receptors in draining lymph nodes [15]. We therefore investigated whether TLR2 or TLR4 are functionally involved in mediating the L. paracasei induced regulatory responses observed in this study. Stimulation of spleen cells from naïve wild type mice with L. paracasei induced production of IL-10 and TGF-b, and both IL-10 and TGF-b production were markedly reduced in TLR4 2/2 derived splenocytes (Fig 7A-B). L. paracasei-stimulated splenocytes from TLR2 2/2 mice had markedly impaired production of TGFb while IL-10 production remained unchanged compared to wildtype splenocytes (Fig 7A-B). As expected, the production of cytokines triggered either by LPS (TLR4 ligand) or Pam3CSK4 (TLR2 ligand) was completely abrogated in spleen cells from TLR4 2/2 or TLR2 2/2 mice, respectively (Fig 7A-B). Thus, L. paracasei mediates the production of regulatory cytokines through TLR4 and partly through TLR2.

Bone Marrow Derived Dendritic Cells May be an Important Source of Regulatory Cytokines Following Activation by L. paracasei
Dendritic cells (DC) are professional antigen presenting cells, which are involved in the polarisation of adaptive T cell immune responses (i.e. Th1, Th2 or regulatory). Stimulation of BM-DC with L. paracasei induced high levels of IL-10 and TGF-b in a dosedependent manner (Fig 7C-D).

Discussion
In this study we have shown that perinatal administration of L. paracasei NCC 2461 to pregnant/lactating mothers markedly attenuated allergic airway inflammation in their offspring. The protective effects were likely mediated by activation of regulatory pathways, signalling through TLR2 and/or TLR4 and were associated with suppression of both local and systemic immune responses.
Our findings are consistent with a recent clinical trial which showed that administration of a combination of probiotic strains to women during pregnancy and lactation without direct infant supplementation reduced the prevalence of eczema at 2 years [23]. The fact that direct infant intervention with probiotics may not be an absolute prerequisite for beneficial outcomes in offspring is an exciting observation, as there are considerable advantages in terms of convenience and compliance offered by maternal as compared to child-directed interventions.
There is a strong body of evidence suggesting a ''window of opportunity'' for modulation of infant immune responses during late gestation and the early perinatal period [13,24,25]. Indeed, farm exposure during pregnancy (as compared to postnatal farm exposure) was associated with the strongest protective effect against development of allergy [26]. We took advantage of this early ''window of opportunity'' and proved that intervention with L. paracasei during the process of maturation of immune system leads to anti-allergic phenotype in offspring. Nevertheless, our experimental approach does not allow us to distinguish whether the L. paracasei-driven effects were due to the prenatal intervention alone or to the combination of prenatal and early postnatal treatment.
Recently, it has been shown that prenatal maternal exposure to Acinetobacter lwoffii F78 prevented the development of asthma in the progeny which was associated with changes in H4 acetylation at the IFNG promoter [27,28,29]. Our findings support the intriguing possibility that stimulation of the maternal immune system during pregnancy with specific probiotic strains may generate signals that induce reprogramming of foetal immune responses, thereby protecting the offspring from development of allergic disease.
We observed that reduced airway inflammation in offspring after aerosol challenge with birch pollen was independent of changes in antibody titres in serum, which were not reduced by maternal exposure to L. paracasei. This situation is similar to human studies, where combined prenatal/postnatal probiotic treatment had no effect on atopic sensitization despite conferring a reduced prevalence of eczema (reviewed by Tang et al. [13]). Furthermore, our observations are in agreement with a study published by Blümer et al., where perinatal Lactobacillus rhamnosus GG (LGG) supplementation suppressed allergic airway inflammation in offspring without having an impact on allergen-specific immunoglobulin titres in serum [30]. Similarly, as described in our study with L. paracasei, L. rhamnosus primarily induces IL-10 production in vitro [31]. On the other hand, other probiotic strains, such as Lactobacillus plantarum, Lactococcus lactis or Lactobacillus casei strain Shirota, eliciting the induction of proinflammatory cytokines in vitro, have been shown to reduce allergen induced IgEdependent basophil degranulation or directly inhibit IgE production in sera [32,33]. In these cases the reduction of allergic immune responses is associated with a shift from typical Th2-based allergic responses to Th1-type immune responses. These observations indicate the complex character of probiotic applications which precise mechanism of action has to be still investigated as well as the clinical consequences.
Furthermore, perinatal L. paracasei treatment increased the production of gut IgA in offspring. Similarly, we have shown previously that intranasal application of probiotic bacteria enhanced production of secretory IgA in BAL of adult mice [14,15]. Secretory IgA might act as ''blocking antibody'' by capturing and neutralizing of the allergens. Recently, Smits et al. proved elegantly that secretory IgA plays a prominent role in preventing the onset of allergic inflammation in a mouse model [21]. In a clinical trial, daily consumption of probiotic bacteria induced beneficial effects on allergic responses which were accompanied by a significant increase in mucosal IgA [34].
However, the importance of increased intestinal IgA in our model remains to be investigated.
With respect to cellular responses, the production of both Th2 (IL-4, IL-5) and Th1 (IFN-c) cytokines was markedly reduced in re-stimulated spleen cells by maternal L. paracasei-treatment. This data are in agreement with our previous work, where the application of L. paracasei to the nasal mucosa led to reduction of allergen-specific Th2 responses as well as of IFN-c in re-stimulated spleen cells [15]. This broad suppression in T cell responsiveness might be an explanation for the observed anti-inflammatory effects in the airways of offspring in the present study.
Surprisingly, maternal L. paracasei-treatment reduced production of IL-10 by allergen-re-stimulated splenocytes from offspring. This phenomenon was also observed in our previous study, where mucosal application of L. paracasei in a model of allergic polysensitization reduced IL-10 production in splenocytes [15], while allergic poly-sensitization and airway challenge itself was associated with increased IL-10 production. We therefore hypothesise that IL-10 is produced as part of a compensatory mechanism to control allergic inflammation, and so may be reduced following L. paracasei treatment if there is associated abrogation of allergic inflammation.
Notably, perinatal application of L. paracasei was associated with suppression of both antigen-specific and mitogen-induced responses in offspring. It remains important to evaluate whether intervention with certain probiotic strains might interfere with active immune responses against infections or vaccines. Even though such an effect has not been reported in published clinical trials, the question deserves to be evaluated in detail and may lead to additional selection criteria for strains destined to perinatal interventions.
De Roock et al. have shown that probiotic bacteria can induce functional Treg cells in vitro and block effector T cells function [35]. We have shown that stimulation of spleen or BM-DC with L. paracasei in vitro induced high levels of IL-10 and TGF-b, and that L. paracasei-induced production of regulatory cytokines by splenocytes is mediated via TLR2 and/or TLR4. Interestingly, certain probiotic bacteria have been shown to prime DC to promote the development of Treg cells [36]. Although we have not directly explored this possibility, it is attractive to postulate that the beneficial effects of perinatal L. paracasei might, at least in part, be explained by the induction of Treg cells through L. paracasei-induced regulatory DC. There is a body of evidence suggesting that the TLR ligands play an important role in prevention of allergic disease [37,38]. Our data clearly demonstrates that TLR2 and TLR4 are important for recognition of L. paracasei and signalling through these receptors leads to induction of regulatory cytokines. In support of this, Conrad et al. have shown that maternal perinatal exposure to cowshedderived bacteria induced resistance to allergic disease in offspring and functional maternal TLR-signalling was prerequisite for the transmission of protection [39].
In summary, our data demonstrate that maternal interventions by way of probiotic supplementation during gestation and the early postnatal period can modulate infant immune responses and thereby protect against unwanted allergic inflammation in the infant. This approach to disease prevention avoids direct interventions to the child, while having the capacity to influence regulation of immune responses in the infant.