Bacillus subtilis Provides Long-Term Protection in a Murine Model of Allergic Lung Disease by Inﬂuencing Bacterial Composition

: Probiotics are an attractive target for reducing the incidence of allergic disease. Bacillus subtilis is a gut-associated probiotic bacteria that can suppress allergic lung disease; however, it is not clear for how long this protection lasts. We exposed C57Bl/6 mice to B. subtilis via oral gavage and challenged them with intranasal house-dust mite for up to 8 weeks. We found that B. subtilis treatment was able to provide protection from eosinophil inﬁltration of the airways for 3 weeks. This loss of protection correlated with an increase in the eosinophil chemoattractant CCL24. Additionally, we demonstrate that B. subtilis treatment altered the bacterial composition by increasing the phylum Bacteroidetes and Verrucomicorbiota. The phylum Verrucomicorbiota was reduced in B. subtilis -treated mice at 8 weeks when protection was lost. These results support B. subtilis as a prophylactic for preventing the production of allergic lung disease and highlights that protection can last up to 3 weeks. This work also expands our understanding of how B. subtilis mediates protection and that in addition to modifying the immune system it is also altering the host microbiota.


Introduction
The number of people with allergies has been on the rise over the last half-century, and currently 20% of the population in developed countries has allergies [1]. The rise in allergic disease correlates with an increasingly sterile living environment and reduced exposure to bacteria, contributing to inappropriate responses by the immune system to non-infectious or innocuous antigens [2]. The rise in allergic disease and reduced early-life exposure to microbes is reflected in the changing bacterial composition of infants. Specifically, children born via caesarean section have reduced bacterial diversity and reduced levels of Lactobacillus and Bifidobacterium, and therefore increased risk of developing allergic disease [3,4]. Probiotic intervention is being pursued as a method to restore this bacterial dysbiosis. Both Lactobacillus and Bifidobacterium haven been shown to reduce the development of allergic disease in mice when administered prophylactically [5,6]. The mechanism of Lactobacillus protection is strain-specific and includes induction of T-regulatory cells, as well as modulation of dendritic cell function [7][8][9]. Although Lactobacillus is generally safe, it requires frequent dosing for efficacy, which poses a risk for immunocompromised patients and can lead to bacteremia [10].
We and others have identified that Bacillus subtilis, a Gram-positive bacterium commonly found in soil and the gastrointestinal tract, can prevent allergic lung inflammation [11,12]. B. subtilis has demonstrated efficacy in protecting from gastrointestinal inflammation, sepsis, and allergic lung disease with infrequent dosing in acute models of disease [13,14]. B. subtilis has been demonstrated to reduce pathogenic bacteria via quorum sensing with implications for reducing infections [15]. In addition, we know that B. subtilis production of exopolysaccharide (EPS) is critical for its immune suppression [13,16]. Previous work has established that B. subtilis-derived EPS induces an anti-inflammatory response by suppressing dendritic-cell function and inducing anti-inflammatory macrophages, as well as via NF-kb and STAT6 pathways [11,12,16].
The process of allergic sensitization is initiated by antigen-presenting cells (APC), such as dendritic cells (DC), which capture and internalize allergens and are essential for HDM-induced type-2 inflammation [17]. Allergen presentation by DCs drives the proliferation of allergen-specific Th2 cells and activation of B cells [18,19]. Eosinophils are recruited to the airway following Th2-produced cytokines IL-4, IL-5, and IL-13 stimulating a variety of cells to produce the eosinophil chemoattractants CCL11 and CCL24 [20,21]. Allergen-specific IgE is produced by B cells after an encounter with IL-4-producing Th2 cells and is dependent on the STAT6 pathway [19].
Our previous work identified that B. subtilis treatment can protect mice from HDMinduced eosinophilia, but we did not determine whether B. subtilis influenced eosinophil chemokines or allergen-specific immunoglobulin production. Additionally, we do not know how long B. subtilis-mediated protection lasts. To determine how long B. subtilis can protect mice from an allergic response, we utilized repeated HDM challenges and found that B. subtilis protects from HDM-induced eosinophilia for 3 weeks. We further determined B. subtilis reduced levels of CCL24, with no impact on immunoglobulin production. Interestingly, we found that B. subtilis exposure altered the composition of bacteria during periods of protection compared to no-treatment mice. These results support the long-term protective effects of B. subtilis exposure, further identify the effects of B. subtilis on the immune response, and point to additional impacts of B. subtilis exposure on the microbiota.

Mice
Male and female C57Bl/6 mice were purchased from Jackson Laboratory (Bar Harbor, ME, USA) and bred in house at Midwestern University, Downers Grove, IL, in a specific pathogen-free facility. All animal protocols were approved by the Midwestern University Institutional Animal Care and Use Committee.

Allergic Lung Disease Model
C57B1/6 mice, age 6-10 weeks old, were treated through oral gavage with 10ˆ9 wildtype B. subtilis spores, 10ˆ9 epsH spores unable to produce exopolysaccharide (EPS), or sterile PBS as a negative control. B. subtilis and epsH spores were generated via exhaustion, as previously described [22]. After 24 h, mice were sensitized intranasally with 100 µg of house-dust mites (HDM, D. pteronyssinus XPB82S3A2.5, Stallergenes Greer, Lenoir, NC, USA) in 100 µL of sterile PBS. After 5 days of initial treatment, mice were once again treated orally with B. subtilis spores as described above. On days 6-9, mice that were in the "week 1" group were sensitized intranasally with 25 µg HDM in 100 µL sterile PBS. These mice were euthanized on day 12. Bronchial alveolar lavage (BAL) was collected by flushing lungs with 0.8 mL BAL fluid (10% FCS, 1 mM EDTA, 1X PBS). Blood was collected through cardiac puncture and spun down at 8000 rpm for 8 min, and serum was collected and frozen for ELISA assays. Lungs were excised and fixed in formalin for 24 h. BALF was counted using a hemocytometer to determine total cells/mL, cytospun onto slides, and assessed for percent eosinophils, neutrophils, lymphocytes, and macrophages using Diff-Quick staining and counting 100 cells per slide (Dade Behring, Deerfield, IL, USA). The total cells/mL of each cell type was determined by calculating the number of cells/mL based on the percent of each cell type. On days 13-16, mice that were in the "week 2" group were sensitized intranasally with 25 µg HDM in 100 µL sterile PBS. On day 19, these mice were euthanized, and tissues were collected as described above. Intranasal sensitizations were repeated for groups "week 3" through "week 8," mice were euthanized, and tissues were collected.

ELISA
CCL24: 96-well plates were coated with 100 µL of biotinylated anti-mouse CCL24 in carbonate buffer at 4 • C overnight (Peprotech, Cranbury, NJ, USA). The plates were washed three times with PBST buffer (1xPBS, 0.05% Tween 20) and blocked with 3% BSA in PBS for 2 h at room temperature. The plates were washed again and BALF was added to the wells along with antibody standard recombinant murine CCL24 (Peprotech). After three washes, the plates were treated with secondary antibody Biotin anti-murine CCL24 (Peprotech) and incubated at room temperature for 1 h. The plates were washed and stained with HRP-avidin (Biolegend, San Diego, CA, USA) in the dark at room temperature for 1 h. After washing the plates, TMB substrate solution (Biolegend) was added to the wells, and once there was a color change the plates were measured on a MultiSkan™ plate reader (Thermo Fisher, Waltham, MA, USA) at OD 405 nm.
HDM-specific IgE and IgG1: 96-well plates were coated with 100 µL of HDM (5 µg/µL) in carbonate buffer at 4 • C overnight. The plates were washed three times with PBST buffer (1xPBS, 0.05% Tween 20) and blocked with 3% BSA in PBS for 2 h at room temperature. The plates were washed again and serum was added to the wells. After subsequent washes, the plates were treated with secondary antibody Biotin anti-mouse IgE or IgG1 and incubated at room temperature for 1 h. The plates were washed and stained with HRP-avidin (Biolegend) in the dark at room temperature for 1 h. After washing the plates, TMB substrate solution (Biolegend) was added to the wells and once there was a color change, the plates were measured on a MultiSkan™ plate reader (Thermo Fisher, Waltham, MA, USA) at OD 405 nm.
Total IgE and IgG1: 96-well plates were coated with 100 µL of rat anti-mouse IgE or anti-mouse IgG1 (BD Biosciences, San Jose, CA) in carbonate buffer at 4 • C overnight. The plates were washed three times with PBST buffer (1xPBS, 0.05% Tween 20) and blocked with 3% BSA in PBS for 2 h at room temperature. The plates were washed again and serum was added to the wells along with antibody standards (BD Biosciences). The plates were incubated at 4 • C overnight. The plates were washed and stained with HRP-avidin (Biolegend) in the dark at room temperature for 1 h. After washing the plates, TMB substrate solution (Biolegend) was added to the wells and once there was a color change, the plates were measured on a MultiSkan™ plate reader (Thermo Fisher, Waltham, MA, USA) at OD 405 nm.

DNA Extraction and 16S rRNA Amplicon Sequencing
Fecal samples were collected from mice during the allergic lung inflammation model on days −1 and 5 (before o.g. treatment) and the day of euthanization across all treatment groups into sterile tubes and stored at -80°C. DNA extraction was performed on thawed fecal samples using the E.Z.N.A. Tissue DNA Kit (Omega BioTek, Norcross, GA, USA) according to the manufacturer's instructions. The extracted DNA was sent to Argonne National Laboratory for 16S rRNA amplicon sequencing using the Illumina MiSeq instrument. The sequences obtained were processed through Illumina-utils software to demultiplex, align, merge, and quality filter the sequence data. Taxonomy was assigned using the SILVA database version 138. Alpha-and beta-diversity was determined using QIIM31.9.1 and QIIME2 to generate PCoA ordinations.

Statistics
All analyses were carried out using GraphPad Prism Software (La Jolla, CA, USA).

Bacillus Subtilis Protection from House-Dust Mite-Induced Eosinophilia Wanes after 3 Weeks
To determine how long B. subtilis can protect against eosinophilia, we challenged mice with HDM for up to 8 weeks after B. subtilis exposure, as shown in Figure 1A. We measured eosinophils in the bronchial alveolar fluid lavage (BALF) and found that at 4 weeks mice receiving B. subtilis were no longer protected from eosinophilia ( Figure 1B,C). Interestingly, the protection from weeks 2 and 3 seemed to be independent of exopolysaccharide production, as demonstrated by the reduced eosinophils in mice treated with epsH spores that were unable to produce EPS ( Figure 1B,C). These results expand on our previous knowledge that EPS is important in the B. subtilis-mediated protection from eosinophilia at week 1 [11]. The reduction in eosinophils was due to a shift to macrophages being the primary cell in the BALF, as shown in Figure 2A,D. Additionally, B. subtilis treatment did significantly alter neutrophil infiltration and had a moderate impact on lymphocyte recruitment ( Figure 2B,C,E,F).

Bacillus Subtilis Protection from House-Dust Mite-Induced Eosinophilia Wanes after 3 Weeks
To determine how long B. subtilis can protect against eosinophilia, we challenged mice with HDM for up to 8 weeks after B. subtilis exposure, as shown in Figure 1A. We measured eosinophils in the bronchial alveolar fluid lavage (BALF) and found that at 4 weeks mice receiving B. subtilis were no longer protected from eosinophilia ( Figure 1B,C). Interestingly, the protection from weeks 2 and 3 seemed to be independent of exopolysaccharide production, as demonstrated by the reduced eosinophils in mice treated with epsH spores that were unable to produce EPS ( Figure 1B,C). These results expand on our previous knowledge that EPS is important in the B. subtilis-mediated protection from eosinophilia at week 1 [11]. The reduction in eosinophils was due to a shift to macrophages being the primary cell in the BALF, as shown in Figure 2A,D. Additionally, B. subtilis treatment did significantly alter neutrophil infiltration and had a moderate impact on lymphocyte recruitment ( Figure 2B,C,E,F).  Data represent mean ± SEM representative (n = 3-10). * p < 0.5, *** p < 0.005, and **** p < 0.001 by Student t test compared to HDM alone in the same week. HDM, house-dust mite; WT, wild-type; epsH spores, B. subtilis spores unable to produce exopolysaccharide.

CCL24 Production Is Suppressed by Bacillus subtilis Treatment for 3 Weeks
Eosinophil recruitment to the lung is driven in part by the production and presence of the chemokine CCL24. We measured the levels of CCL24 in the BALF each week for 4 weeks according to the B. subtilis exposure and HDM treatment outlined in Figure 1A. We found that CCL24 levels were significantly reduced for the first 3 weeks, but not after 4 weeks following B. subtilis exposure (Figure 3). Again, we found that suppressed CCL24 at week 3 was independent of EPS production, but EPS was involved in the CCL24 suppression at weeks 1 and 2 ( Figure 3).

B. subtilis Treatment Does Not Influence Immunoglobulin Production
House-dust mite-induced eosinophilia is driven by the adaptive immune response and includes involvement of immunoglobulins IgG1 and IgE. To determine whether B. subtilismediated protection was influenced by a reduction in IgG1 or IgE production, we measured serum levels of total and HDM-specific IgG1 and IgE. We found that B. subtilis treatment had no effect on HDM-specific IgG1 and IgE or total IgG1 and IgE levels ( Figure 4A-D).  Figure 1A. Data represent mean ± SEM representative (n = 5-10). * p < 0.5, ** p < 0.05, *** p < 0.005, and **** p < 0.001 by Student t test compared to HDM alone in the same week.

CCL24 Production Is Suppressed by Bacillus subtilis Treatment for 3 Weeks
Eosinophil recruitment to the lung is driven in part by the production and presence of the chemokine CCL24. We measured the levels of CCL24 in the BALF each week for 4 weeks according to the B. subtilis exposure and HDM treatment outlined in Figure 1A. We found that CCL24 levels were significantly reduced for the first 3 weeks, but not after 4 weeks following B. subtilis exposure (Figure 3). Again, we found that suppressed CCL24 at week 3 was independent of EPS production, but EPS was involved in the CCL24 suppression at weeks 1 and 2 ( Figure 3). Data represent mean ± SEM representative (n = 3-10). ** p < 0.05 and *** p < 0.005 by one-way ANOVA followed by Tukey's multiple-comparisons test was performed compared to HDM alone in the same week. n.s, no sample.

CCL24 Production Is Suppressed by Bacillus subtilis Treatment for 3 Weeks
Eosinophil recruitment to the lung is driven in part by the production and presence of the chemokine CCL24. We measured the levels of CCL24 in the BALF each week for 4 weeks according to the B. subtilis exposure and HDM treatment outlined in Figure 1A. We found that CCL24 levels were significantly reduced for the first 3 weeks, but not after 4 weeks following B. subtilis exposure ( Figure 3). Again, we found that suppressed CCL24 at week 3 was independent of EPS production, but EPS was involved in the CCL24 suppression at weeks 1 and 2 ( Figure 3). Data represent mean ± SEM representative (n = 3-10). ** p < 0.05 and *** p < 0.005 by one-way ANOVA followed by Tukey's multiple-comparisons test was performed compared to HDM alone in the same week. n.s, no sample.

B. subtilis Treatment Does Not Influence Immunoglobulin Production
House-dust mite-induced eosinophilia is driven by the adaptive immune response and includes involvement of immunoglobulins IgG1 and IgE. To determine whether B. Data represent mean ± SEM representative (n = 3-10). ** p < 0.05 and *** p < 0.005 by one-way ANOVA followed by Tukey's multiple-comparisons test was performed compared to HDM alone in the same week. n.s, no sample.

Bacillus Fecal Composition Is Similar across Mice Regardless of Treatment with B. subtilis Spores
To determine whether the waning protection was due to higher levels of B. subtilis following exposure that dwindled over time, we collected fecal pellets from mice over the course of 8 weeks to assess bacterial composition. We found that mice receiving no B. subtilis and HDM alone and mice receiving WT B. subtilis spores had very similar levels of Bacillus strains identified by 16S rRNA amplicon sequencing ( Figure 5A). Of note, mice treated with epsH B. subtilis spores had elevated levels of Bacillus during the first week following spore administration ( Figure 5A). This elevated Bacillus seemed to be temporarily detrimental, as evidenced by the elevated eosinophilia in epsH B. subtilis-treated mice during week 1 and protection in weeks 2-3 ( Figure 1B).
Allergies 2022, 2, FOR PEER REVIEW 6 subtilis-mediated protection was influenced by a reduction in IgG1 or IgE production, we measured serum levels of total and HDM-specific IgG1 and IgE. We found that B. subtilis treatment had no effect on HDM-specific IgG1 and IgE or total IgG1 and IgE levels ( Figure  4A-D). Data represent mean ± SEM representative (n = 2-4). Oneway ANOVA followed by Tukey's multiple-comparisons test was performed for each immunoglobulin and no statistical differences were found.

Bacillus Fecal Composition Is Similar across Mice Regardless of Treatment with B. subtilis Spores
To determine whether the waning protection was due to higher levels of B. subtilis following exposure that dwindled over time, we collected fecal pellets from mice over the course of 8 weeks to assess bacterial composition. We found that mice receiving no B. subtilis and HDM alone and mice receiving WT B. subtilis spores had very similar levels of Bacillus strains identified by 16S rRNA amplicon sequencing ( Figure 5A). Of note, mice treated with epsH B. subtilis spores had elevated levels of Bacillus during the first week following spore administration ( Figure 5A). This elevated Bacillus seemed to be temporarily detrimental, as evidenced by the elevated eosinophilia in epsH B. subtilis-treated mice during week 1 and protection in weeks 2-3 ( Figure 1B). One-way ANOVA followed by Tukey's multiple-comparisons test was performed for each immunoglobulin and no statistical differences were found.

Bacterial Composition Clustering Correlates with Protection from Eosinophilia
This early difference induced between epsH and WT B. subtilis was also observed when comparing all bacterial compositions, as evidenced by all three treatment groups clustering separately early in treatment and epsH spores clustering more closely with HDM alone during the time of elevated eosinophilia. However, at day 5 epsH spore treatment changed its clustering to be more closely related to WT B. subtilis spores, correlating to a shift to protection from eosinophilia ( Figure 5B). Additionally, we found that bacteria phylum was altered by B. subtilis treatment, specifically increased levels of the phylum Bacteroidetes, as well as elevated Verrucomicorbiota early in disease during B. subtilis-mediated protection ( Figure 5C). In addition to elevated Bacteroidetes and Verrucomicorbiota in B. subtilis treatment, there was reduced representation by the phylum Firmicutes ( Figure 5C).

Bacterial Composition Clustering Correlates with Protection from Eosinophilia
This early difference induced between epsH and WT B. subtilis was also obs when comparing all bacterial compositions, as evidenced by all three treatment g clustering separately early in treatment and epsH spores clustering more closely HDM alone during the time of elevated eosinophilia. However, at day 5 epsH spore ment changed its clustering to be more closely related to WT B. subtilis spores, corre to a shift to protection from eosinophilia ( Figure 5B). Additionally, we found that ba phylum was altered by B. subtilis treatment, specifically increased levels of the ph Bacteroidetes, as well as elevated Verrucomicorbiota early in disease during B. su mediated protection ( Figure 5C). In addition to elevated Bacteroidetes and Verrucom biota in B. subtilis treatment, there was reduced representation by the phylum Firm ( Figure 5C).

Discussion
This work expands our understanding of the duration of protection provided by B. subtilis and offers further evidence for its appeal as a probiotic for preventing allergic disease. Unlike other probiotics that require early and frequent dosing, B. subtilis is effective at 6 weeks of age and last for up to 3 weeks. For example, many of the experiments investigating the efficacy of the probiotic lactobacillus utilize daily dosing [5]. B. subtilis provides a novel bacterium to prevent allergic disease with infrequent dosing required.
Current treatments for allergic disease focus on improving symptoms while rarely addressing the mechanism of disease [23]. Treatments that do address the mechanism, such as desensitization studies, involve high patient risk for anaphylactic reactions, are time consuming and expensive, and have only moderate efficacy [24]. We identified B. subtilis as a candidate for treating allergic disease and further characterized B. subtilis-derived EPS as a bacterial-derived molecule that elicits immunosuppression. Specifically, we found that exposure to B. subtilis WT spores resulted in reduced production of eosinophilia at week 1, whereas epsH spore treatment did not show a significant decrease in eosinophils at this timepoint. Additionally, we found that exposure to B. subtilis EPS-producing WT spores resulted in reduced production of the eosinophil chemokine CCL24 at week 1, whereas EPS was not required for B. subtilis treatment to reduce CCL24 levels or eosinophilia at week 3. CCL24 is produced by many cell types in response to IL-4, including airway epithelial cells, airway smooth-muscle cells, vascular endothelial cells, macrophages, and eosinophils [25][26][27]. We suspect that the mechanism of B. subtilis WT spores and epsH spores reducing CCL24 at different type points is due to suppression of CCL24 production by different cell types.
There is growing evidence that the resident microbiota influences susceptibility to allergic disease [28]. Although we and others have identified the immunosuppressive effects of B. subtilis and have investigated the effects on immune-cell function and activation, no one had previously investigated the impact on the host microbiota. We found that treatment with B. subtilis spores influences the bacterial composition of treated mice. Interestingly we found minimal change in Bacillus strains observed by 16S rRNA amplicon sequencing, which is in line with previous reports showing that B. subtilis spore treatment does not establish elevated colonization, with CFUs/gm feces declining rapidly and reaching the limit of detection around 4 days post exposure [29]. Our work highlights the importance of host microbiota on allergic disease and sheds light on how probiotics can influence bacterial composition.
Although allergies are increasing in prevalence and contribute to reductions in quality of life, as well as significant health-care costs [30], we are pursuing a novel method for reducing the incidence of disease using an abundant, cheap, and safe probiotic, Bacillus subtilis. Future work is needed to determine whether B. subtilis can serve as a treatment option after the onset of disease or whether its effects are limited to prophylactic treatment. Additionally, previous work has identified that B. subtilis-derived EPS immune suppression is TLR4 dependent [13], and future studies are needed to determine the requirement for TLR4 in protection from allergic airway inflammation.