Dysregulation of Allergic Airway Inflammation in the ... - ATS Journals

Dysregulation of Allergic Airway Inflammation in the Absence of Microbial Colonization ¨ rki4, Julia Cahenzli5, Kathy McCoy5, Tina Herbst1*, Anke Sichelstiel2*, Corinne Scha¨r2, Koshika Yadava3, Kurt Bu Benjamin J. Marsland3*, and Nicola L. Harris6* 1 Institute of Microbiology, and 2Institute of Integrative Biology, Swiss Federal Institute of Technology, Zurich, Switzerland; 3Service de Pneumologie, Centre Hospitalier Universitaire Vaudois Lausanne, Faculty of Biology and Medicine, University of Lausanne, Lausanne, Switzerland; 4Institute for Laboratory Animal Science, University Zurich, Zurich, Switzerland; 5Maurice Mu ¨ ller Laboratories, University of Bern, Bern, Switzerland; and 6 Swiss Vaccine Research Institute and Global Health Institute, Ecole Polytechnique Fe´de´rale de Lausanne, Lausanne, Switzerland

Rationale: The incidence of allergic disorders is increasing in developed countries and has been associated with reduced exposure to microbes and alterations in the commensal bacterial flora. Objectives: To ascertain the relevance of commensal bacteria on the development of an allergic response, we used a model of allergic airway inflammation in germ-free (GF) mice that lack any exposure to pathogenic or nonpathogenic microorganisms. Methods: Allergic airway inflammation was induced in GF, specific pathogen–free (SPF), or recolonized mice by sensitization and challenge with ovalbumin. The resulting cellular infiltrate and cytokine production were measured. Measurements and Main Results: Our results show that the total number of infiltrating lymphocytes and eosinophils were elevated in the airways of allergic GF mice compared with control SPF mice, and that this increase could be reversed by recolonization of GF mice with the complex commensal flora of SPF mice. Exaggerated airway eosinophilia correlated with increased local production of Th2associated cytokines, elevated IgE production, and an altered number and phenotype of conventional dendritic cells. Regulatory T-cell populations and regulatory cytokine levels were unaltered, but GF mice exhibited an increased number of basophils and decreased numbers of alveolar macrophages and plasmacytoid dendritic cells. Conclusions: These data demonstrate that the presence of commensal bacteria is critical for ensuring normal cellular maturation, recruitment, and control of allergic airway inflammation. Keywords: commensal dysbiosis; allergic airway inflammation; dendritic cells; alveolar macrophages

Variations in exposure to environmental microbes during the early years of life have been implicated in protecting against, or enhancing susceptibility to, allergic diseases (1). An increased prevalence of allergies in populations living within developed countries has clearly been documented over the last few decades, a phenomenon that has been hypothesized to be a result of decreased exposure to infectious microorganisms (2). More recently, dysbiosis of commensal bacteria has also been postulated to modulate the development of allergic disease based on findings of distinct compositions of bacterial communities (Received in original form October 2, 2010; accepted in final form March 25, 2011) *These authors contributed equally to this work. Supported by the Swiss Vaccine Research Institute (N.H.). B.J.M. is a Cloetta Medical Research Fellow. The project was funded by a research grant from the Swiss Federal Institute of Technology (TH-1507–3); by UBS AG on behalf of a client; and by the Swiss National Science Foundation (310,030.130029). K.M. is supported with funding from the Canada Research Chairs program. Correspondence and requests for reprints should be addressed to Nicola Harris, EPFL-SV-GHI-UPHARRIS, Station 19, CH-1015 Lausanne, Switzerland. E-mail: [email protected] This article has an online supplement, which is accessible from this issue’s table of contents at www.atsjournals.org Am J Respir Crit Care Med Vol 184. pp 198–205, 2011 Originally Published in Press as DOI: 10.1164/rccm.201010-1574OC on March 25, 2011 Internet address: www.atsjournals.org

AT A GLANCE COMMENTARY Scientific Knowledge on the Subject

Epidemiologic studies have indicated that commensal dysbiosis influences the incidence and severity of allergies, which has fueled interest in the use of probiotics or prebiotics. However, investigations into the mechanisms by which commensal bacteria could regulate allergic immune responses have been hampered by the absence of suitable animal models. What This Study Adds to the Field

This study demonstrates that commensal bacteria function attenuates allergic airway inflammation, and that such regulation is associated with alterations in the populations and maturation status of lung immune cells.

found within the stool samples of allergic and nonallergic infants (3–5). Commensal bacteria colonize the mucus membranes and the skin of humans soon after birth with the greatest density found within the intestine (with z1012 organisms per gram of intestinal content, represented by z1,000 different species) (6). The exact number and diversity of an individual’s community of commensal bacteria seems to be determined by factors occurring in early childhood, including the type of birth (natural vs. caesarean) (7); diet (formula vs. breast-milk) (8); early use of antibiotics; and environmental conditions (9). Although the bulk of bacteria reside within the intestine, the upper respiratory tract also harbors bacteria (10) and is likely to be repetitively inoculated with the normal bacterial flora of the pharynx. In addition, it has been suggested that immune responses in the intestine and lung are linked through as yet undefined mechanisms (11). Thus, microbial colonization and exposure to microbial products or metabolites in the intestine may have profound effects on the lung and vice versa. Understanding whether and how commensal bacteria can modulate allergic diseases is crucial for the development of preventative strategies based on the use of bacteria and their products, including probiotics and prebiotics. To address this question we used in-bred germ-free (GF) mice as an experimental model, allowing us to circumvent the impact that varied nutritional status, diverse genetics and lifestyles, and individual histories of medical care including antibiotic treatment can have on epidemiologic studies. GF mice were housed and maintained under strict conditions to ensure a complete lack of exposure to environmental microorganisms (bacteria, fungi, and viruses). The impact of commensal bacteria was then investigated by comparison of allergic airway inflammation in GF mice, specific

Herbst, Sichelstiel, Scha¨r, et al.: Commensal Bacteria Prevent Exaggerated Allergy

pathogen–free (SPF) mice, and GF mice recolonized with a complex SPF microbiota. Age- and sex-matched mice were maintained under the same environmental conditions and fed an identical diet for the entire duration of these studies. Using this model we demonstrate, for the first time, that commensal bacteria play a fundamental role in shaping the type and extent of Th2 inflammatory responses in the lung. The absence of a commensal microbiota leads to dysregulated maturation and recruitment of dendritic cell (DC) and macrophage populations, an increased basophil response, and an overall exaggerated allergic airway inflammation.

METHODS Mice GF C57BL/6 mice were kindly provided by the Institute of Laboratory Animal Science, University of Zurich, Zurich, Switzerland, or from the Clean Animal Facility, University of Bern, Bern, Switzerland. GF mice were housed in flexible isolators until the day of killing. Fecal samples of the GF mice and swabs of the inner wall of the isolator were cultured under aerobic and anaerobic conditions. Additionally, Gram stains and DNA stains were performed of fecal samples collected immediately before export from the isolators. SPF C57BL/6 mice (Harlan Laboratories, Fu¨llinsdorf, Switzerland) were maintained in a pathogen-free animal facility and were given the same food as GF mice and water ad libitum. Recolonization of 5- to 6week-old GF mice was established by putting an SPF mouse into the same cage for at least 3 weeks before the first immunization. Studies were performed with mice aged 9–12 weeks. Animal experiments were performed according to institutional guidelines and to Swiss federal and cantonal laws on animal protection. Experiments involving GF and SPF JHD mice were performed at McMaster University in Hamilton, Canada, and were conducted with approval from the McMaster University Animal Care Committee.

Protocol for Experimental Allergic Airway Inflammation Mice were immunized with 100 mg ovalbumin (OVA) (Sigma-Aldrich, Steinheim, Germany) and 200 ml 2% aluminum hydroxide (SERVA Electrophoresis GmbH, Heidelberg, Germany) via the intraperitoneal route. On Days 9 and 10 postimmunization mice were challenged intranasally with 100 mg OVA suspended in 50 ml sterile phosphatebuffered saline (PBS). Control mice either received immunization with aluminum hydroxide only and were challenged with OVA in sterile PBS, according to the protocol used for allergic mice, or were untreated. Both control groups showed comparable results. Airway hyperresponsiveness (AHR) was assessed on Day 3 after the last intranasal challenge using whole-body plethysmography (Buxco Electronics, Inc., Petersfield, UK) and killed the day after. To analyze the cellular compartment of the airway lumen bronchoalveolar lavage (BAL) was performed. Total cell numbers per BAL were determined by using Coulter Counter (IG Instrumenten-Gesellschaft AG, Basel, Switzerland). Differential cell counts were performed on cytospins stained with QuickDiff (Dade Behring, Siemens Healthcare Diagnostics, Deerfield, IL) and the percentage of eosinophils, neutrophils, macrophages, and lymphocytes within a total population of 200 cells was determined.

Histology Lungs were inflated with 1 ml of 10% formalin and embedded into paraffin. Prepared sections (4 mm) were stained with hematoxylin and eosin and periodic acid–Schiff reaction using standardized protocols and analyzed with Axioskop 2 plus microscope equipped with AxioCam HRc (Carl Zeiss Microimaging GMbH, Jena, Germany).

Flow Cytometry For analysis of cytokine production, 2.5–5 3 105 cells from BAL cells were stimulated with phorbol myristate acetate, ionomycin, and monensin (Sigma-Aldrich) for 3 hours at 378C in complete Iscove’s Modified Dulbecco’s Media (IMDM) medium containing 10% fetal calf serum. Thereafter, cells were stained with PerCP-labeled anti-CD4 mAb; fixed with 2% paraformaldehyde; permeablized in saponin

199

buffer; and then stained with anticytokine antibodies (IL-4, IL-5, IL10, and IFNg; BioLegend, San Diego, CA). For analysis of Antigen presenting cells (APC) subsets lung tissue was digested using collagenase IV (BioConcept, Worthington, Lakewood, NY) in Iscove’s Modified Dulbecco’s Media (IMDM) medium for 45 minutes at 378C. Samples were then filtered through a 70-mm cell strainer (MILIAN), washed with 0.2% bovine serum albumin and PBS, and counted with the help of a Coulter Counter. Cells were then stained with anti-CD11c, CD11b, F4/80, I-A/I-E, PDCA-1, B220, and several activation markers for flow cytometry. The percentage of T-regulatory cells of the whole CD41 T-cell compartment of lung and BAL cells was estimated by staining with antibodies against CD4, CD25, and Foxp3. For detection of basophils cells were stained with antibodies against CD49b and IgE. A detailed description of all flow cytometry is given in the online supplement and gating strategies are shown in Figure E2. All cells were recorded using FACSCalibur or LSR II (BD Biosciences, San Jose, CA). Samples were analyzed using FlowJo 8.8.6 software (Tree Star Inc., Ashland, OR).

ELISA Antibody titers of IgA and IgE in BAL fluid and serum were tested using sandwich ELISA as described (12).

Statistical Analysis Student t test (unpaired, two-tailed) was used to calculate significance levels between treatment groups. P values of less than 0.05 were considered significant. Graph generation and statistical analysis was performed using Prism version 4.0c software (GraphPad, La Jolla, CA).

RESULTS Regulation of Allergic Airway Inflammation by the Commensal Microbiota

To directly assess the impact of commensal bacteria on allergic airway inflammation we induced an OVA-specific Th2 inflammatory response in the lungs of SPF mice, GF mice, and GF mice recolonized with a complex SPF microbiota (Figure 1A). GF mice exhibited increased AHR on methacholine challenge compared with SPF mice as determined using a whole-body unrestrained plethysmograph (Figure 1B). Although this is not a direct invasive measurement of lung function, these data are highly indicative of exaggerated airway constriction and Th2 inflammation. Increased AHR in the absence of commensal bacteria was concomitant with an increase in the total number of cells infiltrating the airways (Figure 1C), which could be accounted for by elevated numbers of eosinophils (Figure 1D) and lymphocytes (Figure 1E). Prior recolonization of GF mice was sufficient to ensure that this exaggerated Th2 response did not develop (Figures 1C–E). These changes could also be observed in histologic sections where OVA-sensitized and airwaychallenged GF mice exhibited increased goblet cell hyperplasia and increased perivascular and peribronchial inflammatory cell infiltration compared with SPF mice (Figure 1F). To investigate the activation state and function of Th2 cells, CD41 T cells isolated from the BAL were stimulated ex vivo with phorbol myristate acetate and ionomycin and the production of IL-4, IL-5, IL-10, and IFN-g determined by intracellular cytokine staining. In keeping with the increased allergic infiltrate, the percentage of CD41 T cells producing IL-4 and IL-5 was increased in GF mice compared with SPF or recolonized groups (Figures 2A and 2B). By contrast, the percentage of cells producing the regulatory cytokine IL-10 (Figure 2C), or the prototypic Th1-type cytokine IFN-g (Figure 2D), were similar for all groups. Additional multiplex or ELISA analysis of BAL fluid confirmed the increase of IL-4 in situ in inflammatory GF samples, whereas no significant differences between GF and SPF mice were noted for IL-5, IL-13, IL-10, IFN-g, or transforming-growth factor-b (see Figure E1).

200

AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE

VOL 184

2011

Figure 1. Commensal bacteria down-modulate exacerbated allergic airway inflammation. (A) Experimental setup of ovalbumin (OVA)-induced allergic airway inflammation in C57BL/ 6 specific pathogen–free (SPF), germ-free (GF), or recolonized mice. (B) Airway hyperresponsiveness (AHR) as assessed in OVA (OVA/alum immunized, OVA challenged) or control (alum only immunized, OVA challenged) mice by wholebody plethysmography (n 5 5–9). Numbers of (C ) total cells, (D) eosinophils, and (E ) lymphocytes in the bronchoalveolar lavage (BAL) as assessed by counting total BAL cells and performing differential cell counting of cytospins (controls, n 5 6; OVA, n 5 10–14). Results are pooled from three experiments and are representative of five independent experiments. Mean 6 SEM is shown. Increased infiltration and mucus production in GF compared with SPF OVA-immunized and airway-challenged mice. (F ) Histologic sections of lungs from SPF and GF control or OVA mice were stained with hematoxylin and eosin (H&E) or periodic acid– Schiff (PAS) and analyzed by light microscopy (black bar 5 100 mm). Pictures show representative samples of four to six mice per group from one experiment and are representative of two independent experiments. *P , 0.05; **P , 0.01; ***P , 0.001.

These results provide evidence for a selective increase in the production of IL-4 within the airways of GF mice after OVA immunization and intranasal challenge, which does not seem to be related to a dysregulated development of type 1 immunity or altered production of regulatory cytokines. Altered Antibody Production in the Absence of Commensal Bacteria Is not Responsible for Enhanced Allergic Airway Inflammation

IgE production has previously been reported to be exaggerated in GF mice (13–16). We therefore investigated the production of antibodies after OVA immunization and challenge. Total IgE levels were nonsignificantly elevated in the serum of GF mice before OVA immunization and challenge (Figures 3A and 3B, control groups). As expected, levels of total IgE in the serum increased after OVA immunization in all groups (Figure 3A). However, GF mice exhibited an augmented IgE response after OVA intranasal challenge compared with the SPF or recolonized groups (Figure 3B). This correlated with enhanced production of IL-4 in the airways and indicated that the most striking impact of com-

mensal bacteria on this Th2-mediated response was consequent to local exposure to OVA in the lung. Consistent with these findings, total IgE levels in the BAL and the production of OVA-specific IgE could only be detected in significant amounts in the serum of GF mice (Figures 3C and 3D). No significant differences were noted for OVA-specific IgG1 in the serum of any group (Figure 3G). Sudo and coworkers (15) also demonstrated increased basal and OVA-induced IgE in GF mice and reported that monocolonization with Bifidobacterium infantis could reverse elevated IgE production but only if the bacterium was present from birth. In the current report we show that exposure of GF mice to a complex commensal flora as adults is also effective at reversing elevated IgE production (Figures 3A–D). These data indicate that exposure to a mixture of bacterial species may be more potent at attenuating IgE production compared with a single species. Accordingly, the findings by Sudo and coworkers (15) that bacterial exposure must occur at birth might simply reflect a difference in the total length of time required for a single bacterium, or bacterial species, compared with a complex mixture of bacteria to modulate IgE production.


201

Figure 2. The absence of commensal bacteria results in enhanced Th2 cytokine production by bronchoalveolar lavage (BAL) T cells. C57BL/6 specific pathogen–free (SPF), germ-free (GF), and recolonized mice were subjected to ovalbumin (OVA) immunization and challenge (OVA) or OVA challenge only (control) as depicted in Figure 1. Percentage of airway CD41 T cells producing (A) IL-4, (B) IL-5, (C ) IL-10, or (D) IFN-g after ex vivo restimulation with phorbol myristate acetate, ionomycin, and monesin were determined by intracellular staining and flow cytometry. Results are pooled from two experiments (n 5 7–10) and are representative of four independent experiments. Mean 6 SEM is shown. *P , 0.05; **P , 0.01.

Smits and coworkers (17) recently reported that the delivery of cholera-toxin pulsed DC to the airways of experimental mice promotes the production of local IgA and provides protection against allergic airway inflammation. We therefore investigated the secretion of IgA into the BAL of OVA-immunized and intranasally challenged GF mice. In keeping with previous reports of attenuated IgA in GF mice (6), significantly decreased amounts of total IgA were detected in the BAL of control GF mice compared with control SPF mice (Figure 3E). However, total and OVA-specific IgA levels in the BAL were similar for all groups having received both OVA immunization and intranasal challenges (Figures 3E and 3F). Taken together, these data indicate that the absence of commensal bacteria leads to attenuated production of IgA in the airways before allergen sensitization, and to exaggerated total and OVA-specific IgE responses after intranasal challenge. To determine whether the observed alterations in antibody production contributed to the enhanced airway inflammation observed in GF mice we investigated OVA-mediated allergic eosinophilia in GF JHD mice, which carry a targeted mutation in the JH region of the IgM locus and cannot generate mature B cells (18). As observed for C57BL/6 mice, the absence of commensal bacteria in JH gene deficient mice (JHD) resulted in enhanced allergic airway eosinophilia (Figure 3H), although this did not reach statistical significance. These data indicate that although commensal bacteria modulate antibody production, this process is independent of their impact on allergic airway inflammation. Commensal Bacteria Alter the Number and Activation Status of Lung APCs

Numerous APC populations have been described in the lung and it is clear that they play a fundamental role in shaping the

Figure 3. Elevated antibody production in the absence of commensal bacteria is not responsible for increased allergic airway inflammation. (A– G ) Antibody titers in C57BL/6 specific pathogen–free (SPF ), germ-free (GF), and recolonized mice subjected to ovalbumin (OVA) immunization and challenge (OVA) or OVA challenge only (control) as depicted in Figure 1. Total levels of IgE are shown for (A) serum before intranasal OVA challenge or (B) serum after OVA challenge (controls, n 5 4–9; OVA, n 5 12–20) and (C ) Bronchoalveolar lavage (BAL) fluid after OVA challenge (controls, n 5 4; OVA, n 5 6–10). (D) OVA-specific IgE in serum after intranasal OVA challenge. (E ) Total and (F ) OVA-specific IgA in BAL after intranasal OVA challenge (n 5 3–6). (G) OVA-specific IgG1 levels in serum after intranasal OVA challenge (n 5 3–6). Results A–C are pooled data from three experiments and are representative of five independent experiments. Results D–G are data from one experiment and are representative of three independent experiments. (H) Total numbers of eosinophils in BAL of JH gene deficient mice (JHD) GF or SPF mice after OVA immunization and intranasal challenge. Results are pooled from two independent experiments (n 5 3–9). Mean 6 SEM is shown. *P , 0.05; **P , 0.01; ***P , 0.001.

polarization and effector function of T cells either locally or after migration to the draining lymph nodes. We hypothesized that the APCs in the lung might be directly influenced by the host’s microbiota. As such, we next proceeded to investigate the

202


VOL 184

2011

Figure 4. Commensal bacteria impact on the number and activation of lung conventional dendritic cells (cDCs) in allergic airway inflammation. cD11c1 cells in the lungs of ovalbumin (OVA) or control C57BL/6 specific pathogen–free (SPF) and germ-free (GF) mice were analyzed by flow cytometry as depicted in Figure E2. (A) The following cell subsets of CD11b1classIIhi cDCs, CD11b1classIIint cells, and CD11b2 cDCs were defined as shown in Figure E2A and as indicated in representative fluorescence-activated cell sorter (FACS) plots of SPF and GF OVA mice. (B–D) Total cell number and activation state of each population was determined by flow cytometry. Ox40L was not detected on CD11b2 cDCs. Activation markers are shown as geometric mean of fluorescence intensity 6 SEM and are normalized to isotype control antibodies or fluorescence minus one controls. Total numbers are shown as mean 6 SEM. Results shown are from one experiment (controls, n 5 3–5; OVA, n 5 6) and are representative of two independent experiments. *P , 0.05; **P , 0.01; ***P , 0.001.

number and functional status of the major APC populations in the lung (19). Dendritic cells from the lung were first distinguished by their expression of CD11c but lack of the macrophage marker F4/80 (see Figure E2A). These CD11c1 F4/802 cells were further separated based on their expression of CD11b and major histocompatibility class (MHC) class II, yielding three distinct subpopulations of cells (Figure 4A). The number of

CD11b1MHCIIhi conventional DC (cDC) showed a reduced trend in total numbers in the lung of GF mice; however, this did not reach statistical significance (Figure 4B). These CD11b1 MHCIIhi cDCs did, however, exhibit overall lower levels of the activation markers CD40, CD80, CD86, CD137, and OX40L in GF mice compared with SPF controls (Figure 4B). The two other cell populations assessed, CD11b1MHCIIint cells (Figure 4C)


203

Figure 5. Commensal bacteria impact on the number and function of basophils, alveolar macrophages, and plasmacytoid dendritic cells (pDCs) in allergic mice. Cell types in the lungs of ovalbumin (OVA) or control C57BL/6 specific pathogen–free (SPF) and germ-free (GF) mice were analyzed by flow cytometry as depicted in Figure E2. The frequency of CD49b1IgE1 basophils in (A) airways (controls, n.d. 5 not detected; OVA, n 5 13–16) and (B) lung (controls, n 5 5–8; OVA, n 5 14–17) are shown. Results are pooled data of three experiments and are representative of three independent experiments. Mean 6 SEM is shown. The frequency of CD41CD251Foxp31 T cells was determined for the (C ) airways and (D) lung (controls, n 5 4–6; OVA, n 5 9–10). Results are pooled data of two experiments and are representative of three independent experiments. The frequency of (E ) interstitial macrophages and (F ) alveolar macrophages in the lung (controls, n 5 5; OVA, n 5 5) were determined by flow cytometry. (G) The frequency of pDCs in lung (controls, n 5 5; OVA, n 5 5) is shown. (E–G) Results are data of one experiment and representative of two independent experiments. Mean 6 SEM is shown. *P , 0.05; **P , 0.01.

and CD11b2 cDCs (Figure 4D), exhibited a striking statistically significant reduction in total numbers. Similar to the CD11b1MHChi cDCs, these cells also had reduced surface expression of CD40, CD80, and CD86 (Figures 4C and 4D). Allergen challenge increased the numbers of CD11b1MHCIIhi (Figure 4B), CD11b1MHCIIint (Figure 4C), and CD11b2 (Figure 4D) cDCs in SPF mice, whereas only CD11b1MHCIIhi cells were significantly increased after allergen challenge of GF mice. Overall, it is tempting to speculate that these differences in cell types and activation states may shape the lung environment such that it is more prone to the development of Th2 immune responses. Although the exact role of these distinct cDC populations, or the implications of their activation status, during allergic asthma remains unclear it is clear that the commensal microbiota can impact on both the number and phenotype of

these cells. Of particular interest is the increased inducible costimulator ligand (ICOSL) expression on CD11b2 cDCs (Figure 4D). Although still controversial, ICOS-ICOSL interactions have been reported to play a role in promoting the expansion and effector function of differentiated Th2 cells within the lung (20). Dysregulated Numbers of Regulatory Immune Cell Types in the Lungs in the Absence of Commensal Bacteria

Numerous other cell types have been described to contribute to, or regulate, Th2-mediated allergic asthma. Basophils are potent producers of IL-4 and have been proposed to contribute to the allergic response (21). By contrast, regulatory T cells (22), plasmacytoid DCs (pDCs) (23), and lung macrophages (24) have all been attributed with a regulatory function during

204


allergic asthma. We found that the recruitment of basophils into the airways and lungs of OVA-immunized and airwaychallenged mice was greater in GF compared with SPF mice (Figures 5A and 5B). No differences were noted for airway or lung CD41CD251Foxp31 T cells (Figures 5C and 5D), or for lung interstitial macrophages (Figure 5E) in GF or SPF mice. However, reduced frequencies of lung alveolar macrophages (Figure 5F) and pDCs (Figure 5G) were evident in OVAsensitized and -challenged GF mice compared with SPF mice. Taken together, our data suggest that GF mice exhibit decreased populations of regulatory alveolar macrophages and pDC and increased frequencies of proallergic basophils. Overall, these findings correlate with, and are likely to contribute to, the exaggerated numbers of CD41 T cells producing Th2 cytokines and heightened IgE levels observed in GF mice after OVA airway challenge.

DISCUSSION Our findings of exaggerated OVA-induced airway inflammation in GF mice provide the first experimental evidence for a functional impact of commensal bacteria on allergic inflammation in the lung. Recolonization of GF mice with a complex SPF microbiota for 3 to 4 weeks before OVA sensitization was sufficient to protect against the increased allergic airway inflammation. Such recolonization presumably models the process of bacterial colonization that occurs in every infant after its birth. Our findings may therefore offer an explanation as to why environmental factors experienced during early childhood, when commensal bacteria are first encountered, exert a strong impact on the development of allergic diseases; in addition, our data provide promise for the ‘‘reconfiguring’’ of susceptibility to allergy by reconstitution or alteration of the commensal flora. The respiratory mucosa is constantly challenged by both pathogenic and nonpathogenic microorganisms in addition to environmental toxins and innocuous protein antigens. Thus, as for the intestine, immune homeostasis must be maintained such that exaggerated or inappropriate immune responses are avoided. It is perhaps not surprising, therefore, that the absence of commensal bacteria results in widespread changes to lung antigen-presenting cells concomitant with decreased frequencies of regulatory alveolar macrophages and pDC. Basophils were also found in increased frequencies in the BAL and lung of GF mice and it is likely that these cells contribute to the increased in situ IL-4 production observed. Although any one of these cellular changes alone could impact on allergic asthma, we believe that the dysregulated cDC activation status, elevated basophilia, and decreased presence of regulatory pDCs and alveolar macrophage cell populations observed in the lungs of GF mice likely act in concert and account for the increased allergic airway inflammation observed in these mice. Our data indicate that the commensal flora ‘‘educates’’ cells in the lung and redirects them from a Th2-prone activation state; such a process might be a fundamental step in the maturation of the cells responsible for controlling the balance between protective immunity and pathology in the lung. CD41CD251Foxp31 T-regulatory cells have also been shown to control AHR and allergic airway inflammation in mice (22). However, previous reports have yielded contrasting data regarding the impact of commensal bacteria on regulatory T cells. Min and coworkers (25) reported no impact on either the number or the function of Foxp31 T cells in GF mice, whereas Ostman and coworkers (26) demonstrated reduced effector function for in vitro suppressor assays and Ishikawa and coworkers (27) reported that GF mice exhibit defective oral tolerance as a result of defective Foxp31 T-cell function. In our

VOL 184

2011

experiments normal infiltration of Foxp31 T cells into the airways of allergic GF mice indicates that defective expansion or recruitment of these cells is not likely to be responsible for the enhanced allergic airway disease observed. However, more subtle differences in the functional capacity of regulatory T cells cannot be ruled out by our experiments. Dysbiosis of commensal bacteria has also been linked to other pathologic states, including type 1 diabetes (28), obesity (29), rheumatoid arthritis (30), and experimental autoimmune encephalomyelitis (31). These data indicate that changes in the timing of exposure or complexity of the commensal flora may impact on a multitude of chronic inflammatory diseases. Intriguingly, the prevalence of these autoimmune diseases has also increased selectively within the developed world over the past few decades (1), indicating a likely link between the environmental cues responsible for providing protection or susceptibility against allergy, inflammatory bowel disease, and autoimmune conditions. Immune modulation by commensal bacteria is likely to be complex and their impact on different inflammatory diseases may be mediated by distinct mechanisms. In this regard, a recent report by Maslowski and coworkers (32) has shown that GF mice exhibit exaggerated inflammation using models of experimental colitis and arthritis, and further showed that colitis could be attenuated in these mice after treatment with acetate, a short-chain fatty acid normally produced during fermentation of dietary fiber by intestinal microbiota. Lastly, although most research on commensals has focused on the intestine, it remains plausible that for lung responses the microbiota of the airways may be the key player. Indeed, Hilty and colleagues (10) recently reported that the microbial flora of the lung changes depending on whether or not individuals have asthma. Whether the different microbiota are causal of, or consequential to, the disease remains to be deciphered. Our observations that exaggerated allergic airway inflammation can be overcome by the recolonization of GF mice provides a novel model by which the impact of commensal bacteria on, and their possible therapeutic capacity for, allergic airway inflammation can be further examined. Many interesting questions remain to be addressed, particularly given the progressive characterization of the microbiota in the lung itself (10). Are respiratory immune responses impacted by commensal bacteria residing in the intestine, airways, skin, or all of the above? Do certain species of commensal bacteria direct the immune system down distinct paths? Providing answers to such questions is of great importance for the understanding of basic immunologic mechanisms and for the development of novel strategies aimed at preventing or treating allergic inflammation. Author Disclosure: T.H. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. A.S. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. C.S. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. K.Y. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. K.B. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. J.C. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. K.M. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. B.J.M. received a sponsored grant from the Swiss National Science Foundation. N.H. received sponsored grants from ETH Zurich, UBS, and the Swiss National Science Foundation. Author contributions: T.H., A.S., B.J.M., and N.H. designed research; T.H., A.S., C.S., K.Y., K.M., and J.C. performed research; K.B. and K.M. contributed new reagents/analytic tools; T.H., C.S., and A.S. analyzed data; and T.H., A.S., B.J.M., and N.H. wrote the paper. Acknowledgment: The authors thank Tim Sparwasser and Annette Oxenius for helpful discussions, Kerstin Wanke and Urs Karrer for access to laboratory space and practical assistance, and Leo Mamaril and Jorum Kirundi for the care of germ-free animals.

Herbst, Sichelstiel, Scha¨r, et al.: Commensal Bacteria Prevent Exaggerated Allergy References 1. Wills-Karp M, Santeliz J, Karp CL. The germless theory of allergic disease: revisiting the hygiene hypothesis. Nat Rev Immunol 2001;1: 69–75. 2. Strachan DP. Hay fever, hygiene, and household size. BMJ 1989;299: 1259–1260. 3. Bjorksten B, Naaber P, Sepp E, Mikelsaar M. The intestinal microflora in allergic Estonian and Swedish 2-year-old children. Clin Exp Allergy 1999;29:342–346. 4. Bjorksten B, Sepp E, Julge K, Voor T, Mikelsaar M. Allergy development and the intestinal microflora during the first year of life. J Allergy Clin Immunol 2001;108:516–520. 5. Bottcher MF, Nordin EK, Sandin A, Midtvedt T, Bjorksten B. Microfloraassociated characteristics in faeces from allergic and nonallergic infants. Clin Exp Allergy 2000;30:1590–1596. 6. Macpherson AJ, Harris NL. Interactions between commensal intestinal bacteria and the immune system. Nat Rev Immunol 2004;4:478–485. 7. Biasucci G, Rubini M, Riboni S, Retetangos C, Morelli L, Bessi E. Mode of delivery affects the bacterial community in the newborn gut. Early Hum Dev 2010;86:13–15. 8. Harmsen HJ, Wildeboer-Veloo AC, Raangs GC, Wagendorp AA, Klijn N, Bindels JG, Welling GW. Analysis of intestinal flora development in breast-fed and formula-fed infants by using molecular identification and detection methods. J Pediatr Gastroenterol Nutr 2000;30: 61–67. 9. Fanaro S, Chierici R, Guerrini P, Vigi V. Intestinal microflora in early infancy: composition and development. Acta Paediatr Suppl 2003;91: 48–55. 10. Hilty M, Burke C, Pedro H, Cardenas P, Bush A, Bossley C, Davies J, Ervine A, Poulter L, Pachter L, et al. Disordered microbial communities in asthmatic airways. PLoS ONE 2010;5:e8578. 11. Kraehenbuhl JP, Neutra MR. Molecular and cellular basis of immune protection of mucosal surfaces. Physiol Rev 1992;72:853–879. 12. Massacand JC, Kaiser P, Ernst B, Tardivel A, Burki K, Schneider P, Harris NL. Intestinal bacteria condition dendritic cells to promote IgA production. PLoS ONE 2008;3:e2588. 13. Hazebrouck S, Przybylski-Nicaise L, Ah-Leung S, Adel-Patient K, Corthier G, Wal JM, Rabot S. Allergic sensitization to bovine betalactoglobulin: comparison between germ-free and conventional Balb/ c mice. Int Arch Allergy Immunol 2009;148:65–72. 14. Sudo N, Aiba Y, Oyama N, Yu XN, Matsunaga M, Koga Y, Kubo C. Dietary nucleic acid and intestinal microbiota synergistically promote a shift in the Th1/Th2 balance toward Th1-skewed immunity. Int Arch Allergy Immunol 2004;135:132–135. 15. Sudo N, Sawamura S, Tanaka K, Aiba Y, Kubo C, Koga Y. The requirement of intestinal bacterial flora for the development of an IgE production system fully susceptible to oral tolerance induction. J Immunol 1997;159:1739–1745. 16. McCoy KD, Harris NL, Diener P, Hatak S, Odermatt B, Hangartner L, Senn BM, Marsland BJ, Geuking MB, Hengartner H, et al. Natural IgE production in the absence of MHC class II cognate help. Immunity 2006;24:329–339.

205

17. Smits HH, Gloudemans AK, van Nimwegen M, Willart MA, Soullie T, Muskens F, de Jong EC, Boon L, Pilette C, Johansen FE, et al. Cholera toxin B suppresses allergic inflammation through induction of secretory IgA. Mucosal Immunol 2009;2:331–339. 18. Chen J, Trounstine M, Alt FW, Young F, Kurahara C, Loring JF, Huszar D. Immunoglobulin gene rearrangement in B cell deficient mice generated by targeted deletion of the JH locus. Int Immunol 1993;5: 647–656. 19. Plantinga M, Hammad H, Lambrecht BN. Origin and functional specializations of DC subsets in the lung. Eur J Immunol 2010;40:2112–2118. 20. Simpson TR, Quezada SA, Allison JP. Regulation of CD4 T cell activation and effector function by inducible costimulator (ICOS). Curr Opin Immunol 2010;22:326–332. 21. Hammad H, Plantinga M, Deswarte K, Pouliot P, Willart MA, Kool M, Muskens F, Lambrecht BN. Inflammatory dendritic cells–not basophils– are necessary and sufficient for induction of Th2 immunity to inhaled house dust mite allergen. J Exp Med 2010;207:2097–2111. 22. Ray A, Khare A, Krishnamoorthy N, Qi Z, Ray P. Regulatory T cells in many flavors control asthma. Mucosal Immunol 2010;3:216–229. 23. Kool M, van Nimwegen M, Willart MA, Muskens F, Boon L, Smit JJ, Coyle A, Clausen BE, Hoogsteden HC, Lambrecht BN, et al. An antiinflammatory role for plasmacytoid dendritic cells in allergic airway inflammation. J Immunol 2009;183:1074–1082. 24. Bedoret D, Wallemacq H, Marichal T, Desmet C, Quesada Calvo F, Henry E, Closset R, Dewals B, Thielen C, Gustin P, et al. Lung interstitial macrophages alter dendritic cell functions to prevent airway allergy in mice. J Clin Invest 2009;119:3723–3738. 25. Min B, Thornton A, Caucheteux SM, Younes SA, Oh K, Hu-Li J, Paul WE. Gut flora antigens are not important in the maintenance of regulatory T cell heterogeneity and homeostasis. Eur J Immunol 2007;37:1916–1923. 26. Ostman S, Rask C, Wold AE, Hultkrantz S, Telemo E. Impaired regulatory T cell function in germ-free mice. Eur J Immunol 2006; 36:2336–2346. 27. Ishikawa H, Tanaka K, Maeda Y, Aiba Y, Hata A, Tsuji NM, Koga Y, Matsumoto T. Effect of intestinal microbiota on the induction of regulatory CD251 CD41 T cells. Clin Exp Immunol 2008;153:127–135. 28. Wen L, Ley RE, Volchkov PY, Stranges PB, Avanesyan L, Stonebraker AC, Hu C, Wong FS, Szot GL, Bluestone JA, et al. Innate immunity and intestinal microbiota in the development of type 1 diabetes. Nature 2008;455:1109–1113. 29. Ley RE, Turnbaugh PJ, Klein S, Gordon JI. Microbial ecology: human gut microbes associated with obesity. Nature 2006;444:1022–1023. 30. Vaahtovuo J, Munukka E, Korkeamaki M, Luukkainen R, Toivanen P. Fecal microbiota in early rheumatoid arthritis. J Rheumatol 2008;35:1500–1505. 31. Ochoa-Reparaz J, Mielcarz DW, Ditrio LE, Burroughs AR, Foureau DM, Haque-Begum S, Kasper LH. Role of gut commensal microflora in the development of experimental autoimmune encephalomyelitis. J Immunol 2009;183:6041–6050. 32. Maslowski KM, Vieira AT, Ng A, Kranich J, Sierro F, Yu D, Schilter HC, Rolph MS, Mackay F, Artis D, et al. Regulation of inflammatory responses by gut microbiota and chemoattractant receptor GPR43. Nature 2009;461:1282–1286.