Mucosal B Cell Deficiency in IgA–/– Mice Abrogates the Development of Allergic Lung Inflammation1
http://www.100md.com
免疫学杂志 2005年第14期
Abstract
We have investigated the consequence of lack of IgA on host immunity using a murine model of allergic lung inflammation. Mice with a targeted disruption of the -switch region and 5' H chain gene (IgA–/– mice), which lack total IgA, developed significantly reduced pulmonary inflammation with fewer inflammatory cells in lung tissue and bronchoalveolar lavage fluids, as well as reduced levels of total and IgG1 OVA-specific Abs and decreased IL-4 and IL-5 in bronchoalveolar lavage fluids compared with IgA+/+ controls, following allergen sensitization and challenge. This defect was attributable to fewer B cells in the lungs of IgA–/– mice. Polymeric IgR-deficient (pIgR–/–) mice, which lack the receptor that transports polymeric IgA across the mucosal epithelium where it is cleaved to form secretory IgA, were used to assess the contribution of secretory IgA vs total IgA in the induction of allergic lung inflammation. pIgR–/– and pIgR+/+ mice had comparable levels of inflammation, demonstrating that IgA bound to secretory component is not necessary for the development of allergic lung inflammation, although this does not necessarily rule out a role for transudated IgA in lung secretions because of "mucosal leakiness" in these mice. The results indicate that Ag-specific B cells are required at mucosal surfaces for induction of inflammation and likely function as major APCs in the lung for soluble protein Ags.
Introduction
Humans synthesize 66 mg/kg IgA/day, which is more than all other Ig isotypes combined (1). Of this, the majority (50–70%) is produced by submucosal B cells and transported into mucosal secretions by the polymeric IgR (pIgR)3 (1, 2). The pIgR is expressed on the basolateral surface of most mucosal epithelial cells, where it binds to polymeric IgA- and IgM-containing J chains (2). Following transcytosis of the Ig/receptor complex to the apical surface of the cell, the receptor is cleaved, releasing the Ig bound to the extracellular portion of the receptor, referred to as secretory component (SC) (2). SC-bound IgA, or secretory IgA (SIgA), is considered an important mediator of immunity at mucosal surfaces. As the majority of functions attributed to SIgA are noninflammatory, it is widely referred to as a neutralizing Ab (3, 4, 5). It has been shown to impede bacterial colonization of the mucosa and neutralize virus particles (3, 4, 5, 6), block the activity of bacterial enzymes and toxins (3, 4, 5, 7), and preserve tolerance to food Ags (3, 5, 8, 9). However, within tissues, IgA is capable of mediating the effector functions of several leukocyte populations through its interactions with CD89 (FcRI), including induction of cytokine production, phagocytosis, and Ab-dependent cellular cytotoxicity by neutrophils, monocytes, and macrophages (3, 5, 10) and degranulation of eosinophils (11). Thus, there is a certain degree of compartmentalization to the function of IgA in immunity (3, 5).
The development of IgA-deficient mice (IgA–/–) (12) has provided a model for the further characterization of the biological activities of IgA. Initial studies using these mice did not demonstrate an increased susceptibility of IgA–/– mice to challenge with infectious agents compared with IgA+/+ mice (7, 12, 13, 14, 15, 16, 17); however, more recent investigations have revealed subtle immune defects that could be linked to deficient APC function (18, 19). In the current study, we investigated the effect of IgA deficiency on the inflammatory response associated with allergic asthma using a murine model of allergic lung inflammation. The ability of SIgA to neutralize allergens and prevent sensitization is contentious; some reports (9, 20, 21) have demonstrated a negative correlation between IgA levels and allergic sensitization, whereas others have found no such association (22, 23, 24). Despite this, a possible link between IgA and the pathophysiology of established allergic asthma has been alluded to in a number of correlative human studies. Allergen-specific IgA can be isolated from the bronchoalveolar lavage (BAL) of asthmatics (25, 26) and is increased during periods of high-allergen exposure (27). IgA is capable of inducing degranulation of human eosinophils, the major cell type present in the inflammatory infiltrate of asthmatics (11). CD89 (FcR1) is up-regulated on eosinophils isolated from allergic patients, and IgA levels correlate with levels of eosinophil granule proteins in BAL fluids from asthmatics (28, 29, 30). Furthermore, Ag-specific IgA in BAL fluids has been correlated with stronger Ag-induced, late-phase asthmatic reactions (31). Although the ability of SIgA to neutralize Ags without the induction of inflammation is well established, the above results, although offering no specific mechanism of activity, suggest that IgA may enhance an ongoing mucosal inflammatory response.
We have now examined the ability of IgA–/– and pIgR–/– mice to respond to intranasal (IN) challenge in a murine model of allergic lung inflammation. Using this model, a defect in B cell expression was identified in the lungs of IgA–/– mice. Furthermore, it was found that although SIgA does not appear to contribute to the development of allergic lung inflammation following systemic (i.p.) sensitization and mucosal (IN) challenge, Ag-specific B cells appear to be an essential population for the development of respiratory inflammation induced by a soluble protein Ags (OVA), likely by functioning as APCs.
Results
Recruitment of inflammatory cells into the lungs of sensitized mice following IN Ag challenge is decreased in IgA–/– mice
The most striking feature of murine allergic lung inflammation is the influx of inflammatory cells into the lung subsequent to allergen priming and challenge. After OVA sensitization and IN challenge, IgA+/+ mice developed profound inflammation (inflammatory score = 2.60 ± 0.53, n = 24) that manifested as PV and peribronchiolar cuffing (Fig. 1a, arrows), although areas of diffuse infiltrate were observed as well. Hypertrophy of the bronchial epithelium and plugging of some airways with mucus, inflammatory cells, and debris was also evident in the lungs of these mice (Fig. 1, a and b). By comparison, identically treated IgA–/– mice had a significantly diminished inflammatory response (inflammatory score = 1.50 ± 0.38, n = 24) (p < 0.01), in which fewer infiltrates were observed, primarily as PV cuffing with very little diffuse inflammation (Fig. 1, c and d). Furthermore, hypertrophy of the bronchial epithelium was mostly absent and plugging of airways was not seen (Fig. 1d). Eosinophils were detected in infiltrates from both groups of mice using a CN–EPO stain (Fig. 1, e and f); however, significantly (p < 0.01) greater numbers were detected within the infiltrates of IgA+/+ mice (307.05 ± 12.23 eosinophils/field) compared with IgA–/– mice (122.85 ± 32.64 eosinophils/field). Saline-sensitized and -challenged, saline-sensitized and OVA-challenged, and OVA-sensitized and saline-challenged mice were used as controls and demonstrated negligible inflammation (data not shown).
Similar to the responses observed in lung tissues, significantly more inflammatory cells were present in BAL fluids from IgA+/+ mice compared with IgA–/– mice (8.11 x 105± 3.10 x 105 total cells vs 2.91 x 105± 1.12 x 105 total cells; p < 0.005). Differential staining of BAL fluid inflammatory cell isolates revealed that eosinophils were the predominant cell population in both groups of mice, but again, there were significantly (p < 0.05) more eosinophils isolated from IgA+/+ mice than IgA–/– mice (Fig. 1g). Overall, eosinophils constituted 10- 20% less of the total inflammatory cell population isolated from BAL fluids of IgA–/– mice compared with IgA+/+ mice (data not shown). Lymphocyte, macrophage, and neutrophil numbers were comparable between both groups of mice (Fig. 1g).
To determine whether the diminished inflammatory response in IgA–/– mice was Ag specific or due to an inherent defect caused by disruption of the -H chain, inflammation was chemically induced by IN challenge with bleomycin sulfate. Intranasal administration of bleomycin resulted in recruitment of neutrophils and macrophages to lung tissues of both groups of mice (Fig. 2). The development of inflammation was also reflected in the large numbers of inflammatory cells isolated from BAL fluids of IgA–/– and IgA+/+ mice (5.61 x 106 ± 0.01 x 106 vs 1.31 x 106 ± 0.53 x 106 total cells, respectively) (p < 0.05). These results indicate that IgA–/– mice are fully capable of manifesting a pulmonary inflammatory response and suggest that the defect observed in IgA–/– mice following OVA sensitization and challenge is a deficiency in responding to antigenic stimulation. Furthermore, the presence of significantly more inflammatory cells in BAL fluids from IgA–/– mice compared with IgA+/+ mice suggests that IgA plays different roles in inflammatory responses induced by different stimuli, e.g., Ag-specific vs Ag-nonspecific inflammatory responses.
Unlike other isotypes tested, levels of total serum IgE were significantly lower in IgA–/– mice compared with IgA+/+ mice (Fig. 3b). To determine whether decreased production of serum IgE was responsible for the diminished inflammatory response observed in these mice, both groups were sensitized in the presence of IL-4, the switch factor for IgE. Inclusion of IL-4 at the time of sensitization resulted in increased production of total IgE in the sera of IgA–/– mice to a level approximately half of that seen in IgA+/+ mice, without a concurrent increase in flammation (data not shown). However, IgE levels were still significantly lower than those in IgA+/+ mice and could not be further increased with higher concentrations of IL-4. Thus, although these results suggest that IgE does not influence the development of inflammation in our model, we cannot completely exclude a role for this Ig isotype in enhancing inflammation in IgA+/+ mice. The data also indicate that disruption of the -H chain has affected the ability of IgA–/– mice to produce Ig of this isotype. This confirms previous results of Harriman et al. (12), who found that naive IgA–/– mice have significantly lower levels of serum IgE compared with wild-type littermates.
The reduced inflammatory response in IgA–/– mice is due to lack of total IgA not SIgA
To establish whether the abrogated inflammatory response observed in IgA–/– mice is the result of a lack of total IgA or SIgA, we used pIgR–/– mice, which are unable to transcytose IgA across the mucosal epithelium. As a result, these mice lack SIgA but have 100-fold greater levels of serum IgA (32). OVA sensitization and challenge of pIgR+/+ and pIgR–/– mice resulted in a comparable influx of inflammatory cells into the lungs (Fig. 4, a and b, respectively) (pIgR+/+ inflammatory score = 2.33 ± 0.72, n = 12; pIgR–/– inflammatory score = 2.13 ± 0.49, n = 12) and BAL fluids (pIgR+/+ = 3.52 x 105 ± 1.97 x 105 total cells/lung and pIgR–/– = 3.93 x 105 ± 1.56x 105 total cells/lung). In addition, total and IgG1 OVA-specific Ab levels were similar in BAL fluids (Fig. 4c) and sera (Fig. 4d) from both groups of mice, as were serum levels of total IgE (Fig. 4d). The only differences observed between the mice were in levels of OVA-specific BAL and serum IgA, which were significantly higher in pIgR–/– mice compared with pIgR+/+ mice (Fig. 4, c and d). pIgR–/– mice do have some IgA in mucosal surfaces because disruption of the receptor leads to increased paracellular leakage of serum proteins into the mucosal lumen (32). Thus, although these results indicate that SC does not contribute to the development of allergic lung inflammation, we cannot completely exclude a role for IgA Abs in the respiratory tract.
Ag-specific B cells are required for enhanced inflammation in IgA+/+ mice
To determine whether B cells are required for the enhanced inflammatory response observed in IgA+/+ mice, cell transfer experiments were performed using sensitized splenocytes or CD4+ T cells as described previously (35). Splenocytes were isolated from 14-day sensitized IgA+/+ and IgA–/– mice and cultured with OVA for 72 h. Whole splenocyte populations or CD4+ T cells isolated from cultured splenocytes were then injected i.p. into naive mice. The recipients were challenged IN with OVA, and the lungs were examined by histology for inflammation. Transfer of IgA+/+-sensitized splenocytes into naive IgA+/+ mice resulted in profound inflammation (inflammatory score = 3.25 ± 1.07, n = 8), similar to that seen in directly sensitized and challenged IgA+/+ mice (Fig. 7a). As expected, this was significantly greater (p < 0.05) than the moderate inflammatory response observed in the IgA–/–IgA–/– cell transfer group (inflammatory score = 1.83 ± 1.00, n = 9) (Fig. 7a). In contrast, transfer of isolated CD4 T cells from sensitized IgA+/+ cells into IgA+/+ recipients resulted in a decreased inflammatory response (inflammatory score = 1.58 ± 0.38, n = 6) comparable to that observed in the IgA–/–IgA–/– group (inflammatory score = 1.58 ± 0.38, n = 6) (Fig. 7b). This indicates that isolation of CD4+ T cells from whole splenocytes resulted in loss of some cell type responsible for the enhanced inflammatory response observed in IgA+/+ mice. To determine the importance of Ag-specific B cells in this response, whole splenocyte populations depleted of CD19+ cells were transferred into naive mice. Depletion of CD19+ cells from the splenocyte population before transfer resulted in a significantly attenuated inflammatory response in the IgA+/+IgA+/+ group (inflammatory score = 1.83 ± 0.76, n = 3), similar to that observed in the IgA–/–IgA–/– group (inflammatory score = 1.67 ± 0.58, n = 3) (Fig. 7c). Differential staining of BAL fluid cells revealed that eosinophil numbers mirrored the inflammatory responses observed by tissue staining (data not shown). Transfer of naive cells followed by IN challenge did not result in the recruitment of inflammatory cells to the lungs of any group of mice. Thus, Ag-specific B cells are required for the enhanced inflammatory response observed in allergen-sensitized and -challenged IgA+/+ mice.
Discussion
Several studies have suggested that IgA plays a role in the pathology of allergic inflammation in asthmatic patients (25, 26, 27, 28, 29, 30, 31). In the present study, using a murine model of allergic lung inflammation, we demonstrate that IgA–/– mice have reduced pulmonary inflammation following sensitization and IN challenge with soluble OVA. However, this defect appears to be linked to the absence of Ag-specific B cells within the pulmonary compartment of IgA–/– mice, rather than a lack of SIgA, as pIgR–/– mice developed inflammation similar to pIgR+/+ mice, although using this model we could not completely exclude a role for IgA Abs that may have transudated from serum into the respiratory tract. An earlier study by Arulanandam et al. (18) showed that the absence of IgA in IgA–/– mice results in defective APC function. Furthermore, Mbawuike et al. (19) demonstrated altered Th1 cell function in IgA–/– mice that might also be attributable to a defect in APC function. The present findings suggest that this defect exists because of a lack of B cells in the lung and that B cells at this site play an important role as APCs.
Sensitization and challenge of IgA–/– mice with OVA resulted in an abrogated inflammatory response compared with IgA+/+ mice, with markedly reduced recruitment of cells to both the tissue and airways, and decreased levels of allergen-specific Ab and Th2 cytokines in BAL fluids. Numbers of eosinophils, the dominant cell type present in allergic lung infiltrates, were significantly diminished in the lungs of IgA–/– mice, likely the result of decreased IL-5 production because IL-5 is a hemopoietic and chemotactic factor for these cells (37). In addition, OVA-specific Ab levels were found to be significantly lower in BAL fluids from IgA–/– mice while serum Ab levels were similar. Ongoing work in our laboratory has demonstrated that Abs found in BAL fluids after IN vaccination are produced by resident pulmonary B cells (D. Albu et al., manuscript in preparation). Thus, decreased levels of allergen-specific Ab in BAL fluids of IgA–/– mice likely reflect a lack of Ag-specific B cells at the site of inflammation, which is consistent with our data demonstrating defective expression of IgG+ B cells in the lungs of these mice.
Total and IgG1 allergen-specific serum Ab levels were similar between both groups of mice, indicating normal systemic B cell responses. However, decreased levels of IgE were observed in IgA–/– mice compared with IgA+/+ mice. Total Ig, rather than IgE Ab, is typically used as a marker for increased isotype production in allergy models because the low level of IgE production coupled with uptake by the high-affinity FcRI usually renders free IgE Ab undetectable via conventional ELISA. Administration of IL-4, the switch factor for IgE, during sensitization resulted in an increase in total serum IgE production to a level approximately half of that observed in IgA+/+ mice, without a concurrent increase in inflammation, suggesting that IgE is not critical for the development of allergic inflammation in our model. These results corroborate other studies using IgE and mast cell deficient mice, which imply that this Ab isotype and its interaction with mast cells does not contribute to the development of allergic lung inflammation in mice (38, 39). However, as we were unable to induce similar levels of IgE production in IgA+/+ and IgA–/– mice, we cannot fully exclude a role for this isotype in the development of allergic lung inflammation. These data also indicate that disruption of the -H chain affects class switching to other isotypes as well. This is in accordance with Harriman et al. (12), who demonstrated significantly lower levels of serum IgE in naive IgA–/– mice compared with IgA+/+ mice. Thus, disruption of IgA leads to a generalized defect in class switching; curiously, this is a phenomenon that is also observed in many IgA-deficient patients and a defect that one would not expect in a system with such a dissimilar etiology from human IgA deficiency (8, 9, 40).
Th2 cytokine levels were significantly lower in IgA–/– mice compared with IgA+/+ mice at early time points during IN challenge with OVA but were similar at later times. This correlates with an earlier influx of inflammatory cells into the lungs of IgA+/+ mice compared with IgA–/– mice, whereas the diminished level of cytokine production during the IN challenge of IgA–/– mice is indicative of the weaker inflammatory response observed in these mice. It is unclear why a discrepancy exists between mRNA transcript and protein levels. There may be a difference in posttranscriptional regulation of cytokine production and/or secretion between IgA+/+ and IgA–/– mice. Alternatively, the isolation of total lung mRNA for transcript analysis, rather than specific cell types, may have muted subtle differences in expression making the ELISA a more sensitive method for detecting differences in cytokine levels between the two groups of mice. In either case, there is a distinct difference in levels of cytokine produced during IN challenge of IgA+/+ and IgA–/– mice.
Bleomycin challenge was used to confirm that the reduced inflammatory response observed in allergen-exposed IgA–/– mice was a deficiency in mounting an Ag-specific immune response. Bleomycin induces inflammation in the lungs and skin due to the lack of the enzyme bleomycin hydrolase in these tissues (41). Although the inflammatory response observed after bleomycin challenge is primarily mediated by neutrophils and macrophages, unlike allergic lung inflammation which involves significant eosinophil infiltration, the resultant response demonstrated that IgA–/– mice are capable of mounting pulmonary inflammation. However, the recruitment of significantly more inflammatory cells to the lungs of IgA–/– mice relative to IgA+/+ mice indicates that IgA and/or mucosal B cells play very different roles in mediating inflammation induced by different stimuli. IgA is widely considered to be an anti-inflammatory Ab at mucosal sites, where it is involved in the neutralization and clearance of bacterial, viral, and other Ags (5). However, as administration of bleomycin sulfate does not induce Ag-specific Abs, any anti-inflammatory effect of IgA in this system would not occur through neutralization but through an, as of yet, unknown mechanism.
Analysis of pulmonary leukocyte populations in IgA+/+ and IgA–/– mice revealed significantly fewer B lymphocytes in the lungs of IgA–/– mice in the presence or absence of allergic lung inflammation. In addition, immunofluorescent tissue staining demonstrated a complete absence of IgG staining cells in lung sections from IgA–/– mice following the induction of allergic lung inflammation, whereas many IgG+ cells were present in the lungs of IgA+/+ mice. This finding correlates with the negligible amounts of IgG1 found in the airways, although it is unclear why disruption of the -H chain in IgA–/– mice results in a lack of IgG+ B cell expression in the lungs. In a recent study, Uren et al. (42) also described disrupted mucosal B cell homeostasis in pIgR–/– mice, which contain 100-fold greater levels of serum IgA than pIgR+/+ mice. However, the altered B cell homeostasis manifested in this case as an increased number of IgA+ plasmablasts in the lamina propria and spleen (42); the mechanism for this defect was not determined. As IgA–/– mice have decreased numbers of B cells at mucosal sites, it is likely that IgA production influences B cell homeostasis; the mechanism(s) responsible for this defect are currently being investigated.
Cell transfer studies were performed in an effort to determine whether the enhanced inflammatory response observed in IgA+/+ mice compared with IgA–/– mice is dependent upon the presence of Ag-specific B cells. Transfer of whole splenocyte populations from IgA+/+IgA+/+ mice resulted in a profound inflammatory response that was visibly more severe than the response seen in IgA–/–IgA–/– mice. However, removal of Ag-specific B cells from the transferred population, either by isolating CD4+ T cells or by depleting CD19+ B cells, resulted in a loss of the response observed in IgA+/+ mice. Transfer of purified CD19+ B cells was not performed as splenocyte populations depleted of CD4+ cells have been shown previously to be incapable of transferring disease to naive animals (35). Unfortunately, we could not perform IgA+/+IgA–/– and IgA–/–IgA+/+ cell transfers because the IgA–/– mice are not on a pure inbred background. It is not clear what role B cells play in the response, i.e., if they enhance the response through local Ab production and/or whether they function as APCs. B cells are potent Ag-specific APCs, especially for soluble protein Ags and/or in cases of limited Ag concentrations (43, 44, 45, 46, 47). They can rapidly take up, catabolize, and present Ags and have been shown to play an important role in optimizing Ag presentation to T cells (44, 47). Furthermore, lack of B cells during presentation has been shown to result in reduced T cell responses (44, 45, 47, 48, 49). Dendritic cells (DCs) are likely responsible for the limited inflammatory response observed in IgA–/– mice; however, the lack of adjuvant during IN challenge in this model would cause suboptimal DC activation and maturation. As it has been reported that alveolar macrophages are poor APCs (50), B cells may be a critical APC population within the respiratory tract. This would explain previous data demonstrating that IN immunization of IgA–/– mice with vaccine + adjuvant resulted in protective immunity, whereas vaccination of these mice in the absence of a DC-stimulating adjuvant resulted in a loss of protection (18). We have found that equivalent numbers of IgA–/– and IgA+/+ CD45+ APC populations can stimulate similar levels of cell proliferation and cytokine production cells by CD4+ T cells purified from OT-II mice, which are specific for OVA (our unpublished observations). Thus, there is no inherent defect in the ability IgA–/– B cells to present Ags; rather, our data indicate that reduced B cell numbers in the lungs of IgA–/– mice leads to defective APC activity. Under normal conditions, such as in IgA+/+ mice, Ag-specific B cells likely traffic through mucosal effecter sites, including the lungs, where they encounter their cognate Ags and, with DCs, initiate activation of memory T cells.
It has been reported previously, using B cell-deficient (μMT) mice, that B cells do not contribute to the induction of murine allergic lung inflammation. However, it should be recognized that μMT mice are not fully deficient in B cells. Several studies have demonstrated the presence of IgA-, IgG-, and IgE-producing B cells in the absence of membrane IgM and IgD expression (51, 52, 53). Furthermore, Macpherson et al. (51) demonstrated that a significant number of IgA+ cells are present in an immune effector site, i.e., the lamina propria of μMT mice. In addition, it has been suggested that B cells are critical for normal secondary organ lymphorganogenesis (45, 54, 55, 56). Lack of B cells during development may result in the formation of compensatory mechanisms that might not occur when only a subpopulation of B cells is missing. Evidence for this comes from experiments by Rivera et al. (45), demonstrating that the ability of μMT APCs to induce functional immunity is profoundly reduced when these cells were placed in a normally structured immune system by reconstituting irradiated C57BL/6 mice with μMT bone marrow. This suggests that the lack of B cells during development in μMT mice resulted in formation of a compensatory pathway that allows adequate APC function and that such a mechanism was not functional in a wild-type environment (45). In addition, while some studies using B cell-deficient mice found that B cells do not contribute to the development of normal immune responses, others have demonstrated defects in CD4+ T cell function generated in the absence of B cells (43, 44, 45, 46, 48, 49, 57). It is possible that differences in immunization protocols, the types of Ags and adjuvants used (43), as well as the background strain of mice (45), are partially responsible for the discrepancies in experimental findings between different groups.
We believe that the lack of mucosal B cells and deficient allergic lung inflammation observed in IgA–/– mice correlates with the previously reported defect in APC function in these mice (18). In that study, Arulanandam et al. (18) noted that a defect in APC function was observed when attempting to elicit protection from lethal influenza challenge via IN immunization with a protein subunit vaccine. However, the defect was not observed when mice were immunized with the vaccine plus a strong adjuvant (IL-12). Similarly, IgA–/– mice do not have an increased susceptibility to infection with influenza virus (13, 19, 58), HSV-2 (16), Shigella flexneri (17), or Helicobacter pylori (15), following immunization with a strong adjuvant or attenuated bacteria/viruses. However, under such conditions, DCs and macrophages would be adequately stimulated to present Ags to T cells, diminishing the need for B cells to function as APCs. In the absence of DC-activating adjuvants, such as in the allergic lung inflammation model used here, B cells may play an essential role in enhancing immune responses. Our findings imply that in the absence of Ag-specific B cells at mucosal sites, there is a need for DC-activating adjuvants for efficacious immunization. It has been reported that some IgA-deficient humans also have defects in mucosal B cell numbers (59, 60); it will be of interest to determine whether such a defect is responsible for the increased incidence of infectious sinopulmonary disease observed in these individuals (8, 9).
Acknowledgments
We thank Dr. Maria Lopez for her guidance in operating the flow cytometer and evaluating flow cytometry data and the Immunology Core of the Center for Immunology and Microbial Disease for histology services.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
1 This work was supported by National Institutes of Health Grant AI41715 and by Philip Morris USA, Inc., and by Philip Morris International. P.M.A. was supported by National Institutes of Health Training Grant AI49822.
2 Address correspondence and reprint requests to Dr. Dennis Metzger, Center for Immunology and Microbial Disease, Albany Medical College, Mail Code 151, 47 New Scotland Avenue, Albany, NY 12208. E-mail address: metzged@mail.amc.edu
3 Abbreviations used in this paper: pIgR, polymeric IgR; SC, secretory component; SIgA, secretory IgA; BAL, bronchoalveolar lavage; IN, intranasal; PV, perivascular; CN–EPO, cyanide-resistant eosinophil peroxidase; DC, dendritic cell.
Received for publication March 22, 2005. Accepted for publication May 12, 2005.
References
Mestecky, J., I. Moro, M. Kerr, J. M. Woof. 2005. Mucosal immunoglobulins. J. Mestecky, and M. E. Lamm, and W. Strober, and J. Bienenstock, and J. R. McGhee, and L. Mayer, eds. In Mucosal Immunology Vol. 1: 153-181 Academic Press, San Diego. .
Johansen, F. E., R. Braathen, P. Brandtzaeg. 2000. Role of J chain in secretory immunoglobulin formation. Scand. J. Immunol. 52: 240-248.
van Egmond, M., C. A. Damen, A. B. van Spriel, G. Vidarsson, E. van Garderen, J. G. van de Winkel. 2001. IgA and the IgA Fc receptor. Trends Immunol. 22: 205-211.
Phalipon, A., B. Corthesy. 2003. Novel functions of the polymeric Ig receptor: well beyond transport of immunoglobulins. Trends Immunol. 24: 55-58.
Russell, M. W., M. Kilian. 2005. Biological activities of IgA. J. Mestecky, and M. E. Lamm, and W. Strober, and J. Bienenstock, and J. R. McGhee, and L. Mayer, eds. Mucosal Immunology 267-289 Academic Press, San Diego. .(Paul M. Arnaboldi*, Melis)
We have investigated the consequence of lack of IgA on host immunity using a murine model of allergic lung inflammation. Mice with a targeted disruption of the -switch region and 5' H chain gene (IgA–/– mice), which lack total IgA, developed significantly reduced pulmonary inflammation with fewer inflammatory cells in lung tissue and bronchoalveolar lavage fluids, as well as reduced levels of total and IgG1 OVA-specific Abs and decreased IL-4 and IL-5 in bronchoalveolar lavage fluids compared with IgA+/+ controls, following allergen sensitization and challenge. This defect was attributable to fewer B cells in the lungs of IgA–/– mice. Polymeric IgR-deficient (pIgR–/–) mice, which lack the receptor that transports polymeric IgA across the mucosal epithelium where it is cleaved to form secretory IgA, were used to assess the contribution of secretory IgA vs total IgA in the induction of allergic lung inflammation. pIgR–/– and pIgR+/+ mice had comparable levels of inflammation, demonstrating that IgA bound to secretory component is not necessary for the development of allergic lung inflammation, although this does not necessarily rule out a role for transudated IgA in lung secretions because of "mucosal leakiness" in these mice. The results indicate that Ag-specific B cells are required at mucosal surfaces for induction of inflammation and likely function as major APCs in the lung for soluble protein Ags.
Introduction
Humans synthesize 66 mg/kg IgA/day, which is more than all other Ig isotypes combined (1). Of this, the majority (50–70%) is produced by submucosal B cells and transported into mucosal secretions by the polymeric IgR (pIgR)3 (1, 2). The pIgR is expressed on the basolateral surface of most mucosal epithelial cells, where it binds to polymeric IgA- and IgM-containing J chains (2). Following transcytosis of the Ig/receptor complex to the apical surface of the cell, the receptor is cleaved, releasing the Ig bound to the extracellular portion of the receptor, referred to as secretory component (SC) (2). SC-bound IgA, or secretory IgA (SIgA), is considered an important mediator of immunity at mucosal surfaces. As the majority of functions attributed to SIgA are noninflammatory, it is widely referred to as a neutralizing Ab (3, 4, 5). It has been shown to impede bacterial colonization of the mucosa and neutralize virus particles (3, 4, 5, 6), block the activity of bacterial enzymes and toxins (3, 4, 5, 7), and preserve tolerance to food Ags (3, 5, 8, 9). However, within tissues, IgA is capable of mediating the effector functions of several leukocyte populations through its interactions with CD89 (FcRI), including induction of cytokine production, phagocytosis, and Ab-dependent cellular cytotoxicity by neutrophils, monocytes, and macrophages (3, 5, 10) and degranulation of eosinophils (11). Thus, there is a certain degree of compartmentalization to the function of IgA in immunity (3, 5).
The development of IgA-deficient mice (IgA–/–) (12) has provided a model for the further characterization of the biological activities of IgA. Initial studies using these mice did not demonstrate an increased susceptibility of IgA–/– mice to challenge with infectious agents compared with IgA+/+ mice (7, 12, 13, 14, 15, 16, 17); however, more recent investigations have revealed subtle immune defects that could be linked to deficient APC function (18, 19). In the current study, we investigated the effect of IgA deficiency on the inflammatory response associated with allergic asthma using a murine model of allergic lung inflammation. The ability of SIgA to neutralize allergens and prevent sensitization is contentious; some reports (9, 20, 21) have demonstrated a negative correlation between IgA levels and allergic sensitization, whereas others have found no such association (22, 23, 24). Despite this, a possible link between IgA and the pathophysiology of established allergic asthma has been alluded to in a number of correlative human studies. Allergen-specific IgA can be isolated from the bronchoalveolar lavage (BAL) of asthmatics (25, 26) and is increased during periods of high-allergen exposure (27). IgA is capable of inducing degranulation of human eosinophils, the major cell type present in the inflammatory infiltrate of asthmatics (11). CD89 (FcR1) is up-regulated on eosinophils isolated from allergic patients, and IgA levels correlate with levels of eosinophil granule proteins in BAL fluids from asthmatics (28, 29, 30). Furthermore, Ag-specific IgA in BAL fluids has been correlated with stronger Ag-induced, late-phase asthmatic reactions (31). Although the ability of SIgA to neutralize Ags without the induction of inflammation is well established, the above results, although offering no specific mechanism of activity, suggest that IgA may enhance an ongoing mucosal inflammatory response.
We have now examined the ability of IgA–/– and pIgR–/– mice to respond to intranasal (IN) challenge in a murine model of allergic lung inflammation. Using this model, a defect in B cell expression was identified in the lungs of IgA–/– mice. Furthermore, it was found that although SIgA does not appear to contribute to the development of allergic lung inflammation following systemic (i.p.) sensitization and mucosal (IN) challenge, Ag-specific B cells appear to be an essential population for the development of respiratory inflammation induced by a soluble protein Ags (OVA), likely by functioning as APCs.
Results
Recruitment of inflammatory cells into the lungs of sensitized mice following IN Ag challenge is decreased in IgA–/– mice
The most striking feature of murine allergic lung inflammation is the influx of inflammatory cells into the lung subsequent to allergen priming and challenge. After OVA sensitization and IN challenge, IgA+/+ mice developed profound inflammation (inflammatory score = 2.60 ± 0.53, n = 24) that manifested as PV and peribronchiolar cuffing (Fig. 1a, arrows), although areas of diffuse infiltrate were observed as well. Hypertrophy of the bronchial epithelium and plugging of some airways with mucus, inflammatory cells, and debris was also evident in the lungs of these mice (Fig. 1, a and b). By comparison, identically treated IgA–/– mice had a significantly diminished inflammatory response (inflammatory score = 1.50 ± 0.38, n = 24) (p < 0.01), in which fewer infiltrates were observed, primarily as PV cuffing with very little diffuse inflammation (Fig. 1, c and d). Furthermore, hypertrophy of the bronchial epithelium was mostly absent and plugging of airways was not seen (Fig. 1d). Eosinophils were detected in infiltrates from both groups of mice using a CN–EPO stain (Fig. 1, e and f); however, significantly (p < 0.01) greater numbers were detected within the infiltrates of IgA+/+ mice (307.05 ± 12.23 eosinophils/field) compared with IgA–/– mice (122.85 ± 32.64 eosinophils/field). Saline-sensitized and -challenged, saline-sensitized and OVA-challenged, and OVA-sensitized and saline-challenged mice were used as controls and demonstrated negligible inflammation (data not shown).
Similar to the responses observed in lung tissues, significantly more inflammatory cells were present in BAL fluids from IgA+/+ mice compared with IgA–/– mice (8.11 x 105± 3.10 x 105 total cells vs 2.91 x 105± 1.12 x 105 total cells; p < 0.005). Differential staining of BAL fluid inflammatory cell isolates revealed that eosinophils were the predominant cell population in both groups of mice, but again, there were significantly (p < 0.05) more eosinophils isolated from IgA+/+ mice than IgA–/– mice (Fig. 1g). Overall, eosinophils constituted 10- 20% less of the total inflammatory cell population isolated from BAL fluids of IgA–/– mice compared with IgA+/+ mice (data not shown). Lymphocyte, macrophage, and neutrophil numbers were comparable between both groups of mice (Fig. 1g).
To determine whether the diminished inflammatory response in IgA–/– mice was Ag specific or due to an inherent defect caused by disruption of the -H chain, inflammation was chemically induced by IN challenge with bleomycin sulfate. Intranasal administration of bleomycin resulted in recruitment of neutrophils and macrophages to lung tissues of both groups of mice (Fig. 2). The development of inflammation was also reflected in the large numbers of inflammatory cells isolated from BAL fluids of IgA–/– and IgA+/+ mice (5.61 x 106 ± 0.01 x 106 vs 1.31 x 106 ± 0.53 x 106 total cells, respectively) (p < 0.05). These results indicate that IgA–/– mice are fully capable of manifesting a pulmonary inflammatory response and suggest that the defect observed in IgA–/– mice following OVA sensitization and challenge is a deficiency in responding to antigenic stimulation. Furthermore, the presence of significantly more inflammatory cells in BAL fluids from IgA–/– mice compared with IgA+/+ mice suggests that IgA plays different roles in inflammatory responses induced by different stimuli, e.g., Ag-specific vs Ag-nonspecific inflammatory responses.
Unlike other isotypes tested, levels of total serum IgE were significantly lower in IgA–/– mice compared with IgA+/+ mice (Fig. 3b). To determine whether decreased production of serum IgE was responsible for the diminished inflammatory response observed in these mice, both groups were sensitized in the presence of IL-4, the switch factor for IgE. Inclusion of IL-4 at the time of sensitization resulted in increased production of total IgE in the sera of IgA–/– mice to a level approximately half of that seen in IgA+/+ mice, without a concurrent increase in flammation (data not shown). However, IgE levels were still significantly lower than those in IgA+/+ mice and could not be further increased with higher concentrations of IL-4. Thus, although these results suggest that IgE does not influence the development of inflammation in our model, we cannot completely exclude a role for this Ig isotype in enhancing inflammation in IgA+/+ mice. The data also indicate that disruption of the -H chain has affected the ability of IgA–/– mice to produce Ig of this isotype. This confirms previous results of Harriman et al. (12), who found that naive IgA–/– mice have significantly lower levels of serum IgE compared with wild-type littermates.
The reduced inflammatory response in IgA–/– mice is due to lack of total IgA not SIgA
To establish whether the abrogated inflammatory response observed in IgA–/– mice is the result of a lack of total IgA or SIgA, we used pIgR–/– mice, which are unable to transcytose IgA across the mucosal epithelium. As a result, these mice lack SIgA but have 100-fold greater levels of serum IgA (32). OVA sensitization and challenge of pIgR+/+ and pIgR–/– mice resulted in a comparable influx of inflammatory cells into the lungs (Fig. 4, a and b, respectively) (pIgR+/+ inflammatory score = 2.33 ± 0.72, n = 12; pIgR–/– inflammatory score = 2.13 ± 0.49, n = 12) and BAL fluids (pIgR+/+ = 3.52 x 105 ± 1.97 x 105 total cells/lung and pIgR–/– = 3.93 x 105 ± 1.56x 105 total cells/lung). In addition, total and IgG1 OVA-specific Ab levels were similar in BAL fluids (Fig. 4c) and sera (Fig. 4d) from both groups of mice, as were serum levels of total IgE (Fig. 4d). The only differences observed between the mice were in levels of OVA-specific BAL and serum IgA, which were significantly higher in pIgR–/– mice compared with pIgR+/+ mice (Fig. 4, c and d). pIgR–/– mice do have some IgA in mucosal surfaces because disruption of the receptor leads to increased paracellular leakage of serum proteins into the mucosal lumen (32). Thus, although these results indicate that SC does not contribute to the development of allergic lung inflammation, we cannot completely exclude a role for IgA Abs in the respiratory tract.
Ag-specific B cells are required for enhanced inflammation in IgA+/+ mice
To determine whether B cells are required for the enhanced inflammatory response observed in IgA+/+ mice, cell transfer experiments were performed using sensitized splenocytes or CD4+ T cells as described previously (35). Splenocytes were isolated from 14-day sensitized IgA+/+ and IgA–/– mice and cultured with OVA for 72 h. Whole splenocyte populations or CD4+ T cells isolated from cultured splenocytes were then injected i.p. into naive mice. The recipients were challenged IN with OVA, and the lungs were examined by histology for inflammation. Transfer of IgA+/+-sensitized splenocytes into naive IgA+/+ mice resulted in profound inflammation (inflammatory score = 3.25 ± 1.07, n = 8), similar to that seen in directly sensitized and challenged IgA+/+ mice (Fig. 7a). As expected, this was significantly greater (p < 0.05) than the moderate inflammatory response observed in the IgA–/–IgA–/– cell transfer group (inflammatory score = 1.83 ± 1.00, n = 9) (Fig. 7a). In contrast, transfer of isolated CD4 T cells from sensitized IgA+/+ cells into IgA+/+ recipients resulted in a decreased inflammatory response (inflammatory score = 1.58 ± 0.38, n = 6) comparable to that observed in the IgA–/–IgA–/– group (inflammatory score = 1.58 ± 0.38, n = 6) (Fig. 7b). This indicates that isolation of CD4+ T cells from whole splenocytes resulted in loss of some cell type responsible for the enhanced inflammatory response observed in IgA+/+ mice. To determine the importance of Ag-specific B cells in this response, whole splenocyte populations depleted of CD19+ cells were transferred into naive mice. Depletion of CD19+ cells from the splenocyte population before transfer resulted in a significantly attenuated inflammatory response in the IgA+/+IgA+/+ group (inflammatory score = 1.83 ± 0.76, n = 3), similar to that observed in the IgA–/–IgA–/– group (inflammatory score = 1.67 ± 0.58, n = 3) (Fig. 7c). Differential staining of BAL fluid cells revealed that eosinophil numbers mirrored the inflammatory responses observed by tissue staining (data not shown). Transfer of naive cells followed by IN challenge did not result in the recruitment of inflammatory cells to the lungs of any group of mice. Thus, Ag-specific B cells are required for the enhanced inflammatory response observed in allergen-sensitized and -challenged IgA+/+ mice.
Discussion
Several studies have suggested that IgA plays a role in the pathology of allergic inflammation in asthmatic patients (25, 26, 27, 28, 29, 30, 31). In the present study, using a murine model of allergic lung inflammation, we demonstrate that IgA–/– mice have reduced pulmonary inflammation following sensitization and IN challenge with soluble OVA. However, this defect appears to be linked to the absence of Ag-specific B cells within the pulmonary compartment of IgA–/– mice, rather than a lack of SIgA, as pIgR–/– mice developed inflammation similar to pIgR+/+ mice, although using this model we could not completely exclude a role for IgA Abs that may have transudated from serum into the respiratory tract. An earlier study by Arulanandam et al. (18) showed that the absence of IgA in IgA–/– mice results in defective APC function. Furthermore, Mbawuike et al. (19) demonstrated altered Th1 cell function in IgA–/– mice that might also be attributable to a defect in APC function. The present findings suggest that this defect exists because of a lack of B cells in the lung and that B cells at this site play an important role as APCs.
Sensitization and challenge of IgA–/– mice with OVA resulted in an abrogated inflammatory response compared with IgA+/+ mice, with markedly reduced recruitment of cells to both the tissue and airways, and decreased levels of allergen-specific Ab and Th2 cytokines in BAL fluids. Numbers of eosinophils, the dominant cell type present in allergic lung infiltrates, were significantly diminished in the lungs of IgA–/– mice, likely the result of decreased IL-5 production because IL-5 is a hemopoietic and chemotactic factor for these cells (37). In addition, OVA-specific Ab levels were found to be significantly lower in BAL fluids from IgA–/– mice while serum Ab levels were similar. Ongoing work in our laboratory has demonstrated that Abs found in BAL fluids after IN vaccination are produced by resident pulmonary B cells (D. Albu et al., manuscript in preparation). Thus, decreased levels of allergen-specific Ab in BAL fluids of IgA–/– mice likely reflect a lack of Ag-specific B cells at the site of inflammation, which is consistent with our data demonstrating defective expression of IgG+ B cells in the lungs of these mice.
Total and IgG1 allergen-specific serum Ab levels were similar between both groups of mice, indicating normal systemic B cell responses. However, decreased levels of IgE were observed in IgA–/– mice compared with IgA+/+ mice. Total Ig, rather than IgE Ab, is typically used as a marker for increased isotype production in allergy models because the low level of IgE production coupled with uptake by the high-affinity FcRI usually renders free IgE Ab undetectable via conventional ELISA. Administration of IL-4, the switch factor for IgE, during sensitization resulted in an increase in total serum IgE production to a level approximately half of that observed in IgA+/+ mice, without a concurrent increase in inflammation, suggesting that IgE is not critical for the development of allergic inflammation in our model. These results corroborate other studies using IgE and mast cell deficient mice, which imply that this Ab isotype and its interaction with mast cells does not contribute to the development of allergic lung inflammation in mice (38, 39). However, as we were unable to induce similar levels of IgE production in IgA+/+ and IgA–/– mice, we cannot fully exclude a role for this isotype in the development of allergic lung inflammation. These data also indicate that disruption of the -H chain affects class switching to other isotypes as well. This is in accordance with Harriman et al. (12), who demonstrated significantly lower levels of serum IgE in naive IgA–/– mice compared with IgA+/+ mice. Thus, disruption of IgA leads to a generalized defect in class switching; curiously, this is a phenomenon that is also observed in many IgA-deficient patients and a defect that one would not expect in a system with such a dissimilar etiology from human IgA deficiency (8, 9, 40).
Th2 cytokine levels were significantly lower in IgA–/– mice compared with IgA+/+ mice at early time points during IN challenge with OVA but were similar at later times. This correlates with an earlier influx of inflammatory cells into the lungs of IgA+/+ mice compared with IgA–/– mice, whereas the diminished level of cytokine production during the IN challenge of IgA–/– mice is indicative of the weaker inflammatory response observed in these mice. It is unclear why a discrepancy exists between mRNA transcript and protein levels. There may be a difference in posttranscriptional regulation of cytokine production and/or secretion between IgA+/+ and IgA–/– mice. Alternatively, the isolation of total lung mRNA for transcript analysis, rather than specific cell types, may have muted subtle differences in expression making the ELISA a more sensitive method for detecting differences in cytokine levels between the two groups of mice. In either case, there is a distinct difference in levels of cytokine produced during IN challenge of IgA+/+ and IgA–/– mice.
Bleomycin challenge was used to confirm that the reduced inflammatory response observed in allergen-exposed IgA–/– mice was a deficiency in mounting an Ag-specific immune response. Bleomycin induces inflammation in the lungs and skin due to the lack of the enzyme bleomycin hydrolase in these tissues (41). Although the inflammatory response observed after bleomycin challenge is primarily mediated by neutrophils and macrophages, unlike allergic lung inflammation which involves significant eosinophil infiltration, the resultant response demonstrated that IgA–/– mice are capable of mounting pulmonary inflammation. However, the recruitment of significantly more inflammatory cells to the lungs of IgA–/– mice relative to IgA+/+ mice indicates that IgA and/or mucosal B cells play very different roles in mediating inflammation induced by different stimuli. IgA is widely considered to be an anti-inflammatory Ab at mucosal sites, where it is involved in the neutralization and clearance of bacterial, viral, and other Ags (5). However, as administration of bleomycin sulfate does not induce Ag-specific Abs, any anti-inflammatory effect of IgA in this system would not occur through neutralization but through an, as of yet, unknown mechanism.
Analysis of pulmonary leukocyte populations in IgA+/+ and IgA–/– mice revealed significantly fewer B lymphocytes in the lungs of IgA–/– mice in the presence or absence of allergic lung inflammation. In addition, immunofluorescent tissue staining demonstrated a complete absence of IgG staining cells in lung sections from IgA–/– mice following the induction of allergic lung inflammation, whereas many IgG+ cells were present in the lungs of IgA+/+ mice. This finding correlates with the negligible amounts of IgG1 found in the airways, although it is unclear why disruption of the -H chain in IgA–/– mice results in a lack of IgG+ B cell expression in the lungs. In a recent study, Uren et al. (42) also described disrupted mucosal B cell homeostasis in pIgR–/– mice, which contain 100-fold greater levels of serum IgA than pIgR+/+ mice. However, the altered B cell homeostasis manifested in this case as an increased number of IgA+ plasmablasts in the lamina propria and spleen (42); the mechanism for this defect was not determined. As IgA–/– mice have decreased numbers of B cells at mucosal sites, it is likely that IgA production influences B cell homeostasis; the mechanism(s) responsible for this defect are currently being investigated.
Cell transfer studies were performed in an effort to determine whether the enhanced inflammatory response observed in IgA+/+ mice compared with IgA–/– mice is dependent upon the presence of Ag-specific B cells. Transfer of whole splenocyte populations from IgA+/+IgA+/+ mice resulted in a profound inflammatory response that was visibly more severe than the response seen in IgA–/–IgA–/– mice. However, removal of Ag-specific B cells from the transferred population, either by isolating CD4+ T cells or by depleting CD19+ B cells, resulted in a loss of the response observed in IgA+/+ mice. Transfer of purified CD19+ B cells was not performed as splenocyte populations depleted of CD4+ cells have been shown previously to be incapable of transferring disease to naive animals (35). Unfortunately, we could not perform IgA+/+IgA–/– and IgA–/–IgA+/+ cell transfers because the IgA–/– mice are not on a pure inbred background. It is not clear what role B cells play in the response, i.e., if they enhance the response through local Ab production and/or whether they function as APCs. B cells are potent Ag-specific APCs, especially for soluble protein Ags and/or in cases of limited Ag concentrations (43, 44, 45, 46, 47). They can rapidly take up, catabolize, and present Ags and have been shown to play an important role in optimizing Ag presentation to T cells (44, 47). Furthermore, lack of B cells during presentation has been shown to result in reduced T cell responses (44, 45, 47, 48, 49). Dendritic cells (DCs) are likely responsible for the limited inflammatory response observed in IgA–/– mice; however, the lack of adjuvant during IN challenge in this model would cause suboptimal DC activation and maturation. As it has been reported that alveolar macrophages are poor APCs (50), B cells may be a critical APC population within the respiratory tract. This would explain previous data demonstrating that IN immunization of IgA–/– mice with vaccine + adjuvant resulted in protective immunity, whereas vaccination of these mice in the absence of a DC-stimulating adjuvant resulted in a loss of protection (18). We have found that equivalent numbers of IgA–/– and IgA+/+ CD45+ APC populations can stimulate similar levels of cell proliferation and cytokine production cells by CD4+ T cells purified from OT-II mice, which are specific for OVA (our unpublished observations). Thus, there is no inherent defect in the ability IgA–/– B cells to present Ags; rather, our data indicate that reduced B cell numbers in the lungs of IgA–/– mice leads to defective APC activity. Under normal conditions, such as in IgA+/+ mice, Ag-specific B cells likely traffic through mucosal effecter sites, including the lungs, where they encounter their cognate Ags and, with DCs, initiate activation of memory T cells.
It has been reported previously, using B cell-deficient (μMT) mice, that B cells do not contribute to the induction of murine allergic lung inflammation. However, it should be recognized that μMT mice are not fully deficient in B cells. Several studies have demonstrated the presence of IgA-, IgG-, and IgE-producing B cells in the absence of membrane IgM and IgD expression (51, 52, 53). Furthermore, Macpherson et al. (51) demonstrated that a significant number of IgA+ cells are present in an immune effector site, i.e., the lamina propria of μMT mice. In addition, it has been suggested that B cells are critical for normal secondary organ lymphorganogenesis (45, 54, 55, 56). Lack of B cells during development may result in the formation of compensatory mechanisms that might not occur when only a subpopulation of B cells is missing. Evidence for this comes from experiments by Rivera et al. (45), demonstrating that the ability of μMT APCs to induce functional immunity is profoundly reduced when these cells were placed in a normally structured immune system by reconstituting irradiated C57BL/6 mice with μMT bone marrow. This suggests that the lack of B cells during development in μMT mice resulted in formation of a compensatory pathway that allows adequate APC function and that such a mechanism was not functional in a wild-type environment (45). In addition, while some studies using B cell-deficient mice found that B cells do not contribute to the development of normal immune responses, others have demonstrated defects in CD4+ T cell function generated in the absence of B cells (43, 44, 45, 46, 48, 49, 57). It is possible that differences in immunization protocols, the types of Ags and adjuvants used (43), as well as the background strain of mice (45), are partially responsible for the discrepancies in experimental findings between different groups.
We believe that the lack of mucosal B cells and deficient allergic lung inflammation observed in IgA–/– mice correlates with the previously reported defect in APC function in these mice (18). In that study, Arulanandam et al. (18) noted that a defect in APC function was observed when attempting to elicit protection from lethal influenza challenge via IN immunization with a protein subunit vaccine. However, the defect was not observed when mice were immunized with the vaccine plus a strong adjuvant (IL-12). Similarly, IgA–/– mice do not have an increased susceptibility to infection with influenza virus (13, 19, 58), HSV-2 (16), Shigella flexneri (17), or Helicobacter pylori (15), following immunization with a strong adjuvant or attenuated bacteria/viruses. However, under such conditions, DCs and macrophages would be adequately stimulated to present Ags to T cells, diminishing the need for B cells to function as APCs. In the absence of DC-activating adjuvants, such as in the allergic lung inflammation model used here, B cells may play an essential role in enhancing immune responses. Our findings imply that in the absence of Ag-specific B cells at mucosal sites, there is a need for DC-activating adjuvants for efficacious immunization. It has been reported that some IgA-deficient humans also have defects in mucosal B cell numbers (59, 60); it will be of interest to determine whether such a defect is responsible for the increased incidence of infectious sinopulmonary disease observed in these individuals (8, 9).
Acknowledgments
We thank Dr. Maria Lopez for her guidance in operating the flow cytometer and evaluating flow cytometry data and the Immunology Core of the Center for Immunology and Microbial Disease for histology services.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
1 This work was supported by National Institutes of Health Grant AI41715 and by Philip Morris USA, Inc., and by Philip Morris International. P.M.A. was supported by National Institutes of Health Training Grant AI49822.
2 Address correspondence and reprint requests to Dr. Dennis Metzger, Center for Immunology and Microbial Disease, Albany Medical College, Mail Code 151, 47 New Scotland Avenue, Albany, NY 12208. E-mail address: metzged@mail.amc.edu
3 Abbreviations used in this paper: pIgR, polymeric IgR; SC, secretory component; SIgA, secretory IgA; BAL, bronchoalveolar lavage; IN, intranasal; PV, perivascular; CN–EPO, cyanide-resistant eosinophil peroxidase; DC, dendritic cell.
Received for publication March 22, 2005. Accepted for publication May 12, 2005.
References
Mestecky, J., I. Moro, M. Kerr, J. M. Woof. 2005. Mucosal immunoglobulins. J. Mestecky, and M. E. Lamm, and W. Strober, and J. Bienenstock, and J. R. McGhee, and L. Mayer, eds. In Mucosal Immunology Vol. 1: 153-181 Academic Press, San Diego. .
Johansen, F. E., R. Braathen, P. Brandtzaeg. 2000. Role of J chain in secretory immunoglobulin formation. Scand. J. Immunol. 52: 240-248.
van Egmond, M., C. A. Damen, A. B. van Spriel, G. Vidarsson, E. van Garderen, J. G. van de Winkel. 2001. IgA and the IgA Fc receptor. Trends Immunol. 22: 205-211.
Phalipon, A., B. Corthesy. 2003. Novel functions of the polymeric Ig receptor: well beyond transport of immunoglobulins. Trends Immunol. 24: 55-58.
Russell, M. W., M. Kilian. 2005. Biological activities of IgA. J. Mestecky, and M. E. Lamm, and W. Strober, and J. Bienenstock, and J. R. McGhee, and L. Mayer, eds. Mucosal Immunology 267-289 Academic Press, San Diego. .(Paul M. Arnaboldi*, Melis)