Airway hyper-reactivity mediated by B-1 cell immunoglobulin M antibody generating complement C5a at 1 day post-immunization in a murine hapten model of non-atopic asthma

Abstract

Contact skin immunization of mice with reactive hapten antigen and subsequent airway challenge with the same hapten induces immediate airflow obstruction and subsequent airway hyper-reactivity (AHR) to methacholine challenge, which is dependent on B cells but not on T cells. This responsiveness to airway challenge with antigen is elicited as early as 1 day postimmunization and can be adoptively transferred to naïve recipients via 1-day immune cells. Responses are absent in 1-day immune B-cell-deficient JH^−/− mice and B-1 B-cell-deficient xid male mice, as well as in recipients of 1-day immune cells depleted of cells with the B-1 cell phenotype (CD19⁺ B220⁺ CD5⁺). As B-1 cells produce immunoglobulin M (IgM), we sought and found significantly increased numbers of anti-hapten IgM-producing cells in the spleen and lymph nodes of 1-day immune wild-type mice, but not in xid mice. Then, we passively immunized naive mice with anti-hapten IgM monoclonal antibody and, following airway hapten challenge of the recipients, we showed both immediate airflow obstruction and AHR. In addition, AHR was absent in complement C5 and C5a receptor-deficient mice. In summary, this study of the very early elicited phase of a hapten asthma model suggests, for the first time, a role of B-1 cells in producing IgM to activate complement to rapidly mediate asthma airway reactivity only 1 day after immunization.

Introduction

Increased airway contractility is a cardinal sign of bronchial asthma and occurs during chronic inflammation in the airways. The pathogenesis of airway hyper-reactivity (AHR) appears to be different in atopic and non-atopic asthma syndromes that are characterized, respectively, by the presence or absence of antigen-specific immunoglobulin E (IgE) in the serum of immunized individuals.¹ Interleukin (IL)-4- and IL-5-positive cells have been found more frequently in the airways of individuals with atopic asthma, corresponding to activated T helper 2 (Th2)-type cells.¹ In contrast, neutrophils staining for IL-8 are more frequent in patients with non-atopic asthma, perhaps corresponding more to T helper 1 (Th1)-driven inflammation.¹

In a murine model of AHR, Th1-type airway inflammation has been demonstrated in mice that are skin-contact-sensitized with the hapten picryl chloride [trinitrophenyl chloride (TNP-Cl)] and airway challenged with the same hapten antigen in an aqueous form.² However, we have shown previously that the cells responsible for AHR in this model are not T cells, but B cells, and that AHR can be elicited as early as 1 day postimmunization.³ Subsequent cell-transfer experiments showed that B cells mediate antigen-induced AHR only 1 day after immunization.³ In fact, the phenotype of 1-day immune cells that transfer AHR (CD3^– TCR-α^– CD4^– CD8^– CD19⁺B220⁺ CD5⁺)³ resembles B-1 cells.⁴ These are a subset of B lymphocytes that are infrequent in lymph nodes and spleen (≤ 1% of total cells), and reside mainly in pleural and peritoneal cavities where they comprise 10–40% of the total cells.⁵ B-1 cells are claimed to self-replicate in the peritoneal cavity throughout life, and to originate in fetal/neonatal liver from progenitors, possibly distinct from those for ‘conventional’ B lymphocytes (B-2 cells), which produce T-cell-dependent immunoglobulin G (IgG) and IgE that mediate acquired immune protection and atopic allergic responses, respectively.^4–7 In contrast, the main function of B-1 cells is thought to be the production of natural background immunoglobulin M (IgM) that are present in normal serum and can bind antigen to then activate complement to mediate the first-line ‘natural’ defence mechanisms during the onset of infection.^4–7

In the current study, we used mice deficient in αβ T cells, or deficient in B cells, to confirm that B cells (and not T cells) mediate AHR in this hapten-induced non-atopic asthma model. We focused on the airway responses of 1-day hapten-immune mice. We showed, employing xid B-1-cell-deficient mice (and also via the specific depletion of CD5⁺ and CD19⁺ cells employed in cell transfer), that the B-1-cell subset of B cells is responsible for AHR. Study of 1-day immune mice allowed analysis at a time when sensitized B-2 cells and T cells do not play any role, as at least 3–4 days are required for their activation.^8–11

We also described a new phenomenon of immediate airflow obstruction, peaking 15 min after airway antigen exposure in 1-day skin-immunized mice. In addition, we confirmed the AHR responses to nebulized methacholine, peaking 48 hr after airway antigen challenge and in 1-day skin-immunized mice, and we determined that both are caused by B-1 cells. As B-1 cells principally produce IgM that strongly activates complement, we investigated the production of anti-TNP IgM by spleen and lymph node cells after only 1 day, and also analysed the role of monoclonal IgM and whether complement C5a was involved in the airway responses.

The results suggest that hapten-specific IgM, probably derived from B-1 cells within 1 day, combines in the airway tissues with hapten–self-protein conjugates generated by airway challenge with reactive hapten antigen. This antigen–IgM complex locally activates complement, which then activates the immediate airway response (IR) and AHR responses. We propose that a newly recognized combined innate and early acquired immune cascade, which consists of B-1-cell-produced IgM that activates complement to generate C5a to activate C5a receptors (C5aR), mediates early airway reactivity in a hapten asthma model just 1 day after immunization.

Materials and methods

Mice

Specific pathogen-free normal 8–12-week-old female mice (unless stated otherwise) were purchased from Jackson Laboratory (Bar Harbor, ME) and were used in groups of n = 4–6 per experiment, unless stated otherwise. They comprised: TCR-α^–/– mice (T-cell deficient, C57Bl/6 background);¹²µMT (deficient in B cells, except for immunoglobulin A-secreting B cells;¹³ Richard Flavell, Yale University, New Haven, CT) and B10.A control mice; JH^–/– (pan B-cell deficient; Mark Schlomchik, Yale University, New Haven, CT)¹⁴ and BALB/c H-2b control mice; CBA/N-xid males (B-1 cell deficient)¹⁵ and CBA/J control mice; B10.D2/o (complement C5-deficient) and B10.D2/n control mice; and C5aR^–/– (C5a receptor deficient) (Craig Gerard, Harvard University, Boston, MA)¹⁶ and B6/129 control mice.

Active immunization and airway antigen challenge

Mice were shaved on the abdomen and chest, 1 day prior to immunization, using a blade and a small amount of soap that was washed off. The following day, 200 µl of 0·5% TNP-Cl (diluted in 100% ethanol; Nacalai Tesque, Inc., Kyoto, Japan) was applied on the shaved abdomen, chest and dorsum of all four paws. Twenty-four hours after immunization with TNP-Cl, mice were lightly anaesthetized with ether. When they reached the stage of rapid and shallow breathing, 50 µl of 0·6% trinitrophenyl sulphonic acid (TNPSA; Wako Pure Chemical Industries, Ltd., Chuo-ku, Osaka, Japan; Fluka, Neu-Ulm, Germany; or Eastman Kodak Co., Rochester, NY), was applied in drops on the nostrils and inhalation of the solution was observed. TNPSA is a water-soluble antigenic equivalent of TNP-Cl and was diluted in phosphate-buffered saline (PBS) to a concentration of 0·6% and the pH adjusted to 7·2 using Na₂CO₃. This TNPSA was aliquoted into 1-ml vials and stored at −20°. For simplicity, TNP-Cl sensitization and TNPSA airway challenge are both referred to as ‘TNP’ hapten in the text.

Adoptive cell transfers

Donors were immunized as described above, and spleen and lymph node cells were harvested and transferred intravenously (i.v.) into naive recipients (≈ 7 × 10⁷ cells/ mouse in 200 µl of PBS). For transfers we combined together cells harvested from six lymph nodes (two axillar, two brachial and two inguinal) and one spleen per mouse, and pooled cells from four to six donors. In some experiments, immune cells were depleted in vitro of CD5⁺ or CD19⁺ cells by incubation at 37° for 45 min with anti-CD5 or anti-CD19 monoclonal antibody (BD Pharmingen, San Diego, CA), at a dilution recommended by the manufacturer, and rabbit serum complement at a predetermined dilution (Pel-Freeze, Brown Deere, WI), and then washed and injected i.v.

Enzyme-linked immunospot (ELISPOT) assay for anti-TNP IgM-producing cells

Separate spleen and lymph node single-cell suspensions were prepared, as described above, from male CBA/J and CBA/N-xid mice 1 day after skin immunization with 5% TNP-Cl, and were seeded in triplicate into 96-well filtration plates (containing Immobilon-P membranes; Millipore, Bedford, MA), precoated with 50 µl of TNP-bovine serum albumin (BSA) (100 µg/ml), at 2 × 10⁶ cells/well. Plates were incubated at 37° overnight, and then cells were discarded and wells washed three times with PBS and then three times with PBS containing 0·05% Tween-20. Plates were then incubated at 25° with 2 µg/ml of biotin-conjugated monoclonal anti-mouse IgM (BD Pharmingen) for 1 hr followed by incubation with streptavidin-horseradish peroxidase (1 : 200; Vector, Burlingame, CA) for 1 hr. Spots were developed by using 3-amino-9-ethyl-carbazole as a substrate, the reaction stopped by washing, and wells then dried at 25° in the dark. Membranes were removed with a scalpel, and then stuck on glass slides for the enumeration of spots under an inverted dissecting phase microscope. The number of spots representing anti-TNP IgM-producing cells were counted and expressed per organ.

Passive immunization

Mice were injected i.v. with purified hybridoma 32.17-derived anti-TNP specific monoclonal IgM (100 µg/ 500 µl/mouse) that was prepared as described previously.¹⁰ One hour after injection, mice were airway challenged with TNP.

Pulmonary function tests

Immediately after airway TNP challenge, conscious, surgically intact 1-day immune mice were transferred into plethysmographs and airflow obstruction was determined by measuring Penh (enhanced pause) for 30–60 min.¹⁷ We empolyed six plethsmograph chambers and six mice per experimental group, and tested one mouse per group each time in an air-conditioned laboratory. Airway reactivity initially was determined by measurement of lung resistance in tracheostomized mice.¹⁸ Thus, 24 or 48 hr after TNP airway challenge, mice were exposed to increasing doses of nebulized methacholine and the resulting AHR was measured by lung resistance (RL) in tracheostomized mice that were anaesthetized with pentobarbital (1–3 mg/mouse).³ AHR was also determined by using the Penh technique, and the findings correlated with RL. AHR is expressed either as PC100, the provocative concentration of methacholine that produces a 100% increase in lung resistance (RL; unit = negative log [g of methacholine/ml]), or as the area under the curve (AUC), calculated from the dose–responses to 1–100 mg/ml methacholine (11 min was measured for each dose; in units = Penh × min). In some experiments, the AUC was calculated from the time course of the AHR response to only one dose of methacholine (100 mg/ml), because repeated measurements of Penh over several days led to tachyphylaxis in naive animals or increased airway reactivity in immunized mice that were airway challenged with TNP. The probable reason for this was the presence of induced inflammatory mediators that may have reached the airways owing to plasma leakage in airway antigen-challenged mice, that may affect airway smooth muscle contractility or reduce the airway diameter as a result of airway wall oedema. Thus, we reduced the methacholine exposure to limit these undesirable consequences. Immediate airway response (IR) was measured by the Penh method over 30 min.

Bronchoalveolar lavage

The airways of dead mice were lavaged twice (with 1 ml of sterile PBS at 4°), the samples were centrifuged at ∼43 g for 5 min, the pellet was resuspended for counting total cells and then cytospun (5 min at ∼43 g; Cytospin-3; Shandon, Astmoor, UK) and stained with Diff-Quik (Dade Behring Inc., Newark, NJ) for differential counts.

Histological techniques

Mice were immunized 1 day previously and airway challenged with reactive hapten, as described above. Forty-eight hours after airway challenge, mice were anesthetized with thiopentobarbital and killed by cervical dislocation. Then, the trachea was intubated and the pulmonary circulation perfused with 5 ml of PBS to remove blood, and then lungs were carefully removed from the chest. The tracheal cannula was connected to a column containing 10% formalin and the lungs were allowed to slowly inflate under a pressure of 20 cm H₂O. The formalin-inflated lungs were submerged in 10% formalin for 24 hr, transferred into 70% ethanol for another 4 days, and then processed into paraffin. The fixed lung was cut into transverse sections and then re-embedded, so that the final slides showed three to four sections per lung that were cut perpendicular to the centriaxial airway. Sections were stained with haematoxylin and eosin, and with periodic acid Schiff.

Statistical analysis

Data were analysed by analysis of variance (anova) (as indicated in the Figure legends) or by the non-parametric Mann–Whitney test using the software program stat-view^® 4.5. A P-value of < 0·05 was considered significant.

Results

B cells, not T cells, mediate AHR in TNP-immune mice

To definitively determine whether T cells or B cells mediate AHR in mice actively immunized with the TNP hapten, we, respectively, examined responses in T-cell-deficient TCR-α^–/– mice and in pan B-cell-deficient µMT mice. Forty-eight hours after TNP airway challenge, significant AHR was elicited in TCR-α^–/– mice (Fig. 1, group C) and in the wild-type immune control (group B), but not in immunized µMT mice (Fig. 1, group E) compared with the control (group D). These results were caused by the absence of B cells and not B-cell-dependent tissue alterations in the lungs of µMT mice, as non-immune wild-type recipients of lymphoid cells from immune µMT mice still did not develop AHR (Fig. 1, group G), while mice receiving cells from immune wild-type mice did (group F). Thus, AHR in this hapten system did not depend on TCR-αβ T cells, or on T-cell-dependent antibodies, but was mediated by B cells.

Figure 1.

Trinitrophenyl (TNP) contact sensitization induces airway hyper-reactivity (AHR) owing to B cells and not T cells. AHR, determined by lung resistance (RL) measurement, developed in 7-day immune and TNP airway-challenged C57Bl/6 mice (group B) compared with non-immune and challenged controls (group A), also occurred in immunized T-cell receptor (TCR-α^–/–) mice (group C). AHR did not develop in immunized B-cell-deficient and airway-challenged µMT mice (group E), but was present in control B10.A mice (group D). Adoptive transfer of TNP immune spleen and lymph node cells from B10.A mice (group F), but not from similarly immunized µMT donor mice (group G), induced AHR in B10.A recipient mice.

Mechanism of airway responses in 1-day TNP-immune mice

Antigen challenge in 1-day immune mice induced an IR that peaked 15 min after challenge (Fig. 2a). We also confirmed that the AHR responses developing 48 hr after airway challenge with antigen in 1-day immune BALB/c mice (Fig. 2b).

Figure 2.

Immediate bronchoconstriction and airway hyper-reactivity (AHR) to methacholine is induced in 1-day trinitrophenyl (TNP) immune mice and is caused by B cells.(a) Immediate airway response (IR) in 1-day TNP immune BALB/c mice that were challenged intranasally with TNP (•) was significantly higher than in non-immune/challenged mice (○), or non-immune/saline-challenged mice (□) (n = 7 per group). (b) AHR to methacholine was induced in 1-day immune BALB/c mice 48 hr after airway TNP antigen challenge (•), compared with non-immune/challenged mice (○). (c) IR in naive BALB/c mice that received intravenous (i.v.) transfer of spleen and lymph node cells from 1-day immune (▴), compared with nonimmune BALB/c donors (△), that were airway challenged with TNP 1 day later (n = 6 per group). (d) IR is absent in 1-day immune pan B-cell-deficient JH^–/– mice (▴) compared with 1-day immune and TNP airway-challenged BALB/c controls (□). The baseline of non-immune/saline-challenged JH^–/– (•) did not differ from non-immune BALB/c (○).

In Fig. 2(a), we demonstrated that saline instillation into the airways of non-immune mice produced no increase in Penh, while inhalation of TNP caused a small, but insignificant, increase in Penh, probably owing to non-specific irritation. In preliminary experiments, we tested whether there was any difference in the effect of saline instillation on Penh between immune and non-immune mice; no difference was observed. Thus, in subsequent experiments we used non-immune mice treated to inhale TNP as a more relevant control.

The immunological nature of the IR was confirmed by cell transfer from 1-day immune BALB/c mice. Naive recipients challenged with TNP had an increased IR, compared with controls (Fig. 2c). Furthermore, the IR was not elicited in pan B-cell-deficient JH^–/– mice airway challenged 1 day postimmunization (Fig. 2d). Also, the AHR did not develop in these JH^–/– mice (data not shown) and similarly was not elicited in B-1-cell-deficient male CBA/N-xid mice compared with intact CBA/J male mice (Fig. 3a, group D versus group C), compared with positive responses in wild-type CBA/J mice (group B versus group A). In addition, lymphoid cells from 1-day immune BALB/c mice transferred AHR to naive recipients (Fig. 3b, group B versus group A), but when the cells were depleted of CD5⁺ or CD19⁺ cells (markers on B-1 and B cells, respectively) prior to transfer, AHR did not develop (Fig. 3b, groups C and D). The obvious differences in basal airway reactivity between CBA/J and BALB/c mice that we observed (Fig. 3a,b, group A), represent commonly found interstrain differences previously studied in detail.^19,20 These differences have been attributed to enhanced signal transduction owing to genetic variations of agonist affinity for muscarinic receptors, and consequent up-regulation of G-protein levels in more responsive strains.²⁰ Also, the production of responses higher than background levels in the adoptive cell-transfer experiments was generally weaker than in actively sensitized mice, particularly when we tested an immune response that had just begun to develop at only 1 day postsensitization.

Figure 3.

Airway hyper-reactivity (AHR) in 1-day trinitrophenyl (TNP)-immune mice depends on B-1 cells. (a) The airways of 1-day trinitrophenyl (TNP)-immune B-1-cell-deficient CBA/N-xid male mice that were challenged with TNP were not hyper-reactive to methacholine 48 hr after the challenge [group D, determined by lung resistance (RL) measurement], compared with non-immune TNP airway-challenged xid mice (group C), while 1-day immune intact CBA/J mice developed AHR to methacholine (group B), compared with non-immune CBA/J controls (P < 0·05 group A versus group B). (b) BALB/c mice that received transfers of mixed spleen and lymph node cells of 1-day TNP-immune BALB/c mice, and 1 day later were airway challenged with TNP, were found to have AHR (group B, as determined by Penh measurement), compared with pretransfer baseline reactivity to methacholine (group A), and AHR did not develop in recipients of 1-day immune cells depleted of CD19⁺ cells (group D) or CD5⁺ cells (group C) prior to transfer by incubation with anti-CD19 or anti-CD5 monoclonal antibody (mAb) plus complement. Statistics: group A versus group B, P < 0·025; group C versus group D, non-significant; group A versus group C, non-significant; group B versus group D, P < 0·01 [analysis of variance (anova) P < 0·04]. (c) Enzyme-linked immunospot (ELISPOT) assay of spleen and lymph node cells producing anti-TNP IgM at 1-day postimmunization shows significant responses in spleens (left) and lymph nodes (right) of wild-type male CBA/J mice (group B), compared with natural background responses in sham-immunized controls (group A), or 1-day TNP-Cl-immunized B-1-cell-deficient xid male mice that did not respond (group D), compared with xid non-immune controls (group C). Results for spleen cells were also analysed by anova: group A versus group B, P < 0·05; group C versus group D, non-significant; group A versus group C, P < 0·01; group B versus group D, P < 0·01. The lymph node results were not statistically analysed as cells producing IgM were undetectable in groups A, C and D.

B-1 cell mediation of AHR in 1-day immune mice suggested the possibility that IgM, the major isotype produced by this B-cell subset, might be responsible. To explore this hypothesis, we therefore first employed an ELISPOT assay to enumerate spleen cells producing anti-TNP IgM on 1-day postimmunization of male CBA/J mice versus non-immune controls, compared to TNP-Cl male xid mice immunized similarly and non-immune controls. Significant anti-TNP IgM responses occurred in the spleen of 1-day immune wild-type CBA/J mice compared to natural background levels in controls (Fig. 3c, left, group B versus group A), while similarly immunized xid mice responded insignificantly at a level nearly 10-fold less (group D versus group C, and group B versus group D). One-day immune lymph nodes also showed anti-TNP IgM responses consisting of far fewer cells, but clearly greater than an undetectable background in non-immune CBA/J and xid mice (Fig. 3c, right).

Experiments carried out entirely in vivo directly tested the ability of anti-TNP IgM to mediate the IR and AHR. Passive i.v. immunization with 100 µg of purified anti-TNP IgM and TNP airway challenge induced an increased IR (Fig. 4, group C), and also AHR in BALB/c and CBA/J mice (Fig. 4, group C versus groups A and B, and group J versus group I). Even 10 µg of anti-TNP IgM was sufficient to induce AHR, 48 hr following airway challenge with TNP. The baseline increase induced by 100 mg/ml methacholine was 265 ± 27% in mice transferred with 10 µg of anti-TNP IgM versus 172 ± 12% in control i.v. saline-treated mice (P < 0·05). In contrast, there was no effect of transfer performed with an irrelevant myeloma IgM following a similar challenge with TNP (Fig. 4, group F versus group E). Together, these findings suggest that IgM antibody derived from B-1 cells may account for the IR and AHR in 1-day TNP-immune mice.

Figure 4.

Immediate airway response (IR) and airway hyper-reactivity (AHR) are elicited in mice passively immunized with trinitrophenyl (TNP)-specific IgM monoclonal antibody (mAb) and airway challenged. BALB/c mice were injected intravenously (i.v.) with anti-TNP IgM mAb and 1 hr later airway challenged with TNP. The IR in these mice (group C, determined by Penh measurement) was stronger than in mice that did not receive anti-TNP IgM Ab, but were airway challenged (group A), or received irrelvant IgM myeloma and were TNP challenged (n = 5 per group). BALB/c mice also developed AHR [determined by lung resistance (RL) measurement] after passive immunization with anti-TNP-IgM mAb and TNP airway challenge (group F), but not controls that were similarly challenged and received non-specific myeloma IgM mAb (group E), or mice that received no IgM at all (group D) (n = 5 per group). Similarly, CBA/J mice that were passively immunized with anti-TNP-IgM and challenged 1 hr later with TNP, elicited AHR 48 hr after the airway challenge (group J, determined by RL measurement), compared with CBA/J mice that either received only anti-TNP IgM mAb and were not challenged (group H), or were just airway challenged with TNP (group I), or were neither passively immunized nor airway challenged (group G) (n = 6 per group).

As the biological effects of IgM are often attributed to the activation of complement, we investigated whether complement may play a role in AHR by comparing C5-deficient B10.D2/o mice with their otherwise syngeneic C5-normal B10.D2/n controls. AHR was induced in 1-day TNP-immune C5-normal mice (Fig. 5, group C versus groups A and B), but was absent in C5^–/– mice (Fig. 5, group F versus groups D and E). Furthermore, AHR did not develop in 1-day TNP-immune C5a receptor-deficient (C5aR^–/–) mice that were airway challenged with TNP (Fig. 5, group J versus group I compared with wild-type group G versus group H). Therefore, AHR induced in this model of hapten-induced asthma was dependent on the stimulation of B-1-cell-derived and 1-day immune anti-TNP IgM that combined with challenge antigen to activate complement C5 to generate C5a which stimulated C5a receptors. The airway responses of 1-day immune wild-type B10.D2/n C5-normal mice and Bl6/129 controls (both H-2^b) were weak (Fig. 5, groups B and H). However, previous studies have shown that the production of hapten-induced cutaneous contact sensitivity (CS) responses following the same TNP skin immunization are weak in H-2^b background mice.²¹ Similarly, such H-2^b mice have been shown to elicit decreased AHR in the ovalbumin model of allergic bronchial asthma.²²

Figure 5.

Absence of airway hyper-reactivity (AHR) in 1-day immune C5-deficient and C5aR-deficient mice. AHR [determined by lung resistance (RL) measurement] developed in 7-day trinitrophenyl (TNP)-immune and TNP-challenged complement C5-normal B10.D.2n mice (group C) compared with challenged (group B) and unchallenged (group A) controls, but not in immune C5-deficient B10.D2.o mice (group F), versus controls (groups D and E). AHR developed in 1-day immune wild-type Bl6/129 mice (group H, determined by measurement of the Penh response induced by 100 mg/ml of methacholine over 10 min), compared with non-immune TNP-challenged mice (group G), but not in immune and challenged C5a receptor-deficient mice (group J), compared with non-immune challenged C5aR^–/– controls (group I).

Bronchoalveolar fluid in 1-day immune mice

Bronchoalveolar lavage (BAL) was performed in 1-day immune BALB/c mice when the IR was elicited at 15 min after challenge and also when AHR was induced 48 hr postchallenge. At 15 min, no cellular changes were noted in the BAL fluid of TNP-immune and challenged mice. The only change was observed at 48 hr, when the percentage of neutrophils among total cells increased from 3 ± 1% to 15 ± 8% in non-immune and TNP airway-challenged mice, and similarly to 12 ± 3% in 1-day immune and challenged mice. As there is no difference in BAL neutrophils between airway antigen-challenged immune and non-immune mice, these responses are probably caused by non-specific irritation owing to challenge with TNP hapten.

Lung histology in 1-day immune mice

Mice that were not challenged, regardless of whether they had been immunized or were entirely normal, had no lung inflammatory infiltrates or mucus. Mice that had been airway challenged with hapten, regardless of whether they had been immunized, showed small and very focal interstitial and intra-alveolar pneumonitis, which consisted predominantly of mononuclear cells with a minor component of eosinophils. In most cases, this small, non-specific infiltrate occupied less than 5% of the total lung area. The infiltrate usually involved the adjacent parenchymal vein and/or airway. However, away from these foci, no infiltrate occurred in bronchovascular bundles or around parenchymal veins, as occurs in conventional models of mice that are immunized and then airway challenged with protein allergens such as ovalbumin.²³ Therefore, it was not possible to distinguish the size or nature of the infiltrate between the immune and non-immune groups that both were airway challenged with hapten. No mucus was seen in any mouse.

Thus, histological airway findings in 1-day immune and challenged mice harvested 48 hr after airway challenge at the time of AHR, confirmed the BAL finding by showing that there was no associated specific cellular inflammation compared with identically airway challenged non-immune controls. Thus, immune cellular inflammatory infiltration probably did not contribute to the airway reactivity detected in 1-day TNP-immune and challenged mice.

Discussion

Antigen-induced airway responses in 1-day immune mice

We demonstrate, for the first time, that the IR and AHR airway responses can be elicited as early as 1 day after immunization. Rapidly sensitized B-1 cells, acting via their derived specific IgM, mediate both antigen-induced IR and subsequent AHR in 1-day immune mice. It is probable that the IR and AHR which occur very early after immunization, proceed as a result of complement activation, via the IgM complexed with hapten–self-protein conjugates, to generate active complement fragments such as C5a.

Figure 6 summarizes the findings and hypotheses resulting from this study. We propose that B-1 cells, perhaps in the peritoneal and pleural cavities, are activated early postimmunization by antigen, which consists of locally formed TNP–self-protein conjugates that are dispersed systemically from the skin site of contact sensitization.²⁴ The activated B-1 cells then migrate to the spleen and lymph nodes^11,25 to produce anti-hapten IgM that circulates within the first day postimmunization.¹¹ During subsequent TNP hapten challenge in the airways, the hapten forms conjugates with local airway self-proteins²⁶ that then are bound by anti-TNP IgM from the circulation. The locally formed IgM–antigen immune complexes then activate complement, leading to the IR and subsequently to AHR. The immediate bronchoconstrictive responses are either induced directly via formation of bronchoconstrictive C5a,^27,28 and/or indirectly via C5a triggering C5a receptors. AHR is probably caused by the late effects of the mast cell and/or platelet mediators.³³ C3a could also be involved similarly.^34,35

Figure 6.

Hypothesized asthma cascade in hapten immune contact sensitized mice, consist of B-1 cell- to immunoglobulin M- (IgM) to complement-driven airway responses. The postulated cascade of events are shown from immunization to airway antigen challenge in order to elicit airway hyper-reactivity (AHR) just 1 day later. Trinitrophenyl (TNP) hapten skin contact immunization (upper left) activates TNP antigen-specific B-1 cells in pleural and peritoneal cavities to migrate to the lymphoid tissues. There, the B-1 cells probably mature into cells that secrete IgM which enter the systemic circulation (centre). Then, subsequent TNP airway challenge at 1 day results in the formation of local immune complexes of anti-TNP IgM binding with TNP-self conjugates formed in the airway tissues. These antigen–antibody complexes (upper right) activate complement, and component C5 generates C5a. The C5a (lower right) then either directly binds and activates C5a receptors in the airways, or acts indirectly by activating local mast cells or platelets via their C5a receptors to release bronchoconstrictive mediators. Either pathway then induces an immediate airway response (IR) peaking at 15–30 min and subsequent AHR measured at 24–48 hr (lower right).

Observations in this hapten asthma model parallel findings of analogous early immune responses rapidly induced within 1 day post-contact immunization and challenge in the skin.^9–11 In murine CS, 1-day immune B-1 cells mediate immediate hypersensitivity-like skin responses elicited by antigen challenge that is required to recruit effector T cells to mediate classical 24-hr responses.³⁶ Thus, mice immunized similarly to the asthma model by TNP skin painting, but subsequently TNP challenged on the ear skin at 1 day postimmunization, elicit rapid antigen-specific ear swelling that onsets at 15 min and peaks just 2 hr after local hapten challenge.⁸ This immediate skin response is caused by B-1 cell-produced IgM^10,21,37 that locally generates C5a from C5.^21,38,39 The generated C5a results in the recruitment of CS-effector T cells for subsequent development of the full 24-hr local inflammatory response^36,37 to elicit this prototypic example of classical delayed-type hypersensitivity (DTH). Importantly, the early 2-hr CS component is also elicited within 1 day after sensitization, and also can be transferred by 1-day immune B-1 cells.¹⁰ These rapidly generated specific B-1 cells are similar to the sensitized cells, identified in the present study, in immune spleen and lymph nodes that are needed for airway responses in 1-day contact immunized mice³ and produce anti-hapten IgM detected via ELISPOT. However, in the skin model, this early response at 1-day postimmunization is not followed by the 24-hr classical late phase, because activation of CS-effector αβ T cells requires 3–4 days.³⁶

Role of B-1 cells in the airway responses

Airway responses were absent in immunized pan B-cell-deficient µMT and JH^–/– mice, and in predominantly B-1-cell-deficient xid male mice. The absence of airway responses is probably caused by the lack of B-1 cells, because adoptive transfer of lymphoid cells from 1-day immune wild-type donors into naive recipients results in the elicitability of the AHR that is prevented by prior depletion of cells that are CD5⁺ or CD19⁺ (phenotypic markers of B-1 cells). To our knowledge, these results are the first indication that B-1 cells may be involved in bronchial asthma, and that asthmatic airway reactivity can occur as early as 1 day postimmunization.

Role of anti-TNP IgM in the airway responses

B-1 cells are the major source of normal serum IgM.^4–7 Pertaining to asthma, increased IgM was noted in some asthmatic patients,^40–43 but the role of IgM in asthma has not yet been investigated. We showed that passive immunization of naïve recipients with purified hybridoma-derived anti-TNP IgM monoclonal antibody resulted in the elicitability of IR and AHR in BALB/c or CBA/J recipients that were airway challenged with TNP. Furthermore, ELISPOT assay showed significant numbers of spleen and lymph node cells producing anti-TNP IgM just 1 day following immunization with TNP-Cl.

These responses were higher than natural background levels, while B-1-cell-deficient xid mice were relatively unresponsive. The ELISPOT results strongly suggest that IgM produced by B-1 cells within 1 day of immunization was responsible for IR and AHR. However, direct proof of mediation of airway responses by B-1-cell-derived IgM in the actively sensitized mice is difficult to obtain. Foremost, there is a high background of anti-TNP IgM in normal mouse serum owing to the presence of natural B-1 B-cell-produced anti-hapten IgM, even prior to immunization.⁴⁴ This was reflected in the ELISPOT data (Fig. 3c, group A). However, it is interesting that, despite the presence of B-1-cell-produced anti-TNP in the serum of non-immune mice, these normal background serum antibodies do not transfer 2-hr responsiveness in skin CS, while serum from 1-day immune mice does,¹¹ and the airways of non-immune mice that are producing these antibodies (Fig. 3c, group A), similarly are unresponsive to airway challenge with TNP hapten (Fig. 3a, group A). We have observed increased immediate airway responses in naïve BALB/c mice that were transferred with serum from 1-day PCl-immune BALB/c donors and then airway challenged, but this was not statistically significant (I. Kawikova and P.W. Askenase, unpublished observations). Furthermore, IR responses were absent when the recipients were transferred with serum from either TNP-Cl-immunized pan B-cell-deficient JH^–/– mice, or from BALB/c mice that were immunized with oxazolone, an irrelevant hapten (data not shown).

Thus, to confirm the involvement of anti-TNP IgM in the complement-dependent airway responses in mice actively sensitized only 1 day previously and appropriately producing small anti-TNP IgM responses (Fig. 3c), will require further detailed studies. Such studies must overcome the difficulty of unequivocally demonstrating relevant and rapidly produced immune IgM in 1-day immune mice, as the sera contain natural antibodies and residual antigen from recent immunization, and therefore probably also immune complexes.

Role of complement in airway responses

TNP-immunized and challenged C5^–/– or C5aR^–/– mice did not develop AHR, in contrast to controls, suggesting that C5, C5a and C5aR mediate the airway responses in 1-day TNP-immune mice. We previously established the role of C5³⁸ generating C5a, to activate C5aR³⁹ in the early immediate hypersensitivity-like phase of CS responses. In studies from other laboratories, considerable evidence has been put forward to suggest a role for complement, particularly C5a and C3a, in the elicited effector phase of asthma. In experimental allergic asthma models, responses are defective in C5- and C3a-deficient guinea pigs⁴⁵ and in C3aR-deficient mice.^46,47 Also, in a classical Th2 allergic asthma model in mice, airway inflammation and hyper-responsiveness are decreased by treatment with a specific complement inhibitor prior to challenge of previously sensitized hosts.⁴⁸ Increased plasma C3a⁴⁷ have been reported, and increases in C3a and C5a are present in the BAL of asthmatic patients after local allergen challenge.⁴⁸

Thus, immediate airflow obstruction in the hapten model may be a result of the local generation of biologically active complement C5a fragments, causing direct stimulation of airway smooth muscle via appropriate receptors to induce muscular contraction (Fig. 6).^27–32 Alternatively, C5a may act indirectly via activation of C5a receptors on local mediator-containing cells, such as mast cells or platelets, to release bronchoconstrictive mediators that also sensitize the airways for subsequent AHR (Fig. 6).³³ Activation of mast cells following cross-linking of their FcεRI receptors via attached IgE bound by multivalent challenge antigen, plays an important role in clinical immediate bronchospasm and also in the late-phase asthmatic responses.⁴⁹ Our findings lead us to suggest that perhaps soon after immunization, allergic asthma also may be mediated by an alternative and/or contributing pathway consisting of IgM antibody-dependent and complement-mediated mast cell activation.

The B-1 cell to IgM to C cascade, in relation to bronchial asthma

AHR in this hapten model was found to occur in the absence of significant BAL cellular responses or histological infiltrate above the background responses of mice just airway challenged with TNP antigen, which is unusual. Although dissociation between airway inflammation and hyper-reactivity have been reported,⁴⁹ it is important to note that our system is very different from nearly all current studies of bronchial asthma involving experimental animal models. The majority of published studies start by design at the point when the asthma model histologically resembles developed human bronchial asthma. This time-point usually follows a series of immunizations and airway challenges. Instead, we investigated the development of IR and AHR, which are the cardinal signs of bronchial asthma, very early after immunization, and after only one challenge. Thus, we tested airway function before full development of overt clinical asthma, and thus without BAL inflammation or an airway inflammatory infiltrate. Therefore, at this stage of understanding, our model could represent a very early phase in bronchial asthma.

The postulated asthma cascade, proceeding from activation of the B-1 cells, to IgM production, and then to local complement activation at challenge, has characteristics of both innate and adaptive immunity. The innate aspect is represented by the unusual shortness of time that is required for eliciting IR and AHR in PCl-immune mice, and by the involvement of complement. The B-1 cell IgM responses represent a bridge from innate to adoptive immunity. Finally, the ability to transfer the AHR with 1-day immune B-1 B cells also suggests a component of the early adaptive immune response.

The relevance of the early antigen-stimulated specific B-1 cell mechanisms that lead to IgM activation of complement to then mediate IR and AHR clinical asthma, is suggested by several observations. First, the model resembles hapten-induced occupational asthma, such as that induced by highly reactive diisocyanates, which does not seem to involve IgE,⁵⁰ and therefore hapten-specific IgM or IgG complement-activating antibodies could participate. Second, increased serum IgM levels have been described in some asthmatic patients.^40–43 Third, complement deposits were found in the airways of some patients who died of asthma.⁵¹ Fourth, increased C3a⁴⁷ were reported in active asthmatic patients, and C3a and C5a were recently detected in the BAL fluid of allergen-challenged asthmatics.⁴⁸ However, further studies are required to link our findings, regarding B-1 cell IgM activation of complement, with clinical asthma.

In summary (Fig. 6), by employing a hapten model, we found that B-1 cells are activated very early after immunization to rapidly produce antigen-specific IgM in just 1 day. Following subsequent airway challenge, on day 1 postimmunization, the 1-day immune IgM probably binds the challenge antigen to form IgM–antigen complexes in the airways that locally activate complement. Consequently, elaborated C5a, and perhaps C3a, may cause immediate airflow obstruction and subsequent AHR. C5a may act directly to stimulate the airways via their C5aR, or act indirectly to stimulate C5aR on local mast cells and/or platelets, resulting in the release of bronchoconstrictive mediators. This may sensitize the airways for AHR, without inducing significant accompanying local cellular inflammation at this very early time-point.