INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Sensitization of human airway smooth muscle can be induced passively by incubation of the tissue with serum from atopic donors (1). This sensitization confers in vitro responsiveness to specific allergen and anti-IgE (1, 2, 5, 6)mimicking the reaction of asthmatic lung tissue (7)and nonspecific hyperresponsiveness to bronchoconstrictor agents such as histamine and neuropeptides (3, 8).

The specific tissue responses to allergen and anti-IgE have been studied extensively and it is widely accepted that these responses are mediated predominantly by sulfidopeptide leukotrienes, and to a lesser extent by histamine (2, 4, 5, 7, 11, 12). The nature of the nonspecific hyperresponsiveness is less well understood, as hyperresponsiveness seems to depend on serum factors other than IgE (13). The dissociation of nonspecific bronchial responsiveness from solely IgE-dependent mechanisms may have clinical correlates in such conditions as nonallergic asthma, chronic obstructive bronchitis, and airway viral infection, in which hyperresponsiveness can be observed in the absence of elevated circulating IgE (14, 15). Thus it is possible that, in passively sensitized human airways, IgE determines the induction of specific responsiveness to allergen while the induction of "nonspecific" responsiveness to histamine depends on other factors present in allergic serum.

We have further investigated the crucial role of IgE in airway bronchoconstriction by sensitizing airways with pure IgE in the absence of other serum factors, and by blocking IgE sensitization with a novel nonanaphylactogenic anti-IgE antibody, 17-9. Pure IgE was used in the form of JW8, a chimeric, humanized, hapten-specific IgE, containing a human  constant region ligated to the murine VH region and the original murine L chain with specificity for the hapten 4-hydroxy-3-iodo-5-nitrophenylacetic acid (NIP) (16); this chimeric IgE antibody can mimic natural allergen-specific IgE reactions in biological systems. The anti-IgE antibody 17-9 (17) binds IgE with high affinity in an isotype-specific way and, owing to its particular epitope specificity, can prevent IgE binding to the Fc RI on basophils and mast cells, thereby blocking IgE-mediated cell responses. In contrast to conventional anti-IgE antibodies, 17-9 does not induce histamine release from IgE-sensitized human basophils and hence is a nonanaphylactogenic anti-IgE antibody. 17-9 has been characterized similarly to other described nonanaphylactogenic antibodies of different clonal origin (18- 20), showing similar properties with a high affinity for IgE.

In the present study, we have used 17-9 to block IgE-dependent mechanisms in tissues sensitized either with high IgE serum or JW8. We demonstrate that neutralization of IgE in sensitizing serum effectively suppresses the specific responsiveness to allergen whereas responsiveness to histamine is unaffected.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Serum Donors

Serum was prepared from whole blood of seven atopic individuals demonstrating high total IgE (> 500 U/ml, up to > 3,000 U/ml) and a range of concentrations of specific IgE for Dermatophagoides farinae (FAST class 1-4: > 0.35 U/ml up to > 17.5 U/ml). Sera were not pooled but were frozen individually in 200- to 250-µl aliquots until used.

Tissue Donors

Macroscopically normal bronchial tissue was obtained from 20 patients undergoing surgery for lung cancer (16 male, 4 female; mean age 61, range 52 to 72 yr). At the time of operation 15 were active smokers and five were ex-smokers, having not smoked for at least 2 yr. Preoperative lung function parameters and medications are given in Table 1. No donors had a history of allergy. Sixteen donors had serum IgE levels in the normal range (0 to 100 U/ml) while four had slightly elevated levels (101 to 238 U/ml); two of these donors had detectable but very low specific anti-D. farinae IgE titers (0.61 to 0.67 U/ml), but this did not render them responsive to allergen in vitro in the absence of passive sensitization (see RESULTS).

Characterization of the Nonanaphylactogenic Anti-IgE Antibody 17-9

From several hybridoma fusions a mouse IgG_2b monoclonal antibody (termed mAb 17-9) has been selected that binds human IgE in an isotype-selective manner and that does not induce histamine release from human basophils and hence is classified as a nonanaphylactogenic anti-IgE antibody, as has been described previously for another antibody (20).

Isotype specificity of 17-9. The isotype specificity of selected antibodies was determined by ELISA (20) using microtiter plates coated with 50 µl of 5 µg/ml of the different purified human myeloma proteins (IgM, Cappel 0001-0870; IgG₁, WHO Hu-0781; IgG₂, WHO G236-0783; IgG₃, WHO Val-1080; IgG₄, WHO Stoe-1184; IgA, WHO Gra-0386; IgE WT [kindly provided by D. Stanworth, Birmingham, UK] and JW8 [16]) for 15 h at 4° C. After washing and quenching wells with phosphate-buffered saline (PBS) containing 1% bovine serum albumin (BSA), 0.05% Tween 20, and 30 µM NaN₃, 100 µl of serial dilutions in a range of 0.1 to 10,000 ng/ml of 17-9 were incubated in the same buffer for 2 h at room temperature and detection was performed with biotinylated rat anti-mouse  antibody R-33-18-12 followed by streptavidin-coupled alkaline phosphatase. The final detection limit was  1 ng/ml on IgE-coated plates.

Determination of histamine release from human basophils. In vitro histamine release from human basophils was determined essentially as described (21). Aliquots of human peripheral blood mononuclear cells containing 10³ basophils were challenged with mAb 17-9 at concentrations of 0.03 to 30 µg/ml and incubation performed in a volume of 150 µl HEPES 25 mM, 0.9% NaCl, 0.5 mM glucose, pH 7.4, for 15 min at 37° C. The monoclonal anaphylactogenic anti-human IgE antibody TN142 (20) served as positive control. Histamine concentration in supernatants (or total content from lysed cells) was determined by enzymatic-isotopic assay. For this purpose 50 µl of supernatants were withdrawn and incubated for 90 min at 37° C with a cocktail of kidney histamine methyltransferase and 0.6 µCi [³H]-S-adenosylmethionine (Amersham International, Amersham, UK). The reaction was terminated by addition of 40 µl 1.5 N HClO₄ and 40 µl 10 N NaOH. The labeled methylhistamine was extracted with 500 µl toluene:isoamylalcohol (80:20 vol/vol) and aliquots of the organic phase were counted in a liquid scintillation counter.

Tissue Sensitization

Immediately after resection, peripheral bronchi (2 to 5 mm internal diameter) were dissected free of alveolar tissue and cut into rings (2 to 4 mm length). Tissues were sensitized by rotating overnight in tubes containing modified Krebs buffer (composition mM: NaCl 118.4, KCl 4.7, MgSO₄ 0.6, CaCl₂ 1.3, KH₂PO₄ 1.2, NaHCO₃ 25.0, D-glucose 11.1) in the absence or presence of either serum from D. farinae-sensitive atopic individuals (10% vol/vol) or the chimeric, NIP hapten-specific IgE, JW8 (10 µg/ml, corresponding to 4,000 U/ml).

To investigate the effect of the anti-IgE antibody 17-9 on the sensitization, tissues from the same individuals were sensitized as above in the absence or presence of purified mouse IgG_2b anti-human IgE antibody, 17-9 (100 µg/ml) or a control mouse antibody of the same isotype (100 µg/ml). As further controls, nonsensitized tissues were similarly treated with 17-9.

Measurement of Bronchial Contraction

After sensitization, rings were transferred to 10 ml organ baths containing oxygenated (95% O₂, 5% CO₂) modified Krebs solution (pH 7.4, 37° C) and tissues were equilibrated for at least 60 min at a resting tension of 250 to 300 mg before the commencement of experimental protocols. All concentration-effect curves were constructed in a cumulative manner, using incremental concentrations spaced at half log₁₀ intervals. At the end of experiments, tissues were exposed to a single concentration of histamine (10 µM) to ensure that negative responses to allergen or hapten-carrier complex were not the result of deterioration of contractile responses in the tissues.

Cumulative concentration-effect curves to histamine (0.01 to 300 µM) were constructed in all tissues after equilibration for at least 30 min, two histamine priming concentrations (10 µM) to determine tissue viability and to enable standardization of histamine maxima, and a further equilibration period of 30 min. After completion of histamine concentration-effect curves, tissues were washed and reequilibrated for 60 min before construction of cumulative concentration-effect curves to D. farinae (0.03 to 30 U/ml) or NIP-BSA (0.001 to 1 µg/ml). In experiments investigating the action of a control IgG_2b antibody, responses to a single concentration of allergen (10 U/ml) or NIP-BSA (0.3 µg/ml) were assessed.

To determine whether 17-9 exerted an effect on the interaction between sensitized tissues and allergen or hapten-carrier complex, tissues were sensitized normally with serum or JW8 in the absence of 17-9 and histamine concentration-response curves were constructed as described previously. Tissues were reequilibrated for 60 min and then incubated for a further 60 min in the absence or presence of 17-9 (100 µg/ml) prior to challenge with D. farinae (10 U/ml) or NIP-BSA (0.3 µg/ml), as appropriate.

To determine the influence of sensitization on bronchial responses to histamine, tissues were sensitized with serum or JW8, in the absence or presence of 17-9, as described earlier, and histamine concentration-response curves were constructed, after which effective sensitization was confirmed by assessing the response to a high concentration of D. farinae (30 U/ml) or NIP-BSA (1 µg/ml), as appropriate.

Measurement and Analysis of Data

All responses were recorded as changes in isometric tension (mg). The potency of histamine was determined by an iterative curve-fitting procedure (Kaleidagraph; Synergy Software, Reading, PA) for each individual tissue and is expressed as the pEC₅₀ value (i.e., log₁₀ of the concentration of histamine giving a half-maximal effect). Statistical analysis of the data was performed using unpaired Student t tests. Analysis of variance was performed where appropriate and, when significance was found, post hoc pairwise comparisons between groups were performed by Newman-Keuls multiple comparisons test. A value of p < 0.05 was defined as significant.

Materials

Histamine dihydrochloride was obtained from Sigma Chemie GmbH (Deisenhofen, Germany). D. farinae allergen was purchased from Allergopharm KG (Reinbek, Germany). JW8, NIP-BSA, 17-9, and purified mouse IgG_2b control antibody were generated at Novartis AG, Basel, Switzerland. All drug solutions were prepared using distilled water, with the exception of allergen and antibody solutions which were prepared using 0.9% saline.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

17-9 Is a Specific, Nonanaphylactogenic Anti-IgE Antibody

The mouse IgG_2b anti-human IgE mAb 17-9 bound specifically to human IgE while exhibiting cross-reactivity < 0.1% against other human immunoglobulin classes (Figure 1a). At concentrations as low as 10 ng/ml, 17-9 resulted in a clear detectable binding to IgE whereas a 1,000-fold excess of 17-9 did not give detectable binding to other human Ig isotypes. 17-9 was also analyzed for its ability to induce histamine release from human basophils. In contrast to a conventional anti-IgE antibody, such as TN142, 17-9 did not induce significant histamine release from human basophils (Figure 1b) even at a high concentration of 30 µg/ml, indicating that it is a nonanaphylactogenic antibody.

View larger version (17K): [in this window] [in a new window]   Figure 1.   Characterization of 17-9 as a nonanaphylactogenic anti-IgE antibody with respect to specific IgE binding and absence of induction of histamine release. (a) The binding of 17-9 at the indicated concentrations to plates coated with human myeloma IgE (closed circles) or other myeloma immunoglobulins isotypes as indicated (open symbols) and as specified in METHODS. (b) Specific histamine release from purified human blood basophils during incubation for 15 min at 37° C with the indicated concentrations of anti-IgE antibody 17-9 (closed circles) or a conventional anaphylactogenic IgE antibody, TN142 (open squares).

Anti-IgE Antibody 17-9 Prevents Responses to Allergen in Tissues Sensitized with High IgE Serum

In tissues sensitized with serum containing high specific IgE for D. farinae, there was a concentration-related contraction to D. farinae that was not observed in nonsensitized controls (Figure 2a). Tissues sensitized in the presence of 100 µg/ml 17-9 did not respond to D. farinae at concentrations up to 30 U/ml.

View larger version (20K): [in this window] [in a new window]   Figure 2.   Contractile concentration-effect curves to allergen and specific antigen in human bronchus sensitized with serum containing high total IgE and high (fluoro-allergosorbent test [FAST] class > 3) D. farinae-specific IgE (a, n = 8) or with 10 mg/ml NIP-specific chimeric IgE JW8 (b, n = 5). Responses to D. farinae (a) and hapten-carrier complex NIP-BSA (b) in nonsensitized (open squares) and sensitized tissues (open circles), and tissues sensitized in the presence of 100 µg/ml 17-9 (closed circles). Contractions are expressed as percent histamine maximum and are the mean ± SEM of the indicated numbers of experiments using tissue derived from different individuals.

Anti-IgE Antibody 17-9 Prevents Responses to Antigen in Tissues Sensitized with JW8

In tissues sensitized with the chimeric IgE, JW8, there was a concentration-related contraction to the specific antigen, NIP-BSA, that was not observed in nonsensitized control tissues (Figure 2b). Tissues sensitized in the presence of 100 µg/ml 17-9 did not respond to NIP-BSA at concentrations up to 1 µg/ml.

Non-IgE-specific IgG_2b Antibody Does Not Prevent Responses to Allergen and Antigen in Tissues Sensitized with High IgE Serum or JW8

While the inclusion of 100 µg/ml 17-9 effectively blocked development of D. farinae responsiveness in bronchial rings sensitized with high-IgE serum, tissues sensitized in the absence or presence of a control antibodyof the same IgG_2b isotype but directed against an irrelevant antigenexhibited no significant difference in their responses to allergen or specific antigen (Figure 3, left). Similarly, there was no difference between contractions induced by NIP-BSA in tissues sensitized with JW8 in the absence or presence of control mouse IgG_2b antibody (Figure 3, right).

View larger version (26K): [in this window] [in a new window]   Figure 3.   Lack of effect of non-lgE-specific control antibody on sensitization with atopic serum or chimeric IgE. Contractile responses to 10 U/ml D. farinae or 0.3 µg/ml NIP-BSA, respectively (open bars), in tissues sensitized with 10% high IgE serum (left, n = 12) or 10 µg/ml JW8 (right, n = 9). Solid bars represent responses in tissues sensitized in the presence of 17-9 (n = 8 for serum, n = 5 for JW8); shaded bars represent contractions in tissues that were sensitized in the presence of a control, non-IgE-specific IgG2b antibody (n = 4). Contractions are expressed as percent of histamine maxima and are the mean ± SEM of the indicated numbers of experiments using tissue derived from different individuals. *p < 0.05 compared with sensitized tissues.

Effect of 17-9 Added after Sensitization

ln tissues sensitized with high IgE serum or JW8, contractile responses to 10 U/ml D. farinae or 0.3 µg/ml NIP-BSA, respectively, were observed. Addition of 17-9 (100 µg/ml) 60 min before in vitro challenge with allergen had no significant effect on the magnitude of these responses. 17-9 itself caused no contraction of sensitized tissues (Figure 4).

View larger version (15K): [in this window] [in a new window]   Figure 4.   Lack of effect of 17-9 added after sensitization. Contractile responses to 10 U/ml D. farinae or 0.3 µg/ml NIP-BSA, respectively (open bars), in tissues sensitized with 10% high IgE serum (left) or 10 mg/ml JW8 (right). Solid bars represent contractions in sensitized tissues that were challenged with allergen 60 min after the addition of 17-9 (100 ng/ml) to the organ bath. Contractions are expressed as percent of histamine maxima. Data are the mean ± SEM of three experiments using tissue derived from different individuals.

17-9 Does Not Alter Histamine Responsiveness in Sensitized Human Bronchus

Histamine produced concentration-dependent contractions in all tissues. Sensitization with high-IgE serum led to an increase in the potency of and the magnitude of maximal contractions to histamine, as described previously (13), with pEC₅₀ increasing from 4.9 ± 0.2 in nonsensitized bronchial rings to 5.3 ± 0.2 in serum-sensitized tissues (n = 15, p < 0.05). In eight experiments in which tissues were sensitized in the absence or presence of 17-9, the anti-IgE antibody had no effect on either the potency or the maximal response to histamine (Figure 5). Similarly, in JW8-sensitized tissues, which exhibited no difference in histamine responsiveness from nonsensitized control tissues (13), the presence of 17-9 during sensitization had no effect on responses to histamine (not shown).

View larger version (23K): [in this window] [in a new window]   Figure 5.   Contractile concentration-effect curves to histamine in nonsensitized (open squares), high-IgE serum-sensitized control tissues (open circles), and tissues sensitized in the presence of 100 µg/ ml 17-9 (closed circles). Contractions are expressed as mg tension changes and represent mean ± SEM from eight experiments using tissue derived from different individuals.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The present study demonstrates that the nonanaphylactogenic, monoclonal anti-IgE antibody 17-9 prevents the IgE-dependent passive sensitization of human airways in vitro. This is observed in tissues sensitized with either allergen-specific high-IgE titer homologous serum or with a purified hapten-specific, chimeric IgE (JW8). This effect is mediated specifically through the interaction of 17-9 with IgE, because a control antibody of the same IgG_2b class directed against an irrelevant antigen did not show any inhibition. While 17-9 was effective when present during sensitization, it failed to inhibit allergen responses when applied to presensitized tissues 60 min prior to allergen stimulation. This suggests that 17-9 blocks IgE in the sensitizing serum/buffer from binding but that, under these experimental conditions, 17-9 is unable to displace bound IgE from its receptors on the tissue. This is in contrast to experiments in which 17-9 was shown to be able to displace IgE from human basophils in vitro, as assessed by its action on IgE-mediated histamine release (C. H. Heusser, unpublished data). This inhibition was dependent on the concentration of 17-9 and the time of incubation and has been attributed to the capture of IgE during its off-on interaction with Fc RI.

Similar studies have shown that anti-IgE and a chimeric construct of Fc RI-IgG were able to prevent sensitization of tissue for allergen-induced bronchial tissue contraction (22). Those studies were based on sensitization with atopic serum and do not preclude participation of other serum factors besides IgE in the sensitization process. Our studies with purified IgE extend these investigations and demonstrate the effect of the nonanaphylactogenic anti-IgE antibody 17-9 to be accounted for purely by prevention of IgE sensitization of tissues.

In contrast to its effects on allergen responses, 17-9 had no effect on responses to histamine in any tissues and did not prevent the induction of histamine hyperresponsiveness in tissues sensitized with IgE-containing serum. We have previously demonstrated that serum sensitization increases airway responses to both allergen and histamine while sensitization with JW8 induces responsiveness to the specific hapten-carrier complex (NIP-BSA) but does not affect responsiveness to histamine (13). Thus, we have speculated that, under these in vitro conditions, the induction of nonspecific hyperresponsiveness is related to serum factors other than IgE while the specific allergen or hapten-carrier responses are entirely dependent on IgE. The findings of the present study support this notion, since neutralization of IgE eliminates only allergen responses and has no effect on histamine responsiveness in sensitized bronchial rings.

Because IgE-mediated responses are central to the pathology of atopic disease, the inhibition of IgE binding to its high-affinity receptors has been suggested as a possible therapeutic intervention in allergic airway diseases such as bronchial asthma (22). Although IgE fragments have been shown to inhibit the process of passive sensitization, excessively high concentrations are required to produce this effect (23). Anti-IgE monoclonal antibodies (mAb) have been developed as an alternative tool but it has been difficult to produce mAb that inactivate IgE in solution without also binding and cross-linking receptor-bound IgE (24). Recently, antibodies such as MaE11 (19), CGP51901 (20), BSW17 (24), and 17-9 (17) have been developed which bind soluble IgE without cross-linking the Fc RI and RII receptor-bound IgE on effector cells, thereby allowing inhibition of IgE-dependent sensitization without provoking anaphylactic reactions. The inability to induce cross-linking of bound IgE has been shown to be due to an inability to recognize receptor-associated IgE (18), as the corresponding epitope presumably is located within the IgE region recognized by the Fc RI receptor. However, in one case a nonanaphylactogenic anti-IgE has been shown to interact with receptor-bound IgE and to provoke the release of IgE from the receptor (25), presumably by inducing a conformational alteration of IgE. Thus different mechanisms may contribute to the nonanaphylactogenic property of anti-IgE antibodies.

Nonanaphylactogenic antibodies clearly can prevent tissue sensitization by IgE and thus are highly attractive therapeutic principles. However, two possible limitations are indicated by the present study. First, application of 17-9 to IgE-sensitized tissues for 60 min prior to allergen provocation and during the administration of allergen produced no significant attenuation of the allergen-induced smooth muscle contraction. Secondly, although 17-9 abolished specific allergen responses conferred by IgE-sensitization of airways, it had no effect on the increase in "nonspecific" responsiveness induced by atopic serum, as illustrated by the enhanced responses to histamine. These potential restrictions may, however, not be relevant in vivo when nonanaphylactogenic anti-IgE antibodies can act for prolonged periods of time. It has been shown that treatment of humans with a humanized nonanaphylactogenic anti-IgE antibody results in an effective, dose-dependent reduction of serum IgE levels (26). The IgE level, in turn, seems to regulate the degree of Fc RI receptor expression as concluded from a study in which patients treated with humanized nonanaphylactogenic anti-IgE antibody for several weeks showed a 96% reduction in basophil Fc RI expression associated with a markedly decreased response of these cells to IgE-mediated triggering ex vivo (27). Thus, as has been shown by studies in animals (28), it is likely that such anti-IgE antibodies may prevent allergic reactions in sensitized individuals. It has been demonstrated that a corresponding nonanaphylactogenic antibody to murine IgE was able to inhibit antigen-induced skin reactions and bronchoconstriction in sensitized mice (29). Moreover, anti-IgE also caused a marked inhibition of pulmonary eosinophil infiltration induced by allergen inhalation in these animals (30). This latter effect of anti-IgE has been shown to be mediated by blockade of IgE interaction with CD23, the low-affinity receptor for IgE (Fc RII) (30). Thus, it is likely that nonanaphylactogenic anti-IgE antibodies could also be effective in allergic asthma. On the other hand, it is not expected that these antibodies would reverse the nonspecific serum factor- mediated airway hyperresponsiveness to histamine.

The bronchial tissue used in these experiments was obtained from surgical resection material and it should be borne in mind that the functional responses of the tissue may be influenced either by the donors' disease or by their smoking history. Airways were in all cases dissected from areas of lung that were macroscopically normal and some distance removed from the tumor. However, the possibility that airways in the lungs of smokers react differently to those from nonsmokers and that airway function may be influenced by cancer existing elsewhere in the same lobe cannot be discounted. In addition, care should be taken when extrapolating the results from our in vitro experiments to the in vivo situation, since the effects of the nonanaphylactogenic anti-IgE antibody on cells and tissues outside the bronchus may influence its in vivo actions.

In conclusion, a novel monoclonal anti-human-IgE antibody, 17-9, blocks passive IgE sensitization of human airways by either atopic serum or by a chimeric anti-NIP IgE without affecting the responses of the tissues to histamine. This confirms that the specific responses of passively sensitized human airways to allergens are dependent on IgE whereas increased responsiveness to histamine is mediated by a component of atopic serum other than IgE.