Lung surfactant proteins A and D can inhibit specific IgE binding to the allergens of Aspergillus fumigatus and block allergen-induced histamine release from human basophils

Clin Exp Immunol 1997; 110:241-249

Lung surfactant proteins A and D can inhibit specific IgE binding to the allergens of Aspergillus fumigatus and block allergen-induced histamine release from human

basophils

T. MADAN, U. KISHORES, A. SHAH*, P. EGGLETON$, P. STRONG$, J. Y. WANGS, S. S. AGGRAWALt, P. U. SARMA & K. B. M . REID$ CSIR Centre for Biochemical Technology, *V.P. Chest Institute and ?College of Pharmacy, Delhi, India, $Medical Research Council lmmunochemistry Unit, Department of Biochemistry, University of Oxford, Oxford, UK, and

$Department of Paediatrics, National Chen-Kung University, Tainan, Taiwan

(Accepted for publication 9 August 1997)

SUMMARY

Aspergillus fumigatus is an opportunistic fungal pathogen which, in the immunocompetent host, causes allergic disorders such as allergic rhinitis, allergic sinusitis, hypersensitivity pneumonitis, and allergic bronchopulmonary Aspergillosis (ABPA). In the present study, the interaction of 3-week culture filtrate (3wcf) allergens and various purified glycosylated and non-glycosylated allergens of A. fumigatus with lung surfactant proteins, SP-A and SP-D, was investigated. Purified SP-A and SP-D, isolated from human bronchoalveolar lavage fluid, bound to the 3wcf allergens and purified allergens, gp55 and gp45, in a carbohydrate-specific and calcium-dependent manner. Both SP-A and SP-D did not bind to deglycosylated allergens, suggesting that the ability of SP-A and SP-D to bind certain allergens is mediated through their carbohydrate recognition domains, interacting with the carbohydrate residues on the allergen. Both SP-A and SP-D could inhibit the ability of allergen-specific IgE from Aspergillosis patients to bind these allergens, suggesting that SP-A and SP-D may be involved in the modulation of allergic sensitization and/or development of allergic reactions. The view that SP-A and SP-D play a protective role against airborne allergens is further supported by the demonstration of their ability to inhibit A. fumigatus allergen-induced histamine release from allergic patients’ basophils.

Keywords surfactant protein A surfactant protein D Aspergillus fumigatus IgE allergen

INTRODUCTION inflammatory response of the respiratory tract and surrounding

Aspergillus fumigatus is a major airborne fungus causing a wide spectrum of clinical conditions (allergic and invasive) depending on the host’s immune status [1,2]. In the immunocompetent host, this opportunistic pathogen causes allergic disorders, such as allergic rhinitis, allergic sinusitis, hypersensitivity pneumonitis, and allergic bronchopulmonary aspergillosis (ABPA). In the absence of adequate host defence, A. fumigatus disseminates to other tissues, such as bone, skin and viscera, resulting in a fatal systemic disease, called invasive aspergillosis (IA) [3]. The normal defence mechanisms of the host act as a formidable barrier to the inhaled spores and result in their elimination. However, some individuals with pre-existing atopic disease may develop ABPA on exposure to A. fumigatus. ABPA results from colonization of the respiratory tract with the fungi, the development of an intense

pulmonary parenchyma in the acute phase, destruction and fibrosis of the inflamed structures in the chronic phase. ABPA can cause complications in up to 30% of asthmatics [4] and most cystic fibrosis patients [5 ] . There is an increasing incidence of pulmonary aspergillosis in HIV patients [6,7].

Treatment for the invasive form of aspergillosis, to provide complete cure, is complicated by the chemotherapy of allergic patients with steroids and of chronically infected patients with aggressive antifungal drugs, which have highly adverse side effects, such as hepatotoxicity and nephrotoxicity. The immune mechanisms involved in the disease are not fully understood, although in the initial stages of A. fumigatus infections, both humoral and cell-mediated immunity have shown little, if any, protective roles [8]. Therefore, innate immunity, mediated by polymorphonuclear leucocytes and macrophages, forms the main line of defence against A. fumigatus infections [9], where the host can call upon not only resident pulmonary alveolar macrophages, but also neutrophils and monocytes recruited from the circulation.

Correspondence: Professor Kenneth B.M. Reid, MRC Irnmunochem- istry Unit, Department of Biochemistry, South Parks Road, Oxford OX1 3QU, UK.

0 1997 Blackwell Science 24 I

242 T. Madan et al.

The mechanisms by which macrophages and neutrophils kill Aspergillus organisms are not completely understood, but several lines of evidence suggest that oxidative mechanisms play an important role [10,1 I]. The demonstration that patients with defective oxidative mechanisms (e.g. chronic granulomatous disease (CGD)) are susceptible to infections with Aspergillus [ 121, and the fact that macrophages and neutrophils from CGD patients show reduced oxidative killing of Aspergillus [ 1 11, support the role of oxidative killing mechanisms. The infectious conidia of A. fumigatus are often resistant to reactive oxygen intermediates generated during stimulation of phagocytes [13]. The role of opsonins in the defence against Aspergillus is controversial. Washburn el al. demonstrated that phagocytosis of conidia was enhanced by opsonization by C3b [ 1 I]. However. Kurup failed to demonstrate enhanced phagocytosis by opsonizing spores with either complement or immune serum in the rabbit [14]. It is also speculated that binding of Aspergillus conidia to macrophages prior to phagocytosis can occur directly through ligand-receptor interaction [lo]. There has, therefore, been an intensive search for molecules present in the lung that can selectively enhance the contribution of innate immune mechanisms of phagocytes against such pathogens.

There is increasing evidence to suggest that two of the surfactant proteins may have protective roles to play in pulmonary defence against pathogens [15]. Lung surfactant proteins A (SP-A) and D (SP-D) are large oligomeric structures, each assembled from multiple copies of a single polypeptide chain, and belong to a family of C-type lectins, called collectins. Their primary structure is organized into four distinct regions: an amino terminal region involved in the formation of interchain disulfide bonds, a collagenous region composed of Gly-Xaa-Yaa repeats. an 0-helical trimerizing neck peptide, and a carboxy terminal, C-type, carbohydrate recognition domain, designated as a CRD. The C-type CRDs are spaced, in a trimeric orientation, at the end of triple helical collagenous stalks. Collectins play a role in innate immunity by virtue of their recognizing carbohydrate targets on pathogens. Recently, we have demonstrated the involvement of SP-A and SP-D in the initial protective immunity against A. fumigatus infections [16]. Both SP-A and SP-D bound to carbohydrate structures on A. ,fumigatus conidia in a calcium-dependent manner, causing agglutination of the conidia and enhancing the binding of conidia to alveolar macrophages and neutrophils. Furthermore, in the presence of SP-A and SP-D, phagocytosis, oxidative burst and killing of conidia were significantly increased by neutrophils. These findings strongly suggest that SP-A and SP- D may have an important immunological role in the early antifungal defence in the lung, through inhibiting infectivity of conidia by agglutination and by enhancing uptake and killing of A. fumigatus by phagocytes.

Aspergillus-specific cellular and humoral immune responses do not appear to play an important role in host defence [ 171. However, antigen-specific antibody and T cell responses are found in many of the pulmonary Aspergillosis syndromes. These responses are believed to play an important role in the pathogenesis of diseases in which hypersensitivity reactions predominate (e.g. ABPA and hypersensitive pneumonitis). Having established antifungal prop- erties of SP-A and SP-D, we sought to investigate whether these surfactant proteins would interact with carbohydrates on the inhaled allergens of A. jiumigatus and modulate type I and type TI1 hypersensitivity reactions. SP-A is known to bind to water- extractable. allergenic glycoproteins from pollen grains in a

calcium-dependent manner, suggesting that the interaction of surfactant proteins with certain allergens may have some role in immediate type hypersensitivity [ 181. Furthermore, SP-A and SP- D can inhibit the binding of allergen-specific IgE to house dust mite extracts [ 191, which is suggestive of a possible involvement of these lung surfactant proteins in the allergen sensitization and/or the development of allergic reactions. In the present study, we examined the interaction of the 3-week culture filtrate (3 wcf) allergens, and various purified glycosylated, enzymatically deglycosylated, and naturally occumng, non-glycosylated allergens, of A. fumigatus with SP-A and SP-D. Purified SP-A and SP-D, isolated from human bronchoalveolar lavage fluid (BALF), bound to allergens of 3 wcf as well as purified allergens, gp55 and gp45, in a carbohydrate-specific and calcium-dependent manner. Binding was inhibited by EDTA as well as by maltose in the case of SP-D, or mannose in the case of SP-A. A self- associating, trimeric recombinant polypeptide, composed of the neck and CRD region of human SP-D [20], also bound to these allergens in a manner similar to that shown by the native surfactant proteins. However, SP-A and SP-D failed to bind to deglycosylated allergens, suggesting that their ability to bind certain allergens is mediated through their CRDs interacting with carbohydrate residues on the allergens. Both SP-A and SP-D inhibited the ability of allergen-specific IgE from Aspergillosis patients to bind these allergens. The possible protective roles played by SP-A and SP- D against airborne allergens are further supported by their ability to block A. fumigatus allergen-induced histamine release from allergic patients’ basophils.

MATERIALS AND METHODS

Strains, human sera, antibodies to SP-A and SP-D Spores and culture filtrate were obtained from the 285 strain of A. fumigatus isolated from the sputum of a patient suffering from ABPA. Cultures were maintained on Saboraud dextrose agar slants. Escherichia coli BL21 was used for expression of a recombinant fragment of SP-D. Sera from patients suffering from allergic bronchopulmonary aspergillosis (ABPA, based on Rosenberg criteria) were supplied by the V. P. Chest Institute (Delhi, India). The serological response against Aspergillus antigens was tested by immunodiffusion, showing precipitating arcs against 3 wcf. Control sera were obtained from consenting healthy donors without any indication of pulmonary disease. No antibodies against Aspergillus antigens were detected in these sera, using the same immunodiffusion tests.The anti-SP-A and anti-SP-D antisera were raised in rabbits against human SP-A and SP-D which had been purified from bronchoalveolar lavage. Antiserum was also raised, in rabbits, against a recombinant fragment of human SP-D. These antibodies were specific for SP-D and SP-A at the working dilutions, as assessed by ELISA and Western blotting.

Purification of SP-A, SP-D and a recombinant fragment of SP-D Native human SP-A and SP-D were purified using a novel method (P. Strong, U. Kishore, K.B.M. Reid, manuscript in preparation) from B ALF obtained from alveolar proteinosis patients. In brief, the fluid was centrifuged at 2000 g to remove aggregates and the resulting supernatant recalcified by the addition of 35 mM CaC12 and the pH readjusted to 7.4. Maltose agarose (Sigma, Poole, UK) was packed into a column, washed and then equilibrated with buffer containing 2 0 m ~ Tris-HC1 pH 7.4, l O m ~ CaC12, and 0.02% (w/v) sodium azide. The SP-A- and SP-D-enriched

0 1997 Blackwell Science Ltd, Clinical and Experimental Immunology, 110:241-249

supernatant was applied to the maltose agarose column. SP-D was eluted with 2 0 m ~ Tris pH 7.4, 1OOm MnC12. The SP-A was eluted with an EDTA gradient (10-50 m). Final purification was achieved by gel filtration on a Superose 6 column. Both preparations were judged to be pure by SDS-PAGE and Western blotting. A recombinant polypeptide, comprising the neck region and CRD of human SP-D, and designated rSP-D (N/CRD), was expressed in E. coli as a fusion protein linked to the E. coli maltose-binding protein. After affinity purification and removal of the maltose- binding protein, the rSP-D(N/CRD) was isolated in the form of self-associating, homotrimer [20]. Purified SP-A, SP-D and rSP- D(N/CRD) were biotinylated by incubating 0.6ml of 1 mg/ml solution of protein in 0.1 M NaHC03 with 34.2 pl of 1 0 m N- hydroxyl succinimidobiotin (BNHS) (Pierce, Rockford, IL,) in 0 . 1 ~ NaHC03 for 4 h at room temperature. Unreacted sulfo- NHS-biotin was removed by dialysis against 10 mM sodium phosphate, 1 5 0 m NaCl pH 7.4, and any insoluble material was removed by centrifugation in a microfuge before use.

Preparation of A. fumigatus allergens/antigens from culture filtrate and deglycosylation of A. fumigatus allergendantigens Three-week culture filtrate allergens/antigens (3 wcf) of 285 strain of A. fumigatus were prepared from an isolate from the sputum of an ABPA patient. The 3 wcf was selected for the studies as it contains both the cell wall and the secretory allergendantigens. The secretory allergendantigens gp45, gp55, and Asp fI were purified from the 3 wcf allergendantigens as reported previously [21,22] (Fig. 1). Deglycosylation of purified allergendantigens gp45 and gp55 was carried out as described earlier [22]. Briefly, the purified glycoprotein (20 pg) was incubated with 1 U of N- glycanase-F (Boehringer, Mannheim, Germany) in a total volume of 1OOpl of 0.1 M sodium phosphate buffer pH 8.2, containing 2 5 m ~ EDTA, 1% (v/v) nonide P-40, 0.2% (wlv) SDS, and 1% (v/v) P-mercaptoethanol. The deglycosylated allergendantigens were extensively dialysed against 0.01 M phosphate buffer pH 7.4.

A 6 C

SP-A and SP-D block IgE binding to A. fumigatus allergens 243


Fig. 1. SDS-PAGE of 3-week culture filtrate of Aspergillusfumigatus (B); purified preparations of gp55 (C), gp45 (D), 33-kD elastase (E) and 18-kD non-glycosylated allergen (F). Lane A shows molecular weight markers (kD; 112, 84, 63, 52, 35, 32).

The native preparations of gp45 and gp55 were treated with the same buffer without N-glycanase - F, and used as appropriate controls, following extensive dialysis. The deglycosylated allergendantigens were analysed by SDS-PAGE and by Western blotting for binding to specific anti-allergen IgE, using sera from ABPA patients as a source of IgE antibody (data not shown).

Interaction of SP-A and SP-D with 3 wcf allergens/antigens and puriJied allergens Microtitre plates (Polysorp; Nunc, Kamstrup, Denmark) were coated with dilutions of 3 wcf allergendantigens or purified native and deglycosylated allergens made in 100 pl coating buffer ( l o o m Na2C03 pH 9.3) at 4°C overnight. Non-specific sites were blocked with bovine serum albumin (BSA) (200 pVwell; 3 mg/ml) in loading buffer (10 m~ Tris-HC1,150 mM NaCl pH 7.4, 0.05% (w/v) Tween 20). After extensive washing with loading buffer, the plate was incubated with serial dilutions of biotinylated SP-A, SP-D or rSP-D (100pl; maximal amount 1 pglwell) in washing buffer containing 5 m CaC12 for 3 h at 37°C. For inhibition studies, 100 pl (1 pg/well) samples of biotinylated proteins were added in presence of 10 mM EDTA, or 100 m~ mannose, or 1 O O m maltose. All incubation steps were carried out at 37°C for 3 h. After washing with the same buffer, bound biotinylated SP- A, SP-D, or rSP-D(N/CRD) were detected by the use of strepta- vidin peroxidase conjugate (1 : 5000) (Sigma) and TMB-peroxidase enzyme immunoassay (EIA) kit (BioRad, Hercules, CA). The reaction was stopped by addition of 2 N H2S04 and the plate was read at 450 nm.

Ligand blot analysis of 3 wcf allergens/antigens using biotinylated SP-A and SP-D or A. fumigatus-specific IgE antibody Ligand blot analysis of 3 wcf allergendantigens and purified native and deglycosylated allergens was camed out with native SP-A and SP-D and rSP-D(N/CRD) in order to identify the allergens bound by lung surfactant proteins. Samples were run on SDS-PAGE after reduction of disulfide bonds, and then transferred to a nitrocellu- lose membrane (Hybond C; Amersham, Aylesbury, UK). The non- specific sites were blocked with BSA (3mg/ml) in 1 0 m Tris- HCl buffer pH 7.4 containing 150 m~ NaC1. The membrane was incubated with biotinylated SP-A or SP-D (10 ml, 1 pg/ml) in incubation buffer (10 mM Tris-HC1 pH 7.4, 150 m NaCl containing 0.1% (v/v) Tween 20 and 5 m ~ CaCI2) overnight at 4°C. Following extensive washing with incubation buffer, the blots were incubated with extravidine peroxidase conjugate (1 5000; Sigma) for 2 h at room temperature in the above buffer. Following three washes with the same buffer (5 midwash), diaminobenzidine and hydrogen peroxide (DAB/H202; Sigma) were used as the substrates. To detect allergen-specific IgE binding, the membrane was incubated with a dilution of pooled sera (150 dilution in l O m ~ Tris-HC1, 1 5 0 m NaCl pH 7.4, containing 0.1% (v/v) Tween 20 and 5 m CaCI2) from ABPA patients. Sera from normal human donors were used as a control. After extensive washing, the membrane was incubated with a mouse monoclonal anti-human IgE conjugated with peroxidase (1:lOOO; Sigma) for 2 h at room temperature and colour development was camed out as above.

Maltose-binding protein and rSP-D(N/CRD) fusion affinity column chromatography The recombinant MBP-rSP-D(N/CRD) fusion protein (200 pg) was applied to an amylose resin column matrix (1 ml) in loading buffer containing 2 0 m ~ Tris-HC1 pH 8.0, 1 0 0 m ~ NaC1, 5 m ~

T. Madan et al.

CaC12, 5% (v/v) glycerol. The MBP-rSP-D(N/CRD)-resin was mixed with 1OOpg of 3 wcf allergens of A . fumigatus at 4°C for I h, poured into a column, and then extensively washed with the loading buffer. The interacting glycoproteins of extract were eluted from the column using 1 ml of buffer containing 20 m~ Tris-HC1 pH 8.0, 1OOm NaCl, lOmM EDTA, 5% (v/v) glycerol. The eluent was loaded on an 10% (w/v) SDS-PAGE and subsequently probed with pooled sera from ABPA patients to detemine if any of the glycoproteins recognized by the rSP-D(N/CRD) were also recognized by IgE antibody. A control column composed of maltose- binding protein (MBP) linked to amylose resin was also used in the study. It should be noted that the rSP-D(N/CRD), on its own, has no affinity for the amylose resin column.

Inhibition by SP-A and SP-D of the binding of sera-conraining allergen-specijic IgE antibody to 3 wcf allergens/antigens The ability of SP-A and SP-D to inhibit allergen-specific IgE binding to 3 wcf allergendantigens or purified glycosylated and deglycosylated allergens was tested. Various concentrations of native human SP-A, SP-D and rSP-D(N/CRD) were incubated with pooled sera (diluted 1:100 with buffer containing 5 m ~ CaCl?), from six ABPA patients for 30min at 37"C, then added to microtitre wells, coated with 3 wcf allergendantigens or purified glycosylated or deglycosylated allergens and incubated for 2 h at 37°C. After extensive washing with buffer, peroxidase conjugated with monoclonal mouse anti-human IgE (Sigma) was added to the wells for 1 h at 37°C. Bound human IgE was detected by colour development with TMB-peroxidase EIA kit (BioRad) and percen- tage decrease in IgE-dependent binding was estimated from the reduction in the optical density (OD) at 450nm observed on addition of SP-A or SP-D.

Histamine release by A. fumigatus-stimulated basophils, in the presence and absence of SP-A und SP-D The effect of SP-A and SP-D on A. fumigatus allergen-mediated histamine release was quantified using sensitized basophils (basophils from ABPA patients based on Rosenberg criteria) and non- sensitized basophils (basophils from a normal subject with no indication of pulmonary disease). Whole blood (1-ml aliquots)

from each subject was collected in heparinized tubes (10 U/ml). For the assay of total histamine, 50p1 of the whole blood were diluted to 100Op1 with distilled water (1:20 dilution). The diluted blood was frozen and thawed twice. The resulting lysed cell suspension was acylated using reagents provided with the histamine immunoassay kit (Immunotech, Marseilles, France). The remaining blood was diluted 1:7 with histamine release buffer provided with the kit. Fifty microlitres of each allergen preparation (0.78-12.5 pg/ml of culture filtrate, 12 pg/ml of purified gp45, gp55, deglycosylated gp45, deglycosylated gp55, with or without 50 p1 of 5 pg/ml SP-A or SP-D) were then added to the wells of a round-bottomed microtitre plate and incubated for 30 min at 37°C. One hundred microlitres of diluted blood (1:7) were added to each well and the covered plate was incubated for 30 min at 37°C. The plate was centrifuged for 5min at 900 g at 4°C. Then 100-pl aliquots of supernatant fluid were collected without disturbing the cell pellet and were acylated using the reagents provided by the histamine assay kit. Next, 5O-pI volumes of the acylated standards (histamine standards with concentrations of 5 0 n ~ , 1 5 n ~ , 5 nM, 1 nM and 0.1 nM) or samples (5 p1 diluted to 5Opl with distilled water) and 200 pl of enzyme conjugate were added to each well of the ELISA plate which had been precoated with an anti-histamine MoAb. One well was left empty for blank correction. The plate was incubated for 18 h at 4°C. The plate was then washed three times in order to remove the unbound conjugate, and 2 0 0 4 of substrate were added to each well including the blank, and the plate was incubated in the dark for 20min. The enzymatic reaction was terminated with 50pl of the stop solution and the plate read at 405 nm against a substrate blank. A standard curve was constructed with absorbance of histamine standard plotted versus concentration in nM. The corresponding histamine concentration present in each sample was calculated from the standard curve.

RESULTS

The glycoproteins in A. fumigatus extract which are recognized by SP-A and SP-D are allergenic To assess the carbohydrate specificity of SP-A and SP-D, competi- tion assays were performed in the presence or absence of 100 mM

Table 1. Binding of surfactant proteins to Aspergillusfumigatus extract in the presence of calcium, sugars and EDTA

Binding conditions SP-A SP-D rSP-D(N/CRD)

(Absorbance at 450 nm)

5 mM CaCI2 1.39 ir 0.09 1.52 2 0.08 0.64 '-c 0.05 100 mM mannosd5 mM CaCI, 0 3 7 t 0.10 0.97 ir 0.04 0.43 ir 0.04 I O O ~ M maltoseA mM CaCI2 0.94 2 0.05 0.28 2 006 0 1 8 ? 0.07 5 miM EDTA 0.14 t- 0.03 0.09 '-c 0.02 0.07 -C 0.03

Solid-phase binding of single concentrations (1 gg/ml) of lung surfactant protein A (SP-A), lung surfactant protein D (SP-D) and recombinant rSP-D(N/CRD) to immobilized .A.fumigarus extract (1 gg/well) was tested in the presence of 5 mM CaCI,, 100 mM mannose with 5 mM CaCIZ, 100 mM maltose with 5 mM CaCI2, or in the presence of EDTA, as described in Materials and Methods. Results are the mean? s.d. of three separate experiments. I t should be noted that the native SP-A and SP-D are oligomeric molecules composed of hexamers and tetramers, respectively, of subunits containing three C-type lectin domains, while the rSP-D(N/CRD) is essentially a single subunit composed of the neck region of SP-D and three C-type lectin domains.

0 1997 Blackwell Science Ltd, Clinical and Experimertful Immunology, 110:241-249


Table 2. Binding of surfactant proteins A and D to various forms of allergens

Immobilized With SP-A antibodies With SP-D antibodies With rSP-D(N/CRD) allergen at 450 nm at 450nm antibodies at 450 nm

3 wcf extract 1.39 ? 0.09 1.52 -C 007 066 t 0.54 gp55 0.85 t 005 0.88 2 0.06 0.44 t 0.07 gp45 073 2 0.04 0.74 t 0.08 0.38 2 0.04 Asp fI(18kD) 0.15 t 0.05 0.14 2 0.03 0.10 2 0.03 Asp fII (29kD) 0.21 2 003 0.19 -C 0.04 0.13 2 0.04 de-gp55 0.13 t 004 0.14? 004 0.08 2 0.04 de-gp45 0.12 t 003 0.13 2 0.03 008 t 003

Binding of fixed amounts (1 pglml) of lung surfactant protein A (SP-A), lung surfactant protein D (SP-D) or rSP-D(N/CRD) to various preparations of Aspergillusfiigutus allergens in the presence of 5 m~ CaClz was carried out as described in Materials and Methods. Results are the mean 2 s.d. of three separate experiments.The de-gp55 and de-gp45 designate the deglycosylated forms of gp55 and gp45, respectively. Asp fI and Asp fII are non-glycosylated allergens.

maltose or mannose (Table 1). SP-A and SP-D have previously been reported to show specific binding to mannose and maltose, respectively. The results in Table 1 suggest that the CRD regions of SP-A and SP-D interact with carbohydrate structures on the conidial cell wall of A. fumigatus and the binding is calcium- as well as sugar-dependent. Human SP-A, SP-D and the recombinant fragment, rSP-D(N/CRD), are able to bind glycosylated allergens derived from A. fumigatus through their CRDs in a calcium- dependent manner (Table 2). The binding of SP-A and SP-D to the 3 wcf allergens seems to be comparable (Table 2). Figure 2a,b show that native SP-A and SP-D isolated from human lung surfactant bind to the A. fumigatus 3 wcf allergendantigens, purified gp55 and gp45 in a dose-dependent manner. The two major non-glycosylated allergens of A. fumigatus, Asp fI and Asp fII, were not recognized by either of the lung surfactant proteins. Further, binding of surfactant proteins to the gp55 and gp45 was abolished on deglycosylation of these allergendantigens (Table 2). These results suggest that SP-A and SP-D interact with carbohydrate structures on the glycoprotein allergens of A. fumigatus.

Protein affinity column chromatography confirmed the interaction of CRD with glycoproteins of 3 wcf of A. fumigatus which are allergenic In an attempt to investigate further the interaction between carbohydrate recognition domains of lung surfactant proteins with glycoprotein allergens of A. fumigatus, we used an in-frame fusion protein, composed of the neck and CRD region of SP-D linked to E. coli MBP, which was bound to an amylose resin column in the presence of calcium. The neck-CRD on its own does not bind to the amylose resin column. A control column containing only immobilized MBP was also prepared. Following coincubation of fusion protein-amylose resin slurry with the 3 wcf of A. fumigatus or the purified gp55 and gp45 preparations, the proteins bound by the CRDs of SP-D were eluted using l o r n EDTA. The 3 wcf allergenslantigens EDTA eluate (Fig. 3 , lanes A, B and C) and gp55, and gp45 EDTA eluates (Fig. 3, lanes D and E) were electrophoresed and examined by ligand blotting involving detec- tion by sera containing allergen-specific IgE (lanes A, D and E), biotinylated SP-A (lane B) or SP-D (lane C) (Fig. 3 ) . The

I (a’ 1.50

0 LD

n 0

0

0 200 400 600 800 1000 1200

SP-A (ng/mI)

0 200 400 600 800 1000 1200

SP-D (ng/ml)

Fig. 2. Solid-phase binding assays of (a) lung surfactant protein A (SP-A) and (b) lung surfactant protein D (SP-D) to 3-week culture filtrate of Aspergillusfumigatus, gp45 and gp55.0, gp55; 0, gp45; W, 3-week culture filtrate; 0 bovine serum albumin.

0 1997 Blackwell Science Ltd, Clinical and Experimental Immunology, 110241-249

T. Madan et al. 246

195 - 112 - 84 - 63 - 52 -

35 - 32 -

A B

195 - 112 - 84 - 63 - 52 - 35 - 32 -

C 1 D

Fig. 3. Immunoblots of 3-week culture filtrate of Aspergillus fumigatus, captured on a recombinant lung surfactant protein D (SP-D) column, eluted with EDTA, and then probed with (A) pooled sera from allergic patients and anti-IgE conjugate; ( B ) biotinylated native lung surfactant protein A (SP-A); ( C ) biotinylated native SP-D. Lanes D and E represent capture of gp55 and gp45, respectively, by a recombinant fragment of SP-D linked to an amylose resin and probed with pooled sera from allergic patients and anti-IgE conjugate.

glycoproteins, captured by recombinant SP-D column, were allergenic as they were also recognized by the IgE antibodies from the sera of ABPA patients.The control column, with only MBP bound to it, did not capture any of the A. fiirnigarus glycoproteins.

1.2

1.0

0.4

0.2 0 1 2 3 4

Concentration of SP-A/SP-D (yg/rnl)

Fig. 4. Inhibition of the binding of Aspergillus fumigatus-specific IgE antibody to 3-week culture filtrate allergens by lung sufactant protein A (SP-A) (0) and lung surfactant protein D (SP-D) (W).

SP-A and SP-D can inhibif the binding of A. fumigatus-specific IgE antibodies to A. fumigatus allergens The inhibition of the binding of A. fumigatus-specific IgE to 3 wcf allergens by surfactant proteins is shown in Fig. 4. Both native SP- A and SP-D showed inhibition, in a dose-dependent manner, of the binding of IgE to the allergens. The rSP-D(N/CRD) was also found to inhibit the binding of IgE to the allergens, but to a lesser extent (data not shown) than that shown by native SP-D (>go% at a concentration of 4 pg/ml). Thus, both SP-A and SP-D efficiently inhibited the binding of allergen-specific IgE to native gp55 and gp45, in an ELISA-based system.

SP-A and SP-D can block histamine release by A. fumigatus- stimulated basophils Inhibition of specific IgE binding of A. fumigatus allergens by SP- A and SP-D prompted us to determine the effect of surfactant proteins on histamine release by allergens from the sensitized basophils. Samples of diluted whole blood from ABPA patients and normals were challenged with A. furnigatus allergens in the presence of SP-A and SP-D. Histamine release was measured by an immunoassay based on acylated histamine. A standard curve was constructed from known concentrations of histamine (50, 15, 5, 1, 0.1 nM histamine gave absorbance readings of 0~08,0.15,055, 1.34 and 1.6, respectively). As seen in Fig. 5, both SP-A and SP-D could reduce, by more than 50%, the histamine release mediated by the 3 wcf allergens (at allergen concentrations of 6.25 pg/ml or less) acting on sensitized basophils, derived from three ABPA patients. Similar results were obtained when purified gp45 and gp55 were used. Histamine release was reduced by approximately 50% when a deglycosylated preparation of gp55 was used instead of the intact gp55 (Table 3). Deglycosylation of gp45 similarily lowered the histamine release with patient 1 cells (Table 3), but no significant decrease was seen with patient 2 cells. The discrepancy could be due to patient-to-patient variation based on antisera composition differences. Control samples, gp55 and gp45 incubated with buffer only without any N-glycanase F treatment, were able to bring about

Table 3. Histamine release from basophils induced by purified allergens of Aspergillus fumigarus, in the presence and absence of lung surfactant protein A (SP-A) and lung surfactant protein D (SP-D). The de-gp55 and

de-gp45 designate the deglycosylated forms of gp55 and gp45

Histamine release (nM)

Patients Concentration of Concentration of allergen/ml SP-A or SP-Dlml 1 2 Control

gp45 12.5 pg

de-gp45 12.5 pg

gp55 12.5 p g

de-gp55 12.5 pg

None 5 pg SP-A 5 pg SP-D

None 5 p g SP-A 5 pg SP-D



285 I60 157 152 I46 145 402 205 155 158 154 1 50

156 52 130 48 136 55 142 50 143 60 139 53 265 46 198 48 156 55 148 45 146 46 152 46

0 1997 Blackwell Science Ltd, Clinical and Experimental hnmunology, 110:241-249


r Patient 1 (u 1.0

-? 0.8 0.8 L m

0.6 0.6 E r n Eg 0.4 0.4 I

m -

.- Y

.- 1 0.2 0.2

0.0 n.n

12.5 6.25 3.125 0.781 w w -

12.5 6.25 3.125 0.781

Patient 2

1.2 r 4 0.8 L m

Patient 3

0 * 0.6 E r n ;g 0.4

0.2 I

.- Y

.- -

n.n " Y

12.5 6.25 3.125 0.781

A1 lerg e n concentration (pg/rn I)

1.2

1.0 Control subject

0.6 o.8 i- 0.4

0.2

0.0 12.5 6.25 3.125 0.781

Allergen concentration (pg/rnI)

Fig. 5. Inhibition of Aspergillus fumigatus 3-week culture filtrate allergen-induced histamine release from sensitized basophils by lung surfactant protein A (SP-A) and lung surfactant protein D (SP-D). 4, Allergen alone; 8, allergen + 5 pg/ml SP-A; 0, allergen + 5 pglml SP-D.

histamine release (data not shown). Furthermore, histamine release observed with non-glycosylated allergens was not inhibitable by SP-A and SP-D (data not shown). Non-sensitized basophils showed background levels of histamine release and SP-A and SP-D had no effect on basophils alone. The results strongly suggest that both SP-A and SP-D are capable of inhibiting histamine release from sensitized basophils, possibly by binding to the glycosylated allergens and thereby interfering with the crosslink- ing of specific IgE molecules to basophils/mast cells.

DISCUSSION

The inhalation of A. firnigatus conidia as a major cause of allergic respiratory illness is generally accepted. Occupational asthma associated with Aspergillus has been described in the baking and food industries, where Aspergillus-derived enzymes (e.g. a-amylase and cellulase) are used as additives in food preparation in powdered form [23-251. Inhalation of enzymes results in allergic respiratory symptoms (rhinitis and asthma). An IgE-mediated pathogenesis has been confirmed in these cases [24,25]. Patients with ABPA have elevated serum levels of A. furnigarus-specific IgE and IgG. In this study, we investigated the binding of purified surfactant proteins, SP-A and SP-D, and a recombinant form of SP- D to various allergens/antigens of A. furnigatus. SP-A and SP-D are synthesized and secreted into the lung airspaces by type I1 pneumocytes and Clara cells. They belong to group III of the C-type lectins which have an overall structural similarity to the complement protein Clq. The C-type CRDs of the surfactant molecules can bind, via specific carbohydrate residues, to the surfaces of pathogenic organisms [26].

Glycoprotein antigens, the major constituents of both the mycelium and culture filtrate of A. fumigatus, have been shown to have strong immunological reactivity [ 81. Located at the interface between host and pathogen cells, the fungal cell wall plays a major role during fungal invasion. It possesses several surface-expressed molecules involved in adhesion of the fungus to host protein and cells. The inhibition of allergen-surfactant protein binding by sugars suggests that specific binding takes place by the CRDs of SP-A and SP-D. This view is supported by the finding that trimeric recombinant CRDs of the SP-D bind to allergenic extracts, and also by the fact that SP-D and SP-A do not bind to the non-glycosylated or deglycosylated allergens. SP-A and SP-D are C-type Iectins which may distinguish non-self carbohydrates from self-glycoproteins via their carbohydrate recognition domains. Little is known about the actual molecular architecture of the cell wall of A. jiunigarus, but hyphal wall preparations have shown abundance of mannose, glucose and amino sugars [27]. Analysis of the purified allergen gp45 by gas liquid chromatography and lectin binding assays has shown the presence of mannose, glucose, galactose and N-acetylglucosamine [22]. A recently described, 58-kD allergen also contains mannose, galactose and glucose residues in a 2:1:2 ratio [28]. The finding that the SP-A and SP-D bind to the infectious conidia surface as well as the culture filtrate allergendantigens is not surprising, since the Aspergillus cell wall contains most of the antigens secreted by the fungus during its active in vitro or in vivo growth. Immunofloures- cence microscopy with MoAb against 58-kD auergedantigen showed that it is abundant on the cell wall surface [28]. An immunoelectron microscopy study involving a panel of six MoAbs showed that similar antigens are present on or within the conidia or hyphae of the fungus [29].

0 1997 Blackwell Science Ltd, Clinical and Experimenful Immunology, 110:241-249

248 T. Madan et al.

To date, the mechanisms of sensitization and provocation by airborne allergens such as fungal spores and pollen grains are largely undefined. The direct participation of surfactant proteins in the allergic inflammatory processes that are frequently seen in asthmatic patients is debatable. There has been no comprehensive analysis of surfactant protein levels and phospholipid compositions in BALF samples from asthmatic subjects. There is one recent study [30] describing a decreased concentration of SP-A relative to phospholipid in BALF from asthmatic patients compared with that from normal controls. Other evidence for a possible involvement of surfactant in asthma is based on animal studies. Recently, Wang er al. [19] reported inhibition of specific IgE binding to mite allergenic extracts using SP-A and SP-D, suggesting that these proteins may be involved in the modulation of hypersensitivity reactions. In the present study, we have shown that both SP-A and SP-D are capable of completely blocking the binding of allergen- specific IgE to 3 wcf allergendantigens, at a concentration of 4 pg/ ml, which is in the physiological range. There are two probable mechanisms by which surfactant proteins inhibit IgE binding to allergens: either by steric hindrance posed by the surfactant molecules already bound to the allergen’s carbohydrate structure, or by the recognition of the same bindmg site by both IgE and the surfactant proteins. The carbohydrate moiety of a glycoprotein is proposed to be important for the specific IgE binding, as we observed loss of IgE binding activity of the gp45 allergedantigen on partial deglycosylation with N-glycanase F. Also, sodium metaperiodate treatment of the 58-kD allergen completely abolished its serological activity [28]. Further studies to dissect the mechanisms by which lung surfactant proteins inhibit allergen-specific IgE binding may aid in development of a new therapy for allergic patients.

Specific IgE-mediated type I and specific IgG-mediated type I11 hpersensitivity reactions are proposed to play an important role in the immunopathogenesis of ABPA caused by A. fumigarus [3 11. Allergens crosslink the specific IgE antibodies, allowing interaction of the immunoglobulin with the IgE receptor, resulting in degranulation of the mast cells and basophils, and subsequent release of histamine follows. The pharmacological actions of histamine lead to hypersensitivity reactions in the patient, causing irreversible damage. The SP-A and SP-D significantly reduced histamine release from the sensitized basophils by the A. fumigatus allergens. However, as some of the A. furnigatus allergens are non- glycosylated, surfactant proteins were not able to bind them and thus would not be expected to completely block histamine release. There is a discrepancy seen in the IgE inhibition (90% at 4 pg/ml SP-A or SP-D, Fig. 4) and the suppression of histamine release (50% at 5 pglml SP-A or SP-D, Table 3 and Fig. 5 ) , but this may be primarily due to differences in the assay systems used. The ability of surfactant proteins to reduce histamine release by the sensitized cells on allergen exposure suggests that the SP-A and SP-D may find a prophylactic and/or therapeutic role in the treatment of allergic disorders like ABPA. The proposed therapeutic role for lung surfactant proteins would be further strengthened by experi- mentation in animal models of Aspergillosis. A study of levels of lung surfactant proteins in allergic and invasive Aspergillosis patients during and after the disease may also help understand a possible up-regulation or down-regulation of SP-A and SP-D under pathological conditions.

ACKNOWLEDGMENTS This work was supported by the Medical Research Council of Great Britain

(P.E., K.B.M.R.), the European Commisssion Biotechnology Programme (U.K., K.B.M.R.), Department of Biotechnology, India (P.U.S.), and Council of Scientific and Industrial Research, India, for a Senior Research Fellowship (T.M.).

REFERENCES

1 Binder RE, Faling U, Pugatch RD et al. Chronic necrotising pulmonary aspergillosis: discrete clinical entity. Medicine 1982; 61: 109-24.

2 Bardana EJJ. The clinical spectrum of aspergillosis. Part 2. Classifica- tion and description of saprophytic, allergic, and invasive variants of disease. CRC Crit Rev Clin Lab Sci 1980; 13:85-159.

3 Rinaldi MG. Invasive aspergillosis. Rev Infect Dis 1985; 5:1061-77. 4 Banejee B, Chetty A, Joshi AP et al. Identification and partial

characterization of diagnostically relevant antigens of Aspergillus fumigatus. Asian Pac J All Immunol 1990; 18:13-18.

5 Becker JW, Burke W, McDonald G er al. Prevalence of allergic bronchopulmonary aspergillosis and atopy in adult patients with cystic fibrosis. Chest 1996; 109:1536-40.

6 Andriole VT. Aspergillus infections: problems in diagnosis and treatment. Infect Agents Dis 1996; 5:47-54.

7 Schaffner A. Pulmonary aspergillosis and AIDS. N Engl J Med 1991; 325:355.

8 Kurup VP, Kumar A. Immunodiagnosis of aspergillosis. Clin Microbiol Rev 1991; 4:434-56.

9 Schaffner A, Douglas H, Brande A. Selective protection against conidia by mononuclear and against mycelia by polymorphonuclear phagocytes in resistance to Aspergillus. J Clin Invest 1982; 69:617-31.

10 Elstad A, Douglas H, Braude E. Aspergillosis and lung defenses. Semin Respir Infect 1991; 6:27-36.

11 Washbum RG, Gallin JI, Bennett JE. Oxidative killing of Aspergillus fumigatus proceeds by parallel rnyeloperoxidase-dependent and -inde- pendent pathways. Infect Immun 1987; 55:2088-92.

12 Corrado MS, Cleri D, Fikring SM. Aspergillosis in chronic granulomatous disease : therapeutic considerations. Am J Dis Child 1980;

13 Levitz SM, Diamond RD. Mechanisms of resistance of Aspergillus fumigatus conidia to killing by neutrophils in vitro. J Infect Dis 1985;

14 Kurup VP. Interaction of Aspergillus fumigatus spores and pulmonary alveolar macrophages of rabbits. Immunobiol 1984; 16654-61.

15 Kishore U, Eggleton P, Reid KBM. Modular organization of carbohydrate recognition domains in animal lectins. Matrix Biol 1997; 15583-92.

16 Madan T, Eggleton P, Kishore U et al. Binding of pulmonary surfactant proteins A and D to Aspergillus conidia enhances phagocytosis and killing by human neutrophils and alveolar macrophages. Infect Immun 1997; 65:3171-9.

17 Waldorf A. Pulmonary defense mechanisms against opportunistic fungal pathogens. Immunol Ser 1989; 47:243-71.

18 Malhotra R, Haurum J, Thiel S et al. Pollen grains bind to lung alveolar type I1 cells (A549) via lung surfactant protein A (SP-A). Biosci Rep 1993; 13:79-YO.

19 Wang JY, Kishore U, Lim BL et al. Interaction of human lung surfactant proteins A and D with mite (Dermutophagoides pteronyssi- nus) allergens. Clin Exp Immunol 1996; 106:367-73.

20 Kishore U, Wang JY, Hoppe H-J et al. The a-helical neck region of human lung surfacatant protein D is essential for the binding of the carbohydrate recognition domains to lipopolysaccharide and phospholipids. Biochem J 1996; 318:505-1 1.

21 Teshima R, Ikebuchi H, Sawada J et al. Isolation and charcterization of a major allergenic component (gp55) of Aspergillus fumigatus. J Allergy Clin Immunol 1993; 92:698-706.

22 Banejee B, Madan T, Sharma GL et al. Characterization of glycoprotein antigen (45 kD) of Aspergillus fumigatus. Serodiagn Immunother Infect Dis 1995; 7:147-52.

23 Baur X, Fruhmann G, Hang B. Role of Aspergillus amylase in Baker’s asthma. Lancet 1986; 1:43.

134~1092-4.

152133-42.



24 Baur X, Sauer W, Weiss W. Baking additives as new allergens in Baker’s asthma. Respiration 1988; 5470-72.

25 Quirce S, Cuevas M, Diez-Gomez ML. Respiratory allergy to Asper- gillus-derived enzymes in baker’s asthma. J All Clin Immunol 1992;

26 Holmskov U, Malhotra R, Sim RB et al. Collectins: collagenous C-type lectins of the innate immune defense system. Immunol Today 1994;

27 Heam VM, Sietsma JH. Chemical and immunological analysis of the

28 Fratamico M, Buckley R. Identification and charcterisation of an

901970-8.

15167-74.

Aspergillusfumigatus cell wall. Microbiol 1994; 140:789-95.

immunodominant 58 kilodalton antigen of Aspergillusfumigatus recog- nised by sera of patients with invasive aspergillosis. Infect Immun

29 Reijula K, Kurup VP, Kumar A ef al. Monoclonal antibodies bind identically to both spores and hyphae of Aspergillus fumigatus. Clin Exp Allerg 1992; 22547-53.

30 van de Gaarf E, Jansen HM, Lutter R et al. Surfactant protein A in bronchoalveolar lavage fluid. J Lab Clin Med 1992; 120:252-63.

3 I Slavin RG, Fischer VW, Levine EA e f al. A primate model of allergic bronchopulmonary aspergillosis. Int Arch Allergy App Immunol 1978;

1991; 591309-15.

56:325-33.


Lung surfactant proteins A and D can inhibit specific IgE binding to the allergens of Aspergillus fumigatus and block allergen-induced histamine release from human basophils

Documents