Antigen 43/Fcε3 chimeric protein expressed by a novel bacterial surface expression system as an effective asthma vaccine

Construction of the Ag43 surface expression system

The Ag43 system consists of an E. coli strain with an Ag43-gene-knockout (Tan109), and a recombinant Ag43 plasmid, pETAg43. Deletion of the Ag43 gene from JM109 was performed with a TargeTron^™ Gene Knockout System (Sigma-Aldrich, St Louis, MO) by insertion of a group II intron into the Ag43 gene at site 1812/1813 as described by the manufacturer. Bacterial strain knockout was confirmed by PCR and sequencing. To construct the recombinant plasmid, pETAg43, the Ag43 gene (598–3210, GenBank Accession no. CP000948) was amplified by PCR using primers 5′-CCAAAGCTTAACGGTGATACCGG-3′ and 5′-CCACTCGAGTCAGAAGGTCACAT-3′. The product was then digested and cloned into the backbone of plasmid pET-22b(+) (Novagen^®, Merck KGaA, Darmstadt, Germany). Plasmid pETAg43 was then verified by restriction enzyme digestion and sequencing.

Induction of Ag43 chimeric protein expression and vaccine preparation

The Fcε3 domain of the IgE gene (517-804, GenBank Accession no. J00476) was amplified with reverse transcription PCR using primers 5′-CCGGATCCTACCTACCTGATCCCACCC-3′ and 5′-CCACTCGAGGGTGATGGAACGCTC-3′. The PCR product was then subcloned into the pETAg43 plasmid. The resultant plasmid – pETAg43/Fcε3 – was verified by restriction enzyme digestion and sequencing. Surface expression of Ag43 chimeric proteins was performed in Tan109 by growing the transformed bacteria in 1 mm isopropyl-β-d-thiogalactoside, and surface expression was measured with anti-IgE (Santa Cruz Biotechnology, Santa Cruz, CA) immunofluorescence. Bacteria were then suspended in purified water, heated to 60° for 5 min and centrifuged at 2300 g for 3 min. Supernatants were separated by SDS–PAGE and subsequent Western blotting was used to confirm the presence of chimeric proteins. The murine recombinant IgE protein (rIgE) was prepared as previously reported.24 Supernatant proteins were directly lyophilized using a lyophilizer (FreeZone 4.5; Labconco, Kansas City, MO) and stored at 4°. Protein endotoxin isolated from bacterial expression was measured with a Limulus amoebocyte lysate assay and endotoxin units were reduced to less than 0·5/μg with a Sartobind Q filter. Before vaccination, proteins were dissolved at 200 μg/ml in PBS and Imject^® Alum adjuvant (Pierce Biotechnology, Rockford, IL) was added, which contained an aqueous solution of aluminium hydroxide (40 mg/ml) and magnesium hydroxide (40 mg/ml) plus inactive stabilizers, dropwise with constant mixing for 30–60 min of each protein solution to a final volume ratio of Imject^® Alum to protein of 1 : 1 as previously described.25

SDS–PAGE and Western blot analysis

SDS–PAGE and Western blot analysis was performed as described previously.24 Briefly, sample proteins were separated with 12% SDS–PAGE and stained with Coomassie Brilliant Blue (Bio-Rad, Hercules, CA). Proteins were then transferred onto a PVDF membrane (Bio-Rad) using a mini Trans-Blot system (Bio-Rad). Membranes were blocked in 5% non-fat dry milk at 4°, washed, and probed with IgE antibodies (Santa Cruz) at 1 : 500, or with sera. Membranes were washed and incubated with a biotinylated secondary antibody (biotinylated rabbit anti-mouse IgG) and stained with the Vectastain ABC Kit (Vector Laboratories, Burlingame, CA).

Animals

Animal care and use were carried out in accordance with the guidelines of the Dutch Committee of Animal Experiments. All experiments were conducted under a protocol approved by the Animal Care and Use Committee (approval ID: HNMCE10012-5). Specific pathogen-free male BALB/c mice between 6 and 8 weeks old were purchased from Guangzhou Animal Centre, China. Mice were housed in makrolon cages under a laminar flow cabinet and provided with ovalbumin (OVA)-free food and water ad libitum.

Mouse asthma model and vaccine immunization

To establish a mouse allergic asthma model, mice were sensitized to OVA (Sigma-Aldrich) using two intraperitoneal injections (7 days apart) of 100 μl alum-precipitated antigen composed of 10 μg OVA absorbed onto 2·25 mg Imject^® Alum (Pierce Biotechnology). Thereafter, mice were exposed to an aerosol challenge containing 2% OVA (2 mg/ml saline) for 60 min/day. The challenge was performed in a Plexiglas exposure chamber coupled to a Pari LC Star nebulizer (particle size 2·5–3·1 μm; PARI Respiratory Equipment, Midlothian, VA) driven by compressed air at a flow rate of 6 l/min. Details are given in Fig. 1(g).

To test anti-asthma effects, sensitized mice were randomly divided into four groups (n = 10). Group 1 (Ag43/Fcε3) mice were first sensitized with OVA at days 0 and 7, challenged with 2% OVA aerosol, and immunized subcutaneously with Ag43/Fcε3 (10 μg/mouse in 100 μl). The other three were control groups (Ag43, rIgE and aPBS) and were sensitized, challenged and immunized with the corresponding proteins (aPBS contained no protein).

Lung histology

For histological evaluation of lung tissues, lungs were removed and fixed in 10% PBS. Left lung lobes were embedded in paraffin, sectioned (3–5 μm), and stained with haematoxylin and eosin. Cellular infiltrates were measured in five randomly selected fields under a microscope (400 ×). Groups were analysed in a blind fashion, as previously described.24

Bronchoalveolar lavage

On day 42, after 2% OVA challenge, mice were killed by CO₂ asphyxiation, and samples of bronchoalveolar lavage fluid (BALF) and lung tissue were collected as previously described.26 Briefly, 10 mouse lungs were perfused through the pulmonary artery and lavaged through the trachea three times with 1 ml PBS. Total number of BALF cells were measured using a haemocytometer. BALF cells were then spun onto slides by cytocentrifugation, fixed and visualized using Hansel stain. Differential cell counts were made on at least 400 cells as previously described.24

Quantification of eosinophils

Peripheral blood samples, lung tissues and BALF were collected at the end of the experiment. Carbol's chromotrope–haematoxylin stain was used to identify eosinophils using morphological criteria, and these were quantified as previously described.24

Measurement of airway hyper-responsiveness

Airway hyper-responsiveness (AHR) was assessed using methacholine-induced airflow obstruction in conscious mice placed in a whole body Plethysmography (model PLY 3211; Buxco Electronics, Wilmington, NC). Measurement of dimensionless parameters such as enhanced pause (P_enh) was performed as previously described.24, 27 Briefly, mice were placed in the plethysmograph chamber and exposed to an aerosol of normal saline water (baseline readings) and then to increasing concentrations of β-methacholine (3 mg/ml). The aerosol was generated by an ultrasonic nebulizer and drawn through the chamber for 2 min. The inlet was then closed, and P_enh readings were taken for 3 min and averaged. Values are reported as the per cent increase over baseline.

Enzyme-linked immunosorbent assay

The autoantibodies against self-IgE, OVA-specific IgE, immunoglobulin subclasses, T helper type 1 (Th1) and and type 2 (Th2) cytokines were measured by ELISA using kits or antibodies sets according to the manufacturer's protocol (BD Biosciences, San Jose, CA). Enzyme activity was measured with an ELISA reader (Bio-Tek, Winooski, VT) as described previously.28

Enzyme-linked immunospot assay (ELISPOT)

ELISPOT was used to quantify spleen B cells secreting anti-IgE, OVA-specific IgE antibody or histamine as reported previously.24 Briefly, PVDF-bottomed 96-well filtration plates (Millipore, Billerica, MA) were coated with 25 μg/ml of murine rIgE protein, OVA, or anti-histamine antibody. Mononuclear cells prepared from spleens or lymph nodes were incubated in plates at 37° in 5% CO₂ for 5 hr. IgG, IgE or histamine (after a 20 μg/ml OVA challenge) bound to the membrane was stained with alkaline phosphatase-conjugated anti-mouse IgG, IgE, or anti-histamine antibodies.

Purification and adoptive transfer of immunoglobulin

Antibodies were purified using affinity chromatography (CM Affi-Gel blue gel kit; Bio-Rad) from pooled sera derived from mice at day 7 after the fourth immunization as published elsewhere.24 To assess antibody efficacy for anti-asthma activity in vivo, purified antibodies (50 mg/kg) were intravenously transferred 1 day before mice were challenged with 2% OVA aerosols (see Materials and methods; Fig. 1g). Animals were treated twice weekly until they were killed. Thereafter, lung histology, inflammatory cells, eosinophils and AHR were evaluated as described.

Histamine assay

To measure histamine release, lymph nodes were separated and cultured with OVA (10 μg/ml) for 2 hr. Then, supernatants were collected and histamine was measured with ELISA (Neogen, Lexington, KY).

Flow cytometry

The mast cell marker FcεRIα and intracellular expression of histamine were measured by flow cytometry after surface staining with an FITC-conjugated anti-mouse FcεRIα antibody (BD Biosciences) and intracellular staining with anti-histamine antibodies conjugated with phycoerythrin (BD Biosciences). For intracellular histamine staining, mononuclear cells prepared from spleens were incubated on six-well plates with OVA (20 μg/ml) at 37° and 5% CO₂ for 5 hr, then fixed, permeabilized and stained with phycoerythrin-conjugated anti-mouse histamine antibodies (BD Biosciences). Flow cytometry was performed with a FACS Canto II (BD Biosciences) and analysed using bd fcsdiva software and fcs epress 4 software (De Novo Software, Los Angeles, CA).

Statistical analysis

Significant differences between experimental groups were analysed with analysis of variance and unpaired Student's t-test. Values were calculated as mean ± SEM. Differences in means were considered to be significant at P < 0·05.

Surface expression of Ag43/Fcε3 and Ag43 proteins and their antigenicity

The Ag43 system consists of an Ag43-deleted strain of E. coli Tan109 and a recombinant plasmid, pETAg43. The Fcε3 domains of the mouse IgE gene were inserted into recombinant plasmid pETAg43, producing a new plasmid, pETAg43/Fcε3. This pETAg43/Fcε3 was then transformed into Tan109 to induce surface expression of Ag43/Fcε3 chimeric protein (Fig. 1b). To confirm Ag43/Fcε3 surface expression, Tan109 cells were fixed on slides and incubated with anti-IgE mAb followed by FITC-conjugated secondary antibody. Green signals were observed only in bacteria transformed with plasmid pETAg43/Fcε3, not in those transformed with plasmids pETAg43 or pETFcε3 (Fig. 1c). Heating Tan109 cells expressing Ag43/Fcε3 and Ag43 proteins to 60° followed by SDS–PAGE confirmed a green signal on pETAg43/Fcε3-expressing Tan109 cells that almost disappeared after heating (Fig 1d). The SDS–PAGE results indicated that bacterial cells expressed Ag43/Fcε3 protein (Fig. 1e, lane 1) and there was loss of a protein band in heat-exposed cells (Fig. 1e, lane 2). However, a single band of a molecular weight similar to the lost band was observed in cell supernatant expressing Ag43/Fcε3 (Fig. 1e, lane 3). Ag43 was similar with respect to SDS–PAGE (data not shown). Hence, heating detached the Ag43/Fcε3 and Ag43 proteins expressed on the surface of Tan109. Figure 1(f) depicts specificity of anti-IgE monoclonal antibody to Ag43/Fcε3 and rIgE protein and binding of Ag43 sera to Ag43/Fcε3 and Ag43, respectively, suggesting that Ag43/Fcε3, Ag43 and rIgE have excellent antigenicity.

Decreased lung inflammation

Haematoxylin and eosin staining of lung sections from mice immunized with Ag43/Fcε3 revealed almost normal lung histology, with only marginal perivascular and peribronchiolar lymphocytic infiltrates (Fig. 2a). In contrast, lung sections from mice immunized with Ag43, rIgE or aPBS vaccine had significant mucus hypersecretion and airway inflammation with peribronchiolar and perivascular infiltrates, consisting of lymphocytes, eosinophils and some neutrophils (Fig. 2a). These results suggest that Ag43/Fcε3 immunization decreased mucus hyperproduction and airway inflammation.

Inflammatory cells (including eosinophils) were observed in BALF, blood and lung tissues. Inflammatory cells in BALF of mice immunized with Ag43/Fcε3 were fewer than in control groups (Fig. 2b). In addition, eosinophils as measured with Carbol's chromotrope–haematoxylin stain were significantly attenuated in BALF, blood and lung tissue of Ag43/Fcε3-immunized mice compared with control mice (Fig. 2c–e). These results indicate that Ag43/Fcε3 can significantly reduce inflammation in an asthmatic mouse model.

Inhibition of airway hyper-responsiveness

Airway hyper-responsiveness to various bronchoconstrictive stimuli is a pathological characteristic of asthma.6, 29, 30 In the present study, AHR was measured by challenging animals with increasing concentrations of methacholine at different time-points (see AHR development in Fig. 3). Mice immunized with Ag43 and rIgE, or aPBS vaccine developed significant AHR. However, mice immunized with Ag43/Fcε3 had less AHR (Fig. 3), indicating that Ag43/Fcε3 can significantly inhibit the development of AHR in an allergic asthma mouse model.

Production of autoantibodies against mouse self-IgE

To investigate autoantibodies against mouse self-IgE, sera were isolated from experimental mice. As shown with Western blot, sera from Ag43/Fcε3-immunized mice recognized not only Ag43/Fcε3, but also Ag43 and rIgE proteins (Fig. 4a), whereas sera from Ag43-immunized mice recognized Ag43/Fcε3 and Ag43, but not rIgE protein (Fig. 4a). In contrast, sera from both rIgE-immunized and aPBS-immunized mice did not recognize Ag43/Fcε3, Ag43, and rIgE (Fig. 4a) as shown with ELISA, sera from mice immunized with Ag43/Fcε3 contained IgG antibodies against self-IgE (rIgE), and major immunoglobulin subclasses responding to Ag43/Fcε3 were significantly elevated in IgG1 and IgG2a (Fig. 4d). However, antibodies against self-IgE were not observed in sera from control mice (Fig. 4d, left). Mouse spleens were collected and anti-IgE antibody-secreting B cells were measured with ELISPOT. Compared with control groups, antibody-secreting B cells secreting antibodies against self-IgE were significantly increased in spleens of mice immunized with Ag43/Fcε3 (Fig. 4b and c). Hence, Ag43/Fcε3 immunotherapy may bypass immune tolerance and induce the production of autoantibodies against mouse self-IgE.

Anti-asthma effects by passive adoptive transfer of antibodies

Passive adoptive transfer of purified antibodies isolated from Ag43/Fcε3-immunized and control mice was performed to investigate anti-asthma effects. Inflammatory cells (including eosinophils) in BALF and peripheral blood and AHR were assayed as described above. Data show that passive adoptive transfer of purified antibodies from mice immunized with Ag43/Fcε3 also reduced inflammatory cells in BALF (Fig. 5a), significantly decreased eosinophils in both BALF (Fig. 5b) and peripheral blood (Fig. 5c) and inhibited AHR development (Fig. 5d). These results agree with those from mice immunized with Ag43/Fcε3, indicating that anti-asthma activity of the Ag43/Fcε3 vaccine is related to autoantibodies against self-IgE.

Down-regulation of IgE production

Total IgE and OVA-specific IgE were measured with ELISA. Sensitization and challenge with OVA elicited significant OVA-specific and total IgE responses in control mice, but not in mice immunized with Ag43/Fcε3 (Fig. 6a,b). Also, ELISPOT was used to measure OVA-specific IgE-secreted B cells. The antibody-secreting B cells in mice immunized with Ag43/Fcε3 significantly decreased compared with control mice (Fig. 6c,d). Hence, vaccination with Ag43/Fcε3 inhibits production of IgE antibody.

Reversion of Th2 cytokine production

To determine whether reduced inflammation and AHR in OVA-induced allergic asthma mice immunized with Ag43/Fcε3 correlated with alterations in T helper cytokines, we measured interleukin-4 (IL-4), IL-5 and IL-13 to represent the Th2 profile, and interferon-γ, IL-2 and IL-12 to represent the Th1 profile. Lymph node cells were stimulated with OVA in vitro and cytokines were detected with ELISA. High IL-4 (Fig. 7a), IL-5 (Fig. 7b) and IL-13 (Fig. 7c) and low interferon-γ (Fig. 7d), IL-2 (Fig. 7e) and IL-12 (Fig. 7f) were measured in the lymph nodes from control mice. However, vaccination with Ag43/Fcε3 significantly decreased Th2 cytokine production and greatly increased Th1 cytokine synthesis (Fig. 7).

Decreased histamine release

Compare with the control mice, histamine released in the serum and culture medium (Fig. 8a) was significantly reduced in mice immunized with Ag43/Fcε3. Also, histamine-secreting cells detected by ELISPOT were significantly decreased in mice immunized with Ag43/Fcε3 (Fig. 8b). Moreover, spleen monocytes double stained with ant-FcεRIα and anti-histamine antibodies and then examined with flow cytometry revealed that FcεRIα-positive and histamine-expressing monocytes were significantly decreased in mice immunized with Ag43/Fcε3 compared with control mice (Fig. 8c,d). As we know, FcεRIa is the major cell marker of mast cells and basophils. However, mast cells are mainly in the organs and basophils in the peripheral blood. Hence, the FcεRIα-positive and histamine-expressing monocytes from spleens in our current study are possibly mast cells. These data indicate that decreased histamine release is strongly related to fewer histamine-releasing mast cells.