Recombinant Dirofilaria immitis Polyprotein That Stimulates Murine B Cells To Produce Nonspecific Polyclonal Immunoglobulin E Antibody

Abstract

Nonspecific immunoglobulin E (IgE) production is an event characteristically observed in parasitic helminth infections, but its mechanisms are still unclear. To define these mechanisms, we prepared a recombinant Dirofilaria immitis protein (rDiAg) and assessed its effect on nonspecific IgE production. rDiAg preferentially induced nonspecific IgE production, without eliciting specific IgE production, as well as a Th2-type cytokine profile (high interleukin-4 [IL-4] and IL-10 production but low gamma interferon production) in BALB/c mice. rDiAg significantly elicited the proliferative response of naive B cells. This response was not abolished by polymyxin B, an inhibitor of lipopolysaccharide (LPS), and rDiAg normally expanded splenic B cells from LPS nonresponder C3H/HeJ mice. Thus, the mitogenic effect of rDiAg was not due to LPS contamination. rDiAg also enhanced levels of CD23 expression on splenic B cells. Splenic B cells produced marked levels of IgE when cultured with the combination of rDiAg and IL-4 (rDiAg-IL-4), whereas peritoneal B cells produced negligible levels of IgE. rDiAg-IL-4-induced IgE production by splenic B cells was synergistically increased by coculture with peritoneal B cells. rDiAg-driven IL-10 secretion was higher in peritoneal B cells than in splenic B cells. IgE production by splenic B cells cocultured with peritoneal B cells was decreased to a level comparable to that by splenic B cells in the presence of a neutralizing anti-IL-10 monoclonal antibody. Collectively, these results suggest that rDiAg-induced polyclonal expansion and IgE class switching of splenic B cells contribute to nonspecific IgE production and that these responses are enhanced by peritoneal B-cell-derived IL-10.

Parasitic helminth infections are characterized mainly by markedly elevated levels of total immunoglobulin E (IgE), including parasitic antigen-specific IgE and significant levels of nonspecific polyclonal IgE. The former is associated with the elimination of invading worms, while the latter, which dose not respond to parasitic antigens, is considered to be involved in the survival of invading parasites in an immunologically intact host (22, 34, 41). Nonspecific IgE production is positively controlled by Th2-type cytokine interleukin-4 (IL-4), as supported by a report that IgE production in the nematode Nippostrongylus brasiliensis infections can be fully blocked with anti-IL-4 monoclonal antibody (MAb) plus anti-IL-4 receptor MAb (12). In addition, soluble crude extracts from this worm are able to stimulate IL-4-dependent nonspecific IgE synthesis (11). However, the mechanisms by which the production of nonspecific IgE is preferentially induced during helminth infections have not yet been fully elucidated, except that this response is dependent on both IL-4 and worm products.

The induction of IgE class switching is dependent on two primary signals. The first one is IL-4 inducing the expression of the germ line ɛ transcript. The second one is the interaction of CD40 on B cells with CD40 ligand expressed on activated T cells inducing the expression of the mature ɛ transcript IgE (4). Furthermore, IgE production is enhanced by IL-5, IL-6, IL-9, or IL-10 (10, 23, 39, 50, 52). Among these four cytokines, only IL-10 can enhance the production of nonspecific IgE induced with IL-4 plus anti-CD40 MAb (23, 35, 50). IL-10 is known to be produced by various cells, including B cells, to enhance the development of Th2-type cells by indirectly suppressing the activation of Th1-type cells and to augment the proliferation and differentiation of activated B cells (13). In helminth infections, IL-10 is associated with the induction and maintenance of antigen-specific hyporesponsiveness (26, 32, 42). This cytokine, therefore, may play an important role in the production of nonspecific IgE observed during helminth infections.

B cells can be subdivided into two subtypes based on tissue distribution, surface markers, cell size, proliferative response, and cytokine profile. Conventional B cells (B-2 cells) are generated in bone marrow, are distributed mostly in lymphoid organs (e.g., spleen) or in systemic circulation, express CD23 concomitant with maturation, react to various exogenous antigens with high affinity, and produce IL-10 with the appropriate stimuli (40). In contrast, peritoneal B-1 cells can develop independently of bone marrow, have a capacity for self-renewal, constitutively express CD5 but not CD23, react to autoantigens or bacterial components with low affinity, and spontaneously produce large amounts of IL-10 (18, 36). It has been shown that B cells are polyclonally stimulated in hosts infected with some helminths (14, 43, 48). Furthermore, soluble crude extracts from several helminths can polyclonally stimulate B cells (30, 47, 53). These observations raise the hypothesis that nonspecific IgE production seen during helminth infections is attributed to polyclonal activation of B cells in response to worm products.

It has been shown that nematode polyprotein allergens (NPAs) are secreted as an excretory-secretory (ES) product by various nematodes in vivo and in vitro. The DNA encoding NPA was characterized by tandemly repetitive sequences, containing up to 20 repeat units (24). NPA precursor resulting from the DNA possesses cleavage sites of subtilisin serine protease, which is composed of a cluster of tetra-basic amino acid residues. Thus, NPAs are initially synthesized as a very large precursor polyprotein. The polyprotein is then processed into single repeat units (ranging from 14- to 15-kDa proteins or glycoproteins) by endoprotease digestion, hence yielding multiple copies of similar or identical proteins. Amino acid sequences of NPAs are similar, but not identical, among nematode species. Owhashi et al. showed that a single repeat unit from polyprotein from the canine filarial nematode Dirofilaria immitis adult worm (DiNCF/Di-NPA) acts as a neutrophil chemotactic factor (38). Interestingly, it has been shown that this protein is not immunoprecipitated by sera from naturally infected hosts (1, 46), indicating that DiNCF/Di-NPA-specific antibodies are not produced in their hosts.

We have previously shown that the ES component from D. immitis (native DiAg) induces increased levels of serum IgE in mice and significantly enhances IL-4-induced IgE production by human spleen cells (15, 54), implicating that this molecule may be an inducer of nonspecific IgE. In our preliminary experiments, DNA sequence analysis showed that DiAg is composed of 387 bp and is completely consistent with DiNCF/Di-NPA. To investigate the molecular mechanisms underlying nonspecific IgE synthesis induced during helminth infections, we prepared recombinant DiAg (rDiAg) using Escherichia coli and then administered this recombinant protein via a micro-osmotic pump into Th2 responder BALB/c mice. We here demonstrated that rDiAg exerts itself as a B-cell mitogen, an inducer of IL-10, and a costimulatory molecule in IgE class switching, thereby inducing nonspecific IgE synthesis.

MATERIALS AND METHODS

Amplification, expression, and purification of rDiAg.

The DNA encoding a single repeat unit of DiAg was prepared from pDi6 (a kind gift from Makoto Owhashi), which encodes six repeat units of DiAg, and ligated linkers including an NdeI or BamHI restriction site. The DNA was amplified by PCR methods. Briefly, pDi6 (100 ng/ml) was added to a solution containing 0.5 μM (each) primer, 80 μM deoxynucleoside triphosphate, and 2.5 U of KOD DNA polymerase (TOYOBO, Osaka, Japan) in the recommended buffer. The primer sequences used were as follows: 5′ primer, including an NdeI restriction site, 5′-GCATATGAATGATCATAATTTAGAAAGC-3′; and 3′ primer, including a BamHI restriction site, 5′-CTAAAGGATCCTATCACCGCTTACGCCGTTCATTCATTG-3′. The PCR was performed under the following conditions: 20 cycles of 95°C for 1 min, 54°C for 1 min, 72°C for 2 min, and 72°C for 8 min. The PCR products were purified from agarose gel, digested with NdeI and BamHI (Takara Shuzo Co., Shiga, Japan), and ligated into the NdeI/BamHI-digested and dephosphorylated-pET3a expression vector (Novagen, Madison, Wis.). The pET3a construct, including DiAg (designated as pDP5), was transformed into E. coli strain HMS174(DE3) (Novagen), and the cells were grown in M9ZB medium to an absorbance of 0.6 to 0.8 U at 550 nm. DiAg expression was induced by 0.5 mM isopropyl-β-d-thiogalactopyranoside (Wako Pure Chemical, Osaka, Japan) at 37°C for 2.5 h. The cells were harvested by centrifugation, resuspended in 50 mM HCl-5 mM EDTA, and then centrifuged. Recombinant DiAg in supernatants was precipitated by 60 to 80% saturated ammonium sulfate, purified by Superdex 200 (Pharmacia Biotech, Uppsala, Sweden), dialyzed against distilled water, passed through a column of polymyxin B-immobilized beads to remove contaminating endotoxins, lyophilized, and stored at −30°C until use. The yield of purified rDiAg was 50 to 100 mg/liter of induced bacterial culture with >98% purity. Purified recombinant 15.2-kDa protein was checked for purity by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) (see Fig. 2A). Briefly, the recombinant protein was boiled in SDS sample buffer (125 mM Tris-HCl [pH 6.8], 2.3% SDS, 10% glycerol, 5% 2-mercaptoethanol, and 10 μg of bromophenol blueml) for 3 min and applied onto a 15% polyacrylamide gel. The control recombinant protein used was the N-half terminal of ABA-1 (rCont; amino acid residues 1 to 67). The cDNA encoding two repeat units of ABA-1 was prepared from Ascaris lumbricoides by reverse transcriptase (RT-PCR). The PCR conditions and primers used were as follows: 35 cycles of 94°C for 30 s, 60°C for 30 s, 72°C for 90 s, and 72°C for 9 min; 5′-TTAAGCATGTTGTTGG-3′ as a 5′ primer; and 5′-AATTTCGCGGCATCCGCC-3′ as a 3′ primer. After purification, the PCR products were amplified by using PCR conditions and primers as follows: 25 cycles of 98°C for 15 s, 55°C for 2 s, 74°C for 30 s, and 74°C for 9 min; 5′ primer, including an NdeI restriction site, 5′-CGACATATGCATCATTTCACTCTTGAAAGC-3′; and 3′ primer, including a BamHI restriction site, 5′-AATTGGATCCAATTTCGCGGCATCCGCCC-3′. Subsequent procedures were performed as described above. Native DiAg and whole ES products were prepared as described previously (54). The recombinant and native proteins used in our studies had negligible endotoxin contamination (<2 pg/mg of protein), as measured by Endotoxin Test-D (Seikagaku, Tokyo, Japan).

FIG. 2.

Specific anti-DiAg antibody is not detected in serum from rDiAg-administrated mice. (A) rDiAg was resolved by SDS-15% PAGE and stained with Coomassie brilliant blue. Molecular mass markers are shown in kilodaltons. rDiAg is marked with an arrowhead. (B) rDiAg-blotted membranes were treated with serum from control mice (lanes a and c) or mice administered rDiAg (lanes b and d) (day 21) and control rabbits (lane e) or rDiAg-immunized rabbits (lane f) and then reacted with HRP-conjugated anti-mouse IgE (lanes a and b), anti-mouse IgG (lanes c and d), or anti-rabbit IgG (lanes e and f). Data represent one typical experiment out of five.

Antisera and immunoprecipitation.

Rabbits (Japan White, female, 2 kg each; SLC Japan, Hamamatsu, Japan) were immunized subcutaneously (s.c.) with rDiAg (200 μg/500 μl in phosphate-buffered saline [PBS]) emulsified in Freund's complete adjuvant (500 μl; Sigma Chemical Co., St. Louis, Mo.). After 2 weeks, boosters was administered with rDiAg (200 μg/500 μl in PBS) emulsified in Freund's incomplete adjuvant (500 μl; Sigma) at 2-week intervals. Rabbits were bled at days 0 (preimmune serum) and 63 (hyperimmune serum). IgG fractions in these rabbit sera were collected using a protein A-IgG purification system (MAPS II; Bio-Rad, Richmond, Calif.). D. immitis-derived whole ES products (100 μg) were incubated with each IgG fraction (200 μg) at 4°C for 8 h, and the mixture was incubated with protein A-immobilized Sepharose beads at 4°C for 16 h. After incubation, supernatants were collected by centrifugation, passed through a column of polymyxin B-immobilized beads, lyophilized, and stored at −30°C until use. The DiAg fraction in DiAg-depleted ES products was not detected by Western blotting.

Mice.

Male BALB/c (H-2^d) mice, lipopolysaccharide (LPS) nonresponder C3H/HeJ (H-2^k) mice, and the wild-type counterpart of the latter, C3H/HeN (H-2^k) mice, were obtained from Clea Japan (Tokyo, Japan). All mice were rested for a week after arrival at the facility and used at 8 to 12 weeks of age.

In vivo treatment.

Administration of rDiAg into mice was carried out using the ALZET micro-osmotic pump (Alza 1002; Alza Co., Palo Alto, Calif.), which is capable of releasing the test compound for 2 weeks (0.25 μl/h). Mice received s.c. implants of a micro-osmotic pump filled with rDiAg, rCont, whole ES, native DiAg-depleted ES products. or ovalbumin (OVA, grade VII; Sigma) (100 μg/100 μl in PBS). Plasma and serum were harvested from individual mice on days 7, 14, 21, and 28 after implantation and were used to determine IgE and cytokine levels, respectively (see below).

Western blotting analysis.

rDiAg was separated in SDS-15% PAGE gels under reducing conditions and transferred to polyvinylidene difluoride membranes (Millipore, Bedford, United Kingdom). The protein blots were blocked with Block Ace (Dainihon Pharm. Co., Osaka, Japan) at room temperature for 30 min and incubated with mouse (on day 21 after implantation) or rabbit (diluted 1:100) sera at 4°C for 16 h. After washing, the blots were incubated with horseradish peroxidase (HRP)-conjugated anti-mouse IgE (diluted 1:500; GAM/IgE [Fc] PO; Nordic Immunological Laboratories B.V., Tilburg, The Netherlands), anti-mouse IgG (diluted 1:500; A90-131P, Bethyl Laboratories, Inc., Montgomery, Tex.), or HRP-conjugated anti-rabbit IgG (diluted 1:1,000, Sigma) at room temperature for 2 h. Bands were revealed by the POD Immunostain Set (Wako).

Cell preparation.

B cells were freshly prepared from spleen and peritoneal cavity samples of naive or osmotic pump-implanted mice (on days 7 and 21 after implantation). Briefly, whole spleen cell and peritoneal cell suspensions were placed on 25-mm² tissue culture flasks (Falcon IBD, Lincoln Park, N.J.) at 37°C for 1 h to deplete adherent cell fractions contaminating macrophages (repeated two or three times). Nonadherent cell fractions (used as macrophage-depleted spleen cells) were treated with rabbit anti-mouse Thy-1.2 antibody (diluted 1:20; Cedarlane Laboratories, Hornby, Ontario, Canada) at 4°C for 1 h and subsequently with Low-Tox-M rabbit complement (diluted 1:10; Cedarlane Laboratories) at 37°C for 1 h to deplete T-cell fractions. After washing, the residual cells were treated with rat anti-mouse B220 MAb conjugated with magnetic beads (diluted 1:10; Miltenyi Biotec, Auburn, Calif.) at 4°C for 15 min. Magnetically labeled cells were applied onto separation columns in a magnetic field, and then the column was washed with PBS containing 0.5% fetal calf serum (FCS) and 2 mM EDTA (isolation buffer) to remove negative cell fractions. The column was taken away from the magnetic field, and positive cell fractions were recovered by washing the column with isolation buffer. The positive cell fractions were used as B cells. This procedure routinely yields cells that are >98% surface IgM⁺ and <1% CD3⁺.

To prepare the T-cell-depleted fraction, whole spleen cells were directly treated with the combination of rabbit anti-mouse Thy-1.2 antibody (diluted 1:20; Cedarlane Laboratories) and rabbit complement (diluted 1:10; Cedarlane Laboratories). To deplete B cells, whole spleen cells were treated with rat anti-mouse B220 MAb conjugated with magnetic beads (1:10; Miltenyi Biotec) and then applied onto separation columns, as described above. After washing of the column, negative cell fractions were used as B-cell-depleted spleen cells. This procedure routinely yields T-cell- and B-cell-depleted spleen cells that are <5% CD3⁺ and <3% surface IgM⁺, respectively. Each of the cell populations was suspended in RPMI 1640 supplemented with 10% heat-inactivated FCS, penicillin G (100 U/ml), streptomycin (100 μg/ml), 2 mM glutamine, 50 μM 2-mercaptoethanol, and 0.5 mM sodium pyruvate, applied in all subsequent experiments.

Assay for cell proliferation.

Each of the cell populations (2 × 10⁶ cells/ml) from naive mice was incubated with various doses of rDiAg, rCont, or LPS (5 μg/ml; from E. coli serotype O55 B5; Sigma) for several hours in a total volume of 100 μl in 96-well flat-bottom plates (Corning, Cambridge, Mass.). Pretreatment of rDiAg with polymyxin B sulfate (10 μg/ml; Sigma) was performed at 37°C for 1 h, and then the mixture was applied to the cell cultures. The cell proliferation was measured by a bromodeoxyuridine (BrdU) incorporation assay (BrdU enzyme-linked immunosorbent assay [ELISA]; Boehringer Mannheim, Montreal, Canada). The BrdU ELISA was performed according to the manufacturer's instructions. Briefly, the cells were pulsed with 10 μl of BrdU solution (100 μM in PBS) for the last 6 h. After cultivation, plates were centrifuged at 300 × g for 10 min, and cells were fixed with 200 μl of FixDenat solution for 30 min. One hundred microliters of peroxidase-conjugated mouse anti-BrdU MAb (diluted 1:100) was added into each well. After incubation for 1 h, 100 μl of substrate solution was applied to each well. The colorimetric reaction was stopped after 15 min by addition of 100 μl of 2 N H₂SO₄. The optical density was measured at 450 nm, a test wavelength, and at 690 nm, a reference wavelength, using a micro plate reader. The blank well corresponded to 100 μl of culture medium with BrdU.

Assay for total and nonspecific or specific IgE production.

To determine IgE production by B cells in vitro, in the coculture system, splenic and/or peritoneal B cells (each at 10⁶ cells/ml) from naive mice were cultured with rDiAg (10 μg/ml), rCont (10 μg/ml), rat anti-mouse CD40 MAb (1 μg/ml; HM40-3, no NaN₃ and low endotoxin content; PharMingen, San Diego, Calif.) and/or IL-4 (200 U/ml; Chemicon International Inc., Temecula, Calif.) in a total volume of 1 ml in 24-well flat-bottom plates (Corning). In the Transwell system, splenic B cells were applied to the lower compartment and peritoneal B cells were added within the upper compartment of the Transwell culture plates (Corning). Culture supernatants were harvested after 8 days of culture. IgE levels in plasma samples or in supernatants were measured by ELISA as follows.

For total IgE levels, 96-well plates were coated with 100 μl of anti-mouse IgE MAb (5 μg/ml in 50 mM carbonate buffer, pH 9.5; LO-ME-2; Experimental Immunology Unit, Brussels, Belgium), as the capture antibody, overnight at 4°C. Plates were then blocked with Block Ace (diluted 1:4; Dainihon Pharm. Co.). Diluted plasma samples (diluted 1:20) or culture supernatants were applied to all wells and incubated for 1 h at 37°C. One hundred microliters of HRP-conjugated goat anti-mouse IgE (diluted 1:10,000; GAM/IgE [Fc] PO; Nordic Immunological Laboratories B.V.), as the detection antibody, was added to each well. After incubation for 1 h at 37°C, the reaction mixtures were visualized with 100 μl of O-phenylenediamine substrate (1 mg/ml in 0.1 M potassium citrate, pH 4.5, containing 0.0125% hydrogen peroxide) (Wako), and the colorimetric reaction was stopped with 100 μl of 2 N H₂SO₄.

For nonspecific and specific IgE levels, 100 μl of biotinylated rDiAg (10 μg/ml) was added to avidin-coated plates and incubated for 30 min at room temperature. Diluted plasma samples were applied to each well. After incubation for 1 h at 37°C, the plasma samples were aspirated and transferred into anti-mouse IgE MAb-coated plates. The subsequent procedures were performed as described above. The optical density was determined at 492 nm. IgE concentration was determined in comparison to a standard curve generated by serial dilutions of monoclonal mouse antidinitrophenol IgE (ranging from 0.1 to 1,000 ng/ml; M-IgE; Yamasa, Chiba, Japan). The lower limit of detection in the ELISA was 100 pg/ml for IgE.

Analysis for expression of CD23 and IL-10 mRNA.

B cells (5 × 10⁶ cells/ml) from naive mice were cultured with rDiAg (10 μg/ml) and/or IL-4 (10 U/ml) or rCont (10 μg/ml) in 24-well flat-bottom plates for 24 h. Total RNA was prepared from splenic B cells by the RNeasy Mini Kit (Qiagen, Chatsworth, Calif.), and then cDNA was prepared and amplified using the RT-PCR high (TOYOBO). To eliminate contaminated DNA, total RNA was pretreated with 0.075 U of DNase I (GIBCO, Grand Island, N.Y.) in the recommended buffer. Reverse transcription of the purified total RNA was carried out in a mixture containing total RNA (10 ng/μl), 1 mM deoxynucleoside triphosphate, 1.25 μM random primer, 0.5 U of RNase inhibitor, 1 U of RT, and the recommended buffer. cDNA was added at 0.4 μM each primer, and 0.025 U of recombinant Taq DNA polymerase was added per μl in the recommended buffer. The primer sequences used were as follows: IL-10, sense chain, 5′-CGGGAAGACAATAACTG-3′, and antisense chain, 5′-CATTTCCGATAAGGCTTGG-3′; CD23, sense chain, 5′-TGGCAAAGCTGTGGATAGAG-3′, and antisense chain, 5′-CGACCATACAAACTCTCCCT-3′; housekeeping gene, glyceraldehyde 3-phosphate dehydrogenase (GAPDH), sense chain, 5′-ACCACAAGTCCATGCCATCAC-3′, and antisense chain, 5′-TCCACCACCCTGTTGCTGTA-3′. The PCR was carried out at 95°C for 5 min before 30 cycles of 95°C for 30 s, 56°C for 30 s, and 72°C for 1 min, followed by 72°C for 9 min. The PCR products were separated by agarose gel electrophoresis and detected by ethidium bromide staining.

Flow cytometric analysis.

To examine the expression of CD23 on splenic B cells, single immunofluorescence staining was performed. Splenic B cells (10⁶ cells/ml) from naive mice were cultured with rDiAg (10 μg/ml) for 48 h in a total volume of 1 ml in 24-well flat-bottom plates. The collected B cells (10⁶ cells) were treated with fluorescein isothiocyanate-conjugated rat anti-mouse CD23 MAb (1 μg/ml; B3B4; PharMingen) in 50 μl of PBS containing 1% FCS and 0.1% NaN₃ (staining buffer) for 30 min at 4°C. After washing, the cells were resuspended in staining buffer and then analyzed on a FACScalibur device (Becton Dickinson, San Jose, Calif.). Analysis of data was performed using CellQuest software (Becton Dickinson). Results are represented as mean relative ratios versus the control level of 1.0.

Quantitation of cytokine protein.

B cells (5 × 10⁶ cells/ml) from naive mice were incubated with rDiAg (10 μg/ml) and/or IL-4 (10 U/ml), rCont (10 μg/ml), or LPS (5 μg/ml) for 48 h. The levels of cytokine released in culture supernatants or in serum were quantitated using the cytokine ELISA kit (for IL-4 and IL-10; Endogen, Woburn, Mass.; for gamma interferon [IFN-γ]; Genzyme-Techne, Cambridge, Mass.) according to the manufacturer's instructions.

Neutralizing IL-10 treatment.

Rat anti-mouse IL-10 MAb (JES5-2A5, no NaN₃ and low endotoxin content; Genzyme-Techne) or an IgG1 isotype-matched control MAb (R3-34; PharMingen) was applied in the experiment for IgE synthesis in vitro. Each antibody (5 μg/ml) was added simultaneously with rDiAg (10 μg/ml) and IL-4 (200 U/ml) into the B-cell cultures. After cultivation for 8 days, IgE levels in culture supernatants were assessed as described above. IL-10 was not detected in any cultures treated with anti-IL-10 MAb, demonstrating that the cytokine had been successfully neutralized.

Statistical analysis.

The statistical significance of values obtained was evaluated by using Student's t test. A confidence level of <0.05 was considered significant.

RESULTS

rDiAg preferentially induces nonspecific IgE production in vivo.

To determine whether rDiAg induces nonspecific IgE production, rDiAg was s.c. administered to BALB/c mice through a micro-osmotic pump. The level of total IgE was already significant at 14 days and was maximal at 21 days after administration of rDiAg (Fig. 1A). In mice administered PBS or rCont, no increases in total IgE levels were detected throughout the experimental periods (<50 ng/ml). On day 21 after administration of rDiAg, nonspecific IgE production presented a pattern similar to that of the total IgE production (Fig. 1B), while rDiAg-specific IgE production was not induced (Fig. 1C). The Western blotting analysis also showed that rDiAg was not recognized by serum IgE and IgG from mice that had been administered rDiAg (Fig. 2B).

FIG. 1.

Kinetics of plasma IgE levels in mice administered rDiAg. BALB/c mice received s.c. implants of a micro-osmotic pump filled with rDiAg, rCont, native DiAg, whole ES, DiAg-depleted ES (DiAg⁻ES), or PBS (each at 100 μg/head), and plasma was collected weekly. Total (A and D), nonspecific (B), and specific (C) IgE levels in plasma samples were measured by ELISA. The data represent the mean values and standard deviations (error bars) from five independent experiments. ∗, P < 0.01, when compared with mice administered PBS or rCont.

To further confirm that DiAg is an inducer of nonspecific IgE, we next prepared DiAg-depleted ES products and assessed the effect of these products on IgE production. Native DiAg induced IgE production at a level similar to rDiAg (Fig. 1A and D). Whole ES products stimulated IgE synthesis in mice, whereas DiAg-depleted ES products only induced a weak production of total IgE (Fig. 1D). Mice administered OVA showed negligible levels of total IgE, even on day 21 after implantation (<50 ng/ml). rDiAg-specific IgE production was not detected in plasma from any mouse tested (data not shown). These results indicate that DiAg is a potent inducer of nonspecific IgE production.

rDiAg favors increases in Th2-type cytokine levels in serum.

It is well-known that IL-4, a Th2-type cytokine, is an inducer of IgE production, while IFN-γ, a Th1-type cytokine, has an inhibitory effect (4). IL-10, another Th2-type cytokine, has been shown to enhance IL-4-induced nonspecific IgE production and to inhibit IFN-γ production (13, 23, 35, 50). To determine cytokine responses induced by rDiAg stimulation, we monitored levels of these three cytokines in serum from BALB/c mice administered rDiAg. IL-4 concentration increased and peaked around 7 to 14 days after implantation (Fig. 3A). IL-10 first became detectable at 7 days after rDiAg administration and reached a peak at 21 days (Fig. 3B). In contrast to Th2-type cytokines, IFN-γ concentration increased transiently at 7 days and then decreased up to the baseline level (Fig. 3C). In PBS-administered mice, the production of these cytokines remained at background levels observed on day 0. These results suggest that rDiAg-induced nonspecific IgE production is dependent not only on up-regulation of IL-4 concomitant with down-regulation of IFN-γ but also on up-regulation of IL-10.

FIG. 3.

Kinetics of cytokine levels in sera from mice administered rDiAg. BALB/c mice received s.c. implants of a micro-osmotic pump filled with rDiAg (100 μg/head) or PBS, and serum was collected weekly. IL-4 (A), IL-10 (B), and IFN-γ (C) levels in serum were measured by ELISA. ∗, P < 0.01, when compared with mice administered PBS.

rDiAg directly stimulates B-cell proliferation.

We first tested whether rDiAg has the ability to stimulate proliferative response of whole spleen cells from naive BALB/c mice. rDiAg significantly induced the proliferation of spleen cells (Fig. 4). To determine which population of spleen cells was responding to rDiAg, each of the cell populations was depleted. B-cell-depleted spleen cells did not respond to rDiAg, while T-cell-depleted populations did respond well. The depletion of macrophages from spleen cells did not affect the proliferative response to rDiAg (Fig. 4). These results indicate that the responder cells to rDiAg are B cells.

FIG. 4.

Proliferation of spleen cells in response to rDiAg is dependent on the presence of B cells. Whole spleen cells or T-cell-, B-cell-, or macrophage-depleted spleen cells (2 × 10⁶/ml) from naive BALB/c mice were cultured with rDiAg (10 μg/ml) for 48 h. Cell proliferation was measured by the BrdU incorporation assay. The data represent the mean values and standard deviations (error bars) from five independent experiments. ∗, P < 0.05, when compared with medium alone.

To clarify whether the molecule directly elicits B-cell proliferation, naive splenic B cells were stimulated with various concentrations of rDiAg. rDiAg induced B-cell proliferation in a dose-dependent manner, with a bell-shaped response curve (starting at 1 μg/ml, increasing at 10 μg/ml, and peaking at 100 μg/ml) (Fig. 5A). Also, B-cell proliferation by rDiAg (10 μg/ml) was similar to that by native DiAg (10 μg/ml). Thus, we applied 10 μg of rDiAg/ml in all subsequent experiments. We subsequently examined the time course of the proliferative response of splenic B cells. The proliferation of B cells reached a peak at 48 h after rDiAg stimulation (Fig. 5B), and this response was comparable to the LPS-induced response (Fig. 5A). Although similar effects of rDiAg were obtained in peritoneal B cells (Fig. 5A and B), the proliferative capacity was lower in peritoneal B cells than in splenic B cells. These results suggest that rDiAg acts as a B-cell mitogen.

FIG. 5.

rDiAg induces the proliferative response of B cells. (A) Splenic and peritoneal B cells (2 × 10⁶/ml) from naive BALB/c mice were cultured with various concentrations of rDiAg, native DiAg (10 μg/ml), or LPS (5 μg/ml) for 48 h. (B) B cells were cultured with or without rDiAg (10 μg/ml) for various periods. Cell proliferation was measured by the BrdU incorporation assay. The data represent the mean values and standard deviations (error bars) from five independent experiments. ∗, P < 0.05, when compared with medium alone.

Mitogenic activity is not due to bacterial molecules.

It is widely known that bacterial LPS is a mitogen for mouse B cells. To exclude the possibility that the mitogenic activity of rDiAg may be due to LPS contamination, rDiAg was pretreated with polymyxin B, an inhibitor of LPS, for 1 h, and this mixture was added into the splenic B-cell cultures. LPS-induced B-cell proliferation was significantly inhibited by addition of polymyxin B, whereas rDiAg-induced proliferation was not prevented with this inhibitor (Fig. 6A). Moreover, B cells from naive C3H/HeJ mice, which lack the response to LPS, retained the responsiveness to rDiAg, but not to LPS. On the other hand, splenic B cells from naive C3H/HeN wild-type mice responded well to both rDiAg and LPS (Fig. 6B), indicating that LPS contamination is not responsible for the mitogenic activity. It has recently been demonstrated that other bacterial molecules, such as CpG oligonucleotides and lipoproteins, are also mitogenic for B cells (28, 31). To further rule out the role of these bacterial molecules in B-cell proliferation, splenic B cells were cultured with rCont prepared in the same procedure as rDiAg. rCont failed to proliferate splenic B cells from both naive BALB/c and C3H mice (Fig. 6). Together, these results excluded the potential role played by bacterial molecules in rDiAg-induced B-cell proliferation.

FIG. 6.

Splenic B-cell activation induced with rDiAg is not due to the contamination of LPS. (A) Splenic B cells (2 × 10⁶/ml) from naive BALB/c mice were cultured with rDiAg (10 μg/ml), rCont (10 μg/ml), or LPS (5 μg/ml) in the presence or absence of polymyxin B (10 μg/ml) for 48 h. rDiAg, rCont, and LPS were pretreated with polymyxin B at 37°C for 1 h before being applied to the cultures. (B) Splenic B cells (2 × 10⁶/ml) from naive C3H/HeN or C3H/HeJ mice were cultured with various stimuli as described above. The data represent the mean values and standard deviations (error bars) from five independent experiments. ∗, P < 0.05, when compared with medium alone or rCont.

rDiAg induces the early differentiation of splenic B cells.

B cells undergo quantitative and qualitative phenotypic changes concomitant with differentiation. CD23 is not expressed on resting conventional B cells but is dramatically induced in the early differentiation stage of B cells following primary stimulation, such as IL-4 (20, 25). Thus, CD23 has been thought to be a key marker of early differentiation in conventional B cells. To further investigate the contribution of B cells in IgE production induced with rDiAg, we next assessed the effect of rDiAg on the expression of CD23 mRNA and surface protein in splenic B cells from naive BALB/c mice in vitro. CD23 mRNA expression was significantly up-regulated within 24 h after rDiAg stimulation (Fig. 7A). When splenic B cells were stimulated with rDiAg for 48 h, the expression of CD23 surface protein on their cells was enhanced (Fig. 7B). As expected, IL-4 significantly caused CD23 expression on splenic B cells. The combination of rDiAg and IL-4 (rDiAg-IL-4) resulted in an additive increase in the level of CD23 expression. rCont failed to induce the expression of CD23 mRNA, and IL-4 was not detected in rDiAg-stimulated B-cell cultures (data not shown). These results suggest that rDiAg directly induces not only proliferation but also differentiation of conventional B cells.

FIG. 7.

rDiAg induces CD23 expression in splenic B cells. (A) Splenic B cells (5 × 10⁶/ml) from naive BALB/c mice were cultured with or without rDiAg (10 μg/ml) for 24 h, and CD23 mRNA expression was detected by RT-PCR. The housekeeping gene GAPDH was amplified as an internal control. The data represent one typical experiment out of five. (B) Splenic B cells (10⁶/ml) were cultured with or without rDiAg (10 μg/ml) for 48 h and were then stained with fluorescein isothiocyanate-conjugated rat anti-mouse CD23 MAb. Fluorescence was analyzed with a FACScalibur device. The data represent the relative ratio, which is equal to the mean fluorescence of rDiAg-treated cells divided by the mean fluorescence of untreated cells, and standard deviations (error bars) from five independent experiments. ∗, P < 0.05, when compared with medium alone.

rDiAg synergizes with IL-4 in IgE class switching.

From these results, it is possible that rDiAg directly induces nonspecific IgE synthesis through the polyclonal expansion and differentiation of B cells. To investigate this, we cultured naive B cells with rDiAg-IL-4 for 8 days. rDiAg-IL-4 induced significant IgE production in splenic B cells (Table 1). The IgE level was comparable to the amount induced with anti-CD40 MAb plus IL-4 (24.5 ± 0.2 ng/ml). Neither rDiAg alone nor IL-4 alone was able to induce significant IgE production. Similar patterns were also observed in peritoneal B cells, but at very low levels compared with splenic B cells (Table 1). rCont did not induce IgE synthesis by both splenic and peritoneal B cells, even in the presence of IL-4 (3.2 ± 0.4 ng/ml and 1.5 ± 0.3 ng/ml, respectively). These results suggest that rDiAg exerts itself as a costimulatory molecule in IgE class switching.

TABLE 1.

rDiAg and IL-4 synergize to induce IgE production by B cells^a

B-cell source	Mean IgE production (ng/ml) ± SDb in:
	Medium	rDiAg	IL-4	rDiAg-IL-4
Spleen	1.6 ± 1.0	2.8 ± 0.4	3.1 ± 0.5	20.4 ± 3.2*
Peritoneum	1.1 ± 0.2	1.4 ± 0.3	1.6 ± 0.8	6.1 ± 1.7*
Spleen + peritoneum	2.5 ± 0.7	3.8 ± 1.1	4.1 ± 1.5	38.6 ± 3.4*,**

^aNaive splenic or peritoneal B cells (10⁶ cells/ml) or both were cultured with rDiAg (10 μg/ml) and/or IL-4 (200 U/ml) for 8 days. IgE levels in culture supernatants were measured by ELISA.

^bThe mean and standard deviations are from five independent experiments. Statistical significance: *, P < 0.05, when compared with medium alone; **, P < 0.05, when compared with rDiAg-IL-4-stimulated splenic B cells.

rDiAg stimulates IL-10 production by B cells.

IL-10 is produced by B cells, including conventional B cells and B-1 cells (16, 37). Peritoneal B cells are known to be enriched for B-1 cells (18). We therefore attempted to examine the effect of rDiAg on IL-10 production by splenic and peritoneal B cells. To determine whether rDiAg induces IL-10 synthesis by B cells, we analyzed the expression of IL-10 mRNA and protein in rDiAg-stimulated B cells. Both splenic and peritoneal B cells stimulated with rDiAg significantly synthesized IL-10 mRNA as well as biologically active IL-10 protein (Fig. 8). LPS-induced IL-10 production was higher in peritoneal B cells than in splenic B cells (354.2 ± 23.5 pg/ml and 138.1 ± 14.5 pg/ml, respectively). IL-10 production in response to rDiAg was also significantly greater in peritoneal B cells than in splenic B cells. IL-4 induced only low levels of IL-10 production by both B-cell subsets and additionally enhanced IL-10 production induced with rDiAg. rCont did not induce the expression of IL-10 mRNA in both B cells (data not shown). These results indicate that rDiAg also possesses cytokine stimulatory property.

FIG. 8.

rDiAg induces IL-10 production by B cells. Splenic or peritoneal B cells (5 × 10⁶/ml) from naive BALB/c mice were stimulated with rDiAg (10 μg/ml) and/or IL-4 (10 U/ml) for 24 h (for RT-PCR) or 48 h (for ELISA). (A) Total RNA from B cells was analyzed for IL-10 mRNA expression by RT-PCR. The housekeeping gene GAPDH was amplified as an internal control. Data represent one typical experiment out of five. (B) IL-10 production in culture supernatants was measured by ELISA. The data represent the mean values and standard deviations (error bars) from five independent experiments. N.D., not detected; ∗, P < 0.05, when compared with medium alone.

Peritoneal B cells enhance rDiAg-induced IgE production by splenic B cells.

Because IL-10 can enhance nonspecific IgE production induced with the combination of IL-4 plus anti-CD40 MAb (23, 35, 50), it was possible that the nonspecific IgE production seen in rDiAg-administrated mice was dependent on IL-10 production by peritoneal B cells. To examine this possibility, naive splenic B cells were cocultured with naive peritoneal B cells in the presence of rDiAg-IL-4 for 8 days. IgE production by splenic B cells induced with rDiAg-IL-4 was synergistically enhanced in the presence of peritoneal B cells (Table 1). Similar results were also obtained in the case of the Transwell culture system (data not shown). These findings suggest that rDiAg-induced IL-10 production by peritoneal B-1 cells may contribute to nonspecific IgE production.

IL-10 is involved in rDiAg-induced nonspecific IgE production.

To confirm that endogenous rDiAg-driven IL-10 secretion by peritoneal B cells is involved in the IgE production, B cells stimulated with rDiAg-IL-4 were cultured with or without a neutralizing anti-IL-10 MAb for 8 days. IgE production induced by rDiAg-IL-4 in splenic B cells was partly prevented by addition of the anti-IL-10 MAb but not by a control MAb (Fig. 9). In addition, the anti-IL-10 MAb inhibited by >50% the enhancement of IgE synthesis by splenic B cells cocultured with peritoneal B cells. In the presence of the neutralizing MAb, the level of IgE produced by splenic B cells was consistent with that produced by splenic B cells cocultured with peritoneal B cells (splenic B cells-peritoneal B cells). These results indicate that autocrine and paracrine production of IL-10 is an important step in the augmentation of nonspecific IgE synthesis by splenic B cells induced with rDiAg-IL-4.

FIG. 9.

Anti-IL-10 antibody inhibits rDiAg-induced IgE synthesis in vitro. Naive splenic, peritoneal B cells (10⁶/ml) or both were cultured with the combination of rDiAg (10 μg/ml) and IL-4 (200 U/ml) in the presence of anti-IL-10 or isotype-matched control MAb (5 μg/ml). IgE levels in culture supernatants were measured by ELISA. The data represent the mean values and standard deviations (error bars) from five independent experiments. ∗∗ and ∗, P < 0.01 and P < 0.05, respectively, when compared with anti-IL-10 MAb culture.

DISCUSSION

It has been shown that nonspecific IgE is induced in rats infected with N. brasiliensis as well as in mice injected with its soluble crude or ES products (11, 22, 49). Lee and McGibbon demonstrated that Ascaris pseudocoelomic body fluid-treated mice display an increased IgE production in response to a third-party antigen (29), suggesting that Ascaris pseudocoelomic body fluid, including Ascaris body fluid allergen-1 (ABA-1; the NPA from the nematode Ascaris suum), induces nonspecific IgE production. We have previously shown that native DiAg, an NPA from D. immitis, polyclonally induces IgE production by human spleen cells in vitro (54), but the molecular basis for this effect has not yet been defined completely. To investigate this, first, we prepared a recombinant protein of DiAg. Second, we used a micro-osmotic pump, but not an adjuvant, to administer rDiAg into mice, because the adjuvant itself has the ability to induce Th1/Th2 responses as well as to preferentially amplify specific antibody production independent of the property of antigens, including NPAs (e.g., ABA-1 and gp15/400 of Brugia species) (2, 5, 7). Indeed, it has not yet been clarified whether NPAs other than DiAg have the ability to induce nonspecific IgE synthesis, because adjuvants have been used to repetitively administer the molecules into mice. We demonstrated that rDiAg preferentially induces nonspecific IgE production, which reaches a peak at 21 days after implantation. This result is consistent with a previous report that markedly elevated levels of nonspecific IgE are seen in the early phase of N. brasiliensis infections (22). The depletion of DiAg from whole ES products failed to induce significant levels of nonspecific IgE. These findings suggest that DiAg is a crucial factor in generating high nonspecific IgE responses in D. immitis-infected hosts.

Both in vivo and in vitro, IgE synthesis by B cells essentially requires at least two signals: IL-4 and CD40 ligand on activated T cells (4). In fact, mice administered rDiAg presented increased serum IL-4 levels with decreased serum IFN-γ levels throughout the experimental period. These results indicate that rDiAg is a selective activator of Th2 responses. We next examined whether rDiAg can directly induce IgE production by naive splenic B cells. As expected, neither rDiAg nor IL-4 alone induced significant levels of IgE production. However, IgE production by splenic B cells was significantly induced with rDiAg-IL-4. These results indicate that IL-4 is essentially required for nonspecific IgE production seen in rDiAg-administrated BALB/c mice and that rDiAg may act as a costimulatory molecule in the induction of IgE class switching.

B cells are polyclonally stimulated in mice infected with helminths, such as the trematode Schistosoma mansoni and the nematode Angiostrongylus cantonensis (14, 48). Soluble crude extracts from A. suum, the nematode Toxocara canis, and the cestode Taenia solium possess B-cell mitogenic activity (30, 47, 53). We found that rDiAg has the ability to proliferate splenic B cells as effectively as LPS, suggesting that rDiAg may also act as a mitogen to B cells. The possibility that B-cell expansion by rDiAg is due to the presence of residual LPS was excluded for three reasons. First, C3H/HeJ mice, which lack the B-cell responses to LPS, normally responded to rDiAg as well as C3H/HeN wild-type mice. Second, rDiAg activity was not affected by polymyxin B, an inhibitor of LPS. Third, the peak response of B cells to rDiAg was observed at 48 h after stimulation, whereas proliferative response to bacterial mitogens maximized 120 h later (45). Furthermore, recombinant control protein failed to stimulate B cells, indicating that the B-cell-stimulating capacity of rDiAg is not due to the contamination of other bacterial components, including CpG oligonucleotides and lipoproteins (28, 31). We have previously shown that native DiAg enhances IL-4-induced CD23 expression on human splenic B cells (54). In the present study, we demonstrated that the expression of CD23 mRNA and surface protein in mouse splenic B cells is up-regulated with rDiAg alone. This may be due to differences in the responsiveness to rDiAg between splenic B cells from humans and mice. It has been shown that CD23 is a ligand for CD21 on B cells and suggested that this interaction may be involved in the regulation of IL-4-induced IgE production by human B cells (3). Thus, these findings indicate that, unlike LPS, rDiAg can induce T-cell-independent IgE production and proliferation in splenic B cells.

We found that rDiAg-induced IL-10 production is higher in peritoneal B cells than in splenic B cells. These results can be explained by a report that peritoneal B-1 cells are the main source of B-cell-derived IL-10 (37). IL-10 acts as an autocrine growth factor of B-1 cells and as a survival factor of conventional B cells in mice (17, 21). Thus, IL-10 from B cells stimulated with rDiAg may be partly involved in the proliferation of their cells in an autocrine or paracrine manner. Around 14 to 21 days after rDiAg administration, although serum IL-4 levels showed a tendency to decrease slightly, serum IL-10 levels were increased parallel with nonspecific IgE levels. IgE production induced by rDiAg-IL-4 in splenic B cells was synergistically enhanced by coculture with IL-10-producing peritoneal B cells. The observation that anti-IL-10 MAb inhibited IgE production by splenic B cells-peritoneal B cells induced with rDiAg-IL-4 confirms the critical role that IL-10 plays in the synthesis of nonspecific polyclonal IgE antibody. IL-10 has been shown to enhance IL-4-plus-anti-CD40 MAb-induced nonspecific IgE production by B cells and to prevent the expansion of memory B cells in the germinal center (6, 23, 35, 50). These findings suggest that IL-10-producing B-1 cells may play an important role in the development of Th2-type responses as well as the enhancement of rDiAg-IL-4-induced nonspecific IgE-producing B cells in vivo.

In helminth infections, invading parasites can survive for many years in an immunocompetent host. Such long-term survival of parasites is considered to be associated with the induction of immunological tolerance against parasitic antigens (26, 27, 33, 34). It is well-established that helminth infections induce markedly elevated levels of nonspecific IgE and immunosuppressive cytokines (34, 41). It is possible that nonspecific IgE produced during helminth infections is attributed to the polyclonal expansion of B cells. In fact, there are several reports that a defined parasite-derived molecule interacts directly with B cells to polyclonally activate their cells in vitro. ES-62, a phosphorylcholine-containing glycoprotein, from the filarial nematode Acanthocheilonema viteae and proline racemase from the protozoan parasite Trypanosoma cruzi are mitogenic for conventional B cells (19, 44). In contrast, lacto-N-fucopentaose III, an oligosaccharide present on Schistosoma eggs and larvae, induces outgrowth as well as IL-10 production of peritoneal B-1 cells (51). Therefore, it has been thought that parasite-derived products have B-cell mitogenic activity, and polyclonal lymphocyte activation is a common phenomenon seen during parasite infections (43). On the other hand, IL-10, an immunosuppressive cytokine, strongly down-regulates the expression of major histocompatibility complex class II and B7 molecules on macrophages and eventually suppresses the antigen-specific activation of T cells (8, 9). In fact, IL-10 produced during helminth infections diminishes the worm antigen-specific T-cell proliferative response (26, 32, 42). These findings suggest that polyclonally activated lymphocytes and immunosuppressive cytokines contribute to the immune evasion by the invading parasites, hence prolonging their survival.

DiAg is not recognized by antibodies from naturally D. immitis-infected hosts (1, 46) or by antibodies from mice exposed to the recombinant protein, indicating that this molecule does not possess allergenic or antigenic properties. Thus, DiAg appears to be mimic the behavior of an endogenous factor-like molecule in the host. Taken together, the findings of the present study demonstrated that rDiAg possesses B-cell mitogenic, IL-10-stimulatory, and IgE-costimulatory properties and suggested that rDiAg selectively induces the production of nonspecific IgE through polyclonal expansion and IgE class switching of B cells in the presence of Th2-type cytokines.