Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

The respiratory tract, a representative site of mucosal tissue, is repeatedly exposed to a broad array of airborne foreign allergens (1). Inhalation of allergens evokes deleterious immune and inflammatory responses in the airway, as in bronchial asthma (2). The characteristic features of bronchial asthma include airway eosinophilia and elevated serum immunoglobulin (Ig)E levels, both of which are reported to be orchestrated by type-2 T-helper (Th2) cells (3). Suppression of Th2-dominated immune responses in this setting could be a potential target of immunotherapy for bronchial asthma.

Control of airway eosinophilia by manipulation of Th2 functions has been achieved by inducing immune tolerance of Th2 cells using several distinct approaches. For example, another CD4⁺ T-cell subset, Th1, exerts inhibitory effects on Th2 cells through the secretion of cytokines such as interferon (IFN)- (10), and those induced upon exposure to Mycobacterium tuberculosis inhibit Th2-mediated eosinophilia (11, 12). Transforming growth factor--producing CD4⁺ T cells induced by oral or tracheal tolerance are another subset that inhibits Th2 cells and eosinophilia (13, 14). In these experimental models, the inhibition of airway eosinophilia was accompanied by neutralizing Th2 activities in mediastinal lymph nodes (LNs) (13, 14).

Oligodeoxynucleotides (ODNs) containing CpG motifs (CpG) found in bacterial but not vertebrate DNA activate macrophages and dendritic cells to secrete interleukin (IL)-12 and induce IFN--secreting Th1 cells (15). Systemic administration of CpG biased the immune responses toward a Th1-dominant phenotype and inhibited Th2- mediated responses, including airway eosinophilia (17, 22- 26). In light of the pathogenic roles of Th1 cells in various types of autoimmunity (27), the effects of CpG on the activation of autoreactive T cells and autoimmunity may be unpredictable and controversial; Gilkeson and colleagues (28) and Mor and associates (29) reported that bacterial and plasmid DNA did not deteriorate autoimmunity, whereas Krieg (30) suggested that microbial DNA could be a pathogenic factor. The toxicity of CpG was another problem. Bacterial DNA was reported to increase the toxicity of lipopolysaccharides (LPS) (31). CpG also increased serum tumor necrosis factor (TNF)- levels and mortality in CpG-treated mice (32). These adverse effects would critically hamper the usefulness of CpG as a therapeutic immunomodulator for allergic diseases.

In this study, we administered CpG via the airway mucosa and examined the effects on Th2 cells and airway eosinophilia. We found that concomitant administration of CpG and antigen (as opposed to either one alone) via a tracheal route efficiently inhibited Th2 cells in parallel with the improvement of airway eosinophilia and hyperresponsiveness, while leaving systemic immune responses unaffected. These results provide an experimental basis for the possible therapeutic application of intratracheal administration of allergen plus CpG to patients with bronchial asthma.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Animals and Immunization

BALB/c mice were bred in our animal facility and were used at 5 to 10 wk of age. These animals were primed intraperitoneally with 10 µg of ovalbumin (OVA) (Sigma Chemical Co., St. Louis, MO) precipitated with 4 mg of aluminum hydroxide (alum) in 200 µl of phosphate-buffered saline (PBS) three times at 1-wk intervals. At 7 d after the last immunization, they were challenged intratracheally with OVA either alone or with ODN (Figure 1, protocol 1). Otherwise, the OVA/alum-primed BALB/c mice were pretreated intratracheally with ODN and/or OVA (10 or 0.1 µg), and after 6 d were challenged with OVA (Figure 1, protocol 2).

View larger version (10K): [in this window] [in a new window]   Figure 1.   Experimental protocols. BALB/c mice were primed with three intraperitoneal injections of OVA/alum at 1-wk intervals. In the experiments shown later in Figure 2, protocol 1, mice were challenged intratracheally with OVA either alone or with CpG on Days 21 and 22. BALF samples for cytokines were collected 6 h after the last challenge. Serum, mediastinal LN cells, or BALF for differential cell counts were prepared on Day 24. In other experiments (protocol 2), the OVA/alum-primed mice were instilled intratracheally with OVA and/or CpG on Days 21 and 22, and then challenged with OVA on Day 28. BALF or serum was collected on Day 30.

View larger version (69K): [in this window] [in a new window]   Figure 2.   Inhibition of Th2 cells and airway eosinophilia by antigen challenge with CpG through airway mucosa. BALB/c mice were primed with OVA/alum three times at 1-wk intervals. At 7 d after the last immunization, these mice were challenged with 10 µg of OVA either alone or with 5 µg of CpG or control ODN on two consecutive days ( protocol 1 in Figure 1). At 3 h after the last challenge, BALF for cytokines was collected (B-E ). Otherwise, 2 d after the last challenge, BALF for eosinophils (A) or mediastinal LN cells was prepared. The LN cells enriched for CD4+ T cells were cultured with APCs and OVA (100 µg/ml) for 2 d, and the concentrations of IL-4 (F ) and IL-5 (G) in culture supernatants were determined by ELISA. Intratracheal administration of CpG at the time of antigen challenge inhibited eosinophilia and Th2 cytokines with parallel increases in IFN- and TNF- in BALF. (*P < 0.05; **P < 0.01; and ***P < 0.001.)

CpG and Control ODNs

The ODNs were designed using published sequences (26). The CpG ODN (1826) consisted of 20 bases containing two CpG motifs (TCCATGACGTTCCTGACGTT). The control ODN (1745) was identical except that the CpG motifs were rearranged (TCCATGAGCTTCCTGAGCTT). Phosphorothioate ODNs were synthesized and purified using HPLC by Nihon Gene Research Laboratories, Inc., Sendai, Japan.

Bronchoalveolar Lavage

The lungs of BALB/c mice primed and challenged with OVA as described above were lavaged twice with PBS (0.25 and 0.20 ml each time) and approximately 0.4 ml of the instilled fluid was consistently recovered. The bronchoalveolar lavage fluid (BALF) was cytospun onto microscope slides and stained with Diff-Quik (International Reagents Corp., Kobe, Japan). Differential cell counts in BAL cells were done with at least 300 leukocytes. For cytokine measurement, BALF was centrifuged, and the supernatants were assayed by enzyme-linked immunosorbent assay (ELISA).

Histologic Evaluation

For photomicrographs, cryosections from the frozen trachea tissue were stained as described previously (13). In brief, 2 d after the last challenge the excised trachea was fixed in 10% formalin and frozen in OCT compound (Miles Laboratories, Naperville, IL). Cryosections (8 µm thick) from the frozen tissue were stained with Diff-Quik. A light-microscopic image of the trachea was captured using Light microscope Axioplan II and TV camera progress 3200 (Zeiss, Oberkochen, Germany).

Cell Culture

BALB/c mice were primed and challenged according to the protocols in Figure 1. At 2 d after the last challenge, mediastinal LN cells were pooled from three mice of each group and cultured in the presence of OVA (100 µg/ml). Where indicated, the CD4⁺ T-cell fraction of the LN cells was cultured with mitomycin C (Wako Chemical Industries, Ltd., Osaka, Japan)-treated spleen cells of BALB/c mice as antigen-presenting cells (APCs). The CD4⁺ fraction was prepared by panning methods as described previously (12, 35). The cultures were incubated in quadruplicate for 2 d in 96-well plates, and the culture supernatants were assayed for cytokines.

Anti-OVA IgE Assay

A 1:10 dilution of sera was applied to 96-well microtiter plates (Becton Dickinson, Oxnard, CA) coated with anti-IgE monoclonal antibody (5 µg/ml) (Experimental Immunology Unit, Brussels, Belgium). Bound IgE antibodies were incubated with biotinylated OVA, followed by peroxidase-labeled streptavidin (Vector Laboratories, Burlingame, CA) and tetramethylbenzidine reagent (Kierkegaard & Perry Laboratories, Gaithersburg, MD). Optical densities were determined at 450 nm, and converted to arbitrary units (U/ml) according to the standard curve obtained with serial dilutions of pooled serum from OVA-hyperimmunized BALB/c mice. The conjugation to OVA of biotin with aminocaproyl spacer (Sigma) was done in our laboratory.

Cytokine Assay

Cytokine concentrations in BALF, culture supernatant, or serum were determined by using ELISA according to the manufacturer's recommendations. Paired anti-IL-4, anti- IL-5, anti-IFN-, and anti-TNF- monoclonal antibodies were purchased from PharMingen (San Diego, CA). Standard recombinant mouse IL-4, IL-5, IFN-, and TNF- were purchased from Genzyme Co. (Cambridge, MA).

Airway Hyperresponsiveness

The OVA/alum-primed BALB/c mice were treated with either PBS or CpG plus OVA (0.1 µg) and then challenged with OVA (Figure 1, protocol 2). After 2 d of OVA challenge, airway responsiveness was assessed as a change in pulmonary resistance (RL) after injections of increasing doses of methacholine chloride (MCh) (Wako) (0.1 to 30 mg/kg), as described elsewhere (36). Mice were anesthetized by intraperitoneal injection of pentobarbital sodium (Wako) (50 mg/kg) and were tracheostomized. The airflow rate at the airway opening during spontaneous breathing was monitored by a pneumotachogram (8430B; Hans Rudolph, Kansas City, MO) combined with a differential pressure transducer (LCVR, 0 to 2 cm H₂O; Celesco, Canoga Park, CA). The esophageal pressure monitored by a water-filled tube and a pressure transducer (Ohmeda, Singapore) was used as the transpulmonary pressure, because the pressure difference generated by the pneumotachogram connected to the tracheal tube was very small (< 1/100) compared with the amplitude of the esophageal pressure. RL was calculated by the subtraction method of Mead and Whittenberger (37). An average RL of three breaths at 3 min after each injection of MCh was calculated and expressed as a percentage of the baseline RL that was measured and calculated in the same way after the injection of saline used as a diluent of MCh. The provocative concentration of methacholine in milligrams per kilogram that caused a 200% increase in RL, designated PC200, was calculated by linear interpolation of the appropriate dose-response curves.

Statistics

Data are expressed as means ± standard error of the mean. For in vivo experiments, each group consisted of four or five mice. Each experiment was repeated at least twice. Student's t test was used in the analysis of the results.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Inhibition of Th2 Cells and Airway Eosinophilia by Antigen Challenge with CpG through Airway Mucosa

We examined the transmucosal effects of CpG on airway eosinophilia. Airway eosinophilia was induced in OVA/ alum-primed BALB/c mice by challenging with two doses of OVA via an intratracheal route (Figure 1, protocol 1). A total of 10 µg of OVA as an optimal challenge dose was used on the basis of our previous experiments (14). The airway eosinophilia was inhibited by concurrent administration of CpG at the time of intratracheal antigen challenge (Figure 2A). This inhibition was specific for CpG inasmuch as the control ODN, which did not contain a CpG motif, failed to inhibit airway eosinophilia (Figure 2A). Histologic analysis of the trachea also showed that CpG treatment attenuated the eosinophilic inflammation in the trachea (Figure 3).

View larger version (52K): [in this window] [in a new window]   Figure 3.   Inhibition of tracheal eosinophilia by CpG. Groups of mice were primed and challenged as in Figure 2A. After 2 d of challenge, the trachea was fixed, cryosectioned, and stained. OVA challenge induced tracheal eosinophilia (A), whereas coinstillation of CpG with OVA challenge inhibited eosinophil infiltration (B).

The extent of airway eosinophilia was in parallel with the Th2 cytokine levels in the BALF. In comparison with the control mice, which exhibited strong airway eosinophilia, those challenged with OVA plus CpG showed reduced levels of IL-4 (Figure 2B) or IL-5 (Figure 2C). In contrast, the level of Th1 cytokine IFN- was enhanced by OVA plus CpG treatment (Figure 2D). The TNF- level in BALF was similarly increased (Figure 2E). The cytokine levels in BALF from the mice challenged with OVA plus control ODN were comparable to those of the control group (Figures 2B-2E).

The production of cytokines by regional LN cells was affected in a similar manner. Mediastinal LN cells from control OVA/alum-primed mice produced significant levels of IL-4 (Figure 2F) or IL-5 (Figure 2G) in response to in vitro antigen stimulation. The challenge with OVA concurrently with CpG inhibited IL-4 (Figure 2F) and IL-5 (Figure 2G) production from antigen-stimulated LN CD4⁺ T cells. No IFN- production from the LN cells was detected (data not shown). Thus, intratracheal administration of CpG together with antigen challenge downregulated Th2 cells in parallel with the inhibition of airway eosinophilia.

Inhibition of Airway Eosinophilia by Pretreatment with CpG plus Antigen in Sensitized Animals

We next examined whether pretreatment with OVA plus CpG in advance of antigen challenge could protect the OVA-sensitized mice from the elicitation of airway eosinophilia. As in the experiments in Figure 2, the BALB/c mice were sensitized with OVA/alum and instilled intratracheally with OVA alone or together with CpG. They were then challenged with OVA (Figure 1, protocol 2). In comparison with the corresponding control mice that were pretreated intratracheally with OVA, those that were pretreated with OVA plus CpG showed reduced levels of airway eosinophilia (Figure 4A). Mice pretreated with OVA plus control ODN showed levels of eosinophilia comparable to those of the control group.

View larger version (78K): [in this window] [in a new window]   Figure 4.   Inhibition of airway eosinophilia by pretreatment with CpG plus antigen. BALB/c mice were primed with OVA/alum three times at 1-wk intervals. At 7 d after the last immunization, these mice were pretreated intratracheally with PBS, OVA (10 µg), CpG (5 µg), or control ODN (5 µg) as indicated for 2 consecutive d ( protocol 2 in Figure 1). After 6 d, mice were challenged intratracheally with OVA. After 2 d, the numbers of eosinophils in BALF and anti-OVA IgE levels in serum were determined. Intratracheal administration of OVA plus CpG in advance of antigen challenge inhibited airway eosinophilia, whereas systemic anti-OVA IgE responses in serum remained unaffected. (*P < 0.05; **P < 0.01.)

In a separate experiment, we compared the preinoculation of CpG plus antigen with that of PBS (Figure 4B). In contrast to the PBS-pretreated control group, OVA treatment deteriorated the airway eosinophilia. This could reflect the booster effects by 10 µg of instilled OVA, the dose that was optimal for inducing airway eosinophilia (14). In spite of this upregulation, the same dose of antigen inhibited eosinophilia when instilled together with CpG. Thus, instillation of CpG alone had no inhibitory effects and instillation of antigen alone had deteriorating effects, whereas the combination of the two inhibited eosinophilia when given in advance of antigen challenge.

CpG Improved Airway Eosinophilia and Hyperresponsiveness upon Pretreatment with a Minimal Amount of OVA That Failed to Boost Eosinophils

From a therapeutic standpoint, it would be better to minimize the antigen dose introduced into the airway in order to avoid a possible deleterious boost. We examined the effects of CpG with lower doses of OVA (Figure 5A). Preinstillation of as little as 0.1 µg of OVA had no ability to boost eosinophilia. Pretreatment with CpG alone failed to inhibit eosinophilia. In contrast, simultaneous instillation of OVA and CpG reduced eosinophilia in response to subsequent OVA challenge. Control ODN plus OVA had no effect on eosinophilia. Thus, pretreatment of antigen-sensitized animals with minimal amounts of antigen and CpG had prophylactic effects against subsequent antigen-induced eosinophilia. The inhibition of eosinophilia was in parallel with the inhibition of Th2 cells of mediastinal LN (Figures 5B and 5C). Airway hyperresponsiveness was also improved by CpG plus OVA pretreatment; the concurrent pretreatment with OVA plus CpG reduced the methacholine reactivity of the airway in comparison with the control PBS treatment (Figure 5D).

View larger version (86K): [in this window] [in a new window]   Figure 5.   CpG improved airway eosinophilia and hyperresponsiveness upon pretreatment with a minimal amount of OVA that failed to boost eosinophils. BALB/c mice were primed and challenged as in Figure 2B, except for the antigen dose for pretreatment. As little as 0.1 µg of OVA was administered either alone or with 5 µg of CpG or control ODN in advance of OVA challenge. After 2 d of OVA challenge, airway eosinophilia (A), hyperresponsiveness (D), or cytokine production from the mediastinal LN cells (B and C ) was examined. The LN cells were cultured in the presence of OVA (100 µg/ml) for 2 d, and the concentrations of IL-4 (B) and IL-5 (C ) in culture supernatants were determined by ELISA. Note that 0.1 µg of OVA failed to boost eosinophils, whereas concurrent instillation with CpG reduced eosinophilia in response to the subsequent antigen challenge. Airway hyperresponsiveness was in parallel with eosinophilia. Th2 inhibition was also induced by this minute amount of OVA plus CpG. (*P < 0.05; **P < 0.01.)

Lack of Effects on Systemic Anti-OVA IgE Responses

We examined the effect of pretreatment with CpG on anti-OVA IgE responses. In spite of the reduced levels of tracheal eosinophilia in mice pretreated with CpG plus OVA (Figure 4B), anti-OVA IgE responses in sera were not inhibited (Figure 4C). Thus, CpG plus antigen inhibited eosinophilia in presensitized mice even under the condition in which systemic immune responses to the antigen were not yet affected.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

Downregulation of unfavorable immune responses against inhaled antigens is one of the possible approaches in immunotherapy for bronchial asthma. Two major components that play pivotal roles in the pathogenesis of bronchial asthma are elevated IgE and eosinophilic inflammation in the airways, both of which are orchestrated by cytokines elaborated by Th2 cells (3). Selective abrogation of Th2 cell-mediated responses specific for allergen with sparing of other immune functions is the most desirable immunotherapy for avoiding adverse effects such as a deterioration of the host defense ability, which is inevitable as long as antigen-nonspecific immunosuppressants are used. In an attempt to meet these requirements, in this study we adopted a novel approach to control Th2 cells and airway eosinophilia. We found that concomitant administration of CpG and antigen (as opposed to either one alone) through airway mucosa efficiently inhibited airway eosinophilia in parallel with the downregulation of Th2 cells in the regional lymph nodes.

CpG activates macrophages and dendritic cells to produce IL-12, which in turn induces Th1 cell development (15). Because IL-12 secretion is induced independent of the attendant antigen, immune responses against diversified antigens are likely to be biased toward a Th1-dominant phenotype. In fact, injection of CpG alone without the accompanying antigen inhibited antigen-induced airway eosinophilia (25). Activation of Th1 cells in an antigen-nonspecific manner may lead to the potential danger of envisaging autoreactive Th1 cells, which would otherwise remain unactivated. Autoreactive Th1 cells are considered to be prime pathogenic cells that deteriorate experimental autoimmune diseases (27). The toxicity of CpG is another problem. Bacterial DNA was reported to increase the toxicity of LPS (31). Systemic administration of CpG is also reported to increase mortality, owing in part to an increase in the serum TNF- level (32). Transmucosal administration of CpG solved these problems. Inhibition of airway eosinophilia was achieved with smaller doses (5 µg) of CpG than those in other reports, where 30 to 100 µg of CpG were injected (25, 26). CpG alone had no effects on eosinophilia (Figures 4 and 5), mirroring the absence of nonspecific Th1 skewing, which was in sharp contrast to the use of higher doses of CpG (25). Surprisingly, such a minute amount of CpG exerted efficient antiallergic effects upon transmucosal coinoculation with antigen (Figures 4 and 5).

Then, were there any harmful effects of antigen inoculation into the trachea? Intratracheal instillation of OVA actually led to an enhancement of eosinophilic inflammation (Figure 4). This hazardous effect might reflect a boost of Th2 cells by 10 µg of OVA, which was an optimal dose for eosinophilic inflammation (14). The boost effect was circumvented by coadministration with CpG (Figure 4). More ideally, even a lower dose of antigen unveiled the anti-Th2 activity of CpG; as little as 0.1 µg of OVA inhibited eosinophilia upon combined intratracheal instillation with CpG, whereas the inoculation of this dose of antigen failed to elicit the boost (Figure 5). Thus, minute amounts of CpG and antigen were safe and efficient regulators of airway eosinophilia when used in combination but not separately. The synergism of CpG and antigen was critical for the transmucosal treatment.

We previously observed a similar synergism in the regulation of airway eosinophilia (12). We reported that both the antigen and M. tuberculosis bacilli, but not either one alone, were essential for the efficient induction of antigen-specific Th1 cells, and suggested that the copresentation of the antigen and M. tuberculosis bacilli would occur at the level of APC secretion of IL-12 (12). In the present report we could not clarify the precise mechanisms for the synergism. Inasmuch as CpG induced IL-12 secretion from dendritic cells (20, 21), targeting of Th2 cells to APCs that phagocytose both CpG and antigen and secrete IL-12 might be a critical step for the synergism.

The effects of transmucosally administered CpG were reported previously. Freimark and coworkers reported that CpG sequences present within the plasmid induced the Th1 cell-promoting cytokines after intratracheal administration of the plasmid/lipid complexes (38). CpG was found to be an enhancer of antibody responses against influenza virus or hepatitis B surface antigen that were administered intranasally with CpG (39, 40). In these two reports, the effects of CpG on T-cell functions were not examined. More recently, Broide and collaborators reported that intratracheal instillation of CpG (50 to 100 µg) without the accompanying antigen inhibited airway eosinophilia and hyperresponsiveness (25). Sur and coworkers reported that intratracheal administration of CpG induced long-term prevention of allergic lung inflammation (41). They also found that no suppression of lung inflammation was observed when CpG were coadministered with the allergen (41). These reports are in sharp contrast to our observations. We found that pretreatment with 5 µg (Figure 2) or even 50 µg (data not shown) of CpG alone did not inhibit eosinophilia. We showed that the administration of a smaller dose (5 µg) of CpG with antigen induces the antigen-specific Th2 inhibition in the mediastinal LN, which improves airway eosinophilia and hyperresponsiveness. We could not reconcile these discrepancies. However, additional evidence that supports our notion comes from our recent experiment using a direct conjugate of CpG to OVA. When the covalently linked CpG-OVA conjugate was instilled into the trachea, the conjugate inhibited airway eosinophilia and cytokine production at smaller doses than did a simple mixture of OVA plus CpG (submitted manuscript). These observations reinforce our view that efficient targeting of CpG effects to antigen-specific T cells would be facilitated when the CpG-phagocytosed APC presented the antigen after coadministration of CpG and the antigen.

In summary, we have described a novel approach to inoculate CpG to inhibit Th2 cells and airway eosinophilia. The important findings are as follows. Administration of CpG through the airway mucosa inhibited local Th2 responses, and the dose of CpG and adverse effects could be minimized. Concomitant transmucosal administration of CpG and antigen, but not either alone, was essential for the inhibition of airway eosinophilia. The amount of antigen could be reduced to avoid the induction of eosinophilia. These findings provide the experimental basis for a possible inhalation therapy of allergen plus CpG to patients with bronchial asthma.