Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Clinical impression is that bacterial respiratory infections, unlike viral infections, do not trigger exacerbations in asthmatics. Living in a "dirty" environment and development of nonwheezing lower respiratory tract infection in early life are associated with lower prevalence of allergic sensitization and asthma later in life. To examine the relationship between these two inflammatory processes, we have developed an in vivo rat model of combined bacterial and allergic inflammation.

Exposure to lipopolysaccharide (LPS) to mimic bacterial inflammation resulted in an inflammatory response maximal after 6 h. This response has been characterized by neutrophil influx into bronchoalveolar lavage fluid (BALF) and lung parenchyma, increased microvascular leakage (MVL) (assessed by Evans blue), and airway and parenchymal hyperresponsiveness (HR) to methacholine (MCh) (1). In this model, the cellular influx is still significantly elevated above control at 24 h; however, the HR and MVL are no longer apparent. In addition, we have characterized allergic inflammation with elevated ovalbumin (OVA)-specific serum immunoglobulin (Ig) E and infiltration of eosinophils, lymphocytes, and macrophages into the lung parenchyma and BALF, as well as MVL and airway/parenchymal HR to MCh 24 h after allergen challenge. By combining these two in vivo models we have shown that exposure of sensitized animals to LPS 18 h after OVA challenge completely inhibits late-phase allergen-induced HR and further potentiates neutrophil influx while reducing the numbers of eosinophils and lymphocytes present in the lavage (1). These results are illustrated schematically in Figure 1. We then studied the mechanisms involved in these phenomena by using nitric oxide (NO) synthase (NOS) isoenzyme inhibitors in vivo and have demonstrated that the NO produced by the inducible NOS (iNOS), localized predominantly in inflammatory cells and epithelial cells, plays an important role in both migration of inflammatory cells and increase in microvascular permeability after allergen challenge. In contrast, we reported that the NO produced by the constitutively expressed neuronal NOS (nNOS) limits bronchial HR to MCh (2).

View larger version (14K): [in this window] [in a new window]   Figure 1.   Role of LPS on allergen-induced inflammatory response. Schematic representation of the effect of LPS (dashed line, n = 6) or saline (solid line, n = 6) given 18 h after OVA challenge in sensitized animals on (A) maximal response to inhaled MCh and (B) number of neutrophils (open circles), lymphocytes ( filled circles), and eosinophils ( filled triangles) recovered from BALF measured 24 h after OVA challenge.

The results of our earlier studies suggest that in both naive animals exposed to LPS and in sensitized animals challenged with OVA there is an upregulation of iNOS and downregulation of constitutive NOS (cNOS) activity. However, these conclusions are based on relatively selective isoenzyme inhibitors. We felt that direct proof was required to understand the potential effects of LPS exposure during the allergen-induced late-phase response (LPR) on NOS isoenzymes. In addition, we have investigated the potential role of interleukin (IL)-10 in LPS-induced inhibition of LPR to OVA.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Animals

Male Piebald-Virol-Glaxo (PVG) rats weighing 200 to 280 g (Research Center, Institute for Child Health Research, Perth, Australia), barrier-housed in a clean animal-house environment, kept on an OVA-free diet with access to water and food ad libitum, were studied at 10 wk of age. The colony was free of known pathogens. The Institutional Animal Ethics Committee approved the study protocol.

Sensitization Procedure and Allergen Challenge

Animals were sensitized on Day 0 (100 µg OVA plus 50 ng ricin, intraperitoneally) and challenged with allergen (1% OVA for 30 min) on Day 12, at the peak of IgE response, using an ultrasonic nebulizer (De Vilbiss Ultra-Neb 2000; Sunrise Medical, Somerset, PA). NOS activity and IL-10 were measured in two separate groups of animals.

LPS Exposure

Naive animals were exposed to nebulized LPS from Salmonella typhimurium (50 µg/ml for 30 min) using an ultrasonic nebulizer (De Vilbiss Ultra-Neb 2000; Sunrise Medical). In naive animals, NOS activity was measured 6 h after LPS exposure and IL-10 levels were measured at 1, 3, 6, or 18 h after LPS exposure. In separate groups of sensitized rats, NOS activity was measured 18 or 24 h after OVA challenge or 24 h after OVA challenge with LPS exposure at 18 h. IL-10 activity was measured in sensitized animals at 1, 3, 6, 18, or 24 h after OVA challenge or in animals exposed to LPS at 19, 21, or 24 h after OVA challenge.

Lung Perfusion

After administration of a lethal dose of anesthesia (pentobarbitone at 100 mg/kg), the lungs were immediately perfused with 25 ml of perfusion media (phosphate-buffered saline [PBS], pH 7.2; 1,000 U/ ml heparin; and 0.2% bovine serum albumin [BSA] at 37°C), removed from the thoracic cavity, separated from the trachea, snap-frozen in liquid N₂, and transported on dry ice for measurement of NOS activity. Blood was collected from the descending aorta for measurements of OVA-specific IgE and IgG titers.

Lung Homogenization

Frozen lung samples were homogenized in 5 vol (wt/vol) of ice-cold homogenization buffer (250 mM tris[hydroxymethyl]-aminomethane [Tris]/HCl, pH 7.4) containing 10 mM ethylenediaminetetraacetic acid (EDTA), 10 mM ethyleneglycol-bis-(-aminoethyl ether)-N,N'-tetraacetic acid (EGTA), and Complete Mini protease inhibitor cocktail tablets (1 tablet/10 ml). The tissue was homogenized (20,000 rpm twice for 30 s each time) using a Polytron tissue grinder (model PT 2000 Ultra-Turrax T25; Janke & Kunkel IKA-Labortechnik, Staufeni, Brussels) with brief shaking in between. The crude homogenates were centrifuged at 9,000 × g for 5 min at 4°C. The pellet was washed, resuspended in 2 vol (wt/vol) of homogenization buffer, and centrifuged once again for 5 min. Supernatants were pooled and stored in 200-µl aliquots at 70°C. The final 7 ml of supernatant contained extract from 1 g of tissue.

Measurement of NOS Activity in Lung Homogenates

Maximal NOS activity was measured by conversion of L-[³H]arginine to L-[³H]citrulline using NOS Assay Kit (Calbiochem-Novabiochem, San Diego, CA). The supernatant (10 µl) was added to prewarmed (37°C) Epindorf tubes containing 40 µl of reaction stock mix (50 mM Tris/HCl [pH 7.4], 6 µM tetrahydrobiopterin, 2 µM flavin adenine dinucleotide (FAD), 2 µM flavin mononucleotide (FMN), 10 mM nicotinamide adenine dinucleotide phosphate [NADPH], and 30 µl 1 µCi/µl stock L-[2,3,4,5-³H]arginine monohydrochloride) in the presence or absence of 6 mM CaCl₂. Samples were incubated for 1 h at 37°C and reactions terminated by embedding the samples in ice and adding 400 µl stop buffer (50 mM N-2-hydroxyethylpiperazine-N'-ethane sulfonic acid (HEPES) and 5 mM EDTA, pH 5.5) to the reaction mixture. To obtain free L-[³H]citrulline, 100 µl of equilibrated cation exchange resin (8% crosslinked Na⁺ form, pH 5.5) was added to eliminate excess L-[³H]arginine from solution. The sample was then transferred to separation columns and centrifuged (13,000 rpm for 30 s) to separate the resin from the fluid phase. The amount of citrulline produced was measured in a follow-through using a liquid scintillation analyzer (Packard Tri-Carb 1500; Packard Instruments, Canberra, Australia). NOS activity was expressed as picomoles of citrulline produced by 1 g of tissue. NOS activity in the presence of Ca²⁺ is referred to as total NOS and in the absence of Ca²⁺ as iNOS, and cNOS is the difference between total NOS and iNOS activity. Homogenate of rat cerebellum extract provided was used as positive control. L-NMMA, a nonselective NOS inhibitor, served as negative control to assess background activity and non-NOS-dependent conversion of L-[³H]arginine to L-[³H]citrulline.

Rat Tracheal Explants

Tracheae were removed from rats and placed in CRML-1066 medium supplemented with 2 µg insulin, 100 U/ml penicillin, 100 µg/ ml streptomycin, 250 ng/ml amphotericin-B, 2 mM L-glutamine, and 2% fetal bovine serum (FBS). All connective tissue and visible blood vessels were removed from the trachea. Tracheal rings (1.5 mm in width) were cut and rings immediately adjacent to the larynx were discarded. Four randomly chosen tracheal rings were placed in 35-mm plastic culture dishes containing 1.5 ml of CRML-1066 medium. Tissue-culture dishes were placed on a tray in a controlled atmosphere chamber which was flushed with a mixture of 45% O₂, 50% N₂, and 5% CO₂ at a flow rate of 4 liters/min for 15 min. The chamber was then placed in a 37°C incubator on a rocking platform set at 10 cycles/min so that the tracheal lumen was intermittently exposed to media and the gas mixture. The tracheal explants were cultured for 24 h.

Measurement of IL-10

IL-10 was measured in the supernatant of tracheal explants using a commercially available rat IL-10 OptEIA Set (PharMingen, San Diego, CA) with a few modifications as outlined later. Microtiter 96-well plates (Maxisorp; NUNC, Rochester, NY) were coated with the capture antibody (antirat IL-10) diluted in coating buffer (0.2 M sodium phosphate, pH 6.5) and incubated overnight at 4°C. On the second day, the wells were aspirated and washed three times for 3 min each time with wash buffer (PBS with 0.05% Tween-20, 200 µL/well). Plates were blocked with assay diluent (PBS with 10% FBS, pH 7, 200 µL/well) for 2 h at room temperature, then washed as described earlier. A standard concentration curve for recombinant IL-10 was made up in assay diluent (PBS with 10% FBS, pH 7) ranging from 15.6 to 4,000 pg/ml. Test samples and standards were added (100 µL/well) and plates were incubated at 37°C for 2 h, then washed (five times for 3 min each time) as described earlier. The working detector (biotinylated antirat IL-10 monoclonal antibody [mAb] and avidin-horseradish peroxidase; 100 µL/well) diluted in assay diluent was added and incubated for 1 h at room temperature. Plates were washed seven times as described earlier, with 3-min soaks between washes. The peroxidase substrate solution (TMB and hydrogen peroxide, 100 µl/well) was added according to manufacturer's instructions and plates were incubated for up to 30 min at room temperature. Reaction was stopped by the addition of 1 M H₂SO₄ (50 µl/well) and plates read spectrophotometrically at 450 nm with an enzyme-linked immunosorbent assay plate reader (Microplate AutoReader EL311; Bio-Tek Instruments, Inc., Winooski, VT). Data were analyzed using linear regression analysis in AssayZap (Biosoft, Cambridge, UK). Assay detection limit was 30 pg/ml.

Drugs and Materials

OVA (Grade V), ricin, LPS (S. typhimurium), penicillin, streptomycin, and amphotericin-B were purchased from Sigma Chemical Co., St. Louis, MO. BSA was purchased from CSL (Parkville, VIC, Australia), Ketamine (Ketamil) from Troy Laboratories (Smithfield, NSW, Australia), xylazine (Rompun) from Bayder (Pymble, NSW, Australia), pentobarbitone sodium (Lethabarb) from Virbac (Peakhurst, NSW, Australia), and heparin from David Bull Laboratories (Mulgrave, VIC, Australia). CRML-1066 media, L-glutamine, and insulin were from Life Technologies (Melbourne, VIC, Australia). Tris was from Astral Scientific (Gymea, NSW), L-[2,3,4,5-³H]-arginine monohydrochloride was from Amersham Pharmacia Biotech (Castle Hill, NSW, Australia), and Complete Mini protease inhibitor cocktail tablets were from Roche Diagnostics (Basel, Switzerland). The NOS assay kitwhich contained the homogenization, stop, and reaction buffers; equilibrated resin; CaCl₂; rat cerebellum extract; and NADPHwas purchased from Calbiochem-Novabiochem. The rat IL-10 OptEIA Set was kindly donated by PharMingen.

Statistical Analysis

The differences in the enzyme activity and IL-10 production between groups were identified by analysis of variance with Student-Newman-Keuls correction for multiple comparisons. All results are expressed as means ± standard error of the mean (SEM). P < 0.05 was regarded as statistically significant.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

NOS Activity in Naive Animals Exposed to LPS

Endogenous iNOS and cNOS activity was detected in lung homogenates of naive, nonsensitized, and saline-challenged animals. In these animals, more than 99% of total NOS enzyme activity was due to cNOS (n = 6, P < 0.001) (Figure 2, upper panel). Exposure to LPS induced a greater than 100-fold increase in iNOS activity from 2.30 × 10⁵ to 2.27 ± 0.08 pmol citrulline/g lung tissue (n = 6, P < 0.001) (Figure 2, lower panel). LPS also caused a 6-fold reduction in cNOS activity from 1.47 ± 0.12 to 0.25 ± 0.05 pmol citrulline/g tissue (n = 6, P < 0.001) (Figure 2, upper panel).

View larger version (19K): [in this window] [in a new window]   Figure 2.   cNOS and iNOS isoenzyme activity in lung homogenates. Measurement of cNOS (top panel) and iNOS (lower panel) enzyme activity in lung homogenates 6 h after saline (n = 6) or LPS exposure (n = 6) in naive animals, 18 (n = 5) and 24 (n = 6) h after OVA challenge in sensitized animals, or 24 h after OVA challenge with prior exposure to LPS (n = 5). Results are expressed as means ± SEM. P < 0.001 versus nonsensitized and saline-challenged animals; *P < 0.001 versus sensitized and saline-challenged animals (n = 7); and §P < 0.001 versus sensitized animals 18 or 24 h after OVA challenge. First column represents the positive control (rat cerebellum extract) supplied with the NOS enzyme kit.

NOS Activity in Sensitized Animals Exposed to OVA

Similar changes in enzyme profiles were observed after OVA exposure in sensitized animals to the changes seen in naive animals exposed to LPS. At 18 h after OVA challenge (OVA₁₈), iNOS activity was upregulated to 2.24 ± 0.09 pmol citrulline/g tissue (n = 5, P < 0.001) (Figure 2, lower panel) and cNOS downregulated 6-fold from 1.68 ± 0.11 to 0.30 ± 0.08 pmol citrulline/g tissue (n = 5, P < 0.001) (Figure 2, upper panel). At 24 h after OVA challenge (OVA₂₄), iNOS activity remained elevated above baseline (P < 0.001); however, it was significantly reduced when compared with iNOS activity at 18 h (P < 0.05). In contrast, cNOS activity was decreased to a similar degree at 18 and 24 h after OVA challenge (Figure 2, upper panel).

NOS Activity in Sensitized and OVA-Challenged Animals Exposed to LPS

Exposure of sensitized animals to LPS 18 h after OVA challenge (OVA₂₄+LPS₆) resulted in greater than 3-fold increase in cNOS activity to 1.00 ± 0.19 pmol citrulline/g tissue (n = 5, P < 0.001) (Figure 2, upper panel) and 5-fold reduction in iNOS activity to 0.48 ± 0.13 pmol citrulline/g tissue (n = 5, P < 0.001) (Figure 2, lower panel) in sensitized animals when compared with a control group with no LPS exposure (OVA₁₈). Constitutive NOS activity in the lungs of OVA₂₄+LPS₆ animals was significantly elevated when compared with cNOS activity in naive animals 6 h after LPS exposure (n = 6, P < 0.01) (Figure 2, upper panel).

IL-10 Measurements in Tracheal Explants

In initial experiments we failed to detect IL-10 in BALF after LPS exposure in naive or in sensitized animals with or without OVA challenge. Using the supernatant of tracheal explants, endogenous IL-10 was detected in naive animals (312 ± 24 pg/ml, n = 9) and in sensitized animals 24 h after saline challenge (369 ± 31 pg/ml, n = 6) (Figure 3). In naive animals, IL-10 was increased 1 h after LPS exposure to 883 ± 71 pg/ml (n = 5, P < 0.001), peaked at 3 h to 1,652 ± 168 pg/ml (n = 9, P < 0.001), and remained significantly elevated above baseline by 18 h at 457 ± 23 pg/ml (n = 11, P < 0.05) (Figure 3).

View larger version (21K): [in this window] [in a new window]   Figure 3.   IL-10 levels in naive and in sensitized animals exposed to LPS. Measurements of IL-10 from supernatant of tracheal explants of naive animals 1 (n = 5), 3 (n = 9), 6 (n = 10), or 18 (n = 11) h after LPS exposure; of sensitized animals 1 (n = 5), 3 (n = 5), 6 (n = 6), 18 (n = 11), or 24 (n = 5) h after OVA challenge; or of sensitized and OVA-challenged animals exposed to LPS 18 h after OVA and measured 1 (n = 10), 3 (n = 5), or 6 (n = 5) h later, i.e., 19, 21, or 24 h after OVA challenge. Results are expressed as means ± SEM. *P < 0.05, **P < 0.01, and ***P < 0.001 versus naive (n = 9) or sensitized and saline-challenged animals (n = 6); P < 0.001 versus sensitized and OVA-challenged animals at 18 h.

In sensitized animals, IL-10 was not significantly increased until 18 h after OVA challenge (706 ± 32 pg/ml) (n = 11, P < 0.001); however, this increase was 2-fold lower than the increase seen with LPS exposure (Figure 3). In sensitized and OVA-challenged animals, exposure to LPS 18 h after OVA challenge produced a 3-fold increase in IL-10 to 2,163 ± 30 pg/ml (n = 5, P < 0.001), 3 h after exposure to LPS (that is, 21 h after OVA challenge). The time course and kinetics of IL-10 release in those animals were similar to those in naive animals (Figure 3).

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

The results of the present study have shown that bacterial inflammation, mimicked by exposure of PVG rats to inhaled LPS, upregulates the activity of iNOS and downregulates the activity of cNOS 6 h after exposure. We have also shown that exposure of sensitized animals to OVA results in similar upregulation of iNOS activity and downregulation of cNOS at both 18 and 24 h after OVA challenge. However, the most striking finding is that exposure to LPS, during an established allergen-induced LPR, is able to reverse the allergen-induced changes in iNOS and cNOS isoenzyme activities.

The data from the present study support our previous conclusions on the basis of isoenzyme inhibitors, that NO produced by NOS isoenzymes have differential effects in regulating OVA-induced allergic responses in rats (2). In vivo inhibition of iNOS with aminoguanidine was effective in abolishing the OVA-induced cellular inflammation and MVL in sensitized rats, producing no effect on their airway and parenchymal HR to MCh, suggesting this isoenzyme to be predominantly involved in cellular migration and/or recruitment after OVA challenge. However, inhibition of nNOS with S-methyl-L-thiocitrulline produced no effect on OVA-induced increase in cellular influx or MVL but further exaggerated airway and parenchymal HR (2). Further, we have shown LPS to reverse these inflammatory changes and, in this study, these changes are reported to be associated with reversal in allergen-induced changes in NOS isoenzyme activity; that is, reduction in iNOS and upregulation in cNOS activity.

The mechanism by which LPS modifies NOS activity during an established allergic response is currently unclear. LPS is known to stimulate production of the anti-inflammatory IL-10. IL-10 is thought to be involved in protection against lung injury after LPS insult in mice (3) and inhibition of OVA-induced cellular recruitment into the airways of sensitized mice (7). IL-10 can inhibit airway inflammation by a variety of mechanisms, including reducing the production of proinflammatory cytokines, chemokines, and transcription factors such as nuclear factor-B in human monocytes (10, 11).

In our in vivo model of bacterial inflammation, IL-10 was rapidly increased in tracheal explants of naive animals as early as 1 h after LPS exposure, peaked at 3 h, and remained elevated above baseline at 18 h. Our results are in agreement with earlier reports showing IL-10 to peak as early as 1.5 h after LPS challenge in aged mice (12), 2 h after systemic exposure of mice to LPS (13), or 3 h after IL-10 gene transfer in LPS-induced mice (14). Such transient and rapid increase in IL-10 implicates preformed release of IL-10. Alternatively, de novo production of IL-10 may occur later than 24 h after bacterial exposure as suggested by Van der Pouw Kraan and colleagues in their murine model of pneumococcal pneumonia, showing increased IL-10 production in the lungs over a period of three days after infection (15).

Inflammatory responses to inhalation of LPS in naive animals are thought to be initiated by activation of CD14/ Toll-like receptor (TLR) 4 complex on alveolar and tissue macrophages (16). Blood monocytes and tissue macrophages may be the most important source of anti-inflammatory IL-10 (17) and are likely to be the major source of IL-10 in vivo upon LPS stimulation in our naive animals (18). Rapid release of IL-10 in our model may be part of a negative feedback mechanism in response to proinflammatory cytokines after exposure to LPS and may represent an important homeostatic mechanism to control inflammation in the lungs.

In our and in similar animal models of LPS-induced inflammation, inflammation is thought to be neutrophil-mediated (1, 19, 20). We have previously shown that exposure to LPS results in large neutrophil influx into BALF and lung parenchyma 6 h after exposure, and results of our present study clearly suggest that LPS upregulates iNOS activity in lung homogenates. As neutrophils accounted for greater than 87% of the total cell count in the BALF of these animals (1), it is reasonable to hypothesize that increased iNOS activity may be a result of iNOS upregulation among neutrophils. Increased production of NO as a result of increased expression of iNOS in rat blood polymorphonuclear neutrophils has been demonstrated in vivo after intravenous administration of LPS (21) and after zymosan-induced inflammation (22).

In our study we have detected significant endogenous IL-10 activity in the airways of naive animals and sensitization per se had no effect on basal cytokine levels. In sensitized animals, IL-10 was not significantly increased until 18 h after OVA challenge and the magnitude of this increase was 2-fold lower than that seen with LPS exposure. These results are consistent with timing of de novo production of cytokines in the literature (23). In agreement with our study, increased IL-10 levels have been demonstrated in BALF after allergen challenge in sensitized mice (24) and in patients with asthma (17, 25). Yssel and colleagues reported that the message for IL-10 does not appear until 8 h after stimulation of human CD4⁺ T cells and is maximal at 12 to 24 h (26). Similar kinetic studies reported that low levels of IL-10 could be detected 7 h after activation of human monocytes and that maximal IL-10 production occurred 24 to 48 h after activation (27).

Exposure of sensitized and OVA-challenged animals to LPS 18 h after allergen challenge induced a large increase in IL-10 3 h after LPS; that is, 21 h after OVA challenge (Figure 3). At 6 h after LPS (or 24 h after OVA challenge), IL-10 levels returned to baseline and in these animals LPS exposure was associated with complete abolition of OVA-induced airway and parenchymal HR to MCh (1). These data indicate that the early release of preformed IL-10 may be responsible for abolition of late-phase airway and parenchymal HR and reduction in macrophage and lymphocyte numbers previously reported in our sensitized and OVA-challenged animals (1).

The mechanism(s) by which release of large quantities of IL-10 could reverse the allergen-induced late-phase increase in MCh HR and cellular influx are unknown. Possibilities include the ability of IL-10 to prevent allergen-specific proliferation and activation of CD4⁺ T lymphocytes in vitro. Both direct and indirect effects have been previously described. IL-10 exerts its effects on T cells indirectly by inhibiting the antigen-presenting capacity of monocytes, macrophages and dendritic cells (DCs). High concentrations of IL-10 strongly downregulate class II major histocompatibility complex (27, 28) and adhesion molecule expression including intercellular adhesion molecule (ICAM)-1, B7-CD28, CD80, or CD86 on the surface of antigen-presenting cells (29, 30), which are essential for T-cell activation. In addition to these indirect effects, IL-10 can also directly reduce antigen-driven accumulation of eosinophils and CD4⁺ T cells (9, 31) and inhibit release of a variety of early inflammatory cytokines through downregulation of IL-2 gene expression in the responding cells (32). In the later study, exogenous addition of IL-10 to the cultures strongly inhibited the cytokine production at the transcriptional level, and the presence of neutralizing anti-IL10 mAbs further increased the cytokine production relative to LPS treatment alone (27).

There is some debate in the literature as to the relative contributions of IgE-dependent versus T cell-dependent mechanisms in the pathogenesis of allergen-induced late-phase responses. The IgE-dependent allergic response is induced by IgE crosslinking resulting in mast cell (MC) degranulation and immediate release of preformed proinflammatory mediators. This response is driven by CD4⁺ T lymphocytes and their release of IL-4, IL-5, and IL-13 cytokines which are thought to be responsible for perpetuation of allergic inflammation. However, extensive evidence also exists to show that T cells themselves can directly cause late-phase phenomenon and chronic asthma (25, 33). Recently demonstrated by Haselden and colleagues, cat allergen (Fel d1) peptides can directly cause T cell-dependent LPR independent of IgE and MC activation in sensitized asthmatic subjects (34). These peptides are too short to crosslink with IgE, however still caused T-cell proliferation, IL-5 production, and LPR to Fel d1 cat allergen. Further, using animal models, eosinophil responses and bronchial HR to inhaled antigen challenge can be adoptively transferred by CD4⁺ T cells alone (35, 36). Given that IL-10 prevents T-cell activation, and if responses to OVA in sensitized PVG rats are T cell-dependent, this may represent a direct mechanism responsible for loss of OVA-induced HR associated with LPS exposure in our allergic animals.

Our results demonstrate that both allergen- and LPS-induced inflammation stimulate production of NO by increasing iNOS activity in the lungs. However, LPS exposure in already sensitized animals reduced iNOS activity. An alternative explanation for the loss of HR may be that iNOS is subject to negative feedback by NO itself; its own product. In support of our findings, it has been demonstrated in vitro that IL-10 can inhibit the induction of iNOS in macrophages and their production of NO in a dose-dependent manner when activated with a low dose of LPS (10 ng/ml) (37). Inhibition of iNOS, as shown by our findings, would result in selective inhibition of OVA-induced cellular inflammation in sensitized animals, and we propose this to be the mechanism involved in inhibition of macrophage and lymphocyte accumulation as well as reduced MVL in sensitized and OVA-challenged animals when LPS was given 18 h after OVA challenge.

Along with IL-10, IL-12 has also been shown to prevent antigen-induced airway HR, lung eosinophilia, and T helper (Th) 2 cytokine expression after OVA exposure in sensitized mice despite the presence of circulating IgE (38, 39). LPS is known to stimulate large production of IL-12 from macrophages and from human dendritic and mononuclear cells (40). This production peaks 12 to 24 h after LPS stimulation and remains elevated for up to 48 h (Upham, unpublished observation). Interestingly, in the presence of IL-4 (as would be found during an allergic response), IL-12 production is significantly upregulated in human DCs (40). Whereas IL-12 is generally thought to modify allergic responses by inhibiting IgE production and the development of Th2-type responses during primary allergic sensitization, these earlier reports and the timing of its production clearly indicate that IL-12 is capable of modifying existing allergic responses and support its role in inhibition of OVA-induced HR in our in vivo model after LPS stimulation.

A novel signaling pathway for IL-12 has recently been described by Collison and colleagues, which may explain how IL-12 is capable of inhibiting allergen-induced late-phase allergic response (41). These researchers have demonstrated that IL-12 is able to increase free Ca²⁺ inside neutrophils by its direct binding to the IL-12 receptor on human neutrophils (41). This evidence for IL-12-activated, Ca²⁺-dependent activation of human neutrophils allows the speculation that a similar mechanism operating in inhibitory nonadrenergic noncholinergic nerves could potentially explain the reversal of OVA-induced downregulation of cNOS by LPS. Because cNOS is a Ca²⁺-dependent enzymeas opposed to iNOS, which is Ca²⁺-independentan increase in intracellular Ca²⁺ would result in increased cNOS activity and relaxation of airway smooth muscle.

In summary, we have shown that LPS in naive animals and OVA challenge in sensitized animals both cause an upregulation in iNOS and a downregulation in cNOS activity in the lungs of these animals. Further, we have shown that LPS is capable of reversing such OVA-induced changes in NOS isoenzyme activity during an already established allergic response, and that LPS-induced release of IL-10 is thought to play an important immunomodulatory role in this model of allergic inflammation. Whether a single intermediatory product is responsible for both effects (unlikely) or different products are responsible for the differential effects on iNOS and cNOS activity (more likely) is unknown. In either case, this area of research clearly warrants further investigation as identification of such a mediator(s) may provide new therapeutic targets for the treatment of asthma. Drugs that are capable of reversing established late-phase allergen reactions would clearly be of major benefit in treating acute exacerbations of asthma.