Abstract
In most cells trans-activating NF-κB induces many inflammatory proteins as well as its own inhibitor, IκB-α, thus assuring a transient response upon stimulation. However, NF-κB-dependent inflammatory gene expression is persistent in asthmatic bronchi, even after allergen eviction. In the present report we used bronchial brushing samples (BBSs) from heaves-affected horses (a spontaneous model of asthma) to elucidate the mechanisms by which NF-κB activity is maintained in asthmatic airways. NF-κB activity was high in granulocytic and nongranulocytic BBS cells. However, NF-κB activity highly correlated to granulocyte percentage and was only abrogated after granulocytic death in cultured BBSs. Before granulocytic death, NF-κB activity was suppressed by simultaneous addition of neutralizing anti-IL-1β and anti-TNF-α Abs to the medium of cultured BBSs. Surprisingly, IκB-β, whose expression is not regulated by NF-κB, unlike IκB-α, was the most prominent NF-κB inhibitor found in BBSs. The amounts of IκB-β were low in BBSs obtained from diseased horses, but drastically increased after addition of the neutralizing anti-IL-1β and anti-TNF-α Abs. These results indicate that sustained NF-κB activation in asthmatic bronchi is driven by granulocytes and is mediated by IL-1β and TNF-α. Moreover, an imbalance between high levels of IL-1β- and TNF-α-mediated IκB-β degradation and low levels of IκB-β synthesis is likely to be the mechanism preventing NF-κB deactivation in asthmatic airways before granulocytic death.
Chronic airway inflammation, associated with persistent overexpression of many proteins involved in immune and inflammatory responses, is a characteristic feature of asthma (for review, see Ref. 1). Protein overexpression depends on increased gene transcription, suggesting that activation of some transcription factors underlies asthma pathogenesis. Transcription factors that are thought to be involved in asthma are NF-κB, AP-1, NF-AT, cAMP response element binding protein, STATs, and GATA-3 (Refs 2 and 3 ; for review, see Ref. 4). All of the inflammatory genes overexpressed in asthma, such as those encoding proinflammatory cytokines, chemokines, adhesion molecules, and inflammatory enzymes, contain κB sites for NF-κB within their promoter (for review, see Ref. 5), suggesting that these genes are controlled predominantly by NF-κB and that NF-κB could be of particular importance in the initiation and the perpetuation of allergic inflammation. This assumption is reinforced by the fact that glucocorticosteroids, the most potent treatment for asthma, strongly inhibit NF-κB in vitro (6, 7).
The NF-κB family is composed of five structurally related DNA-binding proteins, called p50, p52, p65/RelA, c-Rel/Rel, and RelB (for review, see Ref. 8). The most common form of NF-κB is a heterodimer composed of p50 and p65 subunits, although the different family members can associate in various homo- or heterodimers through a highly conserved N-terminal sequence, called the Rel homology domain. Dimerization of various NF-κB subunits produces complexes with different DNA-binding specificities and trans activation potentials. In most cell types, inactive NF-κB complexes are associated with inhibitory proteins of the IκB family, which sequester NF-κB in the cytoplasm. The members of the IκB family are IκB-α, IκB-β, IκB-ε, p100, p105, and Bcl-3, where the most common IκB protein is IκB-α (8, 9). p105 and p100 are the precursors of p50 and p52, respectively. Following various stimuli, such as viruses, bacteria, pro-oxidants, and proinflammatory cytokines, IκB proteins are first phosphorylated, ubiquitinated, and then rapidly degraded by the proteasome, allowing NF-κB nuclear translocation and transcriptional initiation of NF-κB-dependent genes (10).
Macrophages of induced-sputum and bronchial epithelial cells from stable asthmatic patients exhibit increased NF-κB activity compared with cells from healthy patients (11). Mice deficient in p50 or c-Rel are unable to develop eosinophilic airway inflammation when sensitized and challenged with OVA (12, 13). In bronchial brushing samples (BBSs)3 recovered in heaves-affected horses, a spontaneous animal model of asthma (for review, see Ref. 14), NF-κB complexes are mainly atypical p65 homodimers (15). p65 homodimer activity drastically increases in BBSs from heaves-affected horses challenged with moldy hay, which contains the allergens responsible for the disease (i.e., proteins borne by spores of Asperigillus fumigatus, Faenia rectivirgula, and Thermoactinomyces vulgaris). Interestingly, this increased activity is maintained at high or moderate levels for at least 21 days after allergen eviction from the horses’ environment. In this model it has also been demonstrated that p65 homodimer activity found in BBSs is highly correlated to the degree of lung dysfunction and to the level of ICAM-1 expression (15). Although these in vivo observations confirmed that NF-κB is likely to play a crucial role in allergic inflammation and in subsequent airway obstruction, no study was devoted to the mechanisms by which NF-κB activity is maintained in lung cells even when the etiologic agent is absent.
In the present report, we describe studies aimed at identifying the mechanisms by which NF-κB activity is regulated in cells obtained by bronchial brushing in heaves-affected horses after allergen eviction, and we propose a cellular and molecular model that accounts for the persistent NF-κB activity observed in the bronchi of this animal model of asthma.
Materials and Methods
Experimental animals
Nine horses (564 ± 83 kg; 18.6 ± 1.8 years; mean ± SD) with a history and clinical signs of heaves were used. These horses typically developed acute airway obstruction (crisis) when housed in a barn and fed moldy hay, and they entered clinical remission once pastured or stabled in a controlled environment. One month before the experiment they underwent a thorough clinical examination, including an electrocardiogram, arterial blood gas analysis, hematology, endoscopy of the airways, tracheo-bronchial lavage, and pulmonary scintigraphy. This confirmed that they suffered from heaves and were free from any other health problems. Six healthy horses (605 ± 99 kg; 7.6 ± 2.6 years) were used as controls. Experimental horses did not receive any treatment during the month preceding the experiments.
Bronchial cells of heaves-affected horses were obtained by bronchial brushing on two separate occasions: 24 h after the onset of a crisis and then 21 days after removal from the causative environment. To obtain crisis, the horses were stabled and subjected to a natural challenge with moldy hay. The horses were considered to be in crisis when their breathing mechanic variables were within the following limits: maximal difference in pleural pressure ≥2.00 kPa, total pulmonary resistance ≥0.2 kPa/l · s−1, and dynamic compliance ≤8 l/kPa−1. These respiratory mechanic variables were calculated from simultaneous measurements of esophageal pressure, air flow, and tidal volume (for more technical details, see Ref. 16). Eviction of the antigenic agents was obtained by pasturing the horses or stabling them with dust-free bedding and feed. Healthy horses were investigated twice at a 21-day interval. The protocol was approved by the ethics committee of the University of Liege.
Bronchial brushings
Horses were premedicated i.v. with 0.01 mg/kg romifidine (Sedivet; Boehringer Ingelheim, Ingelheim, Germany). Bronchoscopy was performed with a 9-mm diameter bronchoscope (Pentax, Breda, The Netherlands) using a transnasal approach. The brushing was performed in 10 different places, from the main bronchi to the fourth generation airways, by inserting a cytology brush (Cook Veterinary Products, Eight Mile Plains, Australia) into the different segments. Bronchial cells were obtained using 20 gentle upward and downward strokes of the brush against the airway walls. Care was taken to avoid bleeding. Bronchi were not irrigated with physiological serum before brushing, to conserve leukocytes and to ensure that samples were representative of the cellular changes occurring within the bronchi of diseased horses.
Cell processing
After retraction of the brush into its protective sheath and its removal from the bronchoscope channel, collected cells were dislodged by shaking the brush into 15-ml conical tubes containing ice-cold RPMI 1640 medium (Life Technologies, Merelbeke, Belgium) supplemented with 1% glutamine, 10% FBS, 50 μg/ml gentamicin, and 10 μg/ml amphotericin B. The harvested cell suspension was vortexed and filtered through gauze to remove mucus. The cells were then centrifuged at 800 × g for 5 min, and the pellet was resuspended in LHC-8 complete medium without hydrocortisone (Biofluids, Rockville, MD) supplemented with 10 μg/ml amphotericin B. The cells were then incubated at 37°C in a 5% CO2-95% air mixture for different times before protein extraction. The minimal culture time before protein extraction was 3 h. Cell density was assessed by the use of a hemocytometer, and cell viability was evaluated by propidium iodide exclusion (5 μg/ml of culture medium). Cell differentials were performed on cytospin preparations stained with Diff-Quick (Dade Behring, Dudingen, Germany). Where necessary, polymorphonuclear cells were separated from the other cells using Histopaque centrifugation (specific gravity, 1.077; Sigma, Bornem, Belgium).
Cytoplasmic and nuclear protein extraction
Cytoplasmic and nuclear protein extracts were prepared as previously described (17). Cytoplasmic buffer contained 10 mM HEPES (pH 7.9), 10 mM KCl, 2 mM MgCl2, 0.1 mM EDTA, 0.2% (v/v) Nonidet P-40, and 1.6 mg/ml protease inhibitors (Complete; Roche, Mannheim, Germany). The pelleted nuclei were resuspended in 20 mM HEPES (pH 7.9), 1.5 mM MgCl2, 0.2 mM EDTA, 0.63 M NaCl, 25% (v/v) glycerol, and 1.6 mg/ml protease inhibitors (nuclear buffer), incubated for 20 min at 4°C, and centrifuged for 30 min at 12,000 × g (Eppendorf centrifuge 5415C; Eppendorf Scientific, Hamburg, Germany). Protein concentrations were quantified with the Micro bicinchoninic acid protein assay reagent kit (Pierce, Rockford, IL).
Anti-IκB Abs
The anti-IκB Abs used were 1) a mouse mAb directed against IκB-α (a gift from Katrina Wood, University of Oxford, Oxford, U.K.); 2) a rabbit polyclonal Ab recognizing an NH2-terminal peptide of mouse IκB-β (Santa Cruz Biotechnology, Santa Cruz, CA); 3) a mouse mAb directed against aa 1-444 of the human p52 subunit (Upstate Biotechnology, Lake Placid, NY); and 4) a rabbit polyclonal Ab recognizing an NH2-terminal peptide (aa 1–12) of human p50 (Upstate Biotechnology). Immunoblot experiments performed with cytoplasmic extracts prepared from equine lymphocytes showed that all these Abs are equine reactive.
Plasmids
ScaI before coupled in vitro transcription and translation.
Coupled in vitro transcription and translation of IκB-β
Linearized pRc/CMV and pRc/CMV-HA-IκB-β plasmids were in vitro transcribed and translated simultaneously using the TnT T7 Coupled Wheat Germ Extract System (Promega, Madison, WI) according to the manufacturer’s instructions. The reactions were performed either with or without [35S]methionine in the transcription-translation mixture. The [35S]methionine-labeled translated products were analyzed by electrophoresis in a 10% polyacrylamide-SDS gel and autoradiography to verify that they contained proteins of the expected molecular mass of ∼47 kDa for HA-IκB-β (see Ref. 18 for the first characterization of IκB-β). Unlabeled translated products were used in EMSA experiments.
EMSAs
Binding reactions were performed for 30 min at room temperature with 5 μg of nuclear proteins in 20 mM HEPES (pH 7.9), 10 mM KCl, 0.2 mM EDTA, 20% (v/v) glycerol, 1% (w/v) acetylated BSA, 3 μg of poly(dI-dC) (Amersham Pharmacia Biotech, Aylesbury, U.K.), 1 mM DTT, 1 mM PMSF, and 100,000 cpm of 32P-labeled double-stranded oligonucleotide probes. Probes were prepared by annealing the appropriate single-stranded oligonucleotides (Eurogentech, Liege, Belgium) at 65°C for 10 min in 10 mM Tris, 1 mM EDTA, and 10 mM NaCl, followed by slow cooling to room temperature. The probes were then labeled by end-filling with the Klenow fragment of Escherichia coli DNA polymerase I (Roche), with [32P]dATP and [32P]dCTP (DuPont-New England Nuclear, Les Ulis, France). Labeled probes were purified by spin chromatography on Sephadex G-25 columns (Roche). DNA-protein complexes were separated from unbound probe on 4% native polyacrylamide gels at 150 V in 0.25 M Tris, 0.25 M sodium borate, and 0.5 mM EDTA, pH 8.0. Gels were vacuum dried and exposed to Fuji x-ray film (Tokyo, Japan) at −80°C for 12 h. The amount of specific complexes was determined by photodensitometry of the autoradiography (Gel Doc 2000; Bio-Rad, Hercules, CA). To confirm specificity, competition assays were performed with a 50-fold excess of unlabeled wild-type and mutated probes. The sequences of the oligonucleotides used in this work were as follows: wild-type palindromic κB probe (19), 5′-TTGGCAACGGCAGGGGAATTCCCCTCTCCTTAGGTT-3′; and mutated palindromic κB probe, 5′-TTGGCAACGGCAGATCTATTCCCCTCTCCTTAGGTT-3′.
For experiments performed with in vitro translation products, 2 μl of each reaction was incubated with the nuclear extracts obtained from BBSs for 30 min either before or after incubation with the radiolabeled probe.
Immunoblots
Protein extracts (10 μg) were added to a loading buffer (10 mM Tris-HCl (pH 6.8), 1% (w/v) SDS, 25% (v/v) glycerol, 0.1 mM 2-ME, and 0.03% (w/v) bromophenol blue), boiled, and run on a 10% SDS-PAGE gel. After electrotransfer to polyvinylidene difluoride membranes (Roche) and blocking overnight at 4°C with 20 mM Tris (pH 7.5), 500 mM NaCl, 0.2 (v/v) Tween 20 (Tris-HCl/Tween), and 5% (w/v) dry milk, the membranes were incubated for 1 h with the first Ab (1/250 dilution for IκB-α and 1/1000 dilution for the other IκB proteins), washed, and then incubated for 45 min with peroxidase-conjugated rabbit anti-mouse IgG (1/2000 dilution) for IκB-α and p52 (Dako, Glostrup, Denmark), or peroxidase-conjugated goat anti-rabbit IgG (1/5000 dilution) for the other IκB proteins (Kirkegaard & Perry, Gaithersburg, MD). The results of the reaction were revealed with the enhanced chemiluminescence detection method (ECL kit; Amersham Pharmacia Biotech). Equal loading of protein on the gels was confirmed in all experiments by probing the blots for either α-tubulin (Santa Cruz Biotechnology) in the case of cytoplasmic extracts or Oct-1 (Santa Cruz Biotechnology) in the case of nuclear extracts (data not shown).
Neutralization experiments
Neutralizing Abs directed against recombinant human IL-1β and TNF-α were purchased from Sigma. Anti-IL-1β and anti-TNF-α Abs were used at 3 or 8 μg/ml. These were incubated for 180 min before protein extraction with 3-h cultured BBSs obtained from heaves-affected horses, 21 days after allergen eviction.
Statistical analysis
Linear associations between variables were assessed by the use of standard least-square linear regressions. Correlation coefficients (r) were presented as measures of linear association for regression relationships. Significant differences of the slopes from zero were determined using two-tailed Student’s t test. The differences between mean values were estimated using Student’s t tests for unpaired data. p < 0.05 was considered significant.
Results
Cell number, type, and viability
The number of harvested cells averaged 19.9 ± 7.0 (mean ± SD) million cells/animal (range, 11–34 million). Differential cell counts showed a significant increase in the percentage of granulocytes in BBSs obtained from heaves-affected horses compared with healthy horses (Table I⇓). The viability of harvested granulocytes (92.5 ± 4.2%), as determined by propidium iodide exclusion, was significantly greater than the viability of the other cells present in BBSs (24.2 ± 9.3%). Accordingly, total cell viability measured in BBSs from heaves-affected horses was significantly higher than that measured in BBSs from healthy horses (Table I⇓). We previously reported lower percentages of granulocytes in BBSs from heaves-affected horses (i.e., 3.4 ± 0.7% during the crisis and 1.8 ± 1.3% 21 days after allergen eviction), and cell viability that was not significantly different between healthy and diseased horses (15). This could be imputed to the fact that bronchi were irrigated with physiological serum before brushing in this earlier study. Indeed, bronchial irrigation partly eliminates granulocytes, which are present in large quantities in diseased horses at the surface of the airway epithelium.
Characteristics of the cells recovered by bronchial brushing in healthy and heaves-affected horses
Positive correlation between the percentage of granulocytes and NF-κB activity in BBSs
Consistent with our previous studies (15), NF-κB activity was much greater in nuclear extracts prepared from BBSs of heaves-affected horses during crisis (Fig. 1⇓A, lanes 4, 6, and 8), when compared with extracts from BBSs of healthy horses (Fig. 1⇓A, lanes 1–3). Twenty-one days after the eviction of the causative agents, NF-κB activity was maintained at high or moderate levels in BBSs from diseased horses (Fig. 1⇓A, lanes 5, 7, and 9). As the percentage of granulocytes and NF-κB activity simultaneously increased in BBSs from heaves-affected horses, notably during crisis, correlations between the percentage of granulocytes in BBSs and the intensity of NF-κB DNA binding, as measured by photodensitometry, were calculated. These regression analyzes were conducted with the results obtained from four separate EMSAs. Correlation coefficients between the percentage of granulocytes and the intensity of specific NF-κB bands were 0.96 (p < 0.001, first gel), 0.97 (p < 0.001, second gel), 0.93 (p < 0.01, third gel), and 0.89 (p < 0.01, fourth gel). These significant correlations were all positive. Results obtained from a representative gel are shown in Fig. 1⇓.
Association between NF-κB activity and the percentage of granulocytes in BBSs obtained from healthy and heaves-affected horses. A, NF-κB DNA binding activity of nuclear protein extracts obtained from BBSs of three healthy horses (lanes 1–3) and three heaves-affected horses (lanes 4–9). Each heaves-affected horse was investigated on two occasions: during crisis (lanes 4, 6, and 8) and 21 days after removal from the causative environment (lanes 5, 7, and 9). The open arrow indicates specific NF-κB complexes revealed in cells from healthy horses. The solid arrow indicates specific complexes revealed in cells from heaves-affected horses. The filled columns show the amount of specific NF-κB complexes as determined by photodensitometry of the autoradiography. B, Percentages of granulocytes and nongranulocytic cells measured in the BBSs obtained from the six investigated horses. C, Relationships between the percentage of granulocytes and nongranulocytic cells, and the intensity of corresponding NF-κB bands (r is the correlation coefficient). This experiment is representative of four similar experiments performed in nine heaves-affected horses and six healthy horses.
NF-κB activity is increased in both granulocytes and nongranulocytic cells contained in BBSs from heaves-affected horses
The strong correlation between the intensity of NF-κB activity and the percentage of granulocytes presents in BBSs suggested that the increased NF-κB activity observed in BBSs from heaves-affected horses was restricted to polymorphonuclear cells. To verify this hypothesis, BBSs were recovered in heaves-affected horses 21 days after allergen eviction, and the granulocytes were separated from the other BBS cells using Histopaque centrifugation. Because BBSs contained many clusters made of various cell types, the use of specific immunological methods, such as flow cytometry, to separate the granulocytes from the other BBS cells was inadequate. Histopaque centrifugation only allowed partial cell separation. Two fractions were obtained: a granulocyte-enriched fraction, in which the percentage of granulocytes averaged 62.1 ± 12.2%, and a bronchial epithelial cell (BEC)-enriched fraction, in which the percentage of granulocytes averaged 9.5 ± 9.9%. EMSAs performed with total and enriched samples obtained simultaneously from the same heaves-affected horses showed identical NF-κB activities (Fig. 2⇓), indicating that NF-κB activity is similarly increased in both granulocytes and nongranulocytic cells contained in BBSs from diseased horses.
NF-κB activity in total, granulocyte-enriched, and BEC-enriched BBSs. Two BBSs were obtained simultaneously from the same heaves-affected horse 21 days after the crisis. One BBS was unmodified (total BBS, percentage of granulocytes, 28%). Cells from the second BBS were separated using Histopaque centrifugation. Two fractions were obtained: a granulocyte-enriched fraction (percentage of granulocytes, 62%) and a BEC-enriched fraction (percentage of granulocytes, 13%). Nuclear extracts were prepared from total and enriched samples and analyzed for NF-κB-binding activity by EMSA. This experiment is representative of three similar experiments.
Granulocytic death and NF-κB deactivation are concomitant in cultured BBSs from heaves-affected horses
NF-κB activity was maintained at high or moderate levels in the bronchi of heaves-affected horses 21 days after allergen eviction, indicating that NF-κB activity in asthma-like diseases does not necessarily require the continuous presence of the etiologic agent. To determine whether increased NF-κB activity is also sustained ex vivo, three BBSs obtained simultaneously from the same heaves-affected horses (n = 9) 21 days after the crisis were cultured for 3, 24, or 48 h before assessment of NF-κB-binding activity. Nuclear extracts prepared from BBSs cultured for 3 and 24 h demonstrated identical NF-κB activities (examples are given in Fig. 3⇓A, lanes 1, 2, 4, and 5). BBSs cultured for 48 h displayed NF-κB activities that were either similar to those observed at 3 and 24 h (n = 5; an example is provided in Fig. 3⇓A, lane 6) or drastically decreased (n = 4; an example is given in Fig. 3⇓A, lane 3).
Relationship between the kinetics of granulocytic death and the kinetics of NF-κB activity in cultured BBSs obtained from heaves-effected horses. A, Three BBSs obtained simultaneously from the same heaves-affected horses (n = 2) 21 days after allergen eviction were cultured for 3 h (lanes 1 and 4), 24 h (lanes 2 and 5), or 48 h (lanes 3 and 6) before assessment of NF-κB-binding activity by EMSA. The filled columns show the amount of specific NF-κB complexes as determined by photodensitometry of the autoradiography. B, Granulocyte viability and nongranulocytic cell viability measured in the BBSs from the two horses studied at each time point. C, Relationships between granulocyte (left panel) and nongranulocytic cell (right panel) viability and the intensity of the corresponding NF-κB bands (r is the correlation coefficient). This experiment is representative of three similar experiments performed in nine heaves-affected horses and six healthy horses.
At 24 h, granulocyte viability and other BBS cell viability averaged 83.4 ± 10.5 and 20.3 ± 11.2%. At 48 h, these values reached 55.3 ± 24.3 and 18.5 ± 10.3%, respectively. The large variance observed in granulocyte viability at 48 h was due to the fact that the rates of granulocytic death were not consistent among the BBSs. At this time point, granulocytic viability was indeed maintained at high levels in BBSs from five horses (an example is given in Fig. 3⇑B, lane 6), whereas granulocytic viability considerably decreased in BBSs from the four other horses (an example is provided in Fig. 3⇑B, lane 3). Reduced granulocytic viability was always accompanied by a drastic decrease in NF-κB activity (an example is given in Fig. 3⇑, A and B, lane 3). Conversely, when granulocytic viability was maintained at high levels, NF-κB activity was also sustained at high levels (an example is given in Fig. 3⇑, A and B, lane 6). The link between granulocytic death and abolishment of NF-κB activity in cultured BBSs from heaves-affected horses was confirmed by regression analyses, which were conducted with the results obtained from three separate EMSAs. Correlation coefficients between granulocyte viability and the intensity of NF-κB bands were 0.93 (p < 0.01, first gel), 0.89 (p < 0.01, second gel), and 0.87 (p < 0.05, third gel). Correlation coefficients between the viability of the other BBS cells and NF-κB activity were also calculated. These correlation coefficients were not significant and ranged from 0.29–0.39. Results obtained from a representative gel are shown in Fig. 3⇑. These data show that sustained NF-κB activity in cultured BBSs requires the presence of living granulocytes.
IκB-β is the most prominent IκB protein found in equine BBSs
In most cells stimulation leading to IκB-α proteolysis and nuclear translocation of NF-κB also results in the subsequent rapid NF-κB-dependent induction of IκB-α (20, 21). The reaccumulation of IκB-α following its loss allows a fast repression of NF-κB activity, thereby ensuring a transient NF-κB response. In cultured BBSs from heaves-affected horses, NF-κB activity was maintained as long as living granulocytes were present (Fig. 3⇑), suggesting that mutual regulation of NF-κB and IκB-α is impaired in these samples. Two hypotheses could account for this observation: either IκB-α is degraded as soon as it is resynthesized, preventing NF-κB deactivation, or IκB-α is not produced in the BBSs obtained from heaves-affected horses. To explore these hypotheses, the presence of all IκB proteins in cytoplasmic and nuclear extracts prepared from 3-h cultured BBSs of healthy and heaves-affected horses 21 days after allergen eviction was assessed by immunoblots.
Only very low amounts of IκB-α were detected in cytoplasmic and nuclear extracts obtained from BBSs of both healthy and heaves-affected horses (Fig. 4⇓). Similarly, p100 was undetectable (data not shown). Only small amounts of p105, the precursor of p50, were revealed by immunoblot in cytoplasmic extracts obtained from cells of healthy and heaves-affected horses (data not shown). On the contrary, significant amounts of IκB-β were observed in cytoplasmic extracts from BBSs of healthy horses, while considerably lower amounts of cytoplasmic IκB-β were observed in BBSs obtained from heaves-affected horses (Fig. 4⇓). IκB-β was not detectable in the nuclear extracts from BBSs of healthy and heaves-affected horses (Fig. 4⇓). These results indicate that IκB-β, rather than IκB-α, is the most prominent IκB protein found in the BBSs of horses and suggest that IκB-β is degraded in BBSs from heaves-affected horses.
Representative IκB-α and IκB-β immunoblots performed with cytoplasmic and nuclear extracts prepared from 3-h cultured BBSs of healthy and heaves-affected horses 21 days after allergen eviction. Arrows indicate the position of the specific bands. This experiment is representative of five similar experiments.
Neutralizing anti-IL-1β and anti-TNF-α Abs inhibit the persistent IκB-β degradation and NF-κB activation in BBSs from heaves-affected horses
NF-κB stimulates the production of IL-1β and TNF-α. These proinflammatory cytokines induce the degradation of the IκB proteins and the subsequent activation of NF-κB, thus initiating autoregulatory feedback loops (for review, see Ref. 22). To determine whether these loops were involved in the sustained IκB-β degradation and NF-κB activation observed in the BBSs of heaves-affected horses before granulocytic death, neutralizing anti-IL-1β and/or anti-TNF-α Abs were added to the medium of 3-h cultured BBSs obtained from heaves-affected horses 21 days after allergen eviction. The final Ab concentrations was either 3 or 8 μg/ml. Cytoplasmic and nuclear extracts were prepared from treated BBSs 180 min after Ab addition and were subsequently analyzed for IκB protein expression by immunoblot and for NF-κB binding activity by EMSA. Neither the addition of anti-IL-1β Abs (3 or 8 μg/ml) nor the addition of anti-TNF-α Abs (3 or 8 μg/ml) was able to reduce IκB-β degradation and NF-κB activation in BBSs (Fig. 5⇓A). Conversely, when added simultaneously, the anti-IL-1β and anti-TNF-α Abs (3 μg/ml of each Ab) drastically increased the cytoplasmic and nuclear amounts of IκB-β (Fig. 5⇓B) and markedly decreased the NF-κB activity in BBSs (Fig. 5⇓A), indicating that autoregulatory feedback loops involving both IL-1β and TNF-α are at least partly responsible for the sustained IκB-β degradation and NF-κB activation in the cultured BBSs from diseased horses. The cytoplasmic and nuclear amounts of IκB-α, p100, and p105 were not altered by the addition of neutralizing Abs to the culture medium.
Effects of neutralizing anti-IL-1β and/or anti-TNF-α Abs on NF-κB activity (A) and IκB-β expression (B) in BBSs. Three-hour cultured BBSs obtained from heaves-affected horses 21 days after allergen eviction were incubated for 180 min with neutralizing anti-IL-1β Ab (8 μg/ml), neutralizing anti-TNF-α Abs (8 μg/ml), or both neutralizing Abs (3 μg/ml for each Ab). Cytoplasmic and nuclear extracts were prepared from treated BBSs and were subsequently analyzed for NF-κB-binding activity by EMSA (A) and for IκB-β expression by immunoblot (B). This experiment is representative of three similar experiments.
IκB-β prevents p65 homodimer DNA binding and removes bound p65 homodimers from κB sites
We had previously demonstrated that active NF-κB complexes found in BBSs of heaves-affected horses were mainly p65 homodimers, rather than classical p65-p50 heterodimers (15). Knowing that IκB-β is the most prominent IκB protein present in equine BBSs (Fig. 4⇑) and that the appearance of IκB-β in the nucleus of BBS cells and NF-κB deactivation are concomitant (Fig. 5⇑), we hypothesized that IκB-β is able to prevent p65 homodimer DNA binding and to displace bound p65 homodimers from their κB sites. To verify this hypothesis, the effects of recombinant IκB-β on p65 homodimer DNA binding were investigated using EMSAs. First, either the linearized plasmid vector pRc/CMV or this same linearized plasmid containing the IκB-β cDNA insert (pRc/CMV-HA-IκB-β) was used as DNA templates in a coupled in vitro transcription/translation reaction system. The [35S]methionine-labeled translated products were analyzed by electrophoresis in a 10% polyacrylamide-SDS gel and by autoradiography. A major translation product of ∼47 kDa was synthesized from pRc/CMV-HA-IκB-β, consistent with the predicted size of HA-IκB-β (Fig. 6⇓A). Two microliters of each unlabeled lysate was then incubated for 30 min with nuclear extracts obtained from BBSs of heaves-affected horses, either before or after incubation with the radiolabeled κB probe. Afterward, DNA-protein complexes were analyzed by EMSA. Mock-translated products were unable to alter p65 homodimer DNA-binding (Fig. 6⇓B, lanes 2 and 3). On the contrary, in vitro translated IκB-β was able to prevent p65 homodimer DNA binding and remove bound p65 homodimers from their κB sites (Fig. 6⇓B, lanes 4 and 5, respectively).
Effects of in vitro translated IκB-β on p65 homodimer DNA binding. A, In vitro transcription/translation reactions were performed using equivalent amounts of pRc/CMV or pRc/CMV-HA-IκB-β plasmid DNA. The [35S]methionine-labeled translated products were separated by SDS-PAGE and visualized by autoradiography. B, Two microliters of each unlabeled translated product was incubated for 30 min with nuclear extracts known to contain active p65 homodimers (i.e., nuclear extracts obtained from heaves-affected horses 21 days after allergen eviction), either before (b; lane 2 for pRc/CMV and lane 4 for pRc/CMV-HA-IκB-β) or after (a; lane 3 for pRc/CMV and lane 5 for pRc/CMV-HA-IκB-β) incubation with the radiolabeled κB probe. This experiment is representative of three similar experiments.
Discussion
The chronic inflammation of the airways described in asthma is due to persistent overexpression of multiple inflammatory proteins, including proinflammatory cytokines, chemokines, adhesion molecules, and inflammatory enzymes (1, 4). Despite their crucial role in asthma pathogenesis, the precise mechanisms responsible for this sustained gene overexpression have been poorly investigated and remain obscure. We previously demonstrated that NF-κB activity, which plays a key role in the transcriptional initiation of many inflammatory genes, drastically increases in the bronchi of heaves-affected horses after allergen exposure and is maintained at high or moderate levels in the bronchi of diseased horses for at least 21 days after allergen eviction from their environment (15). Here, we show that this increased NF-κB activity also persists for at least 24–48 h in cultured BBSs obtained from heaves-affected horses 21 days after allergen eviction. As NF-κB activity usually returns to basal level within a few hours after the removal of external stimuli (23), our observations indicate that NF-κB activity is aberrantly sustained in the bronchial cells of heaves-affected horses after allergen eviction and strongly suggest that this persistent transcriptional activity could account for the maintained expression of inflammatory genes in asthmatic bronchi.
The accumulation of active granulocytes in the airways is thought to be of particular importance in the development of clinical asthma (24). Indeed, several investigators have demonstrated a positive and significant correlation between eosinophil profusion in the airways and lung dysfunction in asthmatic patients (25, 26). Furthermore, the resolution of eosinophilic inflammation, which depends upon eosinophil apoptosis, is associated with clinical improvement of asthma (27). In the present study, the levels of NF-κB activity were high or moderate in granulocytes and nongranulocytic cells contained in BBSs obtained from heaves-affected horses 21 days after allergen removal. However, NF-κB activity strongly correlated with the percentage of granulocytes present in BBSs and was completely abrogated after granulocytic death, suggesting that the sustained NF-κB activation that occurs in the airways of heaves-affected horses is mainly driven by the inflammatory cells that remain or appear in the bronchi after allergen eviction. As the level of NF-κB activity in the bronchi is closely related to the degree of pulmonary dysfunction (15), our results also provide a new insight into the molecular mechanisms by which the granulocytes impair lung function. First, the level of NF-κB activity in the granulocytes probably determines the amounts of broncho- and vasoactive inflammatory mediators released by these cells. Second, one may assume that activated granulocytes also secrete cytokines that are able to initiate the NF-κB-dependent synthesis of inflammatory mediators by other bronchial cells. Granulocytic death and clearance would be prerequisites for the cessation of the direct and indirect effects of these cells on lung function.
The physiological half-life of the circulating neutrophil, the most abundant granulocyte, is only 6 h (for review, see Ref. 28). In the present study, granulocyte viability was much longer, indicating that protective mechanisms delay inflammatory cell death in asthmatic airways. TNF-α, which was produced by BBSs from heaves-affected horses, induces apoptosis of ex vivo cultured neutrophils at early time points, but inhibits apoptosis after culture for 18 h (29). This delayed protective effect is lost when protein synthesis is inhibited, indicating that TNF-α induces anti-apoptotic proteins that protect neutrophils which avoid early death (for review, see Ref. 30). It is likely that inflammatory cells that invade the site of inflammation are those that express anti-apoptotic proteins and are subsequently protected from death. This possibility might explain the prolonged survival observed in granulocytes from heaves-affected horses, even in the presence of TNF-α. A second hypothesis could account for the increased survival of inflammatory cells from asthmatic bronchi. Indeed, many cytokines present at inflammatory sites, such as GM-CSF, are able to delay granulocyte apoptosis (for review, see Ref. 30). These anti-apoptotic cytokines could counteract the cytotoxic effects of TNF-α.
An intriguing question concerns the maintenance of NF-κB activity in the bronchi before granulocytic death. A hallmark of many NF-κB-dependent genes that are switched on in inflammatory diseases is that their expression can be induced by the proinflammatory cytokines IL-1β and TNF-α (31). Activated granulocytes generate high amounts of IL-1β and TNF-α (32, 33, 34). IL-1β and TNF-α activate NF-κB, which, in turn, induces the expression of these proinflammatory cytokines, thus initiating autoregulatory feedback loops (for review, see Ref. 22). Accordingly, we postulated that these autoregulatory feedback loops might be involved in the granulocyte-mediated persistent NF-κB activity in bronchial cells of heaves-affected horses after allergen eviction. Addition of both neutralizing anti-IL-1β and anti-TNF-α Abs to cultured BBSs from heaves-affected horses resulted in the suppression of NF-κB activity, confirming our hypothesis. These findings are consistent with previous data from Lentsch et al. (35), who showed that NF-κB activity occurs in a time course similar to that for the production of IL-1β and TNF-α during IgG immune complex-induced lung injury. Interestingly, when either the anti-IL-1β or anti-TNF-α Abs was added individually to the medium of cultured BBSs, each was incapable of reducing NF-κB activity. These results indicate that either cytokine is independently able to maximally stimulate the signaling pathway leading to NF-κB activation in the cultured BBSs from heaves-affected horses and emphasize that the antagonization of a single cytokine would probably have a minor effect on allergic inflammation.
IκB-β, rather than IκB-α, was the most prominent IκB protein found in BBSs from healthy and heaves-affected horses. Moreover, the appearance of IκB-β in the nucleus of BBS cells from heaves-affected horses was accompanied by NF-κB deactivation, as observed after treatment with anti-IL-1β and anti-TNF-α Abs. Finally, recombinant IκB-β was able to prevent DNA binding by p65 homodimers, which are the most abundant NF-κB complexes found in equine BBSs (15) and was able to remove bound p65 homodimers from their κB sites. Because BBSs contained mainly BECs in healthy horses and BECs plus granulocytes in heaves-affected horses, our results unambiguously demonstrate that the inhibition of p65 homodimers by IκB-β regulates NF-κB-dependent gene expression in equine bronchial epithelial cells and in equine bronchial granulocytes. However, large amounts of IκB-α are observed in equine blood granulocytes (our unpublished observations), suggesting a shift in IκB protein expression during granulocyte migration and activation. Although unexpected, our results are in accordance with those of Lentsch et al. (35, 36), who demonstrated that deactivation of NF-κB complexes predominantly composed of p65 by secretory leukocyte protease inhibitor is associated with increased levels of IκB-β, but not IκB-α, in a rat model of IgG immune complex-induced alveolitis. Furthermore, previous in vitro studies have demonstrated that IκB-β (37) as well as IκB-ε (38) preferentially inhibit the p65 homodimeric form of NF-κB. As no equine reactive anti-IκB-ε Ab is available, the inhibitory function of IκB-ε in BBSs from heaves-affected horses was not investigated.
IκB-β has been demonstrated to be involved in the persistent NF-κB activity observed in some cells, including activated 70Z/3 pre-B cells (39), WEHI 231 mature B cells (40), HIV-1-infected myeloid cells (41), and T cells infected by the human T cell leukemia virus type 1 (42). In human T cell leukemia virus type 1-infected T cells, the persistent NF-κB activity results from the continuous degradation of IκB-β by the virally encoded Tax protein (42). In the other cells stimulation results in the degradation of IκB-β, which is resynthesized in a hypophosphorylated form that sustains NF-κB activation (39, 40, 41). Indeed, hypophosphorylated IκB-β interacts with NF-κB without masking its nuclear localization signal and its DNA binding site, thus acting as a chaperone for NF-κB nuclear entry and activity. Furthermore, hypophosphorylated IκB-β prevents NF-κB resequestration by IκB-α. However, IκB-β was not associated with active p65 homodimers in the nucleus of BBS cells obtained from heaves-affected horses, indicating that hypophosphorylated IκB-β is not involved in the sustained NF-κB activity found in these cells.
Rapid NF-κB-dependent resynthesis of IκB-α establishes an autoregulatory loop by which NF-κB activation is self-limited (Refs. 20 and 21 ; for review, see Ref. 43). Conversely, NF-κB activation does not induce IκB-β overexpression, indicating that NF-κB complexes exclusively released from IκB-β are not inhibited by an autoregulatory feedback mechanism (18). This observation led Thompson et al. (18) to anticipate that activation of NF-κB would probably be persistent in tissues lacking IκB-α, because no feedback inhibition through increased synthesis of IκB-α would occur in these tissues. In the present report, IκB-α was lacking, and IκB-β was continuously degraded by proinflammatory cytokines in BBSs from heaves-affected horses, providing the first in vivo example of the mechanism of sustained NF-κB activity theoretically described by Thompson et al. (18).
Taken together, our results allow us to propose a cellular and molecular model that accounts for the persistent NF-κB activity occurring in the bronchi of heaves-affected horses after allergen eviction (Fig. 7⇓). In this model the NF-κB activity is maintained as long as there are living granulocytes in the airways. Before granulocytic death, NF-κB activity is maintained in all bronchial cells by granulocyte-dependent autoregulatory feedback loops involving the proinflammatory cytokines IL-1β and TNF-α. IκB-β is expressed at a basal level unaffected by cell stimulation with IL-1β and TNF-α, is rapidly degraded, does not reach the nucleus, and is subsequently unable to stop the cytokine-mediated NF-κB activation. After granulocytic death, the autoregulatory feedback loops involving IL-1β and TNF-α are strongly attenuated, allowing IκB-β-dependent NF-κB deactivation in nongranulocytic cells.
Proposed cellular and molecular model for the persistent NF-κB activity occurring in the bronchi of heaves-affected horses after allergen eviction. Before granulocytic death, NF-κB activity is maintained in all bronchial cells by granulocyte-dependent autoregulatory feedback loops involving the proinflammatory cytokines IL-1β and TNF-α. The basal level of IκB-β expression, which does not increase after cell stimulation with IL-1β and TNF-α, is insufficient to stop the cytokine-mediated NF-κB activation. After granulocytic death, the autoregulatory feedback loops involving IL-1β and TNF-α are strongly attenuated, allowing IκB-β-dependent NF-κB deactivation in nongranulocytic cells.
Acknowledgments
We thank Dr. Katrina Wood (University of Oxford, Oxford, U.K.) for the mouse mAb directed against IκB-α; Drs. Alain Israël and Robert Weil (Institut Pasteur, Paris, France) for the pRc/CMV-HA-IκB-β expression vector; Drs. Pierre Chatelain, Jacques Gielen, Renaud Louis, Roy Massingham, and Jacques Piette for advice; Dr. Charlotte Sandersen for horse management; Renata Turlej for manuscript correction; and Carine Gresse, Martine Leblond, Michel Motkin, Jean-François Rouelle, and Ilham Sbaï for excellent technical and secretarial assistance.
Footnotes
-
↵1 This work was supported by grants from the National Fund for Scientific Research (Belgium), Union Chimique Belge Pharma (Belgium), and the Ministère de la Région Wallonne (Belgium). F.B. is Research Assistant, M.-P.M. and A.V. are Research Associates, and V.B. is Senior Research Associate at the National Fund for Scientific Research (Belgium). G.B. is a fellow from the Biotechnology Program (European Commission). S.D., L.F., and N.K. are supported by the Fonds de la Formation a la Recherche Dans l’ Industrie et l’ Agriculture fellowships (Belgium).
-
↵2 Address correspondence and reprint requests to Dr. Fabrice Bureau, Department of Physiology, Faculty of Veterinary Medicine, University of Liege, Bâtiment B42, Sart Tilman, B-4000 Liege, Belgium. E-mail address: fabrice.bureau{at}ulg.ac.be
-
↵3 Abbreviations used in this paper: BBS, bronchial brushing sample; HA, hemagglutinin; BEC, bronchial epithelial cell.
- Received February 22, 2000.
- Accepted August 23, 2000.
- Copyright © 2000 by The American Association of Immunologists
References
In this issue
