Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Airway inflammation plays a major role in the pathophysiology of many lung-disease states, such as asthma, cystic fibrosis (CF), adult respiratory distress syndrome (ARDS), and bronchopulmonary dysplasia (BPD). Proinflammatory cytokines, such as interleukin-8 (IL-8), are responsible in part for the initiation and maintenance of airway inflammation. IL-8 is the major chemotactic and activating factor for polymorphonuclear neutrophils (PMN) in the airway (1). In a number of inflammatory disease states, the detection of increased IL-8 protein in lower-airway secretions is associated with increased morbidity and mortality (2).

A number of proinflammatory cytokines, such as IL-8, are regulated at the level of gene transcription by nuclear factor-kappa B (NF-B) (7). Studies with nonairway epithelial cells have shown that NF-B is present in unstimulated cells in the cytoplasmic compartment, bound to its inhibitory protein, IB. Upon exposure to inflammatory stimuli, such as tumor necrosis factor- (TNF-), IB is phosphorylated and targeted for degradation via the ubiquitin-proteasome pathway (8). Degradation of IB allows for translocation of NF-B to the nucleus, where it acts as a regulatory element for gene transcription.

The studies presented here focused on the mechanism of NF-B activation, IL-8 gene transcription, and IL-8 protein release in A549 cells, an airway epithelial-like cell line. Previous studies in our laboratory demonstrated that A549 cells respond to viral infection in a manner similar to that of other airway epithelial cells (9). Since the airway epithelial cell is a major source of proinflammatory cytokines, we focused on the regulation of NF-B activation in A549 cells. Our working hypothesis was that inhibition of proteasome-mediated IB degradation would result in a reversal of TNF--induced NF-B activation, IL-8 gene transcription, and IL-8 protein release from A549 cells. The studies presented here demonstrate that the proteasome inhibitor MG-132 reversed the effects of TNF- on IL-8 protein production by airway epithelial cells in a mechanism involving IB degradation and translocation of NF-B to the nucleus.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Cell Culture

Airway epithelial cells from the human lung-carcinoma cell line A549 (ATCC CCL 185, passage numbers 85 to 95; American Type Culture Collection, Rockville, MD) were used for these studies. A549 cells were maintained in Dulbecco's modified Eagle's medium (DMEM; Sigma Chemical Company, St. Louis, MO) supplemented with 8% fetal calf serum (FCS; Gibco BRL, Gaithersburg, MD) and L-glutamine (Gibco BRL) in ventilated plastic flasks. For experiments, cells were detached from the plastic by incubation in 0.05% trypsin and 0.01% ethylenediamine tetraacetic acid (EDTA) for 15 min, and were seeded at 30,000 cells/cm² on the day prior to infection. The cells were generally 70% to 90% confluent on the day after seeding. Cells treated with TNF- were washed once with DMEM, followed by the addition of 100 U/ml of TNF-. Preliminary experiments revealed that TNF- increased NF-B translocation to the cell nuclei within 5 min after addition, with a peak effect noted between 15 and 30 min. Similar results were obtained when evaluating IB degradation in the cytoplasmic extracts. Additionally, the use of 100 U/ml of TNF- had the greatest observed effect without causing toxicity to the cells. Therefore, the cells were treated for 15 min with TNF- at 100 U/ml when nuclear and cytoplasmic extracts were obtained. For studies dependent on protein synthesis by the cells (luciferase studies and IL-8-protein production studies), preliminary experiments demonstrated maximal effects of TNF- at 1 to 3 h after its addition. Therefore, TNF- was added for 2 h for these experiments. TNF- was obtained from R&D Systems (Minneapolis, MN).

Cbz-Leu-Leu-leucinal (MG-132) was obtained from Sigma Chemicals. In studies exploring the use of MG-132, the compound was dissolved in dimethyl sulfoxide (DMSO) on the day of use, and was added to the cells at a concentration of 10 µM for 1 h. After the 1-h preincubation, TNF- or control medium was added for an additional 15 min. In all studies, the concentration of DMSO was always less than 0.1% (vol/vol).

Construction of Plasmids Containing the Luciferase Reporter Gene under the Transcriptional Control of the 200-bp 5' Flanking Region of the IL-8 Gene, and Mutation of the NF-B Site in That Region

The 200-bp 5' flanking region of the IL-8 gene was obtained by polymerase chain reaction (PCR) amplification of human genomic DNA. The amplified region of the IL-8 gene consisted of nucleotides 97 through +103 (+103 = adenine-thymine-guanine [ATG] start site). This 200-bp segment of DNA contains previously described transcription-regulatory sites (i.e., NF-B and NF-IL-6) and the Kozak consensus translation start sequence (10). In order to determine its transcriptional activity with a reporter gene, the 200-bp IL-8 DNA fragment (or the fragment with the mutated NF-B site [11]) was cloned into the Kpn1 and BglII sites of the pGL2_basic vector. This vector contains the coding region of the firefly luciferase gene, but lacks any DNA fragments that confer transcriptional activity on the vector. Thus, introduction of DNA 5' to the luciferase gene allowed us to test the transcriptional activity of the DNA. After the recombinant vector was amplified and purified, restriction-enzyme mapping of the recombinant IL-8/pGL_basic was used to confirm that the 5' flanking region of the IL-8 gene had been inserted, in a single copy, in the proper orientation.

Transfection of A549 Cells with Recombinant pGL2 and Analysis of IL-8 Promoter Activity in Response to TNF- Stimulation

A549 cells were transfected through the technique of calcium phosphate precipitation (12). On the day before transfection, A549 cells were seeded at 12,000 cells/cm². The cells were typically 30% to 40% confluent 16 h later. The A549 cells were cotransfected with 4 pmol of the plasmid pRC/-galactosidase (a plasmid containing the -galactosidase gene under the control of the cytomegalovirus [CMV] promoter, which is constitutively active in A549 cells) and 8 pmol of either the promoterless pGL2_basic or one of the recombinant plasmid constructs. The DNA was resuspended in 125 mM CaCl₂, and an equal volume of 4-(2-hydroxyethyl)-1-piperazine-N'-2-ethanesulfonic acid (HEPES)-buffered saline was added with bubbling over a period of 15 s. After a 10-min incubation at room temperature, the DNA was added to the cells and the cells were incubated at 37°C overnight. The next day, the cells were washed three times with Hanks' buffered saline and DMEM with 8% FCS was added.

At approximately 90% confluence, the cells were treated with TNF- at 100 U/ml as described earlier. Two hours after control medium or TNF- was added, the cells were harvested by lysis according to the Promega (Madison, WI) luciferase-assay-system protocol. Briefly, the cells were incubated in lysis buffer (25 mM Tris-HCl, 10 mM EDTA, 15% sucrose, and 2 mg/ml lysozyme) for 15 min, scraped from the plates, and centrifuged to remove any debris. The supernatant was stored at 70°C until assayed.

To control for transfection efficiency, -galactosidase activity was determined for each of the lysates (13). Cell lysates were diluted in PM-2 buffer (20 mM NaH₂PO₄, 80 mM Na₂HPO₄, 0.1 mM MnCl₂, 2 mM MgSO₄, and 40 mM -mercaptoethanol, pH = 7.3). -Galactosidase activity was developed by adding the diluted lysate to 4 mg/ml O-nitrophenyl--galactopyranoside (ONPG) and incubating at 37°C. Color development was arrested with 1.0 M Na₂CO₃, and absorbance was determined at 420 nm with a spectrophotometer (DU 64; Beckman, Arlington Heights, IL).

Luciferase activity (14) measurements were made with a commercially available system according to the manufacturer's directions (Promega). An aliquot of cell lysate was mixed with 100 µl of 20 mM tricine, 1.0 mM (MgCO₃)Mg- (OH) · H₂O₂, 2.6 mM MgSO₄, 0.1 mM EDTA, 33.3 mM dithiothreitol (DTT), 270 µM coenzyme A, 470 µM luciferin, and 530 µM adenosine triphosphate (ATP), and the sample was placed in a Monolight 2000 Luminometer (Analytical Luminescence Laboratory, San Diego, CA) for determination of light intensity. Measurements were made at 10 s.

Enzyme-linked Immunosorbent Assay for IL-8

The enzyme-linked immunosorbent assay (ELISA) used in the study has been previously described (9). All plates were read on a Molecular Devices Microplate reader (Molecular Devices, Menlo Park, CA) and analyzed using a computer assisted analysis program (SoftMax, Molecular Devices). Typically, standard curves generated with this ELISA were linear in the range of 20 pg to 5,000 pg IL-8/ml. Only assays having standard curves with a calculated regression line (R) value > 0.95 were accepted for further analysis.

Isolation of Cytoplasmic and Nuclear Extracts

For isolation of nuclear extracts, all procedures were performed on ice. Nearly confluent monolayers of A549 cells, which had been treated with 100 U/ml of TNF- or control medium for the appropriate time, were washed with ice-cold phosphate-buffered saline (PBS), harvested by scraping into 1 ml of PBS, and pelleted in a 1.5-ml microfuge tube at 6,000 RPM for 5 min. The pellet was washed twice in ice-cold PBS and pelleted. The pellet was suspended in one packed-cell volume of lysis buffer (10 mM HEPES, pH 7.9; 10 mM KCl; 0.1 mM EDTA; 1.5 mM MgCl₂; 0.25 vol% Nonidet-P40; 1 mM DTT, and 0.1 mM phenylmethylsulfonyl fluoride [PMSF]). After a 5-min incubation on ice, the nuclear pellet was isolated by centrifugation. The supernatant represented the cytoplasmic extract. The nuclear pellet was resuspended in one packed-cell volume of extract buffer (20 mM HEPES, pH 7.9; 420 mM NaCl; 0.1 mM EDTA; 1.5 mM MgCl₂; 25% (vol/vol) glycerol; 1 mM DTT; and 0.5 mM PMSF), and the nuclei were incubated on ice for 20 min. The nuclear debris was removed by centrifugation, and the protein concentration of the nuclear extract was determined. The nuclear extracts were stored at 70°C until further use.

Electrophoretic Mobility-shift Assays

Electrophoretic mobility-shift assays (EMSAs) were done as previously described (11). Equivalent amounts of nuclear protein were incubated on ice for 10 min in a buffer containing 12 mM HEPES, pH 7.9; 4 mM Tris-HCl, pH 7.9; 25 mM KCl; 5 mM MgCl₂; 1 mM EDTA; 1 mM DTT; 50 ng/ml poly[d(I-C)]; and 0.2 mM PMSF. EMSA probes for NF-B and activator protein-1 (AP-1) were obtained from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA). The probes were end-labeled with ³²P, using T4 kinase (Promega), after which unincorporated nucleotides were removed with a G-25 Sephadex separation column (Boehringer Mannheim, Indianapolis, IN). The respective labeled probe (100,000 dpm) was then added to the extracts and incubated with the reaction mixture for an additional 10 min on ice. Bound and free probe were resolved through nondenaturing polyacrylamide gel electrophoresis (PAGE). For experiments involving supershifting, antibodies were purchased from Santa Cruz Biotechnology.

Western Blot Analysis

For Western blot analysis, 50 µg of cytoplasmic protein was heated to boiling for 3 min. The samples were then subjected to electrophoresis on a 10% Tris-glycine gel (Novex, San Diego, CA) at 140V for 1.5 h. The protein was transferred to nitrocellulose membranes (0.45 µm pore size; Novex). The nonspecific protein binding sites on the membranes were blocked with 5% milk in Tris-buffered saline (TBS) (0.02 M Tris; 0.5 M NaCl; pH 7.5) with 0.05% Tween 20 (TBS-T). The membranes were then probed with an antibody to IB (c-21; Santa Cruz Biotechnology) or ubiquitin (Boehringer Mannheim). After washing the membranes in TBS-T, the membranes were probed with a goat antirabbit antibody conjugated to horseradish peroxidase (HRP; Calbiochem, La Jolla, CA). The immobilized protein, bound to the antibody of interest, was then detected through the ECL detection protocol (Amersham, Arlington Heights, IL).

Determination of Toxicity through Neutral Red Uptake

After treatment of A549 cells with either TNF-, MG-132, both or neither, 400 nM neutral red was added to the cultures. As described by Warner (15), only viable cells retain the dye. Two hours after addition of the dye, the cells were washed three times with PBS (pH 7.4) and lysed with 1:1 (vol/vol) 95% ethanol-100 mM sodium citrate (pH 4.2), and absorbance at 540 nm was measured.

Statistics

Analysis of data from experiments involving measurement of luciferase activity in response to respiratory syncytial virus (RSV) was first done through analysis of variance (ANOVA) to determine overall significance (P < 0.05), followed by individual t tests, using Bonferroni's method to make adjustments in the level of significance.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

An Intact NF-B Binding Site Is Necessary for the Effects of TNF- on Increased Transcriptional Activity of the 5' Flanking Region of the IL-8 Gene

To demonstrate that TNF- increased the transcriptional activity of the IL-8 gene, a fragment of the gene containing 200 bp of DNA proximal to the ATG start site (containing the NF-B site and NF-IL-6 sites) was studied with a previously described luciferase gene reporter system (9). This DNA fragment is transcriptionally active in response to TNF- in other cell lines (16), and to RSV in airway epithelial cells (19). The degree of TNF--induced luciferase activity (used as a measure of transcriptional activity of the gene fragment) is reported as an increase over control levels (control being defined as cells that were transfected with the same reporter construct, but not treated with TNF-). As shown in Figure 1 (n = 6 for all experiments), addition of TNF- to A549 cells transfected with the empty pGL2 plasmid resulted in no increase in luciferase activity over that of control cells (pGL2). However, when the luciferase gene was placed under the transcriptional control of the proximal 200 bp of the 5' flanking region of the IL-8 gene (pGL2-IL-8), a 10-fold increase in luciferase activity was observed over that of control cells (cells transfected with the same plasmid, but not treated with TNF-; P < 0.01 by ANOVA). In contrast, mutating the NF-B site resulted in a complete loss of luciferase activity in response to TNF- (pGL2-IL-8M). These data indicate that an intact NF-B site was critical for the increased transcriptional activity of the IL-8 gene in response to TNF-.

View larger version (8K): [in this window] [in a new window]   Figure 1.   Effects of TNF- on transcriptional activity of the 5' flanking region of the IL-8 gene. For each construct, the effects of 100 U/ml of TNF- on luciferase activity is reported as the increase over the control value. All measurements were controlled for transfection efficiency by cotransfecting with a CMV -galactosidase construct and measuring -galactosidase activity in the samples. pGL2 represents the pGL2 plasmid without an insert, pGL2-IL8 represents the plasmid with the proximal 200 bp of the 5' flanking region of the IL-8 gene, and pGL2-IL-8M represents the pGL2 plasmid with the proximal 200 bp of the IL-8 gene containing a mutated NF-B-binding site. Error bars represent the SD. The difference in luciferase activity measured is statistically significant when comparing pGL2-IL-8 with pGL2-IL-8M (P < 0.01).

TNF- Induces Rapid Loss of IB from the Cytoplasm and Increased NF-B Activation in A549 Cells

Having determined that TNF- increased the transcriptional activity of the 5' flanking region of the IL-8 gene in a mechanism dependent on the presence of an intact NF-B site, we focused our next set of experiments on the effects of TNF- on IB degradation and NF-B activation. A549 cells were treated with 100 U/ml of TNF-, and cytoplasmic and nuclear extracts were prepared at 5, 10, 15, 30, and 60 min after treatment. As illustrated in Figure 2A (representative of four experiments), Western blot analysis of the cytoplasmic extracts showed that addition of TNF- to the cells was associated with a rapid loss of IB from the cytoplasm. Western blot analysis of nuclear extracts did not reveal any change in the low levels of IB present in the nuclei in response to TNF-, indicating that the loss of cytoplasmic IB was not due to movement of IB into the nuclei (data not shown). In parallel with this, EMSA analysis revealed a rapid increase in NF-B binding activity in the nuclear extracts from treated cells (Figure 2B, representative of four experiments). These data indicate that TNF- stimulated NF-B activation in A549 cells in a mechanism associated with rapid loss of IB from the cytoplasm.

View larger version (67K): [in this window] [in a new window]   View larger version (48K): [in this window] [in a new window]   Figure 2.   (A) Western blot analysis of 50 µg of cytoplasmic extracts from control cells (C) or cells treated for the indicated time with TNF- at 100 U/ml, done with an antibody to IB. The arrow indicates the IB band. (B) EMSA analysis of NF-B-binding activity in nuclear extracts from control cells (C) or cells treated for the indicated time with TNF- at 100 U/ml. The single arrow corresponds to binding of NF-B to the probe, and the double arrow indicates the free probe. (C) EMSA analysis of cells under control conditions (lane 1), or TNF- stimulation (lanes 2 through 7), demonstrating the effects of adding 100× cold competitor (lane 3), or antibodies to p50 (lane 4), p65 (lane 5), c-Rel (lane 6), or p52 (lane 7).

To determine which components of NF-B were involved in the response to TNF in A549 cells, cold-competition and supershift experiments were performed. In Figure 2C, lane 1 represents control cells and lane 2 represents TNF--stimulated cells. Experiments illustrated in lanes 3 through 7 represent the use of nuclear extracts obtained from TNF- treated cells. Preincubation with a 100-fold excess of the nonradioactive NF-B probe resulted in a loss of signal (Figure 2C, lane 3), indicating that all of the bands observed in lane 2 represent NF-B binding. Nonradioactive probes for the AP-1 and octamer-binding factor-1 (OCT-1) genes did not result in a loss of signal (data not shown). Addition of an antibody to the p65 component of NF-B resulted in supershifting of a significant component of the TNF--induced NF-B signal (lane 5). However, attempts to supershift with antibodies to p50 (lane 4), p52 (lane 6), and cRel (lane 7) did not result in supershifting of the NF-B signal. These data indicate that TNF- induces activation of p65 homodimers in A549 cells.

Reversal of the Effects of TNF- on IL-8 Protein Production by the Proteasome Inhibitor N-cbz-Leu-Leu-leucinal (MG-132)

The next set of experiments focused on the effects of the proteasome inhibitor on IL-8 protein production by airway epithelial cells in response to TNF-. As illustrated in Figure 3, preincubation with 10 µM MG-132 for 1 h blocked the effects of TNF--stimulated IL-8 protein release (solid bars). The effect of MG-132 on IL-8 release did not result from retention of IL-8 in the cells, since the whole-cell lysates treated with MG-132 also contained less IL-8 than the cells treated with TNF- alone (hatched bars). It therefore appears that MG-132 inhibited TNF--induced IL-8 protein production. The effects of MG-132 were not generalized to all protein production, since total protein from the cell lysates was not altered by MG-132 in either control cells or cells treated with TNF- (data not shown).

View larger version (14K): [in this window] [in a new window]   Figure 3.   Effects of TNF- at 100 U/ml and pretreatment with 10 µM MG-132 on intracellular IL-8 (lysates) and IL-8 released into the medium from A549 cells (supernatant). Error bars represent SD. For both lysates and supernatants, the measured differences are statistically significant when comparing control with TNF--treated and TNF- with TNF- + MG-132-treated cells (P < 0.01 for both comparisons). The difference between the measured quantities of IL-8 was not statistically significant for control cells versus those treated with MG-132 alone.

The Proteasome Inhibitor MG-132 Reverses the Effects of TNF- on the Transcriptional Activity of the 5' Flanking Region of the IL-8 Gene

After investigating the effects of Mg-132 on TNF--induced IL-8 protein production, we focused our next experiment on determining whether the effects of MG-132 on IL-8 protein production were related to altered transcriptional activity of the 5' flanking region of the IL-8 gene. A549 cells transfected with the pGL2-IL-8 plasmid responded to TNF- with a 12-fold increase in luciferase activity over control cells (Figure 4). However, preincubation of the cells with 10 µM MG-132 reversed the effects of TNF- on the transcriptional activity of the IL-8 gene. Additionally, although control cells had minimal amounts of detectable IL-8 transcriptional activity, the addition of MG-132 appears to have eliminated all transcriptional activity, as illustrated by comparison of luciferase activity in the control cells treated with MG-132 with that in untreated control cells.

View larger version (6K): [in this window] [in a new window]   Figure 4.   Effects of MG-132 on TNF--induced transcriptional activation of the 5' flanking region of the IL-8 gene. The addition of MG-132 to the media had a statistically significant effect on TNF--induced luciferase production by A549 cells transfected with the pGL2-IL-8 plasmid (TNF-compared with TNF + MG-132-treated cells; P < 0.01). The differences between control cells and TNF + MG-132-treated cells, control cells and MG-132-treated cells, and TNF + MG-132-treated cells and MG-132-treated cells were not statistically significant (P > 0.05 for all three comparisons).

To determine that MG-132 did not affect the detection of luciferase activity in the assay, MG-132 was added (final concentration of 10 µM) to lysates from cells known to have increased luciferase activity. The presence of 10 µM MG-132 resulted in less than a 5% decrease in measured luciferase activity in positive controls (data not shown). This indicates that MG-132 did not alter the detection of luciferase in the assay.

The next experiment focused on determining that MG-132 did not have a nonspecific effect on gene transcription. MG-132 was added to A549 cells transfected with a plasmid containing the -galactosidase gene under the transcriptional control of the CMV early promoter. The addition of 10 µM MG-132 to A549 cells had no effect on -galactosidase activity in the cells (data not shown). This indicated that transcriptional activity of the CMV early promoter was not altered by MG-132.

The Proteasome Inhibitor MG-132 Reverses the Effects of TNF- on IB Degradation and NF-B Activation

The next set of experiments focused on determining whether the proteasome inhibitor blocked the effects of TNF- on IB degradation and NF-B activation. As illustrated in Figure 5A (which is representative of four experiments), addition of 100 U/ml TNF- for 10 min resulted in a rapid loss of IB from the cytoplasmic extracts (control: lane 1, TNF-: lane 2). However, pretreatment with MG-132 for 1 h reversed the effects of TNF- on IB degradation (lane 3). Similarly, a slight increase in IB was seen in cells treated with MG-132 alone (lane 4) as compared with control cells, suggesting that a small degree of IB turnover may have been affected by the proteasome inhibitor. Additionally, in lane 3, a second signal, slightly greater than that for IB, is present. This probably represents the phosphorylated form of IB (^pIB).

View larger version (41K): [in this window] [in a new window]   View larger version (67K): [in this window] [in a new window]   Figure 5.   (A) The addition of 10 µM MG-132 reversed the effects of TNF- on IB degradation. Lane 1 = control; lane 2 = TNF- (100 U/ml for 10 min); lane 3 = preincubation with 10 µM MG-132 for 1 h, followed by 100 U/ml of TNF- for 10 min; and lane 4 = 10 µM MG-132 for 1 h alone. pIB refers to the phosphorylated form of IB. (B) The addition of MG-132 reversed the effects of TNF- on NF-B activation. Lanes 1, 2, 3, and 4 represent samples as in (A). The single arrow indicates retarded migration of the probe as a result of binding of NF-B. The double arrow represents the unbound probe. (C ) EMSA analysis of cells after treatment with TNF- without (lane 1), or with (lanes 2 through 7) MG-132, demonstrating the effects of adding 100× cold competitor (lane 3), or antibodies to p50 (lane 4), p65 (lane 5), c-Rel (lane 6), or p52 (lane 7).

In addition to its effects on IB degradation, MG-132 also reversed the effects of TNF- on NF-B activation. As illustrated in Figure 5B (representative of four experiments), very low levels of NF-B activation were detected in control cells (lane 1), but addition of 100 U/ml TNF- for 10 min resulted in a marked increase in NF-B activation (lane 2). Pretreatment with 10 µM MG-132 for 1 h resulted in a reversal of TNF--induced NF-B activation (lane 3). Additionally, treatment with MG-132 alone resulted in NF-B activation levels slightly below those of control cells (lane 4 compared with lane 1).

To insure that MG-132 did not have a nonspecific effect on transcription-factor binding, EMSAs for AP-1 binding were performed. MG-132 did not decrease the effect of TNF- on AP-1 activation in A549 cells (data not shown).

In Figure 5B, a stray band is present in lane 3. To determine whether this represented a form of NF-B, cold competition and supershift experiments were performed. In Figure 5C, lane 1 represents TNF- induction of NF-B, and lane 2 represents the effects of 10 mM MG-132 on TNF--induced NF-B activation. Lanes 3 through 7 also represent experiments done with nuclear extracts from MG-132- and TNF--treated cells. Preincubation with a 100-fold excess of the nonradioactive NF-B probe resulted in a loss of the signal (lane 3) corresponding to bands a and b in the figure, but not to bands c or d. This indicates that bands a and b represent NF-B binding. Nonradioactive AP-1 and OCT-1 probes did not result in a loss of signal (data not shown). Addition of an antibody to the p65 component of NF-B resulted in partial supershifting of band a, but not bands b, c, or d. However, attempts to supershift with antibodies to p50 (lane 4), p52 (lane 6), and cRel (lane 7) did not result in supershifting of any of the bands.

MG-132 Does Not Alter Ubiquination of IB

The next set of experiments was designed to insure that MG-132 was not interfering with the ubiquination of IB. The results of these studies are presented in Figure 6 (which is representative of three separate experiments). Cytoplasmic extracts from A549 cells under control conditions (lane 1), treated with TNF- for 10 min (lane 2), pretreated with MG-132 for 1 h, and then exposed to TNF- for 10 min (lane 3) or treated with MG-132 alone for 1 h (lane 4), were subjected to Western blot analysis, using either an antibody to IB (left four lanes) or an antibody to ubiquitin (right four lanes). Western blot analysis done with the antibody to IB revealed two bands, evident at approximately 36 to 37 kD, in cells treated with MG-132 (lanes 3 and 4). These bands correspond to IB and phosphorylated IB. Additionally, a band appeared in both lanes at approximately 45 to 48 kD (a), and a second band appeared at approximately 50 to 55 kD (b). Because the molecular weight of a ubiquitin molecule is 8.6 kD (26), these bands probably represented the addition of one ubiquitin molecule (37 + 8.6 = 45.6 kD) in the case of a, and of two ubiquitin molecules (37 + 17.2 = 54.2 kD) in the case of b. The band labeled b was present whether probing was done with an antibody to IB or to ubiquitin. Additionally, with long exposures, a band corresponding to a was also present in lanes 3 and 4 in the blot probed for ubiquitin (data not shown). (The very weak signal for a as compared with b is probably explained by decreased antigenicity in a, which has only one ubiquitin molecule present, as compared with b, with two ubiquitin molecules.) Thus, the detection of an IB species that is ubiquinated (b, and possibly a) indicates that MG-132 does not alter IB ubiquination.

View larger version (55K): [in this window] [in a new window]   Figure 6.   Western blot analysis of cytoplasmic extracts from A549 cells done either with an antibody to IB (left, lanes 1 through 4) or an antibody to ubiquitin (right, lanes 1 through 4), showing that MG-132 does not alter ubiquination of IB. Lane 1 = control; lane 2 = TNF- (100 U/ml for 10 min); lane 3 = preincubation with 10 µM MG-132 for 1 h followed by TNF- 100 U/ml for 10 min; and lane 4 = 10 µM MG-132 for 1 h alone. The band labeled b was present on both blots. Bands labeled a, b, c, and d probably represent addition of increasing numbers of ubiquitin molecules to IB.

Further examination of the Western blot probed with the antibody to ubiquitin revealed additional bands at approximately 60 to 65 kD (c) and 70 to 75 kD (d). These bands probably represented the addition of three (37 + 25.6 = 62.6 kD) and four (37 + 34.4 = 71.4 kD) ubiquitin molecules. The increase in intensity observed in progressing from b to d probably represents increased antigenicity resulting from the addition of increasing numbers of ubiquitin molecules (which also explains why a was detected only after long exposures). Similarly, the addition of increasing numbers of ubiquitin molecules may either block antigenic sites on IB or change the conformation of IB in a way that alters its antigenicity. This would help explain: (1) the decreasing signal intensity of IB molecules represented by bands a and b when probing with an antibody to IB; and (2) why bands c and d were not detected when probing with an antibody to IB.

As a negative control, the blots were probed with antibodies to p65 and c-fos, neither of which detected bands a, b, c, or d. Taken together, these data support the concept that MG-132 does not alter ubiquination of IB.

The Effects of the Proteasome Inhibitor MG-132 Were Not Secondary to Cell Toxicity

To insure that the effects of MG-132 described earlier were not due to cell toxicity, we performed neutral-red-uptake assays on the cells. These studies showed minimal changes in neutral-red-uptake in the cells treated with 10 µM MG-132 as compared with control cells and cells treated with TNF- (data not shown). Therefore, the effects of MG-132 did not result from cell toxicity.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

The studies reported here demonstrate that the proteasome inhibitor MG-132 inhibited TNF--induced IB degradation and NF-B activation. This effect was associated with decreased IL-8 gene transcription and IL-8 protein release. MG-132 did not alter IB phosphorylation or ubiquination. The effects of MG-132 did not result from cell toxicity.

In a number of airway disorders, inflammation is associated with significant morbidity and mortality. A number of proinflammatory cytokines, including IL-8, are involved in the inflammatory process. Although it is unlikely that altering the production of a specific individual cytokine would have a significant effect on inflammation, limiting production of cytokines as a group may have significant effects on airway inflammation. Therefore, we sought to focus on NF-B activation, since a common feature of a number of cytokines (IL-1, IL-2, IL-6, IL-8, granulocyte/macrophage stimulating factor [GM-CSF], regulated on activation, normal T-cell expressed and secreted [RANTES], TNF-) is the role that NF-B plays in regulating their gene transcription (7). The data presented here show that altering NF-B activation via proteasome inhibition results in altered IL-8 gene transcription and protein release in A549 cells. IL-8 was used as a biologic marker because its gene transcription depends on NF-B activation, and it is the major PMN chemotactic agent in the airway.

Studies with nonairway cells have shown that NF-B activation may be regulated at several potential points, including signal transduction (oxygen-radical formation [20] and IB phosphorylation [21]), IB degradation (7), and nuclear translocation of NF-B (7). Our efforts were focused on IB degradation for the following reasons: (1) although antioxidants may inhibit NF-B-mediated cytokine production in some cell lines, data on the effects of NF-B in airway epithelial cells are not conclusive (22, 23); (2) even though TNF--induced stimulation of IB phosphorylation is a signal for the degradation of IB, phosphorylation of IB is not sufficient to activate NF-B (24, 25); and (3) the mechanisms involved in NF-B translocation have not been defined. MG-132 was chosen for our studies because it is a reversible inhibitor of the proteasome, which limits its effects in altering cell-cycle progression (26), it was not toxic to airway epithelial cells in the concentrations tested, and it did not have any effect on the assays used in this study. In our studies, MG-132 inhibited NF-B activation in a mechanism involving reversal of IB degradation. Furthermore, neither phosphorylation nor ubiquination of IB was altered. Thus, MG-132 appears to work specifically at the level of the proteasome.

Eukaryotic cells have two major proteolytic pathways (27): (1) Extracellular proteins enter the cell through receptor-mediated endocytosis and are degraded via the lysosomal pathway; (2) intracellular proteins are degraded by way of a nonlysosomal pathway involving ubiquination of the targeted protein. The results of the studies presented here indicate that TNF--induced IL-8 production is mediated via the second pathway, involving phosphorylation and ubiquination of IB. This pathway offers a rapid mechanism for controlling gene expression. Although the processes of phosphorylation and ubiquination may not be reversible, future studies, aimed at inhibiting proteasome-mediated IB degradation, may reveal a pharmacologic means of reversing inflammation in the airway. Similarly, increased production of IB may also offer a means of limiting airway inflammation.

In summary, we have shown that in A549 cells, TNF--induced IB degradation, NF-B activation, IL-8 gene transcription, and IL-8 protein production can be reversed by preincubating the cells with the reversible proteasome inhibitor MG-132. The effects of this inhibitor appear to be specific to the proteasome, since no alterations in IB phosphorylation or ubiquination were detected, and MG-132 was not toxic to the cells. Reversible proteasome inhibitors may offer a means of limiting NF-B-mediated cytokine production and subsequent airway inflammation.