Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Infiltration of leukocytes is a hallmark of the inflammation process and is mediated by several chemotactic factors, one of which is interleukin (IL)-8. IL-8 has diverse biologic properties, including chemotaxis of neutrophils and T lymphocytes (1), regulation of cell adhesion properties (2), activation of neutrophils (3), and modulation of histamine release (4, 5). These observations implicate IL-8 as a key factor in the pathogenesis of inflammatory diseases. We previously demonstrated, both in vivo and in vitro, that IL-8 messenger RNA (mRNA) and protein in airway epithelial cells are increased in monkeys after ozone exposure (6). This observation was also demonstrated in cultures of human airway epithelial cell (7). The appearance of IL-8 in the airway lumen and surface epithelium is consistent with neutrophil influx in monkeys after ozone exposure (8). Using IL-8-neutralizing antibody, we demonstrated an 80% inhibition of the chemotactic activity in conditioned media of ozone-exposed airway cultures. These observations suggest that an elevated IL-8 gene expression is an important step in regulating the migration and the activation of neutrophils in monkey airways after ozone exposure (6).

IL-8 is regulated at both the transcriptional and post-transcriptional levels. The 3' end of the IL-8 message contains the repetitive AUUUA unit, which is responsible for destabilization of a variety of cytokine mRNAs (9). Within the 5'-flanking region of the human IL-8 promoter are various motifs with potential for binding transcription factors activator protein (AP)-1, AP-2, AP-3, glucocorticoid receptor, nuclear factor (NF)-B, NF-IL-6, and octamer factor, etc. (10). It has been suggested that the region spanning the nucleotides 94 to 70 relative to the transcription start site of the IL-8 gene is essential for both induction and repression by certain stimuli (11). This region contains the putative binding sites for NF-IL-6 and NF-B transcription factors. The binding of NF-B factors is indispensable for the activation of IL-8 gene transcription (11). NF-IL-6 is required to attain maximal expression of the IL-8 gene. Several members of the NF-B family, such as p50 (NF-B1), p65 (RelA), c-Rel, and p52 (NF-B2), have been shown to bind the NF-B motif of the IL-8 promoter (12, 15, 16). Inactive NF-B normally resides in the cytosol, anchored by an inhibitory molecule, I-B (17). The major step in NF-B activation involves the dissociation of NF-B from I-B and its translocation to the nucleus (17). However, the binding of NF-B to the B site in the nucleus is redox-regulated (18). After the dissociation of I-B, reduction of NF-B protein is necessary for the binding of NF-B to its cis element (19). Thioredoxin (TRX), a small multifunctional protein that has a redox-active disulfide/dithiol within the conserved active site sequence Cys-Gly-Pro-Cys, has been demonstrated as the agent necessary for reduction of NF-B (20). Recently, it was shown that TRX protein can act as a potent costimulus of cytokine expression, including the IL-8 gene (21).

To further understand the nature of IL-8 gene induction, the effects of culture conditions on IL-8 gene expression in airway epithelial cells were examined. We observed that the vitamin A derivative all-trans-retinoic acid (ATRA) used in the culture medium has profoundly enhanced IL-8 gene expression in vitro. This result is consistent with other reports in which different cell systems were used, including fibroblasts (22), neuroblastoma cells (23), a human ovarian carcinoma cell line (24), and a human melanoma cell line (15). Studies have suggested an enhanced DNA-protein interaction on the NF-B site of IL-8 promoter region by ATRA through the enhanced synthesis of p50/p65 NF-B proteins (25), or through the removal of the inhibitors that hindered the normal NF-B binding activity (15). Despite these suggestions, the nature of ATRA-stimulated IL-8 gene expression is still not clear. In the present investigation, we demonstrate a novel mechanism that may be involved in the stimulation of IL-8 gene expression by ATRA. This mechanism involves activation of the NF-B transcriptional factor by TRX, whose expression is stimulated by vitamin A and its derivatives in airway epithelial cells (26).

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Sources of Tissues and Cell Culture Condition

Rhesus monkey airway tissues were obtained from the California Regional Primate Research Center, University of California at Davis. Human airway tissues were obtained from the autopsy lab or the organ transplant program of University of California Davis Medical Center. Procedures allowing the use of this excised organ for cell culture study were approved by a campus committee and are reviewed periodically.

Tracheobronchial epithelial (TBE) cells were isolated from monkey and human airway tissues and cultured on tissue culture- grade plastic dishes as described elsewhere (27, 28). After 6 to 10 d incubation, when cultures were almost confluent, ATRA ranging from 0.1 nM to 1 µM was added and the cultures were harvested 24 to 48 h later. An S (serum-sensitive) clonal line of BEAS-2B cells was also included in the study with culture conditions similar to those described earlier. BEAS-2B is an immortalized human bronchial epithelial cell line created through the transfection of a primary culture of human bronchial epithelial cells with simian virus 40 early-region gene (29). The clonal line was kindly supplied by Dr. J. Lechner (Wayne State University, Detroit, MI).

Cycloheximide (CHX) (Sigma, St Louis, MO) was used to inhibit protein synthesis, and it was added at 10 µg/ml, 30 min before ATRA treatment. Actinomycin D (Sigma) was used to inhibit new transcript synthesis, and it was added to the culture medium at 10 µg/ml.

Quantitation of IL-8 by Enzyme-Linked Immunosorbent Assay

The level of IL-8 secretion was quantified by an enzyme-linked immunosorbent assay (ELISA) kit obtained from Biosource International (Camarillo, CA). The assayed value was normalized with the cell number in each dish. Triplicate dishes were used for each assay, and the assay was repeated at least three times in different primary and cell line cultures.

RNA Extraction and Northern Blotting Analysis

Total RNA was isolated by the single-step acid guanidium thiocyanate-phenol-chloroform extraction method as described elsewhere (30). RNA Northern blot hybridization was carried out as described elsewhere (31). Random primer labeling procedure was used to prepare [³²P]-labeled complementary DNA (cDNA) probes for hybridization. PG486, which is a PGEM 4Z-based clone (Promega, Madison, WI) containing a 486-base pair (bp) cDNA insert of the IL-8 coding region, was used as a template for cDNA probe preparation. The control cDNA probe for normalization of Northern blotting was a cDNA insert of 18S ribosomal RNA (rRNA), as described elsewhere (26).

Nuclear Run-On Assay

Nuclear run-on assay was performed according to the method of Greenberg and Ziff (32) with some modification (33). Duplicate slot membranes containing equal amounts of IL-8 (5 µg), small proline-rich protein (SPRR1) (33) (5 µg), and glyceraldehyde-3-phosphate dehydrogenase (GAP DH) cDNA inserts (0.5 µg) were prepared and used for hybridization. Equal radioactive counts of [³²P]RNA from each sample were used to hybridize the cDNA slot membranes using the same conditions as the Northern blot analysis.

Construction of the Chloramphenicol Acetyltransferase Expression Vector

DNA fragment corresponding to the 5'-flanking region of the human IL-8 gene (from +19 to 1,472) was obtained by polymerase chain reaction (PCR) with appropriate primers based on the published sequence (15). Using this fragment as the template, six different sizes of DNA fragments corresponding to the 5'-flanking region were generated. These DNA fragments were then subcloned into pBL-CAT3 vector (34) at the restriction sites BglII and HindIII to place the IL-8 promoter region upstream of the reporter gene, chloramphenicol acetyltransferase (CAT). The resulting clones were named pc 65 (from +19 to 65), pc 165 (+19 to 165), etc. For construction of the DNA fragment encoding three point mutations at the NF-B binding site of the promoter gene, a PCR-based mutagenesis technique was used as described elsewhere (34). Briefly, each external primer and one of two internal mutated primers were used first to generate two partial and overlapping DNA fragments, then these two external primers were used to PCR-amplify the entire region for subsequent cloning. DNA sequencing was used to verify the sequences of all these chimeric construct clones.

DNA Transfection and CAT Assay

BEAS-2B (S clone) cells were transfected with these IL-8 promoter-CAT chimeric constructs by a liposome (GIBCO BRL, Grand Island, NY)-mediated method as described elsewhere (34). Briefly, cultures at around 60% confluency were transfected with a premixed DNA and liposome preparation in serum- and antibiotic-free F12 nutrient. After overnight incubation, the medium was replaced by the normal hormone-supplemented medium. ATRA was added to some of these cultures. Retinoic acid nuclear receptor- (RAR-) expression clone, kindly provided by Dr. R. Evans' lab (The Salk Institute, La Jolla, CA), was included for cotransfection as needed. After an additional 48 h incubation, cell extracts from these cultures were prepared for CAT assay. The CAT ELISA kit from Boehringer Mannheim (Indianapolis, IN) was used to determine the CAT activity. The -galactosidase (-gal) activity of psv--gal (Promega)-transfected cells was used as an internal control for the normalization of transfection efficiency.

Nuclear Extract Preparation

Confluent cultures of primary human TBE cells, monkey TBE cells, or the BEAS-2B S clone cell line were used for nuclear extract preparation with the method described by Kunsch and Rosen (16).

Electrophoretic Mobility Shift Assay

For consensus NF-B binding sequence, the following double-stranded nucleotides (Santa Cruz Biotech, Inc., Santa Cruz, CA) were used: consensus NF-B (cNF-B) sequence, 5'-AGT TGA GGG GAC TTT CC C AGG C-3'; mutant form of CNF-B (CNF-B_m) sequence, 5'-AGT TGA GGCGAC TTT CCC A GGC-3'. For NF-B binding sequence of IL-8 gene, two corresponding DNA fragments were synthesized. One, designated as IL-8-NF-B, contained 83/67 DNA fragment (underlined), 5'-AGC GGA TCC CGT GGA ATT TCC TCT GAC GGA T-3'; the other contained IL-8 88/67 DNA fragment (underlined), 5'-AGC GGA TCC CAA ATC GTG GAA TTT CCT CTG ACG GAT-3'. For probes, cNF-B and IL-8-NF-B DNA were labeled with [-³²P]adenosine triphosphate by polynucleotide kinase, then separated through a Quick Spin column (Boehringer Mannheim) or purified from polyacrylamide gel.

Binding reactions were carried out in a 20-µl binding reaction mixture consisting of 25 mM N-2-hydroxyethylpiperazine-N'-ethane sulfonic acid (pH 7.9), 50 mM NaCl, 5 mM MgCl₂, 0.5 mM ethylenediaminetetraacetic acid (EDTA), 5% glycerol, 0.5 mM dithiothreitol, 0.5 mM phenylmethylsulfonyl fluoride, 1 mg of bovine serum albumin (BSA) per ml, 0.1 µg of poly d(I-C), and 10 µg of nuclear extract. The mixture was then incubated at room temperature for 20 min with 1 ng of [³²P]-labeled oligonucleotide probe. For the competition assay, unlabeled cNF-B, IL-8-NF-B, IL-8 88/67, and cNF-B_m DNA were each added at 3-, 30-, and 100-fold excess. In the supershift experiments, anti-p50 and anti-p65 antibodies (Santa Cruz Biotech) were added to the binding reaction mixtures, and the mixtures were incubated for 1 h at 4°C before adding the oligonucleotide probe. After 30 min, the samples were loaded onto 4% native polyacrylamide gels with 0.5× Tris borate-EDTA electrophoresis buffer. After electrophoresis, gels were dried and autoradiographed.

Western Blot Analysis

Protein of nuclear extracts prepared from airway epithelial cells grown with or without ATRA was electrophoretically separated on sodium dodecyl sulfate (SDS)-polyacrylamide gels and blotted onto nitrocellulose paper as described by Towbin and colleagues (35). Anti-p50 and anti-p65 antibodies (Santa Cruz Biotech) were used as the primary antibody and an ABC kit (Vector Laboratory, Burlingame, CA) was used for the second antibody and the color reaction.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

IL-8 Gene Expression Is Enhanced by ATRA

ATRA dose-dependent stimulation of IL-8 secretion is demonstrated in various airway epithelial cultures (Figure 1). For primary human TBE cells, IL-8 protein increased from 3.5 ng/10⁶ cells/h in the absence of ATRA to 11.5 ng/10⁶ cells/h after 1 µM ATRA treatment. For monkey TBE cells a similar trend was observed, although IL-8 secretion was lower (1 ng/10⁶ cells/h increased to 8 ng/10⁶ cells/h). Compared with primary cells, IL-8 secretion was one order of magnitude less in the immortalized TBE cell line; however, the effect of ATRA was the same. Overall, ATRA treatments stimulated the IL-8 secretion 3- to 8-fold.

View larger version (32K): [in this window] [in a new window]   Figure 1.   Dosage effect of ATRA on IL-8 protein production in various airway cultures. Primary cultures of human TBE (open bars), monkey TBE (shaded bars), and BEAS-2B (S clone) cells (solid bars) were cultured in media with ATRA concentrations from 0 to 1 µM. Culture media were collected for IL-8 quantitation by an ELISA kit. Cell number in each dish was determined and the level of IL-8 protein was then expressed as nanograms per million cells per hour of incubation. Results are the average of three experiments. *P < 0.05, as compared with the control.

We observed a similar dose-dependent increase of steady-state IL-8 mRNA levels after ATRA exposure (Figure 2). After normalization with 18S ribosomal probe hybridization, this increase ranged from 3- to 4-fold in primary human TBE cells (Figure 2A) and BEAS-2B cells (Figure 2C), up to 8-fold in primary monkey TBE culture (Figure 2B).

View larger version (47K): [in this window] [in a new window]   Figure 2.   Dosage effect of ATRA on IL-8 mRNA expression in various airway cultures. Total RNA was harvested from primary cultures of human TBE (A), monkey TBE (B), and BEAS-2B (C) cells, then analyzed for IL-8 mRNA by Northern blot hybridization. Hybridization with 18S rRNA was used to verify the RNA input in the Northern blot membrane. Results are representative of three independent experiments.

As shown in Figure 3, enhanced IL-8 message in primary human TBE cells (Figure 3A) and BEAS-2B cell line (Figure 3C) was seen 12 h after ATRA treatment. For monkey TBE cells, maximal stimulation was not observed until 48 h after ATRA treatment, whereas in human TBE cells a plateau was reached 24 h after treatment.

View larger version (45K): [in this window] [in a new window]   Figure 3.   Time-course effect of ATRA on IL-8 mRNA expression. Primary human TBE cells (A), monkey TBE cells (B), and BEAS-2B cells (C) were treated with 0.1 µM of ATRA, and RNA was harvested at various times after the treatment. Northern blotting analysis was carried out as described in MATERIALS AND METHODS. Results are representative of three independent experiments.

New Protein Synthesis Is Required for ATRA Stimulation

CHX, a protein synthesis inhibitor, was used to treat cultures before ATRA treatment. As shown in Figure 4, ATRA stimulated IL-8 mRNA in the absence of CHX. In the presence of CHX the ATRA-mediated enhancement of IL-8 no longer exists. When CHX was added to cells, there was a superinduction of IL-8 mRNA under both +ATRA and ATRA conditions. This superinduction may have obscured the ATRA-mediated enhancement. However, the degree of superinduction was similar for both +ATRA and ATRA conditions. These results suggest that new protein synthesis is required for ATRA to elevate IL-8 message.

View larger version (52K): [in this window] [in a new window]   Figure 4.   Effects of CHX on ATRA-induced IL-8 mRNA expression. BEAS-2B cells were treated with or without 0.1 µM of ATRA for 24 h in the presence or absence of CHX at 10 µg/ml. Total RNA was harvested for Northern blot analysis. The number at the bottom line indicates the densitometric value of the IL-8 mRNA normalized with the 18S rRNA value. Results are representative of three independent experiments.

ATRA Regulates the IL-8 Message at the Transcriptional Level

To distinguish between transcriptional and post-transcriptional regulation, actinomycin D (which inhibits new transcript synthesis) was used to determine the half-life of IL-8 mRNA of the cultured cells. As shown in Figure 5, IL-8 message has a half-life close to 50 min, and ATRA treatment had no obvious effect on the half-life of IL-8 mRNA.

View larger version (31K): [in this window] [in a new window]   Figure 5.   Effects of actinomycin D (Act. D) on IL-8 mRNA stability. Actinomycin D (10 µg/ml) was added to ATRA-treated (open circles) and untreated ( filled circles) monkey TBE cells. At 0, 30, 60, and 120 min after the addition of actinomycin D, total RNA was collected and analyzed by Northern blot hybridization. Densitometric value of the IL-8 mRNA in each time point was normalized with the 18S rRNA value and plotted for easier visualization. Results are representative of three independent experiments.

In contrast to the results of the actinomycin D experiment, nuclear run-on assay demonstrated that more IL-8 transcripts were transcribed in the nucleus in ATRA-treated human TBE cultures than in ATRA-free human TBE cultures (Figure 6). New transcript synthesis of the control genes SPRR1 and GAPDH was not affected by ATRA. Therefore, at least a part of ATRA-mediated IL-8 message enhancement occurs at the transcriptional level.

View larger version (56K): [in this window] [in a new window]   Figure 6.   Nuclear run-on analysis of ATRA-mediated IL-8 gene expression. BEAS-2B cells were cultured in media supplemented with (right panel) or without (left panel) ATRA (0.1 µM). The nuclear transcripts of SPRR1 and GAPDH were used as control. A similar procedure was repeated in primary cultures of human TBE cells.

Characterization of ATRA-Mediated Transcriptional Regulation of IL-8 Gene

Two approaches were used to elucidate transcriptional regulation of the IL-8 gene by ATRA. One was the promoter-reporter gene transfection approach; the other was to measure the DNA-protein-binding activity by electrophoretic mobility shift assay (EMSA). To elucidate the 5'-flanking region responsible for the promoter activity of the IL-8 gene, various chimeric constructs were prepared as described in MATERIALS AND METHODS. As shown in Figure 7, the relative CAT activity was enhanced in cultures treated with ATRA. Due to the presence of negative cis-element in addition to the positive element in the chimeric construct with 5'-flanking region longer than 200 bp, the CAT activity decreased when the size of IL-8 promoter in the construct was increased. However, the ATRA-mediated enhancement of CAT activity seemed to be independent of the length of the IL-8 promoter, as long as the two essential motifs, NF-IL-6 and NF-B binding sites, were included. Cotransfection with RAR- expression clone greatly increased ATRA-mediated enhancement of IL-8 promoter-dependent CAT activity, whereas RAR- cotransfection without ATRA made no significant difference. Further, a mutation at the NF-B binding site abrogated not only the ATRA and RAR- responses but also the basal promoter activity of IL-8. These results suggest that ATRA-response and RAR--responsive cis elements are present within the region of nucleotide -165 to +19 relative to the transcription start site of the IL-8 gene, and this region contains the NF-B binding site.

View larger version (33K): [in this window] [in a new window]   Figure 7.   Deletional and site-directed mutagenesis analyses of ATRA-responsive cis-acting DNA fragment in the 5'-flanking region of human IL-8 gene. (A) Schematic deletion of the 5'-flanking region of human IL-8 gene and the construction of IL-8 promoter-CAT reporter chimeric constructs. Starting from the 5' end of the IL-8 promoter region, various DNA fragments of nucleotides spanning from +19 to 1,481, 1,325, 481, 343, 165, or 65, relative to the transcription site, were cloned into the promoterless pBL-CAT3 vector as described in MATERIALS AND METHODS. The shaded boxes on the nucleotides represent the essential motifs corresponding to the binding sites of NF-IL-6 and NF-B. Site-directed mutation was carried out on the NF-B binding site, as shown in pc 165m. (B) CAT reporter gene activity in the chimeric construct- transfected cells. BEAS-2B S cells were transfected with 1 µg of IL-8 promoter-CAT DNA via a lipofectin-mediated method as described in MATERIALS AND METHODS. A total of 0.1 µg of RAR- expression clone was included for cotransfection when necessary. Cells were treated with or without 0.1 µM of ATRA for 24 h. Cell lysate was subjected to CAT ELISA and -gal assay. The promoter activity of each tested plasmid is indicated as CAT amount relative to the -gal activity. Values are means ± standard error for three separate experiments.

To investigate whether ATRA treatment affects NF-B binding activity, double-stranded DNA fragments corresponding to the cNF-B sequence and the NF-B binding sequence of IL-8 gene (IL-8-NF-B) were used to react the nuclear extracts prepared from cultures treated with or without ATRA. As shown in Figure 8A, when reacted with [³²P]-labeled cNF-B, nuclear extracts prepared from ATRA-untreated and -treated cells (Figure 8A lanes 1 and 2) caused two retarded bands: "a" and "b." Band "b" is much stronger in the ATRA-treated than in the untreated preparation. The specificity of the binding was supported by cold DNA competition when bands "a" and "b" disappeared due to the addition of unlabeled IL-8-NF-B, IL-8 88/67 DNA fragments (Figure 8A) or cNF-B DNA itself (Figure 8C). This competition was not observed with the addition of unlabeled cNF-B_m that had a mutation on the B binding site (Figure 8B). The specificity of the binding was further supported by the supershift phenomenon when the binding complex was treated with antibodies specific to the subunit proteins of NF-B transcriptional factors p50 and p65. Treatment of the binding complex with anti-p50 caused the supershift of the binding complexes (from "a" to "c" and "b" to "d"). Treatment with anti-p65 caused a disappearance of "b" and the formation of "d." Anti-p65 treatment also increased the binding activity at the "a" band. When a [³²P]-labeled IL-8-NF-B DNA fragment was used as probe (Figure 8C), the binding pattern was slightly different from Figures 8A and 8B. An additional nonspecific band was seen, which was probably due to the difference of a few nucleotides between the cNF-B oligomer and the IL-8-NF-B oligomer. However, ATRA-mediated enhancement of NF-B binding is still very clear, and is especially noticeable in band "b." Both bands "a" and "b" could be completely abolished by unlabeled cNF-B DNA, suggesting that the binding in these complexes is very specific. These results demonstrate that ATRA treatment enhances the retardation of both consensus NF-B and IL-8-specific NF-B sequences by EMSA.

View larger version (59K): [in this window] [in a new window]   Figure 8.   Characterization of the NF-B binding activity induced by ATRA. EMSA was used to assess the NF-B binding activity in human TBE cells treated with or without 0.1 µM ATRA for 24 h. (A) [32P]-labeled cNF-B DNA was used as probe. Lanes 1 and 2 are NF binding patterns of ATRA-treated and untreated cultures, respectively. Two DNA- protein complexes, "a" and "b," were found on the gel. Lanes 3-6 are antibody supershift experiments with 1 µg each of anti-p50 or anti-p65 antibody. Two supershift bands, "c" and "d," were observed. Lanes 7-10 are competition experiments with 0.175 (3×) and 1.75 (30×) pmol of unlabeled IL-8-NF-B DNA fragment. Lanes 11-14 are competition experiments with 0.175 (3×) and 1.75 (30×) pmol of unlabeled IL-8 88/67 DNA fragment. (B) [32P]-labeled cNF-B DNA was used as probe. Lanes 1 and 2, NF binding patterns of ATRA-treated and untreated cultures; lanes 3 and 4, same condition, except 5.8 pmol of unlabeled mutant NF-Bm DNA fragment was added for competition. (C) [32P]-labeled IL-8-NF-B DNA was used as probe. Lanes 1 and 2 represent the binding patterns of cells treated with or without ATRA. Lanes 3 and 4 are the same as lanes 1 and 2, except 1.75 pmol of unlabeled cNF-B as well as NF-Bm DNA were included for competition. Complexes "a" and "b" were competed by unlabeled cNF-B DNA. ATRA treatment enhanced the formation of "b" complex. The same results were observed in three separate experiments.

The Protein Amount of NF-B Subunits in the Nucleus Is Not Affected by ATRA

Western blot analysis was used to identify and quantify the amount of p50 and p65 subunits of NF-B protein in the nucleus. A total of 1 µg/ml of anti-p50 or anti-p65 antibodies was added to a nitrocellulose membrane that contained the nuclear extracts of ATRA-treated and untreated human TBE cells. The resulting data show that the intensity of the bands remained the same regardless of the addition or the absence of ATRA in the human TBE cells (Figure 9).

View larger version (61K): [in this window] [in a new window]   Figure 9.   Amounts of NF-B protein subunits p50 and p65 in the nucleus were determined by Western blot analysis. Equal amounts (20 µg) of nuclear extracts from ATRA- treated and untreated human TBE cells were separated on an SDS polyacrylamide gel. The material on the gel was then transferred to nitrocellulose membrane and allowed to react with 1 µg/ml anti-p50 or anti-p65 antibodies. The intensity of the resulting bands represents the protein amount of p50 and p65 in those nuclear extracts. The same results were observed in three separate experiments.

Effects of TRX on ATRA-Enhanced IL-8 Gene Expression

Because new protein synthesis is required for ATRA- dependent IL-8 gene expression as described earlier, we looked into the functional role of genes that are enhanced or induced early by ATRA in IL-8 gene expression. We have previously demonstrated that induced TRX gene expression was one of the early events induced by ATRA. As described in our previous publication (26), TRX message was elevated within 4 h after retinol treatment. In addition, it has been recognized that TRX plays an important role in the regulation of various transcriptional factors, including NF-B (19, 20), and that TRX protein can act as a potent costimulus of cytokine expression (21). Inasmuch as there is no ATRA-responsive element in the IL-8 165/+19 region, it is possible that a part of ATRA's effect on IL-8 gene expression may be due to an enhanced synthesis of TRX by ATRA. For these reasons we sought to determine whether TRX protein is able to activate the binding activity of NF-B protein in our TBE cells. As shown in Figure 10, TRX treatment was able to elevate the NF-B binding activity in a dose-responsive manner, especially the "b" complex. The supershift experiments with anti-p65 antibody further support the notion that the "b" DNA-protein complex is elevated by TRX protein. Transfection studies were used to determine whether TRX protein could further stimulate the IL-8 promoter activity. As shown in Figure 11, when IL-8 pc165-transfected cells were treated with TRX protein, the relative CAT activity was elevated 3-fold as compared with the control. In contrast, when the NF-B binding site of IL-8 promoter was mutated, TRX-stimulated CAT activity was not observed. These results further support the notion that a part of ATRA's effects on IL-8 gene expression may occur through enhanced TRX gene expression.

View larger version (91K): [in this window] [in a new window]   Figure 10.   Effects of recombinant human TRX protein on NF-B binding activity as assayed by EMSA. Nuclear extracts from ATRA-free primary monkey TBE cells were used in this study, and the preparation of nuclear extract was carried out as described in MATERIALS AND METHODS. The EMSA conditions were as described in Figure 8. Lanes 1, 2, 4, and 6 are the binding patterns of nuclear extracts reacted with [32P]-labeled cNF-B and increasing amounts of TRX. Lanes 3, 5, and 7 are the same as lanes 2, 4, and 6, except 2 µg of anti-p65 antibody was also added. Results are representative of three independent experiments.

View larger version (45K): [in this window] [in a new window]   Figure 11.   Effects of TRX protein on IL-8 promoter-reporter gene activity. BEAS-2B cells at 80% confluence were transfected with pc165-CAT (dark bars) or pc165m-CAT (light bars) by a lipofectin-mediated method as described in MATERIALS AND METHODS. A total of 2 µg/ml of human recombinant TRX or BSA was added 24 h before harvest as needed. PSV- -gal DNA was used as transfection efficiency control. Cell lysate was used for CAT and -gal assays as described in MATERIALS AND METHODS. Relative CAT activity in transfected cells was then normalized with the -gal activity from the same dish. Results are representative of three independent experiments.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

We report here that ATRA can enhance IL-8 gene expression 3- to 8-fold in cultures of various airway epithelial cells, including primary TBE cells derived from human and nonhuman primate tissues. The enhancement is time- and dose-dependent. Both the time course and CHX treatment experiments suggest the involvement of a delayed and, perhaps, an indirect mechanism requiring new protein synthesis for ATRA-mediated stimulation of the IL-8 gene. ATRA has no significant effect on the IL-8 mRNA stability, a conclusion based on the actinomycin D treatment experiment in which a half-life of around 50 min for IL-8 message was observed in cultures with or without ATRA. The nuclear run-on assay demonstrated that ATRA increases the transcription rate of the IL-8 gene. This transcriptional regulation by ATRA is further supported by the transient transfection study with the promoter-reporter construct. The cotransfection experiment with RAR- in the transient transfection study also demonstrated a further requirement for the interaction between ATRA and its receptor to maximize CAT reporter gene activity. This is consistent with an ATRA-dependent transcriptional regulation.

The second observation obtained from this study is the finding that ATRA-dependent transcriptional regulation is related to activation of DNA-protein interaction on the NF-B motif of the IL-8 promoter region. This conclusion is based on deletion and site-directed mutagenesis analyses with transient transfection studies, as well as results of EMSA studies. The deletion analysis of various 5'-flanking regions of the IL-8 gene has demonstrated that the stimulation of CAT activity by ATRA during transient transfection depends on the presence of the NF-B motif in the promoter-reporter construct. EMSA assays further demonstrated that NF-B binding activity in nuclear extract is increased by ATRA treatment. This binding can be completely replaced by the DNA fragment of the IL-8 promoter region from nucleotides 67 to 88, on the basis of competition assays. This region contains consensus DNA sequences corresponding to binding sites for NF-B transcriptional factors (12, 36), which has been implicated in regulating IL-8 gene expression in various cell systems. The supershift patterns with various antibodies specific for NF-B subunits indicate that this ATRA-enhanced NF-B binding involves the participation of p50 and p65 subunits (Figure 8). For p50, at least two DNA-protein complexes were observed, "a" and "b" bands. For the p65 subunit, only the "b" band was involved. More importantly, ATRA specifically enhances the "b" band and has little effect on the "a" complex, indicating that ATRA treatment is more specific for the activation of p65-related complex. Further, the amounts of p50 and p65 in the nuclei did not change after the addition of retinoic acid, as demonstrated in the Western blot assay. This indicates that the availability of both p50 and p65 in the nuclei is the same regardless of the presence or absence of ATRA.

Activation of NF-B binding activity by ATRA, however, cannot explain how ATRA is involved in the transcriptional activation of the IL-8 gene. Normally, retinoic acid exerts its effect through binding to nuclear receptors, namely, the retinoic acid receptors (RAR-, -, -) and the retinoid X receptors (RXR-, -, -) (37). RARs usually form heterodimers with RXRs (43), and these dimers are able to bind to specific DNA sequences, known as retinoic acid-responsive elements (RARE). They act as ligand-inducible transcription factors. The consensus sequence of RARE contains hexanucleotide half-sites arranged as inverted or direct repeats spaced by various numbers of nucleotides (44, 45). Even though the IL-8 gene does not contain the classical ATRA response elements, ATRA has been shown to upregulate IL-8 expression in several cell types, including fibroblasts (22), neuroblastoma cells (23), a human ovarian carcinoma cell line (24), and a human melanoma cell line (15). The exact mechanism by which ATRA regulates this expression is still not clear. Activation of cells by appropriate stimuli results in the dissociation of I-B from NF-B and the translocation of NF-B to the nucleus (46, 47). It is possible that ATRA may either enhance the synthesis of NF-B subunit proteins (25) or suppress the level of I-B, an inhibitor of NF-B nuclear translocation. However, the amounts of both p50 and p65 in the nuclei remained the same regardless of the presence or absence of retinoic acid, as demonstrated by Western blot assays (Figure 9). This implies that the addition of ATRA to cells does not increase the amount of NF-B protein in the nuclei through the removal of I-B (48, 49). Harant and associates (15) have demonstrated in tumor necrosis factor-- stimulated A3 cells an ATRA-induced disappearance of two additional DNA-protein bands, in EMSA assays. They postulated that these bands play an inhibitory role in the regulation of NF-B binding. However, we did not see two such additional bands in TBE cells.

Several groups (18, 50) have demonstrated the importance of redox regulation on the activation of NF-B (19). They have shown that NF-B reduction in the nucleus represents a necessary step for NF-B-DNA binding activity. Matthews and coworkers (20) further demonstrated that TRX increases the binding of NF-B to DNA by reduction of a disulfide bond involving cysteine 62. Schenk and colleagues (21) have demonstrated that recombinant human TRX protein is a costimulus of cytokine gene expression, including the IL-8 gene. We have previously demonstrated that retinol and its derivatives, retinoids, are potent stimulators for TRX gene expression in airway epithelium (26). To determine whether TRX is involved in ATRA-dependent IL-8 gene expression, we demonstrated that purified human TRX protein is able to elevate NF-B activity (Figure 10). This elevation is specific for the "b" complex, as shown by the anti-p65 gel shift experiment. In addition, we have demonstrated an enhanced IL-8 promoter activity on CAT gene expression by treating the cells with TRX protein. Mutations at the NF-B binding site of the IL-8 promoter abrogate TRX stimulation. These results further support the significance of the interactions between TRX and NF-B transcription factors in ATRA-mediated IL-8 gene expression.

TRX has been found in various cellular compartments, including the extracellular medium (21, 51). It has been suggested that TRX protein can go through a leaderless pathway to be translocated from the cytosol to the plasma membrane without the help of intracellular vesicles (53, 54). The nature of the transport is still not completely known. We think that a similar mechanism of transport from media to nucleus may exist when recombinant TRX protein is added to the culture medium. In a separate study, we observed the presence of FLAG antigen in the nucleus after treating cells exogenously with purified FLAG-TRX fusion protein (Harper and colleagues, manuscript in preparation). This result will support the notion that the recombinant human TRX protein is able to transport to the nucleus to exert its redox action on NF-B transcription factor. Taking these data together, we propose an alternative mechanism in which ATRA elevates IL-8 gene transcription possibly through enhanced TRX protein synthesis. Experiments are currently underway in our laboratory to determine whether antisense or dominant negative approaches to TRX gene expression can modulate ATRA-dependent IL-8 gene expression in airway epithelium.

In summary, we have demonstrated that ATRA has stimulatory effects on IL-8 gene expression in various airway epithelial cultures. The stimulation is related to activation of NF-B binding activity, which is essential for enhanced IL-8 gene expression. However, there is no evidence to support a direct interaction between ATRA-mediated transcriptional machinery and the IL-8 gene transcription. We offer instead an alternative mechanism whereby ATRA stimulates TRX protein synthesis, which then activates the IL-8 gene transcriptional complex.