CsA-sensitive purine-box transcriptional regulator in bronchial epithelial cells contains NF45, NF90, and Ku

Yosuke Aoki, Guohua Zhao, Daoming Qiu, Lingfang Shi, Peter N. Kao

Abstract

Human bronchial epithelial (HBE) cells express interleukin (IL)-2 [Y. Aoki, D. Qiu, A. Uyei, and P. N. Kao.Am. J. Physiol. 272 (Lung Cell. Mol. Physiol. 16): L276–L286, 1997]. 16HBE-transformed cells contain constitutive and inducible nuclear DNA-binding activity for the purine-box/nuclear factor (NF) of activated T cell (NFAT) target DNA sequence in the human IL-2 enhancer. Transcriptional activation through the purine-box DNA sequence requires stimulation with phorbol 12-myristate 13-acetate + ionomycin, and this activation is inhibited by cyclosporin A. Immunohistochemical staining of 16HBE cells demonstrates nuclear expression of the purine-box DNA-binding proteins NF45 and NF90 and no expression of NFATp or NFATc. NF90 and NF45 associate with the DNA-dependent protein kinase catalytic subunit and the DNA-targeting subunits Ku80 and Ku70 (N. S. Ting, P. N. Kao, D. W. Chan, L. G. Lintott, and S. P. Lees-Miller.J. Biol. Chem. 273: 2136–2145, 1998). Antibodies to Ku potently inhibit the purine-box DNA-binding complex. The purine-box transcriptional regulator in 16HBE cells likely comprises NF45, NF90, Ku80, Ku70, and the DNA-dependent protein kinase catalytic subunit.

  • cyclosporin A
  • deoxyribonucleic acid-dependent protein kinase
  • nuclear factor of activated T cells
  • interleukin-2

bronchial epithelial cells exist at the interface between the host and the external environment and initiate and regulate the airway inflammatory response to noxious stimuli and pathogens. Inflammatory signaling molecules expressed by human bronchial epithelial (HBE) cells include the cytokines tumor necrosis factor, interleukin (IL)-1, IL-6, and IL-8 (reviewed in Ref. 30); leukotrienes (4); and hematopoietic growth factors including granulocyte-macrophage colony-stimulating factor (GM-CSF) (10) and IL-2 (3). The production of hematopoietic growth factors by bronchial epithelium likely promotes the local proliferation and functional maturation of immune effector cells recruited from the circulation into the microenvironment of the airways. This ability of bronchial epithelial cells to modulate the development of a local immune response is important for effective host defense against foreign pathogens. However, excessive growth factor expression by bronchial epithelial cells likely contributes to the cellular inflammation that is characteristic of noninfectious airway inflammatory conditions such as asthma (6, 15).

Aoki et al. (3) were the first to demonstrate that HBE cells express IL-2 mRNA and protein. IL-2 is expressed primarily by activated T cells and is regulated at the level of transcription. In activated T cells, transcriptional regulation at the IL-2 enhancer involves binding and activation of the specific transcription factors nuclear factor (NF) of activated T cells (NFAT), NF-κB, octamer-1, and activator protein-1 (AP-1) (reviewed in Ref.28). The target DNA sequence for NFAT has a purine-rich core motif, GAGGAAAA, which has been designated as the purine-box (reviewed in Ref.28) or, alternatively, the antigen-receptor response element (ARRE) (29). Multimerized purine-box/ARRE/NFAT target DNA elements confer stimulation-dependent and cyclosporin A (CsA)-sensitive transcriptional activation on a linked reporter gene transfected into T cells (17). Purine-box/ARRE DNA-binding activity and regulated transcriptional activation have been shown to exist in T cells, B cells (33), and endothelial cells (11). Purine-box transcriptional regulatory elements that are inhibited by CsA have also been identified in the promoters of the IL-3, IL-4, GM-CSF, and tumor necrosis factor-α genes (12, 27,28) .

Nuclear proteins that bind specifically to the purine-box in the IL-2 enhancer have been identified and cloned by diverse laboratories. Corthesy and Kao (13) and Kao et al. (19) purified a purine-box/ARRE DNA-binding complex from the nuclei of activated Jurkat T cells and identified and cloned proteins NF45 and NF90. NFATp and NFATc proteins were isolated from the cytoplasms of T cells and are proposed to translocate into the nucleus and combine with the AP-1 proteins Fos and Jun to activate transcription at the purine-box/NFAT target DNA sequence (27; reviewed in Ref. 28).

Ting et al. (31) recently showed that NF45 and NF90 associate tightly with DNA-dependent protein kinase (DNA-PK). DNA-PK consists of a large catalytic subunit, DNA-PKcs, and DNA-targeting subunits Ku80 and Ku70 (20). DNA-PK is involved in double-strand DNA break repair and V(D)J recombination (9, 20). The association of DNA-PK with an RNA polymerase II complex suggests a role for DNA-PK in the regulation of transcription (21). Sequence-specific transcriptional repression has been shown to be mediated by DNA-PKcs and specific binding of Ku to a purine-rich target DNA-sequence (18). In vitro reconstitution of NF90, NF45, Ku80, Ku70, and DNA-PKcs leads to the formation of a large protein complex that binds to DNA in an electrophoretic mobility shift assay (EMSA) (31).

Here we characterize, for the first time, a purine-box transcriptional regulator complex that exists constitutively in the nucleus of 16HBE cells. The purine-box regulator binds to its target DNA sequence in nonstimulated 16HBE cells and is transcriptionally silent. Stimulation of 16HBE cells increases purine-box DNA-binding activity, activates transcription, and induces IL-2 protein secretion. Purine-box DNA-binding activity, transcriptional activation, and IL-2 expression are inhibited by the T-cell immunosuppressants CsA and FK506. We performed immunohistochemistry on 16HBE cells and demonstrate that NF45 and NF90 are strongly expressed constitutively in the nucleus and that NFATp and NFATc are not expressed at all. NF90 and NF45 associate with DNA-PKcs and Ku, and we show that antibodies against Ku potently inhibit the 16HBE cell purine-box DNA-binding complex. We infer that the purine-box transcriptional regulator in 16HBE cells consists of NF90, NF45, Ku80, Ku70, and DNA-PKcs.

EXPERIMENTAL PROCEDURES

Cell culture and stimulation conditions. An SV40 large T antigen-transformed HBE cell line 16HBE14o− (16HBE cells), which retains differentiated morphology and function of normal human airway epithelia (14), was cultured in Eagle’s minimum essential medium (BioWhittaker, Walkersville, MD) supplemented with 10% heat-inactivated fetal bovine serum, 100 U/ml of penicillin, and 100 mg/ml of streptomycin (BioWhittaker) as previously described (3). An adult human T-cell leukemic cell line, Jurkat (clone E6-1), was obtained from American Type Culture Collection (Manassas, VA) and cultured in RPMI 1640 medium (Mediatech, Herndon, VA) supplemented with 10% fetal bovine serum, penicillin, and streptomycin (BioWhittaker). Monolayer epithelial cells grown to 90% confluency or Jurkat T cells (∼1 × 106 cells/ml) were stimulated for the indicated period of time in culture medium containing 20 ng/ml of phorbol 12-myristate 13-acetate (PMA; Calbiochem, La Jolla, CA) or PMA plus 2 μM ionomycin (Iono; Calbiochem). Modulation of 16HBE cell purine-box DNA-binding activity and transcriptional induction was investigated in response to CsA (Sandoz Research Institute, East Hanover, NJ), dibutyryl cAMP (Calbiochem), PGE2, theophylline, isoproterenol, and erythromycin (Sigma, St. Louis, MO). In these experiments, cells were pretreated with each drug for 16 h before stimulation with PMA or Iono for 2 h in the continued presence of each drug, followed by preparation of bronchial epithelial cell nuclear or whole cell extracts.

Nuclear extract preparation and EMSAs.The 16HBE and Jurkat T cells were stimulated for 2 h, pelleted in microfuge tubes, and then rinsed with cold phosphate-buffered saline (PBS). The nuclear extracts were prepared on ice with ice-cold reagents at 4°C as previously described (13). The cell pellets were resuspended with 300 μl of buffer A(10 mM HEPES, pH 7.6, 15 mM KCl, 2 mM MgCl2, and 0.1 mM EDTA, pH 8.0) containing the following supplements: 1 mM dithiothreitol, 0.2 (16HBE) or 0.05% (Jurkat) Nonidet P-40 (NP-40), 1 mM phenylmethylsulfonyl fluoride, 1 mM NaVO3, 10 mM NaF, 1 mM sodium pyrophosphate, 5 mM benzamidine, 0.1 mM sodium molybdate, and 1 mM β-glycerophosphate. After a 10-min incubation on ice, nuclei were spun down at 4,000 rpm for 5 min and were then resuspended in one volume of buffer C (25 mM HEPES, pH 7.6, 50 mM KCl, 0.1 mM EDTA, pH 8.0, and 10% glycerol) containing the supplements except NP-40. DNA-binding proteins were extracted from chromatin by incubating the nuclear suspension with a one-ninth volume of 3 M (NH4)2SO4for 30 min on ice, followed by centrifugation and pelleting of chromatin at 75,000 rpm for 15 min. The nuclear proteins extracted into the supernatant were precipitated with one volume of 3 M (NH4)SO4for 10 min on ice, followed by ultracentrifugation at 50,000 rpm for 10 min. The pellet of precipitated nuclear proteins was carefully resuspended with 100 μl of buffer Ccontaining the supplements and dialyzed against 100 ml ofbuffer C without the supplements for 3 h at 4°C. The concentration of nuclear proteins was measured with the Bradford dye-binding assay (Bio-Rad, Hercules, CA).

Purine-box DNA-binding activity in 16HBE or Jurkat nuclear extracts was assayed by EMSA. Briefly, 5–10 μg of nuclear proteins were incubated for 40 min at 4°C in 20 μl of binding buffer (25 mM HEPES, pH 7.6, 0.1 mM EDTA, 10% glycerol, 50 mM KCl, and 0.05 mM dithiothreitol) containing 1 μg of poly(dI-dC) and 2.5 pg of32P-labeled oligonucleotide probe (∼1 × 105 counts/min) for the purine-box/NFAT sequence in the human IL-2 enhancer (bases −259 to −284; upper strand, 5′-AAGAAAGGAGGAAAAACTGTTTCATA-3′; lower strand, 5′-TCTGTATGAAACAGTTTTTCCTCCTT-3′) (28). The probe was labeled with Klenow DNA polymerase (New England Biolabs, Beverly, MA) to fill in the four-base overhangs on each end of the annealed strands with [α-32P]dCTP (Amersham, Arlington Heights, IL) and nonradioactive dATP, dTTP, and dGTP. The competitor oligonucleotides utilized were the purine-box/NFAT sequence in the human GM-CSF gene (GM-550; 5′-agctGAAAGGAGGAAAGCAAGAGTCATA-3′) (12), the NF-κB sequence from the mouse Ig κ-light chain (Ig NF-κB; 5′-agctAAA GAGGGACTTTCC̲ -3′) (5), and the Sp1 sequence (5′-agctGA TCGGGGCGGGGC̲ GAGC-3′). The lowercase segment agct represents an appended sequence that overhangs the annealed strands and is available for fill-in radiolabeling; the underlined segments indicate the target DNA-binding sequences for NF-κB and Sp1, respectively. Protein-DNA complexes were resolved from free probe on 4% nondenaturing polyacrylamide gels in 0.5× Tris-borate-EDTA (pH 8.3) and visualized by fluorography.

Cell transfections and luciferase reporter gene assays. Two independent 16HBE cell lines carrying the purine-box luciferase reporter gene construct were generated according to the methods previously described (3). The purine-box reporter construct contains three copies of the distal purine-box in the human IL-2 enhancer (bases −255 to −285; NFAT target DNA sequence) in the context of the minimal IL-2 promoter, driving the firefly luciferase cDNA (16). The reporter construct also contains a neomycin resistance gene driven by the SV40 enhancer. Cells were stimulated for the indicated times, and whole cell extracts were prepared and assayed for luciferase activity as previously described (3). A 16HBE cell line was generated with the EF-1α promoter (32) driving expression of histidine-tagged NF45 protein (19). There were no significant differences in the expression of NF45 or in the purine-box EMSA between NF45-transfected cells versus nontransfected 16HBE cells (data not shown).

Immunohistochemistry and Western immunoblotting. Adherent 16HBE cells were harvested by trypsinization, washed, and resuspended in fresh medium, and one drop of the cell suspension was spotted onto a bovine serum albumin-coated diamond pen-scribed microscopy slide and incubated overnight in a moisture box at 37°C and 5% CO2. The cells were fixed over 15 min at 4°C with 4% paraformaldehyde in PBS, then washed twice in PBS. Fixed cells were exposed to blocking buffer (2% sucrose and 0.1 M glycine in PBS) for 30 min, dehydrated in 100% methanol, permeablized with 3% hydrogen peroxide in PBS for 5 min at room temperature, and then rinsed in PBS. Primary antibodies were added in PBS as purified rabbit IgGs (NF45 and NF90) at 4 μg/ml or goat IgGs [NFATp (M-20) and NFATc (K-18)] at 2 μg/ml according to the manufacturer’s instructions (Santa Cruz Biotechnology, Santa Cruz, CA) and incubated in the humidified box at room temperature for 30 min, and then the slides were rinsed in PBS. Detection was incubation with the appropriate biotinylated anti-rabbit or anti-goat secondary antibody (1:400 dilution; Jackson Immunoresearch, West Grove, PA) for 30 min and rinsing with PBS, followed by incubation with streptavidin-horseradish peroxidase (1:400; Jackson Immunoresearch) for 30 min and rinsing with PBS, and then staining with diaminobenzidine (2.5%) for 5 min. The cells were counterstained with hematoxylin for 30 s. Rabbit antisera to NF45 and NF90 were generated by BABCO (Berkeley, CA) with amino-terminal histidine-tagged NF45 (amino acids 1–194) and amino-terminal histidine-tagged NF90 (amino acids 1–151) as immunogens. The NF45 and NF90 IgGs were purified from the immune sera with protein A-agarose (Zymed, South San Francisco, CA). For immunoblotting, nuclear proteins were fractionated by SDS-PAGE (8% separating gel) and transferred electrophoretically to nitrocellulose. Primary incubations utilized antibodies 111 and N3H10 (monoclonal anti-Ku80 and -Ku70, respectively; provided by W. Reeves, University of North Carolina, Chapel Hill), rabbit polyclonal anti-DNA-PKcs (DPK1; provided by S. Lees-Miller, University of Calgary, Calgary, AL), or anti-NF45 and anti-NF90 (rabbit polyclonal sera). Detection was with horseradish peroxidase-conjugated secondary antibodies followed by enhanced chemiluminescence (Amersham).

Data and statistical analysis.Significance of the differences between the experimental conditions was determined by paired two-sample Student’st-test (Microsoft EXCEL).

RESULTS

Aoki et al. (3) previously showed that normal and transformed HBE cells express the T-cell growth factor IL-2. Here we characterize the DNA-binding properties, transcriptional regulation, sensitivity to immunomodulating drugs, and subunit composition of a purine-box transcriptional regulator involved in 16HBE cell expression of IL-2.

Characterization of CsA- and FK506-sensitive purine-box DNA-binding activity in the nucleus of 16HBE cells. We prepared nuclear extracts from 16HBE and Jurkat T cells and compared the recovery of purine-box DNA-binding activity. In EMSAs, we used a32P-labeled oligonucleotide probe that corresponds exactly to the purine-box/NFAT target DNA sequence from the human IL-2 enhancer. It is apparent that 16HBE cells contain a nuclear protein complex that binds specifically to the purine-box/NFAT oligonucleotide probe (Fig.1 A,lanes 13) and migrates with the identical electrophoretic mobility as the complex in the nucleus of Jurkat T cells, which is known as NFAT (Fig.1 B, lanes 2 and 3). In nonstimulated 16HBE cells (Fig. 1 A,lane 1) but not in nonstimulated Jurkat T cells (Fig. 1 B,lane 1), there is substantial purine-box DNA-binding activity. In 16HBE cells, there is a modest induction of the purine-box DNA-binding complex after stimulation with PMA and PMA+Iono (Fig. 1 A,lanes 2 and 3 vs. lane 1); in Jurkat T cells, purine-box/NFAT DNA-binding activity is substantially induced with PMA alone and even more potently induced after stimulation with PMA+Iono (Fig.1 B, lanes 2 and 3 vs.lane 1). In both 16HBE and Jurkat T cells, CsA destabilizes the purine-box/NFAT DNA-binding complex (Fig.1, A andB, lanes 46). These data demonstrate that 16HBE cells contain a nuclear purine-box DNA-binding activity that is induced with PMA+Iono and inhibited by CsA. The 16HBE cell purine-box DNA-binding complex has similar electrophoretic mobility, induction characteristics, and CsA sensitivity as the Jurkat T cell NFAT complex.

Fig. 1.

Electrophoretic mobility shift assays (EMSAs) of purine-box DNA-binding activity in nuclear extracts of human bronchial epithelial (16HBE) cells. A: purine-box DNA-binding activity in 16HBE cells and effects of stimulation (2 h) and cyclosporin A (CsA) treatment. PMA and P, phorbol 12-myristate 13-acetate (20 ng/ml); Iono and I, ionomycin (2 μM); NS, nonstimulated. B: purine-box/nuclear factor (NF) of activated T cell (NFAT) DNA-binding activity in Jurkat T cells and effects of stimulation and CsA treatment.C: FK506 inhibition of 16HBE cell PMA+Iono-stimulated purine-box DNA-binding activity.D: CsA inhibition of PMA-stimulated 16HBE purine-box DNA-binding activity in absence of calcium signaling.E: FK506 inhibition of constitutive 16HBE purine-box DNA-binding activity in absence of calcium signaling.

The 16HBE cell purine-box DNA-binding complex that is induced by PMA+Iono is also potently inhibited by the T-cell immunosuppressant FK506 (Fig. 1 C, lanes 46). In the presence of 1 μM FK506, the PMA+Iono-stimulated purine-box DNA-binding complex is inhibited to a level below that present in nonstimulated nuclear extracts (Fig.1 C, lane 6 vs. lane 1). The 16HBE cell purine-box DNA-binding complex is inhibited by 1 μM CsA (Fig. 1 D, lane 3) and 0.1 and 1 μM FK506 (Fig.1 E, lanes 2 and 3) in the absence of any calcium signaling.

Transcriptional activation through the purine-box DNA sequence in 16HBE cells. After we established that 16HBE cells contain a nuclear protein complex that binds specifically to the purine-box/NFAT sequence and migrates with similar electrophoretic mobility as the complex designated NFAT in T cells, we next investigated whether 16HBE cells show regulated transcriptional activation of a purine-box/NFAT luciferase reporter gene. We generated 16HBE cells that stably express the purine-box/NFAT luciferase reporter gene and evaluated the stimulation requirements for transcriptional activation. It is apparent that there is no transcriptional activation of the reporter construct in nonstimulated or PMA-stimulated 16HBE cells (Fig.2 A), and significant transcriptional activation of the purine-box/NFAT reporter in 16HBE cells is achieved after stimulation with PMA+Iono (Fig.2 A). These induction requirements in 16HBE cells mimic exactly the induction requirements for NFAT transcriptional activation in T cells (28). The results observed in two independent 16HBE cell lines that stably express the purine-box/NFAT luciferase reporter are similar to the results observed in transiently transfected 16HBE cells (Fig. 2 A).

Fig. 2.

Transcriptional activation of purine-box-luciferase reporter gene in 16HBE cells. A: induction requirements for activation of purine-box-luciferase reporter in 16HBE cells assayed 48 h after transient transfection or in 2 separate stably transfected transgenic 16HBE/purine-box-luciferase cell lines. 16HBE cells were either nonstimulated or stimulated for 6 h with PMA alone (20 ng/ml) or PMA+Iono (2 μM), and whole cell extracts were assayed for luciferase activity. Data are means ± SD determined from 3 (transient) or 4 (stable-1 and -2) independent experiments.B: inhibition by CsA of purine-box transcriptional activation in 16HBE cells.

We next examined the effects of CsA on transcriptional activation of the purine-box/NFAT luciferase reporter in 16HBE cells stimulated with PMA+Iono (Fig. 2 B). It is apparent that CsA causes a dose-dependent inhibition in purine-box transcriptional activation in 16HBE cells. The half-maximal effective dose for transcriptional inhibition by CsA in 16HBE cells is ∼4 ng/ml (Fig. 2 B), and this is the same half-maximal effective dose as for CsA inhibition of NFAT transcription in Jurkat T cells (data not shown).

Effects of CsA on 16HBE cell expression of IL-2. We next investigated whether 16HBE cell expression of IL-2 is regulated by stimulation and CsA similar to the activation of the purine-box transcriptional regulator (Fig.3). It is apparent that 16HBE cell expression of IL-2 protein requires stimulation with PMA+Iono (Fig. 3), and this is similar to the requirement for PMA+Iono stimulation for transcriptional activation of the purine-box/NFAT luciferase reporter (Fig. 2 A). Furthermore, 16HBE cell expression of IL-2 protein stimulated by PMA+Iono is completely inhibited by CsA (1,000 ng/ml; Fig. 3), and this is similar to the inhibition of purine-box/NFAT luciferase activity by CsA (Fig.2 B). The similarities in induction requirements and sensitivity to inhibition by CsA of the purine-box transcriptional regulator and in 16HBE cell expression of IL-2 protein imply that the CsA-sensitive purine-box transcriptional regulator is involved in 16HBE cell expression of IL-2.

Fig. 3.

Interleukin (IL)-2 secretion by 16HBE cells and inhibition by CsA. 16HBE cells (90% confluency in 6-well plates) were either nonstimulated or stimulated for 16 h with PMA (20 ng/ml), PMA+Iono (1 μM), or PMA+Iono in presence of CsA (1,000 ng/ml). Cell supernatants were harvested in a minimal volume (1 ml) and used without dilution for IL-2 ELISA. Data are means ± SD from 3 independent experiments. Induction of IL-2 secretion by PMA+Iono compared with nonstimulated 16HBE cells is significant (P < 0.05), and inhibition of P/I-stimulated IL-2 secretion by CsA is significant (P < 0.05).

Effects of drugs on 16HBE cell purine-box regulator induction and transcriptional activation. We next investigated the effects of other immunomodulating drugs on the induction of the 16HBE cell purine-box regulator DNA-binding complex and on transcriptional activation of the purine-box/NFAT luciferase reporter (Fig. 4). By EMSA, it is apparent that dibutyryl cAMP and isoproterenol showed minimal inhibition; PGE2, theophylline, and erythromycin showed moderate inhibition; and CsA showed complete inhibition of the purine-box DNA-binding complex (Fig. 4,A andB). In the purine-box/NFAT luciferase reporter assay (Fig. 4 C), these drugs showed inhibitory effects that corresponded well to their effects on the purine-box DNA-binding activity (compare Fig.4 C with Fig. 4,A andB). After CsA, the most potent inhibitors of purine-box/NFAT-luciferase activity are PGE2, theophylline, and erythromycin.

Fig. 4.

Drug modulation of purine-box DNA-binding activity and transcriptional activation in 16HBE cells. A: EMSA of 16HBE cell purine-box (Pu-box) DNA-binding activity in nonstimulated cells or in 16HBE cells that were pretreated for 12 h with cAMP (100 μM), PGE2 (0.5 μM), theophylline (Theo; 100 μM), isoproterenol (Isopro; 100 μM), erythromycin (EM; 100 μM), or CsA (1 μM). 16HBE cells were then stimulated with PMA (20 ng/ml)+Iono (2 μM) for 2 h in continued presence of immunomodulating drugs, and then nuclear extracts were prepared and assayed for Pu-box DNA-binding activity [7.5 μg of nuclear protein in presence of 1 μg of poly(dI-dC)].B: densitometer analysis of 16HBE Pu-box EMSA in A. Maximal signal elicited by PMA+Iono is set as 100%.C: Pu-box transcriptional activation and modulation by drugs in 16HBE cells. 16HBE cells that stably express Pu-box-luciferase reporter were preincubated for 12 h with immunomodulating drugs described in Aand then stimulated for 6 h with PMA+Iono, and whole cell extracts were assayed for luciferase activity. Data are means ± SD from 3 independent experiments. There was significant inhibition of PMA+Iono-stimulated Pu-box luciferase activity with PGE2(P < 0.05), Theo (P < 0.01), and CsA (P < 0.01).

Immunological analysis of candidate purine-box transcriptional regulator subunits. To identify subunits of the purine-box transcriptional regulator in 16HBE cells, we used antibodies to investigate the expression and subcellular localization of proteins potentially involved in specific binding to purine-box/NFAT target DNA sequence. We performed immunoperoxidase staining of 16HBE cells and demonstrated constitutive nuclear expression of the proteins NF45 and NF90 (Fig.5 A). In 16HBE cells undergoing mitosis, we observed that the pattern of immunoreactivity for NF45 and NF90 moves from the nucleus into the cytoplasm (Fig. 5 A). There are no signficant changes in the level or subcellular localization of NF45 or NF90 immunoreactivity after 16HBE cell stimulation or treatment with CsA (data not shown).

Fig. 5.

Expression of NF45, NF90 and DNA-dependent protein kinase but not of NFATp or NFATc proteins in 16HBE cells.A: immunoperoxidase staining demonstrating nuclear expression of NF45 and NF90 proteins compared with negative expression of NFATp and NFATc proteins. Nonstimulated 16HBE cells were harvested by trypsinization, fixed on glass slides, and exposed to primary IgGs raised against NF45, NF90, NFATc (K-18), and NFATp (M-20). Detection was achieved with the appropriate biotinylated secondary antibodies followed by streptavidin coupled to horseradish peroxidase. Magnification, ×40.B: Western immunoblot analysis demonstrating 16HBE cell nuclear expression of NF45, NF90, DNA-PKcs, and Ku80 and Ku70 proteins. Nuclear extract (40 μg) from nonstimulated 16HBE cells that stably express histidine-tagged NF45 was fractionated by SDS-PAGE (8% separating gel), transferred to nitrocellulose, and then probed sequentially for expression of Ku80 and Ku70, DNA-PKcs, and NF45 and NF90. Detection was with the appropriate secondary antibodies coupled to horseradish peroxidase and enhanced chemiluminescence. There was no Western immunoreactivity observed with antiserum to NFATc (K-18). MW, molecular-mass markers.

In contrast, we observed no immunoreactivity for any NFATp or NFATc proteins detectable in resting or stimulated 16HBE cells (Fig.5 A and data not shown). In particular, the NFATc (K-18) antiserum, described as broadly reactive against NFAT family members, showed nuclear immunoreactivity in Jurkat T cells (data not shown) but showed no immunoreactivity in 16HBE cells (Fig.5 A). These results imply that the purine-box regulator in 16HBE cells does not contain any NFATp or NFATc family polypeptides. The strong expression of NF45 and NF90 proteins in the nucleus of 16HBE cells is consistent with the hypothesis that NF45 and NF90 proteins, but not NFATp or NFATc proteins, contribute to 16HBE cell purine-box DNA-binding activity and transcriptional activation.

By Western immunoblotting, we show that nuclear extracts prepared from 16HBE cells demonstrate substantial expression of NF45, NF90, and DNA-PK subunits Ku80, Ku70 and DNA-PKcs (Fig.5 B). There was no detectable immunoreactivity with the broadly reactive NFATc antiserum (data not shown).

Oligonucleotide competitions and Ku antibody inhibition of specific 16HBE cell purine-box DNA-binding activity. To investigate the specific binding of the purine-box regulator complex in 16HBE cells, we performed competitions with unlabeled oligonucleotides related to, and distinct from, the IL-2 purine-box/NFAT target DNA sequence (Fig.6 A). There is substantial purine-box DNA-binding activity in the nucleus of nonstimulated 16HBE cells, which is further enhanced after stimulation with PMA+Iono (Fig. 6 A,lanes 1 and2). The purine-box regulator DNA-binding complex in PMA+Iono-stimulated 16HBE cells is strongly inhibited by the self purine-box/NFAT oligonucleotide (Fig. 6 A, lane 3) and is partially inhibited by the related GM-CSF promoter NFAT site oligonucleotide (Fig.6 A, lane 4). The mouse Ig NF-κB sequence oligonucleotide also shows partial inhibition of the specific purine-box regulator complex (Fig. 6 A,lane 5), and the unrelated Sp1 oligonucleotide shows essentially no competition for binding to the purine-box regulator complex (Fig. 6 A,lane 6). These oligonucleotide competition experiments establish that the 16HBE cell purine-box regulator complex binds with sequence specificity to the purine-box/IL-2 NFAT DNA target sequence.

Fig. 6.

EMSAs of Pu-box DNA-binding activity in nuclear extracts of 16HBE cells: oligonucleotide competitions and specific inhibition with antibodies against Ku. A: oligonucleotide competitions. Fifty nanograms of indicated competitor oligonucleotides were added to each DNA-binding reaction. GM-CSF, granulocyte-macrophage colony-stimulating factor.B: specific inhibition of Pu-box regulator complex with monoclonal antibodies against human Ku80 and Ku70 (αKu80 and αKu70, respectively), but not with polyclonal antisera against cytoplasmic phospholipase A2(αcPLA2). 16HBE cells that stably express histidine-tagged NF45 were stimulated for 6 h with PMA (20 ng/ml)+Iono (1 μM); then nuclear extracts were prepared, and Pu-box DNA-binding activity was evaluated by EMSA. For antibody studies, nuclear proteins were incubated with ascites (2 μl) containing monoclonal antibodies 162 (αKu162; anti-Ku70 and -Ku86), 111 (anti-Ku86), or N3H10 (anti-Ku70) or with 2 μg of rabbit polyclonal IgG against cPLA2 for 15 min before addition of32P-labeled oligonucleotide probes and a further 30-min incubation followed by native gel electrophoresis.

Ting et al. (31) previously showed that NF45 and NF90 proteins associate with the DNA-PKcs and the DNA-targeting subunits Ku70 and Ku80, and we hypothesized that the purine-box transcriptional regulator in 16HBE cells comprises NF90, NF45, Ku80, Ku70, and DNA-PKcs. To test this hypothesis, we investigated the effects of monoclonal antibodies against Ku80 and Ku70 (34) on the specific purine-box DNA-binding complex in 16HBE cell nuclear extracts (Fig.6 B). We found that antibody 162, which recognizes native Ku70 and Ku80, substantially inhibits the purine-box DNA-binding complex (Fig.6 B, lane 3 vs. lane 2). Antibody 111, which recognizes human Ku80, causes complete inhibition of the purine-box DNA-binding complex in stimulated 16HBE cells (Fig.6 B, lane 4 vs. lane 2). Antibody N3H10, which recognizes human Ku70, causes near-complete inhibition of the purine-box DNA-binding complex (Fig.6 B, lane 5 vs. lane 2). As a control, we show that the polyclonal rabbit IgG directed against cytoplasmic phospholipase A2 shows no significant inhibition of the purine-box DNA-binding complex (Fig.6 B, lane 6 vs. lane 2). The monoclonal antibodies against Ku produce no significant inhibition of the Ig NF-κkB complex in T cells (data not shown), establishing that the Ku antibody inhibition of the purine-box DNA-binding complex in 16HBE cells is specific.

DISCUSSION

We have characterized the DNA-binding properties, induction requirements for transcriptional activation, and subunit composition of a purine-box transcriptional regulator that controls the expression of IL-2 by 16HBE cells.

Previously, Corthesy and Kao (13) and Kao et al. (19) isolated a purine-box DNA-binding complex from nuclear extracts of Jurkat T cells and identified and cloned the polypeptides NF45 and NF90. During the purification, Ku70 and Ku80 were isolated as components of the intact T-cell purine-box transcriptional regulator complex (Kao, unpublished data). NF45 and NF90 are proteins of novel structure: NF45 has sequence similarity to helicases and NF90 contains two domains predicted to bind double-stranded RNA. NF45 and NF90 proteins are localized predominantly in the nucleus of all cell types examined (19). Recently, Ting et al. (31) identified NF45 and NF90 as copurifying polypeptides tightly associated with DNA-dependent protein kinase. In collaboration with Ting et al., we showed that NF45 and NF90 proteins associate with the DNA-PKcs and Ku70 and Ku80 to form a large multisubunit DNA-binding complex.

The purine-box DNA-binding complex that we have characterized in 16HBE cells migrates with identical electrophoretic mobility as the complex defined as NFAT in nuclear extracts prepared from stimulated Jurkat T cells. Furthermore, the 16HBE cell purine-box regulator complex and the Jurkat T cell purine-box/NFAT complex are each induced after stimulation with PMA+Iono and are each inhibited by CsA and FK506. In contrast to the situation in T cells, 16HBE cells express substantial nuclear purine-box DNA-binding activity in resting conditions. This purine-box DNA-binding activity in 16HBE cells is destabilized by CsA and FK506 in the absence of any calcium signaling. This result implies that CsA and FK506 can inhibit the purine-box transcriptional regulator through mechanisms that do not involve calcium signaling or calcineurin. Such calcium-independent effects of CsA were demonstrated in a mast cell line, in which CsA was shown to destablize IL-3 mRNA (26). These authors suggested that CsA inhibits the expression of an RNA-stabilizing protein, without which IL-3 mRNA is rapidly degraded.

We hypothesized that the purine-box transcriptional regulator in 16HBE cells switches from a transcriptional repressor into a transcriptional activator on cell stimulation with PMA+Iono. 16HBE cells that are nonstimulated or stimulated with PMA alone show substantial purine-box DNA-binding activity, yet purine-box luciferase transcriptional activation is silent. Only after stimulation with PMA+Iono is there a modest enhancement in DNA binding, which is associated with significant transcriptional activation. We propose that the transcriptionally silent form of the purine-box DNA-binding complex is a sequence-specific transcriptional repressor in 16HBE cells. In T cells, the purine-box/NFAT target DNA sequence has been shown to be involved in sequence-specific transcriptional repression as well as transcriptional activation (24, 25). We show in 16HBE cells that stimulation with PMA+Iono enhances the purine-box DNA-binding complex, is associated with transcriptional activation of the purine-box luciferase reporter construct, and, importantly, induces expression of IL-2 protein. Because stimulation that triggers transcriptional activation and IL-2 expression is associated with an increased intensity of the purine-box DNA-binding complex, we believe that there is not a replacement of the repressor with a different activator but, rather, a conversion of the DNA-bound purine-box regulator complex from a repressor into an activator.

CsA and FK506 act to destabilize the purine-box regulator complex in 16HBE cells. In contrast to T cells, we can observe this destabilization effect of CsA and FK506 on the purine-box complex in the absence of stimulation with PMA+Iono. If the purine-box regulator can mediate transcriptional repression as we propose, then we predict that CsA will inhibit repression or derepress transcriptional activation. Others (8) have described that there can be cross competitition between transcriptional regulators of the NFAT and NF-κB target sequences. Aoki and Kao (2) have previously shown that CsA can inhibit calcium-mediated repression of NF-κB activity in 16HBE cells. Our results established a mechanism through which CsA, by destabilizing a repressor, can exert proinflammatory effects in nonlymphoid cells.

We evaluated the effects of immunomodulating drugs on transcriptional activation of the purine-box regulator in 16HBE cells. Purine-box transcription that is stimulated with PMA+Iono is potently inhibited by CsA and is partially inhibited by PGE2 and theophylline. PGE2 has been shown to exert anti-inflammatory effects in T cells by inhibiting T-cell activation (1). The mechanisms of PGE2anti-inflammatory effects may partially involve elevation of intracellular cAMP (23); however, because the inhibition of 16HBE cell purine-box transactivation observed with PGE2 was much more profound than that with dibutryl cAMP, it is likely that PGE2 mechanisms involve other pathways beyond elevation of intracellular cAMP. Similarly, theophylline, which can lead to increases in intracellular cAMP, showed substantially more inhibition of purine-box transactivation than did dibutryl cAMP, suggesting again that the anti-inflammatory effects of theophylline (7) likely involve mechanisms beyond the elevation of intracellular cAMP.

Having established that 16HBE cells express a purine-box transcriptional regulator in the nucleus, we sought to identify candidate polypeptides of the purine-box regulator. We performed immunohistochemical experiments and observed strong constitutive expression of the polypeptides NF45 and NF90. NF45 and NF90 polypeptides were detected in the nucleus of nonmitotic 16HBE cells. During mitosis, NF45 and NF90 translocate to the cytoplasm. Similar results have been described for mitotic-phase phosphoprotein 4, which shows an ∼90% sequence identity to NF90 (22). The biological significance of this nuclear to cytoplasmic translocation during mitosis is unknown. There was no detectable reactivity for NFATp- or NFATc-related polypeptides with antisera that are both broadly and specifically reactive with NFAT family members.

NF45 and NF90 proteins associate tightly with the DNA-PKcs and stabilize the interaction of the DNA-PKcs with DNA-targeting subunits Ku80 and Ku70 (31). Ku70 and Ku80 proteins bind specifically to purine-rich DNA sequences (18) and interact with DNA-PKcs to mediate sequence-specific transcriptional repression (18). DNA-PKcs and Ku are found in association with the RNA polymerase II complex (21). We have demonstrated that monoclonal antibodies to Ku80 and Ku70 potently and specifically inhibit the purine-box DNA-binding complex in nuclear extracts of stimulated 16HBE cells.

In summary, we have established that human bronchial epithelial cells express a nuclear complex that binds with high affinity and specificity to a purine-box sequence that is identical to the NFAT target DNA sequence in the human IL-2 enhancer. This purine-box regulator complex activates transcription in bronchial epithelial cells in a CsA-inhibitable manner. CsA and FK506 inhibit the DNA-binding and transcriptional activation of this purine-box regulator complex and also inhibit 16HBE cell expression of IL-2. Monoclonal antibodies against Ku80 and Ku70 potently and specifically inhibit the purine-box DNA-binding complex in 16HBE cells. We propose that the purine-box transcriptional regulator in 16HBE cells consists of NF45, NF90, Ku80, Ku70, and the DNA-PKcs and that this complex switches from a transcriptional repressor into a transcriptional activator in response to cell stimulation.

Acknowledgments

This work was supported by grants from the California Affiliate of the American Lung Association and the Donald E. and Delia B. Baxter Foundation and by National Institute of Allergy and Infectious Diseases Grants K04-AI-01147 and R01-AI-39624 to P. N. Kao.

Footnotes

  • Address reprint requests to P. N. Kao.

  • Y. Aoki received salary support from Saga Medical School, Saga, Japan.

  • Present address of Y. Aoki: Dept. of Internal Medicine, Saga Medical School, 5-1-1 Nabeshima, Saga 849-8501, Japan.

REFERENCES

View Abstract