Interleukin (IL)-1β released after lung injury regulates the production of extracellular matrix components. We found that IL-1β treatment reduced the rate of elastin gene transcription by 74% in neonatal rat lung fibroblasts. Deletion analysis of the rat elastin promoter detected a cis-acting element located at −118 to −102 bp that strongly bound Sp1 and Sp3 but not nuclear factor (NF)-κB. This element mediated IL-1β-induced inhibition of the elastin promoter. IL-1β treatment did not affect the level of Sp1 but did induce translocation of the p65 subunit of NF-κB. Overexpression of p65 decreased elastin promoter activity and markedly reduced elastin mRNA. Immunoprecipitation studies indicated an interaction between the p65 subunit and Sp1 protein. Microarray analysis of mRNA isolated after overexpression of p65 or treatment with IL-1β revealed downregulation of α-smooth muscle actin and calponin mRNAs. Expression of these genes is associated with the myofibroblast phenotype. These results indicate that IL-1β activates the nuclear localization of NF-κB that subsequently interacts with Sp1 to downregulate elastin transcription and expression of the myofibroblast phenotype.
- nuclear factor-κB
elastin isa major structural protein in the lung. It is found most notably in alveolar walls and blood vessels. Tropoelastin, a soluble precursor, is synthesized in alveolar structures by interstitial fibroblasts and in vascular tissue by smooth muscle cells (4, 34, 35). Elastin synthesis in the pulmonary parenchyma of the rodent lung is highest during alveolarization. This process usually begins in the postnatal period and decreases with maturity (4, 25). Disruption of elastin deposition during development results in failure of alveolar formation (19). In the adult lung parenchyma, elastin mRNA is minimally expressed in interstitial structures but can be reactivated during the development of pulmonary emphysema or fibrosis (18, 20).
We previously reported (20) that elastin and collagen mRNA levels are upregulated after bleomycin treatment of rodent lungs. This expression was confined primarily to myofibroblasts. The myofibroblast phenotype is characterized by α-smooth muscle actin expression (28). Myofibroblasts appear to be responsible for matrix deposition during wound healing (31). In the adult lung, elastin mRNA levels can be modulated by effector substances released from macrophages or resident interstitial cells or from the extracellular matrix after proteolytic injury. Elastin mRNA can be upregulated by insulin-like growth factor, transforming growth factor (TGF)-β, and retinoic acid (13, 16, 22) and downregulated by basic fibroblast growth factor (7) and interleukin (IL)-1β (3).
IL-1β affects gene transcription via several different families oftrans-acting factors including AP-1 proteins, nuclear factor (NF)/IL-6-related factors [CCAAT box/enhancer binding protein (C/EBP)-α and C/EBP-β], and NF-κB protein complexes. NF-κB binds to DNA as a homo- or heterodimer in a number of cell types (33). The p50/p65 dimeric complex is the most transcriptionally active and abundant NF-κB isoform found in cells. The p65 component contains two or three independent transactivation domains. In quiescent cells, NF-κB is maintained in the inactive state in the cytosol bound to inhibitor IκBα/IκBβ. In the present studies, we found that IL-1β treatment decreased elastin transcription and induced large increases in the nuclear localization of the p65 subunit of NF-κB levels. Overexpression of the NF-κB component p65 resulted in the downregulation of elastin mRNA levels and the myofibroblast phenotype.
Lung fibroblasts were isolated from the lungs of 8-day-old Sprague-Dawley rats (Charles River Breeding Laboratory, Wilmington, MA) as previously described (3). The fibroblasts were maintained in minimal essential medium (MEM) supplemented with 5% fetal bovine serum (FBS), 0.37 g sodium pyruvate/100 ml, 100 U penicillin/ml, and 100 μg streptomycin/ml in a humidified 5% CO2-95% air incubator at 37°C. Confluent cultures were rendered quiescent by reducing the serum content of the medium to 0.4% for 24 h. The purity of the cultures was assessed with phase microscopy and Oil Red O staining.
Luciferase reporter constructs driven by the elastin promoter were prepared from a 1,028-bp rat elastin promoter (kindly provided by Dr. Charles D. Boyd, University of Hawaii, Honolulu, HI). Elastin promoter fragments −990 to −1, −535 to −1, −216 to −1, −133 to −1, −118 to −1, −115 to −1, −114 to −1, −102 to −1, and −66 to −1 (relative to the elastin translational start site) were inserted into PGL-2 basic luciferase reporter plasmid (Promega, Madison, WI), referred to as 990-Lux, 535-Lux, 216-Lux, 133-Lux, 118-Lux, 115-Lux, 114-Lux, 102-Lux, and 66-Lux, respectively. The 114A mutant involved changing the flanking sequences for the wild-type GC box from CTCCCACCCGCCCTCTC to CTCCATTCCGCCCACTC as previously described (14). The 114C mutant involved changing the wild-type sequence to CTCCGCCCCGCCCCCTC. All constructs were verified by sequencing. Transient transfection and luciferase assays were performed in triplicate as previously described (17).
RNA isolation and analysis.
Total cellular RNA prepared with the RNeasy kit (Qiagen, Valencia, CA) according to the manufacturer's protocol was used for Northern blotting or microarray analysis. We used a rat microarray chip (Affymetrix rat genome RG-U34A containing 7,000 genes). The analysis was performed by the Partners Gene Array Technology Center (Brigham and Woman's Hospital, Boston, MA) with the standard Affymetrix protocol.
Western blot analysis.
Confluent cultures of fibroblasts were treated with IL-1β (250 pg/ml). Cells were homogenized in buffer A [20 mM HEPES, pH 7.9, 1.5 mM MgCl2, 10 mM KCl, 2 mM EDTA, 0.5 mM dithiothreitol (DTT), 0.1% NP-40, 0.25 mM phenylmethylsulfonyl fluoride (PMSF), 1 mg/ml aprotinin, 1 mg leupeptin, 1 mg pepstatin, and 2 mM Na3VO4]. The homogenate was centrifuged (500 g, 5 min), and the pellets were suspended inbuffer C (buffer A with 25% glycerol). Nuclear proteins were extracted in 300 mM NaCl on ice for 30 min followed by centrifugation (17,000 g, 20 min, 4°C). Cytoplasmic and nuclear extracts were separated by SDS-PAGE, transferred to nitrocellulose membrane, and incubated with anti-IκBα (New England Biolabs, Beverly, MA) or anti-p65 antibodies (Santa Cruz Biotechnology, Santa Cruz, CA).
Nuclear run-on analysis.
Isolated nuclei (∼5 × 106 nuclei/sample) were resuspended in 200 μl of glycerol buffer containing 50 mM Tris-Cl, pH 8.3, 5 mM MgCl2, 0.1 mM EDTA, and 40% glycerol and stored in liquid nitrogen. The nuclear run-on transcription assay was performed according to the methods outlined by Greenberg and Ziff (9) and Groudine et al. (10) with modifications as previously described (17). cDNA inserts coding for elastin and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) were bound to a nitrocellulose filter with a slot blot apparatus.
Electrophoretic mobility shift assay.
Nuclear extracts from confluent cultures of fibroblasts that were untreated or treated with IL-1β (R&D Systems, Minneapolis, MN) were allowed to bind to radiolabeled double-stranded DNA containing consensus binding sequences of the targeted transcription factor. Each binding reaction (20 μl) included labeled DNA (1 × 105 cpm/0.5–1 ng) and 10 μg of nuclear extracts in solution with 1 μg of poly(dI-dC), 10 mM HEPES, pH 7.9, 20% glycerol, 0.1% NP-40, 70 mM NaCl, 1 mM EDTA, and 1 mM DTT. After incubation at room temperature for 30 min, DNA-protein complexes were resolved on a preelectrophoresed 4% nondenaturing polyacrylamide gel. For supershift experiments, antibodies (Santa Cruz Biotechnology) and nuclear extracts were incubated for 30 min at 4°C before binding reactions were initiated.
Nuclear extracts (500 μg) were incubated with 10 μg of anti-p65 antibodies (Santa Cruz Biotechnology) in lysis buffer (20 mM HEPES, pH 7.9, 1.5 mM MgCl2, 150 mM NaCl, 2 mM EDTA, 0.1% NP-40, 0.25 mM PMSF, 1 mg/ml aprotinin, leupeptin, and pepstatin, and 2 mM Na3VO4) at 4°C for 16 h with end-to-end rotation. The precipitates were collected by centrifugation and washed with lysis buffer at 4°C, boiled in 50 μl of electrophoresis sample buffer, separated in 8% SDS-PAGE, and transferred to nitrocellulose membrane. Western blot analysis was performed with anti-Sp1 and anti-p65 antibodies.
Oligonucleotide agarose conjugate pull-down assay.
Sp1 consensus or Sp1 mutated oligonucleotide-agarose conjugate (100 μg) was incubated with 500 μg of nuclear extracts (IL-1β treated or untreated) in binding buffer [10 mM Tris, pH 7.5, 50 mM NaCl, 1 mM DTT, 1 mM EDTA, 5% glycerol, 50 μg/ml poly(dI-dC), 1 mg/ml aprotinin, leupeptin, and pepstatin, and 2 mM Na3VO4] at room temperature for 2 h with gentle rotation. Samples were centrifuged (17,000 g) for 3 min and washed three times with binding buffer at 4°C, proteins were eluted with 300 μl of elution buffer (binding buffer supplemented with 150 mM NaCl), and the supernatants were concentrated with Centricon YM-3 (Millipore, Bedford, MA). Proteins were resolved by SDS-PAGE and subsequently identified by Western blot analysis.
We previously reported (3) that IL-1β treatment decreased elastin mRNA and protein levels in rat neonatal lung myofibroblasts. To assess the effect of IL-1β on the rate of transcription of the elastin gene, we performed nuclear run-on assays. We found that IL-1β treatment caused a marked decrease in the rate of elastin gene transcription (Fig.1 A). Densitometry analysis from four such experiments revealed that the rate of transcription of the elastin gene was reduced by 74%, whereas the rate of transcription of GAPDH did not change. No hybridization was detected to plasmids without inserts. These results parallel the effect of IL-1β on steady-state elastin mRNA levels (3).
To identify the IL-1β-responsive elements in the elastin promoter, fragments of the elastin promoter were synthesized by PCR and cloned into the luciferase reporter construct (Lux). We assessed the effect of IL-1β (250 pg/ml) on the transcriptional activity of these various promoter constructs after transfection (Fig. 1 B). High levels of luciferase activity were detected in quiescent fibroblasts that were transfected with constructs containing the elastin promoter sequence from −1 (relative to the translational start site) to −118 bp or larger. Treatment with IL-1β resulted in a 50–60% decrease of luciferase activity generated by these constructs. In contrast, the basal luciferase activities were 15- and 20-fold lower in the fibroblasts transfected with the 102-Lux reporter or the 66-Lux reporter, respectively, and decreased <10% after treatment with IL-1β. These data suggest that a cis-element located between −118 and −102 bp plays a major role in the inhibition of elastin gene expression by IL-1β in neonatal rat lung fibroblasts. IL-1β did not affect the transcriptional activity of a control luciferase vector utilizing an SV-40 promoter (data not shown).
To determine the mechanism by which IL-1β attenuated the basal activity of the elastin promoter, the binding of nuclear proteins from untreated and IL-1β-treated fibroblasts to the promoter was examined by electrophoretic mobility shift assay (EMSA) with the elastin promoter fragment −118 to −61 bp (Fig.2 A). Nuclear proteins from IL-1β-treated fibroblasts formed complexes comparable to the complexes observed with nuclear extracts from untreated fibroblasts. The specificity of Sp1 binding was demonstrated by successful competition with an oligonucleotide containing the Sp1 binding consensus sequence (Fig. 2 A) and by the addition of Sp1 and Sp3 antibodies (data not shown).
To further explore this unique region, we performed EMSA with additional elastin promoter fragments. We found that the −118 to −61 bp and the −115 to −61 fragments bound Sp1 and Sp3. In contrast, the binding of Sp1 and Sp3 to the −114 to −61 bp fragment was markedly attenuated (Fig. 2 B). The elastin promoter region −118 to −102 bp contains three overlapping elements that bind Sp1: CTCCC, CACCC, and CGCCC. Although the affinity of Sp1 varies, each of these elements can serve as partial Sp1 binding sites to control the expression of a wide variety of genes (24). However, the −114 to −61 fragment did not bind Sp1 or Sp3. To further examine the elastin promoter sequence −114 to −102 bp, the flanking nucleotides were mutated so that the GC box was flanked by either an AT-rich sequence (designated 114A) or a GC-rich sequence (designated 114C). The flanking regions were shown to contribute to Sp1 binding (14). Sp1 binding to the 114A fragment was slightly increased, whereas Sp1 binding to the 114C fragment was dramatically increased.
The contribution of Sp1 binding to promoter activity was assessed with luciferase assays. The luciferase reporter driven by wild-type (115-Lux and 114-Lux) or mutated (114A-Lux and 114C-Lux) promoter fragments was transfected into fibroblasts. The 115-Lux reporter generated high basal luciferase activity, whereas the 114-Lux reporter showed a five- to sixfold reduction in basal luciferase activity (Fig.3). Notably, the mutated 114C-Lux reporter displayed a large increase in basal luciferase activity and the 114A-Lux reporter showed a small increase in luciferase activity. The luciferase activities of the 114C-Lux, 114A-Lux, and 115-Lux reporters were significantly decreased by IL-1β treatment, whereas IL-1β treatment minimally decreased luciferase activity of the 114-Lux reporter.
We examined the activation of NF-κB by IL-1β by following the nuclear accumulation of the p65 subunit and the levels of IκBα and IκBβ in the cytosol. Western blot analysis indicated that 15 min after stimulation with IL-1β, p65 appeared in the nucleus with a decrease of p65 in the cytosol (Fig. 4). Concomitant with the movement of p65, the IκBα subunit disappeared from the cytosol. After 4 h, the level of p65 found in the nucleus remained unchanged but the level of p65 in the cytosol increased. After 24 h, the distribution of p65 and IκBα remained comparable to that observed at 4 h. The distribution and level of expression of IκBβ was not affected by treatment with IL-1β.
EMSA did not detect binding of p65 or p50 NF-κB subunits to the elastin promoter (data not shown). However, Sp1 interacts with certain transcription factors including the p65 subunit of NF-κB (2, 5,26, 27). To determine the effect of p65 on elastin gene expression, fibroblasts were transfected with both a p65 expression vector and the elastin promoter driving the luciferase reporter 118-Lux (Fig. 5 A). Cotransfection of p65 but not p50 markedly decreased the luciferase activity of the 118-Lux reporter (73%; P < 0.05). Thus the overexpression of the p65 subunit attenuated luciferase activity of the elastin promoter. In addition, the steady-state level of elastin mRNA was strongly reduced in fibroblasts transfected with the p65 expression vector but only slightly decreased in fibroblasts transfected with the p50 subunit (Fig. 5 B). Addition ofN-tosyl-l-phenylalanine chloromethyl ketone, an inhibitor of IκB degradation (36), blocked the IL-1β-induced decreases in elastin mRNA levels (Fig. 5 C).
To demonstrate an interaction between Sp1 and the p65 subunit of NF-κB, nuclear extracts were immunoprecipitated with an anti-p65 antibody. The immunoprecipitated proteins were Western blotted with anti-p65 antibodies and anti-Sp1 antibodies. Sp1 was observed in the anti-p65 immunoprecipitate of the nuclear extracts from untreated and IL-1β-treated fibroblasts. We found that p65 was increased in nuclear extracts derived from IL-1β-treated cells (Fig.6 A). We also examined the interaction of Sp1 with the p65 subunit by using agarose beads conjugated to a consensus Sp1 binding oligonucleotide or an oligonucleotide that does not bind Sp1. Nuclear extracts from untreated and IL-1β-treated fibroblasts were incubated with the conjugated beads, and Western blotting identified the bound proteins. Sp1 bound to the agarose beads incubated with the nuclear extracts from both the untreated and IL-1β-treated fibroblasts. In contrast, the p65 subunit was not detected in the nuclear extracts from untreated fibroblasts but was detected in the complex formed by the oligonucleotide-agarose and the nuclear extracts of IL-1β-treated fibroblasts (Fig.6 B). Sp1 and p65 could not be detected in the complexes formed by the non-Sp1-binding oligonucleotide-agarose incubated with nuclear extracts from either untreated or IL-1β-treated fibroblasts.
Treatment of fibroblasts with IL-1β is reported to inhibit expression of the myofibroblast phenotype as assessed by levels of α-smooth muscle actin mRNA (37). To determine the effect of p65 overexpression and IL-1β treatment on the myofibroblast phenotype, total RNA was isolated and analyzed by microchip gene array. Overexpression of p65 and treatment with IL-1β increased levels of several genes known to be transactivated by NF-κB including cyclooxygenase-2 and inducible nitric oxide synthase (Table1). In addition, we found that overexpression of p65 and treatment with IL-1β caused dramatic decreases in mRNA associated with the myofibroblast phenotype including α-smooth muscle actin and calponin (23, 31). Northern blot analysis confirmed that IL-1β treatment decreased α-smooth muscle actin and calponin mRNA levels (Fig.7). Results from three separate experiments indicated that α-smooth muscle actin mRNA levels were decreased by 85 ± 7% and calponin mRNA by 63 ± 9% (mean ± SE; n = 3).
Myofibroblasts synthesize extracellular matrix components during alveolar development and fibrogenic reactions (31). Neonatal rat lung fibroblasts express the myofibroblast phenotype in culture (33). We previously reported (3) that IL-1β reduced the steady-state levels for elastin mRNA. In the present study, we found that this effect was mediated primarily by decreases in the rate of elastin gene transcription. IL-1β treatment induced a 74% decrease in transcriptional activity as assessed by nuclear run-on assay. Similar to the human elastin gene, the proximal region of the rat elastin promoter is GC rich and contains an atypical TATA box sequence (ATAAA) (21). This sequence serves as a functional TATA box motif in other genes such as the β-globin gene, as it may in the elastin gene. Multiple transcription initiation sites are located within the first 50 bp upstream of the translational start site. The promoter also contains a CAAT sequence (located at −56 bp from the translational start site) and several Sp1 binding sites (8, 15).
We found that Sp1 binding to the region between −118 bp and −102 bp relative to the transcriptional start site largely determined the activity of the proximal rat elastin promoter. This region of the promoter also mediated inhibition by IL-1β. Promoter activity was decreased and IL-1β sensitivity was abolished by deletions that eliminated Sp1 binding to this region of the promoter. Notably, increases in promoter activity and IL-1β sensitivity were generated by mutations that increased the binding of Sp1 to this previously unresponsive promoter region. IL-1β treatment did not affect the expression of Sp1 and Sp3 or binding of Sp1 to the elastin promoter but upregulated nuclear localization of the p65 subunit of NF-κB. In addition, overexpression of the p65 subunit of NF-κB decreased elastin promoter activity and steady-state levels of elastin mRNA.
Interactions between Sp1 and p65 have been described previously (2, 5, 26, 27). NF-κB interacts with Sp1 to regulate the activity of specific promoters via several distinct mechanisms. These mechanisms include binding to adjacent or overlapping regulatory sites (2, 27) and physical interactions between the transcription factors (5, 11, 26). In vivo, the protein-protein interactions may disrupt DNA binding when the affinity of transcription factors for their binding site decreases because of associated chromatin structures (1). Alternatively, the assembled complex may interfere with interactions with other transcription factors or the transcriptional apparatus itself. Our studies indicate that p65 and Sp1 do not compete for binding because p65 did not disrupt Sp1 binding to the proximal elastin promoter. However, immunoprecipitation studies with nuclear extracts from IL-1β-treated cultures demonstrate an interaction between Sp1 and p65.
IL-1β activates NF-κB by inducing phosphorylation of IκB with subsequent release and nuclear translocation of p65. We find that IL-1β treatment induces a large and sustained increase in intranuclear p65 in lung fibroblasts. A similar finding was reported by others (12). High levels of p65 persisted in the nucleus despite increases in IκB in the cytoplasm in the hours after IL-1β stimulation. Recent studies suggest that intranuclear cytokine-activated acetylation of NF-κB subunits prevents binding of IκB and subsequent inactivation (6).
Gene array analysis indicated that overexpression of p65 and treatment with IL-1β predictably increased several species of mRNA including inducible nitric oxide synthase and cyclooxygenase-2 (30,32). Importantly, inspection of the array data revealed downregulation of many genes associated with the myofibroblast phenotype. Expression of α-smooth muscle actin mRNA, a classic marker for this phenotype (28, 31), was markedly decreased in both arrays. These results are consistent with the previous observation that IL-1β downregulates expression of α-smooth muscle actin in neonatal rat fibroblasts (37). It is possible that NF-κB functions through a common mechanism involving Sp1 to coordinately regulate these genes. It is noteworthy in this regard that NF-κB, particularly the p65 component, can bind to Sp1 proteins and interfere with the transcriptional rate of α1(I) collagen (29). Activation of other transcription factors by IL-1β likely accounts for differences in gene expression between the two arrays. Together, our results suggest that elastin mRNA expression is a feature of the myofibroblast phenotype and that IL-1β-induced nuclear localization of NF-κB dramatically decreases elastin transcription and expression of the myofibroblast phenotype.
This work was supported by National Heart, Lung, and Blood Institute Grants HL-66547 and HL-46902 and the Department of Veterans Affairs Research Enhancement Award Program research program.
Address for reprint requests and other correspondence: R. H. Goldstein, Pulmonary Center, R 304, Boston Univ. School of Medicine, 80 E. Concord St., Boston, MA 02118 (E-mail:).
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2002 the American Physiological Society