Introduction

Top Abstract Introduction Materials and methods Results Discussion References

In asthma and bronchopulmonary dysplasia, increased airway smooth-muscle mass is thought to contribute to airways hyperresponsiveness (1). The potential role of airway smooth-muscle hyperplasia in the pathogenesis of human airways disease has led investigators to examine the early events involved in airway smooth-muscle cell mitogenesis, as well as those pathways that might attenuate smooth-muscle growth. Several laboratories have reported that substances that increase the intracellular concentration of cyclic adenosine monophosphate (cAMP) inhibit tracheal myocyte growth (2).

One potential mechanism for cAMP-induced growth inhibition is the attenuation of mitogen-activated signaling pathways involved in the transition from G₀ to G₁ of the cell cycle. We have shown that catalytic activation of extracellular signal-regulated kinases (ERKs) is required for platelet-derived growth factor (PDGF)-induced DNA synthesis in bovine tracheal myocytes (9), implying that cAMP reduces growth by inhibiting this signaling pathway. Consistent with this notion are studies of vascular smooth muscle and other cell types demonstrating that cAMP may inhibit ERK activation by attenuating the activity of an upstream signaling intermediate, Raf-1 (10). However, we have found that cAMP, though significantly reducing Raf-1 activity, does not inhibit PDGF-induced activation of ERKs in bovine tracheal myocytes (15), suggesting that ERK activation in these cells is Raf-1-independent, and that the effect of cAMP on growth is mediated downstream from this signaling pathway.

After activation by growth factors, ERKs may translocate to the cell nucleus, where they may phosphorylate and activate various transcription factors. These transcription factors, in turn, regulate the expression of genes required for DNA synthesis. Among the delayed early genes induced after mitogenic stimulation is cyclin D₁, a key regulator of G₁ progression in mammalian cells. We have demonstrated that cyclin D₁ is expressed following PDGF stimulation in bovine tracheal myocytes, and that microinjection of cells with a neutralizing antibody against cyclin D₁ inhibits S-phase traversal (16), suggesting that cyclin D₁ is required for DNA synthesis in these cells. Thus, cyclin D₁ represents another potential target for cAMP.

cAMP has been shown to reduce G₁ cyclin steady-state messenger RNA (mRNA) levels in yeast (17) and cyclin D₁ protein levels in human diploid fibroblasts (18, 19) and an astrocytic cell line (20). These data suggest that cAMP inhibits cell-cycle progression by suppressing activity of the cyclin D₁ promoter, which is known to have a cAMP response element (CRE) capable of binding the CRE-binding protein (CREB) (21, 22). Accumulation of intracellular cAMP may induce phosphorylation and activation of CREB by cAMP-dependent protein kinase A.

In the present study, we hypothesized that cAMP inhibits PDGF-induced cyclin D₁ expression in airway smooth muscle. We found that cAMP decreased both cyclin D₁ promoter activity and protein abundance in bovine tracheal myocytes. cAMP accumulation also induced CREB phosphorylation and increased the binding of CREB1 to the cyclin D₁ promoter CRE. Together, these data suggest a model by which cAMP attenuates airway smooth muscle DNA synthesis via the suppression of cyclin D₁ expression.

	Materials and Methods

Top Abstract Introduction Materials and methods Results Discussion References

Cell Culture

Bovine tracheal smooth-muscle cells were isolated as described previously (23). Myocytes of passage number 5 or less were studied; these cells traverse S-phase approximately 18 h after mitogenic stimulation and exhibit a doubling time of 24 to 36 h. Confluent cultures exhibited the typical "hill and valley" appearance under phase-contrast microscopy and showed specific immunostaining with anti--smooth muscle actin (Sigma Chemical Co., St. Louis, MO).

Assessment of DNA Synthesis

DNA synthesis was assessed by measurement of [³H]thymidine incorporation (23). Early passage cells were seeded into 96-well plates at 10,000 cells/well and allowed to grow for 2 to 3 d in Dulbecco's modified Eagle's medium (DMEM) with 10% fetal bovine serum (FBS) added. When confluence was attained, the cells were serum-starved in DMEM/0.5% bovine serum albumin (BSA) for 24 h. Growth arrest was confirmed by both [³H]thymidine incorporation and fluorescence-activated cell sorting (23). The medium was then removed, and the cells were incubated in DMEM/0.5% BSA to which PDGF (1 to 30 ng/ ml) was added. In other experiments, cells were coincubated with PDGF and forskolin (5 or 50 µM). After 4 h of incubation, 4 mCi/ml of [³H]thymidine (Amersham, Arlington Heights, IL) was added to each well, and the cells were allowed to incubate for another 20 h. At 24 h, the plates were washed twice with phosphate-buffered saline (PBS), and the cells were digested in 0.5% trypsin for 30 min. The plates were then frozen and thawed prior to harvesting the contents of individual wells onto fiberglass filters (Cambridge Technologies, Cambridge, MA). [³H]thymidine incorporation (counts per minute [cpm]) was then measured using a liquid scintillation counter (Beckman Instruments, Fullerton, CA).

Preparation of Cell Extracts for Cyclin D₁ Immunoblots

Confluent cell cultures in six-well plates were serum-starved by incubation in DMEM for 24 h. Cells were then treated with either PDGF (30 ng/ml), forskolin (50 µM), or both reagents for a duration of 16 h. After treatment, cells were washed with cold PBS (0.1 M phosphate, pH 7.5), and incubated with 0.3 to 0.5 ml of a cyclin D₁ lysis buffer containing 10 mM Tris (pH 7.5), 150 mM NaCl, 10 mM ethylenediaminetetraacetic acid (EDTA), 1% NP-40 Tergitol, 1% deoxycholate, 0.05% sodium dodecyl sulfate (SDS), 0.1 mM phenylmethlylsulfonyl fluoride (PMSF), 1 mM sodium vanadate, and 1 µl/ml leupeptin. Cell lysates were centrifuged (13,000 rpm for 10 min at 4°C), and the supernatant was transferred to a fresh microcentrifuge tube.

Western Blotting

Extracts were resolved on a 10% SDS-polyacrylamide gel and transferred to nitrocellulose by semidry transfer (Hoefer, San Francisco, CA). After incubation with polyclonal antibodies against cyclin D₁ (Upstate Biotechnology, Lake Placid, NY) or the phosphorylated form of CREB (phosphoCREB; Upstate Biotechnology), signals were amplified and visualized using antirabbit immunoglobulin G (Sigma) and enhanced chemiluminescence (Amersham). The Upstate Biotechnology anti-cyclin D₁ antibody recognizes the PRAD1 oncogene product, which is identical to cyclin D₁ (24). Cyclin D₁ signals were quantified by optical scanning.

Determination of Cyclin D₁ Promoter Transcriptional Activity

Cells were transiently transfected with a plasmid encoding the human cyclin D₁ promoter subcloned into a luciferase reporter. To construct a plasmid, an 1,882-base pair (bp) PvuII fragment of the human cyclin D₁ genomic clone was subcloned into the vector pA₃, to form the reporter 1745CD₁LUC (25). Cells were seeded into 60-mm plates at 50 to 80% confluence and incubated in 10% FBS/ DMEM overnight. After rinsing, cells were incubated with a liposome solution consisting of serum- and antibiotic-free medium, and Lipofectamine (12 µl/plate; Life Technologies, Gaithersburg, MD). Cells were then transiently cotransfected with plasmid DNA (3 µg 1745CD₁LUC and, to assess transfection efficiency, 0.6 µg pCMV -galactosidase). After 5 h, the liposome solution was replaced with 10% FBS/DMEM. At 24 h after transfection, cells were serum-starved in DMEM; 8 h later, cells were incubated with the appropriate stimulus for 16 h and harvested in lysis buffer provided with the Promega Luciferase Assay system (Promega, Madison, WI). After centrifugation (12,000 rpm for 2 min at 4°C), the supernatant was transferred to a fresh tube. A total of 20 µl of cell extract was added to 100 µl luciferase assay reagent, and luciferase activity was measured using a luminometer (Analytical Luminescence Laboratory, San Diego, CA). Cyclin D₁ promoter transcriptional activation was normalized for transfection efficiency by dividing luciferase activity by -galactosidase activity. -galactosidase activity was assessed by colorimetric assay using o-nitrophenyl--D-galactoside (Sigma) as a substrate (26).

Preparation of Nuclear Extracts

Nuclear extracts were prepared by the method of Dignam and colleagues (27) with some modifications. Cultures were trypsinized, rinsed twice with PBS, and incubated on ice for 10 min with 4 vol of buffer A, which consisted of 10 mM N-2-hydroxyethylpiperazine-N'-ethane sulfonic acid (Hepes) buffer (pH 7.9), 1.5 mM MgCl₂, 10 mM KCl, 0.5 mM PMSF, and 0.5 mM dithiothreitol (DTT). After centrifugation (1,000 rpm for 3 min at 4°C), cells were resuspended in one original packed cell volume of buffer A. After centrifugation, cells were suspended in 0.5 packed cell volume of extraction buffer C (20 mM Hepes, 25% glycerol, 1.5 mM MgCl₂, 420 mM KCl, 0.2 mM EDTA, 0.5 mM PMSF, and 0.5 mM DTT) and rocked on a platform for 30 min at 4°C. After centrifugation, supernatants were dialyzed for 1 h against three changes of 400 ml buffer D (20 mM Hepes, 20% glycerol, 100 mM KCl, 0.2 mM EDTA, 0.5 mM PMSF, and 0.5 mM DTT). After dialysis, nuclear extracts were clarified by centrifugation at 14,000 rpm for 20 min. Protease inhibitors (leupeptin, antipain, chymostatin, and pepstatin A, 5 µg/ml each) were added and aliquots stored at 80°C.

Electrophoretic Mobility Shift Assays

Electrophoretic mobility shift assays were performed using nuclear extracts (5 to 10 µg) and binding buffer containing 20 mM Tris HCl (pH 7.8), 100 mM KCl, 1.0 mM EDTA, 10% glycerol, 50 µg/ml poly (dI-dC), and 30,000 to 100,000 cpm of [-³²P]-labeled probe, and incubated on ice for 1 h. The sequence of the cyclin D₁ promoter CRE site oligodeoxyribonucleotide was 5'-AAC AAC AGT CGG AC-3'. The protein-DNA complexes were analyzed by electrophoresis through a 5% polyacrylamide gel. Supershifts were performed with two antibodies to CREB, SC-186 (Santa Cruz Biotechnology, Santa Cruz, CA) and HM93 (28). The gels were dried and exposed to radiographic film.

Statistical Analysis

The effects of PDGF and forskolin on cyclin D₁ expression were assessed by one-way analysis of variance (ANOVA) with repeated measures. Differences identified by ANOVA were pinpointed with a Tukey's multiple-range test.

	Results

Top Abstract Introduction Materials and methods Results Discussion References

Assessment of DNA Synthesis

DNA synthesis was assessed by measurement of [³H]thymidine incorporation. As expected, treatment with PDGF increased bovine tracheal myocyte DNA synthesis in a concentration-dependent manner (Figure 1). When cells were pretreated with an activator of adenylate cyclase, forskolin (5 µM for 15 min), there was an attenuation of PDGF-induced DNA synthesis. At the higher concentration of forskolin (50 µM), maximal PDGF-induced DNA synthesis was reduced by 75%.

View larger version (13K): [in this window] [in a new window]   Figure 1.   Forskolin attenuates PDGF-induced [3H]thymidine incorporation. Closed squares, PDGF alone; open squares, PDGF + 5 µM forskolin; open circles, PDGF + 50 µM forskolin. Data represent means ± SE of three experiments.

Forskolin Inhibits Cyclin D₁ Protein Abundance

Immunoblots utilizing a polyclonal antibody against cyclin D₁ were performed to determine the effect of forskolin on cyclin D₁ protein levels. As shown previously (16), PDGF treatment induced a 4-fold increase in cyclin D₁ protein abundance (Figures 2a and 2b). Forskolin attenuated cyclin D₁ protein levels in a concentration-dependent manner. Treatment with 50 µM forskolin significantly decreased cyclin D₁ protein levels (P < 0.001, ANOVA).

View larger version (21K): [in this window] [in a new window]   View larger version (11K): [in this window] [in a new window]   Figure 2.   Effect of cAMP on cyclin D1 protein levels. (a) Immunoblot using a polyclonal antibody against cyclin D1. Pretreatment either with forskolin (Fsk; 50 µM for 20 min) or albuterol (1 µM for 20 min) inhibited PDGF-induced cyclin D1 protein abundance. (b) Group mean data. Treatment with 50 µM forskolin significantly reduced cyclin D1 abundance (P < 0.001, ANOVA; data represent means ± SE of eight experiments.

To examine the effects of other substances that increase intracellular cAMP on cyclin D₁ expression, we measured PDGF-induced cyclin D₁ protein accumulation after pretreatment with albuterol (Research Biochemical International, Natick, MA), a -adrenergic receptor agonist known to increase adenylate cyclase activity. Like forskolin, albuterol (1 µM for 20 min) substantially attenuated growth factor-induced cyclin D₁ protein accumulation (Figure 2).

Forskolin Inhibits Cyclin D₁ Promoter Transcriptional Activation

To determine the effects of forskolin on cyclin D₁ promoter activity, cells were transiently transfected with a plasmid encoding the full-length cyclin D₁ promoter subcloned into a luciferase reporter. To normalize for transfection efficiency, cells were cotransfected with a plasmid encoding -galactosidase. As shown previously (16), PDGF treatment induced a 3- to 4-fold increase in cyclin D₁ promoter activity (Figure 3). Pretreatment with forskolin significantly decreased PDGF-induced cyclin D₁ promoter activity (P < 0.001, ANOVA).

View larger version (13K): [in this window] [in a new window]   Figure 3.   Cyclin D1 transcriptional activity in bovine tracheal myocytes transiently transfected with 1745CD1LUC. Cyclin D1 promoter transcriptional activation was normalized for transfection efficiency by cotransfection with -galactosidase. Treatment with 50 µM forskolin (Fsk) significantly inhibited PDGF-induced cyclin D1 transcriptional activation (P < 0.001, ANOVA). Data represent means ± SE of five experiments.

Forskolin Increases Phosphorylation of CREB

To test whether forskolin pretreatment induced phosphorylation of the nuclear transcription factor CREB, we performed immunoblots on whole-cell extracts using an antibody raised against a synthetic phosphopeptide corresponding to residues 123-136 of rat CREB. A small amount of phosphorylated CREB was present under basal conditions and after PDGF treatment (Figure 4). Forskolin pretreatment (2 h duration) increased the level of phosphorylated CREB, consistent with the notion that CREB participates in the suppression of cyclin D₁ gene expression by cAMP.

View larger version (8K): [in this window] [in a new window]   Figure 4.   Phosphorylation of CREB 2 h after forskolin (Fsk) treatment. Cell extracts were probed using an antibody raised against a synthetic phosphopeptide corresponding to residues 123-136 of rat CREB. Forskolin-induced phosphorylation of CREB was not attenuated by PD98059, a synthetic inhibitor of mitogen-activated protein kinase/ERK kinase. (Results typical of three experiments.)

PDGF has been demonstrated to stimulate protein kinase A in human arterial smooth-muscle cells through an ERK-dependent pathway (29). To determine whether forskolin-induced CREB phosphorylation was dependent on ERK activation, we preincubated cells with PD98059 (New England Biolabs, Beverly, MA), a synthetic inhibitor of mitogen-activated protein kinase/ERK kinase. We have demonstrated this signaling intermediate to be required and sufficient for PDGF-induced ERK activation in bovine tracheal myocytes (9). Preincubation with PD98059 (30 µM for 15 min) had no effect on CREB phosphorylation (Figure 4), demonstrating that the CREB phosphorylation pathway activated by forskolin treatment is distinct from the ERK pathway.

Forskolin Increases the Binding of Nuclear Protein to the Cyclin D₁ CRE

To assess directly the effect of cAMP on nuclear protein binding to the CRE, we performed electrophoretic mobility shift assays using an oligonucleotide containing the cyclin D₁ CRE. Treatment with forskolin increased the formation of two protein-DNA complexes, with peak complex formation occurring at 2 h (Figure 5a). Incubation with cold excess of CRE attenuated the response (Figure 5b). Furthermore, incubation with antibodies to CREB1 induced the supershift of DNA-protein complexes (Figure 5c), showing that CREB1 is responsible for at least part of the observed CRE-binding activity.

View larger version (68K): [in this window] [in a new window]   Figure 5.   Electrophoretic mobility shift assays. Nuclear extracts from bovine tracheal myocytes were incubated with a 32P end- labeled double-stranded oligonucleotide analogous to the CRE in the cyclin D1 promoter. Complexes were resolved on a 5% acrylamide gel. Gels were dried and exposed to film. (a) Forskolin (Fsk) treatment increased the abundance of two protein-DNA complexes. (b) Incubation of cold CRE attenuated protein-DNA binding. (c) Addition of rabbit antiserum to protein-DNA complexes from forskolin-treated cells caused no change in the complexes (lane 2), whereas incubation with two different specific antibodies against CREB1 (lane 3, SC-186; lane 4, HM93) caused supershift of a protein-DNA complex. (Results typical of three experiments.)

	Discussion

Top Abstract Introduction Materials and methods Results Discussion References

cAMP may exert opposite effects on cell proliferation in different cell types (30). Prolonged exposure to cAMP has been demonstrated to inhibit human (5), bovine (2), and rabbit (4) tracheal myocyte growth. Vasoactive intestinal peptide, prostaglandin E₂, and salbutamol have also been demonstrated to inhibit tracheal myocyte growth, probably via activation of adenylate cyclase (3, 5). However, the manner by which cAMP attenuates growth in these cells has not been elucidated. In the present study, we have confirmed that intracellular accumulation of cAMP inhibits PDGF-induced DNA synthesis in bovine tracheal myocytes. In addition, we found that pretreatment of these cells with forskolin decreases cyclin D₁ protein abundance and promoter activity. Forskolin treatment also induced the phosphorylation of CREB and increased the binding of nuclear proteins to the cyclin D₁ promoter CRE. Finally, incubation of protein-DNA complexes with an antibody against CREB1 induced supershift of at least one complex. Together, these data suggest that cAMP suppresses cyclin D₁ gene expression via phosphorylation and transactivation of CREB.

cAMP has been shown to reduce G₁ cyclin steady-state mRNA levels in yeast (17). In addition, cAMP has been demonstrated to reduce cyclin D₁ protein levels in human diploid fibroblasts (18, 19) and an astrocytic cell line (20). Both Won and colleagues (18) and Sewing and coworkers (19) found that increasing the amount of intracellular cAMP in human lung fibroblasts decreased cyclin D₁ protein expression in unstimulated as well as growth factor- treated cells. Similarly, we found that forskolin decreased cyclin D₁ protein abundance in unstimulated as well as PDGF-treated cells, suggesting that there is a basal level of cyclin D₁ expression that is also cAMP-sensitive. Gagelin and colleagues (20) found in a human astrocytoma cell line that cAMP inhibits not only the abundance of cyclin D₁ but also that of cyclin E, another important G₁ cyclin (31). Neither of the previous studies examined cyclin D₁ promoter activity, leaving open the possibility that the observed reductions in cyclin D₁ levels could have resulted from alterations in translation or protein degradation. Thus, our results extend previous observations by demonstrating that reductions in cyclin D₁ protein abundance are associated with similar reductions in promoter activity.

We found that cAMP-induced inhibition of DNA synthesis and cyclin D₁ transcriptional regulation was accompanied by CREB phosphorylation and increased DNA binding of CREB1 to the CRE in the human cyclin D₁ promoter, which is located 58 to 52 bp from the start site (21). These data strongly suggest that cAMP suppresses cyclin D₁ gene expression via phosphorylation and transactivation of CREB. cAMP has previously been demonstrated to attenuate cyclin A promoter activity through a CRE (32), suggesting that cAMP attenuates the expression of multiple cyclins via similar regulatory pathways. We have demonstrated that microinjection of cells with a neutralizing antibody against cyclin D₁ inhibits S-phase traversal (16), suggesting that cyclin D₁ is required for DNA synthesis in these cells. Because cAMP attenuates cyclin D₁ promoter activation via phosphorylation and activation of CREB1, these data suggest a mechanism by which cAMP may inhibit airway smooth-muscle growth.

One possible alternative mechanism of cAMP-induced growth inhibition is the increased expression of cyclin- dependent kinase inhibitors. Cyclin D₁, cyclin-dependent kinase-4 (cdk4), proliferating cell nuclear antigen, and a cyclin-dependent kinase inhibitor, p21^Cip1, are induced as part of the delayed early response to mitogenic stimulation (33). The cyclin D₁/cdk4 dimer titrates p27^Kip1, another inhibitor of cdk activity (36), and enters into complexes with proliferating cell nuclear antigen and p21^Cip1 (37, 38). Once enough cyclin D₁ and cdk4 are synthesized, steric inhibition by p27^Kip1 is exceeded, leading to phosphorylation and activation of the cyclin D₁/cdk4 holoenzyme by cdk-activating kinase (39, 40). Thus, G₁ progression relies on the stoichiometric ratios of cyclin D₁, cdk4, and p27^Kip1, as well as the activity of cdk-activating kinase. cAMP has been shown to increase expression of p27^Kip1 and inhibit activation of cdk-activating kinase in BAC1.2F5A macrophages (36), suggesting that cAMP may inhibit airway smooth muscle DNA synthesis by similar mechanisms. Our data demonstrating that cAMP attenuates cyclin D₁ transcriptional activation imply that this second messenger may have a tripartite inhibitory effect on cdk4 activity.

Another potential alternative mechanism for cAMP- induced growth inhibition is the modulation of upstream signaling pathways involved in the transition from G₀ to G₁ of the cell cycle. Albanese and colleagues (25) found that overexpression of either ERK or the transcription factor c-Jun increases cyclin D₁ promoter activity, suggesting that these molecules regulate cyclin D₁ expression. However, we (15) and others (41) have found that cAMP does not inhibit PDGF-induced activation of ERKs in airway smooth muscle, suggesting that the effect of cAMP on growth does not involve the ERK signaling pathway. On the other hand, c-Jun is known to be phosphorylated by another member of the mitogen-activated protein kinase superfamily, c-Jun amino-terminal kinase, or JNK (42), and catalytic activation of JNK has been demonstrated to be forskolin-sensitive (41, 43). Together, these data suggest that cAMP inhibits cyclin D₁ expression and DNA synthesis via attenuation of the JNK pathway. We have found that transient expression of a constitutively active SEK1, an upstream activator of JNK, does not increase cyclin D₁ promoter activity (M. Hershenson, unpublished observations). It is therefore unlikely that cAMP decreases cyclin D₁ expression in bovine tracheal myocytes via inhibition of the JNK pathway. cAMP has also been shown to reduce PDGF-induced phosphatidylinositol-3-kinase activity, S6 kinase activity, and DNA synthesis in bovine tracheal myocytes (44), consistent with the notion that cAMP may inhibit cyclin D₁ expression by modulation of this upstream signaling pathway. However, the potential roles of phosphatidylinositol-3-kinase and S6 kinase in the regulation of cyclin D₁ expression have not been studied.

In summary, we have demonstrated in bovine tracheal myocytes that cAMP attenuates cyclin D₁ promoter activation via phosphorylation and activation of CREB1. Because cyclin D₁ is required for DNA synthesis in these cells (16), these results are consistent with the notion that cAMP inhibits airway smooth-muscle growth by this pathway. However, further studies are needed to determine whether this is the primary mechanism of cAMP-induced growth inhibition, or whether additional pathways are also involved.