Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Recent studies have advanced the concept of G protein- coupled receptors (GPCRs) as mediators of cell growth by demonstrating their potential to activate mitogenic signaling pathways more commonly associated with receptor tyrosine kinases (RTKs). Most notably, GPCRs have been shown to use RTK-based scaffolding complexes to activate p42/p44 mitogen-activated protein kinase (MAPK) (1) and can directly activate phosphatidylinositol 3-kinase (PI 3-kinase) gamma through release of G subunits (5). These studies have relied heavily on the use of immortalized or transformed cell types as systems for heterologous expression of receptors, key signaling intermediates, and their disruptant mutant homologues to demonstrate the intermolecular associations that define the pathway. Although this strategy represents the most powerful and logical cell biology approach to defining novel signaling paradigms, the nature of transformed cells precludes the parallel analysis of GPCR effects on growth per se. Moreover, the validity of extrapolating findings from such models to more physiologically relevant cell types, or the in vivo condition, remains to be established.

The more integrative perspective also acknowledges that any mitogenic signal promoted by GPCRs in vivo likely exists against a backdrop of growth factor-induced RTK activation. Although several studies have demonstrated that GPCR activation can effect additive or synergistic increases in RTK-mediated growth (6), the signaling events that mediate such augmented growth remain uncharacterized.

In chronic asthma, hyperplasia and hypertrophy of airway smooth muscle (ASM) occur in the context of hyperresponsiveness to GPCR agonists such as histamine and acetylcholine. Thus, ASM represents a physiologically and clinically relevant cell type in which to examine the interactive effects of RTK and GPCR activation on cell growth. In this study, we detail the effects of concomitant RTK and GPCR activation on proliferation and mitogenic signaling in human ASM cultures. Numerous GPCR agonists are shown to potentiate epidermal growth factor (EGF)- mediated DNA synthesis and cell growth as well as transcription factor activation and cyclin D1 expression. Somewhat surprisingly, this potentiation was not mechanistically linked to EGF receptor (EGFR) transactivation or phosphorylation, and could occur in the absence of altered p42/p44 activation. Instead, GPCR-mediated growth potentiation was consistently associated with sustained activation of p70 S6 kinase for several hours after the initial early phase of activation.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Materials

p20-5XGal4-Luc and MLV.Gal4-Elk-1 were provided by Channing Der (University of North Carolina, Chapel Hill, NC). FosdE6AP-1-Luc was provided by Craig Hauser (The Burnham Institute, La Jolla, CA). [Methyl-³H]thymidine (1 µCi/ml) and enhanced chemiluminescence (ECL) reagents were purchased from Amersham (Arlington Heights, IL). [-³²P] adenosine triphosphate (ATP) was purchased from NEN Dupont (Boston, MA). Phosphorylation state-specific and phosphorylation state- independent antibodies against p42/p44 were purchased from New England Biolabs (Beverly, MA). Anti-EGFR, anti-cyclin D1 antibody, and antiphosphotyrosine antibodies were obtained from Upstate Biotechnology Inc. (Lake Placid, NY). Anti-p70/85^rsk antibody was purchased from Santa Cruz Biotechnology (Santa Cruz, CA). Horseradish peroxidase-conjugated secondary antibodies were purchased from Boehringer-Mannheim (Indianapolis, IN). Protein A-sepharose was purchased from Pharmacia Biotech AB (Uppsala, Sweden). Luciferase assay reagent was purchased from Promega (Madison, WI). Tyrphostin AG 1478 was purchased from Calbiochem (La Jolla, CA). All other reagents were purchased from Sigma (St. Louis, MO) or from previously identified sources (11).

Human ASM Cell Culture

Human ASM cultures were established as described by Panettieri and coworkers (13) from human tracheae obtained from lung transplant donors in accordance with procedures approved by the University of Pennsylvania Committee on Studies Involving Human Beings. Characterization of this cell line with regard to immunofluorescence of smooth muscle actin and agonist-induced changes in cytosolic calcium has been previously reported (13).

Third to fifth passage cells were plated at a density of 10⁴ cells/ cm² in either 24-well ([³H]thymidine assay), 6-well (MAPK, cyclin D1, and cell proliferation assays), or 10-cm plates (p70 S6 kinase and EGFR phosphorylation assays) in fetal bovine serum (FBS)- supplemented medium as described previously (13). Seven days later, cells were growth-arrested by refeeding cells with Ham's F12 medium supplemented with 5 µg/ml each of insulin and transferrin (IT medium) for 48 h.

Assay of [³H]thymidine Incorporation and Cell Proliferation

Confluent, growth-arrested cells were stimulated with various agents as indicated. Standard concentrations of agents were: EGF, 10 ng/ml; thrombin, 1 U/ml; histamine, 10 µM; and carbachol, 1 mM. After 16 h of stimulation, cells in 24-well plates were labeled with 3.0 µCi [methyl-³H]thymidine (1 µCi/ml) and incubated at 37°C for 24 h. Cells were then washed with phosphate-buffered saline (PBS), harvested with 0.05% trypsin-0.53 mM ethylenediaminetetraacetic acid (EDTA), and lysed with 20% trichloroacetic acid. The precipitate was aspirated onto filter paper and counted in scintillation vials.

For assessment of increases in agonist-stimulated cell number, cells in 6-well plates were maintained and stimulated as described previously for [³H]thymidine assays. After 40 h stimulation, cells were harvested with trypsin/EDTA and counted using a Coulter counter (Coulter Electronics, Hialeah, FL). Datapoints from individual [³H]thymidine and cell proliferation experiments represent the mean values derived from six wells.

Analysis of p42/p44 MAPK Phosphorylation and Cyclin D1 Expression

Human ASM cells were plated in 6-well plates as described previously and stimulated with various agents for 0 to 12 h. At the indicated time points, cells were washed once with cold PBS and lysed by addition of 1% sodium dodecyl sulfate (SDS) sample buffer. Lysates were boiled for 5 min, and 20 µl of cell lysate were electrophoresed on a standard 10% SDS polyacrylamide gel. After electrophoresis, proteins were transferred to nitrocellulose membranes. Blots were subsequently probed with antibodies that recognize cyclin D1 or the phosphorylated form of p42/p44, visualized using ECL, and quantitated by densitometry using autoradiographs that depicted bands within a linear range of exposure, as described previously (14). To control for uniformity of gel loading, blots were first stained with 0.2% Ponceau S or parallel blots were run and probed with antibodies that recognize both phosphorylated and nonphosphorylated forms of the respective MAPKs.

Analysis of EGFR Tyrosine Phosphorylation

Human ASM cells grown in 10-cm dishes were stimulated with 1 U/ ml thrombin, 10 µM histamine, 1 mM carbachol, or 10 ng/ml EGF (± 2 µM tyrphostin AG 1478) for 0 to 30 min at 37°C. Cells were washed twice with ice-cold PBS containing 0.2 mM sodium vanadate, then lysed in RIPA buffer (20 mM Tris-HCl, pH 7.4, 150 mM NaCl, 0.5% sodium deoxycholate, 1.0% NP-40, 0.1% SDS, 1 mM ethyleneglycol- bis-(-aminoethyl ether)-N,N'-tetraacetic acid (EGTA), 5 mM EDTA, 1 mM phenylmethylsulfonyl fluoride [PMSF], 10 µg/ml aprotinin, 10 µg/ml leupeptin, and 0.2 mM sodium vanadate). Lysates were centrifuged at 13,200 × g for 10 min and supernatants were collected. Equivalent amounts of supernatant protein were incubated with sheep antihuman EGFR polyclonal antibody overnight. Protein A-sepharose (80 µl) was then added to lysates for 1 h at 4°C. Immunoprecipitates were washed once with RIPA buffer adjusted to 0.5 M NaCl and 1 mM EDTA, then twice with standard RIPA buffer, and finally once with PBS containing 0.2 mM sodium vanadate. Immunoprecipitated proteins were subjected to 8% SDS-polyacrylamide gel electrophoresis and immunoblotting. Blots were incubated with 1 µg/ ml of antiphosphotyrosine antibody 4G10 in Tris-buffered saline (TBS)/0.5% Tween 20 overnight at 4°C. After three washes in TBS/ 0.5% Tween 20, nitrocellulose filters were exposed to matched primary antibody isotype horseradish peroxide-conjugated antimouse antibody at a 1:3,000 dilution. Filters were washed five times in TBS/0.5% Tween 20 and visualized using ECL.

Assay of p70 S6 Kinase Activity

Human ASM cells were grown to confluence, growth-arrested, washed, and stimulated with various agents for 0 to 12 h. Cells were then solubilized in a 50 mM Tris buffer (pH 8.0) containing 120 mM NaCl, 20 mM NaF, 5 mM EGTA, 1 mM EDTA, 10 mM sodium pyrophosphate, 10 mM p-nitrophenyl phosphate, 1 mM benzamidine, 0.1 mM PMSF, and 1% (vol/vol) NP-40 for 30 min at 4°C (lysis buffer) (15). Lysates were centrifuged at 13,200 × g for 10 min, and equal quantities of the supernatant protein were incubated for 2 h with 2 µg of a polyclonal anti-p70 S6 kinase antibody followed by addition of 50 µl of Protein A-sepharose for 2 h. The immunoprecipitates were washed twice in lysis buffer, twice in the same buffer without detergents, twice in 25 mM N-2-hydroxyethylpiperazine-N'-ethane sulfonic acid (pH 7.4) containing 20 mM -glycerophosphate, 20 mM MgCl₂, 3 mM EGTA, 2 mM dithiothreitol (DTT), and then incubated in the same buffer containing 100 µM substrate peptide, 10 µM protein kinase A inhibitor peptide, and [-³²P]ATP (10 µM, 2.5 µCi) in a volume of 30 µl for 10 min at 30°C. The reaction was terminated by the addition of 10 µl of 300 mM H₃PO₄. The phosphorylated peptide was separated from other products by ion exchange chromatography on P81 ion exchange paper using 75 mM H₃PO₄ and quantified by liquid scintillation counting.

Elk-1 and Activator Protein-1 Reporter Assays

Human ASM cells were transfected using a replication-defective adenovirus (Ad5-GPT) as described previously (16) with either (1) MLV.Gal4-Elk-1 and p20-5XGal4-Luc or (2) FosdE6AP-1-Luc. MLV.Gal4-Elk-1 contains the Gal4 DNA binding domain linked to the Elk-1 transcription activation domain (17). Luciferase reporter activity in p20-5XGal4-Luc is under control of five repeats of the Gal4 DNA binding domain. Cells were passaged into 12-well dishes at a density of 2.0 × 10⁴ cells/cm² in Ham's F12 medium supplemented with 10% FBS 12 h after transfection. Eight hours later, the medium was switched to IT medium and cells were maintained for 48 h. Cells were then stimulated for 15 h with various agonists, washed twice in Ca²⁺- and Mg²⁺-free PBS, then lysed in 25 mM Tris-HCl, pH 7.8, 2 mM DTT, 2 mM EGTA, 10% glycerol, and 1% Triton X-100, and harvested. Samples were briefly centrifuged to pellet Triton-insoluble material, and the supernatants were transferred to Eppendorf tubes and frozen at 70°C. Samples were subsequently assayed for luciferase activity using firefly luciferase substrate as per manufacturer's directions.

Data Analysis

Unless noted otherwise, data are presented as mean ± standard error of the mean (SEM). Statistically significant differences among groups were assessed by t test for paired or independent samples, with P values < 0.05 sufficient to reject the null hypothesis.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

The effects of EGF, thrombin, histamine, and carbachol on DNA synthesis and cell proliferation were examined in human ASM cultures. As demonstrated previously (17, 18), EGF as well as thrombin induced significant increases in both [³H]thymidine incorporation and cell number, whereas histamine and carbachol did not (Figure 1). However, when cells were stimulated with histamine, carbachol, or thrombin in combination with EGF, a synergistic effect on both DNA synthesis and cell proliferation was observed. Stimulation with both EGF and thrombin produced a 4.6 ± 0.9-fold higher [³H]thymidine incorporation than that of EGF stimulation alone, and a 1.8 ± 0.2-fold greater effect than the sum of both EGF- and thrombin-stimulated values (Figure 1A). Analysis of cell proliferation demonstrated a 47 ± 6% increase in cell number after stimulation by EGF + thrombin compared with 13 ± 2% and 12 ± 3% increases when stimulated by EGF and thrombin, respectively (Figure 1B). The synergistic effects of both histamine and carbachol on EGF-stimulated mitogenesis were even more striking. Although neither histamine nor carbachol alone induced significant DNA synthesis or cell proliferation, these agents caused a respective 2.6 ± 0.3- and 1.9 ± 0.2-fold enhancement of EGF-stimulated [³H]thymidine incorporation (Figure 1A). Similarly, cell number increased 29 ± 4% (EGF + histamine) and 25 ± 3% (EGF + carbachol) (Figure 1B). Pretreatment of cultures with the antagonists doxepin and atropine inhibited the potentiation of growth stimulated by histamine and carbachol, respectively (see Figure 1 legend). A similar potentiation to that induced by histamine was observed with costimulation with EGF and 5-hydroxytryptamine (5HT) as well as with EGF + U46619 (a stable thromboxane [TX] analog; data not shown). In addition, synergistic effects on human ASM proliferation were also observed with costimulation with platelet-derived growth factor (PDGF) and thrombin, histamine, or carbachol (data not shown). These results suggest that activation of receptor pathways linked to either Gi (m2 muscarinic acetylcholine, 5HT, TXA₂, protease-activated receptors) or Gq (H1 histamine, m3 muscarinic acetylcholine, 5HT, TXA₂, protease-activated receptors) activation, including those not capable of stimulating growth by themselves, can effectively complement RTK signaling to potentiate mitogenesis in human ASM cultures.

View larger version (16K): [in this window] [in a new window]   Figure 1.   Potentiation of EGF-mediated mitogenesis by GPCR activation. (A) Human ASM cultures were grown to confluence, growth-arrested, and subsequently stimulated with 10 ng/ml EGF, 1 U/ml thrombin (THR), 10 ng/ml EGF + 1 U/ml THR (EGF+THR), 10 ng/ml EGF + 10 µM histamine (EGF+HIST), or 10 ng/ml EGF + 1 mM carbachol (EGF+CARB) for 16 h, after which [3H]thymidine was added and incorporation was determined as described in MATERIALS AND METHODS. Data represent mean ± SEM values from seven (HIST, CARB) and thirteen (all other stimuli) observations. In two separate experiments, inclusion of 10 µM atropine with 1 mM CARB inhibited the increase in EGF-stimulated growth by two-thirds, whereas 1 µM doxepin completely eliminated growth potentiation by HIST (data not shown). (B) Cells grown in 6-well plates were growth arrested then stimulated as described previously. After 40 h stimulation, cells were harvested and counted. Data represent mean ± SEM values from three to five experiments. *P < 0.05, t test for paired samples.

To investigate the mechanisms mediating the synergy produced by EGFR and GPCR activation, we examined the activation and expression of key nuclear elements linked to growth and cell cycle regulation. Although human ASM cells are resistant to conventional methods of transfection, adenovirus-assisted transfection (19) provides sufficient transfection efficiency to allow reporter-based analyses of the nuclear transcription factors activator protein (AP)-1 and Elk-1 in human ASM despite causing cell cycle arrest through an unknown mechanism (17). Human ASM cultures transfected with luciferase reporter constructs responsive to intracellular activation of the transcription factor AP-1 exhibited increased luciferase activity in cells costimulated with EGF and thrombin, histamine, or carbachol when compared with cells stimulated with EGF alone (Figure 2A). A similar augmentation of Elk-1 reporter activity was observed in cells costimulated with EGF + thrombin or EGF + histamine (Figure 2B). A trend was also observed with costimulation with EGF and carbachol (P = 0.17), this lesser effect possibly related to the smaller effect of carbachol on EGF-stimulated mitogenesis (Figure 1).

View larger version (27K): [in this window] [in a new window]   Figure 2.   AP-1 and Elk-1 activation in human ASM. Cultures were transfected as previously described (16) with either FosdE6AP-1-Luc (A) or MLV.Gal4-Elk-1 and p20-5XGal4-Luc (B). Twelve hours after transfection, cells were passaged into 12-well dishes at a density of 2.0 × 104 cells/cm2 in Ham's F12 medium supplemented with 10% FBS. Eight hours later the medium was switched to serum-free medium and cells were maintained for 48 h. Cells were then stimulated for 15 h with indicated agonists. Triton-soluble extracts were prepared and assayed for luciferase activity as described in MATERIALS AND METHODS. Mean EGF-stimulated luciferase activity was 121 ± 42 × 103 SEM relative light units (RLU)/well/s for cells transfected with FosdE6AP-1-Luc and 14.8 ± 5.6 × 103 SEM RLU/well/s for cells transfected with MLV.Gal4-Elk-1/p20-5XGal4-Luc. These values represented a respective 1.9 ± 0.3- and 2.2 ± 0.3-fold increase over vehicle-stimulated (basal) luciferase activity. Values depicted reflect the mean ± SEM values normalized to EGF-stimulated luciferase activity (n = 7). *P < 0.05, t test for paired samples.

Additional studies examined cyclin D1, whose upregulated expression by mitogens, in coordination with titration of the cyclin-dependent kinase inhibitor p27^kip, enables mitogen-induced cells to enter the cell cycle (20) and is a sensitive indicator of promitogenic signaling in ASM (21). Cyclin D1 expression in human ASM cells was significantly increased by stimulation with EGF in a time-dependent manner (Figure 3A). After 6, 9, and 12 h costimulation with EGF + thrombin or EGF + histamine and 12 h stimulation with EGF + carbachol, cyclin D1 expression was significantly higher than that determined for stimulation with EGF alone (Figures 3A-3C). Collectively, these data suggest linkage between the activation of key nuclear elements and the potentiation of EGF-stimulated ASM proliferation, and that the putative mechanisms mediating the synergy either converge upon or lie upstream of nuclear signaling molecules.

View larger version (41K): [in this window] [in a new window]   Figure 3.   Induction of cyclin D1 expression in human ASM cultures. (A) Cultures grown in 6-well plates were stimulated with 10 ng/ml EGF ± 10 µM histamine (HIST) for 0 to 12 h. Reactions were stopped by addition of SDS sample buffer, and cyclin D1 levels were subsequently determined by immunoblotting as described in MATERIALS AND METHODS. (B) Cultures were stimulated with 10 ng/ml EGF ± 1 U/ml thrombin (THR), 10 µM HIST, or 1 mM carbachol (CARB) for 6 and 12 h. (C) Graphic representation of data (mean ± SEM) from five to eight experiments. Bands representative of cyclin D1 expression after stimulation with HIST or CARB (up to 12 h) were visualized only by overexposure of blots and were not significantly elevated over basal expression (data not shown). *P < 0.05, t test for paired samples.

We therefore analyzed transmembrane/cytosolic signaling junctures previously demonstrated to be important in EGF- and GPCR-mediated mitogenesis. Studies by Daub and colleagues (1, 2) have revealed that GPCR activation can induce EGFR intrinsic tyrosine kinase activity via transactivation; this transactivation appears to be the principal mechanism by which many GPCRs activate p42/p44 MAPK in non-neuronal cells (1, 2, 4). We therefore examined this potential phenomenon in human ASM by stimulating cells with EGF, thrombin, carbachol, or histamine, and assessing tyrosine phosphorylation of the EGFR. As shown in Figure 4A, EGF stimulation of human ASM cells resulted in a strong autophosphorylation of the EGFR that was inhibited by tyrphostin AG1478, a specific inhibitor of EGFR tyrosine kinase. However, none of the GPCR activators (thrombin, histamine, or carbaof the EGFR. Consistent with these results, AG1478 essentially eliminated EGF-stimulated p42/p44 activation and mitogenesis but failed to inhibit either thrombin stimulated p42/p44 activation (Figure 4C) or DNA synthesis (Figure 4D), and effectively reduced the [³H]thymidine incorporation induced by costimulation with thrombin and EGF to values obtained with stimulation by thrombin alone. Pretreatment with AG1478 also reduced [³H]thymidine incorporation in cells costimulated with EGF and either histamine or carbachol to near-basal levels (data not shown).

View larger version (27K): [in this window] [in a new window]   Figure 4.   Analysis of tyrosine phosphorylation of EGF receptor in human ASM cultures. (A) Cells were stimulated with 10 ng/ml EGF (± 2 µM AG1478), 1 U/ml thrombin (THR), 10 µM histamine (HIST), or 1 mM carbachol (CARB) for the indicated times. Lysates were prepared and EGFRs were immunoprecipitated with anti-EGFR antibody (4 µ/ml). Antigen-antibody complexes were subsequently precipitated by addition of protein A-sepharose, washed, and separated on an 8% SDS polyacrylamide gel. The gel was transferred to nitrocellulose and probed with an antiphosphotyrosine monoclonal antibody (1 µg/ml). The position of the EFGR is indicated by an arrow. Autoradiograph is representative of duplicate experiments. (B) EGFR tyrosine phosphorylation after stimulation with 10 ng/ml EGF ± vehicle, 1 U/ml THR, 10 µM HIST, 1 mM CARB, or 100 nM PMA. (C) Human ASM cultures were pretreated with either vehicle (0.02% dimethyl sulfoxide) or 2 µM AG1478 for 30 min then stimulated with 10 ng/ml EGF, 1 U/ml THR, or 10 ng/ml EGF + 1 U/ml THR for 10 min. Immunoblotting of phosphorylated p42/p44 was performed as described in MATERIALS AND METHODS. (D) Cultures were pretreated with ± 2 µM AG1478 for 30 min as described in B, simulated with indicated agents, and [3H]thymidine incorporation was determined as described in MATERIALS and METHODS. Data represent mean ± SEM values from four experiments. Open bars represent control, solid bars represent AG1478.

Costimulation of cells with EGF and thrombin, histamine, or carbachol resulted in a decrease in tyrosine phosphorylation of the EGFR, an effect that was mimicked by stimulation with phorbol myristate acetate (Figure 4B) and may thus reflect protein kinase C (PKC)-mediated desensitization of the EGFR (25). In light of the increase in ASM proliferation associated with these stimulation conditions, these data suggest the presence of spare EGFRs in human ASM cells whose functional loss fails to impact EGF-promoted mitogenesis. Regardless of the mechanism by which thrombin, histamine, and carbachol promote the observed reduction in tyrosine phosphorylation of the EGF-activated EGFR, the collective data demonstrate that in human ASM, GPCRs do not cause increased EGFR tyrosine phosphorylation or use EGFR transactivation to effect potentiation of growth.

Previous studies by our laboratory have identified the requirement of a strong and sustained activation of p42/ p44 MAPK for inducing human ASM proliferation in which the dose-dependent effect of EGFs on ASM proliferation correlates with the magnitude of p42/p44 MAPK activation (17). Although data previously discussed also suggest that EGFR transactivation cannot contribute to p42/p44 activation in human ASM, previous studies have identified EGFR-independent pathways linked to PKC (26), focal adhesions (4), Src (27, 28), or arrestin2 (28) by which GPCRs might activate p42/p44. To determine the potential role of p42/p44 in mediating the observed potentiation of human ASM mitogenesis, we examined the time-dependent activation of p42/p44 in human cultures. As shown in Figure 5, there was little or no increase in the magnitude of EGF-mediated p42/p44 activation by addition of histamine or carbachol, and when the integrated p42/p44 response was calculated over the 12 h of stimulation, values among groups varied less than 10%. Costimulation with thrombin had a variable response. Whereas the mean phospho-p42/p44 signals were slightly greater than that of EGF at the 1, 9, and 12 h time points, signals were increased at 3 h (> 40% increase in five of 11 paired observations) and 6 h (> 40% in seven of 15) such that mean values were statistically different (P < 0.05) at these time points. These differences resulted in the mean integrated p42/p44 response (0 to 12 h) for the EGF + thrombin group to be 30% greater than that of the EGF group.

View larger version (35K): [in this window] [in a new window]   Figure 5.   Effects of costimulation with EGF and GPCR agonists on p42/p44 activation in human ASM cultures. Cultures grown in 6-well plates were stimulated as described in Figure 1 and lysed in SDS sample buffer at the indicated time points. Levels of phospho-p42/p44 were assessed by immunoblotting as described in MATERIALS AND METHODS. (A) Blots from a representative experiment in which no alterations in EGF-stimulated p42/p44 phosphorylation were observed upon costimulation with thrombin (THR), histamine (HIST), or EGF. (B) Blots from a representative experiment in which costimulation with thrombin resulted in an increase in p42/p44 phosphorylation at the 3 and 6 h time points. (C) Phospho-p42/p44 signals were quantitated by densitometry and plotted as a function of duration of stimulation by indicated agonists. Data presented are mean values from 11 to 15 experiments. Error bars are omitted for graph clarity. *P < 0.05, t test for paired samples.

We have also previously demonstrated that PI 3-kinase- dependent activation of p70 S6 kinase is required for mitogen-stimulated human ASM proliferation (29). We therefore considered whether augmented activation of p70 S6 kinase could serve as a potential mediator of the synergistic effects of EGF and GPCR agonists. Growth-arrested human ASM cultures were stimulated with EGF, thrombin, or EGF + thrombin for 0 to 4 h, and kinase activity using immunoprecipitated p70 S6 kinase was subsequently assessed in vitro. Kinase activity stimulated by EGF or thrombin peaked at 30 min, after which activity declined and approached basal levels by 4 h (Figure 6A). In contrast, when cells were costimulated by EGF and thrombin, p70 S6 kinase activity was relatively more sustained than that elicited by EGF alone, being significantly (P < 0.05) higher at the 2 and 4 h time points. A similar but less pronounced increase in p70 S6 kinase activity was observed over this time frame in cells stimulated with EGF + histamine and EGF + carbachol. However, a more extended analysis (up to 12 h) clearly demonstrated an increase in p70 S6 kinase activity throughout the late phase in EGF + histamine-stimulated and, to a lesser extent, EGF + carbachol-stimulated cells. Statistically significant increases in the EGF + histamine group (compared with EGF alone) were observed at the 6, 9, and 12 h time points (Figures 6B and 6D), and in the EGF + carbachol group at 6 and 9 h (Figures 6C and 6D), with a trend toward significance (P = 0.2) at the 12 h time point. Thus, an increase in late-phase p70 S6 kinase activity, but not p42/p44 MAPK activation, appears to be a consistent finding associated with GPCR-mediated potentiation of human ASM growth.

View larger version (43K): [in this window] [in a new window]   Figure 6.   Effects of costimulation with EGF and GPCR agonists on p70 S6 kinase activity in human ASM cultures. Growth-arrested cultures grown in 10-cm dishes were stimulated with 10 ng/ml EGF ± 1 U/ml thrombin (THR; A), 10 µM histamine (HIST; B), or 1 mM carbachol (CARB; C), and lysed at the indicated time points. p70 S6 kinase was immunoprecipitated from lysates and used in an in vitro kinase assay as described in MATERIALS AND METHODS. Data from additional experiments analyzing the effect of concomitant HIST or CARB addition on 6, 9, and 12 h time points are presented in D. Data represent mean ± standard deviation from two (B and C), three (A), and four to six (D) experiments. *P < 0.05, t test for paired samples.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

The present study offers two important findings. One is the identification of potentiation of EGF-stimulated human ASM proliferation by GPCR agonists coupled to Gi or Gq. Importantly, these agonists (thrombin, histamine, carbachol, serotonin, and thromboxane) are associated with ASM contraction or airway inflammation, thus suggesting that the phenomena that contribute to acute bronchoconstriction in asthma may also play a role in developing the more chronic features of the disease. Previous studies have also observed similar potentiation of EGF-stimulated mitogenesis in human ASM by endothelin-1 (11), lysophosphatidic acid (10), and leukotriene D₄ (12). However, the mechanisms underlying these effects were not explored.

The second major finding is the association of a consistent increase in late-phase p70 S6 kinase activity, but not increased p42/p44 activity, with the observed synergistic effect of EGFR and GPCR activation on growth. This finding is somewhat surprising given that the proposed central role of p42/p44 MAPK in proliferative pathways (30), combined with the numerous recent studies elucidating the mechanisms by which GPCRs activate p42/p44 (15, 31), has served to implicate p42/p44 as an important pathway for GPCR-mediated growth. However, it should be emphasized that the mechanistic studies to date analyzing GPCR activation of p42/p44 do not include parallel analysis of growth, because cell lines amenable to heterologous expression are typically transformed or immortalized, thus possessing an overriding regulatory defect that renders most superimposed promitogenic signals difficult to perceive or interpret. Thus, the manner in which GPCRs mediate growth under various physiologic conditions in relevant cell types remains undetermined.

Having established the effects of concomitant RTK and GPCR activation on human ASM growth, we sought to characterize the cellular and molecular signaling events associated with this phenomenon. Consistent with their respective enhancement of EGF-stimulated mitogenesis, thrombin, histamine, and carbachol each augmented EGF-stimulated Elk-1 and AP-1 activation and cyclin D1 expression. Thus, putative mechanisms mediating the interactive effects of GPCRs and RTKs likely exist upstream of these nuclear signals. Although numerous pathways have the capacity to regulate transcription factor activation and cyclin expression through diverse mechanisms, our subsequent analysis focused on three pathways previously demonstrated to be mechanistically important in mitogenic signaling by GPCRs.

One potential mechanism involves enhanced activation of the EGFR by GPCR-mediated transactivation. Daub and coworkers (1, 2) have shown that in Rat-1 fibroblasts and COS-7 cells, GPCR activation can induce the transphosphorylation of the EGFR. In human ASM, we found that EGF, but not histamine, carbachol, or thrombin, promoted EGFR tyrosine phosphorylation. EGFR phosphorylation by EGF was abrogated by AG1478, a tyrphostin that specifically inhibits EGFR autophosphorylation. Moreover, AG1478 eliminated growth in response to EGF. In the presence of AG1478, the level of [³H]thymidine incorporation in human ASM costimulated with EGF and thrombin was reduced to levels observed in human ASM treated with thrombin alone, and AG1478 had no effect on thrombin-mediated mitogenesis. Therefore, in contrast to that observed in Rat-1 fibroblasts and COS-7 cells, GPCRs do not appear to activate EGFR tyrosine kinase activity in human ASM.

The ability of different GPCRs to transduce signals to the p42/p44 MAPK pathway has been established in multiple cell types (15, 30). In many cases, p42/p44 MAPK activation by GPCRs is mediated by subunits from heterotrimeric G-proteins that contribute to activation of various receptor or nonreceptor protein tyrosine kinases such as EGFR, Src, or Pyk2 (15, 30). In addition, other pathways, including those linked to PI 3-kinase, p70 S6 kinase, or myc, that may or may not interact with MAPK pathways can be activated by GPCRs and promote mitogenesis (30, 32, 33). However, the mechanisms by which GPCRs complement EGFR signaling to enhance RTK-mediated mitogenesis are unknown.

We have previously demonstrated a relationship between sustained p42/p44 activation and human ASM growth (17). Despite the observed inability of GPCRs to promote EGFR transactivation, we hypothesized that the potentiation of growth by GPCRs might be caused by increased activation of p42/p44, perhaps through a pathway not dependent on EGFR involvement. Interestingly, we did not observe a significant increase of p42/p44 activation in cells stimulated with EGF in combination with histamine or carbachol compared with cells stimulated with EGF alone. Although we did observe a significant effect of thrombin on p42/p44 activation at the 3 and 6 h time points, in approximately one-half of the experiments the phospho-p42/p44 signals were unchanged at these times, and similar growth potentiation by thrombin occurred irrespective of the effect on p42/p44 activation. Moreover, the difference in integrated p42/p44 activation (30%) over the 12-h stimulation does not appear sufficient to explain the large associated increase in growth, given that similar differences occur upon stimulation with different concentrations of EGF or different mitogens and result in relatively small changes in human ASM growth (17). Thus, although increases in p42/p44 activation may variably contribute to growth synergy effected by thrombin, p42/p44 does not appear to be the principal pathway by which GPCRs potentiate RTK-stimulated growth.

As is the case with p42/p44 MAPK, the PI 3-kinase/p70 S6 kinase pathway has been shown to be an important pathway for transducing mitogenic signals from both RTKs and GPCRs (34). In addition, we have previously demonstrated that inhibition of PI 3-kinase activity with wortmannin or LY294002, or inhibition of p70 S6 kinase with rapamycin ablates both EGF- and thrombin-stimulated human ASM growth (29). We thus examined the time- dependent effect of costimulation with EGF and GPCR agonists on p70 S6 kinase activity in human ASM cultures. Although the initial, rapid activation occurring within the first 30 min of stimulation with EGF did not appear significantly altered by costimulation with thrombin, a clear increase in activity was observed at later time points (1, 2, and 4 h). The effect of costimulation with histamine or carbachol was evidenced in a more extended time course, where time points  3 h exhibited increased p70 S6 kinase activity. Of note, previous studies by Susa and colleagues (35) demonstrated the existence of a late-phase p70 S6 kinase activity (occurring after 30 min of stimulation) that was elicited by mitogens (EGF, PDGF) but not by non-mitogens in Swiss 3T3 fibroblasts. A similar observation by Simm and coworkers (38) suggests that prolonged p70 S6 activity in response to PDGF-AB may represent a "progression" factor that complements early-phase "competence" factors (e.g., p42/p44 activity) to enable cycling in AKR-2B cells. Results from the present study suggest that GPCR activation can further augment this late-phase activity to mediate an even greater mitogenic effect.

What upstream signals mediate the observed increase in late phase p70 S6 kinase activity? PI 3-kinase has been frequently characterized as an upstream regulator of p70 S6 kinase, acting indirectly by regulating other kinases or possibly phosphatases capable of modulating p70 S6 kinase phosphorylation state (34, 39). In numerous cell types (including human ASM), the PI 3-kinase inhibitors wortmannin and LY29400 have been shown to effectively block both the activation of p70 S6 kinase and associated mitogenic effects (29, 41). Despite the clear linkage between PI 3-kinase and p70 S6 kinase, preliminary data from our lab have yet to establish a role for PI 3-kinase in the observed increase in late-phase p70 S6 kinase activity. PI 3-kinase activity in antiphosphotyrosine immunoprecipitates is decidedly out of phase with p70 S6 kinase activation, with both EGF- and thrombin-stimulated activity being very transient and rapidly returning to near-basal levels within 10 and 30 min of stimulation, respectively (29). Moreover, no obvious increase is observed in EGF-stimulated PI 3-kinase activity when costimulated with GPCR agonists (data not shown). Thus, any further augmentation of PI 3-kinase signaling by GPCRs may be associated with an activity independent or downstream of that activity observed in antiphosphotyrosine immunoprecipitates. PI 3-kinase , the class 1b PI 3-kinase isoform activated by subunits of heterotrimeric G proteins (5), could account for such activity, but both immunoblotting and reverse transcriptase/polymerase chain reaction demonstrate an absence of PI 3-kinase  in human ASM (data not shown).

Given the diversity of signaling pathways activated by the various GPCRs, it is probable that mechanisms independent of p70 S6 kinase activation also contribute to the effect of GPCRs in augmenting RTK-mediated growth. The identification of such pathways, as well as the upstream modulators of p70 S6 kinase involved in GPCR-mediated growth potentiation, will likely require the development of methodology that overcomes the present difficulties in transfecting ASM cells. Recent advances in microinjection techniques (44) offer an additional means of molecular manipulation compatible with growth analysis. Alternatively, primary cultures of other cell types amenable to transfection procedures that do not significantly disrupt cell cycling may represent suitable models for examining the interactive effects of GPCR and RTK activation on cell growth.

In summary, the present study demonstrates that GPCR activation by inflammatory and contractile agents can synergize with RTK activation to augment human ASM growth. In EGF-stimulated cells, GPCR-mediated potentiation does not appear mechanistically linked to increased EGFR or p42/p44 MAPK activation but is associated with sustained activation of p70 S6 kinase for several hours after the initial early phase of activation. These findings not only provide insight into mechanisms by which inflammation contributes to ASM hyperplasia/hypertrophy in diseases such as asthma but also suggest a general mechanism by which GPCRs and RTKs interact to promote cell growth.