INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

THE CYTOSKELETON of smooth muscle cells is a filamentous network consisting largely of filamentous actin (F-actin), which provides a scaffold on which motor proteins such as myosin translocate to generate internal stress and alter the mechanical properties of the cells. Extracellular stimuli induce cytoskeleton reorganization (stress-fiber formation) in cells (4, 12, 19, 25, 34) and Ca²⁺ sensitization (increased muscle tension under constant-calcium conditions) in intact smooth muscle preparations (4, 5, 9, 17, 27) by activating signaling pathways that involve Rho proteins, a subfamily of the Ras superfamily of monomeric G proteins. A downstream effector pathway linking Rho proteins to the actin cytoskeleton and Ca²⁺ sensitization has recently been elucidated (20). Rho-associated protein kinase, a product of activated Rho (Rho-GTP), phosphorylates and inactivates the myosin-binding subunit of myosin light chain phosphatase, thereby inhibiting myosin light chain dephosphorylation (20). As a result, the phosphorylated form of myosin light chain accumulates, leading to contraction of the actomyosin-based cytoskeleton and Ca²⁺ sensitization of smooth muscle preparations.

The signaling pathways upstream from Rho are cell-type specific and poorly understood, particularly in airway smooth muscle. Cholinergic agonists (1, 5, 8), endothelin-1 (5), and histamine (8) all sensitize intact airway smooth muscle preparations to Ca²⁺ via a Rho-mediated pathway (5). Lysophosphatidic acid (LPA), a lipid mediator which induces actin reorganization in Swiss 3T3 cells via a pathway involving Rho proteins (12) and stimulates mitogenesis in airway smooth muscle cells (3), enhances contractility of airway smooth muscle preparations in response to cholinergic stimulation (33). Airway smooth muscle preparations express M₂ and M₃ muscarinic cholinergic receptors (6, 30), endothelin type A (ET_A) and type B (ET_B) receptors (14, 39), and presumably at least one type of LPA receptor (3).

Heterotrimeric G proteins are the major upstream entity involved in Rho activation induced by agonists. These G proteins consist of an -subunit and two smaller, tightly coupled subunits,  and . The -subunits are unique to each G protein, conferring functional specificity. The -subunits are subdivided into four major families on the basis of their amino acid sequence homology: 1) G_s and G_olf; 2) G_i-1, G_i-2, G_i-3, G_o, G_z, transducins 1 and 2, and gustducin; 3) G_q, G₁₁, G₁₄, and G₁₅; and 4) G₁₂ and G₁₃. Splice variants of G_s, G_i-2, and G_o as well as additional subfamily members of -subunits have recently been identified (16). In addition, five -subunits and at least 10 -subunits have so far been described (16).

M₃ muscarinic receptors couple to phospholipase C to produce increases in inositol trisphosphate and diacylglycerol via the heterotrimeric G protein G_q, whereas activation of M₂ muscarinic receptors inhibits adenylyl cyclase via interaction with members of the pertussis toxin-sensitive G protein family G_i. ET_A receptors are coupled to the inhibition of adenylyl cyclase via G_i (28) and to the production of inositol trisphosphate via G_q (36). LPA receptors couple to at least three G proteins, including G_q, which links the receptor to phospholipase C; G_i, which triggers Ras activation and adenylyl cyclase inhibition; and G_{12 /13}, which mediates Rho activation (24).

Togashi et al. (34) recently demonstrated that carbachol exposure led to actin reorganization in human airway smooth muscle cells that express mainly M₂ muscarinic receptors (37). Moreover, this actin reorganization was blocked by pretreatment with atropine, Clostridium botulinum C3 exoenzyme, or pertussis toxin (34), implicating muscarinic receptors, monomeric G proteins of the Rho family, and pertussis-sensitive heterotrimeric G proteins, respectively, in this pathway. In a subsequent study using antisense oligonucleotides designed to specifically bind to the mRNA encoding G_i-2, G_i-3, or G_q, Hirshman et al. (18) showed that antisense oligonucleotide depletion of G_i-2 protein but not of G_i-3 or G_q protein blocked carbachol-induced increases in actin reorganization in the same cells. These data indicate that G_i-2 proteins couple M₂ muscarinic receptors to Rho proteins and actin reorganization in human airway smooth muscle cells.

The goal of the present study was to investigate whether receptors that couple to the G protein G_q also activate Rho and induce actin reorganization in human airway smooth muscle cells. Using cultured human airway smooth muscle cells that express endothelin, LPA, and M₂ muscarinic receptors but not M₃ muscarinic receptors, we evaluated the ability of carbachol, endothelin-1, and LPA to induce actin reorganization in these cells. Subsequently, we evaluated the ability of pertussis toxin and C3 exoenzyme to block the receptor-mediated effects. Finally, we used an antisense oligonucleotide approach capable of downregulating individual G protein -subunits. Antisense oligonucleotides were designed to specifically bind to mRNA encoding G_i-2, G_i-3, and G_q. We found that LPA and endothelin-1, like carbachol, induced actin reorganization. Moreover, pretreatment of the cells with C3 exoenzyme but not with pertussis toxin inhibited endothelin-1- and LPA-induced actin reorganization; and antisense oligonucleotide depletion of both G_i-2 and G_q proteins but not either one alone inhibited endothelin-1- or LPA-induced actin reorganization in these cells. This study provides the first evidence that both G_i-2 and G_q couple endothelin and LPA receptors to Rho proteins in human airway smooth muscle cells.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Cell culture. Primary cultures of previously characterized human tracheal smooth muscle cells (kindly provided by Dr. Ian Hall, Nottingham, UK) (13, 37) were maintained in medium 199 containing antibiotics (100 U/ml of penicillin G, 100 µg/ml of streptomycin, and 0.25 µg/ml of amphotericin B) and 10% fetal bovine serum at 37°C in an atmosphere of 5% CO₂-95% air. Preliminary immunohistochemical studies performed in our laboratory confirmed that >90% of the cells expressed -actin. Moreover, immunoblot analysis of these cells identified expression of both -actin and desmin, confirming the smooth muscle phenotype of the cells. The cells were plated on eight-well microscope slides (Nunc Chambers, Naperville, IL) and incubated until the cells achieved confluence. The cells were extensively washed and maintained in serum-free medium 199 for 48 h. Quiescent, serum-starved cells were stimulated with either 100 µM carbachol, 1 µM endothelin-1, or 1 µM LPA for 5 min. Because activation of the cytoskeleton by nonspecific stimuli is a major concern, physical manipulation of the slides was kept to a minimum. Termination of agonist activation was achieved by the addition directly to the medium of an equal volume of 7.4% fresh paraformaldehyde in PBS for 15 min.

In some experiments, the cells were pretreated with pertussis toxin to inactivate G_i proteins or with C3 exoenzyme to block Rho activation before exposure to receptor agonists. The cells were exposed to pertussis toxin (final concentration 100 ng/ml) in serum-free medium for 4 h at 37°C in a cell culture incubator, after which the cells were left untreated or treated with agonist for 5 min. Termination of agonist activation was achieved by the addition of an equal volume of 7.4% fresh paraformaldehyde in PBS for 15 min. C3 exoenzyme pretreatment was performed for 72 h. The cells were incubated with C3 exoenzyme (final concentration 10 µg/ml) for 24 h in medium 199 containing 10% serum and subsequently with 10 µg/ml of C3 exoenzyme for 48 h in serum-free medium. This pretreatment with C3 has been previously shown (34) to block carbachol-induced Rho activation in these cells. The cells were either left untreated or treated with agonist for 5 min. Termination of agonist activation was achieved by the addition directly to the medium of an equal volume of 7.4% fresh paraformaldehyde in PBS for 15 min.

In some experiments, the cells were treated with specific antisense oligonucleotides (final concentration 10 µM) directed against the G protein -subunits G_i-2, G_i-3, or G_q. The concentration and time of treatment (6 days) has been previously shown by us and others (18, 31, 32) to result in decreased expression of the respective G protein -subunit by immunoblot analysis. These high concentrations of relatively small oligonucleotides are apparently taken up by cells in the absence of lipid or viral vectors, resulting in quantitative decreases in the respective targeted proteins (31, 32). In some experiments, the cells were treated with a combination of antisense oligonucelotides (G_i-2 and G_q or G_i-3 and G_q). The oligonucleotide sequences were phosphorothioated at the first and last four nucleotides to impair intracellular degradation and enhance resistance to exo- and endonucleases. The oligonucleotides were commercially synthesized as follows: 5'-CTT GTC GAT CAT CTT AGA-3' for G_i-2, 5'-AAG TTG CGG TCG ATC AT-3' for G_i-3, and 5'-GCT TGA GCT CCC GGC GGG CG-3' for G_q (18, 31). Antisense oligonucleotide pretreatment was performed for a total of 6 days. The medium was changed and the oligonucleotide was redosed every 2 days. The first 4 days of incubation were performed in medium 199 with 1% fetal bovine serum, whereas the last 2 days of incubation were performed in serum-free medium. This pretreatment has been shown to result in decreased protein expression of the respective G protein -subunit in these cells (18).

Fluorescence microscopy. The staining protocol for F-actin and globular actin (G-actin) was a modification of previous methods of Knowles and McCulloch (21). Preliminary studies were performed comparing the methods of fixation (methanol versus paraformaldeyde) and the addition of fluorescent stains [FITC-labeled phalloidin (FITC-phalloidin) and Texas Red-labeled DNase I (Texas Red-DNase I)] sequentially or concurrently. Methanol fixation resulted in high background staining and was considered unsatisfactory as previously reported by Knowles and McCulloch. We obtained similar results when stains were added sequentially or concurrently. Therefore, the stains were added concurrently to ensure that all wells received identical concentrations of both stains. Preliminary studies also suggested that agonist-induced increases in F-actin staining was time dependent, and, therefore, the fixative was added immediately after 5 min of agonist exposure to stop the agonist effects equally in all wells. After fixation in 3.7% (final concentration) fresh paraformaldehyde in PBS for 15 min, the wells were washed twice with PBS, excess aldehyde was quenched with 50 mM NH₄Cl for 15 min, and then the cells were permeabilized with 0.5% Triton X-100 in PBS for 5 min. After treatment with blocking solution (1% BSA and 0.1% Triton X-100 in PBS) for 10 min, the cells were stained with FITC-phalloidin (1 µg/ml) in blocking solution for 20 min in a dark room at room temperature to localize F-actin and Texas Red-DNase I (10 µg/ml) to localize monomeric (G) actin (21). The slides were washed twice with 0.1% Triton X-100 in PBS and once with PBS alone, each for 5 min. Incubation and washing were performed in parallel for all wells on a slide. A coverslip was mounted on the slide with Vectashield H-1000 (Vector Laboratories, Burlingame, CA). Actin was visualized with a fluorescence microscope (Olympus BHT, Tokyo, Japan), and the image was stored with Image-Pro Plus software (Medica Cybernetics, Silver Spring, MD) on a personal computer.

The fluorescence intensities of FITC-phalloidin and Texas Red-DNase I were simultaneously calculated from a view containing >15 cells. Measurements were taken from three fields for each treatment and averaged for a single data point. The excitation and emission wavelengths for FITC-phalloidin were 490 and 525 nm, respectively, whereas the excitation and emission wavelengths for Texas Red-DNase I were 596 and 615 nm, respectively. To standardize the fluorescence intensity measurements among experiments, the time of image capturing, image intensity gain, image enhancement, and image black level in both channels were optimally adjusted at the outset and kept constant for all experiments. Images at a maximum diameter were digitized (640 × 484 pixels) with an eight-bit gray-level resolution of 0 (minimum) to 256 (maximum) intensity. Cumulative fluorescence intensities for FITC-phalloidin and Texas Red-DNase I were recorded with Image-Pro Plus software. An increase in the F- to G-actin ratio indicated an increase in actin reorganization.

Materials. Carbachol, LPA, endothelin-1, and FITC-phalloidin were obtained from Sigma (St. Louis, MO). Texas Red-DNase I was obtained from Molecular Probes (Eugene, OR). Phosphorothioate-modified oligonucleotides were obtained from GIBCO BRL (Life Technologies, Gaithersburg, MD). Pertussis toxin was purchased from List Biological Laboratories (Campbell, CA). C3 exoenzyme was purchased from Calbiochem (La Jolla, CA).

Statistical analysis of data. To control for day-to-day variations in staining intensity, untreated cells were always compared with treated cells on the same microscope slide because cells on the same slide undergo identical culture, fixation, permeabilization, staining, and microscopy conditions, allowing meaningful comparisons between samples. All data are presented as means ± SE. F- to G-actin ratios were compared by two-way ANOVA with Bonferroni posttest comparisons with Instat software (GraphPad, San Diego, CA). P < 0.05 was considered significant.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Exposure of serum-deprived human airway smooth muscle cells to 100 µM carbachol, 1 µM endothelin, or 1 µM LPA for 5 min resulted in an increase in the FITC-phalloidin staining intensity of F-actin and a decrease in the Texas Red-DNase I staining intensity of G-actin (indicative of reorganization of G-actin into F-actin fibers) compared with that in the untreated cells (Fig. 1). The F- to G-actin fluorescent-staining ratio indicative of actin fiber reorganization significantly increased from 2.4 ± 0.3 in the untreated cells to 6.7 ± 0.8, 7.2 ± 0.8, and 7.4 ± 0.9 in the carbachol-, LPA-, and endothelin-1-treated cells, respectively (P < 0.001 for each agonist compared with control group; n = 14 experiments). The decrease in G actin staining intensity in response to carbachol is presented in Fig. 2.

View larger version (127K): [in this window] [in a new window]   Fig. 1.   Representative photomicrographs of cultured human airway smooth muscle cells stained with FITC-phalloidin to illustrate filamentous (F) actin fibers. Carbachol (100 µM; B), lysophaphatidic acid (LPA; 1 µM; C), or endothelin-1 (1 µM; D) for 5 min induced increased F-actin staining compared with that in untreated cells (A). Pretreatment with C3 exoenzyme (E) for 72 h blocked carbachol (F)-, LPA (G)-, or endothelin-1 (H)-induced increases in F actin staining.

View larger version (41K): [in this window] [in a new window]   Fig. 2.   Representative photomicrographs of cultured human airway smooth muscle cells stained with Texas Red-DNase I to illustrate globular (G) actin fibers. Under unstimulated control conditions (A), heaviest staining occurred in perinuclear region, whereas carbachol (100 µM; B) decreased staining intensity.

View larger version (19K): [in this window] [in a new window]   Fig. 3.   Fluorescent-staining ratios of F- to G- (F/G) actin in cultured human airway smooth muscle cells. Fluorescent intensity of F-actin staining by FITC-phalloidin and G-actin staining by Texas Red-DNase I were measured in the same field in triplicate for each treatment. Carbachol, LPA, and endothelin-1 (n = 14 experiments) each increased ratio of F/G actin, indicating actin reorganization. P < 0.001 for each effector compared with control. C3 exoenzyme pretreatment (n = 5 experiments) blocked effect of all 3 agonists, indicating that small G protein Rho is an intermediate in the pathway. * P < 0.05 compared with no C3 pretreatment for each effector.

View larger version (119K): [in this window] [in a new window]   Fig. 4.   Representative photomicrographs of cultured human airway smooth muscle cells stained with FITC-phalloidin to illustrate F-actin fibers. Carbachol (100 µM; B), LPA (1 µM; C), or endothelin-1 (1 µM; D) for 5 min induced increased F-actin staining compared with that in untreated cells (A). Four hours of pertussis toxin pretreatment (E) blocked carbachol (F)- but not LPA (G)- or endothelin-1 (H)-induced increases in F-actin staining.

View larger version (17K): [in this window] [in a new window]   Fig. 5.   Fluorescent-staining ratios of F/G actin in cultured human airway smooth muscle cells. Fluorescent intensity of F-actin staining by FITC-phalloidin and G-actin staining by Texas Red-DNase I were measured in the same field in triplicate for each treatment. Carbachol, LPA, or endothelin-1 (n = 14 experiments) each increased ratio of F/G actin, indicating actin reorganization. P < 0.001 for each effector compared with control. Pertussis toxin pretreatment (n = 5 experiments) blocked effect of carbachol but not of LPA or endothelin-1, indicating that Gi proteins are predominant intermediary heterotrimeric G proteins activated by carbachol, but that other or additional G proteins are intermediates for LPA and endothelin-1. * P < 0.05 for pertussis effect on carbachol only.

View larger version (95K): [in this window] [in a new window]   Fig. 6.   Representative photomicrographs of cultured human airway smooth muscle cells stained with FITC-phalloidin to illustrate F-actin fibers. Cells were left untreated (control; A-D) or pretreated for 6 days with antisense oligonucleotides directed against Gi-2 (E-H). Subsequently, cells were left untreated (A and E) or were treated with carbachol (100 µM; B and F), LPA (1 µM; C and G), or endothelin-1 (1 µM; D and H) for 5 min before fixation and staining. Carbachol-induced increases in F-actin staining were blocked by pretreatment with Gi-2 antisense oligonucleotide, but LPA- and endothelin-1-induced increases were unaffected. P < 0.05 for carbachol compared with carbachol plus Gi-2 antisense oligonucleotide.

View larger version (102K): [in this window] [in a new window]   Fig. 7.   Representative photomicrographs of cultured human airway smooth muscle cells stained with FITC-phalloidin to illustrate F-actin fibers. Cells were left untreated (control; A-D) or pretreated for 6 days with 2 different antisense oligonucleotides directed against Gi-2 and Gq (E-H). Subsequently, cells were left untreated (A and E) or were treated with carbachol (100 µM; B and F), LPA (1 µM; C and G), or endothelin-1 (1 µM; D and H) for 5 min before fixation and staining. Carbachol-, LPA-, or endothelin-1-induced increases in F-actin staining were blocked by pretreatment with combined Gi-2 and Gq antisense oligonucleotides. P < 0.05 for each effector compared with effector plus combined antisense oligonucleotide.

In separate experiments, pretreatment of the human airway smooth muscle cells with C3 exoenzyme for 72 h totally inhibited actin reorganization by either carbachol, endothelin-1, or LPA, indicating that Rho proteins are intermediates in the signaling pathway leading from cell surface receptors to actin reorganization. The F- to G-actin fluorescent-staining ratio averaged 2.5 ± 0.1 in the control cells, 1.9 ± 0.3 in the control cells pretreated with C3 exoenzyme, 4.6 ± 0.5 in the carbachol-treated cells, 2.6 ± 0.2 in the carbachol-treated cells pretreated with C3 exoenzyme, 5.7 ± 0.7 in LPA-treated cells, 2.8 ± 0.4 in LPA-treated cells pretreated with C3 exoenzyme, 5.9 ± 0.9 in endothelin-1-treated cells, and 2.7 ± 0.6 in endothelin-1-treated cells pretreated with C3 exoenzyme (P < 0.05 for C3 effect on each agonist; n = 5 experiments; Figs. 1 and 3).

View larger version (121K): [in this window] [in a new window]   Fig. 8.   Representative photomicrographs of cultured human airway smooth muscle cells stained with FITC-phalloidin to illustrate F-actin fibers. Cells were left untreated (control; A-D) or were pretreated for 6 days with antisense oligonucleotides directed against Gq (E-H). Subsequently, cells were left untreated (A and E) or were treated with carbachol (100 µM; B and F), LPA (1 µM; C and G), or endothelin-1 (1 µM; D and H) for 5 min before fixation and staining. Carbachol-, LPA-, or endothelin-1-induced increases in F-actin staining were unaffected by pretreatment with Gq antisense oligonucleotide alone. P > 0.05 compared with no antisense oligonucleotide for each effector.

Pretreatment of airway smooth muscle cells with pertussis toxin for 4 h to inactivate the heterotrimeric G protein G_i led to the inhibition of carbachol- but not of endothelin-1- or LPA-induced actin reorganization. The F- to G-actin fluorescent-staining ratio averaged 2.9 ± 0.3 in the control cells, 2.9 ± 0.5 in the pertussis toxin-pretreated controls cells, 6.2 ± 1.2 in the carbachol-treated cells, 3.5 ± 0.5 in the carbachol-treated cells pretreated with pertussis toxin, 6.4 ± 1.6 in the LPA-treated cells, 5.7 ± 1.2 in the LPA-treated cells pretreated with pertussis toxin, 5.5 ± 0.9 in the endothelin-1-treated cells, and 5.5 ± 0.9 in the endothelin-1-treated cells pretreated with pertussis toxin (P < 0.05 for pertussis toxin effect on carbachol only; n = 5 experiments; Figs. 4 and 5). These data suggest that endothelin-1 and LPA couple to actin reorganization via a pathway independent of G_i proteins (e.g., G_q).

View larger version (117K): [in this window] [in a new window]   Fig. 9.   Representative photomicrographs of cultured human airway smooth muscle cells stained with FITC-phalloidin to illustrate F-actin fibers. Cells were left untreated (control; A-D) or were pretreated for 6 days with 2 different antisense oligonucleotides directed against Gq and Gi-3 (E-H). Subsequently, cells were left untreated (A and E) or were treated with carbachol (100 µM; B and F), LPA (1 µM; C and G), or endothelin-1 (1 µM; D and H) for 5 min before fixation and staining. Carbachol-, LPA-, or endothelin-1-induced increases in F-actin staining were unaffected by pretreatment with combined Gi-3 and Gq antisense oligonucleotides. P > 0.05 compared with no antisense oligonucleotide for each effector.

To further characterize the heterotrimeric G proteins that are intermediates leading from receptor activation to actin reorganization, depletion of G protein -subunits with antisense oligonucleotides was performed in a separate series of experiments. Pretreatment of the airway smooth muscle cells with 10 µM antisense oligonucleotide for 6 days resulted in a decrease in protein expression of the respective G protein -subunit (18). G_i-2 antisense oligonucleotide pretreatment significantly blocked carbachol-induced actin reorganization but had no significant effect on either LPA- or endothelin-1-induced actin reorganization. The F- to G-actin fluorescent-staining ratio averaged 6.7 ± 0.8 in the carbachol-treated cells, 3.8 ± 0.07 in the carbachol-treated cells pretreated with G_i-2 antisense oligonucleotide, 7.2 ± 0.8 in the LPA-treated cells, 6.6 ± 2.3 in the LPA-treated cells pretreated with G_i-2 antisense oligonucleotide, 7.4 ± 0.9 in the endothelin-1-treated cells, and 7.1 ± 2.7 in the endothelin-1-treated cells pretreated with G_i-2 antisense oligonucleotide (P < 0.05 for G_i-2 antisense oligonucleotide effect on carbachol only; n = 3 experiments; Figs. 6 and 10). In contrast, pretreatment with both G_i-2 and G_q antisense oligonucleotides blocked not only carbachol-induced actin reorganization but also LPA- and endothelin-1-induced actin reorganization. The F- to G-actin fluorescent-staining ratio averaged 6.7 ± 0.8 in the carbachol-treated cells, 3.8 ± 0.9 in the carbachol-treated cells pretreated with both G_i-2 and G_q antisense oligonucleotides, 7.2 ± 0.8 in the LPA-treated cells, 4.0 ± 0.7 in the LPA-treated cells pretreated with both G_i-2 and G_q antisense oligonucleotides, 7.4 ± 0.9 in the endothelin-1-treated cells, and 3.8 ± 0.4 in the endothelin-1-treated cells pretreated with both G_i-2 and G_q antisense oligonucleotides (P < 0.05 for the combined G_i-2 and G_q antisense oligonucleotide effect on each agonist; n = 5 experiments; Figs. 7 and 10). Neither pretreatment with G_q antisense oligonucleotide alone (Figs. 8 and 10) nor the combination of G_q and G_i-3 antisense oligonucleotides (Figs. 9 and 10) had a significant effect on actin reorganization induced by carbachol, LPA, or endothelin-1. The F- to G-actin fluorescent-staining ratio averaged 6.7 ± 0.8 in the carbachol-treated cells, 6.1 ± 1.0 in the carbachol-treated cells pretreated with G_q antisense oligonucleotide, 8.0 ± 2.0 in the carbachol-treated cells pretreated with both G_i-3 and G_q antisense oligonucleotides, 7.2 ± 0.8 in the LPA-treated cells, 6.9 ± 1.6 in the LPA-treated cells pretreated with G_q antisense oligonucleotide, 9.1 ± 2.2 in LPA-treated cells pretreated with both G_i-3 and G_q antisense oligonucleotides, 7.4 ± 0.9 in endothelin-1-treated cells, 6.6 ± 1.9 in endothelin-1-treated cells pretreated with G_q antisense oligonucleotide, and 8.6 ± 2.8 in endothelin-1-treated cells pretreated with both G_i-3 and G_q antisense oligonucleotides (n = 3 experiments). Antisense oligonucleotide pretreatment had no significant effect on the F- to G-actin ratios in the untreated (control) cells. The F- to G-actin fluorescent-staining ratios averaged 2.4 ± 0.3 in the untreated cells, 3.8 ± 0.7 in the untreated cells pretreated with G_i-2 antisense oligonucleotide, 2.6 ± 0.6 in the untreated cells pretreated G_q antisense oligonucleotide, 2.4 ± 0.4 in the untreated cells pretreated with both G_q and G_i-2 antisense oligonucleotides, and 3.9 ± 1.0 in the untreated cells pretreated with G_q and G_i-3 antisense oligonucleotide (n = 3 experiments). Taken together, these data suggest that carbachol-induced actin reorganization couples predominantly through a G_i-2 pathway, whereas LPA- and endothelin-1-induced actin reorganization couples through both G_i-2 and G_q pathways.

View larger version (40K): [in this window] [in a new window]   Fig. 10.   Fluorescent-staining ratios of F/G actin in cultured human airway smooth muscle cells. Fluorescent intensity of F-actin staining by FITC-phalloidin and G-actin staining by Texas Red-DNase I were measured in the same field in triplicate for each treatment. Effects of antisense oligonucleotides directed against Gi-2, Gq, or Gi-3 or combined treatment with Gq and Gi-2 or Gq and Gi-3 were measured in presence of carbachol, LPA, or endothelin-1. Carbachol-induced increases in F/G actin ratio were blocked by pretreatment with antisense oligonucleotides against Gi-2 or combined treatment with antisense oligonucleotides against Gi-2 and Gq. In contrast, LPA- or endothelin-1-induced increases in F/G actin ratios were only blocked by pretreatment with a combination of antisense oligonucleotides directed against Gq and Gi-2. * P < 0.05 compared with no antisense oligonucleotide for each effector.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

This study demonstrates for the first time that activation of receptors activated by either LPA or endothelin-1 led to actin reorganization in human airway smooth muscle cells. Pretreatment of these cells with C3 exoenzyme blocked the ability of these agonists to induce actin reorganization, indicating a role for the monomeric G protein Rho in this pathway. Moreover, the combined antisense oligonucleotide depletion of both the G_q and G_i-2 proteins blocked LPA- and endothelin-1-induced actin reorganization, whereas antisense oligonucleotide depletion of the G_q, G_i-2, or G_i-3 protein alone or the combined depletion of G_q and G_i-3 was insufficient to block either LPA- or endothelin-1-induced actin reorganization in these cells. These data implicate for the first time that both G_q and G_i-2 are linked to Rho in human airway smooth muscle cells.

The present investigation used dual labeling with FITC-phalloidin and Texas Red-DNase I adapted from the method of Knowles and McCulloch (21) to image and quantify actin reorganization in these cells. This method allows for simultaneous observation of the relative amounts and configuration of F- and G-actin in the same cell because the staining patterns of F- and G-actin are thought to be spatially separate and distinct. The labeling of F-actin with FITC-phalloidin and of G-actin with Texas Red-DNase I is known to be specific (21), although variations in staining intensity often occur between studies. Factors that contribute to different absolute values between experiments include the affinity (dictated by the stability of the fluorescent dyes in storage) of the phalloidin-FITC and Texas Red-DNase I conjugates, the amount of photobleaching that occurs during analysis and storage of slides, and the sensitivity of the camera settings. To control for these variations in staining intensity, untreated cells were always compared with treated cells on the same microscope slide because cells on the same slide undergo identical culture, fixation, permeabilization, staining, and microscopy conditions, allowing meaningful comparisons between samples. These fluorescent-staining probes for G- and F-actin do not allow us to differentiate between the different isoforms of actin that might be participating in the polymerization process.

The present study is an extension of previously published data from our laboratory demonstrating that in human airway smooth muscle cells, carbachol induces actin reorganization by a signaling pathway that involves G_i-2 (18) and Rho proteins (34). Pretreatment with either C3 exoenzyme, pertussis toxin (34), or antisense oligonucleotides directed against G_i-2 (18) almost totally blocked carbachol-induced actin reorganization because human airway smooth muscle cells used in this study express mainly M₂ muscarinic receptors (37), in contrast to native airway smooth muscle that expresses both M₂ and M₃ muscarinic receptors.

The present study measuring the F- to G-actin ratios is consistent with previously published functional data from our laboratory demonstrating that in porcine tracheal smooth muscle, endothlin-1 and muscarinic agonists induce Ca²⁺ sensitization by a pathway involving Rho proteins (5). Pretreatment with C3 exoenzyme inhibited both acetylcholine- and endothelin-1-induced Ca²⁺ sensitization in porcine tracheal smooth muscle (5). Moreover, pretreatment with pertussis toxin inhibited acetylcholine- but not endothelin-1-induced Ca²⁺ sensitization (5) in the tracheal muscle tissue similar to that seen in human airway smooth muscle cells. The present results in airway smooth muscle cells agree with a study (22) in astrocytes showing that endothelin-1-induced actin reorganization was inhibited by C3 exoenzyme but not by pertussis toxin.

In the present study, results with antisense oligonucleotides directed against both G_i-2 and G_q demonstrate that in human airway smooth muscle cells, endothelin receptors couple to both G_i-2 and G_q. ET_A and ET_B receptors are known to couple to several heterotrimeric G proteins including G_i, G_s, and G_q (10). Endothelin-1 increases inositol trisphosphate via the ET_A receptor in cultured rat (15) and canine (38) airway smooth muscle cells, presumably via G_q, and activates p21^ras via a pertussis-sensitive G protein, presumably G_i, in human airway smooth muscle cells (7). The present study identifies a novel signaling pathway for endothelin-1 in airway smooth muscle cells: the activation of Rho with pathways mediated by either G_q or G_i-2.

The lack of effect of antisense oligonucleotides directed against G protein -subunits in the control cells demonstrates that human airway smooth muscle cells maintain some basal level of actin polymerization independent of heterotrimeric G proteins. This suggests that additional regulatory pathways dictate the state of actin polymerization in these cells and that activation of cell surface receptors coupled to heterotrimeric G proteins is only one of several pathways modulating actin polymerization in these cells.

The present results in airway smooth muscle cells agree with previously published studies in other cell types in which LPA induces actin reorganization by signaling pathways involving Rho proteins. This LPA-induced actin reorganization occurs in Swiss 3T3 (29), JIC9 (a subclone of Chinese hamster embryo fibroblasts) (11), and mouse NIE-115 neuroblastoma cells (23) and in rat-1 and hamster lung fibroblasts (35). The heterotrimeric G protein(s) used by LPA to induce Rho-mediated actin reorganization appear to be cell-type specific. LPA receptors have been shown to couple to G_q, G_i, and G_{12 /13}. In fibroblasts, LPA-induced actin reorganization is inhibited by pertussis toxin (11, 35), suggesting that G_i couples LPA receptors to Rho-induced actin reorganization, whereas in Swiss 3T3 and neuronal cells, G_{12 /13} couples LPA receptors to Rho-induced actin reorganization (2, 24). The results from the present study differ from those of previously published studies (2, 11, 24, 35) in that either of two heterotrimeric G proteins, G_q or G_i-2, is capable of coupling LPA receptors to Rho-induced actin reorganization in human airway smooth muscle cells. Our study agrees with a study by Nogami et al. (26), who showed that LPA-activated receptors coupled to both pertussis-sensitive and pertussis-insensitive pathways in human airway smooth muscle cells.

The ability of LPA and endothelin-1 to induce actin polymerization by both G_i and G_q implies signaling pathway redundancy in these cells. It is likely that activation of G_i and G_q also leads to activation of other signaling intermediates (characteristic of G_i and G_q, respectively) that are not redundant for these receptors. For example, LPA-induced activation of G_i may lead to inhibition of adenylyl cyclase, whereas its coupling to G_q may lead to activation of phospholipase C. These would be independent and nonredundant functions of LPA-receptor activation. It is possible that important cellular signaling pathways are redundant so that dysfunction of one pathway does not lead to cell death. Because cytoskeletal reorganization is a pivotal process in cell motility, division, secretion, and contraction, it is likely that redundant pathways are a cellular protective mechanism.

In conclusion, this study demonstrates that endothelin-1 and LPA induce actin reorganization in human airway smooth muscle cells via a pathway involving Rho proteins and that antisense oligonucleotide depletion of both G_q and G_i-2 proteins significantly inhibited endothelin-1- and LPA-induced actin reorganization in these cells. This study provides the first evidence that G_q as well as G_i-2 is linked to Rho proteins in human airway smooth muscle cells.