β-Agonist-associated Reduction in RGS5 Expression Promotes Airway Smooth Muscle Hyper-responsiveness*

Although short-acting and long-acting inhaled β2-adrenergic receptor agonists (SABA and LABA, respectively) relieve asthma symptoms, use of either agent alone without concomitant anti-inflammatory drugs (corticosteroids) may increase the risk of disease exacerbation in some patients. We found previously that pretreatment of human precision-cut lung slices (PCLS) with SABA impaired subsequent β2-agonist-induced bronchodilation, which occurred independently of changes in receptor quantities. Here we provide evidence that prolonged exposure of cultured human airway smooth muscle (HuASM) cells to β2-agonists directly augments procontractile signaling pathways elicited by several compounds including thrombin, bradykinin, and histamine. Such treatment did not increase surface receptor amounts or expression of G proteins and downstream effectors (phospholipase Cβ and myosin light chain). In contrast, β-agonists decreased expression of regulator of G protein signaling 5 (RGS5), which is an inhibitor of G-protein-coupled receptor (GPCR) activity. RGS5 knockdown in HuASM increased agonist-evoked intracellular calcium flux and myosin light chain (MLC) phosphorylation, which are prerequisites for contraction. PCLS from Rgs5−/− mice contracted more to carbachol than those from WT mice, indicating that RGS5 negatively regulates bronchial smooth muscle contraction. Repetitive β2-agonist use may not only lead to reduced bronchoprotection but also to sensitization of excitation-contraction signaling pathways as a result of reduced RGS5 expression.

Asthma is a complex disorder of unknown etiology characterized by increased bronchial smooth muscle contraction, lung inflammation, and tissue remodeling. These and other abnormalities reduce airway luminal diameter and increase stiffness. A cardinal feature of asthma is "airway hyper-responsiveness" (AHR), 2 which refers to the increased bronchial con-traction to inhaled procontractile agonists such as methacholine that is observed in asthmatics relative to healthy controls. AHR manifests as reduced airflow (1)(2)(3)(4). In allergic asthma, mediators secreted by activated inflammatory leukocytes or resident lung cells including histamine, leukotrienes, bradykinin, and thrombin promote ASM contraction by activating GPCRs linked to activation (GTP binding) of G␣ q (5). This initiates a signaling route culminating in generation of inositol (3,4,5)-trisphosphate and release of Ca 2ϩ from endoplasmic reticulum by activation of IP3 receptors. Hydrolysis of GTP by G␣ q promotes pathway deactivation through formation of inactive G␣ q -GDP-G␤␥ heterotrimers. G␣ q signaling is required for AHR in mouse models of allergic airway inflammation (6).
Receptor-related events increase the frequency of intracellular Ca 2ϩ oscillations in ASM, which induces Ca 2ϩ -calmodulindependent protein kinase (CaMK)-mediated activation of myosin light chain kinase (MLCK). Phosphorylation of myosin light chain (MLC) on Ser-19 by MLCK promotes actin-myosin filament interactions and muscle fiber shortening through the mechanism known as cross-bridge cycling (7). Sustained ASM contraction requires continual and/or increased MLCK activity in the presence of constant Ca 2ϩ oscillation frequency, which is referred to as Ca 2ϩ sensitization. This mechanism is controlled independently of Ca 2ϩ concentration by kinases including Rho-associated kinase through inhibition of MLC phosphatase activity, resulting in prolonged MLC phosphorylation (8).
Inhaled SABA and LABA, which are commonly prescribed for the treatment of asthma, stimulate ␤ 2 -adrenergic receptors (␤ 2 AR) on ASM, promoting relaxation and enhanced airflow (9). ␤ 2 AR activation of G␣ s stimulates activity of the membrane enzyme adenylyl cyclase, which converts adenosine triphosphate (ATP) to cyclic adenosine monophosphate (cAMP). A primary function of cAMP is to activate protein kinase A (PKA), which in turn activates myosin light chain phosphatase, reducing MLC phosphorylation, and generating relaxation (5,7). However, chronic administration of either of these agonists induces tolerance due to receptor desensitization (10 -11). Mechanisms underlying receptor desensitization include decreased ␤ 2 AR transcription, uncoupling of receptors from adenylyl cyclase through down-regulation of G␣ s (12)(13), internalization of uncoupled receptors, and increased phosphorylation and degradation of internalized receptors (9). Regular clinical use of these agents has been linked to heightened sensitivity to allergen-evoked AHR, exacerbation of asthma symptoms, and even death (14 -15). For these reasons, the Food and Drug Administration recommends that ␤ 2 AR agonists primarily be reserved for moderate to severe asthma and used in combination with inhaled corticosteroids (15)(16)(17).
Unfortunately, the mechanism(s) responsible for the seemingly paradoxical effects of SABA and LABA on bronchial contractility/AHR remain unresolved. Several studies have demonstrated that chronic application of ␤ 2 -agonists to whole animals (12), mammalian tracheal smooth muscle strips ex vivo (18 -20), or cultured ASM (21) increased contraction or Ca 2ϩ responses to agonists (histamine, methacholine, or endothelin-1). In some cases, hyper-responsiveness was accompanied by up-regulation of cognate receptors for the procontractile ligand (18 -19). Transgenic overexpression of ␤ 2 AR in mice has also been utilized to evaluate chronic, persistent receptor activity. Surprisingly, tracheas isolated from ␤ 2 AR transgenic mice contracted more than wild-type counterparts, and cultured smooth muscle cells had enhanced IP3 accumulation in response to contractile agonists (22). Mechanisms underlying these responses included up-regulation of PLC␤ in smooth muscle cells from ␤ 2 AR-overexpressing mice. These and other studies demonstrate that in addition to its acute effect on signaling pathways, prolonged ␤ 2 -receptor activation may also elicit transcriptional changes in ASM (20).
Here, we investigated the importance of transcriptional adaptations for the activity of excitation-contraction coupling in ASM. We found that exposure of ASM to the high-affinity ␤ 2 -agonist isoproterenol enhanced Ca 2ϩ responses to several GPCR ligands independently of changes in receptor levels by triggering decreased expression of RGS5. RGS proteins inhibit GPCR responses downstream of receptor-ligand interactions by accelerating the intrinsic GTPase activity of G␣ i and G␣ q . G protein deactivation leads to enhanced formation of inactive G␣␤␥ heterotrimers and reduced signal output (23). Knockdown of RGS5 phenocopied the effects of isoproterenol on ASM, resulting in increased intracellular Ca 2ϩ flux and myosin light chain phosphorylation. These studies suggest that ␤ 2 -agonists directly sensitize the contractile pathways of ASM through mechanisms that include a reduction in RGS5 transcription.

EXPERIMENTAL PROCEDURES
Cell Culture-Human ASM cells were isolated from lung transplant donors as previously described in accordance with procedures approved by the University of Pennsylvania Committee on Studies Involving Human Beings (11). The cells were routinely maintained in Ham's F-12 medium (Invitrogen) supplemented with 10% fetal bovine serum, 100 units/ml penicillin, 100 g/ml streptomycin, and 25 mM HEPES buffer. Cultured medium was replaced every 72 h. Cells in subculture during passage 2-4 were used. HEK293T cells were cultured in complete Dulbecco's modified Eagle's medium (Invitrogen) con-taining 10% (v/v) fetal bovine serum, 1% penicillin, and streptomycin at 37°C with 5% CO 2 .
PCLS Contraction-Lung slice studies were performed as described (24). Rgs5 Ϫ/Ϫ mice (25) were backcrossed onto a C57BL/6 background for more than 5 generations. WT littermates were used as controls. All mouse experiments were performed in accordance with Animal Study Protocol LAD-3E, approved by the NIAID Animal Care and Use Committee, NIH.
RNA Isolation, RT-PCR, and Quantitative RT-PCR-RNA was isolated using Trizol reagent (Invitrogen) per the manufacturer's instructions. Total RNA (1 g) was reverse transcribed into cDNA using the SuperScript III first strand synthesis kit (Invitrogen). Reverse transcriptase-polymerase chain reaction (RT-PCR) amplification was performed using 35 cycles consisting of a 30-s denaturation step at 94°C, followed by annealing for 30 s at 56°C, and extension for 30 s at 72°C. The sequences of the forward and reverse primers used for identification of human RGS mRNAs are shown in supplemental Table S1. Quantitative real-time PCR (qPCR) was performed with Taq-Man gene expression assay probes for each specific RGS using an ABI 7500 HT thermal cycler (Applied Biosystems). The 20-l reaction contained 10 l of 2ϫ TaqMan Universal Master Mix, 1 l of specific RGS or GAPDH gene expression assay probe, and 1 1 of cDNA. The reactions were run for 2 min at 50°C, 10 min at 95°C, followed by 40 cycles of 15 s at 95°C and 1 min at 60°C. The average cycle threshold (C t ) value from each RGS probe was applied to calculate specific gene expression and normalized by the human GAPDH C t value. Results represent values from six individual donors assayed in duplicate.
Lentiviral Amplification and Transduction-The human RGS5 coding sequence was amplified by PCR using pEGFP-RGS5 plasmid as a template and subcloned into pLenti7.3/V5 TOPO (Invitrogen). This construct or pLenti7.3/V5 TOPO-LacZ (negative control) was transfected into HEK293T cells together with a lentiviral packaging mix (Sigma) using Lipofectamine 2000. After 72 h virus was harvested in the supernatant and concentrated 10-fold by centrifugation. Concentrated virus was added directly to HuASM cells, and gene expression and function were assayed 48 -72 h after transduction.
Flow Cytometry-HuASM cells were suspended in FACS buffer (PBS plus 0.2% BSA) and incubated for 30 min at 4°C with Alexa Fluor 647-conjugated anti-thrombin receptor antibody (Santa Cruz Biotechnology sc-13503) or mouse IgG1 as a control. Cells were washed several times followed by analysis on a FACSCalibur flow cytometer (BD Biosciences).
Subcellular Fractionation-Cells lysates were prepared in TE buffer and homogenized by passage through a 25-gauge ␤-Agonist Regulation of Bronchial RGS5 Expression APRIL 1, 2011 • VOLUME 286 • NUMBER 13 syringe. Cells were centrifuged at 500 ϫ g to obtain postnuclear supernatants, followed by ultracentrifugation at 100,000 ϫ g to separate membrane and cytosolic fractions.
Measurement of Intracellular Ca 2ϩ Mobilization-HuASM cells (2.5 ϫ 10 4 cells/well) were seeded in Biocoat poly-lysine 96-well plates (BD Biosciences), allowed to attach overnight and serum-starved in Ham's F-12 medium plus 0.5% BSA for 48 h prior to measurement of intracellular Ca 2ϩ concentration. Ca 2ϩ indicator dye (FLIPR Calcium 3 Assay Component A (Molecular Devices)) was added to each well followed by incubation for 1 h at 37°C with 5% CO 2 . Serial dilutions of agonists (bradykinin, thrombin, endothelin-1, and histamine (Sigma-Aldrich)) were prepared in a separate plate at 10ϫ concentrations for robotic addition to cells by the FLEX Station II fluorimeter (Molecular Devices). Relative fluorescence units (RFUs) were measured every 2 s for a period of 180 s after agonist addition. Results were plotted as the maximum-minimum signal for each agonist concentration over this time period. Graphs represent results (mean Ϯ S.E.) of three or more independent experiments using cells derived from separate individual donors.
Immunofluorescence-HuASM cells were plated into chamber wells at 0.1 ϫ 10 6 /well overnight to allow cell adherence, followed by fixation in neutral buffered formalin. Fixed cells were permeabilized briefly in PBS containing 0.5% Triton X-100 followed by a blocking step in Ready-to-Use blocking buffer (DakoCytomation). RGS5 was detected with anti-RGS5 primary antibody (Prosci, 1:250) followed by incubation with AlexaFluor 594-conjugated goat anti-chicken antibody (Invitrogen). Cells were washed with PBS/0.5% Triton X-100 and fluorescence visualized with a Leica DMI 4000B confocal microscope.
Immunohistochemistry-Frozen human lung sections were fixed in cold acetone and blocked with hydrogen peroxidase followed by incubation with anti-RGS5 primary antibody (ProSci, 1:750) overnight. Slides were then washed extensively with Tris-buffered saline plus 0.05% Tween followed by incu-bation with biotinylated anti-chicken IgY (Abcam ab6876, 1:500) and horseradish peroxidase-conjugated streptavidin (DakoCytomation, 1:400) for 30 min. each. Staining was detected with 3, 3Ј-diaminobenzidine (DAB) followed by hematoxylin counterstaining. Slides were dehydrated in graded ethanol and cleared in xylene prior to mounting with coverslips. Image shown is representative of lung sections derived from 8 individual donors.
Statistical Analysis-Data were analyzed by one-or two-way ANOVA or Student's t test using PRISM software (GraphPad, CA). We considered a p value Ͻ 0.05 statistically significant.

RESULTS
␤ 2 -Agonist Exposure Augments Agonist-evoked Ca 2ϩ Responses of ASM-To address the molecular mechanisms underlying enhanced contraction of bronchi after chronic ␤ 2 -agonist exposure, we evaluated excitation-contraction coupling pathways in HuASM cells treated for 48 h with the high affinity ␤-agonist isoproterenol. We loaded cells with Ca 2ϩsensing fluorescent dye and measured accumulation of intracellular Ca 2ϩ after stimulation with a range of concentrations of the GPCR ligands bradykinin, thrombin, and histamine, which activate the G␣ q signaling pathway (27). Treatment of ASM cells with these compounds evoked a transient increase in cytosolic calcium lasting Ͼ120 s, with a peak increase at ϳ10 -20 s, suggesting an IP3-mediated release of Ca 2ϩ from intracellular stores (Fig. 1, A-C). Although baseline Ca 2ϩ levels were similar in untreated and isoproterenol-treated cells, peak intracellular Ca 2ϩ concentrations in response to bradykinin and thrombin were significantly higher in cells incubated with isoproterenol than in control cells in response to several different concentrations of agonist (Fig. 1, A-C and Table 1). Although we observed an increased histamine-evoked E max in isoproterenol-treated cells, it did not achieve statistical significance, which we attribute to reduced membrane expression of H1 receptors (see Fig. 2C). We also noted that basal phosphorylation of MLC (i.e. without re-stimulation) was higher in cells exposed to isoproterenol than in untreated cells (Fig. 1D). These results support the hypothesis that procontractile signaling pathways are amplified in ␤ 2 -agonist-experienced ASM cells.
To determine whether up-regulation of GPCRs by ␤ 2 -agonist could account for sensitization of Ca 2ϩ responses, we evaluated receptor expression in the presence or absence of isoproterenol by immunoblotting. Drug treatment had no effect on the total cellular expression of thrombin (PAR1), bradykinin (B2), or histamine (H1) receptors ( Fig. 2A). Surface expression of PAR1 was also similar in untreated and isoproterenol-treated cells by flow cytometry (Fig. 2B). As we were unable to assess surface levels of B2 or H1 receptors by flow due to lack of available reagents, we performed cell fractionation and immunoblotting. The subcellular distribution of PAR1 and B2 receptors was similar in untreated and isoproterenol-treated cells (Fig.  2C). In contrast, membrane levels of H1 receptor decreased somewhat after isoproterenol treatment. We also determined whether isoproterenol exposure affected expression of intracellular signaling components downstream of GPCRs. The levels of G␣ q , PLC␤1, and MLC were similar in isoproterenol-treated ␤-Agonist Regulation of Bronchial RGS5 Expression and untreated HuASM cells ( Fig. 2A). Thus, exposure of HuASM cells to ␤ 2 -agonist does not substantially influence expression of key signaling components yet induces amplified Ca 2ϩ responses to pro-contractile agonists.
Expression of RGS Proteins in ASM-We noted that isoproterenol pretreatment primarily altered the efficacy of the agonists to induce intracellular Ca 2ϩ flux (i.e. maximal signal at high doses, E max ) rather than their potency (EC 50 ) ( Fig. 1 and   Table 1). We observed a similar phenotype in mast cells depleted of RGS13 (28). We hypothesized that ␤ 2 -agonists could lead to changes in expression of RGS proteins, rather than components that propagate the G q signaling pathway, given that the abundance and activity of RGSs is often more dynamic depending on cell type and microenvironment (29 -30). We focused on RGS proteins of the R4/B subfamily, which includes RGS1-5, 8, 10, 13, 16, 18, and 21, because their function in regulating GPCR signaling in peripheral tissues is the most well-defined (31). We isolated total RNA from cultured ASM cells derived from healthy donors and evaluated RGS expression by RT-PCR and real-time quantitative PCR. We found expression of RGS2, 3, 4, 5, and 8 in these cells while RGS1, 10, 13, and 16 were absent or minimally detected (Fig. 3, A and B). Although the relative expression of each RGS could not be compared strictly by this method given the varying affinity of various primer probe sets, RGS2-5 were detected readily, and RGS3 appeared to be most abundant. Because RGS2 and 3 are widely expressed (32-34), we studied RGS5 further since its expression is restricted to only a few cell types in humans (33,35). We detected RGS5 protein in HuASM by immunoblotting (Figs. 3D and 4C). RGS5 localized predominantly in the cytosolic fraction by immunofluorescent staining (Fig. 3C) or by subcellular fractionation and immunoblotting (Fig. 3D). The RGS5 antibody appeared to be specific as no signal was detected on immunoblots probed with antibody pre-incubated with recombinant RGS5 (supplemental Fig. S1). We also identified RGS5 in bronchial smooth muscle and epithelial cells of human lung by immunohistochemical staining (Fig. 3E).

Changes in RGS Expression
Evoked by ␤ 2 -Agonist Treatment of ASM-We investigated the effect of ␤ 2 -agonists on the expression of RGS5 in HuASM cells. We treated cells with iso-FIGURE 2. Effect of isoproterenol exposure on expression of receptors and signaling proteins in HuASM cells. A, whole cell lysates were prepared from untreated and isoproterenol-treated cells followed by immunoblot analysis with the indicated antibodies. B, PAR1 levels were quantified in untreated cells or cells exposed to isoproterenol for 24 h by flow cytometry. C, membrane and cytosolic fractions were prepared from cells treated with medium alone or isoproterenol followed by immunoblot analysis. Results are from a single experiment representative of 2-3 independent experiments with similar results.

FIGURE 3. Expression of RGS proteins in cultured HuASM cells and human lung.
A, RNA was extracted from primary HuASM cells derived from a single donor followed by cDNA synthesis using reverse transcriptase. RGS mRNAs were detected by polymerase chain reaction using gene-specific primers. B, RNA was extracted from cultured human ASM cells (five independent donors), followed by cDNA synthesis and quantification of mRNA abundance. Results are presented as the mean Ϯ S.E. of each RGS mRNA relative to RGS5, set as 1 for each donor. C, immunofluorescent staining of HuASM cells with RGS5-specific antibody and fluorophore-conjugated anti-chicken IgY (left panel) or secondary antibody alone (right panel). 4Ј-6-Diamidino-2-phenylindole (DAPI) was used to identify nuclei. D, membrane and cytosolic fractions were prepared from untreated and isoproterenol-treated cells followed by immunoblot analysis with the indicated antibodies. E, human lung sections were stained with hematoxylin and eosin (left panel), anti-RGS5 antibody (middle panel), or biotinylated anti-chicken IgY alone (right panel) followed by incubation with horseradish peroxidase-conjugated streptavidin and detection with the HRP substrate DAB (brown staining). Antibody-stained sections were counterstained with hematoxylin. RGS5 is expressed in the smooth muscle layer of the airway (black arrows) and epithelium (white arrows).

␤-Agonist Regulation of Bronchial RGS5 Expression
proterenol followed by isolation of total RNA and quantification of gene expression by real-time RT-PCR. RGS5 expression decreased following isoproterenol exposure in a concentrationdependent manner (Fig. 4A). After 4 h of incubation with ␤ 2 -agonist, RGS5 mRNA levels decreased by ϳ40%, and declined further to 25-40% of initial levels at 24 -36 h (Fig. 4B). Isoproterenol treatment also led to reduced expression of RGS4 (supplemental Fig. S2) but not RGS2 and RGS3, in ASM cells (supplemental Fig. S3). The expression of RGS5 protein was also diminished ϳ50% 48 h following addition of isoproterenol (Fig.  4C). The partial ␤ 2 -agonist salbutamol, which is more commonly used clinically for the treatment of asthma, also induced a significant reduction in RGS5 expression (Fig. 4D).
To determine if the changes in RGS5 expression induced by ␤ 2 -agonist treatment were responsive to corticosteroids, we treated cells with salbutamol and dexamethasone. We used a dexamethasone concentration previously shown to prevent ␤ 2 -agonist desensitization in PCLS without affecting growth factor-induced ASM proliferation (11,36). RGS5 expression was reduced in salbutamol-treated cells compared with control ( Fig. 4D). However, pretreatment with dexamethasone did not affect RGS5 expression in the presence or absence of salbutamol (Fig. 4E). Thus, since RGS proteins are known to negatively regulate output of G i -and G q -coupled GPCRs, their reduced expression in HuASM exposed to ␤ 2 -agonist could account for the enhanced signaling we observed in these cells after GPCR stimulation.
RGS5 Knockdown Enhances Agonist-induced Intracellular Calcium Flux and Protein Phosphorylation-To evaluate the consequences of reduced RGS5 levels on pro-contractile signaling pathways in HuASM, we used RNA interference to extinguish RGS5 expression. We transfected cells with a scrambled siRNA or two distinct RGS5-specific siRNAs and measured RGS5 mRNA abundance by real-time PCR. Compared with cells expressing the control siRNA, cells transfected with either RGS5 siRNA had 80% less RGS5 mRNA (Fig. 5A). Similar to isoproterenol treatment, transfection of RGS5 siRNA led to substantially reduced RGS5 protein expression compared with control (Fig. 5B). We evaluated the rise in intracellular calcium concentration after stimulation with GPCR agonists in control

␤-Agonist Regulation of Bronchial RGS5 Expression
and RGS5-depleted cells. Compared with control cells, cells expressing RGS5 siRNA had a significantly higher cytosolic Ca 2ϩ concentration in response to thrombin, bradykinin, and histamine (Fig. 5, C-F and Table 2). In contrast, the Ca 2ϩ response to endothelin-1 (which also activates a Gq-linked GPCR) or the Ca 2ϩ ionophore ionomycin was similar in control  1 way ANOVA). B, RGS5 protein amounts in cells exposed to isoproterenol or transfected with RGS5 siRNA determined by immunoblotting. RGS5 amounts (normalized to ␤-actin) relative to control (non-treated (NT) or CTL siRNA, set as 1) are indicated numerically below the blot. C-H, siRNA-transfected cells were loaded with Ca 2ϩsensing dye followed by determination of intracellular Ca 2ϩ concentrations after stimulation with the indicated agonists. Each figure represents mean Ϯ S.E. of 3 or more independent experiments using cells derived from separate donors (C-D): *, p Ͻ 0.05, t test, scrambled v. RGS5 siRNAs.

␤-Agonist Regulation of Bronchial RGS5 Expression
and RGS5 knockdown cells (Fig. 5, G and H). These studies show that reduced RGS5 expression selectively enhances Ca 2ϩ signaling induced by activation of thrombin, histamine, and bradykinin receptors but has no effect on endothelin-1 or receptor-independent responses. Thus, RGS5 negatively regulates G␣ q -linked signaling pathways in HuASM.
To evaluate the effect of RGS5 siRNA on the excitation-contraction pathway downstream of Ca 2ϩ , we examined protein phosphorylation. We treated cells with bradykinin for up to 15 min and assessed MLC phosphorylation by immunoblotting cell lysates. Compared with cells expressing the scrambled siRNA, cells with extinguished RGS5 expression had more basal and bradykinin-evoked MLC phosphorylation (Fig. 6A). The increased basal protein phosphorylation in RGS5-depleted cells is concordant with previous studies (28,37). Consistent with the function of RGS5 as an upstream negative regulator at the level of G protein activation, phosphorylation of other GPCR-activated kinases (Erk and Akt) was also increased in cells expressing RGS5 siRNA compared with control (Fig. 6, B and C).
RGS5 Overexpression Inhibits Intracellular Calcium Flux and MLC Phosphorylation-To examine the effects of increased RGS5 levels on procontractile signaling in HuASM, we measured agonist-evoked Ca 2ϩ flux in cells infected with lentiviruses encoding either a control protein (␤-galactosidase) or RGS5. Overexpression of RGS5 in HuASM blunted bradykinin-induced increases in cytosolic Ca 2ϩ and MLC phosphorylation compared with control cells (Fig. 7, A and B). To address the role of RGS5 in the enhanced excitation-contraction coupling induced by isoproterenol, we expressed ␤-gal or RGS5 in HuASM cells and treated them with medium alone or isoproterenol for 48 h prior to measurement of bradykininevoked Ca 2ϩ flux. Cells overexpressing RGS5 had reduced intracellular Ca 2ϩ concentrations after bradykinin stimulation compared with control in the presence or absence of isoproterenol pretreatment (Fig. 7C). Thus, RGS5-overexpressing cells are resistant to isoproterenol-induced hypersensitivity. Collectively, these studies indicate that RGS5 inhibits G q -linked procontractile pathways in HuASM. Reduced RGS5 expression likely contributes to the augmented signaling observed in cells chronically exposed to isoproterenol.
Increased Bronchial Contractility of Rgs5 Ϫ/Ϫ Mice-To investigate how RGS5 regulates AHR, we compared carbacholevoked contraction of PCLS from WT and Rgs5 Ϫ/Ϫ mice. We detected Rgs5 transcripts in whole lung from WT but not KO mice (Fig. 8A). Bronchial slices from RGS5-deficient mice contracted more than WT airways to equivalent concentrations of carbachol (Fig. 8B). Similar to the Ca 2ϩ responses of primary human ASM cells expressing RGS5-specific siRNA, RGS5 deficiency in mice was associated with increased maximal bronchial contraction (E max : WT ϭ 70.8 Ϯ 5.7%; KO ϭ 93.5 Ϯ 2.9% (p ϭ 0.001)) while the EC 50 of carbachol was unaffected (WT ϭ

DISCUSSION
This is the first study to define a function for RGS proteins in bronchial smooth muscle. We demonstrated expression of several RGS proteins in primary HuASM cells and showed that exposure to ␤ 2 -adrenergic agonists lead to reduced expression of both RGS4 and RGS5. Depletion of RGS5 in these cells either by prolonged exposure to isoproterenol or by RNA interference sensitized excitation-contraction pathways, as evidenced by enhanced intracellular Ca 2ϩ flux and MLC phosphorylation. Thus, ␤ 2 -agonists may inadvertently "prime" signaling routes in HuASM by depleting them of negative regulatory components such as RGS proteins, whose activity controls contraction of naïve cells under physiological conditions. Evidence suggests that chronic exposure to ␤-agonists alters ASM contraction and gene expression (21, 38 -39). Liggett and colleagues found significant changes in expression of Ͼ500 genes in mice overexpressing ␤-adrenergic receptors. As but one example, ASM cells from these mice had reduced expression of phospholamban, a protein that inhibits activity of the sarco-(endo)plasmic reticulum Ca 2ϩ ATPase2 (SERCA2), which promotes reuptake of cytosolic Ca 2ϩ into the ER (38). Mice deficient in phospholamban had reduced airway contraction to methacholine. Decreased phospholamban abundance was surprising given the enhanced contraction of tracheas from these mice (22), indicating that ␤-agonists may influence the transcriptional program and contractile properties of ASM in complex ways.

␤-Agonist Regulation of Bronchial RGS5 Expression
Published work has demonstrated changes in receptor expression in intact bronchial tissue after prolonged ␤-agonist exposure. Two separate studies showed up-regulation of either histamine H1 (19) or endothelin-1 (18) receptors in human or bovine tracheobronchial tissues after exposure to fenoterol, which occurred by both transcriptional and post-transcriptional mechanisms and correlated with enhanced contraction of tracheas to the cognate ligand. The absence of up-regulated receptors after ␤-agonist treatment in the current study might be attributed to the fact our ASM cells were cultured in the absence of other cell types in vitro. In fact, the principal increases in ET receptor expression in the prior study occurred in epithelial cells rather than ASM.
In addition to their possible effects on receptor levels, ␤-agonists may also modulate expression and/or activity of downstream signaling components. Others demonstrated increased Ca 2ϩ levels in response to methacholine in cells exposed to albuterol (21). This phenotype was attributed to increased expression of G␣ i1 and increased total G protein activity (GTP␥ S binding) induced by the inactive (S)-albuterol present in the racemic (RS)-albuterol mixture, which is used clinically. However, because M3 muscarinic receptors (M3R) are coupled to G␣ q , it is unclear whether the increased Ca 2ϩ flux was due to a direct effect of (S)-albuterol on the procontractile signaling pathway. In fact, McGraw et al. (40) found that transgenic overexpression of G␣ i2 in mice reduced airway contraction to methacholine. We observed that although incubation of ASM with the high affinity agonist isoproterenol increased Ca 2ϩ flux evoked by several agonists, it did not lead to substantial changes in G protein (G␣ i , G␤␥, G␣ q ) or effector (PLC␤) expression ( Fig. 2 and data not shown). These differences might be explained simply by the different agonists used.
In another series of reports, Grunstein and co-workers showed that chronic exposure of ASM to salmeterol induced transcriptional up-regulation of phosphodiesterase 4D5, which promotes cAMP degradation, through a G␣ i -Erk-MKP1 dependent mechanism (20,(41)(42). Although procontractile signaling elements such as IP3 and Ca 2ϩ flux in ASM cells exposed to salmeterol were not examined in these studies, rabbit tracheas treated with salmeterol contracted more to acetylcholine, which was reversed by either glucocorticoid or phosphodiesterase inhibitor pretreatment. Although it has been shown recently that cAMP may promote Ca 2ϩ signaling by directly sensitizing IP3 receptors to IP3 independently of PKA (43), it is unclear from these studies whether PDE4 up-regulation alterd the procontractile signaling pathway in ASM cells directly.
In contrast to these studies, we provide evidence that exposure to ␤ agonists directly affects propagation of G q -mediated procontractile signaling in cultured ASM cells by decreasing expression of key negative regulators of the pathway. The reduced expression of RGS4 and RGS5 in isoproterenol-treated ASM is concordant with recent studies of Penn and co-workers (39), who performed microarray studies on HuASM cells expressing GFP or cells expressing a heat-stabile inhibitor of PKA kinase activity, PKI, and cultured in the presence of epidermal growth factor and IL-1␤. Cells expressing PKI had 14-fold more RGS4 and 4-fold more RGS5 than control cells, suggesting negative regulation of RGS4/5 transcription by PKA. However, while we found that ␤-agonist-induced changes in RGS4/5 expression in HuASM were unaffected by steroid pretreatment, in the previous study glucocorticoid pretreatment of the cells before stimulation with growth factors induced a marked down-regulation of RGS4 and RGS5 (39). In addition, stimulation of cardiomyocytes with high doses of isoproterenol did not affect RGS4 or 5 expression in a recent report (44). Although these studies collectively suggest involvement of the cAMP-PKA axis in transcriptional regulation of RGS4/5 in ASM, other cell-type specific factors could be involved. Thus, further dissection of pathways involved in regulation of RGS expression is needed.
Loss of RGS GAP activity enhances downstream output in response to GPCR stimulation (28), which we observed in RGS5-depleted ASM cells. However, endothelin-1-induced Ca 2ϩ responses were unaffected by the loss of RGS5, suggesting that it does not regulate signaling through ET-B receptors in ASM. Receptor-selective regulation of GPCR signaling by RGS proteins has been observed, which may be a result of direction interactions with GPCRs or with linkers or activity-modifying proteins such as spinophilin (45). Although such interactions have not been extensively characterized for RGS5, our results suggest that it specifically regulates a subset of GPCR responses in ASM.
RGS4 and RGS5 are closely related isoforms, which display similar G␣ GAP activity (46 -47). RGS4 has been shown to inhibit bradykinin-and thrombin-induced signaling in other cell types (48 -49). Although we did not formally FIGURE 8. Increased bronchial reactivity of RGS5-deficient mice. A, Rgs5 expression in whole lung. RNA was extracted from lungs of WT or KO mice, followed by analysis of relative Rgs5 expression by quantitative PCR (mean Ϯ S.E. of 3 mice of each genotype measured in duplicate). B, precision-cut lung slices were prepared as described under "Experimental Procedures." Airways were contracted to various concentrations of carbachol (Cch) followed by measurement of airway luminal diameter by microscopy. Graph represents the mean Ϯ S.E. of 6 WT and 7 KO mice analyzed in three independent experiments. APRIL 1, 2011 • VOLUME 286 • NUMBER 13 assess the role of RGS4 in regulating ASM signaling in this study using siRNA, we hypothesize that the reduced expression of both RGS4 and RGS5 induced by isoproterenol leads to a cumulative loss in GAP activity. This would be predicted to enhance GPCR signaling responses, which have been demonstrated in cells from knock-in mice expressing a mutant G␣ i2 resistant to all RGS GAP activity (50). The unique, non-redundant function of RGS5 in the regulation of AHR is confirmed by the increased bronchial contraction of PCLS from RGS5-deficient mice compared with WT. It is also possible that GCPR/G␣-independent signaling mechanisms could contribute to increased excitation-contraction coupling in ASM cells exposed to isoproterenol.

␤-Agonist Regulation of Bronchial RGS5 Expression
Studies of RGS4/5 knock-out mice in models of allergeninduced pulmonary inflammation are needed to fully assess their role in regulating AHR, and these are underway. However, interpretation of such experiments is likely to be challenging given expression of RGS4 and -5 in multiple cell types in lung including bronchial smooth muscle and epithelial cells (our study and (51)), vascular smooth muscle (52), and leukocytes (53). Leukocyte recruitment and activation induced by allergen exposure has an integral function in generating inflammation in the asthmatic lung. Given the capacity of RGS proteins to regulate cell migration (54), RGS4/5 deficiency may also affect the inflammatory component of asthma. Nonetheless, our studies indicate that the downregulation of both RGS4 and RGS5 and resultant hyper-responsiveness of ASM after chronic ␤ 2 agonist exposure should be considered when evaluating the appropriateness of these agents for the treatment of asthma.