cAMP Regulation of Airway Smooth Muscle Function

Abstract

Agonists activating β₂-adrenoceptors (β₂ARs) on airway smooth muscle (ASM) are the drug of choice for rescue from acute bronchoconstriction in patients with both asthma and chronic obstructive pulmonary disease (COPD). Moreover, the use of long-acting β-agonists combined with inhaled corticosteroids constitutes an important maintenance therapy for these diseases. β-Agonists are effective bronchodilators due primarily to their ability to antagonize ASM contraction. The presumed cellular mechanism of action involves the generation of intracellular cAMP, which in turn can activate the effector molecules cAMP-dependent protein kinase (PKA) and Epac. Other agents such as prostaglandin E₂ and phosphodiesterase inhibitors that also increase intracellular cAMP levels in ASM, can also antagonize ASM contraction, and inhibit other ASM functions including proliferation and migration. Therefore, β₂ARs and cAMP are key players in combating the pathophysiology of airway narrowing and remodeling. However, limitations of β-agonist therapy due to drug tachyphylaxis related to β₂AR desensitization, and recent findings regarding the manner in which β₂ARs and cAMP signal, have raised new and interesting questions about these well-studied molecules. In this review we discuss current concepts regarding β₂ARs and cAMP in the regulation of ASM cell functions and their therapeutic roles in asthma and COPD.

INTRODUCTION

3′-5′-Cyclic adenosine monophosphate (cAMP) is the archetypal second messenger whose specific biochemical and signaling properties in numerous cell types have been intensively researched for more than half a century. The resulting wealth of data clearly highlights the importance of cAMP in diverse physiological and pathophysiological processes via its modulation of diverse cellular events. cAMP plays a key role in the functions of many airway cells including controlling ciliary beat frequency (critical for mucus clearance) in airway epithelial cells [1] and suppressing the pro-inflammatory activity of various immune and inflammatory cells. However it is arguably the inhibitory effect of increased cytosolic cAMP levels on the contraction of airway smooth muscle (ASM) which constitutes the most profound cAMP-mediated effect in the lung.

As discussed in greater detail below, cAMP-mediated pathways are most commonly initiated following the binding of specific ligands to G protein-coupled receptors (GPCRs) of the G_s family. Perhaps the most exhaustively studied GPCR is the β₂-adrenoceptor (β₂AR) and it is specifically the β₂ARs on ASM cells which, due to their ability to rapidly promote bronchorelaxation, constitute the frontline target for asthma therapy. In addition to driving this clinically critical effect, cAMP also modulates many other aspects of ASM function including proliferation, migration, secretion of inflammatory mediators, and deposition of extracellular matrix (ECM). Beyond these acute effects, cAMP also appears to influence the “phenotype” of ASM cells; i.e., properties relating to their capacity to function as a contractile cell or one more capable of proliferation and secretory/immune-modulatory functions.

Whilst the molecular events occurring between β₂AR activation and bronchorelaxation were thought to be well understood, several recent studies have introduced layers of complexity which have the potential to profoundly alter our understanding of the β₂AR signaling pathway and how asthma is treated clinically.

In this review we aim to explore some of the myriad ways in which cAMP regulates ASM function with an emphasis placed on novel or controversial observations. A brief overview of cAMP regulation will be given encompassing modulation via GPCRs, phosphodiesterases (PDEs), and two direct downstream effectors of cAMP namely protein kinase A (PKA) and exchange proteins directly activated by cAMP (Epac). We shall then consider the ways in which diverse ASM functions are mediated via cAMP and, where possible, address whether these effects are PKA- or Epac-driven. Finally we will consider the therapeutic implications both in terms of future targets and problems with current therapy and ask “What Next?”

1. ROLE OF G_S-COUPLED RECEPTORS AND PDES IN cAMP ELEVATION

cAMP generation in most cells typically occurs through a classical GPCR transmembrane signaling paradigm. A specific subfamily of GPCRs couple to G_α subunits of the G_s subfamily of heterotrimeric G proteins, which in turn activate the enzyme adenylyl cyclase which hydrolyzes ATP to cAMP [2]. In ASM, numerous G_s-coupled receptors have been identified, including the EP2 and EP4 prostanoid (PGE₂) (discussed in detail below), the A2b adenosine, and the IP (PGI₂) receptors [2]. However, the most well-recognized and studied is the β₂AR, and (inhaled) agonists of the β₂AR (β-agonists) are the drug of choice for relief from acute asthmatic attacks, based on their ability to prevent and reverse ASM contraction (bronchoconstriction). Moreover, long-acting β-agonists (LABAs) are now widely used in maintenance therapy for asthma as well as COPD [3]. Although LABA monotherapy is available for treatment of COPD, a combination of LABAs with inhaled corticosteroids (ICS) is recommended for the regular use of LABA in controlling asthma. Despite concerns (discussed below) regarding the loss of therapeutic benefit with prolonged use, and possible mortality effects, β-agonists along with corticosteroids remain at the cornerstone of asthma therapy.

Subsequent to the G_s-coupled receptor-induced accumulation of cAMP, the principal process by which intracellular cAMP levels are reduced is via their hydrolysis – a function performed by a subset of members of the PDE family. The cAMP-elevating role of PDE inhibitors has been exploited for centuries in the treatment of asthma and COPD therapy and, despite possessing a narrow therapeutic window, PDE inhibitors remain in clinical use and a subject of intense interest today [4-6]. PDEs which preferentially hydrolyze the intracellular cyclic nucleotides, cAMP or cGMP, are grouped into 11 families [7]. The presence of PDE1-5 classes has been found in human ASM cells [8, 9]. Among those classes, PDE3 and PDE4 are the two major cAMP-hydrolyzing enzymes [8] and inhibitors of PDE3 and PDE4 cause relaxation of human ASM [9]. Rabe et al. demonstrated that the selective PDE3 inhibitor SKF94120 was more potent in inhibiting contraction of human bronchi than a selective PDE4 inhibitor rolipram [9]. However, the PDE4D subfamily, particularly the PDE4D5 isoform, appears to play a pivotal role in controlling cAMP degradation in human ASM cells [10]. Recently, Trian and colleagues reported that human ASM cells from asthmatics exhibit increased PDE4D expression which is associated with impaired β₂AR-stimulated cAMP production [11]. This work invites consideration as to how the effect of current therapies could be suboptimal when administered against a background of upregulated PDE activity.

2. cAMP EFFECTORS: PKA AND EPAC

Historically, cAMP has been believed to mediate its action on ASM contractile state via activation of the effector PKA. The phosphorylation of numerous targets in ASM by PKA (shown in the context of β₂AR activation in Fig. 1) has been proposed to cause either reduced intracellular Ca²⁺ concentrations ([Ca²⁺]_i) or reduced Ca²⁺ sensitivity in ASM [12-15] with both effects leading to impaired ability to promote myosin light chain (MLC) phosphorylation which enables ASM to contract. In addition, heat shock protein 20 (HSP20) has recently been identified as another potentially important PKA target capable of inhibiting ASM contraction via an ill-defined mechanism that appears independent of MLC regulation [16, 17].

Figure 1. Proposed model of signaling by the β₂ AR, cAMP, PKA, and Epac involved in regulation of airway smooth muscle contraction.

Signaling events elicited by agonist activation of the β₂AR in ASM promote numerous regulatory effects that inhibit pro-contractile signaling in ASM, as well as events that feedback inhibits β₂AR signaling. See text for details.

It is important to note however that although for years now PKA has been assumed to be the critical effector through which cAMP causes ASM relaxation, the role of PKA has never been directly demonstrated, in part due to difficulties in demonstrating that inhibition of PKA inhibits the proposed mechanisms and reverses the effect of cAMP on ASM contraction. Yan et al. demonstrated that inhibitory effects of β-agonists and PGE₂ on cell growth of human ASM cells are PKA-dependent using an effective molecular strategy for inhibiting PKA [18]. Using the same approach, Goncharova et al. showed that PKA was required for the inhibition of human ASM cell migration by albuterol a short-acting β₂-agonist [19]. To date, no studies have provided direct evidence of the role of PKA in mediating the relaxant effect of β-agonists or other cAMP-generation agents on ASM. Indeed, Spicuzza et al. provided intriguing evidence to the contrary, and suggested a cAMP-dependent yet PKA-independent mechanism by which β-agonists antagonize ASM contraction [20]. Of note, cAMP can cross-activate cGMP-dependent protein kinase (PKG) in smooth muscle cells [21] and Torphy et al. demonstrated that cAMP activates both PKG and PKA in canine trachealis [22]. However, the capacity of cAMP-mediated PKG activation to promote bronchodilatation is unclear, given cGMP-mobilizing agents such as nitroglycerin and zaprinast appear to be much less potent in inhibiting ASM contraction than β-agonists [8, 9, 23].

Recently, an alternative mechanism has emerged that challenges the dogma of PKA as the important cAMP effector in ASM. In 1998, Epac, a Rap1 guanine nucleotide exchange factors (GEF), were found as another downstream effector of cAMP action [24]. Epac that can be activated by either PKA or by cAMP in a PKA-independent manner, has been proposed as sufficient to cause ASM relaxation and to inhibit cell proliferation in the absence of PKA activation [25-32]. In an h-TERT ASM cell line, expression of both Epac isoforms, Epac1 and Epac2, was observed, and Epac1 and Epac2 exhibited different cellular localization but cooperated each other [33]. The main Epac effector is GTPase Rap1 and proposed downstream pathways involve mitogen-activated protein kinases, transcriptional factors, and integrin signaling [25, 26, 33]. Roscioni et al. recently demonstrated that Epac activation using an Epac-selective cAMP analog decreased contraction, MLC phosphorylation, and RhoA activities in response to methacholine in human and guinea pig tracheal smooth muscle tissues [31]. Similarly, Zieba et al. demonstrated the sufficiency of an Epac-selective cAMP analogue to relax permeabilized mouse bronchi [32]. Collectively, these studies suggest our long-held assumptions regarding β₂AR and cAMP effector mechanisms in ASM need to be revisited, and set the stage for future work defining the relative importance of PKA versus Epac in mediating the functional effects of β-agonist and cAMP in ASM.

3. MECHANISMS OF β₂AR-MEDIATED ASM RELAXATION

The stimulation of ASM with β-agonists has been shown to inhibit many signaling events that are activated by contractile stimuli to promote ASM contraction. Although the majority of these pro-contractile signaling events ultimately converge at and regulate the phosphorylation status of MLC₂₀, the molecules involved are diverse with the main contributors being G_q-coupled receptors, phospholipase C (PLC), plasma membrane and sarcoplasmic reticulum (SR) ion channels and pumps, RhoA, Ca²⁺-calmodulin and MLC kinase (MLCK). As each of these effectors is inextricably linked to aspects of Ca²⁺ signaling in ASM cells, we will discuss them in terms of the different stages of intracellular Ca²⁺ signaling in which they are pertinent.

3.1. Pro-contractile signaling

Although roles for specific G_i-coupled receptors have recently been implicated in cooperative crosstalk with G_q-coupled receptors [34, 35], it is ligands binding to GPCRs of the G_q-subfamily which are frequently the initiators of contraction in ASM with the most widely recognized contractile G_q-coupled receptors being the M3 muscarinic acetylcholine, histamine H1, bradykinin B2, cysteinyl leukotriene 1, and endothelin-A/B receptors [2]. Following the agonist binding, a conformational change in the receptor occurs which enables coupling to and activation of G_αq, subsequent activation of PLC and production of inositol 1,4,5-trisphosphate (IP₃), which binds to IP₃ receptors on specialized intracellular stores to promote Ca²⁺ release from these stores. This flux combines with Ca²⁺ release from ryanodine-sensitive stores and influx through plasma membrane Ca²⁺ channels to elevate [Ca²⁺]_i to levels that stimulate MLCK, leading to MLC₂₀ phosphorylation, myosin ATPase activity, cross bridge cycling and sarcomere shortening (reviewed in [36, 37]). Moreover, Ca²⁺ “sensitization” mechanisms are also invoked, via RhoA/Rho-kinase, that augment the effect of Ca²⁺ via regulation of MLC₂₀ phosphorylation/dephosphorylation dynamics. Both the mechanisms of Ca²⁺ mobilization and Ca²⁺ sensitization can be regulated by β-agonists/cAMP.

3.2. Inhibition of Gq-coupled receptor transmembrane signaling

PKA can regulate G_q-dependent contractile signaling at multiple junctures. PKA-mediated phosphorylation and inhibition of G_q-coupled receptors, G_αq, or PLC resulting in reductions in agonist-induced phosphoinositide (PI) hydrolysis and Ca²⁺ mobilization has been demonstrated in numerous cell systems. In ASM, β-agonists and other cAMP-generating agents have been shown to inhibit both PI hydrolysis and Ca²⁺ mobilization, yet mechanistic detail and a demonstrated role of PKA or Epac are lacking. Both methacholine and histamine-induced Ca²⁺ mobilization has been observed to be inhibited following exposure to isoproterenol in bovine tracheal smooth muscle [38], whilst in guinea-pig tracheal smooth muscle tissue isoproterenol, forskolin and db-cAMP have been demonstrated to inhibit methacholine-induced contraction via both Ca²⁺ mobilization and Ca²⁺ sensitization [15]. When PI production is specifically assessed, different routes of G_αq-mediated stimulation impart different outcomes with histamine- but not methacholine-induced PI production inhibited by elevators of cAMP [38, 39]. Conversely, in human cultured ASM cells, a small increase in histamine-induced PI hydrolysis was reported following exposure to salmeterol potentially reflecting the somewhat modified ASM phenotype observed following repeated cell culture [40].

Recently, another novel mechanism by which the β₂AR inhibits human ASM contraction mediated by G_q-coupled receptors was proposed by Holden et al [41]. They found that the expression of regulator G-protein signaling 2 (RGS2) which is a member of GTPase-activating proteins (GAPs) and specific for G_q is transcriptionally upregulated by LABAs in human ASM cells. Moreover, RGS2 expression was associated with a decrease of intracellular Ca²⁺ concentrations elicited by G_q-agonists such as histamine and methacholine [41]. These findings demonstrated the involvement of RGS2 in the β₂AR–mediated ASM relaxation of G_q–mediated signaling. Importantly, RGS2 expression was enhanced by LABA treatment, and further enhanced by combined (LABA plus corticosteroid) treatment. These data suggest an important mechanism underlying the efficacy of ICS/LABA combination therapy as a long-term management of asthma and COPD [3].

3.3. Inhibition of Ca²⁺ influx channels by β-agonist and cAMP

It is known that cAMP modulates regulation of intracellular Ca²⁺ homeostasis, one of the principal drivers of ASM responses. The ASM cell has multiple Ca²⁺ influx pathways and channels, including voltage-dependent Ca²⁺ channels, receptor-operated Ca²⁺ entry (ROC), store-operated Ca²⁺ entry (SOC), stretch-activated Ca²⁺ entry, and the reverse mode of Na⁺/Ca²⁺ exchanger [14, 42-47]. Blockade of voltage-dependent Ca²⁺ channels has failed to improve asthma symptoms and inhibit ASM contraction induced by contractile agonists and inflammatory mediators [44, 46, 48]. Therefore, it is now considered that the major Ca²⁺ influx pathways in ASM cells are ROC, possibly TRPC family channels, and SOC which is mediated by the SR Ca²⁺ sensor STIM1 and plasma-membrane Ca²⁺ channel Orai1 [44, 45, 47, 49].

Ay et al. have shown that cAMP modulates these Ca²⁺ entry pathways in ASM cells [50]. For example, induction of SOC by store depletion with sarco/endoplasmic reticulum Ca²⁺-ATPase (SERCA) antagonist, cyclopiazonic acid in the presence of β-agonist or cAMP synthetic analogue confirms inhibition of Ni²⁺- and La³⁺- sensitive Ca²⁺ channels that constitute SOC. This inhibition was attenuated by a pharmacological PKA inhibitor, suggesting a cAMP-PKA mediated effect. Interestingly, the effects mediated by cAMP also appear to activate protein kinase G to indirectly inhibit SOC [50].

3.4. Regulation of Ca²⁺ oscillations by β-agonist and cAMP

Intracellular Ca²⁺ homeostasis involves ion channels and transporters and some of these mediate Ca²⁺ influx or efflux [14, 44, 51]. These events are important in the regulation of ASM contraction and occur in a dynamic process that oscillates between Ca²⁺ entry/release and Ca²⁺ removal in the ASM cell. Bai and Sanderson demonstrated that cAMP-elevating agents forskolin, isoproterenol, and 8-bromo-cAMP decreased contractility, frequency of Ca²⁺ oscillations and methacholine-induced Ca²⁺ release from the SR in murine lung slices [52]. In addition, they observed diminished sensitivity of the IP₃ receptor to IP₃, and this was reversed using caged-IP₃ [52]. This outcome fits with the proposed structure-function of the IP₃ receptor, as the IP₃ receptor has at least two PKA binding sites in the modulatory domain exposed to the cytosolic side [53].

Further exploratory studies to elucidate the effects of cAMP on Ca²⁺ oscillations have also shown that oscillation frequency induced by BAYK8644 via voltage-gated Ca²⁺ channel activation was attenuated by β-agonist in porcine ASM cells [54]. Whilst these studies continue to shed more light on cAMP regulatory effects, similar studies by Janssen et al. using bovine and porcine bronchial smooth muscle found instead that cAMP- or isoproterenol-mediated relaxation remained unaffected in the presence of L-type Ca²⁺ channel blocker, nifedipine [55].

Other studies suggest the effects of β-agonists or cAMP raising agents on Ca²⁺ oscillations and relaxation are in fact occurring through cAMP-dependent activation of large-conductance Ca²⁺-activated K⁺ (BK_Ca) channels. These studies build on original observations from the Kotlikoff lab noting the ability of β-agonist, cAMP, and PKA to regulate gating of BK_Ca channels in ASM [56, 57]. In a recent paper by Zhou and colleagues, BK_Ca currents could be stimulated by PKA provided serine site (S⁶⁹⁵) in the regulator of K⁺ conductance (RCK) domains had not been phosphorylated by protein kinase C [58]. A current kinetics study by Ishikawa and colleagues revealed low dose isoproterenol and forskolin both mediate a modest Ca²⁺ inward current, however a synthetic cAMP analogue elicited the opposite effect [59]. This finding implies adenylyl cyclase-dependent cAMP synthesis may regulate Ca²⁺ homeostasis at low doses through small Ca²⁺ inward currents via L-type Ca²⁺ channels, but at high doses attenuates the amplitude of inward Ca²⁺ currents. However, since voltage-dependent Ca²⁺ channel inhibitors are not effective for asthma therapy or bronchodilation, the contribution of BK_Ca and L-type Ca²⁺ channels to β₂AR-mediated ASM relaxation is likely minimal.

3.5. Modulation of Ca²⁺ removal system by β-agonist and cAMP

Ca²⁺ removal from the cytosol constitutes a major event during Ca²⁺ homeostasis. Recent findings indicate cAMP raising agents or β-agonist also modulate Ca²⁺ removal. The SERCA transporter sequesters Ca²⁺ back into SR stores after Ca²⁺ release, and this event is predominantly responsible for the removal of intracellular Ca²⁺. cAMP-elevating including β-agonists indirectly modulate Ca²⁺ homeostasis through phospholamban, which inhibits SERCA activity despite phospholamban expression being relatively low in ASM cells are [60]. Notably, a previous work reported ASM cells from patients with asthma to already have decreased levels of SERCA2, the predominant isoform in ASM cells [61]. Unpublished reports by Ojo et al. suggest that chronic stimulation of ASM cells with short and long acting β-agonists reduces SERCA2 protein expression [62]. Similar effects of LABAs have also been reported on SERCA2 expression in cardiac myocytes [63]. Collectively, these studies suggest that SERCA may play an important role in ASM Ca²⁺ homeostasis, and SERCA expression or activity may be regulated by β-agonist therapy.

3.6. Reduced Ca²⁺ sensitivity by β-agonist and cAMP

ASM contraction is also regulated by modulation of Ca²⁺ sensitivity or Ca²⁺-independent mechanisms [64]. Contraction of ASM occurs in a phasic manner followed by a tonic contractile response. The phasic response is dependent on Ca²⁺ release from the SR, and the tonic or sustained response relies on Ca²⁺ entry as well as Ca²⁺ sensitivity via inactivation of MLC phosphatase (MLCP) by RhoA/Rho-kinase [48, 65-68]. Several studies have shown that cAMP-elevating agents reduce ASM Ca²⁺ sensitivity. Oguma et al. simultaneously measured intracellular Ca²⁺ concentrations and isometric tension using guinea pig tracheal smooth muscle tissue strips and showed that relaxation induced by isoproterenol or forskolin was indeed not fully Ca²⁺ dependent [15]. This observation is in agreement with similar work by Janssen and colleagues, showing isoproterenol-induced MLCP activity [55]. Although MLCP activity was not investigated by Ise et al., the differential response of theophylline further supports the regulation of Ca²⁺ sensitivity by cAMP-elevating agents. Using porcine ASM, they showed an attenuated tension but minimal change to intracellular Ca²⁺ transient for high KCl evoked contractions. Yet in carbachol treated ASM strips both tension and Ca²⁺ transients were attenuated in the absence of extracellular Ca²⁺ and store release [69]. Janssen et al. indicated these effects only applied to non-human tissue and were not reproducible in human ASM cells [55]. However, investigations using α-toxin permeabilized rabbit and porcine tracheal smooth muscle tissue strips showing attenuation of Ca²⁺-induced contractions in the presence of theophylline and/or cAMP analogue further supports a role for cAMP or PKA in Ca²⁺ sensitivity [12, 69]. Although MLCP phosphorylation by Rho-kinase is the main mechanism of Ca²⁺ sensitivity in ASM contraction, whether the RhoA/Rho-kinase inactivation is involved in the mechanisms of reduced Ca²⁺ sensitivity by cAMP is controversial [12, 15, 31].

Protein kinase C (PKC)-dependent phosphorylation of the 17-kD myosin phosphatase inhibitor protein (CPI-17) [70] is another pathway reported to lead to increased Ca²⁺ sensitivity in ASM contraction [71]. Morin et al. demonstrated that siRNA-induced knockdown of CPI-17 resulted in an inhibition of contraction of human bronchial smooth muscle by decreasing Ca²⁺ sensitivity [71].

Another possible mechanism of reduced sensitivity to Ca²⁺ by β-agonists is myosin-independent effects on actin polymerization. Actin dynamics are essential for maintaining contractile force of smooth muscle. Previous studies have implicated the cAMP-PKA pathway in inhibiting the actin cytoskeleton in ASM cells [17, 72]. Komalavias et al. showed that PKA-induced HSP20 phosphorylation decreased phosphorylation of cofilin and disrupted actin polymerization, leading to ASM relaxation [17].

4. PGE₂-MEDIATED MODULATION OF ASM FUNCTION

Prostaglandins are mediators derived from arachidonic acid by cyclooxygenase activation [73] with PGE₂ being the most abundant and, arguably the most physiologically relevant in the lung [74, 75]. In ASM cells PGE₂-mediates anti-proliferative, anti-migratory and pro-relaxant effects mainly via cAMP-dependent pathways in a similar fashion to other cAMP-increasing agents [18, 26, 27, 76-85].

The biological effects of PGE₂ are mediated via four distinct prostanoid EP receptor subtypes, EP1, EP2, EP3, and EP4 receptors, from the GPCR superfamily [73, 86]. Recent studies have explored the concept of EP receptor targeting as a means of improving the therapeutic potential of E prostanoids and the expression of mRNA for EP2, EP3, and EP4 receptors has been identified in human ASM cells [80, 82, 87]. Burgess et al. demonstrated that ASM cells isolated from patients with asthma had more expression of EP3 and EP2 receptors than those from control subjects [82]. Interestingly, the anti-mitogenic effects of PGE₂ were enhanced in patients with asthma compared with control subjects [82].

There are no previous reports demonstrating functional expression of EP1 receptor in human ASM cells [18, 80, 88]. Indeed, either EP1 receptor agonist SC19220 or ONO-DI-004, did not affect proliferation of human ASM cells [18, 80]. In contrast, EP1 receptor activation increases ASM tone in guinea pigs [89] and reduces the broncho-dilatory function of β-agonists receptor in mouse ASM cells [90]. Thus the role of the EP1 receptor in regulating ASM cell functions appear to be species-specific [87, 89-92].

Stimulation of G_s-coupled EP2 and EP4 receptors increases intracellular cAMP levels via adenylyl cyclase activation, whereas EP3 receptor inhibits adenylyl cyclase [86]. The EP2 receptor has been reported as the functionally dominant EP receptor isoform in the mechanisms of relaxation by PGE₂ in human ASM [91]. However, a role for the EP4 receptor in ASM relaxation has also been reported. Data from Buckley et al. suggest a component of PGE₂-mediated relaxation of human and rat airways is mediated by the EP4 receptor, while the EP2 receptor-selective agonists AH13205 failed to relax contracted human ASM tissue [93]. In addition, Mori et al. using subtype-selective agonists/antagonists suggested that both EP2- and EP4- receptors appear important in mediating the anti-proliferative effects of PGE₂ on fetal bovine serum (FBS)-induced human ASM cell growth [80]. Whilst EP3-mediated signals were observed to induce increased [Ca²⁺]_i concentrations, the specific EP3 receptor agonist ONO-AE-248 did not affect ASM cell growth. Yan et al. also reported EP2-mediated PGE₂ signaling to be anti-proliferative, whereas the EP3-selective agonist sulprostone was pro-mitogenic in human ASM cells [18].

Previous findings highlight the potential therapeutically advantageous effects of EP2 and EP4 agonists on airway remodeling and add to the intriguing literature examining EP receptor subtype function in ASM [18, 88, 93-96]. Although phase 1 clinical trials previously conducted on the EP2 agonist AH13205 [97] were disappointing as this compound did not elicit bronchorelaxation and exhibited side effects including cough and airway irritation, research on prostanoid EP receptors specifically EP2 and EP4 receptors may still lead to novel therapeutic strategies for the treatment of asthma and COPD.

5. cAMP-MEDIATED MODULATION OF ASM SYNTHETIC AND MITOGENIC FUNCTIONS

5.1. Proliferation and migration

Elevators of cAMP including β-agonists, PDE inhibitors, forskolin, and PGE₂ have been observed to inhibit ASM cell proliferation and migration [76, 83, 84, 98-103]. This is of particular clinical relevance, especially in chronic asthma where an increase in ASM mass is a key feature of airway remodeling [104-107]. This increase in ASM mass can be attributed to altered proliferation, migration, or both. Anti-mitogenic effects of cAMP involve multiple mechanisms, including inhibition of ERK1/2 and phosphoinositide 3′-kinase (PI3K), via PKA activation and Epac in ASM cells [27, 100, 105, 108, 109] (also see Section 2 in this review). Roscioni et al. demonstrated that activation of PKA and Epac prevents PDGF-induced ASM cell proliferation and cell cycle progression [108]. It has been shown that IL-1β and TNF-α inhibit mitogen-stimulated ASM cell proliferation although they have mitogenic effects by themselves [110, 111]. Following IL-1β and TNF-α receptor stimulation, activation of cAMP/PKA pathway occurs via endogenous cyclooxygenase activation and PGE₂ production, leading to reduced ASM DNA synthesis and cell proliferation [95, 110, 112].

ASM cell migration is enhanced by the activation of specific GPCRs and receptor tyrosine kinases such as PDGF receptors that trigger remodeling of the cytoskeleton and change focal adhesion dynamics [107]. Recently, Goncharova et al. demonstrated that β-agonists and PGE₂ inhibit PDGF-induced human ASM cell migration associated with vasodilator-stimulated phosphoprotein (VASP) phosphorylation via PKA activation [19]. Since VASP belongs to a conserved family of actin-regulatory proteins [113], inhibition of the actin cytoskeleton may be a mechanism by which ASM migration is inhibited by cAMP/PKA. The role and sufficiency of Epac as a cAMP effector in regulating ASM cell migration is uncertain. Glucocorticoids have been observed to enhance the inhibitory effects of β-agonists and PGE₂ on human ASM cell migration [76]. Again, this in vitro finding may identify a cooperative effect underlying the utility of LABA/ICS combination therapy for asthma and COPD.

5.2. Synthesis of cytokines and chemokines

ASM cells have the ability to produce inflammatory cytokines and chemokines. Therefore, the ASM cell not only contracts but also functions as an immune-modulator in the airway. cAMP-elevating agents can modulate these synthetic properties of ASM cells [114, 115]. TNF-α-induced expression of eotaxin and RANTES is inhibited by cAMP-mobilizing agents including β-agonists, PDE inhibitors, and PGE₂ [85, 116, 117]. In contrast, it has been reported that cAMP upregulates the synthesis and release of cytokines and chemokines such as IL-6, IL-8, and GM-CSF [85, 116, 117]. In IL-1β-stimulated human ASM cells, Kaur et al. demonstrated that enhanced IL-8 release evoked by IL-1β was significantly reduced by PGE₂, β₂-agonists and forskolin via a PKA-dependent mechanism [79]. Thus cAMP and β-agonists can have both pro- and anti-inflammatory effects on ASM cell function. Importantly, the combination of β-agonist and corticosteroid was observed to be more effective in inhibiting TNF-α-induced IL-8 release in human ASM cells although β-agonist itself increased IL-8 production [117].

5.3. cAMP-mediated ASM phenotype switching

In addition to measuring altered human ASM cell functions in vitro, much interest lies in the mechanisms of human ASM phenotype switching, i.e. identifying and quantifying human ASM cells with increased “proliferative” or “synthetic” rather than “contractile” phenotypes [118-120]. The phenotype is regulated by environmental factors such as cell-cell interactions, ECM, cytokines, growth factors and mechanical forces. It is known that smooth muscle specific proteins including smooth muscle myosin heavy chain (smMHC), SM22α, α-actin, and calponin are markers for the contractile phenotype and differentiation in ASM cells [120]. It has been shown that serum response factor (SRF) is a crucial transcriptional factor of smooth muscle cell differentiation including ASM cells [121, 122]. In vascular smooth muscle cells, PKA negatively regulates SRF and inhibits differentiation [123, 124]. Although the effects of PKA and Epac on SRF activation have yet to be studied in ASM cells, the observations by Roscioni et al. appear to contrast with those from studies of vascular smooth muscle cells [108, 123-125]. Roscioni and colleagues investigated the relative requirement for both PKA and Epac in proliferative/contractile phenotypic switching in human tracheal smooth muscle strips. Using selective activators of each, both Epac and PKA were able to prevent the phenotype altering to a PDGF-induced proliferative one. Collectively, these studies suggest a hitherto unrealised extended therapeutic benefit of asthma therapy utilising PKA- and Epac-mediated signaling.

6. THERAPEUTIC IMPLICATIONS: NOVEL TARGETS VERSUS CURRENT THERAPIES

6.1. Problematic features of β₂AR-agonist/cAMP signaling

As mentioned in the Introduction, despite the reported clinical benefits of β-agonists in asthma therapy, regular use of β-agonist may in fact promote harmful effects in the longer term [77, 126]. Recent studies have drawn considerable attention to limitations of such therapy. Chronic or repeated exposure to β-agonist results in loss of its relaxing effects against ASM contraction, in at least a significant subpopulation of asthmatics [127]. Numerous studies suggest that the β₂AR dysfunction is brought about by inflammatory cytokines and lipid mediators [81, 96, 128-130]. β₂AR desensitization has been presumed to underlie the reduction of β-agonist therapeutic effects associated with both airway inflammation and chronic β-agonist use. One of the main mechanisms of desensitization to β-agonists is mediated by phosphorylation of β₂AR. Benovic et al.found that β₂AR of the hamster lung is phosphorylated by PKA [131]. In the asthmatic lung, products of inflammation such as PGE₂ represent powerful agents capable of stimulating PKA activity in ASM. In cultured ASM cells, either direct application of PGE₂ or induction of autocrine PGE₂ following treatment with the cytokines IL-1β and TNF-α [132], results in heterologous desensitization of human β₂AR signaling [96, 133] that is reversed in cells expressing the PKA inhibitory peptide PKI. Moreover, chronic treatment of murine tracheal rings with the cytokines IL-1β and TNF-α attenuates the relaxant effect of β-agonist, suggested a functional heterologous β₂AR desensitization [96].

It is well recognized that the agonist-specific desensitization of many GPCRs is mediated by G protein coupled receptor kinases (GRKs) in many cell types. β₂AR phosphorylation by GRKs induces receptor binding to β-arrestin, creating steric inhibition between the β₂AR and G_αs [134, 135], effectively “arresting” signaling. Using HEK-293 cells transfected by mutated β₂AR lacking PKA and/or GRK phosphorylation sites on ASM, Wang et al. demonstrated that isoproterenol-induced binding to β-arrestin-2 to the β₂AR was regulated by GRK rather than PKA [136]. GRK inhibition using either G_βγ sequestrants or GRK2/3 siRNA has demonstrated the role of GRK2/3 in agonist-specific (homologous) desensitization of the β₂AR in human ASM cells [88], whereas siRNA targeting β-arrestin1/2 augments β₂AR signaling in human ASM, and genetic ablation of β-arrestin-2 increases both β₂AR signaling and the ability of β-agonist to relax contracted murine ASM ex vivo and in vivo. Thus, GRK/arrestin-dependent mechanisms play a clear role in the biochemical and functional desensitization of the β₂AR in ASM.

6.2. Future directions: what next?

To date, attempts to co-opt or exploit critical elements of β₂AR signaling have not translated into alternative treatments for asthma, and β-agonists along with corticosteroids remain the most frequently used asthma drugs. Whether by design or coincidence, salmeterol has in cell-based studies exhibited resistance to agonist-induced β₂AR desensitization, raising the possibility that the intrinsic activity of an agonist may influence mechanisms of homologous desensitization [137, 138]. However, this issue remains unclear as is the question as to whether functional tachyphylaxis is different in those who use different inhaled β-agonists, be they LABAs such as salmeterol, indacaterol, and formoterol, or short acting β-agonists such as albuterol.

However, as we continue to advance our understanding of β₂AR signaling and regulation, and the mechanisms it employs to regulate ASM contraction and other pathological features of asthma, we can envision “better” β-agonists or new drugs that similarly mediate their beneficial effects. For example, β-agonists with biased signaling properties [77] that enable avoidance/minimization of β₂AR desensitization, or adjuncts that target β₂AR desensitization mechanisms such as GRK inhibitors. Alternatively, drugs that effect similar bronchorelaxant mechanisms as those mediated by β-agonists yet are not subject to such profound negative feedback or exhibit tachyphylaxis. Such drugs might possibly target specific EP receptor isoforms, specific PDE isoforms or their key localized activities, or include other agents capable of mediating critical PKA or Epac activities. Such alternative agents might be necessary if we are ever to improve our protection from or treatment of asthma, in order to avoid potentially deleterious effects of β₂AR signaling that may be permissible for asthma to develop [139, 140] and possibly contribute to increased mortality [77, 126, 141]. Clearly however, such therapeutic advances will rely on additional basic science in the form of β₂AR molecular pharmacology, cell biology and biochemistry providing insight into mechanisms mediating antagonism of ASM contraction, and complementary physiology linking such signaling to ASM function.