Current and Novel Bronchodilators in Respiratory Disease

Domenico Spina


Curr Opin Pulm Med 

In This Article


β2-Agonists reduce airflow limitation by improving airway diameter as a consequence of a direct action on airway smooth muscle. β2-Adrenoceptors are located throughout the airways, principally found on airway smooth muscle but also on a variety of pulmonary cells including epithelium, submucosal glands and mast cells, although to what extent activation of β2-adrenoceptors on non-airway smooth muscle cells contributes to reducing airway obstruction remains a subject of debate.[1,2] β2-Agonists can be broadly classified according to their duration of action: short-acting β2-agonists (SABAs), including salbutamol, terbutaline and fenoterol, have pharmacodynamic half-lives between 2 and 6 h;[1] long-acting β2-agonists (LABAs), including salmeterol and formoterol, require twice daily treatment;[1] and ultra-LABAs (e.g. indacaterol) require once a day dosing.[19] Other long-acting β2-agonists that are currently being developed as once a day treatment include vilanterol,[20] olodaterol,[21] carmoterol,[22] abediterol[23] and milveterol.[24]

Reversible binding of these agonists to the β2-adrenoceptor causes the activation of the canonical Gαs-protein–cyclic AMP pathway, and hence relaxation of airway smooth muscle (Fig. 1).[25–28,29,30–32] Activation of β2-adrenoceptors, which serve as guanosine exchange factors, leads to the replacement of guanosine diphosphate (GDP) for guanosine triphosphate (GTP) on the Gαs subunit of trimeric G protein. The subsequent dissociation allows Gαs-GTP to bind to transmembrane spanning adenylyl cyclase and Gβγ to a myriad of signalling proteins (e.g. GRK2, PLCβ, PI3K).[25] The hydrolysis of GTP to GDP by the endogenous GTPase activity of Gαs leads to the reassembly of the trimeric G protein for subsequent activation. Conversion of ATP to cyclic AMP by adenylyl cyclase results in the stimulation of protein kinase A (PKA) and Epac (exchange proteins activated by cyclic AMP) dependent pathways leading to dephosphorylation of myosin light chain (MLC20) and airway smooth muscle relaxation.[26,27] Other mechanisms for relaxation have also been proposed including the opening of potassium channels to induce hyperpolarization, although the importance of this mechanism has been questioned in human airways[7,28] (Fig. 1).

Figure 1.

Illustration of the potential mechanism of clinically effective bronchodilators (β2-agonists and muscarinic antagonists) and potentially novel targets including ion channels, other G-protein-coupled receptors and intracellular proteins (Table 1). It has been proposed that regular treatment with β2-agonists might promote b-arrestin and/or ERK dependent pathways in the lung to cause inflammation and bronchial hyperresponsiveness (BHR). Nadolol might interfere with this process to promote anti-inflammatory/anti-BHR action [31,32]. AC, adenylyl cyclase; ASM, airway smooth muscle; ATP, adenosine triphosphate; cAMP, cyclic adenosine monophosphate; CM, calmodulin; Epac, exchange proteins activated by cyclic AMP; ERK, extracellular receptor kinase; GDP, guanosine diphosphate; GEF, guanosine exchange factor; GRK, G protein receptor kinase; GTP, guanosine triphosphate; JNK, janus kinase; MAPK, mitogen activated protein kinase; MLC20, myosin light chain; MLCK, myosin light chain kinase; MLCP, myosin light chain phosphatase; MYPT1, myosin phosphatasetargeting substrate-1; P3IK, phosphoinositide 3-kinase; PKC, protein kinase C; PLC, phospholipase C; RhoGEF, rhoA guanosine exchange factor; ROCK, rhoA kinase; RTK, receptor tyrosine kinase.

Activation of the β2-adrenoceptor can also lead to the phosphorylation of the cytoplasmic tail of the receptor by G protein receptor kinase (GRK) and/or PKA leading to the binding of β-arrestin, which can interdict Gαs–adenylyl cyclase signalling, promote receptor internalization/recycling and serve as a platform to recruit a variety of intracellular signalling proteins (eg Raf/MEK/ERK)[29,30,33–36] (Fig. 1) (Table 2). This can give rise to a number of cellular responses including cell chemotaxis, apoptosis, proliferation, regulation of protein synthesis and metastasis.[29,30] It is clear that ligands can induce different conformational states in G-protein-coupled receptors resulting in the phenomenon of biased agonism.[33,37] For example, β2-agonists including salbutamol and salmeterol display different efficacy for Gαs-stimulated cyclic AMP generation, β-arrestin-2 translocation to the receptor and activation of ERK (Table 2), which may be related to the different ways these agonists stabilize the receptor. Similarly, relative to epinephrine, both formoterol and salmeterol are weakly biased towards β-arrestin[34] (Table 2), and to what extent this affects the clinical effectiveness and safety of these drugs is not known at present.

In contrast, carvedilol, propranolol and nadolol do not cause Gαs-stimulated cyclic AMP generation but behave as inverse agonists with negative efficacy for cyclic AMP production. However, carvedilol and propranolol are partial agonists for the activation of ERK1/2 but via different mechanisms. Nadolol differs from carvedilol and propranolol and does not activate ERK1/2 signalling[33,36] and is another illustration of the flexibility of receptors in adopting ligand-dependent conformational states that can exert distinct functional effects of varying efficacy (Table 2). The potential functional consequence of differences in the effect of these ligands on the β2-receptor signalling pathway in airway inflammation will be described later.

To what extent the clinical effectiveness of β2-agonists is determined by their different efficacy for these distinct signalling pathways at native β2-adrenoceptors on human airway smooth muscle cells and whether biased agonists 'selective' for the canonical Gα–cyclic AMP-dependent pathway would be clinically more effective than currently used β2-agonists remain to be determined.[31]

Mechanism of Sustained Bronchodilation

The chemical evolution of β2-agonists to once daily administration is seen as a major advance in the bronchodilator treatment of respiratory disease and various mechanisms have been proposed to account for the pharmacodynamics of LABAs. Radioligand binding studies performed under physiological conditions revealed that the association or dissociation rate constants for various β2-agonists cannot explain the property of onset of action and long-effect duration. For example, the bronchodilator onset of action of salmeterol is slower compared with other LABAs, yet the association rate constant on human recombinant β2-adrenoceptors is fast and highlights a discordance between receptor binding kinetics and clinical effect.[38]

Similarly, the long duration of action of these agonists is also not explained by slow dissociation half-life from the receptor, which in general occurs within minutes (Table 3).[20,21,38,39–41] There is an exception to this rule in the case of olodaterol (t1/2 = 17 h);[39] however, the magnitude of this value has been questioned on the grounds that the binding studies were not undertaken under physiological conditions (e.g. physiological Na+ concentration, presence of GTP) to maintain the receptor in a predominantly low affinity state as might be anticipated in living cells.[38] Similarly, although the dissociation of vilanterol from the high affinity state of the β2-adrenoceptor is relatively long (47 min), this would still be insufficient to account for sustained bronchodilation with once a day treatment.[20] Hence, factors such as efficacy and retention within specific domains of cell membranes are likely to account for long duration of action rather than receptor kinetics. For example, although indacaterol is equally hydrophobic as salmeterol, this agonist documents a two-fold difference in preference for partitioning in lipid rafts that has been suggested to contribute towards the longer clinical duration of action of indacaterol over salmeterol.[42] Furthermore, the efficacy of indacaterol is also an order of magnitude greater than salmeterol and is likely to be a factor concerning duration of action (Table 3).

Another characteristic feature of LABAs is persistent 'retention' by cells and tissues in vitro long after termination of drug exposure. The contraction of airway smooth muscle by endogenously released acetylcholine following the activation of parasympathetic postganglionic nerve terminals is functionally antagonized by LABAs. This action is reversed by β-adrenoceptor antagonists because of competition with LABAs for β2-receptor binding sites. However, the functional antagonism resumes upon removal of the antagonist. Ligands like salmeterol, indacaterol and vilanterol demonstrate 're-assertion behaviour', although this feature is lost for both formoterol and carmoterol in vitro[20,43]. Hence, the long duration of action of LABAs observed clinically is a consequence of efficacy and retention within the vicinity of receptor sites via: binding to an 'exosite' as in the case of salmeterol and possibly vilanterol;[44] preferential retention in lipid domains coupled with very high efficacy (e.g. formoterol);[42,45] and 're-binding' of ligands (e.g. moderate-to-high efficacy ligands, olodaterol, indacaterol and carmoterol).[46,47]

β-receptor Polymorphisms

Single nucleotide polymorphisms and haplotypes have been described for the β2-adrenoceptor and their functional consequence on receptor expression and sensitivity to desensitization has been investigated in vitro, and not surprisingly their role in disease morbidity and clinical effectiveness is the subject of considerable debate.[48] Of the numerous receptor polymorphisms that have been identified, the Arg16 → Gly16 and Ile164 → Thr164 alleles have been more widely investigated. The Gly16 and Thr164 alleles correspond to receptors with greater susceptibility to desensitization and reduced signalling in vitro, respectively.[48] Only the Thr164 receptor polymorphism was associated with reduced lung function and increased risk of COPD, and neither this nor the Gly16 or Glu27 polymorphism was associated with increased risk of asthma.[49,50] Various studies have implied allele specific differences in treatment responsiveness;[48] however, a number of large well controlled clinical trials have shown that there is no evidence of differences in improvement in baseline lung function, exacerbation rate and symptom free days following monotherapy or combination LABA/glucocorticosteroids treatment in Arg16 versus Gly16 genotypes.[50–52] Similarly, there was no difference in improvement in lung function in individuals with the Arg16 genotype treated with salmeterol or tiotropium bromide in combination with steroid.[53] This suggests that earlier studies reporting deterioration in lung function following regular β2-agonist therapy, although paradoxically in the Arg16 and not the Gly16 genotype, was not observed with LABAs.[48] The Salmeterol Multicentre Asthma Research Trial (SMART) investigating the safety of salmeterol was halted because of safety concerns in African-American patients; however, there was no evidence of a difference in asthma control between races,[50] although there was some discrepancy concerning the lack of bronchoprotection afforded by salmeterol in African-Americans with the Arg16 genotype.[50,52] Similarly, airway responsiveness in combination with LABA/glucocorticosteroids treatment was not different in patients with β2-adrenoceptor haplotypes that included the Gly16, Glu27 and Ile164 alleles.[50] Therefore, it is unlikely that β2-receptor polymorphisms play a major role in asthma morbidity and/or responsiveness to β2-adrenoceptor agonists.

Loss of Bronchoprotection

Regular treatment with SABAs has been linked to increasing morbidity and, controversially, mortality in asthma. Similarly, a number of studies have shown that regular treatment with SABAs and LABAs is associated with a loss in bronchoprotection to direct and indirect acting stimuli.[54,55]

Regular treatment with salmeterol did not cause bronchodilator tolerance as assessed by improvement in forced expiratory volume in 1 s (FEV1) in response to this bronchodilator; however, there was a significant loss in bronchoprotection against methacholine challenge from 3.3 doubling doses on day 0, decreasing to 1 doubling dose after 8 weeks regular treatment,[56] and a similar observation has been observed for formoterol.[57] This loss in bronchoprotection is also observed in studies that examined the ability of salbutamol to reverse acute bronchospasm in response to methacholine in individuals regularly receiving LABAs, in an attempt to replicate circumstances in which asthma patients might undergo an exacerbation of their asthma during maintenance LABAs, who require rescue medication. An approximate two-fold loss in 'rescue' bronchodilator potency is observed under these conditions.[58–60]

It is generally considered that β2-adrenoceptor desensitization is responsible for this loss of bronchoprotection despite the fact that bronchodilation per se is unaltered presumably because of the very significant β2-receptor reserve on human airway smooth muscle[61] and the rapid rate at which Class A receptors like β2-adrenoceptors recycle to the cell surface.[62] Moreover, glucocorticosteroids can reverse β2-adrenoceptor desensitization in vitro, but do not appear to protect against this loss in bronchoprotection.[54,55] Furthermore, there is no difference in the loss of bronchoprotection following regular LABA/glucocorticosteroid treatment in individuals with Arg16 or Gly16 genotype, the latter being associated with a greater probability of receptor desensitization.[63]

A potential mechanism to account for this loss in bronchoprotection might involve non Gαs–cyclic AMP-dependent pathways. Genetic ablation of β-arrestin-2 led to the inhibition of eosinophil recruitment and reduction in T helper 2 (Th2)-like cytokine levels presumably reflecting the importance of this protein for Th2 cell chemotaxis to the lung via G-protein-coupled chemokine receptor signalling mechanisms and chemokine release from pulmonary cells.[64] Bronchial hyperresponsiveness (BHR) was also inhibited but was not due to impairment in airway smooth muscle contraction per se, suggesting interdict of an upstream process in knockout mice (e.g. increased epithelium thickness, loss in airway surface tension).[64] The expression of β-arrestin-2 in cells within the lung (e.g. epithelium and/or airway smooth muscle) was sufficient to drive antigen-induced BHR.[65] Regular treatment with β2-agonists could lead to the recruitment of β-arrestin and subsequent activation of intracellular signalling cascades [e.g. MAPK-dependent pathways, phosphodiesterase 4 (PDE4)], which could drive inflammatory processes,[29,30,66,67] increase BHR and reduce the clinical effectiveness of β2-agonists.[32] It remains to be established whether the development of an allergic inflammatory phenotype in mice following regular β2-agonist exposure is via a β-arrestin and/or ERK-dependent process or via the canonical Gαs–cAMP pathway as many of these indices of the allergic inflammatory response are also suppressed in PDE4B gene knockout mice, implicating signalling pathways downstream of cyclic AMP.[68]

Intriguingly, genetic ablation of the β2-adrenoceptor also suppressed the development of allergic inflammation, and increased epithelium mucin content and BHR in mice,[69] which was mimicked by chemical depletion of endogenous catecholamines or in mice lacking the ability to synthesize epinephrine.[70] A major conclusion drawn from this study is that constitutive β2-adrenoceptor activity does not promote the asthma phenotype in allergic mice but requires ligand-induced activation of the β2-adrenoceptor.[70] Moreover, the allergic phenotype could be re-capitulated following chronic treatment with formoterol in mice unable to synthesize epinephrine.[70] The finding that formoterol demonstrates bias towards the β-arrestin pathway relative to epinephrine (Table 1) suggests that signalling via this pathway could account for these findings. The allergic inflammatory phenotype in mice is also pre-empted to varying degrees of effectiveness following chronic treatment with inverse agonists (nadolol = ICI-118551 > carvedilol), but not the partial agonist alprenolol.[69,71,72] It has been suggested that the beneficial action of nadolol may not be due to inverse agonism per se, but possibly by impairing β2-adrenoceptor signalling via β-arrestin and/or ERK signalling[69,70,71,72] (Table 2) (Fig. 1).

The clinical relevance of these findings is being explored with an open-label trial reporting a decrease in BHR following chronic treatment with nadolol,[73,74] although there appeared to be no beneficial effect on this variable following chronic treatment with propranolol in asthma patients concurrently treated with glucocorticosteroids and muscarinic antagonist.[75] These different clinical outcomes would be consistent with the hypothesis that nadolol and not propranolol impairs ERK signalling pathways, and therefore nadolol would be anticipated to exert a beneficial action compared with propranolol.[31]