Inhaled β2 Adrenergic Agonists and Other cAMP-Elevating Agents: Therapeutics for Alveolar Injury and Acute Respiratory Disease Syndrome?

II. β2-Adrenergic Receptor Signaling and Pharmacology

β-ARs are arguably the best studied GPCRs. The three β-AR subtypes (β1-AR, β2-AR, and β3-AR) are class A/rhodopsin-like GPCRs that are differentially expressed in tissues and cells in humans and many other species. In humans, β1-ARs are preferentially expressed in the heart and adipose tissue, whereas β2-ARs are expressed in smooth muscle cells and many other cell types (including immune, epithelial, skeletal muscle, and glandular cells). Β3-ARs are preferentially expressed in adipose tissue and the urinary bladder.

Figure 1 shows expression of β2-AR mRNA (the ADRB2 gene) in human tissues, based on data from the Genotype-Tissue Expression project (Lonsdale et al., 2013), curated by the Human Protein Atlas (http://www.proteinatlas.org; Uhlén et al., 2015). β2-AR expression in the lung is among the highest of all tissues. β-ARs are expressed in multiple regions of the lung, including airway and alveolar epithelia, smooth muscle, and endothelia. Autoradiographic quantification of β-ARs and of β1- and β2-ARs (Carstairs et al., 1985) (Fig. 2) indicate that β-AR expression is most abundant in alveoli and that most of these receptors are β2-ARs. Subsequent work has shown that type 1 and type 2 alveolar epithelial cells (AECs) highly express β2-ARs (Liebler et al., 2004; Mutlu and Factor, 2008).

• Download figure
• Open in new tab
• Download powerpoint

Fig. 1

Expression of ADRB2 mRNA in human tissues. mRNA expression [normalized as transcripts per million (TPM)] from RNA sequencing studies in the Genotype-Tissue Expression database (Lonsdale et al., 2013) and were downloaded from the Human Protein Atlas (Uhlén et al., 2015).

• Download figure
• Open in new tab
• Download powerpoint

Fig. 2

Normalized expression of β-ARs in different regions of the human lung. The data (from tabulated values in Carstairs et al., 1985) are presented as the total abundance of β-ARs and relative proportion of β1- and β2-ARs, based on autoradiography using [I¹²⁵] iodocyanopindolol to label β-ARs in lung sections and with β1- and β2-selective inhibitors to identify the β-AR subtypes.

Norepinephrine (noradrenaline) and epinephrine (adrenaline) are the physiologic agonists of β-ARs. Norepinephrine is primarily produced by neurons in the central nervous system and postsynaptic neurons of the sympathetic nervous system in peripheral tissues, whereas epinephrine is predominantly synthesized by chromaffin (neuroendocrine) cells (e.g., in the adrenal medulla) and is primarily a circulating hormone. Norepinephrine has much greater affinity for β1-ARs than β2-ARs and regulates effector cells innervated by postsynaptic sympathetic neurons. By contrast, as a circulating hormone, epinephrine can activate β1-ARs, β2-Ars, and β3-ARs on target cells.

β-AR activation (by norepinephrine, epinephrine, or pharmacological agonists) activates the heterotrimeric (αβγ) GTP binding protein Gs, which stimulates adenylyl cyclase (AC) activity and promotes the synthesis of cAMP (Fig. 3). Although β-ARs, as 7-transmembrane GPCRs, were thought to be activated exclusively in the plasma membrane, work in recent years has provided evidence for receptor signaling at intracellular sites (Irannejad et al., 2013; Tsvetanova and von Zastrow, 2014; Irannejad et al., 2017; Calebiro et al., 2021). β-AR–stimulated increase in cellular cAMP concentration is blunted by the hydrolysis of cAMP to 5′AMP by PDEs and extrusion of cAMP from cells by multidrug resistance proteins 4 and 5 (ABCC4/5) (Sassi et al., 2012). The β-AR–promoted increase in cAMP regulates cellular physiology by activation of one more of at least three effector proteins: 1) protein kinase A (PKA) and PKA-mediated phosphorylation of protein targets; 2) exchange proteins activated by cAMP (EPACs), which are nucleotide exchange factors for the Rap subfamily of low molecular weight GTP binding proteins; and 3) cyclic nucleotide-gated (CNG) ion channels. Those canonical signaling mechanisms alter many cellular functions via changes in activities of proteins/enzymes, ion channels, and gene expression (Fig. 3). In addition to activation of Gs, β-ARs can also regulate cellular function via β-arrestins, which facilitate endocytosis of activated/phosphorylated β-ARs and can serve as a platform for intracellular signaling (Ippolito and Benovic, 2021).

• Download figure
• Open in new tab
• Download powerpoint

Fig. 3

Signaling mechanisms/pathways through which β2-ARs regulate multiple aspects of cell physiology. CREM, cAMP responsive element modulator; Csk, C-Terminal Src Kinase; GSK-3β, Glycogen Synthase Kinase 3 Beta; ICER, Inducible cAMP early repressor; IP3R, Inositol 1,4,5-trisphosphate receptor; MRP4/5, multidrug resistance-associated protein; RAP1/2, Ras-related protein 1/2.

Many approved drugs target β-ARs. Initial therapeutics were norepinephrine and epinephrine, which activate α-ARs in addition to β-ARs and thus can produce numerous adverse effects (AEs). Subsequent efforts led to the discovery and use of α-AR– and β-AR–selective agonists and antagonists and then β1-AR– and β2-AR–selective drugs. The recognition that inhalational delivery of β2-AR agonists preferentially targets the airways and lungs and minimizes systemic AEs was an important advance in the treatment of pulmonary disorders, especially asthma and COPD (Larsson and Svedmyr, 1977; Heel et al., 1978).

β2-AR agonists with relatively short duration of action (<6 hours) were initially developed and approved for administration to patients with bronchoconstriction, in particular asthma and COPD. These SABAs rapidly relax airway smooth muscle (bronchodilation) and improve symptoms. Other β2-AR agonists were synthesized that had similar actions but longer half-lives (∼12 hours) and thus were termed long-acting beta-agonists (LABAs). Use of LABAs was viewed as a way to decrease administration of corticosteroids in patients, especially those with mild to moderate asthma, to treat ongoing symptoms in patients with COPD and/or to prevent acute exacerbations in patients with asthma or COPD (e.g., Larj and Bleecker, 2004; Mapel et al., 2006). However, LABA administration as monotherapy is associated with an increased risk of fatal exacerbations in asthma (McMahon et al., 2011; Cates et al., 2012; Kersten et al., 2017). Hence, guidelines for treating patients with asthma dictate that LABAs should be administered together with an ICS [e.g., budesonide, fluticasone, etc., which are glucocorticoids (GCs)], to decrease inflammation and prevent sequelae of an LABA administered alone (Cloutier et al., 2020; Global Initiative for Asthma, 2020). The possible development of β-AR agonists for asthma and COPD that have signaling bias to enhance efficacy and decrease toxicity has been proposed (Matera et al., 2018).

Food and Drug Administration–approved inhaled agonists for β2-ARs include LABAs and SABAs, with differences in pharmacological properties, including potency, selectivity, and duration of effect. Table 1 summarizes certain key aspects of pharmacological properties for inhaled SABAs and LABAs that are commonly used for asthma and COPD. LABAs typically show greater potency for β2-ARs, greater selectivity in the activation of β2-ARs (compared with activation of β1-ARs), and longer duration of effect (typically ∼2–3 times longer). In general, all of these drugs have rapid onset of action, typically ∼5–15 minutes (Cazzola et al., 2013, Billington et al., 2017, Cazzola et al., 2019). SABAs are typically considered partial agonists of β2-ARs, whereas the LABAs are usually considered full agonists (Altosaar et al., 2019). This determination is typically assessed by comparing maximal β2-AR activation induced by these agents, relative to isoproterenol (isoprenaline). A range of results have been reported (e.g., Battram et al., 2006; Cazzola et al., 2013; Baker, 2010; Cazzola et al., 2019) but the data generally indicate nearly 100% maximal activation of β2-ARs by most LABAs (relative to isoproterenol), with lower values for SABAs (typically 50% or less). Hence, in terms of producing potent, prolonged, and selective β2-AR–mediated effects in cell types that may participate in pulmonary pathobiology, LABAs have key pharmacological advantages over SABAs.

A pharmacological rationale exists for combining β2-AR agonists with GCs (Barnes, 2002; Page et al., 2017) and/or LAMAs (Dale et al., 2014; Calzetta et al., 2018), and such combinations are approved as inhaled therapy for COPD (Calverley et al., 2017). LAMAs primarily inhibit M3 muscarinic receptors in the lungs but also block M1 and M2 receptors (depending on selectivity of the drugs) (Ejiofor and Turner, 2013; Alagha et al., 2014). M1 and M3 receptors couple to Gq/G11 signaling mechanisms [Inositol trisphosphate (IP3)-mediated calcium and protein kinase C–mediated responses], whereas M2 receptors couple to Gi/Go (reducing cAMP) (Guide to Pharmacology database; Alexander et al., 2019). Inhibition of M1–M3 receptors together with activation of β2-AR receptors can yield additive or perhaps synergistic (i.e., more than additive) responses, such as in the relaxation of airway smooth muscle (Calzetta et al., 2018). Synergy between β2-ARs and GCs can also occur: GCs can reduce β2-AR desensitization and enhance β2-AR functional responses, whereas β2-AR signaling increases GC receptor translocation, transcriptional response, and receptor expression (Barnes, 2002; Johnson, 2004; Tamm et al., 2012, Newton and Giembycz, 2016; Page et al., 2017). In addition, as discussed below, GC signaling enhances actions/expression of ion channels regulated by β2-ARs in epithelial cells. LAMAs (Alagha et al., 2014; Beeh et al., 2017) and GCs (Cain and Cidlowski, 2017) also exert anti-inflammatory actions on immune cells, suggesting additive or perhaps synergistic effects when combined with LABAs.

III. Acute Lung Injury/Acute Respiratory Distress Syndrome: An Overview

ARDS is a syndrome defined by acute onset hypoxemic respiratory failure with bilateral pulmonary edema that is not due to heart failure and that has an identifiable provocation. (Matthay et al., 2019). The revised Berlin criteria stratify ARDS by the ratio of partial pressure of arterial oxygen to fraction of inhaled oxygen (P/F). Severe disease has a P/F <100, moderate disease 100–200, and mild disease <300 (ARDS Definition Task Force et al., 2012). More recently other stratification has been proposed: severe disease, P/F <150; nonsevere disease, P/F 150–300 (Maiolo et al., 2018). Acute lung injury (ALI) was historically used to categorize patients with a P/F ratio of 200–300 (Johnson and Matthay, 2010) and to describe pathologic injury of ARDS when not referring to the clinical syndrome. ALI is no longer commonly used in the clinical literature, but studies of animal models of ARDS still frequently use ALI as an alternative for ARDS. In this review, when discussing ALI/ARDS, we refer to a spectrum of pulmonary pathology (including edema and decreased lung function) in animal studies that is consistent with mild-to-severe ARDS in humans, per the Berlin definition (ARDS Definition Task Force et al., 2012; Ferguson et al., 2012).

ARDS develops in numerous clinical settings, including pneumonia, sepsis, trauma, aspiration, pancreatitis, and after exposure to various medications or toxins. Approximately 40% of patients with ARDS die (Bellani et al., 2016). The lungs of such patients typically have increased stiffness, resulting from diffuse alveolar damage that involves airway epithelial cells and the pulmonary endothelia with edematous fluid in the alveolar space. ARDS is heterogeneous; new classifications of the subphenotypes of patients may help advance understanding of ARDS and improve its outcomes (e.g., Standiford and Ward, 2016; Carla et al., 2020). The necessity to better understand ARDS pathobiology has been underscored by the coronavirus disease 2019 (COVID-19) pandemic: severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection can result in alveolar injury that progresses to ARDS in severe cases, which can lead to cytokine storm and systemic organ failure (Sriram and Insel 2020a; Sriram and Insel 2021; Batah and Fabro 2021; Hasan, 2021).

ARDS arising from COVID-19 has also underscored the importance of coagulopathy and thromboinflammation (i.e., a pathologic combination of excessive inflammation and thrombosis) in driving lung injury (reviewed in Sriram and Insel 2021). Numerous strategies have been proposed to tackle these aspects of COVID-19, including a range of agents that target aspects of coagulation, inhibit cytokine signaling [e.g., antibodies for cytokines such as interleukin (IL)-6] (Liu et al., 2020), or reduce inflammation more generally, including corticosteroids (Scavone et al., 2020; Zhang et al., 2020). The RECOVERY trial (RECOVERY Collaborative Group et al., 2021) revealed beneficial effects of systemic dexamethasone treatment of patients receiving oxygen or on mechanical ventilation. In addition, recent data indicate benefits using ICS (Ramakrishnan et al., 2021), suggesting the possible utility of combining these drugs with inhaled LABAs; this topic is discussed further in subsequent sections.

Current therapy of ARDS emphasizes the need to treat the underlying condition and to employ appropriate use of ventilators (in particular, with positive end-expiratory pressure), prone positioning, management of fluids, and supportive care. Neuromuscular blocking agents, typically nicotinic cholinergic blocking agents, have been employed to aid in the use of ventilators, but a recent systematic review questioned the impact of such agents on patient-important outcomes (Tarazan et al., 2020). “Rescue therapies” to treat patients with progressive clinical deterioration include extracorporeal membrane oxygenation and extracorporeal CO₂ removal. Systemically administered GCs are commonly administered to patients with ARDS, but their use is controversial: mixed results were obtained in clinical trials prior to the current period in which lung-protective ventilation is a key approach (Matthay et al., 2019). As indicated by its high mortality, effective pharmacological therapies do not currently exist for ARDS.

Figure 4 summarizes the pathobiology of ARDS resulting from viral infections [e.g., influenza, severe acute respiratory syndrome (SARS), Middle East respiratory syndrome, SARS-CoV-2] of alveolar epithelial cells that induce lung injury. Numerous cell types in the lung contribute to pathophysiology and may be associated with secondary (typically bacterial) infection. The summary in Fig. 4 is based on descriptions of ARDS pathobiology in SARS virus infections (Sriram and Insel, 2020a; Sriram et al., 2020; Horie et al., 2020), models for ALI/ARDS (e.g., Herrero et al., 2018), thromboinflammation in ARDS (e.g., Frantzeskaki et al., 2017), ARDS in influenza (Herold et al., 2015), and thromboinflammation, epithelial injury and complement activation in COVID-19 ARDS (Sriram and Insel, 2020a, 2020b; Sriram and Insel, 2021). In brief, alveolar inflammation induced by infection can produce inflammation of nearby endothelial cells and fibroblasts, increasing cytokine release from these cell types. Immune infiltration (primarily of innate immune cells, especially early in the infection) increases; immune cell activation and adhesion to epithelia further promote inflammation. In severe infections with virus-induced inflammation (perhaps along with underlying inflammatory conditions) hyperinflammation can ensue as a consequence of proinflammatory events that prevent viral clearance by innate and adaptive immune responses and, thereby, promote injury. Multiple mechanisms (complement signaling, tissue factor release, ADP release, platelet activation, and coagulopathy) are induced; platelet activation further contributes to inflammation via secreted factors and interactions with endothelial cells and neutrophils. In addition, epithelial disruption leads to reduced surfactant production and susceptibility to secondary bacterial pneumonia. Thus, a series of feedback loops is established, whereby multiple cell types interact to promote injury, inflammation, and coagulopathy, resulting in progressive alveolar damage and edema, ultimately leading to ARDS.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 4

Interactions between host cell types in driving pathobiology in viral/bacterial pneumonia, leading to alveolar injury, hyperinflammation, and coagulopathy, with progressive injury leading to ARDS in severe disease.

This progressive pathobiology is counteracted by protective immune mechanisms, in particular adaptive responses that follow seroconversion and can lead to pathogen clearance. Clinical features and patient outcomes thus result from the competing actions between progressive inflammation and injury (which can be enhanced by comorbidities associated with age and other factors) and protective responses (Sriram and Insel, 2020a; Sriram et al., 2020; Sriram and Insel, 2021). In this framework, drugs with anti-inflammatory effects and that target multiple cell types are predicted to be beneficial, especially with early administration prior to induction of hyperinflammation. These mechanisms provide a rationale for re-examining β2-AR agonists in the treatment of ALI/ARDS— in particular, inhaled LABAs administered early in viral respiratory infections.

IV. Evidence Regarding the Use of Long-Acting β-Adrenergic Agonists in Acute Lung Injury/Acute Respiratory Distress Syndrome

Are β2-AR agonists (including LABAs) useful in treating ALI/ARDS? Evidence indicating a protective effect of β2-AR agonists in enhancing alveolar fluid clearance (AFC), thereby reducing pulmonary edema (reviewed below and in Morty et al., 2007; Mutlu and Factor, 2008) led to the hypothesis that β2-AR agonists would be beneficial and protect from lung injury. Data showing a protective effect on the pulmonary endothelium (Morty et al., 2007; Mutlu and Factor, 2008) strengthened the rationale for testing β2-AR agonists in patients with ALI/ARDS. This hypothesis was also supported by evidence that prophylactic use of β2-AR agonists provided protection from high altitude pulmonary edema (HAPE) (Sartori et al., 2002). A prospective study in pediatric patients found that administration of inhaled β-AR agonists on the first day of onset of ALI reduced mortality compared with those not treated (Flori et al., 2005). Retrospective analysis of data from mechanically ventilated patients also indicated a lower severity of ALI and longer survival in patients receiving a β2-AR agonist (Manocha et al., 2006). Other data in patients support a protective role for β2-AR agonists in ALI (Roca et al., 2008): the investigators assessed exhaled markers of nitrosative and oxidative stress and found that inhaled salbutamol reduced inflammation in ventilated patients. Systemic administration of salbutamol also improved AFC in a small cohort of patients with COPD, who had greater improvements than did healthy control subjects (Di Marco et al., 2012). Initial data from the first β-Agonist Lung Injury (BALTI) trial indicated improved lung function (i.e., reduced fluid accumulation) of patients with ALI/ARDS in an intensive care unit (ICU) systemically administered β2-AR agonists but lacked statistical power to detect improvement in mortality (Perkins et al., 2006).

However, subsequent larger randomized controlled trials (RCTs) undermined the case for using β2-AR agonists in ARDS. The BALTI-2 trial sought to expand on BALT-1 trial results by evaluating 28-day mortality and clinical assessments in patients with ARDS admitted to the ICU and administered salbutamol by infusion (Gao Smith et al., 2012; Gates et al., 2013). This trial found no improvement in mortality with salbutamol infusion and, in addition, noted AEs (especially cardiovascular AEs), thus indicating a net harm associated with the treatment. Similarly, the ALTA trial (National Heart, Lung, and Blood Institute et al., 2011) evaluated the effects of aerosolized salbutamol in patients with ALI receiving mechanical ventilation. No statistically significant differences were noted between treatment and placebo control groups, including ventilator-free days and 60-day mortality, although there was little evidence of AEs.

These trials led to reevaluation of the potential utility of β2-AR agonists in the treatment of ALI/ARDS. Subsequent analyses sought to reconcile encouraging preclinical and early clinical data with the negative results (Levitt and Matthay, 2012; Bassford et al., 2012). Explanations for the different results include possible off-target actions of β2-AR agonists, adverse cardiovascular effects, and concerns about whether pulmonary edema was reversible in the context of severe injury (Bassford et al., 2012). In the BALTI-1 and BALTI-2 trials, salbutamol was administered intravenously, a route that would be predicted to lead to greater cardiovascular AEs than would inhalational therapy. Another key issue involved the timing of β2-AR agonist administration: early administration of β2-AR agonists was suggested as more likely to produce benefits (Bassford et al., 2012). This idea reflects a paradigm shift in ARDS research, with greater emphasis on early treatment, including prophylactic approaches in at-risk patients (Ortiz-Diaz et al., 2013; Artigas et al., 2017; Griffiths et al., 2019).

Other studies necessitate re-examination of inhaled LABA and ICS treatment of ALI/ARDS. A retrospective study at the onset of ARDS among hospitalized, at-risk patients found that prehospital use of β2-AR agonists (but not ICS) was associated with reduced incidence of ARDS (Mangi et al., 2015). Recent retrospective data of patients with ARDS administered inhaled bronchodilators (most of which were a SABA/LAMA combination) found that increased cumulative doses of bronchodilators were associated with lower mortality, although the contributions of SABA versus LAMA were not analyzed (A.M. Scott et al., preprint, DOI: https://doi.org/10.21203/rs.3.rs-143660). A prior retrospective analysis by those authors indicated that patients with moderate or severe ARDS administered salbutamol had a dose-dependent improvement in survival (Scott and Ouellette, 2019). By their retrospective design, these studies were unable to control for certain confounders, such as patients with known reactive airway disease likely having been more frequently receiving β-agonist treatment.

In contrast with these supportive data, a 2019 Cochrane review of pharmacological agents for treating ARDS in adults found no statistically significant benefit of β-agonists, or any other agent evaluated, in reducing mortality (Lewis et al., 2019). The authors also noted a (non-statistically significant) indication of increased harm. This Cochrane review only included data from the BALTI-2, BALTI-1, and ALTA trials, which contrast with the approach that we propose: early administration prior to ICU admission.

A prospective study (Festic et al., 2016; Festic et al., 2017) compared the use of a LABA [formoterol/ICS (budesonide)] with placebo in patients presenting to emergency departments with hypoxemia and risk factors for ARDS (high prediction scores for lung injury). The key difference, compared with previous studies (e.g., the BALTI trials), was drug administration prior to the onset of ARDS/ALI, i.e., prophylaxis to mitigate lung injury. The authors found that compared with placebo, LABA/ICS treatment reduced the incidence of ARDS, need for mechanical ventilation, length of intensive care, and duration of hospitalization. A recent study (Fouad et al., 2020) reported similar results: inhaled salbutamol as monotherapy or combined with ICS decreased the development of ARDS in hospitalized patients at risk for developing ARDS. Prioritizing early treatment, the same authors are currently conducting the ARREST Pneumonia trial that is testing whether early treatment of LABA/ICS in pneumonia can prevent exacerbation and respiratory failure (Levitt et al., 2021). The effect size LABAs in ARDS trials may also be reduced by concurrent administration of inotropes and vasopressors that have β-agonism (i.e., dobutamine, epinephrine, norepinephrine). Of note, an RCT with patients with COPD found that increased doses of LABA/ICS administered early in the onset of upper respiratory tract infection reduced the occurrence of severe exacerbations, thus highlighting the potential protective benefit of those drugs in that setting (Stolz et al., 2018).

Observational, retrospective studies have been conducted in patients with COPD, evaluating the effects of different inhaled therapies (various combinations of β-AR agonists, ICS, and LAMAs) and yielded ambiguous findings regarding the association between types of therapy and COVID-19 mortality (Bloom et al., 2021; Schultze et al., 2020). LABA use in the COVID-19 pandemic has received relatively little attention. Two RCTs have been initiated to investigate the use of inhalational LABA/ICS in COVID-19 and are recruiting patients as of March 2021 (NCT04331054 and NCT04331470).

The findings above thus imply the benefit of early treatment, perhaps with LABA/ICS, prior to severe lung injury. In subsequent sections, we review findings in pulmonary cell types that support early treatment with inhaled LABAs and ICS. The rationale for this approach is based on the following ideas:

1. Pulmonary pathology from infections with hyperinflammatory phenotypes results from interactions of multiple cell types and an interplay between cell death, inflammation, and coagulopathy that can lead to ARDS.

2. Feedback loops that drive disease progression are opposed by protective innate and adaptive immune responses, which contribute to tissue repair and clearance of pathogens. However, the protective responses are impeded by excessive inflammation.

3. β2-AR agonists have beneficial effects on multiple cell types (epithelial, immune, endothelial cells and fibroblasts) involved in alveolar injury by certain infections, effects that may be additive or synergize with actions of GCs.

4. Injurious mechanisms “compete” with protective effects of β2-AR agonists and GCs. Initial/moderate injury is potentially reversible, but severe injury may result in irreversible pathology and clinical progression, thus implying the need for early treatment before the onset of severe disease/injury.

5. Early treatment with inhaled LABAs (perhaps combined with ICS) prior to the onset of ARDS may counteract the pathologic feedback loops. By decreasing inflammation and cell death, those drugs help promote tissue repair and an effective immune response before severe pulmonary pathology occurs and progression to ARDS ensues.

6. This prophylactic strategy—early treatment with inhaled LABAs (and perhaps ICS)—for ALI/ARDS in pulmonary infections/damage in which progression of pathology involves an interplay between inflammation, injury, and coagulation thus represents a novel, potentially beneficial therapeutic approach.

V. Effects of β-Adrenergic Receptor Agonists on Pulmonary Cell Types in Alveolar Injury and Acute Respiratory Distress Syndrome

β-AR agonists preferentially activate β2-AR in cells in the lung (Figs. 1–3), and β2-ARs are expressed on multiple pulmonary and immune cell types that mediate diverse physiologic responses. These cell types include alveolar epithelial cells, pulmonary fibroblasts (FBs), endothelial cells (ECs), inflammatory cells associated with the innate immune response (e.g., dendritic cells, neutrophils, and macrophages), and adaptive immune cells (T cells and B cells).

E. Reduced Susceptibility to Epithelial Damage from Infectious Agents

Complementary early data showed a protective effect of β2-AR activation on epithelial damage involving AECs by bacterial products (Salathe, 2002). Supernatants from P. aeruginosa can reduce epithelial integrity in human AECs by decreasing expression of tight-junction proteins (Coraux et al., 2004), effects abolished by preincubation with an LABA. Human AECs infected with S. aureus have increased expression of inflammatory markers, such as IL-6, IL-8, and TNFα, mediated by nuclear factor κB (NF-κB) and activator protein 1 signaling; incubation of cells with a LABA/ICS combination blocks these responses (Fragaki et al., 2006). Similar results were obtained with human BECs treated with β2-AR agonists and corticosteroids (Edwards et al., 2006): mRNA expression of a range of inflammatory chemokines was reduced after exposure to RV16 rhinovirus. Beneficial effects of β2-AR agonists and corticosteroids were also noted in studies of RV16 rhinovirus infection in human BECs, with reduced production of the pro-remodeling factors fibroblast growth factor 2 and vascular endothelial growth factor (Volonaki et al., 2006). Studies with rhinovirus-infected primary human BECs support these results: β2-AR agonists plus corticosteroids reduced expression/release of proinflammatory factors without affecting viral replication; combination of the two drugs was synergistic (Bochkov et al., 2013). In vivo studies that tested the combined effects of β2-AR agonist and corticosteroid in a respiratory syncytial virus (RSV) model in mice also revealed a reduction in epithelial disruption and inflammation (Singam et al., 2006).

In studies with human BECs, β2-AR agonists reduced Mycoplasma pneumoniae titers in infected cells while increasing expression of genes associated with antimicrobial defense (Gross et al., 2010). These protective effects of β2-AR agonists were reduced by IL-13 treatment and were less in cells from patients with COPD, results consistent with the paradigm presented in Fig. 5 regarding a balance between β2-AR–promoted protective effects and severity of pathology. Protective responses of β2-AR agonists and corticosteroids were also shown in human airway glandular cells infected with S. aureus: treatment restored mucus secretion, increased ion transport, and decreased inflammatory cytokine secretion (Zahm et al., 2010).

• Download figure
• Open in new tab
• Download powerpoint

Fig. 5

The balance between AFC and alveolar edema plus inflammation. β2-AR agonists increase AFC and blunt edema by stimulating ion channel function directly and indirectly. Increases in edema and alveolar inflammation can reduce these protective effects by decreasing ion channel expression/activity, thereby worsening edema. Corticosteroids prevent this decrease in ion channels and may synergize with β2-AR agonists. β2-AR agonists and corticosteroids also reduce alveolar inflammation, cell death, and susceptibility to infection, thus protecting against edema and other consequences of alveolar injury.

Other studies with human AECs have demonstrated that β2-AR agonists alone and synergistically with corticosteroids reduce viral titers and inflammatory readouts in cells infected with RV14 rhinovirus (Yamaya et al., 2011, 2013, 2014). Multiple mechanisms appear to mediate these effects, including decreased expression of viral receptors, increased acidification of endosomes (which reduces viral action) and decreased activity of pathways, such as NF-κB, that promote inflammation. Similar findings have been obtained with human BECs infected with the coronavirus HCoV-229E (Yamaya et al., 2020).

Those findings are supported by studies with RSV-A2–infected human BECs treated with forskolin or cAMP analogs, which protected the epithelial barrier by decreasing cytoskeletal and morphologic changes induced by viral infection and prevented disassembly of intercellular junctions (Rezaee et al., 2017). Viral titer and mRNA replication were also reduced. BECs from patients with asthma indicated an antiviral role for cAMP in RV16 rhinovirus infection with reductions in cell death, viral replication, and inflammatory marker expression (Roth et al., 2017). Moreover, LABA-promoted increases in cAMP sensitized human BEC response to corticosteroids, lowering the concentration dependence of GC response element–mediated transcriptional effects (Kaur et al., 2008).

In summary, substantial data from studies with human AECs indicate that β2-AR agonists reduce barrier dysfunction and epithelial cell death, apoptosis, and microbial titer and can restore ion transport. Such effects in AECs and other pulmonary epithelial cells imply the potential utility of β2-AR/cAMP in blunting alveolar injury and potential progression to ARDS.

VI. Fibroblasts

Cyclic AMP has antifibrotic actions that include reduced profibrotic gene/protein expression and decreases in FB proliferation, migration/invasion, and secretion of proinflammatory factors. Such effects occur in FBs from multiple tissues, including lung, heart, and skin (Insel et al., 2012b; Lu et al., 2013; Zuo et al., 2019; Delaunay et al., 2019). Substantial data indicate that β2-AR agonists, via cAMP, promote an antifibrotic phenotype, including in studies of lung injuries that induce pulmonary fibrosis.

β2-AR agonists exert antifibrotic effects (decrease FB proliferation and expression of adhesion marker proteins associated with a profibrotic/remodeling phenotype) in human lung FBs in vitro (Silvestri et al., 2001). In that study, corticosteroids also reduced expression of adhesion markers. Treatment of human lung FBs with a β-AR agonist (isoproterenol), other agents that increase cAMP (PGE2, forskolin, PDE inhibition), or overexpression of AC6 decreased FB proliferation, collagen synthesis, and expression of collagen genes (Liu et al., 2004; Liu et al., 2010). β2-AR agonist and corticosteroid treatment reduce TGFβ-stimulated differentiation of human lung FBs to myofibroblasts (myoFBs) and TNFα-mediated IL-6 secretion (Baouz et al., 2005). LABA treatment of human lung FBs also blocks profibrotic and proinflammatory effects of TGFβ (Goulet et al., 2007). Combining SABAs/LABAs with corticosteroids was antifibrotic and antiproliferative in human bronchial FBs (Descalzi et al., 2008). Reduction in human lung FB proliferation is PKA-dependent (Kohyama et al., 2009).

β2-ARs are highly expressed in human lung FBs and FB cell lines, and LABAs reduce proliferation, collagen synthesis, and expression of α-smooth muscle actin (a fibrotic marker) (Lamyel et al., 2011). Roflumilast (a PDE4 inhibitor) potentiates LABA-promoted antifibrotic and anti-inflammatory effects in human lung FBs (Tannheimer et al., 2012b). Antifibrotic effects of combined β2-AR agonist/corticosteroid treatment also occur in fibrocytes: the drugs blunted the transformation of peripheral blood mononuclear cell–derived fibrocytes (isolated from patients with severe asthma) into myoFBs (Lo et al., 2017). A broad range of studies thus demonstrate antifibrotic effects of β2-AR stimulation, including in tandem with corticosteroids.

Other studies document the antifibrotic role of cAMP in lung FBs. Studies prior to 2012 (Insel et al., 2012b) identified antifibrotic effects by cAMP that include decreased proliferation, reduced FB-myoFB transformation, increased death of FBs/myoFBs, and reduced FB motility and expression/synthesis/release of extracellular matrix (ECM) components, including various collagens. Similar findings occur in FBs from other tissues, including heart, kidney, and skin. These effects appear to be mediated by EPAC (in particular, EPAC1) and PKA, with fibrotic transformation able to blunt such antifibrotic signaling.

The antifibrotic effects of PGE2, via EP2 and EP4 prostanoid receptors that raise cAMP, has been extensively studied (Bozyk and Moore, 2011) and reveal similar responses in lung FBs to those induced by β2-AR agonists and cAMP-elevating agents. PGE2 inhibits α-smooth muscle actin expression in human lung FBs by reducing nuclear accumulation of myocardin-related transcription factor-A and serum-response factor complexes and expression of serum-response factor mRNA, thereby inhibiting p38 MAP kinase activation and TGFβ response (Penke et al., 2014). PGE2 can exert antifibrotic effects (via EP2 and EP4) in human lung FBs from healthy donors and patients with idiopathic pulmonary fibrosis (IPF) by inhibiting Ca2+ signaling and decreasing TGFβ-promoted Ca2+ oscillations and activation of downstream kinases (Mukherjee et al., 2019). PGE2 can also reverse TGFβ-stimulated profibrotic changes in gene expression, suggesting actions in myoFBs (Wettlaufer et al., 2016). Lung epithelial cells produce PGE2; co-culture of epithelial cells and FBs reduces TGFβ-stimulated profibrotic effects in a PGE2-dependent manner (Epa et al., 2015).

Additional work on prostanoid receptors has shown that activation of prostacyclin receptors (PTGIR, a Gs-coupled GPCR) has antifibrotic effects in lung FBs by increasing cAMP, which can inhibit YAP/TAZ-mediated profibrotic transcriptional signaling and differentiation into myoFBs (Zmajkovicova et al., 2019). Inhibition of YAP/TAZ profibrotic signaling has also been observed for Gs-coupled dopamine D1 receptors, suggesting that YAP/TAZ inhibition may be a mechanism for antifibrotic effects of cAMP (Haak et al., 2019, 2020). Treprostinil (an agonist for EP2, EP4, and prostacyclin receptors) can inhibit platelet-derived growth factor-B and TGFβ-mediated profibrotic effects (e.g., expression of ECM proteins) and reduce lung FB proliferation (Lambers et al., 2018). Prostaglandin D2 and agonists for the prostanoid DP1-receptor (a Gs-coupled GPCR) have similar cAMP-dependent antifibrotic effects (e.g., decreasing TGFβ-driven transformation to myoFBs) (Ayabe et al., 2013). Thus, multiple Gs-coupled GPCRs exert cAMP-mediated antifibrotic effects in human lung FBs ex vivo.

In vivo studies using β2-AR agonists and other cAMP-elevating agents support these ex vivo data. Early evidence of the importance of β2-ARs in vivo in lung fibrosis was shown in models of lung injury in mice (induced by bleomycin, hyperoxia, or butylated hydroxytoluene): prior treatment of mice with the β-AR antagonist propranolol increased lung collagen (Lindenschmidt and Witschi, 1985). Similar results were obtained with rats treated with propranolol for 1–3 weeks (Smith and Smith, 1988). Treatment with propranolol increased collagen and interstitial fibrosis in the lung and produced alveolar remodeling. Subsequent studies indicated that propranolol treatment increases lactate dehydrogenase release in vivo and in fibroblasts cultured ex vivo (Smith and Smith, 1989) and induces fibrotic effects that involved ECs (Sommers Smith and Smith, 2002). LABA treatment has antifibrotic effects in lung FBs from normal subjects and patients with IPF (Herrmann et al., 2017). The latter study also tested inhalational LABA treatment of bleomycin-induced lung fibrosis in mice, using both a “preventive” treatment regimen (administration of LABA at the start of bleomycin treatment) or a “therapeutic” regimen (7 days after bleomycin exposure, i.e., once injury and fibrosis occurred). The authors found that LABA treatment in both regimens reduced weight loss, immune cell infiltration, and inflammatory and fibrotic marker expression in bronchoalveolar lavage fluid. Early LABA treatment also reduced collagen deposition and enhanced forced vital capacity. The investigators also tested LABA treatment in TGFβ1-overexpressing mice that spontaneously develop fibrosis; early, but not late, treatment with LABA had strong antifibrotic effects.

Treatment with inhaled LABA/ICS (formoterol/beclomethasone) has been tested in a 4-week pilot study of patients with IPF (Wright et al., 2017). LABA/ICS-treated patients had reductions in several readouts of coagulopathy and platelet activation, reduced sputum eosinophil counts, and improved 1-second forced expiratory volume. These pilot study data provide evidence in support of efforts to test LABA/ICS in settings that can cause lung injury associated with fibrosis and underscore the potential complementarity/synergy between LABAs and corticosteroid.

In vivo studies with other cAMP-raising agents (including other Gs-coupled GPCR agonists) support the findings with β2-AR agonists. Examples are as follows: dibutyryl cAMP was tested prophylactically in a bleomycin pulmonary fibrosis model in hamsters, with treatment commencing 48 hours prior to bleomycin exposure (O’Neill et al., 1992). Treatment with dibutyryl cAMP reduced the bleomycin-induced increase in inflammation and the rate of fibrotic lesion formation and growth. A more recent study that tested dcAMP in a silicosis model of pulmonary fibrosis in rats (Liu et al., 2017) showed reductions in Gαi expression and the extent of fibrotic scar formation along with enhanced abundance of PKA and phospho-CREB protein; ex vivo data indicated a cAMP-PKA-CREB mediated effect, with reductions in phosphorylated SMAD2/3 binding to CREB binding protein. In vivo inhibition of fibrosis by cAMP has been demonstrated in FB-specific AC6-overexpressing mice with bleomycin-mediated lung injury (Liu et al., 2010). AC6 overexpression also reduced immune cell infiltration in vivo, with ex vivo data showing enhanced response to β2-AR agonist stimulation; larger increases in cAMP produced stronger antifibrotic effects.

Substantial in vivo data support a role for prostanoid signaling in mitigating lung fibrosis via Gs-cAMP–driven mechanisms. Using a synthetic PGE2 analog in a bleomycin mouse model, Failla et al. (2009) treated the mice at the onset or 72 hours after bleomycin administration and observed a decrease in inflammation, lung injury, and fibrosis. By contrast, bleomycin-treated cyclooxygenase-2–deficient mice exposed to bleomycin had more severe inflammation and fibrosis, likely because of reduced PGE2 generation (Hodges et al., 2004; Lovgren et al., 2006). Protective effects of PGE2 are likely time-sensitive: Dackor et al. (2011) showed such effects if PGE2 or the PGE2 analog, iloprost, were administered to mice 7 days prior to, but not 14 days after, bleomycin treatment. Studies using iloprost (Zhu et al., 2010; Aytemur et al., 2012) or the corticosteroid methyl-prednisone (Aytemur et al., 2012) also showed that early administration blunted inflammation and fibrosis in bleomycin-treated mice. Orotracheal administration of mice with treprostinil prior to bleomycin exposure had protective antifibrotic and anti-inflammatory effects and reduced vascular remodeling and may protect from pulmonary hypertension induced by bleomycin injury (Nikitopoulou et al., 2019).

Novel delivery methods for PGE2 in fibrosis models further highlights this protective effect of prostanoids (and likely other cAMP-raising agents) in decreasing fibrosis in vivo. An example is inhalation of liposome-containing PGE2 after bleomycin exposure (Ivanova et al., 2013), which delivers drug to the lung and reduces fibrosis. Inhaled PGE2-containing nanoparticles (used with small interfering RNAs against certain profibrotic genes) can also protect from bleomycin-induced fibrosis in mice (Garbuzenko et al., 2017). Thus, multiple inhalational approaches may be ways to increase cAMP and treat lung fibrosis.

Other studies underscore the protective effects of prostanoid signaling against fibrosis in vivo. Examples include the following: 1) upregulation of PGE2 signaling can protect plasminogen activator inhibitor 1–deficient mice from lung fibrosis (Bauman et al., 2010); 2) synergistic protection from lung fibrosis by hepatocyte growth factor and PGE2s (Yoon et al., 2013); 3) inhibition of angiotensin II signaling by losartan enhances PGE2 production and associated antifibrotic effects (Molina-Molina et al., 2006); 4) mice deficient in microsomal prostaglandin E2 synthase-1 are susceptible to more severe bleomycin-induced fibrosis (Wei et al., 2014); 5) intravenous injection of retrovirally transfected FBs with prostaglandin synthase into bleomycin-exposed mice reduces lung injury, edema, and fibrosis (Ando et al., 2003); 6) greater susceptibility to bleomycin-induced fibrosis in SLCO2A1 (prostaglandin transporter)–deficient mice (Nakanishi et al., 2015); and 7) fibroblasts from patients with IPF are less susceptible to apoptosis, but PGE2 induces apoptosis in profibrotic cells and protects AECs from FAS-ligand–induced apoptosis, implying both an antifibrotic and epithelial-protective effect of PGE2 and cAMP (Maher et al., 2010).

Other Gs-coupled GPCR agonists and PDE inhibitors have antifibrotic actions in vivo. A peptide targeting relaxin receptors (RXFP1/LGR7) administered at the time of or 7 days after bleomycin treatment had protective effects; ex vivo studies in human cells showed that these effects are cAMP-mediated, consistent with antifibrotic responses to relaxin in other cell types (Pini et al., 2010). Relaxin also reduced fibrosis in bleomycin-exposed mice 14 days after exposure; ex vivo experiments in FBs derived from patients with IPF and in murine FBs showed that relaxin reduces myoFB contractility and RhoA signaling (Huang et al., 2011). Treatment of bleomycin-exposed mice with a DRD1 agonist 10 days after exposure reduced FB contractility, enhanced cAMP, and inhibited YAP/TAZ signaling (Haak et al., 2019). Inhibition of PDE4 immediately after bleomycin exposure can exert antifibrotic and anti-inflammatory effects in rats (Pan et al., 2009) and mice (Udalov et al., 2010). Treatment of bleomycin-exposed rats 10 days after exposure with a PDE4 inhibitor also alleviated fibrosis and vascular remodeling in vivo (Cortijo et al., 2009).

Thus, numerous studies with human and rodent lung FBs ex vivo and studies in vivo demonstrate that cAMP and Gs-coupled GPCRs, including β2-ARs and certain prostanoid receptors, reduce the profibrotic phenotype of lung FBs and blunt inflammation, vascular remodeling, and lung injury. The ex vivo data imply that raising cAMP concentrations reduces fibrotic signaling in FB/myoFB. In vivo findings likely result from actions of cAMP-elevating agents on multiple cell types besides fibroblasts, including AECs, endothelial cells, and immune/inflammatory cells. Antifibrotic effects appear to optimally occur if cAMP is increased near the onset of profibrotic stimulation, emphasizing the need for early intervention with β2-AR agonists and other agents and consistent with data above for alveolar epithelial injury and edema.

VII. Endothelial Cells

Numerous studies document a protective effect of cAMP on the pulmonary endothelium by enhancing the integrity of the endothelial barrier. Increase in cAMP reduces endothelial permeability, thereby decreasing the severity of edema and immune cell infiltration (Sayner, 2011; Birukova et al., 2013; Birukov and Karki, 2018; Millar et al., 2016; Liu et al., 2015; Claesson-Welsh et al., 2021). These barrier-protective effects of cAMP appear to be mediated by both PKA and EPAC and oppose the effects of RhoA, which enhances endothelial permeability. The barrier-protective effects of cAMP on ECs are of particular relevance in the context of pneumonia that leads to ALI/ARDS in which endothelial integrity is compromised (Gonzales et al., 2015; Millar et al., 2016; Matthay et al., 2019). EPAC has been shown to have protective effects on the endothelium, reducing vascular permeability and disruption and modulating endothelial inflammation in a range of settings (Roberts and Dart, 2014).

In ECs of other tissues, cAMP increases (including by β2-ARs) may promote angiogenesis and protect from apoptosis (Dormond and Rüegg, 2003; Namkoong et al., 2005; O’Leary et al., 2015; Sorriento et al., 2011). However, other findings suggest that cAMP may have opposite, concentration-dependent actions, perhaps mediated by p38 and AKT, respectively (Sorriento et al., 2011). Angiogenesis, a critical process during recovery after lung injury, including in ARDS, restores function (e.g., gas exchange) in the recovering lung (Voelkel et al., 2007), underscoring a potential benefit of β2-AR agonists during this postinjury phase.

The endothelial barrier-protective effects of cAMP are complicated by its role in subcellular compartments. Certain ACs increase cAMP concentration near the plasma membrane and produce barrier-protective effects (Sayner, 2011). By contrast, cytosolic pools of cAMP, perhaps derived from soluble AC, can have opposite actions in ECs. β2-ARs are functional in these cells, but previous reviews (e.g., Sayner, 2011; Millar et al., 2016; Birukov and Karki, 2018; Claesson-Welsh et al., 2021) that discuss effects of cAMP on the pulmonary endothelium have limited discussion of β2-ARs and such effects. β2-AR agonists regulate microvascular endothelial permeability, which is complementary to β2-AR agonist actions on AECs in regulating edema. Below we discuss examples of endothelial barrier-protective effects of β2-AR agonists.

β2-AR agonists can reduce pulmonary vascular permeability in humans (Basran et al., 1986): ten patients with ARDS were systemically administered a β2-AR agonist; radiolabeled transferrin was used to test for plasma protein extravasation and accumulation (PPA). β2-AR agonist treatment reduced PPA in patients who survived; nonresponders succumbed to ARDS. Thus, in a subset of patients, β2-AR agonists reduce vascular permeability. Administration of β2-AR agonists to sheep reduced thrombin-induced increases in protein flux (Minnear et al., 1985). In studies with bovine pulmonary arterial ECs, β2-AR agonists reduced endothelial permeability, including if enhanced by α-thrombin or thrombin peptides; response to a β2-AR agonist was abolished by a β2-AR antagonist (Minnear et al., 1993). An inhaled LABA inhibited histamine-induced increases in pulmonary PPA and reduced LPS-induced neutrophil accumulation in guinea pig lungs, indicating that inhaled LABAs can regulate microvascular endothelial permeability in vivo (Whelan et al., 1993). Systemic administration of a LABA also reduced endothelial gaps in rat tracheal ECs (Baluk and McDonald, 1994).

Data from isolated, perfused rat lungs showed a similar endothelial barrier protective effect, as assessed by high pulmonary venous pressures to measure increased capillary filtration rates (Parker and Ivey, 1997). β2-AR agonist infused in the pulmonary vein reduced capillary filtration rates and protected against fluid accumulation. Intravenous β2-AR agonist also reduced microvascular leakage/permeability (evaluated by estimating plasma protein extravasation) induced by antigen (Ikezono et al., 2005). Studies in rats, using a model of high-volume ventilation, yielded similar results: intraperitoneal β2-AR agonist treatment reduced microvascular permeability and helped reduce lung fluid accumulation (deProst et al., 2008). β2-AR agonists thus can decrease edema not only by effects on AFC (discussed above) but also by regulation of endothelial permeability. However, continuous systemic administration of β2-AR antagonists in vivo can induce EC death (Sommers Smith and Smith, 2002).

β2-ARs may have an endothelial-protective role in infection. Influenza A infection of ex vivo human pulmonary microvascular ECs increased permeability of endothelial monolayers, with similar increases in microvascular permeability noted in vivo in mice (Armstrong et al., 2012). Treatment with a LABA (ex vivo or in vivo) largely abrogated these influenza-induced changes in permeability and protected infected mice from pulmonary edema. As noted above, a complex relationship exists between the ability of β2-ARs to promote AFC and the decrease in this β2-AR–mediated effect by edema-associated pathology. Similarly, in the pulmonary endothelium, LPS treatment of cultured rat ECs reduces β2-AR expression at the cell surface (by promoting internalization) and increases endothelial permeability (Yang et al., 2015). Conversely, pretreatment of ECs with a β2-AR agonist prevents LPS-mediated increases in endothelial permeability.

Additional protective effects of β2-AR in pulmonary ECs may result from increased nitric oxide synthase (NOS) activity and nitric oxide (NO) production (Horvath and Wanner, 2006). Inhaled NO has been tested as a therapeutic approach in ALI/ARDS (Gebistorf et al., 2016). Increased NO production and vasodilation in response to β2-AR agonists was shown in pulmonary arteries of normoxic mice and mice exposed to chronic hypoxia (mimicking a pulmonary hypertensive phenotype) (Leblais et al., 2008). Similar effects were noted in pulmonary arteries of mice, with β2-AR stimulation increasing endothelial NO production and cGMP levels and reducing PDE5 expression (Davel et al. 2015). Certain effects of β2-AR signaling in ECs may occur by signaling from Src-kinase in murine pulmonary arterial caveolae, leading to PI3K/Akt phosphorylation and NOS activation (Banquet et al., 2011). Evidence in pulmonary microvascular ECs is sparse, but β2-AR–mediated NO production has been shown in microvascular ECs from other tissues (Conti et al., 2013), suggesting that this effect also occurs in microvascular ECs in the vicinity of alveoli.

A contributing factor to angiogenesis promoted by β2-ARs is their stimulation of endothelial progenitor cells (EPCs), as precursors for angiogenesis and re-endothelialization. β2-AR agonists stimulate proliferation and migration and increase vascular endothelial growth factor release in EPCs isolated from mice and rats (Galasso et al., 2013). Injection of EPCs from wild-type, but not β2-AR knockout, mice exposed to ischemic injury increased angiogenic response. Ke et al. (2016) observed a similar proangiogenic effect with overexpression of β2-ARs in human-derived EPCs studied in vitro, with increases in proliferation, migration, adhesion, and endothelial nitric oxide synthase activity. Further experiments in mice using a carotid artery injury model showed that β2-AR–overexpressing EPCs induce greater re-endothelization. The latter findings were replicated and extended in studies that revealed enhancement by β2-ARs in shear stress–induced increases in AKT phosphorylation and endothelial nitric oxide synthase activity (Hu et al., 2020), implicating β2-ARs in promotion of EPC-mediated endothelial restoration.

Both ex vivo and in vivo studies thus document endothelial barrier-protective effects of cAMP and β2-ARs. The findings include regulation of endothelial permeability that contributes to protection from edema along with other protective effects on the endothelium and alveoli. β2-ARs/cAMP also likely promote NO production, angiogenesis, and re-endothelialization, thus helping to restore function after inflammation while possibly also inhibiting endothelial apoptosis.

VIII. Immune Cells

β2-ARs are expressed on numerous immune cell types, including dendritic cells, macrophages, monocytes, neutrophils, and others (Theron et al., 2013; Padro and Sanders, 2014; Arumugham et al., 2017). β2-ARs are the predominant AR expressed on B cells and T cells and play a modulatory role on the adaptive immune response (Padro and Sanders, 2014). The general paradigm for the actions of β2-AR activation and cAMP in these cell types is suppression of inflammation. The localized delivery of β2-AR agonists (i.e., inhaled LABAs) may thus help mitigate increases in pulmonary hyperinflammation that can occur in pulmonary infections, such as certain viral and bacterial infections, thereby decreasing disease severity. Other approaches to increase cAMP, e.g., PDE inhibition, can also reduce inflammation in the lungs, for example, infections by rhinovirus (Edwards et al., 2016) or influenza (Sharma et al., 2013). We discuss the evidence for immune-modulating effects of β2-AR activation and cAMP in the following text and accompanying tables. Figure 6 summarizes actions of β2-ARs/cAMP signaling in adaptive and innate immune cells. These observations are relevant to pulmonary inflammation and, more generally, as effects of β2-AR/cAMP signaling.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 6

The effect of β2-AR/cAMP signaling in adaptive and innate immune cells. IFNγ, Interferon gamma; LTB4, Leukotriene B4; MIP1-α, macrophage inflammatory protein 1-alpha; TCR, T cell receptor.

IX. Do Inhaled Long-Acting β-Adrenergic Agonists Reduce Pulmonary Inflammation in Patients?

Preclinical studies of the effects of β2-AR/cAMP in immune cells reviewed above indicate that β2-AR agonists exert multiple anti-inflammatory effects on both innate and adaptive immune cell types. Consistent with those observations, numerous studies have examined the impact of inhaled LABAs on inflammatory markers in patients (primarily those with COPD or asthma). Anti-inflammatory effects have been observed, as discussed in examples below.

Bronchial biopsies and sputum from patients with COPD treated with LABA/ICS had a decrease in the number of neutrophils, eosinophils (Bathoorn et al., 2009), CD45+ cells (leukocytes), CD4+ T cells, and TNFα- and interferon-γ–expressing cells, implying multiple anti-inflammatory effects in vivo (Barnes et al., 2006). Moreover, bronchial biopsies from patients with COPD indicated that LABA/ICS, but not ICS treatment alone, reduced the numbers of infiltrating macrophages and CD8+ T cells (Bourbeau et al., 2007). Inhaled LABAs/ICS can reduce inflammatory cell number in sputum while increasing the proportion of circulating FOXP3+ T reg cells among CD4+ T cells (Yang et al., 2011). Use of inhaled LABA by patients with COPD has been associated with reduced ROS production in peripheral blood neutrophils (Santus et al, 2012). Similar effects were noted in circulating lymphocytes: administration of inhaled LABA/ICS acutely reduced inflammatory (LPS) response of lymphocytes isolated from patients with asthma (Rudiger et al., 2013), consistent with ex vivo preclinical data that showed that LABAs blunt LPS-mediated inflammation. These results agree with other data that show inhalation of LPS increased proinflammatory factors (IL-6, IL-8, TNFα, etc.) in bronchiolar lavage fluid, along with neutrophil influx, protein leakage, and macrophage activation, effects suppressed in subjects pretreated with an inhaled LABA (Maris et al., 2005). Administration of LABA/ICS in patients with COPD exacerbation was associated with reduced levels of proinflammatory IL-8 and increased anti-inflammatory IL-10 in sputum (Feng et al., 2013), findings akin to preclinical data showing that LABAs increase IL-10. LABA/ICS administration also reduces eosinophil number and exhaled NO (markers of inflammation) in patients with asthma (Nolte et al., 2013). Acute exposure of patients with COPD to inhaled LABAs increased nuclear translocation of GC receptors in sputum macrophages, implying sensitization/synergy between LABA and ICS (Haque et al., 2013). Treatment of asthmatic patients with LABAs/ICS reduced small airway inflammation and peripheral pulmonary inflammation, as assessed by impulse oscillometry (Hozawa et al., 2014).

Inhaled LABAs (formoterol or salmeterol) can reduce exhaled alveolar fraction of NO in patients with COPD (Santus et al., 2018). Recent advances in metabolomic techniques have facilitated the analysis of airway breath condensates using “breathomics”: extrafine LABA/ICS formulations show anti-inflammatory effects associated with changes in the composition of exhaled breath condensates (Montuschi et al., 2018). Extrafine LABA/ICS formulations may reduce peripheral, small airway inflammation in the lung (Pirina et al., 2018) and would be of interest to assess in ARDS.

Treatment of patients with PDE4 inhibitors recapitulates many of the effects of LABAs on pulmonary inflammation: e.g., reductions in sputum neutrophils and eosinophils (Grootendorst et al., 2007; Rabe et al., 2018), infiltration of those cells into airways (Hohlfeld et al., 2008), circulating eosinophils (Martinez et al., 2018), and proinflammatory factors in sputum (Singh et al., 2019). These observations support the idea that anti-inflammatory effects in human lung of β2-AR agonists result from elevated cAMP levels. A number of cell-specific effects remain to be resolved in humans (e.g., Th1, Th2, and Th17 differentiation, certain effects on DCs, M1–M2 switching, etc.), but the anti-inflammatory actions of β2-AR agonists translates from preclinical to clinical settings. This response likely occurs via effects on airway and alveolar cells, plus by direct effects on immune cells.

X. Do β2-Adrenergic Receptors Cause Pathologic Pulmonary Immune Suppression in Patients?

As described above, β2-AR agonists suppress a range of immune cell types but in some cases promote immune response (e.g., adaptive responses), such as in B cells. An important concern regarding the potential utility of inhaled LABAs in treating/preventing ARDS is whether such treatment might worsen infections and/or increase the risk of secondary infections, thereby enhancing disease severity. This concern is especially relevant, given preclinical data discussed above, indicating that LABAs reduce bacterial phagocytosis by various immune cell types.

Numerous clinical trials, especially with patients with COPD, have tested if LABA monotherapy or in combination with ICS or LAMAs increase the frequency of respiratory infections, perhaps as a consequence of chronic treatment being immunosuppression in the lungs. Based on concerns regarding fatal exacerbations in patients with asthma administered LABA monotherapy, LABAs are typically only administered in combination with other drugs (e.g., ICS) (Cloutier et al., 2020; Global Initiative for Asthma, 2020). Limited information is available regarding whether LABAs alone increase the risk of respiratory infections in patients with asthma. In the text below and Table 8, we discuss recent RCTs and meta-analyses that have focused on patients with COPD.

Therapeutic approaches in such RCTs include inhaled LABAs, ICS, and LAMAs as monotherapies, as well as combinations (including LABA/LABA/LAMA, triple therapy). Table 8 lists selected RCTs [in particular, large multicenter RCTs, notably the TORCH (Calverley et al., 2007) and SUMMIT trials (Vestbo et al., 2016)] and recent meta-analyses. The overall conclusion is that LABA treatment as monotherapy or combined with LAMAs does not increase the risk of respiratory infections. Moreover, increasing LABA dose does not correlate with an increase in risk of infections. By contrast, prolonged ICS therapy appears to increase the risk of pneumonia, both when used alone or in combination with LABAs, although the SUMMIT trial suggests no increase in pneumonia by ICS treatment (Vestbo et al., 2016). Meta-analyses in the Cochrane Database of Systematic Reviews have considered the risk of serious AEs in COPD and have compared LABAs with placebo (Kew et al., 2013) and LABA/LAMA to placebo (Maqsood et al., 2019). Neither study found an increase in serious AEs of LABAs versus placebo; pneumonia risk was not quantified separately, but the studies included this risk among the serious AEs. A meta-analysis of 20 RCTs reached a similar conclusion: no risk of infections with LABA monotherapy in COPD and minimal risk of serious AEs (Decramer et al., 2013).

Although a theoretical concern exists regarding the immunosuppressive effects of inhaled LABAs, the findings in patients with COPD appear to allay these concerns. Perhaps such results derive from immune-enhancing effects of LABAs discussed above or from their prevention of excessive inflammation, which helps protect immune response while also promoting epithelial barrier integrity. Hence, available evidence implies that inhaled LABAs are unlikely to meaningfully compromise immune response to infection in the setting of acute infection–induced ALI/ARDS. However, this may not be the case if inhaled LABAs are combined with ICS in such settings, especially once seroconversion begins. The examples discussed above in which addition of ICS is associated with increased risk of pneumonia involve chronic exposure to ICS. Data for acute administration of ICS/LABA combinations in ARDS do not indicate increased risk of pneumonia (e.g., Mangi et al., 2015; Festic et al., 2017). Of note, emerging evidence from trials in patients with COVID-19 indicates the benefit of early ICS administration in preventing severe disease (Ramakrishnan et al., 2021; the STOIC trial), supporting the hypothesis that combinations of inhaled medications that suppress inflammation may protect patients from progressive pulmonary injury. Further support for this hypothesis is provided by interim results from the PRINICIPLE trial (Lee2021), which is testing the effects of adding ICS to the standard of care for patients with COVID-19 with risk factors for developing severe disease. ICS had a significant improvement in the time to recovery of these patients and also indicated the likelihood of reduced mortality and rates of admission to ICUs.

XI. Platelets

Elevation of cAMP inhibits platelet activation, aggregation, adhesion, and proliferation. These responses from increased cAMP occur via PKA, inhibition of Ca2+, and RhoA signaling (Noé et al., 2010; Aburima et al., 2013; Fuentes and Palomo, 2014; Nagy and Smolenski, 2018). Gs-coupled GPCRs thus inhibit platelet activation, whereas Gq/11-coupled GPCRs (which increase Ca2+), G12/13-coupled GPCRs (which promote RhoA signaling), and Gi-coupled GPCRs (which lower cAMP) promote platelet activation and coagulation. Inhibitors of Gi- and Gq-coupled GPCRs [e.g., protease-activated (thrombin) receptors and P2Y12 purinergic (ADP) receptors] are widely prescribed as antiplatelet drugs (Sriram and Insel, 2021). PDE inhibition has also been studied as a means to reduce platelet activation; platelets highly express PDE3, which is cAMP-selective (Fuentes and Palomo, 2014). Raising platelet cAMP via agonists for Gs-coupled A2A and A2B receptors has also been suggested (Wolska and Rozalski, 2019).

Older studies reported that β2-AR activation inhibits platelet aggregation (Woulfe, 2005; Anfossi et al., 1996). Study of β2-ARs in platelets has been sparse in recent years, perhaps because epinephrine promotes platelet activation via α2a-ARs, a Gi-coupled GPCR that is the more prominently expressed AR in platelets (Motulsky and Insel, 1982; Woulfe, 2005; Anfossi et al., 1996).

β2-ARs were identified on human platelet membranes using radioligand ([I125]-iodopindolol) binding, together with β1-AR– and β2-AR–selective antagonists (Wang and Brodde, 1985). Isoproterenol, likely via β2-ARs, can increase cAMP and blunt aggregation of human and rat platelets (Yu and Latour, 1977; Kerry and Scrutton, 1983; Winther et al., 1985; Andersson and Vinge, 1991). Isoproterenol inhibits ADP-induced aggregation of platelets from healthy and diabetic rats (Yu and Latour, 1977; Austin et al., 1995) and thrombin-induced release of serotonin from rat platelets (Koutouzov et al., 1985). Subsequent work demonstrated that activation of β2-ARs in human platelets enhances NOS activity (NO production); β2-ARs inhibit thrombin-induced platelet adhesion to endothelial cells in a platelet-expressed, NOS-dependent manner; β2-ARs also reduced thrombin-mediated platelet aggregation in a NOS-independent manner (Queen et al., 2000). Human subjects acutely given systemic doses of β2-AR agonists had reduced platelet aggregation (Larsson et al., 1992) and platelet number (Fredén et al., 1978). Recent data indicate disruption of platelet indices in patients with COPD: reduced mean platelet width and volume, interpreted as markers of thromboinflammation, associated with increased platelet function at sites of inflammation (Hlapčić et al., 2020). LABA/ICS therapy appears to restore these platelet indices, although whether this is an action on platelets or an indirect anti-inflammatory effect requires further study.

The regulation of platelets by β-AR agonists is a potentially overlooked effect of inhaled LABAs, i.e., on platelets in the lung, especially in capillaries in the vicinity of alveoli, that merits further study. Patients with IPF show improvements in markers of coagulation and platelet activation after treatment with inhaled LABAs/ICS (Wright et al., 2017). Crosstalk between signaling pathways associated with inflammation may promote coagulation and pulmonary thromboinflammation in certain viral infections (Sriram and Insel, 2020b, Sriram and Insel, 2021). If LABAs exert anti-inflammatory effects in the lung, those agents may reduce platelet activation and help explain the findings in patients with IPF. Existing evidence thus provides a basis to suggest that LABAs may have inhibitory effects on platelet aggregation, but this idea requires additional study.

XII. Summary and Conclusions

Figure 7 summarizes the effect of β2-AR activation/cAMP on a range of cell types involved in ARDS pathobiology (Fig. 4). In this review, we identify evidence (from experimental ex vivo and in vivo studies plus clinical data) in support of potential efficacy of β2-AR agonists, in particular LABAs, and perhaps other agents that increase cAMP, e.g., other Gs-coupled GPCR agonists (Sun and Ye, 2012; Wendell et al., 2020), and PDE [e.g., PDE4] inhibitors (Phillips, 2020; Giorgi et al., 2020), in the setting of ALI/ARDS. Beneficial effects of β2-AR agonists/cAMP include protection of epithelia (including in AECs) and reductions in profibrotic response of lung FBs, in inflammatory responses in multiple cell types, activation of innate immune cells (with possible potentiation of adaptive immune responses), and EC injury (accompanied by increased endothelial repair). In addition, some evidence implies a reduction in platelet activation. Thus, thromboinflammation and accompanying tissue injury—characteristics of ARDS—are likely blunted by β2-AR agonists/cAMP-elevating agents. Key advantages of β2-AR agonists include the expression of β2-ARs on multiple cell types that participate in ARDS pathobiology and the widespread use of inhaled LABAs as approved therapies, thus offering the opportunity to repurpose such agonists.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 7.

Based on the framework for cell-cell interactions and cell-based effects that drive alveolar injury presented in Figure 4, the effects of β2-AR agonists (mediated by cAMP) on the same cell types that are involved in ARDS/ALI pathobiology.

β2-AR agonists (especially inhaled LABAs) can reduce pulmonary inflammation apparently without harmful immune suppression. Inhaled LABAs are widely used as monotherapies in treating COPD with few serious AEs, but safety concerns, especially regarding LABA monotherapy in asthma, led to the use of combination therapies with ICS and/or LAMAs. What are the implications with respect to LABAs in ARDS? The safety risks associated with LABA use in asthma are largely skewed toward fatal exacerbations in juvenile patients (McMahon et al., 2011; Cates et al., 2012; Kersten et al., 2017). By contrast, serious viral infections, e.g., influenza and SARS coronaviruses, that can result in ARDS occur more commonly in older subjects.

The evidence we have discussed suggests that inhaled LABA monotherapy and/or combination therapies should be tested in settings of infections associated with ARDS. Such an approach would be especially appropriate for adults, especially those over age 60, who have higher mortality with infections that can lead to ARDS. LABAs in combination with ICS and/or LAMAs might be assessed for their ability to prevent disease progression and improve the outcome of patients with ARDS.

Importantly, we interpret the available data on action of LABAs on various cell types, combined with current understanding of the pathobiology of ARDS, as providing a strong rationale to administer LABAs early in the course of the disease. Such an approach will likely maximize the probability that beneficial effects—such as restoration of AFC, prevention of epithelial and endothelial injury, inhibition of fibroproliferative effects, and suppression of hyperinflammation—can occur before injury reaches an irreversible stage. Early treatment with inhaled LABAs is thus a potential prophylaxis against severe lung injury. Limited evidence in recent studies supports this hypothesis (Festic et al., 2017; Fouad et al., 2020) and contrasts with previous trials [e.g., the BALTI trials (Perkins et al., 2006; Gao Smith et al., 2012; Gates et al., 2013)] in which treatment was initiated in patients with ARDS who were already in intensive care. A prophylactic strategy that focuses on early treatment with LABAs may offer a means to improve the currently poor outcomes and high mortality of patients with ARDS. This approach includes treatment of COVID-19, where disease progression to ARDS is characterized by the development of hyperinflammation and thromboinflammation (Sriram and Insel, 2021). Those features may be slowed/prevented by early treatment with LABAs, potentially in combination with ICS or LAMAs. Indeed, initial data from clinical trials suggests the utility of ICS treatment in patients with COVID-19, including use early in the disease (Ramakrishnan et al., 2021; Lee, 2021).