Abstract
Bronchial asthma (BA) is a heterogeneous disease. Some patients benefit greatly from glucocorticoid (GC) treatment, whereas others are non-responders. This could be attributable to differences in pathobiology. Thus, predicting the responses to GC treatment in patients with BA is necessary to increase the success rates of GC therapy and avoid adverse effects. The sustained inflammation in BA decreases glucocorticoid receptor (GR, NR3C1) function. Meanwhile, GRβ overexpression might contribute to GC resistance. Important factors in decreased GR function include p38 mitogen-activated protein kinase-dependent GR phosphorylated at Ser226, reduced expression of histone deacetylase 2 following activation of the phosphatidylinositol 3-kinase-δ signaling pathway, and increased nuclear factor-kappa B activity. MicroRNAs, which are involved in GC sensitivity, are considered biomarkers of the response to inhaled GCs. Some studies revealed that inflammatory phenotypes and disease-related modifiable factors, including infections, the airway microbiome, mental stress, smoking, and obesity, regulate individual sensitivity to GCs. Therefore, future investigations are warranted to improve treatment outcomes.
Keywords
History of glucocorticoid use in bronchial asthma (BA)
The discovery of glucocorticoids (GCs) in 1949 enabled the successful treatment of patients with rheumatoid arthritis. In the following decades, systemic and aerosol GCs, which have been developed since the 1970s, have been widely administered to treat BA owing to their proven efficacy in mitigating symptoms.1,2
From structure-based design, the therapeutic and side effect profiles of highly potent GCs, such as VSG158, 3 as well as selective glucocorticoid receptor (GR) agonists and modulators, 4 have been revealed. GW870086X, AZD5423, and AZD7594, which retain the anti-inflammatory effects of classical GCs but cause fewer or less severe side effects, have been studied in clinical trials for asthma treatment.5,6 These drugs are expected to improve the effectiveness of treating GC-resistant cases.
Poor GC sensitivity in BA
Therapeutic doses are often titrated upward or downward to achieve the desired clinical effects. However, resistance to GCs has occurred at therapeutic doses.7,8 Approximately 5% to 10% of patients with asthma respond poorly to the maximal dose of inhaled GCs. 9
Mechanism of action of GCs
GCs are e anti-inflammatory agents. 10 The main mechanism of action is interference with pro-inflammatory transcription factors such as nuclear factor-kappa B (NF-κB) and activator protein 1, a mechanism also known as GR-dependent transrepression. This action is also mediated by the binding of GR to glucocorticoid response elements (GREs) to regulate the transcription of specific genes, including mitogen-activated kinase phosphatase (MKP-1), which selectively inactivates p38 mitogen-activated protein kinase (MAPK), a process termed GR-dependent transactivation.
Inflammatory phenotypes involved in GC sensitivity in BA
Differences in the inflammatory patterns cause heterogeneity in clinical phenotypes and responsiveness to GCs in asthma. T-helper type 2 (Th2) inflammation, which is mediated by type 2 cytokines such as interleukin (IL)-4, IL-5, and IL-13, is a common type of BA generally associated with eosinophilic inflammation. GCs decrease the number of circulating eosinophils within hours of administration through the C-X-C chemokine receptor type 4 (CXCR-4)-dependent migration of eosinophils to bone marrow. 11 GCs also induce eosinophil apoptosis within hours of cell culture in vitro; however, IL-5 protects cells from GC-induced cell death. 12
Innate lymphoid type-2 cells (ILC2s) contribute to inflammation by enhancing the activities of Th2 cells, eosinophils, and their cytokines. IL-13+ ILC2s in the peripheral blood are associated with asthma severity and GC resistance in humans. 13 The activation cytokines, namely IL‐25 and IL‐33, and co-activation cytokine thymic stromal lymphopoietin (TSLP), produced by epithelial cells, result in GC resistance in ILC2s. 14 TSLP is involved in GC resistance in ILC2s through the activation of the anti-apoptotic molecule Bcl-xL via the signal transducer and activator of transcription 5 (STAT5) pathway and induction of STAT5 and MAPK, which induces resistance to GCs. 15
Non-Th2 inflammation encompasses the neutrophilic and paucigranulocytic categories. Interferon (IFN)-γ or IL-17 derived from Th1 cells, Th17 cells, or ILC3s is involved in this process. 16 High IL-17A and IFN-γ immunophenotypes were observed in patients with poor responsiveness to GCs. 17 Th17 cells cause GC-resistant airway inflammation when adoptively transferred to immunodeficient ovalbumin-challenged mice. 18 Neutrophils from GC-resistant patients with asthma exhibited reduced induction and inhibition of MKP-1 and IL-8, respectively, following exposure to dexamethasone. 19 Dexamethasone treatment exacerbated neutrophilic airway inflammation when colony-stimulating factor 3 was highly induced, which has been associated with GC resistance. 20 Neutrophil extracellular traps during neutrophil death weaken the efficacy of GC treatment by promoting inflammation and delaying tissue repair. 21 Although GC suppresses neutrophil apoptosis, the mechanism underlying GC resistance is unclear. 22
Modulation of GR function is involved in GC sensitivity
GR mediates the action of GCs by acting as a ligand-dependent transcription factor. GRα is a classically functional GR, whereas GRβ, which has a divergent carboxyl terminal in the ligand-binding domain that prevents it from binding to GCs, interferes with GRα-mediated transactivation. GR-β controls the expression of histone deacetylase 2 (HDAC2) by inhibiting GREs in its promoter. 23 Increased GRβ expression in airway T cells is associated with GC resistance. 24
The modulation of site-specific GR phosphorylation by various kinases, including MAPK, is involved in GC responsiveness. 25 GR phosphorylation at Ser211 enhances GR transcriptional activity, whereas phosphorylation at Ser226 is associated with reduced GR function.
Protein phosphatases (PPs) regulate GR phosphorylation in immune and non-immune cells. PP2A enhances GR nuclear translocation by dephosphorylating Jun N-terminal kinase, thereby decreasing GR-Ser226 phosphorylation. 26 In certain conditions in which striatin-3 is present, PP2A can also dephosphorylate GR at Ser211, which results in diminished GR transcriptional activity. 27 Meanwhile, PP5 is believed to negatively affect GR function. Eosinophil resistance to GC-induced apoptosis is correlated with the activation of PP5 and dephosphorylation of GR. 28 The interference of cytokine-induced PP5 by GR phosphorylation at Ser211 partly explains GC resistance in airway smooth muscle cells. 29
Mucin 1 (MUC1) inhibits the phosphorylation of GR at Ser226 and also forms a complex with GR to promote its translocation to the nucleus, thereby increasing the expression of anti-inflammatory genes. GR phosphorylation at Ser226 was increased, and the MUC1-CT–GR complex was downregulated in human bronchial epithelial (HBE) cells and blood neutrophils from patients with severe uncontrolled asthma. 30 GC resistance, which was caused by MUC1 deficiency in tumor necrosis factor (TNF)-α-induced necroptosis in 16HBE cells, was attributable to impaired GR nuclear translocation and inhibited NF-κB phospho-p65 expression. 31
MicroRNAs, which are small non-coding RNA molecules, operate as post-transcriptional regulators. Previous reports demonstrated that treatment with GC induces the expression of miR-124 but fails to correct miRNA abnormalities in airway epithelial cells. 32 For several models of GC-resistant airway hyperresponsiveness, miR-9 inhibition restored GC sensitivity by reducing PP2A activity. 33 Circulating miR-155-5p and miR-532-5p are prognostic biomarkers of the response to inhaled GCs, with miR-155-5p and miR-532-5p decreasing and increasing the transrepression of NF-κB, respectively. 34 A recent report demonstrated that benralizumab restores the expression of genes and miRNAs involved in GC sensitivity. 35
The overexpression of P-glycoprotein and reduced expression of HDAC2 play a pivotal role in the regulation of GC action. 36 The phosphorylation and inactivation of HDAC2 are mediated by phosphoinositide 3-kinase δ (PI3K δ), and PI3K inhibitors ameliorate GC insensitivity. 37
Non-modifiable factors of GC sensitivity: Genetic susceptibility
The response to GCs is too complex to be conditioned by a few genetic variants. 38 Variations in the genes participating in GC secretion and GR signaling are being increasingly studied because they are suspected to elicit the heterogeneity in response to GC therapy in patients with asthma. The responses involve corticotrophin-releasing hormone receptor 1, NR3C1, stress-induced phosphoprotein 1, FK506-binding protein 51, glucocorticoid-induced 1, histone deacetylase 1, HDAC2, and dual-specificity phosphatase 1. Genes associated with BA susceptibility and GC-resistant BA include orosomucoid-like 3, vascular endothelial growth factor, T-box transcription factor 21, Fc fragment of IgE receptor II, and HSD3B1.
Most notably, a large study of 2672 patients with asthma treated with fluticasone furoate or fluticasone propionate did not reveal any single-nucleotide polymorphisms that were determinative of the outcomes following the inhalation of GCs. 39
Modifiable factors of GC sensitivity
Viral infections
Viral infections in the respiratory tract are regarded as key triggers of asthma exacerbation. Prior studies reported that viral infections interfere with the GR response in the host by modulating the production of inflammatory cytokines, IFNs, and growth factors, 40 activating inflammatory pathways involving MAPK and NF-κB, 41 and manipulating the output of viral proteins 42 and RNA. 43
Influence of airway microbiota on responses to GCs
Dysbiosis attributed to a shift toward certain phyla and genera owing to the enhanced replication and persistence of certain bacteria is observed in patients with asthma. 44 The respiratory microbiome of patients with asthma has lower bacterial diversity than that of healthy individuals. Studies have reported an abundance of Proteobacteria relative to Bacteroidetes, Firmicutes, and Acinetobacter in the bronchoalveolar lavage (BAL) or sputum of patients with BA.45,46 Neutrophilic asthma is associated with the bacterial burden. 47 Proteobacteria can elevate the production of IL-8 and TNF-α, as well as recruit TSLP and neutrophils to inflamed airways. 48 Patients with asthma and increased levels of type 1 mediators reportedly have high levels of Actinobacteria and Firmicutes. 49
Bacterial diversity and certain airway microbiota might be involved in the response to GCs. 50 The bronchial epithelial expression of the FKBP5 gene is correlated with bacterial diversity and Actinobacteria counts in patients with stable BA, whereas the expression of Th17-related genes was associated with Proteobacteria. 51 Haemophilus parainfluenzae, which can activate Toll-like receptor 4 (TLR4), can reduce the cellular responses to GCs by activating transforming growth factor-β-associated kinase-1 to suppress GR inhibition. 52 However, whether such findings are related to GC resistance in vivo remains controversial. 53
The aforementioned organisms largely belong to the same bacterial phylum, and they are implicated in different phenotypic features. However, the extent to which the change in microbiota composition reflects inflammation or affects the response to GC treatment remains unclear.
Mental stress in GC-resistant asthma
Although stress can make a person more adaptable to various circumstances, it can also cause mental and physical disorders. Chronic stress results in resistance to GCs.54,55 Mental stress increases the risk of developing BA 56 and the occurrence of asthmatic attacks in children. 57 Possible mechanisms include the affected hypothalamic–pituitary–adrenal axis, the dysregulated immune system, 58 and an attenuated response. In an in vivo murine model of combined social disruption stress and allergic sensitization, exposure to repeated social stress before allergen inhalation enhanced and prolonged airway inflammation and altered corticosterone responsiveness, with GR binding to GREs and reduced GR mRNA and protein expression. 59 An association between exposure to maternal distress and attenuated GC responsiveness in children with BA has also been reported. 60 Children with BA experienced a 5.5-fold reduction in GR mRNA transcription when exposed to acute and chronic stressors simultaneously in leukocyte specimens. 61 Moreover, children who do not feel supported or understood by their parents were more resistant to the anti-inflammatory effects of GCs and demonstrated higher levels of eosinophil cationic proteins. 62
Smoking
Oxidative stress is a major cause of chronic airway inflammation, and it can significantly suppress GC sensitivity. Oxidative stress-induced GC insensitivity is associated with p38 MAPK, PI3K/Akt, and nuclear factor erythroid 2-related factor 2 signaling. 63 Smoking results in oxidative stress, and has been reported to increase the incidence and severity of and enhance the resistance to GCs in patients with asthma. However, patients who quit smoking displayed improved lung function and responsiveness to GC following smoking cessation, 64 suggesting that the resistance mechanisms are reversible.
The sputum cytokine concentrations of IL-17 and IFNα did not change in response to dexamethasone in smokers. 65 Cigarette smoking reduces HDAC2 expression, enhances cytokine expression, and inhibits GC activity in alveolar macrophages 66 and cultured bronchial cells. 67 Moreover, the GRα:GRβ ratio in peripheral blood mononuclear cells (PBMCs) is reduced in cigarette smokers. 68 CpG oligodeoxynucleotides, which are unmethylated CpG dinucleotides that mimic the immunostimulatory effects of bacterial DNA and stimulate TLR9, restore GC sensitivity, and block retinoid-related orphan nuclear receptor t-induced upregulation of IL-17 in patients with cigarette smoke-induced asthma in vivo, as well as in cigarette smoke extract-induced HBE cells, possibly through the restoration of HDAC2 levels and activity. 69
Obesity
Obesity is characterized by inflammation resulting from the secretion of cytokines and macrophage infiltration into adipose tissue. Inflammatory processes and immune responses bridge the gap between obesity and asthma. 70 Among the main adipokines, pro-inflammatory leptin and anti-inflammatory adiponectin are involved in the development of obese asthma.
Asthma in obese patients can be classified into late- and early-onset phenotypes. Late-onset obese asthma occurs mostly in female patients, and it has less allergic airway inflammation involving neutrophilic inflammation. Meanwhile, early-onset obese asthma is characterized by a type 2 phenotype, in which eosinophils might play an important role.
A recent study suggested that autophagy contributes to the exacerbation of eosinophilic inflammation in obese asthma by the development of TSLP- and IL33-dependent eosinophilic inflammation. 71 Mouse studies suggested that the nucleotide-binding domain, leucine-rich repeat-containing family, and pyrin domain-containing-3–IL1β–IL17 axis elicit the relationship between excess adiposity and non-atopic adult-onset asthma.72,73
Obesity is associated with poorer BA control.74,75 GRα, GRβ, and mineralocorticoid receptors are pivotal candidates for lipid metabolism and GC resistance in obesity. 76 Individuals with BA and obesity have reduced expression of MKP-1 and GR in PBMCs and BAL cells following exposure to dexamethasone. 77 They display enhanced expression of TNF-α in peripheral and lung immune cells. 78 Moreover, the GC resistance observed in patients with obese asthma may be caused by altered nitric oxide metabolism and induced airway remodeling. 79
Conclusions
Considerable heterogeneity exists in GC responsiveness among patients with BA. Although evidence exists supporting that inflammatory phenotype and disease-related environmental factors regulate individual sensitivity to GCs, further elucidation of the mechanisms underlying GC-resistant asthma is warranted.
