Genetics of Allergic Asthma and Current Perspectives on Therapeutic Management

Globally, more than 300 million people are asthmatics and this number has been estimated to become 400 million by 2025. Asthma is a chronic inflammatory condition, which, although has no cure, is treatable in most patients. The most common structural alterations in asthmatic airways include thickening of the epithelial and sub epithelial layers, increased airway smooth muscle mass, and angiogenesis. Several genetically control‐ led factors greatly influence the predisposition and severity of allergic asthma. Twin studies have attributed as much as 75% of asthma susceptibility to heredity. Particular‐ ly, genome-wide association studies (GWASs) have discovered several asthma and/or atopy susceptibility genes. Current treatment protocols for managing asthma involve the use of corticosteroids and β-agonists. Over the last 40 years, there has been a marked development-targeted therapy for asthma, such as anti-leukotrienes, anti-immunoglo‐ bulin (Ig)E, anti-tumor necrosis factor (TNF)-α, and anti-interleukins (ILs) (Th2 cytokines). To identify novel targets and to develop newer drug generations, better understanding of asthma molecular pathophysiology is required. Furthermore, the pharmacogenetic studies, focusing on better understanding of beneficial or/and adverse effects to anti-asthma drugs, will also facilitate the development of more effective and targeted treatments in specific subpopulations of patients suffering from asthma. thickening, airway smooth muscle hyperplasia, and angiogenesis [6].


Introduction
Asthma is an inflammatory chronic condition that has reached globally epidemic levels. Although no cure exists, symptoms are treatable in most patients [1]. Statistically, the number of asthmatic cases has been on the rise over the past 10 years and affecting up to 10% of adults and 20% of children worldwide [2]. Globally, more than 300 million people are asthmatics, and this estimate is predicted to become 400 million by 2025 [3]. The worldwide economic burden caused by asthma is predicted to be more than that of both acquired immunod eficiency syndrome (AIDS) and tuberculosis combined together. For example, in the United States of America, the annual asthma care costs exceed US$6 billion [4]. Moreover, these numbers are due to the fact that more than 50% of asthmatic cases are poorly controlled by medication, since the response to treatments varies considerably among patients despite having similar clinical features [3,5]. Asthma is characterized by altered and distinct clinical changes in the lung airways obstructing the flow of air into the lungs. The most prominent airway remodeling features include epithelial and subepithelial layer thickening, increased airway smooth muscle (ASM) mass, and angiogenesis [6]. Different classes of asthma therapies address one or more of the phenotypes of asthma; however, the heterogeneous nature of the disease prevents homogeneous clinical outcomes in response to the current treatment guidelines [7].
In the past two decades, the field of human genetics has evolved due to the advancements in the human genome project and high-throughput sequencing technologies [8,9]. Currently, the advances in genetic, pharmacodynamic, and pharmacokinetic studies, analyzing responsiveness of patients to various therapies, may eventually allow to prescribe personalized treatment and to shift asthma therapies from classical standards, using mostly corticosteroids and βadrenergic agonists, to a highly tailored approach [10]. Future genetic profiles of the population would form the basis of tomorrow's treatments in order to potentiate the required therapeutic benefits, and to diminish any possible adverse effect risks. Overall, there remains a great need for comprehensive drug research, paralleled with high-throughput genetic profiling, in order to treat asthma in a personalized or stratified manner [11].

Genetic control of airway hyperresponsiveness, atopy, and allergic asthma
The heritable nature of asthma has been demonstrated through various types of studies over the past decades. Family and twin studies indicate that 60-70% of asthma cases are due to genetic factors. Moreover, it has been proven that the concordance of asthma is greater among monozygotic twins rather than among dizygotic ones. Adoption studies have shown greater disease prevalence within biological relatives of the affected people compared to the adopted family [12].
Higher prevalence of allergic disease phenotypes is observed among relatives of atopic individuals compared to nonatopic individuals. Overall, the heritability estimates remain in between the range of 30-66% for airway hyperresponsiveness, 35-95% for asthma, and 35-84% for total serum IgE levels [13]. It is clear that both the inter-genetic individual differences and the degree of allergen exposure contribute to these variations in heritability. Heritability of asthma is linked to both disease susceptibility and severity. While the main concern of asthma genetic studies has been on disease susceptibility, increasing evidence shows that many genetic variants are important in asthma progression and severity as well [14]. Lung function tests in asthma showed that genes in the T-helper lymphocyte 1 (Th1) pathway affect asthma severity; meanwhile, T-helper lymphocyte 2 (Th2) pathway genes relate to susceptibility [14]. Based on these hypotheses, genes associated with asthma susceptibility differ from those related to asthma severity; hence, it is important to define both groups distinctly.
By knowing the genetic signature associated with allergic asthma, geneticists can help to better understand the molecular mechanism of this disease, and the shared and distinct pathways among other allergic diseases. Moreover, the genetic signature of asthma-associated genes with altered expression during the peak of asthmatic episodes may help predict the severity and response to therapy. Unfavorable response might be identified and, consequently, more targeted and personalized treatments can be considered for this complex trait. The human genome project and the ongoing advancements in sequencing technologies, both, resulted in more successful gene discovery over the last years, even in diseases as complexed as asthma. Since then, dozens of susceptibility genes were identified through a large variety of methods and rationales. ADAM33 is the first asthma susceptibility gene to be discovered through positional cloning [15]. ADAM33 (also known as Disintegrin and metalloproteinase domaincontaining protein 33) is a membrane-bound metalloproteinase enzyme that has been involved in several cellular interactions involving cell-cell and cell-matrix events [16]. Variants in this gene have been correlated to asthma susceptibility and bronchial hyperresponsiveness, but not atopy. Due to its clinical significance, ADAM33 studies were conducted among 33 different asthmatic population samples all over the world. Additionally, numerous studies have suggested that altered ADAM33 DNA methylation patterns could result in diversely unbalanced biological effects in the airways [17]. Studies focused on candidate genes have examined a number of genes involved in asthma and highlighted more than 100 interesting genetic spots; however, the role of those loci in asthma susceptibility remains largely unexplored [18].
Genome-wide association studies (GWASs) extensively investigate the unknown genetic bases of many intricate disorders including asthma [19,20]. In the first reported GWAS study for asthma susceptibility, Moffatt et al. [21] identified the 17q21 locus, containing several genes, for example, ORMDL3 and GSDMB as being associated with childhood asthma. Importance of this region was later on replicated in numerous subsequent studies [22][23][24]. Expression levels of the gene ORMDL3 are differentially regulated by distinct haplotypes in this region. This gene encodes protein acting as an inhibitor of sphingolipid biosynthesis and in general Orm family proteins were shown to be implicated in the control of sphingolipid homeostasis [25]. Dysregulated sphingolipid formation in the respiratory tract instigates airway hyperreactivity [26] although exact molecular steps are still not known. The results of these studies suggested that the mechanisms of asthma development are linked with genetically determined abnormalities in some patients resulting in their inability to control balance between oxidative and anti-oxidative reactions. The mechanisms of asthma development are linked with genetically determined abnormalities in the functioning of antioxidant defense enzymes. These alterations seem to be accompanied by a systemic imbalance between oxidative and antioxidative reactions with the shift of the redox state toward increased free radical production, oxidation of proteins and phospholipids, and eventually to their selective degradation.
To increase the power of detection of modest alleles due to the large sample size, the results of individual GWAS need to be gathered into a meta-analysis. The scientific literature recognizes two meta-analyses of asthma GWAS. One was done by the GABRIEL Consortium [27] of the European investigators, and the other was conducted by the EVE Consortium of the US investigators [22]. While the EVE meta-analysis included diverse subjects from different ethnic background, US and Mexico population backgrounds, the GABRIEL meta-analysis included only European subjects. Overall, these two thorough meta-analyses present a comprehensive overview of the genetic associations for asthma. Some associations are shared among different populations; by contrast, others are specific to one race. Grouping GWAS in this way increases the power of genetic detection, contrasts different ethnic groups' genotypes, and highlights the worldwide populations' genetic patterns. Overall independent GWASs have identified large number of candidate loci that deserve further testing. Replication studies help to prioritize which genes deserve further study, based on their identification in multiple populations.
Additionally, more loci were identified to be associated with asthma; these include interleukin (IL)-33 (on 9p24), HLA-DR/DQ (on 6p21), IL1RL1/IL18R1 (on 2q12), TSLP (on 5q22), and IL13 (on 5q31) [22,27,28]. Collectively with ORMDL3/GSDMB (on 17q21), these are the most remarkable and consistent loci, which are identified for asthma. Since Moffatt et al. had published the first GWAS results for asthma, identifying ORMDL3 as a candidate gene, numerous other studies have been conducted investigating an array of phenotypes which are observed in allergic diseases. In particular, FCER1A, RAD50, and STAT6 have been associated with total serum IgE levels [29].

Environmental factors contributing to asthma
Parallel to genetic factors, environmental factors are also involved in the development and progression of asthma (Figure 1). The exposure to some environmental factors was shown to contribute not only to asthma but also to other related respiratory disorders, for example, emphysema development. By contrast, there are also some other environmental factors that seem to be solely linked to the development of asthma but not to other inflammatory or/and respiratory disorders [30]. Various studies assessed the risk factors of asthma and found evidence that allergen exposure, respiratory tract infections, gastroesophageal reflux disease (GERD), and physical and psychosocial stress might represent individual risk factors. It is important to keep in mind that some other environmental factors are protective, such as maternal diet, breastfeeding, and farming conditions [31].
Allergen exposure is the major factor impacting sensitization and constitutes the most common cause of asthmatic exacerbations in adults and children. A wide variety of inhaled allergens may trigger asthma symptoms, for example, house dust mite [32], pollens [33], cockroaches [34], and animal fur [35]. Respiratory tract infections have been implicated in asthma occurrence and exacerbation as well. Examples include infection with viruses [36,37], Mycoplasma [38], and Chlamydia species [39]. Based on the conclusions from the Japanese study, which included 3085 patients, the change in weather followed by smoking was identified as two leading asthma-exacerbating factors [40]. Although (passive) smoking is a predominant contributing factor for the development of asthma [41], one occupational study [42] has shown that nonsmokers might also develop asthma due to occupational air pollutant exposure. Additionally, a correlation has been observed between the presence of asthma and gastroesophageal reflux-induced disease, with reports showing one-third of asthmatic patients also diagnosed with GERD [43,44]. Although the coexistence of GERD in asthmatic patients did not affect asthma severity, the airway resistance was significantly higher in asthmatic patients with GERD [45]. Some other psychosocial factors such as parental stress during childhood [46] and the socioeconomic status [47] are reported to influence allergic inflammation severity. It is estimated that psychopathology is six times more common among asthmatics, and accordingly it correlates more closely with the asthmatic quality of life, rather than with lung physiological functions [48,49]. In both directions, psychopathology is supposed to precipitate asthma or vice versa; psychopathology may develop as a consequence of asthma [50].

Asthma pathophysiology
Scientists tried to uncover alterations related to asthma since a long time ago. One of the oldest publications that discussed asthma pathophysiology was in 1873 [51]. Later on, in 1886, F.H. Bosworth concluded a possible relation between asthma and hay fever [52]. Clearly, it is well known that asthmatic patients suffer from reversible airway obstruction resulting from an allergen exposure, consequently releasing multiple broncoconstricting mediators that stimulate airway muscles to contract. Furthermore, airways narrow results from past and current mucus and edema occlusion [53]. The chronic inflammation and associated repair of lung airways leads to structural changes, referred to as "airway remodeling." Airway remodeling (Figure 2) usually involves lung epithelial layer injury and includes features such as subepithelial thickening, airway smooth muscle hyperplasia, and angiogenesis [6]. Asthma and COPD (chronic obstructive pulmonary disease) are now considered to be discrete respiratory disorders. Although both share several similar underlying mechanisms, driving airway obstruction in COPD, and hyperresponsiveness in asthma, core molecular pathology remains to be mostly different for both [54]. Pauwels et al. [55] defined COPD as "a disease state characterized by airflow limitation that is not fully reversible. The airflow limitation is usually progressive and associated with an abnormal inflammatory response of lungs to noxious particles and gases." One important reason of asthma and COPD overlap is the effect of aging. Asthma-COPD overlap syndrome (ACOS) is a medically recognized coexisting syndrome of both asthma and COPD [56]. Some other health conditions may occur more frequently in asthmatic patients. Rhinosinusitis [57], obstructive sleep apnea [58], or GERD [59] are the most common documented comorbidities. Substantially, they can contribute to the same pathophysiological process, which is already triggered by allergic response or alter asthma phenotype detrimentally. The impact of these diseases on asthma is variable and still not fully clear [60].

Epithelial/subepithelial layer thickening
Epithelial changes are not unique to asthma, they are also observed, in more or less of similar manner, in lungs of smokers and cancer patients [61]. Epithelial layer damage in asthma includes loss of ciliated cell layer, shedding of the epithelium, goblet cell hyperplasia and growth factors, cytokine and chemokine upregulation [62].
One important feature of asthma, which has been routinely used as an asthma severity index, is the thickening of the subepithelial airways layer. The epithelial and subepithelial layer thickening is caused by the overdeposition of extracellular matrix (ECM) proteins [63]. Roche et al. observed that intensive layers of collagen sedimentation contribute to the thickened subepithelial basement membrane. Through immunohistochemistry, they have shown that the commonly involved collagen types are collagen I, III and V, and fibronectin [64]. Additionally, the cells that are responsible for ECM protein production are myofibroblasts and fibroblasts, as both are embedded in the sophisticated ECM which they secrete [65]. Meanwhile, some inflammatory cells, for example, T cells, mast cells, and eosinophils also accumulate in the submucosal layer [66]. Moreover, transforming growth factor-β (TGF-β), and some similar growth factors, is usually secreted by the lung epithelial cells echoing any ongoing lung injury, and consequently directly impress the matrix proteins' production by fibroblasts/ myofibroblasts. By increasing the airway rigidity, however, Holgate et al. suggested that the airway thickening due to the ECM proteins precipitation may in fact have a remodeling protective effect via postponing long-term bronchoconstriction events [62]. Collectively, the ECM proteins, the lung structural cells (i.e., epithelial cells and fibroblasts), and the immune system inflammatory cells, all interact together and control the overall airway remodeling and fibrosis [67].

Hyperproliferation of airway smooth muscle mass
Hyperproliferation of airway smooth muscle mass is a common event in asthma and has been suggested to be implicated in its pathophysiology. Hyperplasia and hypertrophy of the ASM in the bronchial airways of asthmatics can be observed by three-dimensional (3D) morphometric studies [68]. Airways smooth muscle layer is estimated to be increased by 25-55% in nonfatal asthma and up to 50-200% in fatal asthma [69]. Meanwhile, in response to some growth factors like TGF-β, vascular endothelial growth factor (VEGF), and connective tissue growth factor (CTGF), ASM cells actively participate in the remodeling process through the process of ECM synthesis [70]. ASM cells also express cellular adhesion molecules (CAMs), receptors for cytokines (e.g., tumor necrosis factor-α), Toll-like receptors, and chemokines (eotaxin, macrophage inflammatory protein 1α, and interleukin 8) presenting multiple mechanisms for the inflammatory and remodeling process [71]. Additionally, one characteristic event of the airway remodeling is the ASM cells migration toward the epithelium [72]. Since ASM cells are crucial in asthma, Zuyderduyn et al. suggested that these cells should be targeted, rather than targeting inflammation or dealing with the symptoms [73].

Angiogenesis
Accumulating evidences indicate that there is an abnormal elevation in the size and number of blood vessels, as well as microvessels vascular leakage within the bronchial tissue in remodeled airways [74]. It is assumed that VEGF strongly affects airways remodeling via its angiogenic effects, but the exact molecular mechanism linking the increase in the VEGF expression to remodeling of the airways has not been fully understood [75].
Correlation between angiogenesis and asthma severity has also been documented. Dense vascularity occurs in severe asthmatics, followed by moderate, and then finally mild asthmatics, who experience less angiogenesis events [76]. This pattern was also observed in fatal asthmatics compared with nonfatal asthmatics [77]. While current asthma therapeutics is not directly targeting vascular remodeling, recent trials investigate some anti-angiogenic therapies as a new approach for asthma. Yuksel et al. showed that Bevucizamab, which significantly neutralizes VEGF, results in a reduced thickening of lung epithelium, a reduced ASM, and a reduced basement membrane thickness compared with untreated ovalbumin (OVA)-challenged mice [78].

Therapies for asthma
Modern treatments for asthma have been tested and used since the early twentieth century [79]. However, the oldest documented drug for asthma dates back to ancient Egypt. Kyphi, an incense mixture drink, was used inside the temples by the priests as a multipurpose lung medicament. There was more than one recipe for Kyphi; each may include as many as 10 herbs [80]. Following this, about 4000 years ago, Atropa Belladonna alkaloids, also called "deadly nightshade" because of their poisonous properties ("Natural Medicinal Herbs"), were derived from the leaves of thorn-apple plant and smoked by the Indians as cough suppressant [82]. Till today, natural and synthesized entities related to the tropane alkaloids class are still widely used. This includes anticholinergics (e.g., natural atropine, hyoscyamine (the levo-isomer of atropine), acopolamine, and the synthetic Ipratropium Bromide and stimulants (e.g., cocaine and hydroxytropacocaine) [83]. In 1872, one of the first papers published on asthma states that rubbing the chest of asthmatics with chloroform liniment can resolve airway constriction [84]. Adrenergic stimulants were in use for asthma over 100 years ago. In 1901, the adrenaline isolated from sheep and oxen adrenal glands was used to treat asthma [85]. The first documented publication of adrenaline as a bronchodilator therapy for asthma was written in 1903 by James Burnett, a physician in Edinburgh [86]. One year later in 1904, adrenaline was synthesized in the laboratories of Friedrich Stolz and Henry Drysdale Dakin, independently [87].
As suggested by the Global Initiative for Asthma (GINA) [88], a five-level step-down approach is widely recognized among the medical practitioners (Figure 3). The GINA approach assigns two types of drug classes for managing asthma: • Relievers (bronchodilators) cause immediate dilatation effects on the airways obstruction, mainly by acting on lung's smooth muscle.
• Controllers (preventers) provide long-term control of symptoms, mainly by suppressing the underlying disease process. β2-agonists and anticholinergics are considered to be bronchodilator relievers. Asthma controllers include corticosteroids, anti-leukotrienes, and anti-IgE. Theophylline is casually classified as both a bronchodilator and a reliever. The following book section will briefly discuss each therapeutic class.

Corticosteroids
Nowadays, most popular protocols for managing asthma involve the use of corticosteroids and β-agonists [1]. Anti-inflammatory corticosteroids, which are one of most trusted treatments for asthma, were introduced in mid-twentieth century [79]. The principle mode of action of corticosteroids in asthma is through their direct anti-inflammatory effect in different white blood cells including T cells, mast cells, and eosinophils. Among leukocytes, corticosteroids suppress chemotaxis and adhesion, and prevent inflammatory cytokines recruitment [89]. In vitro, corticosteroids reduce human ASM proliferation directly [90] by stimulating p21 gene expression [91], an important regulator of cell cycle progression. Moreover, corticosteroids improve vast majority of vascular remodeling aspects in asthma, reducing angiogenesis, excess blood flow, and vascular leakage [92]. This is mainly mediated by decreasing VEGF activity within the airway wall cells [93].
Various studies describe contradicting effects of corticosteroids on the lung epithelial abnormalities in asthmatics. Dorscheid et al. [94] reported that Dexamethasone treatment resulted in increased epithelial apoptosis and shedding. Similar results were obtained when treating guinea pigs with Budesonide, which did not improve the tracheal epithelium [95]. By contrast, some in vivo studies showed that inhaled corticosteroid (ICS) treatment resulted in improvement of epithelial damage in severe asthmatics [96,97].
ICS has been used around for the past couple of decades. Its idea dates back to the nineteenth century when the hand-held glass bulb nebulizer was used; however, pressurized metereddose inhaler (pMDI) came to the clinic in 1956. After seeing his daughter's suffering while using the hand-held nebulizer, George Maison, a medical consultant at 3M Pharmaceuticals, had advocated the use of pMDI. In 1959, George Maison and Irvine Porush were awarded a patent on the first pMDI [98].

β-adrenergic agonists
Long-acting β-agonists (LABAs), for example, Formoterol [99] and Salmeterol [100], offer a longer period of bronchodilation compared to the short-acting beta agonists (SABAs), for example, Salbutamol [101] and Terbutaline [102]. LABAs persist in the airway tissues for long periods due to their lipophilic nature and they provide a good umbrella of asthma bronchodilation and control, particularly at night [99,100]. However, until recently, the medical literature lacked supporting studies reporting the positive effect of β 2 agonists on the chronic airway remodeling [103]. Addition of a β-agonist to the corticosteroid therapy allows a "steroid-sparing" effect, that is, maintains asthma control using lower doses of corticosteroids [104]. LABAs are not used as monotherapies anymore and they must be used in combination with ICS [105], because there have been cases of severe exacerbations and death when LABAs are administrated solely.

Antimuscarinic agents
Inhaled antimuscarinic agents, also known as inhaled anticholinergics, are considered another alternative bronchodilator group to β-agonists. The bronchodilation effect is functionally mediated via muscarinic receptor subtypes M1, M2, and M3, although five muscarinic receptors have been revealed in the lungs M1, M2, M3, M4, and M5 [106]. It is widely known that parasympathetic stimulation via the vagus nerve leads to immediate smooth muscle contraction and mucus secretion in the airways [107]. It is also suggested that M receptors interact with β2-adrenergic receptors (ADRB2) on the airways smooth muscle, leading to a reduced bronchodilator response of the β-agonists [108]. For years, in both adults and children, short-acting antimuscarinic agents use, for example, Ipratropium [109], has been limited to acute asthma management, in addition to inhaled SABA [110,111]. Long-acting antimuscarinic agents, for example, Tiotropium [112], appear to have more benefits in difficult-to-control asthma. Adding Tiotropium to the standard asthma therapy significantly reduces asthma symptoms and highly increases the clinical outcomes [113,114].

Targeted therapies
Over the last 40 years, there has been a marked increase in the development of targeted treatments for asthma-anti-leukotrienes, anti-IgE, anti-interleukins, and anti-TNF-α [115].
Obviously, as more of the biological basis of asthma is uncovered, more effective targeted asthma treatments might be developed.

Anti-leukotrienes
Leukotrienes are lipid eicosanoids with a wide range of biological activities. They are derived from arachidonic acid through the enzymatic action of 5-lipooxygenase, and play a crucial role in asthma inflammatory pathogenesis, and in other allergic diseases such as allergic rhinitis, rhinosinusitis, atopic dermatitis, and urticaria [116]. Leukotrienes class includes three main types: cysteinyl leukotrienes (CysLTs), LTB4, and LTG4. LTG4 is the metabolite of LTE4 in which the cysteinyl moiety has been oxidized to an α-keto-acid [117]. Since, very little is known about the LTG4-putative leukotriene, most clinical research studies focus on CysLTs and LTB4. CysLTs are strong bronchoconstrictors that powerfully affect airway remodeling, whereas LTB4 is a strong chemoattractant for most leukocyte subsets [118]. Over the last 20 years, since leukotriene antagonists were introduced to the clinic for asthma management, Montelukast [119,120] and Zafirlukast [121] are the most frequently used drugs in this class.

Anti-IgE
At the moment, Omalizumab, which is the only approved targeted monoclonal antibody against IgE, is used to treat allergic asthma in clinical practice. It can significantly decrease serum IgE levels (up to 99%) within 2 h following subcutaneous administration, and diminish serum, sputum, and tissue eosinophilia [122]. Recently, Omalizumab has also been reported to have steroid-sparing effect, reducing the rate of asthma exacerbations up to 50%, and hence improving the quality of life [123]. However, nearly 45% of patients treated with Omalizumab had adverse reaction at the local injection site, which is considered the most commonly observed adverse event for Omalizumab. Some other minor upper respiratory tract infections and sinusitis have also been reported as well. Patients treated with Omalizumab display a very low (0.09%) frequency of anaphylaxis reaction. Importantly, there are no data reporting any correlation between cancer and Omalizumab treatment [124].

Anti-ILs
Three interleukin pathways are of physiological importance for asthma: IL-5, IL-9, and IL-4/ IL-13 pathways. IL-5 is pivotal for both eosinophil differentiation and maturation in the bone marrow. Subsequently, it controls eosinophil mobilization, activation, and survival [125]. Hence, antagonizing IL-5 has been proposed to be beneficial for asthma therapy, particularly for predominantly eosinophilic asthma. A number of anti-IL-5 and anti-IL-5 receptor monoclonal antibodies are in the process of development for allergic diseases: Reslizumab [126], Mepolizumab [127], and Benralizumab [128]. IL-9 is one of the T-helper 2 (Th2) pro-inflammatory cytokines that promote mast cell proliferation and T-cell growth [129]. In mouse models, IL-9 causes several common features of chronic asthma: excessive mucus production, eosinophilic airway inflammation, smooth-muscle cell hyperplasia, and aryl hydrocarbon receptor (AHR) [130]. Currently, a phase IIb clinical trial evaluates the efficacy and safety of subcutaneous Medi-528, a humanized IgG1 anti-IL-9 mAb, in adults with uncontrolled asthma (NCT00968669). Activated mast cells, eosinophils, basophils, and dendritic cells secrete IL-4 and IL-13. IL-4 and IL-13 both play an important role in asthma mainly by enhancing IgE production. They also control mast cells' growth and development, eosinophil recruitment, and AHR [131]. The first trial aimed at antagonizing the IL-4 used a soluble recombinant human IL-4 receptor antagonist (IL-4RA), altrakincept, which blocked the binding of IL-4 to its cellular receptors [132]. Several humanized IL-13-neutralizing antibodies have entered asthma phase I/II clinical trials-anrukinzumab [133], QAX576 [134], and CAT354 [135].

Anti-TNF-α
TNF-α, a cytokine produced by Th1 cells and macrophages, has diverse biological functions. TNF-α shows crucial, and previously extensively documented, role in Crohn's disease, rheumatoid arthritis, and psoriasis pathogenesis. The association between TNF-α increase and these disease progressions had inspired studies aiming to extend anti-TNF-α therapies also for the treatment of severe asthma and COPD [136]. Infliximab and Golimumab, two anti-TNFα mAbs, and Etanercept, a decoy soluble TNF-α receptor, are both able to biologically neutralize TNF-α cytokine, and blunt the immune response, thereby abolishing TNF-α effects in asthma [137].

Pharmacogenetics of asthma
The US Food and Drug Administration definition for pharmacogenomics is "the study of variations in DNA and RNA characteristics as related to drug response" [138]. "Pharmacogenomics" differs from "Pharmacogenetics" in that the former is concerned with the whole genome, its components, and regulators, while the latter is focused only on the DNA sequences of individual gene. Thus, in sense, pharmacogenetics is thought to be a subset of pharmacogenomics [139].
Because it is a complex trait, the drug response to asthma is diversely heterogeneous even among patients with apparently similar clinical profiles [7]. It is estimated that up to 50% difference in therapeutic response has been attributed to genetic variations between individuals [140]. Although several possible mechanisms have been postulated, genetic variants affect the pharmacogenetic response to drugs in two different ways: 1. Pharmacodynamic genetic variations are variations in which the receptor binding the drug ligand or another member of the drug target pathway is altered resulting in different drug effect. Most of the current pharmacogenetic research fall into this mechanistic category.
Populations are stratified into responders and nonresponders, and then analyzed for DNA polymorphisms, which distinguish these two groups apart.

2.
Pharmacokinetic genetic variations are related to altered uptake, distribution, and/or metabolism of the administered drug. Fewer examples fall in this category; however, the most common research subfield here is the area of investigating drug-catabolizing orexcreting enzymes. An important example here is the cytochrome P450 (CYP450) family, a widely recognized metabolizing enzyme with several variable pharmacogenetic patterns.
Single nucleotide polymorphism, SNP, denoted by a reference sequence (rs) number, represents a class of polymorphism that is derived from a one-base point mutation in which a single nucleotide is substituted with another one. SNPs may be located in the gene regulatory or coding regions, and so it may affect the gene expression in more than one way; however, in majority of cases, most discovered SNPs do not change the gene function in a significant manner [141]. Consequently, it is essential to investigate whether the DNA sequence variances would actually cause significant functional impacts (i.e., resulting in an altered observed biology), or is a linkage disequilibrium marker of another DNA variant, which is the real cause of the response variability, or is generally nonsignificant. Because of its strong importance, since 15 years, catalogs of SNPs have started to outline the most common genetic polymorphisms among different population groups [141,142], and this process has attracted more attention during the last couple of years [143]. All genes contain huge number of SNPs and copy-number variations (CNVs). CNVs are another form of structural variations, which account for 13% of the human genome bulkiness, and manifest as kilo-to-mega bases of deletions or duplications [144]. Conjointly, it is challenging to outline which polymorphism is influencing the treatment response and which are not relevant. Two major approaches declaiming this challenge have been practiced so far: candidate gene approach and GWAS. As it combines transcriptomic, proteomic and metabolomic profiling traits, a third approach, the integrative system biology approach, had led to a more comprehensive pharmacogenetic view [3]. To differentiate, candidate gene approach is based on a prior evidence according to the knowledge of the drug pharmacody-namics or/and pharmacokinetics, by contrast, GWAS methodology identifies new associations with the null hypothesis being that no associations exist. GWAS picks the variations which are associated with observable phenotypes by scanning SNP markers that tag, via linkage disequilibrium, the complete human genome. GWAS and integrative system biology approaches are modern tools contributing to the recent advancements of genotyping and statistical technologies.
Current pharmacogenetic studies of the corticosteroids, β-adrenergic, and leukotriene pathways are mostly candidate gene studies, with some GWAS, however, altogether have identified several genetic loci in strong association with therapeutic responsiveness to asthma. Figure 4 summarizes the pharmacogenetically significant genes with relevance to the corticosteroids, β-adrenergic, and leukotriene biological pathways.

Corticosteroid pathway pharmacogenetics
In cytosol, the glucocorticoids bind to their corresponding glucocorticoid receptor, forming a hetero-complex that is activated by ligand binding, and translocate into the nucleus. In the nucleus, this complex binds to the glucocorticoid response elements in some target genes' promoter region resulting in their expression regulation. The core role of glucocorticoids is mediated via activating the transcription of anti-inflammatory genes, and suppressing the transcription of pro-inflammatory genes [145,146]. The glucocorticoid pharmacogenetic studies formerly focused on candidate gene approach. Those candidate genes covered functions related to the corticosteroid biosynthetic pathway, the hetero-complex receptor formation, and the related chaperone proteins.
Corticotrophin-releasing hormone (CRHR1), stress-inducible protein 1 (STIP1), TBX21, CYP3A4, GLCCI1, T gene, and FBXL7 are the most up-to-date potential pharmacogenetic biomarker targets for predicting patients' response to ICS [147]. Studies of the corticotrophinreleasing hormone gene are considered to be as one of the oldest and remarkable footsteps in asthma pharmacogenetics. CRHR1 protein, also known as CRF1, is the primary receptor controlling the adrenocorticotropic hormone release; hence, it plays a pleiotropic and vital role in steroid actions. A candidate gene study of CRHR1 in 1117 asthmatics administrating ICS therapy, from three clinical cohorts, revealed two SNPs (rs242941 and rs1876828) associated with different response in lung functions [148]. Tantisira et al. [148] found that CRHR1 gene variation was frequently related to augmented therapy response in each of the three studied cohorts. Since 2004, CRHR1 gene studies opened the doors for all other corticosteroid pharmacogenetics and the possible future therapeutic outcomes.
STIP1 or HOP (abbreviated for Hsp70-Hsp90-Organizing Protein) gene mainly functions to reversibly link Hsp70 and Hsp90 together as a co-chaperone [149]. STIP1 pharmacogenetic studies in one adult cohort revealed three SNPs (rs2236647, rs6591838, and rs1011219) within this heat shock-organizing protein and related to improved lung response during ICS therapy [150]. STIP1 rs2236647 variant analysis in healthy and asthmatic children showed that this SNP could serve as an asthma marker for choosing the population who receives corticosteroid therapy [151]; however, further replication studies should be held to confirm those results.
One significant aspect of pharmacogenomics is that it investigates the interactions with genes of other pathways. TBX21 gene is one good example for observing the ICS response outside the glucocorticoid pathway. TBX21 is one of the conserved genes of a family sharing a common DNA-binding domain; the T-box encodes T-box transcription factor TBx21 protein. Tbx21 protein is a Th1 (T-helper1) transcription factor, which regulates one of the Th1 cytokine expression, interferon-gamma (IFNG). In 2004, a nonsynonymous SNP rs2240017 (His 33 Glu) in the TBX21 gene was linked to improvements in bronchial hyperresponsiveness or "bronchoprotection" in response to ICS in individuals participating in the Child Asthma Management Program (CAMP) cohort [152]. This finding was also observed in an independent Korean cohort in 2009 [153]. Thus, TBx21 may be an important determinant pharmacogenetic candidate gene for predicting asthmatics' response to inhaled corticosteroid therapies.
In 2005, another example demonstrated the glucocorticoid pathway interactions with one other pathway. ADCY9, adenylyl cyclase type 9, gene encodes a membrane-bound enzyme in the β2-adrenergic receptor pathway, which catalyzes the production of cyclic adenosine monophosphate (AMP) from adenosine triphosphate (ATP). This candidate gene contains a pharmacogenetic nonsynonymous SNP, Met 772 Ile, which was correlated to enhanced Salbutamol (SABA) bronchodilator effects only in patients treated with ICS [154]. An independent Korean cohort replicated the trial, using Formoterol (LABA) treatment in combination with ICS, and confirmed those results [155].
Cytochromes P450s belong to a heme cofactor-containing superfamily of metabolizing enzyme proteins that potentially control the metabolism of drug (i.e., pharmacokinetics), and consequently treatment response in many diseases. For asthma, CYP3A4, CYP3A5, and CYP3A7 candidate genes have been studied among a retrospective analysis of 413 asthmatic children treated with the ICS Fluticasone propionate [156]. The three candidate CYPs of all subjects were genotyped for nine SNPs. Results showed that asthmatics with the CYP3A4*22 allele demonstrated a significant symptom control compared with those lacking that allele. This study included a small number of participants (n = 20), so further large-scale replication is required.
Tantisira et al. [157] conducted the first pharmacogenetic GWAS for ICS treatment in asthma and identified an SNP (rs37972) in the promoter of the glucocorticoid-induced transcript-1 gene (GLCCI1), which significantly associates with lung functions. Replicated in four independent populations (935 persons in total), this candidate SNP was linked to substantial decrements in the response to the ICS in asthmatics. The wild-type allele homozygotes (CC) showed greater forced expiratory volume in 1 s (FEV1) in response to the ICS compared with those identified with the homozygote variant allele (TT). Another functionally correlated SNP (rs37973) in the promoter of the same gene was further validated within in vitro studies [157]. Results showed declined luciferase reporter activity in cells with the minor allele. GLCCI1 GWAS outlines that drug response to asthma treatment is subjected to wide inter-individual variation, and GWAS would uncover more novel pharmacogenetic associations in the future. factor protein) compromised two SNPs (rs3127412 and rs6456042) that were associated, out of the successfully genotyped 47 SNPs, with altered lung function response to ICS [158].

β-adrenergic receptor pathway pharmacogenetics
β2-adrenergic receptor gene remains to be the most studied pharmacogenetic loci among the beta-agonist pathways. ADRB2 gene has several polymorphic variants that were discovered in multi-ethnic genetic asthma cohorts [159,160]. ADRB2 protein is a cell membrane-spanning receptor that binds epinephrine, but not norepinephrine, unlike the other adrenergic receptors, and consequently mediates both smooth muscle relaxation and bronchodilation [161,162].
Early ADRB2 studies showed that Gly 16 Arg, a prevalent coding variant of the amino acid at position 16 of ADRB2, is associated with altered bronchodilator response to SABAs [163].
The BARGE (Beta-Agonist Response by Genotype) study [164], held by the National Heart, Lung and Blood Institute Asthma Clinical Network, was one of the first genotype "stratified" pharmacogenetic studies for asthma. In this study, only Gly 16 Arg homozygotes for ADRB2 were included (i.e., Arg/Arg and Gly/Gly). Participants were randomly receiving either intermittent or regular albuterol, and then crossed over to receive the alternative treatment dose. For statistical stratification, this study ensured that the Arg 16 homozygotes, who are less frequent, were appropriately randomly distributed to both SABA intermittent and regular protocols. Compared to Gly 16 homozygotes, the BARGE study showed that the Arg 16 homozygotes were good responders only to acute intermittent SABAs rather than to long-term regular treatments, a finding that does not coincide with the current clinical asthma treatment guidelines [165] which recommend SABA as for on-demand intermittent usage. Since the 16th amino acid of ADRB2 controls regular response to albuterol, bronchodilator medications other than SABAs would be more appropriate for Arg/Arg asthmatics.
Collectively, the BARGE study [164], along with some other pharmacogenetic studies [163,[166][167][168] of Gly 16 Arg and SABAs' exposure, provided insights for further studies [169][170][171] on LABAs. In contrast to SABAs, a large cohort [169] of 2250 asthmatics, randomly assigned to formoterol plus budesonide, demonstrated no pharmacogenetic action due to ADRB2 variation on therapeutic response. Furthermore, a multicenter trial [170] showed that asthmatics with both Arg/Arg and Gly/Gly genetic signatures had improved airway functions, if they received combination treatment with Salmeterol and ICS, when compared with ICS therapy alone. Similarly, the results of another prospective trial cohort [171] of 544 subjects, also randomized by genotyping, demonstrated no evidence of any pharmacogenetic action due to ADRB2 variation in response to Salmeterol. Together, these findings, confirmed among several asthma populations, suggest that in contrast to SABAs, asthmatics can still be treated with LABAs plus ICS irrespective of their genotyping status.
Genetic variants' occurrences among different ethnic groups are quantified by their percentage of allele frequencies. Usually, frequent and common variants have only little or modest impacts on disease susceptibility and, subsequently, therapeutic response. On the other hand, as the variant is characterized to be rare or more "private," its effect size on disease progression and therapeutic response dramatically increases [172]. Early in vitro studies had investigated a rare polymorphism of ADRB2 within the fourth transmembrane domain, the Thr 164 Ile variant. For the Ile 164 genotype, results showed significant lowering in Gs-protein signaling and different SABA-and LABA-binding affinities [173,174]. While the Thr 164 Ile polymorphism is pointed out to be a rare coding variant (i.e., <5%), population studies showed that this variant is more common in non-Hispanic white populations [159,160], a finding that requires further pharmacogenetic investigation in different and larger populations. To replicate results, a study of two large Copenhagen population cohorts [175], with more than 55,000 participants, was held to investigate the relation of Thr 164 Ile variation and lung responses. Among the general population, the Copenhagen study reported that the Thr 164 genotype was associated with decreased FEV1, diminished lung function, and increased the overall COPD risk.
In addition to ADRB2 Gly 16 Arg and Thr 164 Ile variants, the (-376 In-Del) polymorphism was extensively studied as another significant pharmacogenetic ADRB2 variant. Presented primarily among African Americans and Puerto Ricans [159,160], the 24-bp promoter insertion at −376, related to the start codon, is associated with asthma-related hospitalization in asthmatics treated with LABA [160]. Altogether, these variants, being unique to different populations, highlight the increasing need of personalized-based treatments.
Adenylyl cyclase type 9, encoded in humans by ADCY9 gene, is a membrane-bound enzyme that catalyzes the formation of cyclic AMP from ATP. ADCY9 is a widely abundant adenylyl cyclase, and it is stimulated via beta-adrenergic receptor activation [176]. Ile 772 Met is a coding variant of ADCY9 gene that has been associated with both acute FEV1 bronchodilation in response to SABAs [154] and long-term FEV1 response for LABAs [155]. CRHR2 (which is more commonly known as CRF2) is a type-2 G protein-coupled protein receptor for the corticotropin-releasing hormone [177]. Out of the 28 studied SNPs in CRHR2, five SNPs were significantly correlated with acute bronchodilator response in one, or frequently more than one, cohort. Among those, variant rs7793837 was associated with altered SABA response in all three cohorts of the CRHR2 study containing 607, 427, and 152 participants, respectively [178].
Different variants of ARG1 (Arginase 1) and ARG2 (Arginase 2) show altered acute response to SABAs, while the endothelial nitric oxide synthase (NOS3) shows altered acute response to LABAs. NO (nitric oxide), an endogenous vasorelaxing bronchodilator, is generated by the action of NOS3 on L-arginine. Since ARG1 and ARG2 are metabolizing L-arginine, so it is expected that the entire three genes, ARG1, ARG2, and NOS3, might be implicated in asthma pharmacogenetics. Combined association evidence, surviving Bonferroni correction for multiple testing from the CAMP four asthma cohorts [179], points to SNP rs2781659 in ARG1. C-allele homozygotes for SNP rs2781667 in arginase 1 showed significantly less response to the inhaled corticosteroid treatments [180]. Arginase-2 variants rs17249437 and rs3742879 correlated with increased airway obstruction and airway hyperresponsiveness, and lower reversibility of airway constriction following treatment with beta-2 agonists [180]. A small candidate gene study [181] of NOS3 had revealed one possible variant (Asp 298 Glu) correlated with lung function response to ICS/LABA combined treatment; however, this result still needs to be replicated in larger cohorts. THRB [182], SLC24A4 [183], SLC22A15 [183], SPATS2L [184], and SNPs (rs892940, rs77441273, rs1281748/rs1281743, and rs295137, respectively) show promising loci for further pharmacogenetic investigations.

Leukotriene pathway pharmacogenetics
Relative to the corticosteroid and β-adrenergic pathways, the cysteinyl leukotriene pathway pharmacogenetic studies are generally fewer and have smaller sample sizes. The oldest of these studies [185], held in 1999, had investigated the tandem repeat polymorphism in ALOX5 promoter. Among 114 asthmatics, it has been shown that the ALOX5 promoter repeat is associated with altered lung functions in response to a 5-LO inhibitor [185]. It has been shown in children that those who had more or less than five repeats (3, 4, and 6) of the ALOX5 promoter-binding motif experienced increased urinary leukotriene E4 (the terminal cysteinyl leukotriene metabolite) concentrations and reduced FEV1 baseline than the wild-type genotype with five repeats [186]. Further pharmacogenetic studies revealed that the ALOX5 promoter polymorphism, along with the ALOX5 SNPs rs892690, rs2029253, and rs2115819, influences leukotriene pathway antagonist therapy [187][188][189][190]. Moreover, variants of LTC4S, encoding Leukotriene C4 synthase, and MRP1 (or ABCC1), encoding multidrug resistanceassociated protein 1, have been linked to lung function response while treatment with Zileuton and Montelukast [189,190].
Arg 312 Gln, rs12422149, which is a coding variant in SLCO2B1 (solute carrier organic anion transporter family member 2B1 gene), has been related to symptom control during Montelukast therapy. This fact was due to the interindividual variability of carrier-mediated Montelukast transport in the intestines, and consequently its plasma levels [191]. By contrast, two other studies, probably due to their small sample sizes, were unable to replicate similar SLCO2B1 pharmacokinetic effects [192,193]. Overall, larger replicate cohorts, for the leukotriene pathway identified loci, are still needed.

Current and future challenges facing asthma pharmacogenetics
As demonstrated above, there has been fundamental progress in the field of asthma pharmacogenetics; however, these efforts have not yet been introduced into clinical practice to guide physician. There are several reasons that account for this gap. Most important is the limited number of asthma pharmacogenetics-focused GWAS, which would compare common candidate gene methodology that would allow combining all patients from small cohorts studied. Small sample sizes prevent any expansion of the pharmacogenetic research of asthma, which needs a large number of subjects for statistical significance. Along with limited cohort size, study defects due to poor ancestry structuring and stratification substantially result in replication inconsistencies. Furthermore, genes interact together in networks; therefore, simply attributing phenotypic variation to individual genes is not appropriate. Epigenetics studies investigate the changes in gene activities, which are heritable to the subsequent generations, but are independent of any DNA sequence alterations [194,195]. Epigenetic tuning of the genes associated with asthma has a significant impact on determining the drug response. Several mechanisms, related to epigenetics, are currently being investigated for both biomarker tagging and therapeutic innovation intervention [196]. Moreover, epigenetic changes have the ability to override the genetic effects of time, environment, tissue specificity, and other conditions such as age and gender of a patient, nutrition and hygiene, and intestinal microflora, which all highly influence the drug response in addition to the genetic factors. The collective impact of all combination of these factors requires the application of complicated algorithm that could take into consideration each of these factors and their interplay. The prospective genetic profile of an asthmatic should compromise a set of common and rare variants, on ancestral basis, which will be predictive of the pattern of his/her therapeutic responsiveness to different treatment options. The current human variant catalog continues to grow in an exponential manner because of the lower costs associated with whole genomic sequencing. Despite the steep decline in sequencing costs, the technology of sequencing, in terms of speed and quality, enormously increases. The future pharmacogenetic profile would also predict any possible adverse response associated with the chosen line of treatment. Genetic biomarkers are needed to warn the physician about any potential adverse side effects which can be life threatening. It is very important for typical genetic profiling to also consider genegene and gene-environment interactions. Gene-gene interactions are predominately crucial in the framework of combination therapies, for example, ICS and β-adrenergic agonists. Interactions between the surrounding environment and the patients' genes are assumed to be an additional element, because environmental stress, apart from the genetic makeup, contributes to the development of asthma exacerbations. Future pharmacogenetic directions need to cover also the pharmacokinetic side of the patient profile. Altered drug absorption, metabolism, distribution, or excretion extensively influence drug dosing and even drug selection. All in all, the complete asthma pharmacogenetic catalog has many aspects to cover, before being introduced into the clinical practice.

Conclusion
Asthma is a complex respiratory and immune disease. Inadequate (or exaggerated) ability of genetically predisposed individuals to control inflammation, induced by innate and environmental factors, results in asthma. Further, studies using allergic asthma and atopy models enable to better understand several interacting gene products and variable responsiveness of asthmatic subjects to current therapies. Eventually, thorough investigation of the complexity of asthma might lead to successful designing of personalized therapies for patients suffering from allergic asthma.