Transforming growth factor β (TGF-β) is a pluripotent cytokine expressed by all somatic human cell types (1), which plays several important roles in maintaining homeostasis (2). TGF-β occurs in three isoforms (TGF-β1, 2 and 3). All forms of TGF-β are receptor ligands of three receptor types (TβRI, II and III), have similar biological activity and participate in processes such as regulation of proliferation, migration, differentiation and apoptosis (3).

TGF-β regulates cell proliferation, contributes to epithelial-to-mesenchymal transition (EMT), regulates immune cell function by modulating immune responses, contributes to the conversion of fibroblasts to myofibroblasts, and causes overproduction of the extracellular matrix (ECM) in tissues undergoing fibrosis (4). TGF-β stimulates expression of proteins such as collagens, basal lamina proteins, and ECM proteins (5). Additionally, in the early stages of tissue fibrosis, TGF-β stimulates myofibroblasts and other stromal cells to increase the synthesis of collagen-crosslinking enzymes (6). Furthermore, TGF-β decreases the synthesis of matrix-depleting proteins, such as matrix metalloproteinases. Ultimately, the increased synthesis of matrix proteins and decreased activity of matrix metalloproteinases due to TGF-β activity contribute to the remodelling of epithelial tissue and may result in fibrosis (7–9).

The canonical TGF-β signalling pathway is activated by the binding of TGF-β ligands to their receptors. TGF-β is secreted in a latent, inactive form (L-TGF-β) and activated in an integrin-dependent manner (10). Active TGF-β binds to TGF-β type II receptors (TβRII), which are transmembrane receptors with serine/threonine kinase activity. TGF-β-TβRII interactions lead to the recruitment and phosphorylation of TGF-β type I receptors (TβRI) (11). Upon activation, TβRIs phosphorylate SMAD2 and SMAD3, which then assemble into heterodimers and trimers with SMAD4. Complete SMAD protein complexes translocate to the nucleus and influence the expression of TGF-β target genes. SMAD3 and SMAD4 bind DNA directly, with low affinity, and require the presence of other transcription factors to fully regulate target gene expression. SMAD2 does not bind DNA directly (12). Target genes regulated by TGF-β include plasminogen activator inhibitor-1 (PAI-1) and Smad7, which acts as an inhibitor of TGF-β/SMAD signalling by labelling activated TβRI for proteasomal degradation (13). Figure 1 is a simplified visual representation of the TGF-β signalling scheme.

Dysfunction of TGF-β signalling influences the development of many diseases and disorders, including cancer and organ fibrosis (14, 15). Excessive TGF-β activity promotes organ fibrosis by increasing connective tissue growth factor (CTGF) expression, inhibiting extracellular matrix (ECM) degradation, and increasing collagen synthesis. TGF-β also induces fibroblast proliferation and differentiation (16). The induction of organ fibrosis by TGF-β is mediated by Smad-dependent pathways, which directly increase the transcription of genes involved in ECM production, and Smad-independent pathways, including MAP kinase pathways (17).

Asthma is the most common chronic inflammatory disease of the respiratory tract (18), characterized by bronchial hyperresponsiveness to a number of factors, which leads to bronchoconstriction and reversible airflow limitation. Asthma is a heterogeneous disease with a significant degree of variability in phenotypes (clinical symptoms in conjunction with patient characteristics, such as age of asthma onset, clinical response to pharmacological therapies and presence of comorbidities) and endotypes (pathophysiological mechanisms leading to the development of the disease) (19). Characteristic symptoms of asthma include wheezing, shortness of breath, cough, and chest tightness. Symptoms vary in severity and frequency and occur mainly during disease exacerbations (20). Bronchial asthma can be divided into two main endotypes based on the presence of inflammatory responses mediated by T helper type 2 (Th2) lymphocytes. The two classically recognized endotypes are type 2 (eosinophilic) asthma and non-type 2 (non-eosinophilic) asthma (21). The discovery that innate lymphoid cells type 2 (ILC2) are capable of releasing Th2 cytokines has led to a more precise classification of asthma endotypes into T2-high, T2-low, and non-T2 (22).

The clinical picture of asthma depends on complex gene-gene and gene-environment interactions. These components make asthma a dynamic disease, and its classifications can change both due to treatment and independently of it. At the outset of asthma development, airway hyperresponsiveness to normally harmless antigens, such as pollen or mites, causes epithelial cells to release inflammatory mediators, including many interleukins (23). Inflammatory mediators trigger an immune activation cascade that leads to activation and recruitment of ILC2s, mast cells, Th2 cells, eosinophils, and dendritic cells to the site of inflammation (24). Repeated exposure to allergens can lead to chronic airway inflammation and airway remodelling, which ultimately leads to partial loss of respiratory function (25). Airway remodelling in asthma includes subepithelial fibrosis, abnormal extracellular matrix deposition, goblet cell proliferation, smooth muscle and mucous gland hypertrophy, and epithelial damage (26).

The disease develops under the influence of variables that, according to the Global Strategy for Asthma Management and Prevention (GINA) report (27), can be divided into exacerbation-inducing (inducing disease exacerbations in patients with pre-existing asthma) and asthma-causing (causing the development of asthma in patients without pre-existing asthma). According to the current GINA guidelines, the Asthma Control Questionnaire (ACQ) (28) or the Asthma Control Test (ACT) (28) can be used to assess symptoms. The severity of the disease is assessed not based on the severity of symptoms before treatment, but only after several months of treatment, when the level of severity and the intensity of treatment required to achieve and maintain asthma control are established (29).

Asthma has also been studied in the context of pharmacogenomics. In a systematic review of pharmacogenomic studies, García-Menaya et al. focused on four genes with the potential to be implemented in pharmacogenomics for asthma therapy. FCER2 was shown to be related to inhaled corticosteroid (ICS) response, ABCC1 and LTC4S were related to anti-leukotriene agents, and ADRB2 was related to beta-agonist response (30). In another systematic review which examined 30 studies involving over 6,000 subjects, variant rs1042713 of the ADRB2 gene was consistently associated with response to long-acting beta-agonists (LABA) (31). Another systematic review focused on the main findings of pharmacogenetic studies of pediatric asthma published in 2018–2019. The findings of the studies discussed in this review led to the validation of previously established gene-treatment response associations in asthma, and more importantly—allowed the identification of novel associations (PRKG1, DNAH5, IL1RL1, CRISPLD2, MMP9, APOBEC3B-APOBEC3C, EDDM3B, and BBS9). Critically, many of the results were not consistent between studies, underlining the need for large studies in diverse populations and a harmonization of treatment response definitions (32). It is important to note, that despite over 35 guidelines being published for pharmacogenomics-based optimization of therapy by the Clinical Pharmacogenetics Implementation Consortium (CPIC), there are no guidelines for asthma (33). In spite of much effort put towards researching pharmacogenomic biomarkers for asthma therapy, more compelling, robust clinical evidence is required for diagnostic implementation.

For the following section (“TGF-β gene polymorphisms and asthma etiopathogenesis”), literature research has been performed in the PubMed and Google Scholar databases using the following keywords: “snp” or “polymorphism” and “tgf” or “tgfb” or “tgf beta” and “asthma”. The inclusion criteria were as follows: (1) studies focused on the effects of transforming growth factor-β gene polymorphisms on asthma risk and severity, and (2) articles written in English. The exclusion criteria were as follows: (1) reviews, editorials, opinion articles or case studies; (2) animal or in vitro model studies; (3) papers that mentioned but did not directly study TGFB gene polymorphisms. A stepwise evaluation of each position in the search results according to the inclusion and exclusion criteria has been conducted based on the title, the abstract, and lastly the full text. Each paper has been independently evaluated by both authors.

Importantly, all of the studies described in this manuscript so far were candidate gene studies or meta-analyses of such studies. This type of studies focuses on specific genes hypothesized to be involved in a particular disease, selected based on established knowledge or biological plausibility. This approach involves examining only a limited number of polymorphisms within the chosen genes of a sample population. It is useful in cases of a known strong association between a gene and a disease, but not for discovering novel associations (112). Indeed, this approach has several limitations. It requires prior knowledge, which limits the rate of novel association discoveries, and can introduce bias. What is more, a small range of hand-picked genes may not reflect the complexity of the pathophysiology and etiology of a disease by not examining the wider genomic context (112).

On the other hand, genome-wide association studies (GWAS) utilise a hypothesis-free approach, and scan the entire genome via modern techniques such as next-generation sequencing (NGS), which has revolutionised the field of genetics. This approach examines millions of SNPs across the whole genome, and usually requires a large study population, which can be difficult to recruit. The major advantage of GWAS is its ability to discover numerous novel associations and create a comprehensive view of the entire genomic complex of a studied disease. Additionally, GWAS can detect variants with small effect sizes or rare variants with larger effects far more efficiently than candidate gene studies (113, 114).

Numerous GWAS have been published on asthma, however most of them do not describe any associations between the disease and TGF-β genes (115–117). Nevertheless, a number of genome-wide studies have found associations of asthma and genes related to TGF-β, or TGF-β genes themselves and other conditions affecting the airways. A GWAS conducted on a UK Biobank, including over 50,000 asthmatic subjects and over 350,000 controls led to the discovery of a causal association between TGFBR1 and asthma development, however no particular gene variants were described (118). Another GWAS described an association between asthma and two SNPs in the GSDMB gene (rs2305480 and rs11078927), which is implicated in TGF-β signalling (119). Several SNPs of SMAD3 (rs16950687, rs1729363, rs744910) the protein product of which is a downstream signalling component of TGF-β and important for regulatory T-cell and TH17 cell pathways, have been associated with asthma in several GWAS (120–122). Importantly, in a GWAS conducted on Mexican subjects with childhood asthma, rs2241715 in TGFB1 has been associated with an increased risk of asthma development. However, the study included only 492 subjects, which is not sufficient for a comprehensive GWAS analysis (123). In a 2011 GWAS conducted on 48,201 European subjects, TGFB2 was found to be a locus strongly associated with pulmonary function (124).

Lastly, two GWAS analyses have found associations between chronic obstructive pulmonary disease (COPD) and loci within TGFB2 (125, 126).

However, despite the numerous candidate gene studies and several GWAS linking TGF-β and related genes to asthma, so far this data has not been applied clinically. In general despite many novel asthma target genes being identified recently, not many of these discoveries have been used in the clinical environment. A major hurdle in applying GWAS findings is the unclear relationship between the novel gene-disease associations and biological mechanisms underlying them (117). Identification of drug targets, such as classical surface receptors and genomics-based drug repositioning, for drugs such as soproxil, olamkicept or inarigivir have been proposed as potential application of GWAS data in asthma (117). However, much GWAS data is still missing for asthma. Many asthma-related genes are missing functional annotations in GWAS, and many cell types relevant to asthma have not had expression quantitative trait locus (eQTL) mapping carried out. Mechanistic links between SNPs identified through GWAS and gene expression in asthma have to be established and validated in cell types involved in asthma pathogenesis (117).

Epigenetic studies are another recent approach used to study relationships between asthma, pharmacological interventions and the human epigenome. These studies have improved our understanding of the molecular basis of treatment response to corticosteroids, bronchodilators and immunotherapy, and how those pharmaceuticals interact with immune and inhibitory pathways, such as IL-2, TNF-α, NF-κB, and C/EBPs (127). Histone modifications are another layer of the epigenome, for which associations with allergic phenotypes, including those related to asthma, were discovered (128).

Importantly, several of these studies have discovered associations with TGF-β signalling pathways. In an epigenetic study of nasal epithelia taken from asthmatic Puerto Rican children, epigenomic responses to albuterol treatment in 22 genome-wide significant CpGs were characterised. TGFB1 was found to be one of the genes, the methylation patterns of which were influenced by albuterol treatment (129). RUNX3, a gene involved in TGF-β signalling, was found to have disparate methylation patterns in peripheral blood leukocytes from children with atopic asthma and non-asthmatic children (130). A genome-wide DNA methylation study examining leukocytes from asthmatic Latino children identified associations between epigenetic regulation of leukocytes and inflammatory signalling pathways. The strongest associations in this study were found for the TGF-β signalling pathway (131). A comprehensive evaluation of these epigenetic associations has the potential to contribute significantly to precision medicine development and new therapeutic target discovery. What is more, epigenetic patterns (including those of the TGF-β pathways) may provide novel tools in the diagnosis of allergic disorders and asthma.

Asthma exacerbations remain a significant burden for patients, especially young children (132–135). The TGF-β gene polymorphisms discussed in this review show potential as diagnostic factors. In particular, T allele of the −509C/T and C allele of the +869T/C polymorphisms in TGFB1 were shown to strongly associate with the risk of asthma development. Diagnostic use of these polymorphisms may facilitate earlier and more accurate diagnosis of asthma, allowing appropriate therapy to be initiated and the frequency of exacerbations to be reduced. However, nearly all of the studies published on the associations between TGF-β gene polymorphisms are based on candidate gene studies, and the representation of different ethnic groups should be improved. More studies based on robust and modern methods such as GWAS or NGS techniques, and focused on both genomic and epigenomic associations, need to be carried out on genetically diverse populations to provide stronger evidence and apply TGF-β gene polymorphisms in clinical practice as asthma development risk biomarkers.