Expression of eosinophil β chain-signaling cytokines receptors, outer-membrane integrins, and type 2 inflammation biomarkers in severe non-allergic eosinophilic asthma

Asthma is a common lung disorder that usually lasts the sufferer’s lifetime. Despite the wide range of treatment options, in some cases, patients encounter a significant loss of quality of life, reduced work productivity, greater mortality risk, and higher health care resource use and costs [1, 2]. Severe asthma affects less than 10% of all asthma cases, but the treatment costs are high [3], disease severity is a major factor for clinical and economic burden [4]. This rare phenotype has sub-classifications driven by distinct pathophysiological processes. One of the severe asthma subtypes is a severe non-allergic eosinophilic asthma (SNEA), which is characterized by severe clinical course, frequent exacerbations of the disease, resistance to treatment including high doses of systemic glucocorticoids, and poorer clinical outcome [5, 6] as well as a good response to biologicals such as treatment with anti-IL-5 antibodies. The basis of this disease is eosinophilic airway inflammation, which plays a pivotal role in the pathogenesis of both allergic and non-allergic asthma, but the paths through which the genesis is initiated varies. In allergic asthma cases, Th2 cells mainly release prototype cytokines such as interleukin (IL) 4, IL-5 and IL-13, thereby stimulating type 2 immunity, which has high antibody titers and eosinophilia. Non-allergic asthmatics have increased numbers of type 2 innate lymphoid cells (ILC2s), that upon stimulation with the epithelial cytokines produce Th2-associated cytokines, including high amounts of IL-5, which is important for eosinophilic maturation and migration [7, 8]. This might explain severe eosinophilic inflammation in the absence of classical Th2-mediated allergy [9]. However, the eosinophilic properties themselves may be responsible for some eosinophilic inflammation features.

Eosinophils are circulating granulocytes that are produced via granulopoiesis in the bone marrow. IL-3, IL-5, and granulocyte-monocyte colony-stimulating factor (GM-CSF) are cytokines that are particularly important in regulating eosinophil development [10]. These three cytokines, also called eosinophilopoietins, are responsible for eosinophil maturation, activation, survivability, and apoptosis [10]. To recruit circulating eosinophils from the blood into the lungs, they need to be activated [11, 12]. For the most part, it depends on cytokines IL-3, IL-5, GM-CSF, as well as their receptors [11, 13]. The receptors for IL-3, IL-5 and GM-CSF each have unique α -chains but share a common β-chain. The β-chain is essential for signal transduction and explains the overlapping activities of these cytokines. The overexpression of these cytokines or their receptors can lead to excessive or aberrant initiation of signaling, which results in pathological conditions [14]. A marked increase in eosinophil count in the tissues and blood is observed in SNEA cases, but data are lacking on whether this is due to the quantitative values of these three major cytokines or the change in expression of the receptor on the eosinophil.

Another important part of the inflammatory process in the asthma pathogenesis is direct eosinophil adhesion to a target cell, which is relevant for eosinophil activation and functions. Integrins are cellular receptors that contain an α and a β subunits that regulate extravasation of eosinophils from the bronchial circulation to the airway wall and airspace. Such movement into the asthmatic lung depends on integrins on circulating eosinophils. Eosinophils tether in flow and roll on bronchial endothelial cells, integrins on rolling eosinophils become further activated because of exposure to cytokines, eosinophils arrest firmly to adhesive ligands on the activated endothelium, and eosinophils transmigrate to the airway in response to chemoattractants [15, 16]. α4β1 and αMβ2 are likely the two most important integrins that mediate eosinophil adhesion and movement [16]. There is evidence that a change in the integrin expression can attenuate eosinophil-induced airway smooth muscle remodeling in asthma [17].

Currently-used type 2 inflammation biomarkers are peripheral blood eosinophils, sputum eosinophil count, exhaled nitric oxide concentration (FeNO), immunoglobulin E (IgE), serum periostin concentration, but some of them have limitations [18]. Sputum eosinophil count is calculated only in large centres, tissue-specific, time-consuming test. FeNO is cheap, not specific, loses specificity in smokers. No clear association has been identified between IgE or allergy as a biomarker of treatment responses or clinical outcome. Periostin measurement is poor available, confounded by growth in childhood, pregnancy, and dental disease. There is an increasing need for useful type 2 inflammation biomarkers, controlled the recruitment of activated eosinophils from the bloodstream into tissues, like CCL11 (eotaxin), CCL17, with predictive and prognostic value for the progression of the disease in SNEA patients and their link with clinical treatments.

We hypothesized that differences exist in eosinophil biological properties during SNEA compared with non-severe AA. Therefore, it was important to investigate factors that reflect eosinophil activity like specific cytokines and their receptors as well as integrin expression. The purpose of our study was to evaluate the expression of the main eosinophilopoietins IL-3, IL-5 and GM-CSF receptors and integrin subunits α4, αM, β1, β2 at the surface of eosinophils. Therefore, we investigated serum levels of type 2 inflammation biomarkers as eotaxin, periostin, and CCL17 chemokines.

Several clinical parameters, like eosinophil count, decreased lung function, uncontrolled course and frequent disease exacerbations, and resistance to glucocorticoid treatment, help clinicians to differentiate severe and non-severe courses of asthma. However, in many cases, these clinical data could overlap or be insufficiently expressed, which limits clinicians’ ability to appoint appropriate medication. Therefore, new biological differences between asthma phenotypes should be revealed. In this study, we aimed to find which disease pathogenesis-associated biological aspects of eosinophils in SNEA patients could help to explain more severe disease course. We found that IL-3, IL-5, GM-CSF receptors, and α4, β1 and αM integrin subunit expression by eosinophils is increased in SNEA patients compared with non-severe AA patients. Moreover, IL-5Rα changes function from downregulation in AA patients to upregulation in SNEA patients. Furthermore, we identified higher eotaxin and CCL-17 serum levels, which in combination with validated eosinophilic inflammatory biomarkers could help to distinguish severe eosinophilic asthma, and predict the disease course, treatment effectiveness, and outcome.

Currently, the eosinophil count is the best-identified marker to determine eosinophilic asthma severity [20]. However, the number of eosinophils in patients with different types of asthma can overlap significantly and cause confusion among clinicians about appropriate treatment. Increased eosinophil count in peripheral blood is important, but not the only factor that contributes to disease pathogenesis because only pre-activated cells could significantly contribute to disease pathophysiological features. Eosinophil infiltration from blood to asthmatic airway depends on primed blood eosinophils, which leads to their arrest on activated endothelium, extravasation into the airway wall, and migration through airway tissues to the airway lumen. Eosinophil pre-activation is associated with maturation and activation mediators like β receptor-signaling cytokines IL-3, IL-5, and GM-CSF in the blood [21]. These eosinophilopoietins are responsible not only for eosinophil maturation in bone marrow but they also contribute to the regulation of eosinophils surface proteins, including integrins, intracellular adhesion molecule-1, cytokines receptors, L-selectin, CD-13 and others [22]. In our study, we measured mRNA levels of these cytokine α receptors in eosinophils in different investigating groups. The relationships between gene expression of these receptors and its ligands – IL-3, IL-5 and GM-CSF is not completely clear, confirming that signal transmission through receptor depends not only on ligand concentration but also on the number of receptors. IL-3 serum levels did not differ between investigating groups, but IL-3Rα gene expression in the SNEA group was significantly increased, which may show increased IL-3 signaling. GM-CSF serum level was similar in SNEA and AA patients but significantly increased compared with HS. However, the gene expression of GM-CSFRα in SNEA patients did not differ compared with healthy controls but significantly decreased in the AA group. This suggests that though GM-CSF serum levels were similar in SNEA and AA patients, its altered receptor expression shows increased GM-CSF signaling in SNEA patients. Intriguing results were observed regarding IL-5 concentration and its receptor expression. The highest serum concentration of IL-5 was in the SNEA group; this significantly differed from those in the AA and HS groups. IL-5Rα also increased dramatically in SNEA patients, while it significantly decreased in the AA group. This reveals a flip from downregulation to upregulation, which highlights the important role of IL-5 signaling in SNEA patients. However, recent therapeutic approaches targeting IL-5 alone have not completely ablated tissue accumulation of eosinophils and have had limited effects on disease progression, which suggests important roles for IL-3 and GM-CSF. β-chain cytokines change their roles at a different stage of disease development; moreover, expression of their receptors is regulated by different signaling pathways. Our result partly agrees with the works of Gregory et al., and Uchiyama et al., who investigated how IL-3, IL-5, and GM-CSF cross-regulate expression of these cytokine α receptors in healthy eosinophils [23, 24]. These teams revealed that IL-5Rα was down-regulated and IL-3Rα was upregulated by all β-chain cytokines, while GM-CSFRα was downregulated by GM-CSF, but was not affected by IL-3 or IL-5. However, the situation in asthma could be different due to complex pathogenesis interrupted by chronic inflammation conditions. Our results showed that GM-CSF could not downregulate its receptor expression in SNEA patients. More interestingly, according to Gregory et al., and Uchiyama et al., IL-5Rα expression in SNEA patients decreased dramatically. However, we found that IL-5 serum levels increased over 7-fold, but there was an important shift in IL-5Rα, with a visible significant increase in its mRNA levels by over 5-fold compared with healthy controls and 8-fold compared with AA. There is no clear explanation about these significant changes. As blood eosinophils count in SNEA patients is significantly increased compared with AA and HS, signaling through IL-5Rα might take an important part in enhanced eosinophilic inflammation and their survival. Increased mRNA levels of IL-5Rα could be associated with the soluble form of this receptor that can be generated by differential splicing of mRNA transcripts [25] or by cleavage of surface receptors [26]. Wilson et al. demonstrated that increased serum IL-5 levels and blood eosinophil count, as observed in our SNEA patients group, elevates soluble IL-5Rα [27]. Moreover, receptors expression is primarily controlled by specific transcription factors, which role in the regulation of eosinophils IL-5Rα transcription is almost unknown in asthma. Therefore, this altered process can lead to increased mRNA levels of IL-5Rα and different AA and SNEA pathogenesis. This may have implications with respect to the use of novel therapeutic agents targeting IL-5 and its receptor in patients with eosinophilia. However, IL-5 is recognized as an effective biomarker for other diseases in which eosinophilia is detected [28, 29]; therefore, its role as a biomarker for asthma severity is not yet defined.

Increased expression of eosinophil integrins is an important factor for more stable and enhanced adhesion mediating cell activation by interacting with counter-receptors on other cells or an extracellular matrix [15, 30]. Eosinophil expresses seven-transmembrane heterodimeric integrins, but α4β1 and αMβ2 receive the most important in asthma [15, 31]. Previously, enhanced expression of these integrins was identified in newly-diagnosed AA patients with the significant blocking effect of small tetrapeptide Arg-Gly-Asp-Ser (RGDS) [17]. Moreover, higher numbers of α4β1 and αMβ2 integrins in an activated state are related with better eosinophil arrest on endothelium cells that express elevated amount of vascular cell adhesion molecules-1 (VCAM-1) – the ligand for α4β1 [32] and intercellular adhesion molecule-1 (ICAM-1) for αMβ2 [33] thus contributing to increased eosinophilic airway inflammation. Although evidence exists about eosinophil integrin expression during AA, there are none about the expression of these integrins during SNEA. In our study, we found that expression of αM, α4, and β1 integrin subunits in eosinophils from SNEA patients significantly differ from AA patients. This suggests increased adhesive properties of these integrins required for enhanced infiltration into the lungs. Studies with integrin-deficient mice revealed that α4 and β2 integrins enhance eosinophil recruitment to the airways [34]. However, our study did not show significant β2 expression differences between SNEA and AA groups, which suggests that an amount of β2 subunit in eosinophils is sufficient and that only increased expression of the αM subunit is required for enhanced eosinophil activation through αMβ2 integrin, as was found in the SNEA group (Fig. 1). Additionally, Johansson et al. discovered that blood eosinophils stimulated with IL-5, IL-3, or GM-CSF specifically adhere and migrate to periostin via activated aMβ2 [35]. We confirmed that serum levels of periostin increased in SNEA and AA patients compared with healthy controls, as well as by IL-3, IL-5, and GM-CSF depending signaling in SNEA. However, because in SNEA only IL-5 levels and the expression of its receptors significantly increases, this may be the main trigger that activates aMβ2 integrins in eosinophils, leading to their enhanced recruitment into the lungs. Periostin, as a ligand for aMβ2, may act as an activator for eosinophils, but more important is their role modulating eosinophil transmigration, chemoattraction, and adhesion [36]. All these features could be activated in an autocrine manner because eosinophil is a predominant periostin producing cells in the blood [37].

Some overlap in asthma phenotypes can lead to an overlap in eligibility for the different biological therapies. Therefore, more specific biological markers should be identified which, together with well-known clinical properties, could help to separate SNEA in early-stage investigations. Blood eosinophil count is a good marker to identify eosinophilic asthma phenotype, together with sputum eosinophils count, which is the best-identified marker for airway eosinophilia. The number of eosinophils in sputum can be used as a marker of disease severity, with strong associations among airway eosinophilia, the severity of disease symptoms, and worsened lung function [38]. Unfortunately, sputum eosinophil count and their activity tend to be influenced by the current, mostly corticosteroid therapy and are not consistent, and airway eosinophilia could prevail in many asthma phenotypes. Asthma has a multicomponent pathology, and sputum eosinophil count alone does not disclose variability in asthmatics [39]. Currently, sputum eosinophil measurements are used only in places with limited clinical equipment because of its complexity and practical difficulties.

Many studies revealed the periostin as a significant marker to identify tissue eosinophilia, but our research data suggest that periostin alone is not suitable to identify asthma severity because of similar possible corticosteroid-affected periostin levels in SNEA and steroid-free AA patient groups. Moreover, periostin, as a marker, according to its expression mechanisms, is linked to anti-IL-13 [40] or anti-IgE treatment [40], but not with IL-5 associated biological therapies. FeNO is weakly affected by corticosteroid treatment and could be a more effective biomarker for asthma severity [41]. We tested our subjects and found significant differences in FeNO concentration among SNEA or AA and healthy control study groups. Increased FeNO level in the exhaled gas of asthma patients is linked to T-helper type 2 (Th2)-mediated inflammation [42]. FeNO strongly correlates with a sputum eosinophilia and lung functions and could help to identify eosinophil recruitment to asthmatic lungs. However, our data indicate that though differences in FeNO level exist between SNEA and AA groups, the disparity is insufficient to distinguish these two asthma phenotypes. This finding is in agreement with several reports that indicated the discordance between treatment response or adjustment and FeNO levels, especially after anti-IL-5 treatment [43‐45].

In our study, eotaxin and CCL-17 protein act as the most selective markers because their levels in blood serum samples vary depending by asthma severity. One important study was made, indicating eotaxin as a potential diagnostic marker for asthma. Data analysis of thirty different studies revealed that eotaxin sputum level is a potential biomarker for asthma severity, but eotaxin serum level is also associated with lung function, sputum eosinophil count, eosinophil cationic proteins level, and FeNO [46]. Unfortunately, no anti-eotaxin approaches have been made so far. As Th2-mediated inflammation is associated with specific chemokines expression, like CCL-17 protein, by lung tissue cells for specific Th2 cell recruitment and activation, these chemokines in serum could act as specific biomarkers for disease severity. Significant correlations between CCL-17 and the severity of various diseases were published [47‐49], highlighting its usage in allergic conditions, but no data exist on non-allergic asthma with predominant eosinophilia. We think that the release of eosinophilopoietins by CCL-17 activated Th2 cells leading to eosinophil maturation in SNEA patients reveals the role of CCL-17 as a potential biomarker. Recently was demonstrated that CCL17 and eotaxin are regulated by IL-4/IL-13 pathway and using anti–interleukin-4 receptor α monoclonal antibodies significantly reduce serum levels of eotaxin and CCL-17 [50]. Different concentrations of eotaxin and CCL-17 between SNEA and AA patients can be explained by Th2 cells, or ILC2 generated IL-4 and IL-13 cytokine levels, leading to altered activity of IL-4/IL-13 signaling pathway and its role in SNEA and AA pathogenesis.

There are some limitations to our study. First, a small number of patients with SNEA were investigated. However, this was sufficient to find significant differences between our study groups. Moreover, ICS treatment could also affect gene expression and protein levels, but to date, there is only evidence about the inhibitory effect of ICS that indicates that our received differences should be even higher. Also, almost all SNEA patients were older age women; consequently, it is necessary to assess the possible effects of ageing and sex on study objectives. To date, it has not been determined whether eosinophils differ by gender. Adhesion of eosinophils and chemotaxis upon eotaxin stimulation do not differ significantly between younger and older subjects [51]. However, no data show how eosinophil receptor expression changes with ageing. It has been noted that eosinophil degranulation in response to IL-5 stimulation is significantly decreased in older patients [51], but data were obtained not in a disease context. Furthermore, eosinophil receptor and integrin expression were possible to investigate only in mRNA levels that in rare cases, do not correlate with protein levels due to several parameters [52].

We found that eosinophils in SNEA patients differ from those in AA because of increased gene expression of IL-3, IL-5, GM-CSFR receptors α-chains and α4, β1, αM integrins subunits, in conjunction with enhanced serum levels of IL-5. Moreover, increased serum levels of eotaxin and CCL-17 may become potential SNEA biomarkers.