Leptin favors Th17/Treg cell subsets imbalance associated with allergic asthma severity

Abstract Background Obesity has often been associated with severe allergic asthma (AA). Here, we analyzed the frequency of different circulating CD4+T‐cell subsets from lean, overweight and obese AA patients. Methods Mononuclear cells from peripheral blood were obtained from 60 AA patients and the frequency of different CD4+T‐cell subsets and type 1 regulatory B cells (Br1) was determined by cytometry. The effect of obese‐related leptin dose on cytokine production and Treg cell function in AA‐derived CD4+ T cell cultures was evaluated by ELISA and 3H thymidine uptake, respectively. Leptin levels were quantified in the plasma by ELISA. According to the BMI, patients were stratified as lean, overweight and obese. Results AA severity, mainly among obese patients, was associated with an expansion of hybrid Th2/Th17 and Th17‐like cells rather than classic Th2‐like cells. On the other hand, the frequencies of Th1‐like, Br1 cells and regulatory CD4+ T‐cell subsets were lower in patients with severe AA. While percentages of the hybrid Th2/Th17 phenotype and Th17‐like cells positively correlated with leptin levels, the frequencies of regulatory CD4+ T‐cell subsets and Br1 cells negatively correlated with this adipokine. Interestingly, the obesity‐related leptin dose not only elevated Th2 and Th17 cytokine levels, but also directly reduced the Treg function in CD4+ T cell cultures from lean AA patients. Conclusion In summary, our results indicated that obesity might increase AA severity by favoring the expansion of Th17‐like and Th2/Th17 cells and decreasing regulatory CD4+T cell subsets, being adverse effects probably mediated by leptin overproduction.


| INTRODUCTION
Asthma is a chronic condition of the lower airways classified as allergic and non-allergic. 1 This clinical condition can significantly reduce the patient's quality of life, in addition to generating economic impacts on health care systems. 2,3 Depending on the number of episodes and the therapeutic treatment required to control exacerbations, persistent asthma can be classified as mild, moderate or severe, with the last being potentially fatal due to the irreversibility of bronchial hyperresponsiveness, 4 even with the standard treatment applying long-acting β2 agonists (bronchodilators) and oral corticoids.
Asthma is a not a single entity, it comprises of complex disease involving different immune mechanisms (endotypes) and variable clinical presentations (phenotypes). Classically, allergic asthma (AA) is endotype triggered by innocuous environmental substances called allergens and involves the activation of allergen-specific T helper 2 (Th2) cells and production of immunoglobulin E (IgE). 5,6 The hallmark of this endotype is the production of high levels of interleukin (IL)-4, IL-5 and IL-13 that favor not only IgE production, but also activation of mast cells and eosinophils in the respiratory tract of AA patients following allergen exposure. Moreover, IL-5 and IL-13 produced by local group 2 innate lymphoid cells  in response to epithelial cell-derived IL-25, IL-33 and thymic stromal lymphopoietin protein (TSLP) amplify Th2-mediated AA. 6,7 In this endotype, the symptoms are particularly a result of biological actions of newly synthetized sulphidopeptide leukotrienes (LTC4, LTD4, and LTE4) and plateletactivating factor (PAF), 8 mainly produced by activated eosinophils.
These pro-inflammatory lipids play a pivotal role in the pathogenesis of acute attacks due to their ability to provoke local vasodilatation, edema formation, local neurogenic stimulation, smooth muscle contraction and mucus hypersecretion. However, as aforementioned, the pathogenesis of asthma is even more complex, and some patients, particularly those with resistance to inhaled corticosteroids, present an endotype non-T2 of the disease characterized by intense infiltration of neutrophils in the respiratory airways during exacerbation, 1,4,9 suggesting the involvement of Th17 cells in severe forms of asthma. 10 In humans, pathogenic Th17 differentiation appears to be induced by IL-23 high DCs, 11 with increased levels of both IL-17 and IL-23 being found in the serum and lungs of patients with severe asthma. 10,[12][13][14][15] Moreover, the mixed-granulocytic endotype, characterized by elevated levels of eosinophils and neutrophils in bronchialalveolar lavages, involves the induction of dual IL-17 and IL-4secreting CD4 + T-cells in some patients with severe asthma. 16,17 Interestingly, in vitro studies have indeed revealed a greater sensitivity of Th2 cells to glucocorticoids when compared to Th17 and IL-4 + Th17 cells. 18 Furthermore, in addition to effector CD4 + T cell subsets, asthma severity should also be associated with functional impairment of regulatory lymphocyte compartment capable of producing IL-10, such as T cells that express (Tregs), or not (Tr1), the FoxP3 marker, as well as type 1 regulatory B cells (Br1). [19][20][21] The existence of several endotypes of the disease with different responses to therapy might be associated with a complex and poorly understood relationship between genetic factors and environmental events, such as obesity.
Obesity can be defined as a chronic disease mediated by abnormal and excessive accumulation of fat in the body, leading to adverse effects on physical, social and mental wellbeing. 22 The prevalence of both asthma 23 and obesity 24 has increased over the past 20 years, and prospective epidemiological meta-analysis studies indicate that asthma and obesity coexist in many patients. 25 In these patients, obesity is an important risk factor for greater frequency and severity of asthma exacerbation and poor response to therapy. 26 This adverse association may involve increased production of different adipokines, such as leptin.
In addition to the increased frequency of both pro-inflammatory M1 macrophages and effector CD4 + and CD8 + T-cells in the adipose tissue, visceral obesity is normally characterized by hyperleptinemia. 27 Apart from its role in regulating balance energy expenditure and nutritional status, leptin is also critical for normal T cell response. 28 Nonetheless, elevated leptin levels have been associated with hypersensitive reactions mediated by both Th2 29 and Th17 30 phenotypes. Moreover, by inducing tumor necrosis factor-α (TNF-α) and IL-6 production, obesity-associated leptin levels also damage Treg function, which reduces the production of antiinflammatory cytokines such as IL-10. 31,32 In asthmatic patients, high leptin levels were inversely correlated with lung function in asthmatic patients. 32 In a recent study published by our group, 30 a direct correlation between AA severity with both plasma leptin concentrations and levels of IL-5, IL-6 and IL-17, released by purified CD4 + T-cells, was observed. Nonetheless, this study did not analyze the relationship between different CD4 + T cell phenotypes and AA severity. Therefore, the objective of the present study was to characterize the proportions of different effector and regulatory CD4 + T cell subsets, as well as Br1, according to AA severity. Also, the direct effect of obese-related leptin dose on the cytokine production by CD4 + T cells was additionally investigated. It is possible that immune imbalance associated with obesity and elevated leptin levels favor the expansion of different CD4 + T cell subsets associated with AA severity.  [6][7][8][9][10][11][12][13][14][15][16][17]. Patients were allowed to receive treatment with inhaled corticosteroids for 2 months prior to the study, but not with systemic steroids. All patients had a positive skin prick test, defined as a >5-mm diameter skin wheal response to at least 1 of 6 common allergens (Dermatophagoides pteronyssinus, D. farinae, Alternaria, mixed grass pollen, dog and cat hair). The great majority of patients are polysensitized (15% of patients had positive reaction to 1 or 2 allergens and 85% had a positive reaction to 3 or more allergens). The presence of rhinitis was observed in 65% (n = 39) and the severity of symptoms was determined by using total nasal symptom score (TNSS) (sneezing, congestion, itching, and rhinorrhea) 33 (Table 1). Of note, relevant clinical allergens were recorded to dust mites (n = 31), cat dander (n = 12), dust mites and cat dander (n = 13), Alternaria (n = 2) and mixed grass pollen (n = 2).

| Subjects
The occurrence of infectious or other autoimmune diseases were excluded by clinical and serological tests. Twenty healthy subjects (15 women and 5 men), matched by age, gender and body mass index (BMI) were also recruited for the control group. BMI is calculated from the mass (weight in Kg) and height (in meters) of an individual adopting the formula (BMI = kg/m 2 ). Subjects were stratified as lean (BMI from 18.5 to 24.9), overweight (BMI from 25 to 29.9) and obese class I (BMI from 30 to 35) according to BMI. In the present study, all subjects included were nonsmokers, with no history of upper or lower airway infectious diseases 4 months prior to recruitment in the study. We also excluded individuals taking oral or intravenous steroids, theophylline, long-acting β2-agonists, leukotriene antagonists or antihistamines 2 months prior to the study. The Ethics Committee for Research on Human Subjects at the Federal University of the State of Rio de Janeiro (UNIRIO) approved the study, and blood was collected only after written informed consent was obtained from each individual.

| Leptin quantification
Circulating leptin levels were measured using a commercial ELISA kit

| The effect of leptin on cytokine production by CD4 + T cells
The CD4 + T cells were obtained from PBMC via negative selection using magnetic columns according to manufacturer's instructions

| The role of leptin in modulating Treg function
In

| Statistical analyzes
All statistical analyzes were conducted using the Prism 8.0 pro-  were recruited and subsequently stratified by body mass index (BMI).

| Impact of clinical status and obesity on different effector CD4 + T-cell subsets in AA patients
As expected, most patients also suffered from rhinitis (Table 1).
Twenty healthy subjects (15 females and 05 male) were also recruited as control (Table 1). Of note, no difference in the CD4 + T cell and B cell counts was observed between healthy subjects and AA patients in the different clinical subgroups. Our primary objective was to evaluate the proportion of total IL-4 + CD4 + T cells in AA patients. Following the gating strategy shown in Figure 1A Figure 1C). Interestingly, in patients, the severity of the disease was associated with an expansion of IL-4 + IL- Figure 1D) rather than classic Th2-like cells ( Figure 1C). This hybrid Th2/Th17 cell subtype was practically absent in the control group ( Figure 1D). Among CD4 + T cells negative for IL-4, the proportion of IFN-γ + IL-17 -(type Th1) was significantly higher in healthy individuals in comparison to patients ( Figure 1E). In the AA group, an important decrease in the frequency of Th1 cells was observed in patients with severe AA ( Figure 1E). In contrast, a higher frequency of Th17-like cells was detected in patients with severe AA when compared to all other individuals ( Figure 1F). No difference was observed in the proportion of this phenotype between the control group and patients with mild or moderate AA ( Figure 1F). Finally, the percentage of IL-4 − CD4 + T cells capable of co-expressing IL-17 and IFN-γ was low and no difference between the different groups of subjects was observed ( Figure 1G).
Regarding the BMI of patients, although the percentage of total CD4 + T cells had not been associated with fat mass, the frequency of Th2-like cells was significantly lower in obese individuals with severe disease when compared to mild and moderate AA (Figure 2A). No significant difference was observed in the groups with mild and moderate AA in terms of Th2 cells (Figure 2A). On the other hand, a higher percentage of the hybrid Th2/Th17 phenotype was observed in obese patients with severe AA, with no significant difference between patients with mild or moderate forms of the disease ( Figure 2B). Similarly, the highest percentages of Th17-like cells were observed in obese AA patients, especially those with severe disease ( Figure 2C). In contrast, both the severity and excessive weight gain negatively impacted the frequency of Th1-like cells in AA patients ( Figure 2D).

| Obesity amplifies damage in the regulatory lymphocyte compartment in patients with AA
Following the gating strategy for identifying CD4 + T cells that express the FoxP3, IL-10 and CD39 markers demonstrated in Figure 3A, we observed a lower frequency of IL-10 + Foxp3 + CD4 + T cells (Tregs) in patients with severe AA when compared to the control group. Among IL-10 + FoxP3 − CD4 + T cells, known as Tr1 cells, 34 a lower percentage was observed in patients with moderate and severe AA when compared to the control group ( Figure 3B), without statistical difference between healthy individuals and mild AA patients.
Amongst the patients, the lowest percentage of Tr1 cells was observed among those with severe AA ( Figure 3B). The acquisition of the CD39 marker has been associated with high functional capacity of FoxP3 + CD4 + T-cells. 35 Figure 3C shows that the percentage of these cells was significantly higher in healthy individuals when compared with patients with moderate and severe AA, but not when compared to those with mild AA. Amongst patients, the frequency of this subtype of regulatory T cell was highly variable with no significant difference according to AA severity ( Figure 3C). As for Br1 cells, identified as IL-10 + CD19 + cells ( Figure 3D), a lower percentage was identified in the samples obtained from patients with moderate and severe AA when compared to the control group ( Figure 3E). Among patients, the frequency of Br1 cells was significantly higher in those with mild AA compared to individuals with severe forms of the disease ( Figure 3E).
Regarding BMI, obesity was associated with a lower percentage of IL-10 + FoxP3 + CD4 + T-cells only in patients with severe AA when compared to other patients and the control group ( Figure 4A). In addition, the percentage of CD39 + FoxP3 + CD4 + T-cells was significantly lower in overweight or obese patients who presented moderate or severe forms of AA ( Figure 4C). As for Tr1 cells, a lower frequency of this cell subtype was observed only in obese patients with moderate and, mainly, severe AA ( Figure 4B). As demonstrated in Figure 4D, the presence of obesity was also associated with a lower percentage of IL-10 + CD19 + cells among patients with mild, moderate and, mainly, severe AA.

| Elevated plasma leptin levels are related to an imbalance of CD4 + T-cell subsets and Br1 cells implicated in AA severity
Many of the immune disorders in obese patients have been associated with high production of certain adipokines, particularly leptin. [28][29][30][31][32]36,37 As expected, plasma leptin levels were significantly lower in lean subjects from the control group and AA patients when compared to their overweight/obese counterparts ( Figure 5A). However, and interestingly, among lean subjects, leptin concentrations were higher in severe VOLLMER ET AL.  Figure 5F). In the control group, similar results were observed.
While leptin levels positively correlated with the percentage of Th17like cells, a negative correlation was observed between this adipokine and the frequency CD39 + FoxP3 + CD4 + T-cells and Tr1 cells ( Figure S2).

| Effect of leptin on in vitro cytokine production and Treg function in CD4 + T cells from AA patients
Previous findings demonstrated the relationship between the plasma leptin levels and the frequency of different effector and regulatory T cells in AA patients. As demonstrated in Figure 6, higher levels of IL-5 ( Figure 6B), IL-6 ( Figure 6C), IL-13 ( Figure 6E), and IL-17 ( Figure 6F), associated with a lower release of IL-10 ( Figure 6D), were observed in purified CD4 + T cells from more obese than leaner AA patients.
Interestingly, obesity-related leptin concentration significantly increased the ability of purified CD4 + T cells from lean AA patients to produce IL-5, IL-13, IL-17 and IL-6, but diminished IL-10 release. No significant difference was observed with regard to IL-4 ( Figure 6A) and IFN-γ (data not shown). Although no significant difference was observed for pro-inflammatory cytokines released by obese-derived cell cultures in response to leptin, this adipokine reduced IL-10 levels ( Figure 6D). Interestingly, in addition to reducing the proportion of IL-10-secreting CD4 + T cells, leptin significantly damages the ability of Tregs to inhibit Tresp proliferation (Figure 7) in cell cultures from lean (n = 5) and obese (n = 5) AA patients.

| DISCUSSION
Classically, AA has been described as an adverse immune event  In addition to the involvement of different effector CD4 + T cell subtypes, development of AA is also associated with damage in IL-10 production by CD4 + T and B cells. 42,43 Here, both AA severity and obesity negatively affected the frequency of Tregs and Tr1 cells. Moreover, higher plasma leptin levels correlated inversely with those cells. Also, obesity-related leptin level not only decreased IL-10 production by CD4 + T cells, but also reduced the in vitro suppressive function of these lymphocytes from lean and obese AA patients.
Like IL-10, a cytokine that plays a potent anti-inflammatory role, the CD39 molecule has also been described as a marker of Treg cell function. 35 CD39, in association with CD73, metabolizes the extracellular adenosine triphosphate (ATP) molecule into adenosine monophosphate (AMP), a potent immune inhibitor. 35 Here, lower frequency of CD39 + FoxP3 + CD4 + T-cells was observed in patients with moderate and severe AA. Further, a lower frequency of these cells was found in patients with higher plasma levels of leptin. This finding is new for AA, but a study by Cortez-Espinosa et al. 44 observed a negative correlation between the proportion of CD39 + Treg cells with IBM in patients suffering from type 2 diabetes.
In addition to T cells, IL-10-secreting B (Br1) have been associated with protection against allergic diseases. 42,43 In the present study, we observed a decrease in the proportion of Br1 cells in patients with moderate and, mainly, severe AA. Among these patients, the lowest proportion of Br1 was seen in obese patients with high plasma leptin levels.
Although preliminary, our findings showed that obesity favors the expansion of circulating Th17-like cells and hybrid Th2/Th17