Associations between Th17-related inflammatory cytokines and asthma in adults: A Case-Control Study

Chronic airway inflammation is recognized as an essential process in the pathogenesis of asthma. Cytokine profiles derived from immune and inflammation cells such as T-helper (Th) cells, eosinophilia and neutrophilia are not limited to the Th2 type in asthma. However, little is understood about associations between Th2-low inflammatory cytokine profiles and risk of asthma in adults. A case-control study of 910 adult asthma and 881 healthy controls was conducted. Inflammatory cytokines screening was undertaken by high-throughput protein microarray technology, and Th17-related inflammatory cytokines (IL17A, IL-9, adipsin and CCL11) were finally selected. Associations between these four cytokines and adult asthma risk were analyzed by multivariate logistic regression models. We observed that plasma IL-17A and IL-9 levels were significantly increased in asthmatics when compared with controls. However, the plasma expressions of adipsin and CCL11 in asthmatics were significantly lower than that in health controls. The adjusted ORs (95%CI) of association between IL-17A, IL-9, adipsin and CCL11 expressions and adult asthma were 3.08 (1.91, 4.97), 1.93 (1.41, 2.64), 10.02 (6.99, 14.37) and 3.29 (2.36, 4.59), respectively (all Ptrend < 0.0001). Our results suggested that elevated IL-17A and IL-9 expressions and decreased levels of adipsin and CCL11 were positively associated with adult asthma.

IL-4, IL-5 and IL-13 had the potential to stimulate epithelial cells, smooth muscle cells and fibroblasts that would contribute to airway hyper-responsiveness and remodeling [9][10][11] . Recently, accumulating evidences have shown that non-Th2 cytokines including IL-8, IL-10 and IL-17 are related with severity of asthma 7,12 . However, little is understood about associations between Th2-low cytokine profiles and risk of asthma in adults.
In this study, we applied high-throughput protein microarray technology to screen 280 plasma cytokines in adult asthmatics and healthy controls, and selected some specific inflammatory cytokines involved in Th2-low inflammation such as Th17-related signaling pathway. Information of occupational exposure, family history of asthma and lifestyle habits including smoking, drinking, and physical activity was collected for each participant. The aim of this study is to investigate associations of Th17-related inflammatory cytokines with asthma risk in adults.

Results
Basic characteristics. The basic characteristics of all participants are summarized in Table 1. The age of 1791 participants ranged from 18 to 86 years. The mean age was 45.64 ± 14.11 years for asthmatics and 46.63 ± 14.40 years for controls. Compared with healthy controls, asthmatics were more likely to be less physical activity, drinking, smoking, passive smoking and had a higher percentage of asthma family history and poor sleeping quality (P < 0.0001). Lung function parameters such as Forced vital capacity (FVC), FVC%, forced expiratory volume in one second (FEV 1 ), FEV 1 %, FEV 1 /FVC and peak expiratory flow (PEF) in asthmatic patients were significantly lower than those in healthy controls (P < 0.0001). There were no statistical differences in body mass index (BMI), occupational hazards exposure, keeping pets and planting flowers between asthmatics and healthy controls.
Among asthma patients, symptoms of cough and dyspnea in persistent asthmatics were significantly more severe than those in intermittent asthmatics (P < 0.01). However, there is no significant difference in the rate of cyanosis and wheezing sound between intermittent and persistent asthma cases (Table 2). Inflammatory cytokines levels in plasma samples. Table 3 displays the plasma levels of four inflammatory cytokines in the study population. Without adjustment for confounders, plasma IL-17A and IL-9 levels were higher in both persistent and intermittent asthmatics, however, plasma adipsin and CCL11 expressions were significantly lower than those in healthy controls (all P < 0.0001). Furthermore, elevated IL-9 levels and decreased expressions of adipsin and CCL11 in persistent asthmatics were statistical significance in comparison with intermittent asthmatics (P < 0.05).
In asthma patients, plasma IL-17A level was negative correlated with adipsin expression (r s = −0.13, P < 0.001). Moreover, there were significant correlations between plasma adipsin expression and levels of IL-9 and CCL11 with r s for −0.22 and 0.35, respectively. (all P < 0.001) ( Table 4). We also found similar relationships between these inflammatory cytokines in intermittent asthma and persistent asthma patients (see Supplementary  Tables S1 and S2). Table 5 presents the adjusted odds ratios (ORs) and 95% confidence (CIs) for asthmatic patients by quartiles of four inflammatory cytokines. After adjustment for age, sex, BMI, education, smoking, passive smoking, drinking, physical activity, family history of asthma, keeping pets, planting flowers and sleeping quality, we observed that ascending IL-17A and IL-9 levels and descending adipsin and CCL11 expressions were significantly related to increased risk of asthma (all P trend < 0.0001). The adjusted ORs (95%CI) of IL-17A, IL-9, adipsin and CCL11 for asthmatics in the highest quartile were 3.08 ( Supplementary Table S3).

Association between four inflammatory cytokines and asthmatics.
In addition, the abnormal IL-17A, adipsin and CCL11 levels were significantly associated with adult asthma. The adjusted ORs (95%CI) of abnormal expressions of elevated IL-17A, decreased adipsin and CCL11 for adult asthma patients were 12.50 (5.52, 28.34), 19.55 (11.74, 32.55), and 4.13 (1.78, 9.55) in comparison with normal level, respectively. And there were significant relationships between abnormal levels of elevated IL-17A or decreased adipsin and risk of persistent asthma in adults, with the adjusted ORs (95%CI) of 24.59 (9.69, 62.43) or 37.06 (19.40, 70.78) when compared with normal level, respectively.

Discussion
Asthma is a chronic airway inflammatory disease characterized by T-helper cell immune response, eosinophilic and/or neutrophilic inflammatory response. Recent studies reported that Th17 cells, known as one of Th2-low Clinical symptoms and signs Intermittent asthmatics (n = 537) Persistent asthmatics (n = 373) P value  cells, have a powerful influence on the pathogenesis of asthma [13][14][15][16] . Th17 cytokines such as IL-17A could result in airway remodeling by increasing neutrophil infiltration, mucous cell metaplasia, and smooth muscle mass 13,17,18 . Moreover, higher IL-17A expression was associated with expressions of IL-9, adipsin, and CCL11 [19][20][21] .
Our results found that plasma IL-17A was only significantly negative correlated with adipsin expression, but there were significant correlations between plasma adipsin expression and levels of IL-9 and CCL11. Actually, the process of cytokines interaction is a complex regulatory network in the development of diseases such as asthma. IL-9 is effective in inducing IL-17-producing cells when synergizing with TGF-β. Nevertheless, the IL-9 level could be negatively regulated by IL-23, which is independent of IL-17 signaling 19 . Thus, although IL-17A was not significantly correlated with IL-9 or CCL11, the influence of IL-17A on IL-9 or CCL11 expression may be mediated by adipsin or other regulators.
In this study, we observed that the plasma levels of IL-17A and IL-9 were significantly increased in asthma cases when compared with controls. However, the expressions of adipsin and CCL11 in asthmatics were significantly lower than that in health controls. The relationships became stronger when compared persistent asthmatics with healthy controls.
We found significantly positive association between plasma IL-9 production and adult asthma risk. This is in line with previous studies showing increased IL-9 and IL-9R expressions in lung tissue of asthmatics rather than health controls 22 . Anti-IL-9 monoclonal antibody was effective in treatment of mild asthma 23 . Recently, Elyaman' study suggested that IL-9 exerts a promoting effect on the differentiation of Th17 cells 19 . It is possible that IL-9 may play a crucial role in development of asthma, combining with Th17-related cytokines.
To our knowledge, this is the first report that plasma adipsin expression was associated with adult asthma. Our results indicated that plasma adipsin levels were statistically lower in asthmatics than healthy controls. Significant negative dose-response relationship was observed between plasma adipsin level and risk of adult asthma. Few previous publications reported adipsin expression in asthmatic patients. Leivo-Korpela and colleagues did not observe difference in adipsin level between female asthmatics and health controls in adults 24 . Fewer sample examples (35 non-smoking female asthmatics) prevented further analysis in their study. The exact mechanism of adipsin in the development of adult asthma is not clear. Adipsin, primarily expressed in monocyte-macrophages and adipocytes, could be obviously inhibited by IL-17 25 . Elevated plasma IL-17A expression in asthma patients   may partly contribute to low adipsin level in this study. Furthermore, adipsin, also called as complement factor D, could promote C5 to liberate C5a through alternative complement pathway. C5a overexpression has a critical impact on the pathogenesis of asthma after allergen provocation 26 . Maybe decreased plasma adipsin combined with depressed C5a expression could result in increased risk of asthma. The exact mechanism of adipsin interacting with Th17-related inflammation or complement factors needs to be further confirmed in pathogenesis of adult asthma. In addition, we observed that decreased plasma CCL11 level was associated with increased risk of adult asthma. Our result is partly supported by Dent's study, which suggested no difference in sputum CCL11 level between mild asthma and healthy controls 27 . Conversely, some clinical studies indicated that plasma CCL11 expression was higher in asthma attack, uncontrolled and severe asthmatic patients than that in health controls [28][29][30][31] . CCL11 (also known as extaxin-1) plays a pivotal role in recruitment of eosinophils into airways in asthma. Adult asthma predominated by eosinophilic response often develops into severe asthma, however, it seems to be mild or moderate asthma when regulated by Th2-low inflammation such as neutrophilic response through Th17 cells pathways 6 . Thus, decreased plasma CCL11 level might be regulated by IL-17 overexpression in mild to moderate adult asthma. Supporting by Schnyder-Candrian's study, CCL11 expression was largely inhibited by IL-17 administration and could be partly reversed by neutralizing IL-17 antibody treatment in lung tissues of allergic asthma 21 .
This study has several strengths. Firstly, we evaluated associations of Th17-related cytokines with risk of adult asthma in a relatively large sample size. Secondly, total of 280 cytokines expression profiles was measured to select valuable cytokines. Thirdly, potential confounding factors which are supposed to have impacts on asthma were considered in multivariate logistic regression models. However, in this study, there are two limitations. On the one hand, we did not analyze the influence of atopy on the associations between cytokine levels and risk of asthma because of only 9.12% asthma patients caused by atopy in this study. On the other hand, approximately 12.3% asthmatic patients might take medication to control symptoms before they went to hospitals, because they were previously diagnosed as asthma. Accordingly, prospective studies are essential to explore whether atopy or drug medications have influences on the associations of four inflammatory cytokines with risk of asthma in adults.
In conclusion, we found that elevated plasma IL-17A and IL-9 levels and decreased plasma expressions of adipsin and CCL11 were positively associated with asthma risk in adults.

Methods
Study population. All subjects were aged more than 18 and living at least 5 years in Wuhan city in China.
The asthmatics were outpatients and diagnosed by respiratory specialists from two hospitals in Wuhan between October 2010 and January 2012. The diagnosis of asthma was based on the presence of asthma symptoms and spirometry demonstrating airway hyperresponsiveness (Provoking Dose 20 (PD20) to methacholine <2.5 mg) and/or bronchodilator responsiveness (an increase of FEV 1 more than 12% and 200 ml after inhaling 200 μg salbutamol) according to the Global Initiative for Asthma (GINA) guidelines 2 . We excluded the asthma patients who had any infectious diseases, heart diseases and other kinds of diseases such as hypertension, diabetes and cerebral vessel disease. A total of 1009 eligible asthmatics were collected in this study.
According to living area of the cases, the health controls without respiratory diseases, allergic diseases, infectious diseases, heart diseases and other kinds of diseases such as hypertension, diabetes and cerebral vessel disease were recruited from the same residential areas. A total of 1009 health controls were matched with asthmatics for gender and age (±5 years).
After excluded 227 participants for inadequate plasma samples, 910 asthmatics and 881 healthy controls were enrolled in the final analysis. All the asthmatics were divided into two categories by clinical manifestations of asthma, including 537 intermittent asthmatics and 373 persistent asthmatics. The intermittent asthmatics were defined as patients with asthma symptoms less than once a week, night waking due to asthma no more than twice a month and normal FEV l . Patients were identified as persistent asthmatics when they presented these clinical features exceeding the criteria of intermittent asthma. And the persistent asthmatics were classified into three categories such as mild (n = 246), moderate (n = 96) and severe (n = 31) asthma patients based on the clinical manifestations and the current level of treatment required to control symptoms and exacerbations, according to GINA guidelines 2 .
Data collection. Basic data were collected by trained investigators via standard questionnaires by face-to-face interview, which covered information on socio-demographic and lifestyle such as education, occupational hazards exposure, family history of asthma, smoking, passive smoking and drinking status, physical activity, self-reported sleeping quality, and habits of keeping pets and planting flowers. Education level was categorized as low (below senior high school, ≤9 years), middle (senior high school, 9~12 years) and high (college degree or beyond, >12 years). Occupational hazards exposure was defined as whether individuals were exposed to industrial dust or chemicals in their workplaces. Physical activity was defined as regular exercise for more than 30 minutes at each time, at least once a week during the previous 6 months.
Clinical symptoms and signs of asthma patients such as cough, sputum, dyspnea, cyanosis and wheezing sound were collected by respiratory physicians. The cough status was classified into two types: occasional and chronic. The sputum status was classified into three categories: occasional, chronic and purulent. The occasional status indicated that the symptoms happen no more than three times once a week. The chronic status denoted that the symptoms happen last for more than two months. The degree of dyspnea was rated into 5-point (0 to 4) Grades according to modified Medical Research Council (mMRC) scale 32 . Then the severity of dyspnea was categorized as three types: mild (Grade 0 and 1), moderate (Grade 2) and severe (Grade 3 and 4).
Standing height and body weight were measured when individuals were standing with light indoor clothing and without shoes. Body mass index (BMI) was calculated as weight in kilograms divided by the square of height in meters.

Ethics Statement. The study was approved by the Ethnics and Human Subject Committees of Tongji
Medical School at the Huazhong University of Science and Technology . The methods were carried out in accordance with relevant guidelines and regulations. All participants enrolled in this study were given written informed consents for participation, storage and use of blood samples. Spirometric examinations. The spirometric examinations were conducted when participants were in sitting position with a nose clip after normal breathing at least 5 minutes by an electronic spirometer (Chestgraph HI-101, CHEST Ltd., Japan). Each participant was required to accomplish three acceptable curves and the optimal data were selected for analyses according to the recommendations of the American Thoracic Society 33 .
Inflammatory cytokines selection and determination. Approximately 5 mL fasting venous blood were collected in a tube with Ethylenediaminetetraacetic acid for each participant. On the same day of taking blood samples, the samples were centrifuged at 1500 rpm for 10 minutes at room temperature, then the plasma samples were isolated, aliquoted and stored at −80 °C until analysis.
Statistical analysis. Socio-demographic and lifestyle characteristics were compared by Chi-square test for categorical variables and Student's t-test for continuous variables. In the screening stage, differences of 280 cytokines antibody array data were analyzed using Significance Analysis of Microarray (SAM) 3.00 algorithm (http://statweb.stanford.edu/~tibs/SAM/index.html) 34 . SAM assigns a d-score to each cytokine on the basis of a two class paired analysis of expression changes and indicates significance by FC and q-value. The levels of four cytokines (IL-17A, IL-9, adipsin and CCL11) were not presented on normal distribution, so the differences in health controls, intermittent and persistent asthmatics were analyzed by kruskal-wallis-nemenyi test 35 . The correlations among these cytokines in asthma patients were analyzed by Spearman's rank correlation. The associations between these four cytokines and adult asthma risk were evaluated by multivariate logistic regression models, with adjustment for potential confounders including age, sex, BMI, education, smoking status, passive smoking status, alcohol consumption, physical activity, family history of asthma, keeping pets, planting flowers, sleeping quality. In addition, we also divided the levels of four cytokines into two categorical variables such as normal and abnormal levels. The abnormal levels of IL-17A and IL-9 were more than the P 90 levels of health controls, whereas the abnormal levels of adipsin and CCL11 were less than the P 10 levels of health controls. After adjustment for age, sex, BMI and family history of asthma, the abnormal levels of IL-17A, IL-9, adipsin and CCL11 with adult asthma risk were analyzed by multivariate logistic analysis. The adjusted odds ratios (ORs) and 95% confidence (CIs) for the risk of adult asthma were calculated. The statistical analyses were computed by using SAS, version 9.3, software (SAS Institute Inc., Cary, North Carolina). Two-sided values of P-value < 0.05 were considered as statistically significance.