The Role of Chitinases in Chronic Airway Inflammation Associated with Tobacco Smoke Exposure

Chitinases and chitinase-like proteins are thought to play a role in innate inflammatory responses. Our study aimed to assess whether chitinase concentration and activity in induced sputum (IS) of patients exposed to tobacco smoke are related to the level of airway inflammation including the level and activity of chitinases and chitinase-like proteins. The study included 22 patients with chronic obstructive pulmonary disease (COPD), 12 non-COPD smokers, and nine nonsmoking subjects. Sputum CHIT1 and YKL-40 levels and chitinolytic activity were compared with sputum IL-6, IL-8, IL-18, and MMP-9 levels. A hierarchical cluster analysis was also performed. Sputum YKL-40 was higher in COPD patients than in the control groups. Sputum CHIT1 and YKL-40 levels correlated with IS inflammatory cell count as well as with MMP-9 and IL-8 levels. Two main clusters were revealed: Cluster 1 had lower chitinase levels and activity, lower IS macrophage and neutrophil count, and lower IS IL-8, IL-18, and MMP-9 than Cluster 2. Comparison of COPD patients from both clusters revealed significant differences in the IS inflammatory profile despite comparable clinical and functional data. Our findings seem to confirm the involvement of chitinases in smoking-associated chronic airway inflammation and show that airway chitinases may be a potential novel marker in COPD phenotyping.


Introduction
Chitinases, represented by chitotriosidase (CHIT1) and acid mammalian chitinase (AMCase), are a group of proteins involved in the innate immune response to chitincontaining organisms (i.e., parasites and fungi) [1,2]. They function by binding to chitin chains and hydrolyzing them to low molecular weight chitooligomers [3]. Chitinase-like proteins (CLPs) are a group of proteins which do not have hydrolytic activity but are still capable of binding chitin or chitin particles [3]. The most prominent CLP representant is chitinase-3-like protein 1 (CHI3L1), also known as YKL-40 [1]. An imbalance of chitinase activity has been reported in several inflammatory lung disorders, including asthma, chronic obstructive pulmonary disease (COPD), and fibrosing lung diseases [1][2][3][4][5]. Recent studies are focused not only on the role of chitinases and CLPs in airway diseases but also on chitinases as a therapeutic target [6,7].
The primary cell sources of CHIT1 and YKL-40 in the lung are macrophages and neutrophils [8,9]. YKL-40 is also produced by epithelial cells [9]. Multiple triggers, including environmental toxins and tobacco smoke exposure, may enhance chitinase and CLP production [1]. It has been shown that CHIT1 and YKL-40 may simultaneously stimulate alveolar macrophages and epithelial cells to secrete proinflammatory and profibrotic mediators resulting in tissue inflammation, alveolar destruction, and tissue remodeling [1,5].
Smoking is known to be associated with an upregulation of chitinase gene expression and chitinase levels were found to be higher in smokers than those who had never smoked. This phenomenon was demonstrated not only in studies comparing patients with airway diseases with both smokers and non-smokers [8,10], but also in studies conducted in smoking and non-smoking patients affected by extrapulmonary disorders [11,12].
The role of chitinases in the pathogenesis of COPD is a relatively novel point of interest. With the growing evidence for their contribution to COPD-related inflammatory pathways and their potential application as a marker for smokers at risk of developing the disease [12,13], chitinases are a promising area of research. However, there is still little known about the mutual relationships between chitinases and other COPD inflammatory markers. Another understudied area is the role of chitinases and CLPs as potential novel treatment targets for COPD; considering that the efficacy of chitinase inhibitors is currently being investigated for the treatment of lung diseases in in vitro and animal models [14] and that therapeutic options in COPD are still limited and unsatisfactory in terms of effectiveness, targeting chitinases and CLPs is an attractive possible solution.
In the light of the above, we undertook a study aimed at assessing the levels and activity of CHIT1 and YKL-40 in induced sputum, as well as the mutual relationships between CHIT1/YKL-40 and various features of airway inflammation in patients with COPD, smokers without COPD, and people who have never smoked. Our working hypothesis, based on previously published data and our own results, was that chitinases/CLPs are not specifically related to COPD, but rather to the level of airway inflammation due to exposure to noxious factors, mainly those present in tobacco smoke. Thus, the specific aims were to: (1) compare chitinase concentration and activity in induced sputum (IS) in the investigated groups, (2) search for potential relationships between sputum chitinase concentration/activity and other inflammatory biomarkers, and (3) correlate chitinase concentration/activity with the clinical profiles of the investigated patients.

General Study Design and Participants
This prospective, observational study was performed by the Department of Internal Medicine, Pulmonary Diseases and Allergy of the Medical University of Warsaw between May 2018 and February 2019 and included mild-to-moderate COPD patients and controls who agreed to participate in the project. The ethics committee of the Medical University of Warsaw (KB/67/A/2018) approved the study. All participants signed an informed consent form before participating.
The scope of this study included: medical history, with particular attention given to signs and symptoms assessed using the modified Medical Research Council (mMRC) scale and the COPD Assessment Test (CAT), smoking and exacerbations history, physical examination, spirometry, and sputum induction.

Subjects
Forty-three subjects were enrolled, including 22 patients with COPD, 12 smokers without COPD, and 9 healthy nonsmokers. Inclusion criteria were as follows: (1) age above 40 years, and (2) diagnosis of COPD based on past medical history, data on noxious inhalants exposure, typical signs and symptoms, and irreversible airway obstruction identified by spirometry. We defined airway obstruction using a z-score of −1.645 (5th percentile) as the lower limit of normal [15] and divided COPD patients into four categories (A-D) depending on symptom severity and exacerbation history as recommended in the Global Initiative for Chronic Obstructive Lung Disease (GOLD) 2018 guidelines [16].
The inclusion criteria applied for the control groups (smokers and nonsmokers) comprised of: age above 40 years, a negative history of any chronic lung diseases, and an absence of respiratory symptoms and normal spirometry. Smokers were defined as individuals with 10 or more pack-years of smoking history. Patients who had previously smoked 10 or more pack-years but denied smoking one year before the study onset were defined Cells 2022, 11, 3765 3 of 13 as ex-smokers. Patients with no smoking history (less than 1 pack-year) were classified as non-smokers.
Universal exclusion criteria for all subgroups included: symptoms of acute respiratory tract infection in the previous 6 weeks, oral and inhaled steroid treatment within 6 weeks preceding sample collection, active cancer disease, and any worsening of comorbidities (i.e., heart failure) within 6 weeks before the examination.

Spirometry
Spirometry testing was performed using a Lungtest 1000 spirometer (MES, Cracow, Poland) according to the American Thoracic Society/European Respiratory Society (ATS/ERS) guidelines [17]. Pre-and post-bronchodilator forced expiratory volume in the first second (FEV 1 ), forced vital capacity (FVC), and FEV 1 /FVC% were measured in all COPD patients, while only baseline measurements were performed in control subjects (no bronchodilators administered).

Induced Sputum Collection and Processing
Sputum induction was preceded by inhalation of 400 µg of salbutamol. Next, inhalations of sterile hypertonic saline (NaCl) were applied at increasing concentrations (3,4, and 5% solutions) via an ultrasonic nebulizer (ULTRA-NEBTM2000, DeVilbiss, Somerset, PA, USA), following ERS recommendations [18]. Induced sputum samples were processed as described elsewhere [19]. Sputum smears were used to assess total and differential leukocyte count. Induced sputum supernatants were stored at −80 • C and then used to measure cytokine concentrations, YKL-40 and CHIT1 levels, and chitinolytic activity.

Chitinolytic Activity Analysis
Chitinolytic activity was detected using the procedure described previously [20]. The concentrations of some substrates and volumes of analyzed samples were modified to adjust the procedure to human materials [6]. Substrate hydrolysis product 4-methlyumberlliferone was measured fluorometrically using a Tecan Spark 10M microplate reader (Tecan, Switzerland) (excitation 355 nm/emission 460 nm). Chitinolytic activity in human samples was measured at pH 2 for the AMCase activity only and at pH 6 for both AMCase and CHIT1. Each sample was tested twice.

Statistical Analysis
Estimation of the sample size was based on data from an earlier study by Létuvé S et al. [21]. To detect between-group difference with a power of 80% and a significance level of 5%, the required sample size was estimated to be 42 subjects (21 COPD patients and 21 controls).
Statistical analysis was performed using Statistica 13.3 software (StatSoft Inc., Tulsa, OK, USA) or R environment (version 4.0.5; R Foundation for Statistical Computing, Vienna, Austria). Data were presented as median and interquartile range (IQR) or percentage (%). Differences between groups were tested using a Pearson's chi-squared test for categorical variables, a Mann-Whitney U test for comparisons between two independent groups, or a Kruskal-Wallis test followed by a Dunn's post hoc test for multiple comparisons between groups. A p-value of less than 0.05 was accepted as indicating significance. Correlations between data were assessed using Spearman rank tests. Hierarchical clustering was performed on standardized data using Euclidean distance Ward D linkage. Outliers were normalized to mean +/−3SD (standard deviations). Logistic regression was performed using data prepared using the same method as for hierarchical clustering (i.e., standardized and outliers clipped).
Data on cellular compositionconcentrations of cytokines, chitinases, and chitinolytic activity in IS are given in Table 2. Patients with COPD were characterized by the highest sputum IL-8 and YKL-40 levels ( Table 2 and Figure 1). YKL-40 was undetectable in 75% of smoking controls, 33% of non-smoking controls, and only 14% of COPD patients. There were no differences in CHIT1 concentration and chitinolytic activity between the investigated groups (Table 2 and Figure 1). Chitinolytic activity was detected only at pH 6 and was highly correlated with sputum CHIT1 levels (r = 0.71, p < 0.05), suggesting that CHIT1 is the primary source of chitinolytic activity in the airways.
Using a logistic regression model with selected clinical and biochemical parameters that we created, we found that YKL-40 is an important inflammatory factor in COPD pathobiology (r-coefficient of 1.69 and p-value of 0.02) ( Table 3).
Analysis of the entire investigated group revealed a significant negative correlation between IS YKL-40 and spirometry parameters (FEV 1 % of predicted and FEV 1 /FVC %, r = −0.56, r = −0.66, respectively, both p < 0.0001). Chitinase levels and activity in IS did not correlate with any other clinical features ( Figure 2). Table 2. Sputum parameters in study participants.

COPD (n = 22) p-Value
Total cell count (×10 6 /g)  was highly correlated with sputum CHIT1 levels (r = 0.71, p < 0.05), suggesting that CHIT1 is the primary source of chitinolytic activity in the airways.

COPD (n = 22) p-Value
Total cell count (×10 6 /g)    pathobiology (r-coefficient of 1.69 and p-value of 0.02) ( Table 3). Analysis of the entire investigated group revealed a significant negative correlation between IS YKL-40 and spirometry parameters (FEV1% of predicted and FEV1/FVC %, r = −0.56, r = −0.66, respectively, both p < 0.0001). Chitinase levels and activity in IS did not correlate with any other clinical features (Figure 2). A significant correlation between sputum CHIT1, YKL-40, and chitinolytic activity and both total cell count and the number of sputum neutrophils was found. Moreover, significant positive correlations were found between sputum macrophage counts, CHIT1 level, and chitinolytic activity. There were also positive correlations between IS MMP-9 Cells 2022, 11, 3765 7 of 13 and chitinase levels and chitinolytic activity, and between sputum IL-8 and YKL-40 levels ( Figure 2).
A clustering analysis involving IS data from COPD patients and control groups was performed to discover any differences in the features of airway inflammation (Figure 3). Two main clusters were revealed; the clinical characteristics of both clusters are shown in Table 4.
A clustering analysis involving IS data from COPD patients and control groups was performed to discover any differences in the features of airway inflammation (Figure 3). Two main clusters were revealed; the clinical characteristics of both clusters are shown in Table 4. Cluster 1 included patients with COPD (38%), control smokers (35%), and non-smokers (27%) with a preserved gender balance. Cluster 2 included only patients with COPD (100%) and showed male predominance. There was no difference in age and BMI between the two clusters. Patients from Cluster 2 had worse lung function and greater tobacco smoke exposure (Table 4). This cluster also had a significantly higher proportion of total  Cluster 1 included patients with COPD (38%), control smokers (35%), and non-smokers (27%) with a preserved gender balance. Cluster 2 included only patients with COPD (100%) and showed male predominance. There was no difference in age and BMI between the two clusters. Patients from Cluster 2 had worse lung function and greater tobacco smoke exposure (Table 4). This cluster also had a significantly higher proportion of total sputum cells and higher macrophage and neutrophil counts. Moreover, chitinase concentration (CHIT1 and YKL-40) and chitinase activity were significantly higher in Cluster 2 than in Cluster 1. The concentration of sputum cytokines IL-8, IL-18, and MMP-9 was elevated in Cluster 2, but the level of IL-6 was higher in Cluster 1 (Table 5). In the following step, we compared the clinical and biochemical profiles of patients with COPD from Clusters 1 and 2. No differences in sex distribution, age, BMI, smoking exposure, or exacerbation frequency and symptom severity (as reflected by mMRC and CAT scores) were found. There was no statistically significant difference in spirometry test results (Table 6). Patients with COPD from Cluster 2 were characterized by a significantly higher total IS cell count and a greater number of sputum neutrophils and macrophages. They also had higher levels of sputum IL-8, IL-18, and MMP-9 and higher chitinase levels and activity. Interestingly, the concentration of CHIT1 in sputum COPD patients from Cluster 1 was below the detection threshold (Table 7). Table 7. Sputum cellular characteristics and mediator levels in COPD patients from the two identified clusters.

Discussion
Our study showed an interesting relationship between the level of airway inflammation due to chronic tobacco smoke exposure, expressed as the number of different inflammatory cells in IS, and both CHIT1 and YKL-40 levels and chitinolytic activity. This was demonstrated by a significant positive correlation between the sputum total cell, neutrophil, and macrophage counts and the levels of sputum YKL-40, CHIT-1, and chitinolytic activity in all the investigated groups. Moreover, our study revealed a significant positive correlation between the concentration of sputum MMP-9 and IL-8 and the levels of sputum CHIT1, YKL-40 levels, and CHIT1 activity. Sputum YKL-40 was significantly higher in patients with COPD than in both smoking and non-smoking controls, while CHIT1 and chitinolytic activity levels were comparable in the three investigated groups. A logistic regression model including a combination of clinical and biochemical variables confirmed that elevated IS YKL-40 level is a risk factor associated with COPD.
At the same time, cluster analysis showed that significant differences in chitinase concentration and activity among patients with COPD may occur. To our knowledge, hierarchical clustering analysis was not used in previous studies on chitinases in lung diseases and our study is the first to apply this approach. By demonstrating significant differences between the identified clusters, we confirmed different inflammatory patterns in the disease, supporting a potential role of precision medicine in COPD management.
Bearing in mind that macrophages, neutrophils, and epithelial cells are the main source of airway chitinases [8,9], the positive correlation between sputum cell count, neutrophils, and macrophages and chitinase levels and activity could be expected, however the existing data for COPD are not unequivocal. Otsuka et al. showed that sputum YKL-40 correlated with sputum neutrophils in both COPD and asthma and with sputum macrophages in COPD, although such an association was not found for controls [9]. In a study comparing sputum biomarkers in idiopathic pulmonary fibrosis (IPF) in which patients with COPD comprised one of the control groups, Guiot et al. demonstrated that sputum neutrophil count was significantly higher in patients with IPF and those with COPD compared to healthy controls, however, both sputum levels of YKL-40 and YKL-40 expression did not differ between the groups [22]. Majewski et al., who compared patients with COPD with smoking and non-smoking controls, found a higher sputum neutrophil and macrophage count in patients with COPD vs. non-smoking controls. This was also the case with sputum YKL-40 and both chitotriosidase levels and chitotriosidase expression, but not with the expression of YKL-40 [10]. It should be emphasized that our study is different than all previous studies on airway chitinase/CLPs levels in terms of the general approach. Namely, we were not focused on the chitinase and CLP levels in specific airway diseases but rather on their levels and activities in patients exposed to noxious gases and fumes that can induce and sustain airway inflammation. This was a consequence of our working hypothesis that the level of chitinases/CLPs in the airways is less associated with the specific disease, i.e., COPD, and more with the level of airway inflammation. As we found that participants with chronic airway inflammation (manifested by a higher number of total sputum cells, macrophages, and neutrophils) had significantly higher chitinase concentrations and activity, we believe our results can be construed as supporting this hypothesis. This observation also seems to align with the results of some earlier studies which suggest that CHIT1 and YKL-40 can be markers of inflammation associated with fibrosis and matrix remodeling in the lungs [8,22]. The lack of correlation between the level of cigarette smoking exposure and chitinase levels probably results from the complexity and diversity of airway inflammation. Tobacco smoke exposure is one of the numerous triggers and driving factors of airway inflammation. Moreover, the inflammatory effect of noxious gases and inhaled particles varies and depends on the individual susceptibility and characteristics of airway inflammatory responses.
Our analysis of chitotriosidase and YKL-40 in the context of other sputum inflammatory markers showed a significant positive correlation between CHIT1, YKL40, and chitinolytic activity and MMP-9 and a positive correlation between YKL-40 and IL-8. Furthermore, one of the two identified clusters characterized by elevated chitinase levels and chitinolytic activity also had significantly higher sputum MMP-9, IL-8, and IL-18 levels. These results are in line with earlier findings on the relationship between YKL-40, MMP-9, and IL-8. It has been documented that bronchial epithelial cells treated with YKL-40 show increased IL-8 production [23] and exposure of alveolar macrophages to YKL-40 or CHIT1 promoted the release of IL-8 and MMP-9 [8,22]. Our results confirm these observations. Cluster analysis of the investigated subjects revealed some interesting findings. Two main clusters were identified. Cluster 1 comprised participants from all investigated groups (COPD, non-COPD smokers, and non-COPD non-smokers) and was characterized by lower sputum cell counts, lower inflammatory cytokine levels, and lower CHIT1, YKL-40, and chitinolytic activity in sputum. Cluster 2 was formed exclusively by patients with COPD and demonstrated more pronounced airway inflammation reflected by a higher sputum inflammatory cell count, higher cytokine concentrations, and higher CHIT1, YKL-40, and chitinase activity than Cluster 1. Interestingly, when only COPD patients from Cluster 1 and Cluster 2 were compared, both groups were comparable in terms of age, gender distribution, BMI, smoking history and pulmonary function, despite relevant differences in the sputum inflammatory profile as presented above. This supports the concept of treatable traits in obstructive lung diseases, highlighting the need to identify individual disease features in patients with similar clinical manifestations to tailor precise treatment. [24,25]. The extremely low values of CHIT1 and YKL-40 and reduced chitinolytic activity in COPD patients from Cluster 1 may be, at least partially, responsible for the lack of differences in chitinase levels/activity between the COPD group and controls.
One limitation of our study was the small COPD sample size used for cluster analysis. The COPD group was also too homogenous to reveal any significant differences in lung function and severity of the disease. Moreover, more prospective longitudinal studies should be conducted to characterize COPD phenotypes and assess if these clusters influence the risk of exacerbation and the effectiveness of inhaled therapy. Although in the current study, the level of airway inflammation was attributed to only one factor, i.e., tobacco smoke exposure, we could have missed other crucial factors possibly contributing to the level of airway inflammation (e.g., working in a toxic environment, passive smoking).
Another significant limitation is that the elderly participants in our study had different comorbidities which could have affected our results. CHIT1 and YKL-40 are not specific to lung pathology and their levels can be elevated in multiple conditions, such as cardiovascular disease, cancer, inflammation, diabetes, and others [3]. Chitinase expression and activity may also be influenced by genetic polymorphism, although literature data on this matter are not uniform.

Conclusions
In conclusion, elevated IS YKL-40 levels in the COPD group and the relationship between some IS chitinases/CLPs (YKL-40) and some proinflammatory cytokines (IL-8 and MMP-9) suggest that chitinases may play an active role in chronic inflammation and tissue remodeling. Our study points to an association between the expression of chitinases and CLPs in the airways and the level of airway inflammation due to cigarette smoke exposure. However, this relationship seems to be ambiguous and needs further elucidation. Two main clusters were identified in the study. Cluster 2 included only COPD patients and was characterized by higher chitinase levels, chitinolytic activity, and more intense airway inflammation than Cluster 1. This difference should be considered when planning personalized treatments and precision medicine strategies for chronic airway diseases. Funding: This study was supported by the project "Development of a 'first-in-class' small molecule drug candidate for treatment of idiopathic pulmonary fibrosis through chitotriosidase inhibition" (POIR.01.01.01-00-0551/15) and co-financed by the European Union through the European Regional Development Fund within the Smart Growth Operational Programme.

Institutional Review Board Statement:
The study was conducted in accordance with the Declaration of Helsinki and approved by the Institutional Ethics Committee of the Medical University of Warsaw (KB/67/A/2018).

Informed Consent Statement:
Informed consent was obtained from all subjects involved in the study.

Data Availability Statement:
The reported data are available from the corresponding author upon reasonable request. reports a research grant from the National Science Centre, Poland, honoraria for lectures from Chiesi, AstraZeneca, Polpharma and MSD, fees for Advisory Board participation from MSD and AstraZeneca; all the above outside the submitted work. Boehringer Ingelheim, Chiesi, AstraZeneca and MSD have covered his fee and travel expenses for international conferences, outside the submitted work. K.G. reports a research grant from the National Science Centre, Poland, honoraria for lectures by Roche, Astra Zeneca, Polpharma and Boehringer Ingelheim, support for attending meetings and travel by Roche and Boehringer Ingelheim and participation in Advisory Boards by Boehringer Ingelheim, outside the submitted work. B.D. and A.Z. at the time of the study were employees and shareholders of Molecure SA (previously OncoArendi Therapeutics SA), which develops inhibitors of chitinases.