Innate immunity but not NLRP3 inflammasome activation correlates with severity of stable COPD

Background In models of COPD, environmental stressors induce innate immune responses, inflammasome activation and inflammation. However, the interaction between these responses and their role in driving pulmonary inflammation in stable COPD is unknown. Objectives To investigate the activation of innate immunity and inflammasome pathways in the bronchial mucosa and bronchoalveolar lavage (BAL) of patients with stable COPD of different severity and control healthy smokers and non-smokers. Methods Innate immune mediators (interleukin (IL)-6, IL-7, IL-10, IL-27, IL-37, thymic stromal lymphopoietin (TSLP), interferon γ and their receptors, STAT1 and pSTAT1) and inflammasome components (NLRP3, NALP7, caspase 1, IL-1β and its receptors, IL-18, IL-33, ST2) were measured in the bronchial mucosa using immunohistochemistry. IL-6, soluble IL-6R, sgp130, IL-7, IL-27, HMGB1, IL-33, IL-37 and soluble ST2 were measured in BAL using ELISA. Results In bronchial biopsies IL-27+ and pSTAT1+ cells are increased in patients with severe COPD compared with control healthy smokers. IL-7+ cells are increased in patients with COPD and control smokers compared with control non-smokers. In severe stable COPD IL-7R+, IL-27R+ and TSLPR+ cells are increased in comparison with both control groups. The NALP3 inflammasome is not activated in patients with stable COPD compared with control subjects. The inflammasome inhibitory molecules NALP7 and IL-37 are increased in patients with COPD compared with control smokers. IL-6 levels are increased in BAL from patients with stable COPD compared with control smokers with normal lung function whereas IL-1β and IL-18 were similar across all groups. Conclusions Increased expression of IL-27, IL-37 and NALP7 in the bronchial mucosa may be involved in progression of stable COPD.

3 function. Values of FEV 1 (% predicted) and FEV 1 /FVC (%) were significantly different in the groups with mild/moderate and severe/very severe COPD compared to both control groups (healthy smokers and healthy non-smokers). Severe/very severe COPD patients also differed significantly from mild/moderate COPD patients (for overall groups, ANOVA test: p<0.0001 for FEV 1 % predicted and FEV 1 /FVC% values). Forty-three percent (n=14) of the total COPD patients and 41% (n=5) of healthy smokers with normal lung function also had symptoms of chronic bronchitis. There was no significant difference when COPD patients and healthy smokers were compared for the presence of chronic bronchitis.
The clinical details of the subjects undergoing bronchoalveolar lavage (BAL) are summarized in Table 2 of the main manuscript. We analyzed the BAL fluid obtained from 26 COPD and 18 control smokers with normal lung function but due to the necessity to concentrate the BAL supernatants for many ELISA assays we were unable to perform all these assays in all subjects. The results provided for each ELISA are the data from 15 COPD and 14 control smokers with normal lung function.
Of the 44 subjects included for the bronchoalveolar lavage (BAL) analysis, 21 patients were recruited in the Veruno's Hospital and 23 subjects in Poland and Ferrara. All the BAL procedures were well standardized in accordance with standard guidelines. All the bronchoscopists in the three centers followed the same SOP for BAL collection and processing and the ELISAs on the BAL supernatants were run in a single center (Ferrara).

Lung function tests and volumes
Pulmonary function tests were performed as previously described (E2) according to published guidelines (E3). Pulmonary function tests included measurements of FEV 1 and FEV 1 /FVC under baseline conditions in all the subjects examined (6200 Autobox Pulmonary Function Laboratory; Sensormedics Corp., Yorba Linda, CA). In order to assess the reversibility of airflow obstruction and post bronchodilator functional values the FEV 1 and FEV 1 /FVC% measurements in the groups of subjects with FEV 1 /FVC%≤70% pre-bronchodilator was repeated 20 min after the inhalation of 0.4 mg of albuterol.

Fiberoptic bronchoscopy, collection and processing of bronchial biopsies
Subjects were at the bronchoscopy suite at 8.30 AM after having fasted from midnight and were pre-treated with atropine (0.6 mg IV) and midazolam (5-10 mg IV). Oxygen (3 l/min) was administered via nasal prongs throughout the procedure and oxygen saturation was monitored with a digital oximeter. Using local anaesthesia with lidocaine (4%) to the upper airways and larynx, a fiberoptic bronchoscope (Olympus BF10 Key-Med, Southend, UK) was passed through the nasal Thorax-2012-203062R2 online data repository 4 passages into the trachea. Further lidocaine (2%) was sprayed into the lower airways, and four bronchial biopsy specimens were taken from segmental and subsegmental airways of the right lower and upper lobes using size 19 cupped forceps. Bronchial biopsies for immunohistochemistry were extracted from the forceps and processed for light microscopy as previously described (E2).
Two samples were embedded in Tissue Tek II OCT (Miles Scientific, Naperville, IL), frozen within 15 min in isopentane pre-cooled in liquid nitrogen, and stored at -80°C. The best frozen sample was then oriented and 6µm thick cryostat sections were cut for immunohistochemical light microscopy analysis and processed as described below.

Immunohistochemistry
Two sections from each sample were stained applying immunohistochemical methods with a panel of antibodies specific for inflammatory cells, innate immune mediators and inflammasome components (Table E1). Briefly, after blocking non-specific binding sites with serum derived from the same animal species as the secondary antibody, primary antibody was applied at optimal dilutions in TRIS-buffered saline (0.15M saline containing 0.05 M TRIS-hydrochloric acid at pH 7.6) and incubated for 1hr at room temperature in a humid chamber. Antibody binding was demonstrated with secondary antibodies anti-mouse (Vector, BA 2000), anti-rabbit (Vector, BA 1000) or anti-goat (Vector, BA 5000) followed by ABC kit AP AK5000, Vectastain and fast-red substrate (red color) or ABC kit HRP Elite, PK6100, Vectastain and diaminobenzidine substrate (brown color). Slides were included in each staining run using human tonsil, nasal polyp or breast cancer, as a positive control. For the negative control slides, normal non-specific goat, mouse or rabbit immunoglobulins (Santa Cruz Biotechnology, Santa Cruz, CA, USA) were used at the same protein concentration as the primary antibody.

Scoring system for immunohistochemistry
Morphometric measurements were performed with a light microscope (Leitz Biomed, Leica Cambridge, UK) connected to a video recorder linked to a computerized image system (Quantimet 500 Image Processing and Analysis System, Software Qwin V0200B, Leica). Light-microscopic analysis was performed at a magnification of 630x.
The immunostaining for all the antigens studied was scored (range: 0 = absence of immunostaining to 3 = extensive intense immunostaining) in the intact (columnar and basal epithelial cells) bronchial epithelium, as previously described (E2). The final result was expressed as the average of all scored fields performed in each biopsy. A mean±SD of 0.70±0.26 millimeters of epithelium was analyzed in COPD patients and control subjects.
Thorax-2012-203062R2 online data repository 5 Immunostained cells in bronchial submucosa (lamina propria) were quantified 100 µm beneath the epithelial basement membrane in several non-overlapping high-power fields until the whole specimen was examined. The final result, expressed as the number of positive cells per square millimeter, was calculated as the average of all the cellular counts performed in each biopsy.

Double staining and confocal microscopy
Four patients with COPD (FEV 1 , 64±15%; FEV 1 /FVC, 61±8%) and 4 control smokers (FEV 1 , 104±14%; FEV 1 /FVC, 81±3%) were used for double staining immunofluorescence and confocal microscopy. Double staining was performed as previously reported (E2). For confocal microscopy sections were fixed with 4% paraformaldehyde, washed with phosphate buffered saline (PBS) and incubated (1 hour) with PBS containing 5% bovine serum albumin and 5% donkey serum. After blocking, sections were incubated 1 hour with the primary antibodies diluted as indicated in Table   E1 in PBS containing 5% bovine serum albumin. The following antibodies were used: rabbit antihuman IL-27 (LS B2565, R&D); and mouse anti human CD4, CD8 and CD68, M716, M7103 and M814, respectively, Dako). After washing with PBS, the preparations were incubated for a further 30 min with the appropriate secondary Alexa Fluor 488-or Alexa Fluor 647 conjugated antibodies diluted 1:200 in PBS. Negative controls included non-specific mouse and rabbit immunoglobulins revealed as for primary antibodies. Slides were mounted using a specific mounting medium (Vector, H-1400, Vectashield Hard Set). The slides for confocal microscopy were analysed using a three-channel Leica TCS SP5 laser scanning confocal microscope. The Leica LCS software package was used for acquisition, storage, and visualization. The quantitative estimation of colocalized proteins was performed calculating the "co-localization coefficients" (E4, E5).

Bronchoalveolar lavage
Bronchoalveolar lavage (BAL) was performed from the right middle lobe using four successive aliquots of 50ml of 0.9% NaCl. BAL cells were spun (500×g; 10min) and washed twice with Hanks' buffered salt solution (HBSS). Cytospins were prepared and stained with May-Grünwald stain for differential cell counts. Cell viability was assessed using the trypan blue exclusion method.
BAL supernatants were aliquoted and left at -80°C before its use for the ELISA assays summarised in Table E2. These assays have been performed according to the manufacturers' instructions.

Cell culture and treatments
We used normal human bronchial epithelial cells (NHBE), of non-smoking subjects obtained from Lonza (Cologne, Germany) grown in BEGM media (Lonza) with Singlequot supplement (Lonza) following the suppliers instructions. Cells were passaged using the ReagentPack ™ Subculture Reagents (Lonza) following the manufacturer's instructions. Passage number of cells used in this study ranged from 3 to 6. Prior to all the experiments, 70-80% cell monolayers were incubated in supplement-free medium for the 18hrs. The cells were treated with hydrogen peroxide (H 2 O 2 , 100µM), IL-1β (1 ng/ml), combined IL-1β and H 2 O 2 , cytomix (50ng/ml of TNFα, IL-1β and IFNγ) or combined cytomix and H 2 O 2 for 2hrs to analyse the messenger RNAs or 24 hours to measure protein expression. All experiments were performed at least three times.

Extraction and quantification of RNA and qRT-PCR from NHBE
Total RNA was isolated from cells using the RNeasy RNA extraction kit following manufacturer's instructions (Qiagen, UK). cDNA was made from quantified RNA by reverse transcription using the high capacity cDNA kit following manufacturer's instructions (Applied Biosystems, UK). The expression of genes of interest was measured using Syber green (Qiagen, UK) for qPCR in a Corbett Rotorgene 6 (Corbett, Cambridge, UK). We detected the expression of IL-27p28 and EBI3 (IL-27B) using the following primers (E6): IL-27p28 forward, agc tgc atc ctc tcc atg tt; reverse, gag cag ctc cct gat gtt tc; EBI3 (IL-27B) forward, tgt tct cca tgg ctc cct ac; reverse, gct ccc tga cgc ttg taa c. mRNA was normalized using a housekeeping gene 18S, using the following primers 18S forward ctt aga ggg aca agt ggc g; reverse acg ctg agc cag tca gtg ta, for each experimental condition.

Extraction and quantification of IL-27 protein from NHBE
IL-27 levels in the cell culture supernatant following stimulation with H 2 O 2 (100µM), IL-1β (1 ng/ml), combined IL-1β and H 2 O 2 , cytomix (50ng/ml of TNFα, IL-1β and IFNγ) or combined cytomix and H 2 O 2 for 24 hours was quantified by sandwich ELISA (R&D Systems Europe, Abingdon, UK) exactly according to the manufacturer`s instructions.

Statistical analysis
Group data were expressed as mean (standard deviation) for functional data or median (range) or interquartile range (IQR) for morphologic data. We tested for a normal distribution for functional data (i.e. FEV1%, FVC, age etc.) and for a non normal distribution for morphological parameters.
Normality tests were performed on all group data. The Grubb's outlier test was used to determine Thorax-2012-203062R2 online data repository 7 whether individual values were outside the rest of the group. Then we applied the analysis of variance (ANOVA) for functional data in comparing subgroups of patients and control subjects for functional data. The non parametric Kruskal Wallis test was applied for multiple comparisons, without application of Bonferroni correction, when morphologic data were analysed followed by the Mann-Whitney U test for comparison between groups. The statistical guide to GraphPad Prism recommends that the Bonferroni correction should not be used when comparing more than 5 variables due to the conservative nature of the test and the subsequent likeliness of missing real differences. We believe that this comparative analysis is of value and represents part of our informative findings. For this reason we applied specific non parametric statistical tests to our data of Table 3 without including the Bonferroni correction. To verify the degree of association between functional or morphological parameters, in all smokers with and without COPD or in smokers with COPD alone the correlation coefficients between functional and morphological and morphologicalmorphological data were calculated using the Spearman rank method. Probability values of p<0.05 were considered significant. Data analysis was performed using the Stat View SE Graphics program (Abacus Concepts Inc., Berkeley, CA-USA) and GraphPad Prism software (www.graphpad.com/scientific-software/prism/ ).

Measurement of inflammatory cells in the bronchial submucosa
The results are summarized in Table E3. The number of CD8 positive lymphocytes was significantly increased in severe/very severe (p=0.021) and mild/moderate (p=0.027) stable COPD compared to control non-smokers. The number of CD4 positive lymphocytes did not differ significantly between the four groups of subjects. Compared with control non-smokers, the number of CD68 positive macrophages was significantly higher in severe/very severe (p=0.033) and mild/moderate (p=0.036) stable COPD. The number of neutrophils was also significantly higher in severe/very severe stable COPD patients compared with control smokers (p=0.011) and nonsmokers (p=0.010). Stable COPD patients with chronic bronchitis had a similar number of neutrophils when compared with stable COPD patients without chronic bronchitis (data not shown).

Immunohistochemistry for innate immunity and inflammasome pathways in the bronchial submucosa
For all the proteins studied mononuclear cells (lymphocytes and macrophages) and endothelial cells were the most represented immunostained cells in the submucosa. The number of IFNγ+ cells was significantly higher in mild/moderate (p=0.010) and severe (p=0.008) COPD compared to control non-smokers, confirming previously reported data (E2). The number of IFNγRI+ cells was also increased in severe COPD compared to mild COPD (p=0.031), control smokers (p=0.0035) and control non-smokers (p=0.006). IL-18Rβ showed a slight increase in severe COPD compared to mild/moderate COPD (p=0.045) and control smokers (p=0.039) but did not differ in comparison with control non-smokers. The number of IL-7+ ( Figure E1) cells was higher in severe (p=0.008), mild/moderate COPD (p=0.010) and in control smokers (p=0.012) compared to control nonsmokers. In addition the number of IL-7Rα+ cells was significantly higher in severe compared to mild/moderate COPD (p=0.040), control smokers (p=0.009) and control non-smokers (p=0.002).
IL-10 was poorly expressed but number of IL-10+ cells was higher in severe (p=0.005), mild/moderate COPD (p=0.047) and in control smokers (p=0.054) in comparison with control nonsmokers.
The number of IL-27+ (Figure 1) as well as pSTAT1+ ( Figure E2) cells was significantly higher in severe COPD (p=0.032 and p=0.018, respectively) compared to control smokers but did not differ in comparison with the other groups. Interestingly, compared to control smokers and non-smokers, the number of IL-27R+ cells was higher in severe (p=0.010 and p=0.002, respectively) and mild/moderate COPD (p=0.054 and p=0.009, respectively). Similarly, the number of total STAT1+ cells was significantly higher in severe (p=0.0043 and p=0.015, respectively) and mild/moderate COPD (p=0.022 and p=0.029, respectively).

Double staining and confocal microscopy
Double staining for identification of CD4+, CD8+ lymphocytes and macrophages (CD68+) coexpressing IL-27 was performed in four representative healthy smokers with normal lung function and in four patients with COPD. We found no difference in the percentage (mean±SD) of CD4+IL-

IL-27 and inflammasome mRNAs expression in NHBE cells induced by oxidative and inflammatory stimuli "In vitro"
(1ng/ml) and H 2 O 2 + IL-1β at the same concentrations as for single treatments, cytomix alone (TNFα, IL-1β and IFNγ each at 50ng/ml) and combined cytomix + H 2 O 2 at the same concentrations as for single treatments and quantified the expression of IL-27p28 ( Figure E7A) and IL-27B ( Figure   E7B) mRNA by qRT-PCR. IL-27B mRNA was significantly increased ( Figure E7B and E7C) after combined treatment with cytomix plus H 2 O 2 (n=3, paired T test, p<0.05). The same stimulation did not significantly stimulate the expression of inflammasome-related IL-1β, IL-18 and caspase 1 encoding mRNAs ( Figures E8A, E8B and E8C). The 2hr time point was selected after an initial time-course study was performed (data not shown).

IL-27 protein expression in NHBE cells induced by oxidative and inflammatory stimuli "In vitro"
The levels of IL-27 protein was measured in the supernatant and whole cell extract of NHBE cells,  Figure E9).

Figure E1
Photomicrographs showing the bronchial mucosa from (a) control non-smoker, (b) control healthy smoker with normal lung function, (c) mild/moderate stable chronic obstructive pulmonary disease (COPD) and (d) severe stable COPD immunostained for identification of IL-7+ cells (arrows) in the bronchial submucosa. Results are representative of those from 11 non-smokers, 12 healthy smokers, 14 mild/moderate COPD and 18 with severe COPD. Bar=30 microns.

Figure E2
Photomicrographs showing the bronchial mucosa from (a) control non-smoker, (b) control healthy smoker with normal lung function, (c) mild/moderate stable chronic obstructive pulmonary disease (COPD) and (d) severe stable COPD immunostained for identification of pSTAT1+ cells (arrows) in the bronchial submucosa. Results are representative of those from 11 non-smokers, 12 healthy smokers, 14 mild/moderate COPD and 18 with severe COPD. Bar=30 microns.

Figure E3
Photomicrographs showing the bronchial mucosa from (a) control non-smoker, (b) control healthy smoker with normal lung function, (c) mild/moderate stable chronic obstructive pulmonary disease (COPD) and (d) severe stable COPD immunostained for identification of NALP7+ cells (arrows) in the bronchial epithelium and submucosa. Results are representative of those from 11 non-smokers, 12 healthy smokers, 14 mild/moderate COPD and 18 with severe COPD. Bar=30 microns.

Figure E4
Representative double-labelled confocal fluorescence images showing double staining for CD68+ macrophages and IL-27 in the bronchial mucosa from four healthy smokers with normal lung function (A-B) and four patients with stable chronic obstructive pulmonary disease (COPD) (C-D).
Images A and C were obtained from one healthy control smoker and one patient with severe COPD, respectively. Arrows in panel C indicate double stained cells. Images B and D show the coexpression levels of IL-27 (Alexa Fluor 488-green) and CD68 (Alexa Fluor 647-red) and represent the correlation cytofluorogram of the images in A and C, respectively.

Figure E5
BAL supernatant levels of (A) IL-6R, (B) IL-7, (C) IL-27 and (D) IL-37 in stable COPD patients (n=15) compared to the control smokers with normal lung function (n=14). Exact p values are shown above each graph.

Figure E7.
In vitro expression of IL-27p28 (A) and IL-27B (B and C)  Panel C shows a direct comparison between results from 3 subjects to indicate level of variability.

Figure E8
In vitro expression of expression of the inflammasone-related genes (IL1B, IL18 and CASP1) in normal human bronchial epithelial cells were measured in cells treated with H 2 O 2 (100µM), IL-1β (1ng/ml), cytomix (50ng/ml TNFα, IL-1β and IFNγ) and combined IL-1β + H 2 O 2 or cytomix + H 2 O 2 . Whilst treatment with cytomix alone did significantly increase IL1B gene expression (n=3, paired T test, p<0.05) no other gene or treatment showed significant increase in expression relative to the untreated controls.