Deoxyribonuclease 1 reduces pathogenic effects of cigarette smoke exposure in the lung

Our aim was to investigate if deoxyribonuclease (DNase) 1 is a potential therapeutic agent to reduce pathogenic effects of cigarette smoke exposure in the lung. Cigarette smoke causes protease imbalance with excess production of proteases, which is a key process in the pathogenesis of emphysema. The mechanisms responsible for this effect are not well-defined. Our studies demonstrate both in vitro and in vivo that cigarette smoke significantly increases the expression of neutrophil and macrophage extracellular traps with coexpression of the pathogenic proteases, neutrophil elastase and matrix metalloproteinases 9 and 12. This response to cigarette smoke was significantly reduced by the addition of DNase 1, which also significantly decreased macrophage numbers and lung proteolysis. DNase 1, a treatment currently in clinical use, can diminish the pathogenic effects of cigarette smoke.


Study design
were not blinded to allocation of samples/mice or analysis of the results.

Bronchoalveolar lavage (BAL) patients
Patients were referred to have a bronchoscopy at Monash Medical Centre (MMC)/Monash Health. Of the 34 patients, 32 of the subjects were referred to MMC/Monash Health after assessment by an affiliated respiratory physician, whilst two other subjects were referred from other medical institutions. Subjects had lung function testing performed using GOLD guidelines 1 with the best post-bronchodilator response used for results and lung diffusing capacity performed. Two of the subjects were inpatients and the rest were outpatients.
Twelve subjects of the total cohort had a significant smoking history (> than 10 pack-years), but only three subjects of the total group were active smokers at the time of bronchoscopy. Of the 12 subjects, six had had computed tomography scans which were reported by a consultant 3 radiologist to show emphysema. In the other 6 subjects, all had low lung diffusing capacity on lung function testing (without any other detectable lung disease that would explain a low lung diffusing capacity). A low lung diffusing capacity in smokers has been previously shown to correlate with the presence of emphysema 2,3 . Two of these subjects had lung cancer. In the twenty-two patients without a significant smoking history, no subject had evidence of emphysema. Two subjects had bronchiectasis, one had pulmonary fibrosis and two subjects had asthma.
The primary indications for performing the bronchoscopies on the subjects were for assessment of: chronic cough (twelve subjects), airway infection (eight subjects), haemoptysis (seven subjects), possible interstitial lung disease (three subjects), exclusion of malignancy (two subjects), investigation of possible foreign body (one subject) and chest pain (one subject). Informed written consent was obtained from subjects prior to the bronchoscopy. Details of the subjects are listed in Table E1.
Bronchoscopy was performed in the bronchoscopy suite at MMC/Monash Health. BAL was performed as previously described 4 with multiple warmed aliquots of 25 ml of normal saline injected into the right middle lobe and aspirated. Louis, MO), supplemented with 10% fetal calf serum, 0.1% L-glutamine (Sigma-Aldrich) and antibiotics (1% Penicillin (Sigma-Aldrich), 1% Streptomycin (Sigma-Aldrich) and 180 µg/ml Metronidazole (Claris, Chachamadi, Ahmedabad, India)). Cells were counted with a haemocytometer with trypan blue exclusion to obtain a viable cell count. Ammonium chloride (0.08%) was used to lyse red blood cells. More than 90% of the cells were viable with a mean macrophage count of 2.8 × 10 6 per subject. For confocal imaging, macrophages were seeded onto coverslips in 24-well flat-bottomed plates at a concentration of 1 to 3 × 10 5 cells per well and rested overnight.

Neutrophils
Peripheral blood neutrophils were obtained from healthy donors (the authors and laboratory staff) with informed consent prior to sample being taken. Neutrophils were isolated by density gradient using Polymorphprep (Progen Biotechnik, Heidelberg, Germany) 4 and seeded onto poly-l-lysine (Sigma-Aldrich) coated coverslips in 24-well plates for adhesion.

Cigarette smoke extract
An established method was used for the generation of cigarette smoke extract (CSE) for the in vitro experiments 5 . Cigarettes of the following composition were used: 16 mg of tar, 1.2 mg of nicotine, and 15 mg of CO 2 . Smoke was generated in 50 ml tidal volumes over 10 seconds by use of a timed draw-back mimicking normal smoking inhalation volume and cigarette burn rate into media. One cigarette was used per 40 ml of media. Based on our previous work and preliminary experiments, undiluted CSE media was used with a one-hour incubation period. Preliminary experiments using both trypan blue exclusion and propidium iodide excluion, showed no significant differences in cell viability between control and CSEexposed macrophages and neutrophils. 5

Smoking mouse model
To study the in vivo effect of cigarette smoke exposure an established smoking-mouse model was used 6 . This was a whole-body smoke exposure model. Male BALB-C mice (Animal Resource Centre, Perth, Australia), 6-8 weeks old were smoked three times a day for 4 days; with three cigarettes (each time) per 18 liter perspex enclosure (using same cigarettes that were listed above in the CSE section). Sham mice were only exposed to normal air. On the 4 th day, BAL samples were taken (3 hrs after exposure to cigarette smoke) or for analysis of whole lung samples, lung tissue was inflated and fixed in 10% neutral formalin. To assess the effect of DNase, 1 × 10 4 international units per kg of deoxyribonculease (DNase) 1 (Dornase alfa, Roche Genentech, San Francisco, CA) was administered twice a day intraperitoneally.
As a vehicle control for DNase group, mice were smoked and were administered with solution containing 0.15 mg/ml calcium chloride dehydrate and 8.77mg/ml sodium chloride at pH 6.3 (i.e. dornase alfa with DNase not present) at the same volume as DNase.
Preliminary experiments using both trypan blue exclusion and propidium iodide excluion, showed no significant differences in cell viability between control and DNase-exposed macrophages and neutrophils. The doses of DNase used (both for in vitro and in vivo work) were based on our previous studies 4, 10 .

Flow cytometry
We used a well-established flow cytometry method to measure the production of ROS from lung phagocytic cells 4,7 . The production of ROS causes cleavage of the dye

Confocal Microscopy
Previously described methods were used 4,[8][9][10] . BAL macrophages (1 to 3 × 10 5 per sample) or peripheral blood neutrophils from healthy donors were seeded (4 × 10 5 per sample) on poly-L-lysine coated coverslips and incubated for one hour 1), untreated 2), with CSE 3), with CSE and DNase (Roche Genentech) at 2 µg/ml. For these three experimental conditions, we also had 1), control-background (with secondary antibody only) and 2), isotype control with primary antibodies. 7 Staining of lung sections was performed on formalin-fixed, paraffin-embedded tissue specimens cut at 4 µm using Microtome and mounted on superfrost plus slides (Menzel, Braunschweig, Germany) as previously described by O'Sullivan KM et al 10 Table E2). Sections were washed in PBS and mounted with DAPI prolong gold for visualisation using NIKON C1 confocal microscope using the excitation 405, 488, 561 and 647 nm lasers and a 40x 1.0N.A oil objective.
In situ zymography was used to measure the area of the lung with proteolysis 11,12 . We used noninflated, nonfixed tissue for this method. were taken from each mouse for each stimulation (i.e. from each mouse; 10 FOV for control, 10 FOV for smoke, 10 FOV for smoke/DNase then results were averaged for each sample).
To determine the area of lung with evidence of proteolysis we used Image J to measure the area of lung that had staining above the background (this was done on minimum of 10 high power FOV for each sample) of the quenched sample and as another control, in lung tissue in which fluorescent gelatin had not been added. 10 The length of extracellular traps of macrophages and neutrophils was measured in ImageJ

Statistical analysis
Statistical analysis was performed using Prism 6 software (GraphPad Software, San Diego, CA). Comparison between control and stimulated groups was done using paired or unpairedtesting with parametric or non-parametric methods as appropriate. Paired and unpaired t testing was used for parametric data (2 tailed). For non-parametric data; the Wilcoxon matched-pairs rank test was used for paired data and for unpaired data the Mann-Whitney test was used. A P value of less than 0·05 was considered to indicate statistical significance.
Parametric results were represented graphically by mean and standard error of the mean, whilst non-parametric results were represented graphically by medians. Footnote: Non-smokers = < 10 pack-years of smoking, smokers = > 10 pack-years of smoking, FEV 1 = forced expiratory volume in 1 second, FVC = forced vital capacity, FER = forced expiratory capacity, DLCO = lung diffusing capacity for carbon monoxide. Results are expressed as mean ± standard deviation or median and interquartile range.    µm). Panel K shows neutrophils exposed to culture medium without CSE (as control).