Proteins. By these means we have examined the expression of citrullinated antigens and the PAD enzymes, and their relationship to COPD and smoking. MethodsLung samplespulmonary lesions, mostly lung tumours (n = 41). Only uninvolved lung tissue, at maximal distance from the lung tumour, has been used for the current study. All patients attended Ghent University Hospital and samples were donated with informed consent and approval from the Medical Ethics Committee (Belgium registration number: B670201110668). Based on spirometry, according to the Global Initiative for Chronic Obstructive Lung Disease (GOLD) guidelines, and smoking history, patients were divided into four groups: smokers with (n = 13) and without (n = 10) COPD, ex-smokers with COPD (n = 8) and never smokers without COPD (n = 10). For details of the study subjects see Additional file 1: Table S1. The lung tissue was homogenised in RIPA buffer with protease inhibitors, centrifuged and the supernatants used for further experiments.Control tissue samples from other organsUninvolved surgically removed human tissue samples (spleen (n = 5), skeletal muscle (n = 2), liver (n = 4), ovary (n = 4), lymph node (n = 4), kidney (n = 4) and heart (n = 1)) were provided by Imperial College Healthcare NHS Trust Tissue Bank with appropriate informed consent and ethical approval (Imperial College Healthcare Tissue Bank HTA licence: 12275; and Research Ethics Committee Wales approval: 12/WA/0196). These control tissues were homogenised as described above.Immunoblotting and quantificationLung and control tissue lysates (20 g/sample) were resolved on 4 to 12 Bis-Tris NUPAGE gels (Invitrogen, Carlsbad, CA, USA), transferred to nitrocellulose membranes (Invitrogen) and western blotted using a standard protocol. The following primary antibodies were used: rabbit anti-fibrinogen (1:5,000; Calbiochem, San Diego, CA, USA), rabbit anti-PAD2 (1:500; Abcam, Cambridge, UK), goat anti-PAD4 (1:800; Abcam), rabbit anti-vimentin (1:1,000; Santa Cruz Biotechnology, Dallas, TX, USA), rabbit antiCEP1 (1:10,000; in house,[12]) and AMC antibody (Merck Millipore, Darmstadt, Germany, according to the manufacturer's instructions). All secondary antibodies were horseradish peroxidase (HRP)-conjugated and diluted 1:5,000. Recombinant PAD2, β-Amyloid (1-40) (TFA) PAD4, and in vitro citrullinated recombinant -enolase were used as positive controls where needed. Band intensities were quantified semi-quantitatively by three independent individuals using a four point (0 to 3) scoring method. The mean and median scores were analysed by statistical software GraphPad Prism (GraphPad Software, La Jolla, CA, USA).StatisticsLung tissue was carefully excised from lobectomy specimens removed from patients diagnosed with solitaryStatistical differences between intensity scores in each lung group were calculated by either one-way analysis ofLugli et al. Arthritis Research Therapy (2015) 17:Page 3 ofvariance (ANOVA) when PubMed ID:https://www.ncbi.nlm.nih.gov/pubmed/10435414 comparing all groups together or by two-tailed Mann-Whitney T-test using GraphPad Prism software. A value of P <0.05 was considered significant.Mass spectrometryRepresentative tissue lysates from 12 lung samples and one sample from each of the control tissues were selected according to their levels of citrullination on the AMC blots. A 5 mm band in the region of 52 kiloDaltons (kDa) was excised from a Coomassie-stained gel and used for mass spectrometry analysis. Sample preparation was performed as described previously [13]. Briefly, pr.