CITRULLINATION IN HEALTHY AND INFLAMED LUNG TISSUE AS A PRIMING SITE FOR AUTOIMMUNITY IN RA

Background Anti-citrullinated peptide/protein antibodies (ACPAs) are the key pathogenic autoantibodies in rheumatoid arthritis (RA). Because smoking is a risk factor for RA and ACPA often appear years before the onset of disease, it has been proposed that the lung may be a site for priming the ACPA response. Previous studies using immunohistochemistry suggested that smoking upregulates the expression of PAD2 and PAD4 with the resultant increased expression of citrullinated proteins. However these studies are limited by the availability of healthy lung tissue. In this study we used lung tissue taken at a distance from the primary tumour in lobectomy specimens and antibody reactivity to PAD2 and PAD4 and to two important precursor antigens in RA, alpha-enolase and fibrinogen, was defined by immunoblotting to ensure specificity. Methods Lobectomy specimens from 40 subjects undergoing surgery for tumours or bronchiectasis (10 never smokers, 10 smokers without airflow limitation, 10 COPD ex-smokers and 10 COPD current smokers) were carefully dissected to remove a sample of uninvolved lung. The tissue samples were examined by immunoblotting with an anti-modified citrulline (AMC) antibody and scored for the level of citrullination with a semi-quantitative score from 0–3 by two blinded observers. The presence of PAD2, PAD4, alpha-enolase and fibrinogen was also determined by immunoblotting and scored. Recombinant proteins were used as positive controls and blots with secondary antibodies only were carried out to exclude non-specific cross-reactivity. Results Citrullinated proteins were detected in 37 out of the 40 lung tissue samples, including 9 out of 10 never smokers. The number of bands and intensity was slightly increased in the COPD smokers (mean score 1.7), followed by COPD ex-smokers (1.4), smokers (1.3) and never smokers without airflow limitation (1.1). PAD2 was detected in all samples, with the band intensity scores correlating roughly with those seen for citrullinated proteins i.e. highest amongst the COPD smokers (mean score 1.9) and lowest amongst never smokers (mean score 1.1). PAD4 and the RA antigens alpha-enolase and fibrinogen were observed in all lung tissue in comparable amounts regardless of disease and smoking status. There was also evidence of citrullination of alpha-enolase provided by co-migration of a ~50KD band recognised by AMC and an anti-enolase antibody, and the demonstration that alpha-enolase in lung tissue ran at several isoelectric spots by 2D electrophoresis. Conclusions We have shown that there is widespread citrullination of proteins in lung tissue from never smokers, and that there is a modest increase with smoking and COPD. This pattern of expression corresponds to that of PAD2. There is also expression of at least two major RA autoantigens and of PAD4. This supports the hypothesis that the lung is a site for priming the ACPA response, which is enhanced by smoking and and COPD.