Background/Purpose: CD146 is a transmembrane glycoprotein belong to immunoglobulin superfamily, acts as adhesion molecule for the maintenance of cell monolayer. Human endothelial cells constitutively express CD146 which is involved in angiogenesis and inflammation. Recently, we established a sandwich ELISA for detecting soluble CD146 (sCD146) in human serum and reported presence of elevated level of sCD146 in patients with Behçet’s disease and systemic sclerosis. However, an association of sCD146 with PM/DM remains unknown. The aims of this study are to examine serum levels of sCD146 in patients with PM/DM and their association with clinical features and to clarify the mechanism of the difference in levels of sCD146.

Methods: Serum levels of sCD146 were quantified in 102 patients with PM/DM who visited our hospital from January 2001 to 2017 and compared with those of 22 healthy controls (HC). Recombinant CD146 protein was used to obtain a standard curve to measure quantity of sCD146. We also established another ELISA for detecting autoantibody against CD146 by using human recombinant CD146 protein.

Results: Serum levels of sCD146 were higher in the PM/DM subset than in HC but not statistically significant difference. Especially, the levels of sCD164 was significantly higher (P < 0.01) at mean 12.2 ng/mL in the PM subset, as compared to the DM subset or HC (Figure 1). Inverse correlations were observed between the levels of sCD146 and those of C-reactive protein in the PM/DM subset, and between the levels of sCD146 and those of creatine kinase in the PM subset. The levels of sCD146 were significantly higher (P < 0.05) in the patients with anti-ARS antibody than in patients with anti-TIF1-gamma antibody. There was no correlation between the levels of sCD146 and other clinical features such as complication of interstitial lung disease or malignancy. We hypothesized the existence of CD146 autoantibody in patients with DM and its involvement in the mechanism of the difference in the levels of sCD146. We measured serum CD146 autoantibody by established ELISA and found out the presence of serum CD146 autoantibody in PM/DM and significant high levels (P < 0.05) in the DM subset at 2 folds higher than mean level of the PM subset. These findings suggested the presence of CD146 autoantibodies could make disturbance to detect sCD146 in DM. Additionally, the levels of CD146 autoantibody showed a significant correlation with disease activity in patients with DM.

Conclusion: We identified higher levels of sCD146 in patients with PM than DM, and higher levels of CD146 autoantibody in DM than PM. CD146 could be one of the key factors involved in the pathophysiology of PM/DM.