Serum proteome analysis identifies a potential biomarker for axial psoriatic arthritis

Abstract

Abstract Background To identify potential serum biomarkers for differentiating between axial psoriatic arthritis (axPsA) and peripheral psoriatic arthritis (pPsA). Methods Serum samples were collected from patients with PsA to create a biomarker discovery cohort and a verification cohort. Patients with PsA were classified into axial or peripheral subtypes based on imaging criteria. Untargeted proteomics technology was used in the discovery phase to screen for biomarkers, and candidate biomarkers were evaluated using enzyme-linked immunosorbent assay (ELISA) in the verification phase. Results We identified 45 significantly differentially expressed proteins (DEPs) between axPsA (n = 20) and pPsA (n = 20) with liquid chromatography-mass spectrometry. Among these DEPs, serum pigment epithelium-derived factor (PEDF) was identified as a candidate biomarker using the Boruta algorithm and lasso regression. Results of ELISA further confirmed that the level of serum PEDF expression was significantly higher in axPsA (n = 37) than in pPsA (n = 51) at the verification cohort (37.9 ± 10.1 vs. 30.5 ± 8.9 μg/mL, p < 0.001). Receiver operating characteristics analysis showed that PEDF had an area under the curve (AUC) of 0.72. Serum PEDF was positively correlated with body mass index and C-reactive protein. Additionally, there was a tendency towards a positive correlation between PEDF and the Bath Ankylosing Spondylitis Disease Activity Index. Conclusions This study provided a comprehensive characterization of the proteome in axPsA and pPsA and identified a candidate biomarker, PEDF, that may contribute to early diagnosis for axPsA.