Reina Sofia Hospital / IMIBIC / University of Cordoba Cordoba, Spain
Elena Carnero-Montoro1, Ivan Arias-de la Rosa2, Clementina Lopez-Medina3, lourdes Ladehesa-Pineda4, Rafaela Ortega-Castro2, Daniel Toro-Domínguez1, Manuel Martinez-Bueno1, Raúl López-Domínguez5, Olivia Castellini-Pérez1, Guillermo Barturen1, Nuria Barbarroja2, Eduardo Collantes2 and Marta Alarcon-Riquelme1, 1Center for Genomics and Oncological Research (GENYO), Granada, Spain, 2IMIBIC/University of Cordoba/Reina Sofia Hospital, Cordoba, Spain, 3Reina Sofia University Hospital, Rheumatology Department, Jaén, Spain, 4Reina Sofia University Hospital/Rheumatology Department/Maimonides Institute for Biomedical Research (IMIBIC), Cordoba, Spain, 5University of Granada, Granada, Spain
Background/Purpose: An important role of DNA methylation changes in Spondylitis Ankylosing has been recently demonstrated by epigenome-wide studies. However, there is a lack of genome-wide studies that examine the role of epigenetics in the physiopathology of a wider spectrum of Spondyloarthritis and in Psoriatic Arthritis and that further explore its clinical utility.
Methods: Here, we measured blood genome-wide DNA methylation (DNAm) at more than 800,000 CpG sites using the Illumina MethylationEPIC array in 207 patients diagnosed with Spondyloarthritis (SpA) and 56 patients diagnoses with psoriatic arthritis (PsA) and coupled this information with genotyping and clinical data. We performed epigenome-wide association studies accordingly to disease status, clinical forms, comorbidities and disease progression.
Results: At a False Discovery Rate of 5% we identified 24 differentially methylated CpG sites (DMS) between SpA and PsA. The largest DNAm differences were observed at the chromosome 6 class I HLA region (HLA-B, MICA and HCP5 genes) and could be solely explained by HLA-b27 positivity. The remaining DMS were not dependent on HLA-b27 positivity, and interestingly were enriched in genes expressed in the large intestine (as for example CASZ1, TP73 or SAMD11 genes). Principal component analyses based on non-HLA DMS show that DNAm has the ability to differentiate SpA from PsA. We observed higher number DMS between Axial SpA and PsA and recognized important genes such as the cytokine signaling molecule JAK2, implicated in osteoclast differentiation and inflammation upon IL17-A stimulation. Besides, we identified DNAm differences between peripheral SpA patients and PsA at the TRIM25 gene, which is implicated in antiviral response. Lastly, we identified up to 16 DMS associated with SpA disease activity score (ASDAS), some of them lying in genes with important function such as CNKSR3 (a negative regulator of ERK1/ERK2 cascade and implicated in eye and vascular disease), SULF2 (involved in cell signalling and associated with autoimmune diseases) or COL4A2 (type IV collagen associated with coronary artery disease and autoimmune diseases). Logistic regression and ROC curves show that the epigenetic profile based on these ASDAS-associated DMS has a high discriminatory utility to differentiate patients with active and non-active SpA patients (AUC = 0.94).
Conclusion: This study highlights the importance of studying epigenomic signatures of Spondyloarthritis in relation to its many different clinical outcomes and underscores the utility of epigenetic biomarkers in rheumatology future clinical practice.
Disclosures: E. Carnero-Montoro, None; I. Arias-de la Rosa, None; C. Lopez-Medina, None; l. Ladehesa-Pineda, None; R. Ortega-Castro, None; D. Toro-Domínguez, None; M. Martinez-Bueno, None; R. López-Domínguez, None; O. Castellini-Pérez, None; G. Barturen, None; N. Barbarroja, None; E. Collantes, None; M. Alarcon-Riquelme, None.
