Session Information
Session Type: ACR Abstract Session
Session Time: 2:30PM-4:00PM
Background/Purpose: Approximately 30% of psoriasis patients develop psoriatic arthritis (PsA), an inflammatory musculoskeletal disease, usually after psoriasis onset. A large proportion of individuals with PsA remain undiagnosed. An understanding of how the epigenome changes during the transition to PsA could yield predictive biomarkers and facilitate PsA diagnosis. We hypothesized that DNA methylation changes occur early in PsA pathogenesis, prior to overt clinical symptoms, and can be used as biomarkers for disease prediction. The specific aim was to characterize the epigenomic landscape of psoriasis patients who later developed PsA (converters) and compare it to psoriasis patients who did not develop PsA (non-converters).
Methods: We performed an epigenome-wide comparison of DNA methylation in baseline whole blood samples from psoriasis converters (n=58) and non-converters (n=59) from a longitudinal cohort. Post-conversion samples taken at the time of PsA diagnosis were available on 23 of the converters. Patients were matched for age, sex, psoriasis duration, and duration of follow-up. No patients were taking biologic agents. DNA was bisulfite converted was analyzed on Human MethylationEPIC BeadChips using the Bioconductor package ChAMP. Cell type heterogeneity was corrected using RefbaseEWAS. Differentially methylated probes (DMPs) and regions (DMRs) were identified using limma and DMRcate. The FEM package was used to infer differentially methylated gene modules (subnetworks) within a protein-protein interaction (PPI) network.
Results: Converter baseline samples were collected a median of 4.2 (interquartile range 1.9-6.3) years prior to the onset of PsA, while non-converters samples were collected a median of 4.3 (1.2-7.3) years prior to the most recent clinic visit. Between converters and non-converters, there were 129 DMPs and 15 DMRs containing at least 4 significant CpGs (FDR< 0.05). DMRs were identified in FBXO27 (fold change [FC]=0.06, FDR=2.2×10-24), a ubiquitin ligase involved in lysosomal degradation, RCAN1 (FC=0.05, FDR=4.2×10-12), a protein involved in bone homeostasis, and PMAIP1 (FC=0.04, FDR=6.0×10-12), which encodes the NOXA protein involved in mediating apoptosis of activated B cells. Differentially methylated promoters mapped to PPI subnetworks involved in canonical toll-like receptor/NFKB signaling (MYD88, TICAM1, TLR6, TOLLIP), alternative NF-KB signaling (MYD88, TICAM1, TNFRSF13C, TNFSF13B), and MAPK signaling (p38, MEF2D, DUSP10, PTPN7, STK39, MKNK2). Between non-converters and post-conversion samples, 25 DMPs and 3 DMRs in genes such as SERPINF1 (FC=-0.04, FDR=1.13×10-18) was found. Significant PPI subnetworks included notch signaling (APH1A, MAML3, NOTCH4, PSEN1) and ephrin signaling (EFNA1, EFNA2, EPHA7, EPHA2).
Conclusion: Changes in individual CpGs, DMRs, and inflammatory pathways were detected in baseline samples of psoriasis converters compared to non-converters and between non-converters and post-conversion samples. These preliminary data support our hypothesis that DNA methylation changes occur early in PsA pathogenesis and can potentially serve as prognostic biomarkers of future onset of arthritis in psoriasis patients.
To cite this abstract in AMA style:
Pollock R, Machhar R, Chandran V, Gladman D. Characterizing the Epigenomic Landscape of Psoriasis Patients Destined to Develop Psoriatic Arthritis [abstract]. Arthritis Rheumatol. 2019; 71 (suppl 10). https://acrabstracts.org/abstract/characterizing-the-epigenomic-landscape-of-psoriasis-patients-destined-to-develop-psoriatic-arthritis/. Accessed .« Back to 2019 ACR/ARP Annual Meeting
ACR Meeting Abstracts - https://acrabstracts.org/abstract/characterizing-the-epigenomic-landscape-of-psoriasis-patients-destined-to-develop-psoriatic-arthritis/