Work

Integrated metabolomics and DNA methylation signatures and their association with flare remission in SLE

Year: 2026

Type: article

Abstract: Objective Identify whether changes in metabolite abundance are associated with SLE flare remission and changes in DNA methylation (DNAm). Methods 40 multiethnic patients with SLE were recruited during a rheumatologist-confirmed flare and returned to the clinic 3 months later. Per visit, we obtained blood samples and generated untargeted metabolomics (ultra-high performance liquid chromatography coupled to high-resolution mass spectrometry) and DNAm (Illumina EPIC array) profiles. We identified chemical features whose change between visits was associated with remission at the follow-up visit using the limma package and a time × remission interaction term, adjusting for follow-up time and medications. We tested for correlations between changes in significant chemical features and DNAm at 291 sites previously associated with remission. Results 16 patients with SLE were in remission at the follow-up visit. Remitters and non-remitters had similar demographics and disease activity at baseline. We identified 62 chemical features whose changes over time were associated with remission. 19 were assigned a putative consensus chemical formula. Two were confirmed with reference standards—adenosine (interaction coefficient (β): 2.29, 95% CI 1.44 to 3.14) and bilirubin (β: 2.02, 95% CI 1.09 to 2.94). We identified 23 significantly correlated chemical feature-DNAm pairs (p<0.05/62×291). This included a strong correlation between adenosine and four DNAm sites, including three within immune-related genes , EBF1 (r=−0.77, p=1.35×10 −7 ), IL12B (r=0.69, p=1.87×10 −6 ) and GRAMD1C (r=−0.71, p=1.00×10 −6 ). Conclusion Changes in chemical features, including confirmed metabolites bilirubin and adenosine, were associated with flare remission and changes in DNAm at SLE-relevant loci. Metabolites and their correlations with DNAm might serve as indicators of immune function and improve our knowledge of determinants of response to treatment. Future studies are needed to confirm the identities of metabolites and robustness of findings.

Source: Lupus Science & Medicine

DOI: 10.1136/lupus-2026-002017