F80. A DNA METHYLATION STUDY OF COMMON AND SPECIFIC TREATMENT EFFECTS OF THREE ANTIPSYCHOTICS IN THE TREATMENT OF PSYCHOSIS

• Published in: European neuropsychopharmacology 2023-10, Vol.75, p.S262-S263
• Main Authors: Villar, Jonelle, Stavrum, Anne-Kristin, Spindola, Leticia M., Zayats, Tetyana, Melle, Ingrid, Andreassen, Ole A., Hellard, Stephanie Le
• Format: Article
• Language: English
• ISSN: 0924-977X; 1873-7862
• DOI: 10.1016/j.euroneuro.2023.08.462

Antipsychotic (AP) medication is prescribed to alleviate symptoms of psychosis in schizophrenia (SCZ), bipolar disorder (BPD), and major depressive disorder (MDD). Epigenetic studies have previously revealed altered DNA methylation (DNAm) patterns associated with APs. Many of these studies, however, fail to distinguish the treatment effects of polypharmacy from the specific effects of monotherapy. Furthermore, the specific DNAm patterns associated with these disorders are not differentiated from AP treatment effects. Developing models that differentiate the DNAm patterns associated with psychosis from pharmacological treatment effects are lacking. Recent data-driven studies of receptor profile categories may inform study design. Due to similar AP receptor profiles, we hypothesize that quetiapine (Que) and olanzapine (Ola) will share greater commonality than either AP with risperidone (Ris). Therefore, we aim to identify DNAm signatures specific to each of the three medications or common among them. Patients were recruited through the TOP Study (Thematically Organised Psychosis, Oslo, Norway). We selected European patients diagnosed with SCZ, BPD, and MDD, with either confirmed adherence to AP monotherapy (serum values > 0) or medication-free (MF). DNA methylation data derived from peripheral blood samples were assessed genome-wide using the Illumina 850K EPIC array. Differentially methylated positions (DMPs) were identified using β-values with limma. Medicated patients (n=131) were compared to MF (n=128) for the common effects. For the specific effects, DNAm values from monotherapy samples were compared to MF: Que (n=27) vs. MF, Ola (n=89) vs. MF, and Ris (n=25) vs. MF. Differentially methylated regions (DMRs) were identified using comb-p with a distance cutoff of 750 base pairs. The statistical model was corrected for age, sex, smoking score, estimated cell-type composition, diagnosis, and technical batch effects. Shared genes and DMPs with p-values < 10E-04 were visualized in Venn diagrams. All analyses were performed using R Statistical Software (v4.2.2). A descriptive analysis of the identified shared genes and DMPs suggested a slight increase of shared genes and DMPs between Ola and Que as hypothesized, compared to Ris. No significant DMPs, however, were identified for either the common or specific effects. Ten DMRs with z-sidak p-value < 0.05 (seed 0.001) were identified for the specific models, and 28 DMRs in the common effects model (seed 0.05). We