P.2.010 - In search for predictive biomarkers: dissecting the molecular pathways in brain and blood underlying poor and good antidepressant treatment response

Abstract

Major depression poses a serious social and economic threat to modern societies, as it accounts for more lost productivity compared with any other disorder. There are currently two major problems calling for innovative research approaches: 1. The absence of biomarkers predicting antidepressant response and 2. The lack of conceptually novel antidepressant compounds. Identification of biomarkers could allow patient stratification and enable the selection of pathophysiologically distinct patient subgroups to allow optimized treatment choices based on biology. In search for conceptually novel antidepressants, the hippocampal dentate gyrus is a region of particular interest, as there is a large body of evidence suggesting that stimulation of neurogenesis by antidepressant drugs could be one of their major common targets [1,2]. Aiming to identify the molecular pathways shaping response to antidepressant treatment and to identify predictors for antidepressant response, we used our animal experimental approach to model heterogeneity in response to antidepressant treatment (Carillo-Roa et al., PLoS BIOL, in revision). Male, adult DBA/2J mice show a strong innate anxiety and depression-like phenotype. Mice were treated for 14 days with customized pills containing either paroxetine or vehicle, providing a stress-free and oral treatment paradigm. Efficacy of antidepressant treatment was measured using the Forced Swim Test and results analyzed using a Student’s t-test. For the identification of peripheral biosignatures predicting response, we collected blood samples at baseline and following 14 days of treatment (longitudinal design), whereas for the identification of novel antidepressant targets, we specifically dissected the hippocampal dentate gyrus using a microdissecion protocol for native hippocampal tissue. We extracted stranded total RNA from dentate gyrus samples (paroxetine vs vehicle and good vs poor responders) and whole blood samples (baseline and 14 days in good vs poor responders) and sequenced. Analysis of differentially expressed genes (DESeq analysis) was conducted applying Benjamini-Hochberg correction with a padj 2.68, and fold changes >1.5. Pathway analysis was conducted using STRING software. qPCR of selected genes was conducted and analyzed using delta-delta-ct method. Paroxetine treatment led to antidepressant-like (p With this study, we provide additional strong evidence that neurogenesis drives antidepressant treatment effects and determines response status in rodents. We also expect to identify transcriptome signatures predictive of later antidepressant response in whole blood. Finally, we aim at validating our murine predictive biomarkers in human studies on antidepressant response.