Identification of Pharmacogenomic and Functional Biomarkers for Antidepressant Treatment Outcome in Patients-Derived Lymphoblastoid Cell Lines

Abstract

Major depression is the most prevalent mental disorder worldwide. Pharmacological management of the disease is considered a cornerstone in treatment guidelines adopted by many industrialized countries. However, response and remission rates remain moderate to low, leaving a major proportion of patients without sufficient symptomatic improvement. Moreover, almost 15% of patients will develop treatment resistant depression resulting in a substantial burden to the health and social systems. An early clinical evaluation of individual therapy outcome is hampered by a delay in clinical improvement of several weeks. Therefore, the aim of this work was to identify rapidly determinable predictive biomarkers for the clinical outcome of antidepressant therapy and to provide insights in the molecular mechanisms underlying antidepressant effects. For this purpose, the observational Munich Antidepressant Response Signatures (MARS) study was chosen as a study cohort, in which depression patients had been previously treated in real-world practice settings and documented for their individual weekly clinical improvement in addition to their clinical response and remission status. From MARS patients, lymphoblastoid cell lines (LCLs) generated from B-lymphocytes were employed in this work as an in vitro model for transcriptional profiling using directed candidate gene and undirected whole-transcriptome analyses. Differential gene expression was tested under control and antidepressant short-term incubations using citalopram as a prototype antidepressant drug. After step-wise analyses of gene hits using predetermined methodologies and a validation in a larger MARS cohort, GAD1, TBC1D9 and NFIB could be determined as tentative predictive genes for clinical response, clinical remission and for improvement in depression scores, respectively. Pathway analysis of citalopram-altered gene expression revealed response-status-dependent transcriptional reactions. Whereas in clinical responders neural function pathways were primarily up- or downregulated after incubation with citalopram, deregulated pathways in non-responders LCLs were less specific to the nervous system and mainly involved cell adhesion and immune response. For determination of predictive biomarkers for treatment-resistant depression, LCLs from the interventional Sequenced Treatment Alternatives to Relieve Depression (STAR*D) study were employed. The particular sequential 4-level-treatment design of the study enabled us to select patients representing the clinical edge groups, i.e. first-line responders and treatment-resistant patients, for pursuing transcriptional biomarker analyses. Hits detected from MARS cells showed only a tendency to association of NFIB in STAR*D cells, whereas associations with further genes were not significant. <br /> Furthermore, the pharmacologically functional target of antidepressants, the serotonin transporter (SERT) was systematically studied on the genetic variants level and the transcriptional level, as well as on the total and surface protein expression levels in LCLs from the MARS and STAR*D studies for associations with response and treatment resistance status of the donor patients. Genetic investigations of the SERT-coding gene (SLC6A4) polymorphisms 5-HTTLPR and rs25531 did not reveal associations with the clinical outcome of the donor patients. Transcription and total protein analysis did not show static or reactional (i.e. changes upon antidepressant incubation) differences in SERT expression between LCLs from patients with different clinical outcomes. However, surface SERT demonstrated a stably higher expression in LCLs from treatment-resistant patients than in those from first-line responding patients from the STAR*D cohort. Ubiquitination of SERT did not reveal definite patterns to be associated with clinical outcome. <br /> This work provides deeper insights in personalization of treatment in depressed patients based on response patterns. Our whole-transcriptome results propose the existence of distinct pathway regulation mechanisms in responders vs. non-responders and suggest that tentative predictors for clinical response, full remission, and improvement in depression scale, do not overlap as predictors of different therapy outcome phenotypes. Whereas no transcriptional biomarker for treatment resistance could be identified, our SERT analyses suggest an association of this clinical phenotype with higher cell surface expression of SERT.