The current therapy success of depressive disorders remains in need of improvement due to low response rates and a delay in symptomatic improvement. Reliable functional biomarkers would be necessary to predict the individual treatment outcome. On the basis of the neurotrophic hypothesis of antidepressant's action, effects of antidepressant drugs on proliferation may serve as tentative individual markers for treatment efficacy. We studied individual differences in antidepressant drug effects on cell proliferation and gene expression in lymphoblastoid cell lines (LCLs) derived from patients treated for depression with documented clinical treatment outcome. Cell proliferation was characterized by EdU (5-ethynyl-2'-deoxyuridine) incorporation assays following a 3-week incubation with therapeutic concentrations of fluoxetine. Genome-wide expression profiling was conducted by microarrays, and candidate genes such as betacellulin—a gene involved in neuronal stem cell regeneration—were validated by quantitative real-time PCR. Ex vivo assessment of proliferation revealed large differences in fluoxetine-induced proliferation inhibition between donor LCLs, but no association with clinical response was observed. Genome-wide expression analyses followed by pathway and gene ontology analyses identified genes with different expression before vs after 21-day incubation with fluoxetine. Significant correlations between proliferation and gene expression of WNT2B, FZD7, TCF7L2, SULT4A1 and ABCB1 (all involved in neurogenesis or brain protection) were also found. Basal gene expression of SULT4A1 (P=0.029), and gene expression fold changes of WNT2B by ex vivo fluoxetine (P=0.025) correlated with clinical response and clinical remission, respectively. Thus, we identified potential gene expression biomarkers eventually being useful as baseline predictors or as longitudinal targets in antidepressant therapy.

The neurotrophic hypothesis of depression suggests an association between effects on neuroplasticity and clinical response to antidepressant drug therapy. We studied individual variability in antidepressant drug effects on cell proliferation in lymphoblastoid cell lines (LCLs) from n=25 therapy-resistant patients versus n=25 first-line therapy responders from the Sequenced Treatment Alternatives to Relieve Depression (STAR*D) study. Furthermore, the variability in gene expression of genes associated with cell proliferation was analyzed for tentative candidate genes for prediction of individual LCL donor's treatment response. Cell proliferation was quantified by EdU (5-ethynyl-2′-deoxyuridine) assays after 21-day incubation of LCLs with fluoxetine (0.5 ng μl−1) and citalopram (0.3 ng μl−1) as developed and described earlier. Gene expression of a panel of candidate genes derived from genome-wide expression analyses of antidepressant effects on cell proliferation of LCLs from the Munich Antidepressant Response Signature (MARS) study was analyzed by real-time PCR. Significant differences in in vitro cell proliferation effects were detected between the group of LCLs from first-line therapy responders and LCLs from treatment-resistant patients. Gene expression analysis of the candidate gene panel revealed and confirmed influence of the candidate genes ABCB1, FZD7 and WNT2B on antidepressant drug resistance. The potential of these genes as tentative biomarkers for antidepressant drug resistance was confirmed. In vitro cell proliferation testing may serve as functional biomarker for individual neuroplasticity effects of antidepressants.

Functional imaging studies of affective disorders have demonstrated abnormal activity in the amygdala in response to emotionally salient stimuli. Since in other studies this response has been shown to habituate during the scanning session, it is not clear if it may be of use in monitoring disease progression or remission, or in monitoring the effects of therapy, as habituation may confound normalisation of response. We investigated here amygdala activation in healthy participants exposed to displays of emotional facial expressions in a sample of N = 31 individuals assessed twice in an interval of three weeks. At this interval no habituation could be detected, suggesting the validity of this imaging assay in repeated assessments of amygdalar reactivity. However, the fusiform gyrus and the inferior frontal lobes showed decreases in activations that may be related to the role of these areas in encoding visual and emotional aspects of the stimuli.

HighlightsOur study highlights the use of neuroimaging methods to detect and analyze therapy associated side effects, and the main finding of our study is that interferon, known to induce depression as adverse drug effect, leads to selective blunting of appetitive rather than affective stimuli.Our study shows the utility of sophisticated imaging methods to assess and clarify the nature of subtle drug effects on emotional and cognitive behavior.We think that the use of neuroimaging methods for the prediction and detection of off-target drug effects on mental health is an important topic in personalized medicine.