Substantial evidence highlights the hippocampus as one of the brain regions that is critically implicated in the pathogenesis of depression as well as in the mediation of antidepressant actions (MacQueen & Frodl, 2011). Furthermore, the hippocampus is a major site of action for stress-related adrenal glucocorticoids, levels of which are elevated in patients with major depression. Following the seminal observation that glucocorticoids inhibit cell division in the subgranular zone of the hippocampal dentate gyrus (Gould et al., 1992), a growing body of data has supported the influential ‘neurogenesis hypothesis of depression’. This hypothesis posits that reduced hippocampal neurogenesis contributes to the pathophysiology of depression and that enhanced hippocampal neurogenesis underlies the effectiveness of antidepressant treatments (Samuels & Hen, 2011). The available evidence partially supports this hypothesis, particularly the role of neurogenesis for treatment; however, major questions remain unresolved.
In this issue of EJN, Van Bokhoven et al. (2011) report two important novel findings. They employed a stress model of depression, guided by evidence that stress experience is a major vulnerability factor for mood disorders (Sandi & Richter-Levin, 2009). First, they show that stress can have long-term effects on neurogenesis, at a time when depression-like behaviors are still present. This observation suggests that suppression of neurogenesis could be a key mechanism in the maintenance of depression. Second, they report that antidepressant treatment, administered long after the stressor, normalizes the effects of stress on neurogenesis. This finding has important therapeutic implications, specifically in terms of extending the effectiveness of antidepressant treatments on neurogenesis to much later time points than suggested in previous studies (Sahay & Hen, 2007).
An additional observation concerns the type of cells affected by antidepressant treatment. In their study, Van Bokhoven et al. (2011) evaluated hippocampal neurogenesis by stereologically quantifying the number of doublecortin (DCX)-immunopositive cells. DCX is a protein selectively expressed in young, immature neurons from 4 to 14 days after their birth. The authors differentiated between two types of DCX-positive cells in accordance to their specific morphologies: the younger class I and the more mature class II DCX-positive cells. Class II DCX-positive cells were identified, for the first time, as particularly susceptible to the long-term effects of stress. However, the two classes of DCX cells responded similarly to the enhancing effects of imipramine, in agreement with previous studies in which antidepressants increased several stages of the adult neurogenesis process (Sahay & Hen, 2007).
Importantly, at the time of the analyses, glucocorticoid levels did not differ between control and previously stressed animals. Therefore, this study provides evidence that a reduction in neurogenesis might underlie structural (such as reduced hippocampal volume) as well as functional hippocampal deficits observed during late stages of depression. These long-term effects do not merely reflect the immediate impact of stress or the associated acute elevation of glucocorticoids. The findings described by Van Bookoven et al. also suggest that antidepressants can reverse long-term effects of stress on neurogenesis. To confirm the significance of these findings, it will be important to show that these newborn neurons are functionally integrated into existing neural circuitry (Zhao et al., 2008). Finally, and while their relevance for the human condition should be taken with a note of caution, the results of Van Bokhoven et al. (2011) suggest a link between stress, antidepressants and neurogenesis during the maintenance phase of depression.