Glutamatergic NMDA receptor antagonists, notably ketamine, have emerged as promising candidates for next-generation fast-acting antidepressants (33
). The results of the present study demonstrate that a single dose of ketamine completely reverses the behavioral deficits caused by long-term exposure to CUS. Similar effects were observed with a selective NR2B antagonist, Ro 25-6981, which also rapidly reversed the anhedonic and anxiogenic behaviors resulting from CUS. This is consistent with a recent report that administration of a selective NR2B receptor antagonist, CP-101,606, produces rapid antidepressant effects in depressed patients (34
). Reversal of the CUS effects by ketamine and Ro 25-6981 are sustained for up to 7 days, similar to the time course for the therapeutic actions of ketamine (3
). Although previous studies have reported antidepressant actions of ketamine in models that are responsive to acute or sub-chronic administration of typical antidepressant agents (7
), the current study provides a rigorous test of the rapid actions of NMDA antagonists in a CUS paradigm that requires chronic (21d) antidepressant treatment (11
). These findings differ from a recent negative CUS study (38
), although the latter measured consumption instead of preference, which is generally considered a better measure of anhedonic behavior.
We recently reported that ketamine increases synaptic protein levels and produces antidepressant behavioral actions in non-stressed naïve animals, and that these effects are blocked by rapamycin, a selective inhibitor of mTOR (25
). Similarly, we show here that ketamine-reversal of the behavioral deficits caused by CUS is blocked by rapamycin, indicating that the effects of ketamine are mTOR-dependent alterations in spine/synapse. In support of this hypothesis, we found that CUS exposure decreased levels of synapsin I, GluR1 and PSD95, effects that were rapidly reversed by ketamine in a rapamycin-dependent manner. The requirement for mTOR signaling is consistent with our recent findings that ketamine directly activates several components of the mTOR pathway (25
). Another study has reported that rapamycin paradoxically produces an antidepressant response (39
), although rapamycin was administered systemically, which could influence mTOR signaling and other kinases in peripheral tissues that could indirectly impact behavior. In addition, the latter paper only used the forced swim and tail suspension, which are typically used as rapid drug screens and have limited validity as measures of depressive behavior.
We examined the possibility that mTOR inhibition underlies the decrease in synaptic proteins caused by CUS, but found no significant differences in levels of phosphorylated/activated mTOR, 4EB-P1, and p70S6 kinase (Figure S3 in the Supplement
). It is possible that there is a complex time course for stress regulation of mTOR signaling (e.g., S6 kinase is up-regulated by exposure to acute restraint stress) (40
). Alternatively, the decrease in synaptic proteins could be mediated by other mechanisms, including increased degradation or other via other mTOR signaling pathways (e.g., neurotrophic factor stimulated kinases are decreased by exposure to repeated stress or corticosterone) (41
). Further studies of these and other pathways are currently being conducted to characterize the mechanisms underlying the neuronal atrophy caused by chronic stress exposure.
Decreased levels of synaptic proteins resulting from CUS is consistent with previous studies demonstrating that chronic stress causes dendritic atrophy of neurons in limbic brain regions, including the PFC (14
). This possibility is confirmed in the present study, which shows that CUS decreases the density of apical spines in layer V pyramidal neurons of the PFC. Notably, CUS caused a marked reduction in the population of large-diameter mushroom spines, which correlates with a significant reduction in the amplitude as well as frequency of both 5-HT- and hypocretin-induced EPSCs. These transmitter induced synaptic currents involve both cortical-cortical (5-HT) and thalamocortical (hypocretin) synaptic inputs (13
), indicating that diverse pathways are affected. Administration of ketamine rapidly and completely reversed these deficits in spine density and function, indicating a causal relationship between the morphological and physiological responses. However, it is possible that the 5-HT and hypocretin-induced EPSPs reflect only a functional change in synaptic efficacy and further studies of mini EPSPs, measured under depolarization blockade, may be useful to directly test this causal interaction.
Dendritic atrophy of PFC neurons in response to chronic CUS is consistent with previous reports with other paradigms such as chronic restraint stress (15
), and provides a cellular mechanism that could explain the decreased PFC volume reported in depressed patients (20
). Neuronal atrophy and loss of spines/synapses in the PFC could underlie some of the behavioral deficits observed in depressed patients (46
), including reduced cognitive functions and loss of inhibitory control of emotions, perhaps mediated by amygdala circuits (47
The ability of ketamine to increase synaptogenesis represents a fundamental, conceptual frame shift in our understanding of synaptic plasticity and the treatment of depression. First, the results demonstrate a basic characteristic of brain plasticity, the ability to rapidly stimulate the formation and function of new spines/synapses. Second, the findings show that the dendritic atrophy resulting from long-term stress exposure is reversible. This also implies that the structural deficits caused by, and implicated in the pathophysiology of mood disorders, including treatment resistant depression, are reversible and that the rapid actions of ketamine are mediated by increased synaptogenesis. Brain imaging studies that directly or indirectly measure synaptogenesis and/or glutamate transmission will be required to directly test this hypothesis in depressed patients.
In summary, the results provide direct evidence that NMDA receptor antagonists rapidly reverse the behavioral, morphological, and physiological deficits resulting from chronic stress via a rapamycin-sensitive, mTOR pathway. These findings, together with earlier studies, provide promising targets for development of next-generation fast-acting antidepressant medications that are safer than ketamine. This possibility is supported by the finding that a NR2B selective antagonist also produces rapid antidepressant actions in the CUS model (present study), as well as other paradigms (7
), and also increases mTOR-dependent synaptogenesis (25
). Studies are currently underway to test additional targets in the glutamate-mTOR signaling pathway that could also be developed as rapid acting antidepressants.