Stress, Corticosteroids and Excitatory Neurotransmission

Traumatic experiences and genetic background are key etiologic variables in numerous psychiatric diseases, including major depression. Recent animal and human data suggest that neuroplasticity, specifically, an adaptive set of remodeling changes underlying efficient learning and memory, is altered in depression, with traumatic/stressful experiences playing a major role therein. In the hippocampus, stress negatively impacts on synaptic transmission efficacy (synaptic plasticity), cell morphology (dendrite remodeling) and cell integrity/number (neurotoxicity and neurogenesis).

These processes are thought to play a key role in depression (e.g., depression-related amnesia) and the most widely used antidepressants have been shown to exert intrinsic positive effects on these processes and/or to reverse stress-elicited dysregulations in neuroplasticity. Stress-elicited release of corticosteroid hormones (e.g., corticosterone in rodents and cortisol in humans), through its actions on AMPA receptor (AMPAR)-mediated excitatory transmission, contributes to a major extent to the aforementioned negative impact of stress on neuroplasticity. However, how acute and chronic stress effect synaptic AMPAR trafficking, and especially synaptic AMPAR lateral mobility (a key event during plasticity processes), is unknown.

Depression & neuroplasticity

The criteria that help to diagnose major depression show that depression is related to a diminished ability to think and concentrate, illustrating the tight links between depression and cognitive processes [1]. Hence, depression is associated with a severe impairment in learning and retrieval of explicit memory, and the observations that the size of the hippocampus, a key substrate for learning and memory retrieval, is reduced in recurrent depressed patients point to hippocampal neuroplasticity as one important target of depression [2,3]. However, whether this depression-related atrophy of the hippocampus is accounted for by decreases in dendritic branching and/or reduced adult neurogenesis (subgranular zone of the dentate gyrus) and/or neuron loss is not known with certainty [4]. Animal models, based essentially on the inescapable exposure to chronic/repeated stressors, suggest that the hippocampal atrophy associated with severe recurrent depression is accounted for by a combination of these three phenomena [4]. This illustrates how the use of stress in laboratory animals, despite its obvious limitations (see below), has brought major advances in our quest for the biological factors of depression.

Stress & neuroplasticity

Data gathered in depressed humans indicate that the onset of major depression depends on the interaction between the genetic background and exposure to stressful life events (i.e., number and intensity of these events), that is, depression is more likely to arise in individuals that are unable, owing to susceptibility genes, to cope and thus adapt to stressful stimuli (whether these are applied in utero or thereafter) [5]. In keeping with this important, albeit not unique, role of stress in the etiology of depression, numerous studies have investigated the impact of stress on neuroplasticity.

Acute/chronic stress has been shown to affect the whole neuroplasticity process. Thus, stress-elicited changes in emotional behaviors and cognitive abilities are associated with cellular modifications in the hippocampus, which range (according to stress intensity and duration) from functional alterations in synaptic efficacy (synaptic plasticity, through which new memories are likely to be encoded or not) to morphological alterations (dendritic remodeling) and cell viability/number (neurotoxicity and neurogenesis) [6,7]. As far as synaptic efficacy is concerned, in the long term, stress alters the ability of the hippocampus (CA1 and dentate gyrus) to undergo synaptic plasticity, as illustrated by stress-induced impairments in high frequency stimulation-elicited long-term potentiation (LTP) and stress reductions in the threshold for low-frequency stimulation-elicited long-term depression (LTD) in both slice preparations and intact animals [6,7] (note that in the short term, stress may facilitate LTP [8,9]). Interestingly, low-frequency stimulation-elicited LTD, which is mediated by NMDA receptors (NMDAR) – one class of ionotropic glutamatergic receptors (which also comprise AMPA and kainate receptors) – is not the sole form of LTD to be facilitated by prior stress (1–5 h beforehand), as the LTD mediated by another glutamatergic receptor – the metabotropic glutamate receptor type 5 – is also facilitated in CA1 of hippocampal slices from stressed animals [10]. On repeated exposure, stress additionally shortens and debranches apical dendrites from CA3 pyramidal neurons, possibly to compensate for the stress-elicited increase in the excitatory input from mossy fibre terminals [11,12]. Albeit less dramatic in its amplitude, a similar dendritic remodeling takes place in the CA1 region of stressed rats [13]. Lastly, stressful events have been shown to impair adult neurogenesis in the dentate gyrus whilst facilitating the death impact of neurotoxic insults on CA3 neurons [13].

Taken together, these data show that stress is endowed with major consequences on hippocampal neuroplasticity. In keeping with the finding that chronic administration of antidepressants prevents most of these consequences in stressed animals [14,15], the aforementioned indications of weakened memory storage/retrieval and hippocampal atrophies in severely depressed subjects (see above) strengthen the need to explore the relationships between stress and hippocampal neuroplasticity. However, this can only be achieved if the means by which stress exerts these effects are identified. Indeed, stress-elicited corticosteroid release, by acting directly on excitatory transmission, plays a major role in neuroplasticity.

Stress, corticosteroids & neuroplasticity

Activation of the hypothalamo–pituitary–adrenal (HPA) axis is one of the hallmarks of stress. Thus, on exposure to a stressor, corticosteroid hormones (corticosterone in rodents and cortisol in humans) are rapidly released from the adrenal cortex into the circulation to exert their numerous effects in peripheral organs and in the brain. Corticosterone binds first to the high-affinity mineralocorticoid receptor (MR) and, when concentrations of the hormone are high enough, as during stress or during the active circadian period, the low-affinity glucocorticoid receptor (GR) is then activated [16]. These two intracellular receptors are ligand-activated nuclear regulators that mediate the slow (within hours) actions of corticosterone. Recent findings in the hippocampus point, however, to the additional existence of membrane MRs that mediate rapid (within minutes) nongenomic actions of corticosterone [17].

The finding that corticosterone plays a major role in neuroplasticity comes, first, from behavioral studies indicating a major role for corticosterone in the time-dependent modifications in learning and memory retrieval elicited by stress [18]. Second, from in vitro and in vivo electrophysiological studies showing that stress-elicited corticosterone release or local (intrahippocampal) application of corticosterone at concentrations mimicking those reached during stress, act in a manner similar to stress the thresholds required to observe LTP and LTD and with a similar time-dependency [6,19,20]. Finally, from histochemical studies revealing that corticosterone release has significant impact on acute/chronic stress-elicited changes in dendrite remodeling, in neurogenesis rates and in the sensitivity to neurotoxic insults (see above). Stress- and corticosterone-elicited changes in synaptic efficacy (LTP and LTD) underline the tight relationships between the HPA axis and excitatory transmission, a finding reinforced by the observation that within minutes, application of corticosterone to CA1 pyramidal neurons reversibly increases AMPAR mini-excitatory postsynaptic currents (mEPSC) frequency (through membrane MRs) and potentiates LTP [17,21], and also that, within hours, corticosterone increases AMPAR (but not NMDAR) mEPSC and evoked EPSC amplitudes (through intracellular GRs) [22] and hampers LTP whilst facilitating LTD.

Cellular pathways involved in these physiological adaptations

The understanding of how corticosterone modulates synaptic plasticity has thus captured a lot of attention over the last decades. In the study of Groc et al.[23], the use of single nanoparticle tracking [24] unravels that corticosterone alters the strength of excitatory synapses through changes in the surface mobility of the AMPARs [23]. Indeed, the glutamatergic receptors, for example, AMPARs and NMDARs, embedded within the plasma membrane of neurons are highly dynamic [25]. In the short term, corticosterone increases AMPAR mobility, enabling glutamatergic synapses to adapt their strengths following appropriate patterns of activation (e.g., LTP induction).

This discovery links, for the first time, the surface mobility of AMPARs with the plasticity of the synaptic transmission. Importantly, these results were obtained from hippocampal neurons that play a key role in the learning and memory processes and are highly sensitive to the effects of corticosteroids [26].

A few hours after the exposure to corticosterone, the number as well as the membrane diffusion of surface AMPARs are increased. Interestingly, the outnumbered surface AMPARs are more retained within the synapse, increasing de facto their number. An important consequence of these long-term effects of corticosterone is observed at the level of the synaptic function – the impairment of synaptic potentiation [23]. As a consequence, the plastic range of the synapse is reduced.

In conclusion, a significant increase of corticosteroid concentration in the brain, such as during the circadian cycle or after a stressor event, leads to two opposite effects on the synaptic function. First, in the short term, corticosterone increases AMPAR surface diffusion without affecting synaptic transmission. However, if the synapse is exposed to specific stimulation patterns in the presence of corticosterone, then its plasticity range is increased. Second, in the long term, corticosterone increases surface AMPAR trafficking and occludes synaptic plasticity [23]. The unraveling of these cellular pathways will surely open new avenues of research for drug discovery in stress-related disorders.

Financial & competing interests disclosure

The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

No writing assistance was utilized in the production of this manuscript.