Discussion
Previous studies have shown that in the brain the κ-OR is involved in motivation, stress-related responses and adult neurogenesis and that κ-ΟR agonists and antagonists exert potent pro- and anti-depressant effects, respectively, in rodents (6, 26). However, the mechanistic details of the aberrant synaptic function and resulting behavior mediated by κ-OR upon stress remain elusive. In the present study, we demonstrate that κ-OR plays a role in stress-induced autophagy, which leads to synaptic alterations. κ-ΟR-induced autophagy occurs primarily in the hippocampus despite κ-ΟR’s high expression levels in cortex and striatum. Gi/o proteins and U50,488H-induced ERK1,2 activation are responsible for κ-OR-mediated autophagy. An intriguing finding of the present study has been that U50,488H-dependent-CREB activation regulates the transcription of the becn1 gene, confirming that κ-ΟR activation leads to transcriptional induction of specific autophagic genes. We are thus proposing a putative G protein dependent signaling pathway for the control of autophagy by κ-ΟR. Activated κ-OR leads to activation of ERK1,2 which phosphorylates CREB, with the later promoting alterations of autophagic genes leading to synaptic protein changes (Fig. 8). This proposal is in agreement with previous observations which have shown that heterotrimeric G proteins control autophagic sequestration in HT-29 cells (27, 28), and that Gαi3, which is activated by κ-OR (8), plays a crucial role in autophagosomal membrane compartmentalization (29) and autophagy initiation (30). It is also compatible with previous findings suggesting that a dynamic interplay between Gαi3, the activator of G-protein signaling 3 and Gα-interacting vesicle-associated protein (GIV), are signaling components that determine whether autophagy is induced or inhibited (31), and that Gαi3 interacts with RGS4 to inhibit autophagy (32).
A number of different GPCRs have been shown to regulate autophagy albeit through different mechanisms (15). Thus, dopamine D2 and D3 receptors were shown to be positive regulators of autophagy, involving the Akt-mTOR and AMP-activated protein kinase (AMPK) signaling pathways (33). Similarly, activation of the β2-adrenergic receptor upregulates autophagy and increases collagen degradation in order to maintain cardiac extracellular matrix homeostasis (34). Methamphetamine exposure also induces autophagy via the κ-OR as a pro-survival response against apoptotic endothelial cell death, an effect that is also mediated by ERK1,2 activation and inactivation of the Akt/mTOR pathway (16). In the mouse hippocampus, acute or chronic morphine administration upregulates autophagic flux as protective response towards morphine-induced neuronal death and consequent spatial memory deficits (35). Moreover, chronic morphine administration alters synaptic plasticity in the hippocampus and results in spine and excitatory synapse density reduction, via generation of reactive oxygen species leading to Endoplasmic Reticulum (ER) stress activation (18). Finally, it was noted that morphine-mediated autophagy involves activation of ER stress with subsequent downstream astrocyte activation via the μ-opioid receptor (36), a finding confirming that opioids are indeed potential positive regulators of autophagy.
The immediate responses to stressful stimuli include neuro-morphological changes in multiple brain areas including the hippocampus (37). Acute stress reduces the density of dendritic spines, alters the location of postsynaptic elements of excitatory synapses, and impairs long-term potentiation and memory (14, 37). Chronic stress has been reported to enhance autophagy in rodents (38, 39). Furthermore, a role for autophagy in depressive-like behaviors and cognitive impairment has been demonstrated following prenatal stress (40). Our current findings extend the existing evidence by demonstrating a plausible scenario whereas the dynorphin/κ-ΟR system initiates autophagy, which leads to stress-induced synaptic alterations.
Neuronal autophagy plays a major role in brain function by modulating synaptic organization and morphogenesis (12, 13). It contributes to synaptic plasticity by degrading specific synaptic proteins such as PSD-95, PICK1 and SHANK3, which play important roles in synaptic function and spine modeling. This implies a direct link between autophagy and pruning of synaptic connections during postnatal development (13). In agreement with these predictions, our findings demonstrate that U50,488H-κ-ΟR activation of primary neuronal hippocampal cultures reduces the number of neurite branches. In addition, sub-chronic U50,488H administration in mice led to degradation of the key scaffolding synaptic proteins, spinophilin, PSD95 and SNAP25, particularly in the hippocampus, but not in the cortex or striatum. We specifically chose to examine these proteins as they are implicated in dendritic spine remodeling. Spinophilin localizes in the postsynaptic compartment, is enriched in dendritic spines, and modulates spine morphogenesis and maturation through the regulation of the actin cytoskeleton (41). It also interacts directly with opioid receptors and other GPCRs to regulate their trafficking and signaling that leads to synaptic alterations (41-43). On the other hand, SNAP-25 plays a crucial role pre-synaptically by mediating synaptic vesicle fusion. Of note is that SNAP-25 and PSD-95 are substrates of autophagic degradation modulating dendritic spine morphology and function (44) .
Stress blocks long term potentiation through release of endogenous opioids including the release of the endogenous κ-opioid neuropeptide, dynorphin. Activation of κ-ΟR in vivo promotes aversion, dysphoria, depression, and anxiety-like behaviors (5, 6, 45, 46). Conversely, κ-OR antagonists prevent many effects of stress and counteract stress-induced behavioral responses and for this reason, are considered as novel therapeutics for stress-related disorders (4, 47). Forced swim stress in rats elevates dynorphin A levels in the hippocampus [2], while chronic autophagy deficiency in dopamine neurons results in increased size of axon profiles, increased evoked dopamine release and rapid presynaptic recovery [12]. Another interesting finding of the present study has been that FST in mice promoted autophagy, as indicated by the elevated levels of autophagic markers and this effect was prevented by administration of the κ-OR antagonist, nor-BNI. This suggests that the endogenous dynorphin/κ-OR system is involved in stress-induced autophagy and could be part of the orchestration of structural changes observed in the hippocampus under stress exposure. Interestingly, a concomitant decrease of the three synaptic proteins, spinophilin, PSD-95 and SNAP-25, was also detected in the hippocampus but not the cortex of stressed animals. We thus postulate that degradation of these synaptic proteins could be attributed to their engulfment in the autophagosome. These findings are further supported by the observation that LC3 interacts with these three key protein substrates that possess putative LIR motifs in their sequence. In turn, this suggests that under acute stress, the release of dynorphin triggers the autophagic machinery leading to synaptosomal alterations in the hippocampus. It is interesting to note that similar synaptosomal alterations that are crucial for dendritic spine remodeling and are caused by autophagic degradation have also been reported previously under conditions of nutritional stress, through a BDNF-regulated mechanism (13). Autophagy was also reported to play a crucial role in postnatal spine pruning in layer V pyramidal neurons (48), suggesting that it plays a significant role in synaptic organization and morphogenesis.
Based on the present findings, we conclude that an interplay exists between κ-ΟR-mediated autophagy and stress-mediated synaptosomal alterations. Indeed, we propose the existence a signaling pathway (Fig. 8) correlating κ-OR-induced autophagy in neurons with synaptic hippocampal alterations under stress conditions. This κ-OR-mediated autophagy mechanism results in synaptic dysfunction in hippocampus that may contribute to the cognitive changes observed upon stress exposure. Given that κ-OR antagonists (47, 49) are in phase II clinical trials for stress-related mood and anxiety disorders, it would be interesting to explore whether these drugs effectively alleviate stress-related pathologies via κ-OR-mediated-autophagy.