Ana Maria Magarinos

Repeated restraint stress facilitates fear conditioning independently of causing hippocampal CA3 dendritic atrophy

This study investigated whether 21 days of restraint stress (6 hr/day) and the subsequent hippocampal dendritic atrophy would affect fear conditioning, a memory task with hippocampal-dependent and hippocampal-independent components. Restraint-stressed rats were injected daily (21 days) with tianeptine (10 mg/kg; to prevent hippocampal atrophy) or vehicle then tested on fear conditioning (Days 23-25, with 2 tone-shock pairings) and open field (Day 25). Restraint stress enhanced freezing to context (hippocampal-dependent behavior) and tone (hippocampal-independent) and decreased open-field exploration, irrespective of whether tianeptine was given. Results confirmed that stress produced CA3 dendritic atrophy and tianeptine prevented it. Moreover, CA3 dendritic atrophy was not permanent but reversed to control levels by 10 days after the cessation of restraint stress. These data argue that different neural substrates underlie spatial recognition memory and fear conditioning.

Stress Effects on Morphology and Function of the Hippocampusa

The hippocampal formation, which contains high levels of adrenal steroid receptors, is vulnerable to insults such as stroke, seizures, and head trauma, and it is also sensitive and vulnerable to the effects of stress. We have discovered that the hippocampus of rodents and tree shrews shows atrophy of pyramidal neurons in the CA3 region. Psychosocial stress and restraint stress produce atrophy over approximately 3-4 weeks. Atrophy is blocked by inhibiting adrenal steroid formation and by blocking the actions of excitatory amino acids using Dilantin or NMDA receptor inhibitors. Glucocorticoid administration also blocks CA3 atrophy, but Dilantin administration blocks this as well, indicating that excitatory amino acid release mediates the atrophy, which likely involves disassembly of the dendritic cytoskeleton. Studies with in vivo microdialysis in several laboratories have shown that glutamate release in the hippocampus increases in stress and that stress-induced glutamate release is reduced by adrenalectomy. Recent electron microscopy of mossy fiber terminals on CA3 neurons has revealed a depletion of synaptic vesicles as a result of repeated stress. The mossy fiber terminals appear to be responsible for driving atrophy of CA3 neurons, which involves principally atrophy of the apical dendrites. These results are discussed in relation to data from MRI showing atrophy of the whole human hippocampus in Cushings disease, recurrent depressive illness, PTSD, and normal aging as well as dementia.

Effects of antidepressants and benzodiazepine treatments on the dendritic structure of CA3 pyramidal neurons after chronic stress

Both repeated stress and corticosterone administration induce remodeling of apical dendrites of hippocampal CA3 pyramidal neurons. Circulating glucocorticoids are involved in the mechanism that produces atrophy, along with excitatory amino acids and serotonin (5-hydroxytryptamine, 5-HT). We used 5-HT-related antidepressants and a benzodiazepine in order to explore indirectly the role of serotonin and GABA(A)-benzodiazepine receptors in the stress-induced structural changes visualized by the Golgi impregnation of the rat hippocampus. The 5-HT reuptake enhancer (+/-)-tianeptine prevented the dendritic atrophy caused by repeated restraint stress in a non-stereoselective fashion and two 5-HT reuptake antagonists, fluoxetine and fluvoxamine, failed to block dendritic atrophy. Tianeptine also functions as a therapeutic tool since it reversed the already established hippocampal atrophy caused by treatment with corticosterone for 3 weeks. Finally, the benzodiazepine agonist adinazolam was effective in preventing the stress-induced dendritic atrophy. These findings suggest that the synaptic availability of 5-HT is involved in the mechanism leading to stress-induced dendritic remodeling and supports the idea that the hippocampal inhibitory GABAergic tone may play a regulatory role.

Chronic social stress reduces dendritic arbors in CA3 of hippocampus and decreases binding to serotonin transporter sites.

Male rats housed in mixed‐sex groups in a visible burrow system (VBS) form a dominance hierarchy in which subordinate animals show stress‐related changes in behavior, endocrine function and neurochemistry. Dominants also appear to be moderately stressed compared to controls, although these animals do not develop the more pronounced behavioral and physiological deficits seen in the subordinates. In the present study, we examined the effects of chronic psychosocial stress on the morphology of Golgi‐impregnated CA3 pyramidal neurons. In addition, since serotonin has been implicated in the mechanisms mediating the dendritic remodeling seen with other chronic stress regimens, we used quantitative autoradiography to measure binding to the serotonin transporter (5HTT) in hippocampus and dorsal and median raphe. Chronic social stress led to a decrease in the number of branch points and total dendritic length in the apical dendritic trees of CA3 pyramidal neurons in dominant animals compared to unstressed controls; subordinates also had a decreased number of dendritic branch points. [3H]paroxetine binding to the 5HTT was decreased in Ammons horn in both dominants and subordinates compared to controls, while 5HTT binding remained unchanged in dentate gyrus and raphe. The similarity of the changes in 5HTT binding and dendritic arborization between both groups of VBS animals, despite apparent differences in stressor severity, suggests that these changes may be part of the normal adaptive response to chronic social stress. The mechanisms underlying dendritic remodeling in CA3 pyramidal neurons are likely to involve stress‐induced changes in glucocorticoids and in 5HT and other transmitters. Synapse 36:85–94, 2000.

Estrogen alters hippocampal dendritic spine shape and enhances synaptic protein immunoreactivity and spatial memory in female mice.

Estrogen (E) treatment induces axospinous synapses in rat hippocampus in vivo and in cultured hippocampal neurons in vitro. To better explore the molecular mechanisms underlying this phenomenon, we have established a mouse model for E action in the hippocampus by using Golgi impregnation to examine hippocampal dendritic spine morphology, radioimmunocytochemistry (RICC) and silver-enhanced immunocytochemistry to examine expression levels of synaptic protein markers, and hippocampal-dependent object-placement memory as a behavioral readout for the actions of E. In ovariectomized mice of several strains and F1 hybrids, the total dendritic spine density on neurons in the CA1 region was not enhanced by E treatment, a finding that differs from that in the female rat. E treatment of ovariectomized C57BL/6J mice, however, caused an increase in the number of spines with mushroom shapes. By RICC and silver-enhanced immunocytochemistry, we found that the immunoreactivity of postsynaptic markers (PSD95 and spinophilin) and a presynaptic marker (syntaxin) were enhanced by E treatment throughout all fields of the dorsal hippocampus. In the object-placement tests, E treatment enhanced performance of object placement, a spatial episodic memory task. Taken together, the morphology and RICC results suggest a previously uncharacterized role of E in synaptic structural plasticity that may be interpreted as a facilitation of the spine-maturation process and may be associated with enhancement of hippocampal-dependent memory.

Restraint stress reversibly enhances spatial memory performance

The effects of restraint stress on performance of a spatial memory task, the eight arm radial maze, was examined in rats. When stress was given for 6 h/day for 7 days and performance evaluated days 10-13 post stress, no effect on performance was noted; however, daily restraint stress for 13 days caused a small, but significant, enhancement of performance days 10-13 post stress. Stressed rats performed better than controls: their number of correct choices in the first 8 visits was higher than the controls, and stressed rats took fewer total choices to finish the maze than controls. Stress-dependent, enhanced performance does not appear permanent since further maze testing on days 14 and 15 post stress showed no differences between the groups. Performance of the stressed rats significantly correlated with their stress-induced, serum corticosterone levels measured after 6 h of restraint on the last day of restraint, day 13 (r = -0.63, P < 0.05); rats with higher levels of CORT took fewer choices to finish the task. Examination of hippocampal CA3c pyramidal neurons with Golgi techniques showed no effect of stress on the basal or apical dendritic arbors. Since our previous study showed that 21 days of restraint stress is associated with impaired spatial memory performance (10), these results suggest that the duration of stress may differentially affect learning/memory with shorter periods of stress serving an adaptive function while longer durations causing maladaptive changes.

Tissue plasminogen activator in the amygdala is critical for stress-induced anxiety-like behavior

Although neuronal stress circuits have been identified, little is known about the mechanisms that underlie the stress-induced neuronal plasticity leading to fear and anxiety. Here we found that the serine protease tissue-plasminogen activator (tPA) was upregulated in the central and medial amygdala by acute restraint stress, where it promoted stress-related neuronal remodeling and was subsequently inhibited by plasminogen activator inhibitor-1 (PAI-1). These events preceded stress-induced increases in anxiety-like behavior of mice. Mice in which the tPA gene has been disrupted did not show anxiety after up to three weeks of daily restraint and showed attenuated neuronal remodeling as well as a maladaptive hormonal response. These studies support the idea that tPA is critical for the development of anxiety-like behavior after stress.

Effect of Brain-Derived Neurotrophic Factor Haploinsufficiency on Stress-Induced Remodeling of Hippocampal Neurons

Chronic restraint stress (CRS) induces the remodeling (i.e., retraction and simplification) of the apical dendrites of hippocampal CA3 pyramidal neurons in rats, suggesting that intrahippocampal connectivity can be affected by a prolonged stressful challenge. Since the structural maintenance of neuronal dendritic arborizations and synaptic connectivity requires neurotrophic support, we investigated the potential role of brain derived neurotrophic factor (BDNF), a neurotrophin enriched in the hippocampus and released from neurons in an activity‐dependent manner, as a mediator of the stress‐induced dendritic remodeling. The analysis of Golgi‐impregnated hippocampal sections revealed that wild type (WT) C57BL/6 male mice showed a similar CA3 apical dendritic remodeling in response to three weeks of CRS to that previously described for rats. Haploinsufficient BDNF mice (BDNF±) did not show such remodeling, but, even without CRS, they presented shorter and simplified CA3 apical dendritic arbors, like those observed in stressed WT mice. Furthermore, unstressed BDNF± mice showed a significant decrease in total hippocampal volume. The dendritic arborization of CA1 pyramidal neurons was not affected by CRS or genotype. However, only in WT mice, CRS induced changes in the density of dendritic spine shape subtypes in both CA1 and CA3 apical dendrites. These results suggest a complex role of BDNF in maintaining the dendritic and spine morphology of hippocampal neurons and the associated volume of the hippocampal formation. The inability of CRS to modify the dendritic structure of CA3 pyramidal neurons in BDNF± mice suggests an indirect, perhaps permissive, role of BDNF in mediating hippocampal dendritic remodeling.

Morphological changes in the hippocampal CA3 region induced by non-invasive glucocorticoid administration: a paradox.

Repeated stress induces atrophy, or remodeling, of apical dendrites in hippocampal CA3 pyramidal neurons. In rats, the stress effect is blocked by adrenal steroid synthesis inhibitors, and mimicked by daily injection of corticosterone. We report that non-invasive administration of corticosterone in the drinking water (400 micrograms/ml) also produced atrophy of apical dendrites in CA3. Unexpectedly, the combination of daily stress and oral corticosterone negated the effects of either treatment alone, and no changes in the apical dendritic length or branching pattern of CA3 pyramidal neurons were observed compared to control unstressed rats.

Prevention of stress-induced morphological and cognitive consequences

Atrophy and dysfunction of the human hippocampus is a feature of aging in some individuals, and this dysfunction predicts later dementia. There is reason to believe that adrenal glucocorticoids may contribute to these changes, since the elevations of glucocorticoids in Cushings syndrome and during normal aging are associated with atrophy of the entire hippocampal formation in humans and are linked to deficits in short-term verbal memory. We have developed a model of stress-induced atrophy of the hippocampus of rats at the cellular level, and we have been investigating underlying mechanisms in search of agents that will block the atrophy. Repeated restraint stress in rats for 3 weeks causes changes in the hippocampal formation that include suppression of 5-HT1A receptor binding and atrophy of dendrites of CA3 pyramidal neurons, as well as impairment of initial learning of a radial arm maze task. Because serotonin is released by stressors and may play a role in the actions of stress on nerve cells, we investigated the actions of agents that facilitate or inhibit serotonin reuptake. Tianeptine is known to enhance serotonin uptake, and we compared it with fluoxetine, an inhibitor of 5-HT reuptake, as well as with desipramine. Tianeptine treatment (10 mg/kg/day) prevented the stress-induced atrophy of dendrites of CA3 pycamidal neurons, whereas neither fluoxetine (10 mg/kg/day) nor desipramine (10 mg/kg/day) had any effect. Tianeptine treatment also prevented the stress-induced impairment of radial maze learning. Because corticosterone- and stress-induced atrophy of CA3 dendrites is also blocked by phenytoin, an inhibitor of excitatory amino acid release and actions, these results suggest that serotonin released by stress or corticosterone may interact pre- or post-synaptically with glutamate released by stress or corticosterone, and that the final common path may involve interactive effects between serotonin and glutamate receptors on the dendrites of CA3 neurons innervated by mossy fibers from the dentate gyrus. We discuss the implications of these findings for treating cognitive impairments and the risk for dementia in the elderly.