Stress-Induced Depression and Alzheimer’s Disease: Focus on Astrocytes

According to Hans Selye, “stress is the nonspecific response of the body to any demand” [ 1 ]. Stress is also defined as a state of threatened (or perceived as threatened) internal dynamic balance (“homeostasis”) caused by external or internal stimuli (“stressors”) [ 2 ]. To achieve homeostasis, the highly conservative regulatory neuroendocrine system, the “stress system”, is activated through synchronized interaction between the hypothalamic–pituitary–adrenal axis (HPAA) and the autonomic nervous system [ 2 ]. In principle, stress is necessary to adapt to changing environmental or internal conditions and increase the chance of survival. It is known that moderate stress is able to activate mental and behavioral processes to find solutions to the challenges facing an individual. However, excessive and/or prolonged stressors and the consequent chronic deregulation of the stress system can lead to a wide range of chronic pathological conditions, including pathologies of the cardiovascular, endocrine, immune and nervous systems. One of the stress-related pathologies of the nervous system is depression (major depressive disorder). Currently, depression is a most widespread mental disorder worldwide, which, according to World Health Organization, affects approximately 280 million people in the world, or about 4.0% of the population, including at least 5% of adults [ 3 ]. The incidence of depression increases with age, so it is about 27% in the age group of 75–80 years, 33% in the age group of 81–85 years and reaches 46% in the age group of 91 years and older [ 4 ]. Considering that the frequency of neurodegenerative diseases (in particular, dementia) also increases with age and that depression often manifests itself in dementia, it is possible that pathological changes in the course of dementia are associated with the development of depression. Depression, as a mental disorder, is characterized by two core symptoms, depressed mood and loss of interest or pleasure in nearly all activities, and may be accompanied by other symptoms such as cognitive impairments, sleep disturbance, psychomotor retardation or agitation, feelings of worthlessness or excessive or inappropriate guilt [ 5 ]. Depression significantly worsens the quality of life. A large percentage of suicides, especially among young people, is associated with depression. Figure 1 schematically shows the general sequence of events leading to the development of chronic stress, depression and, ultimately, to the degeneration of nerve cells. Among the main factors leading to the development of chronic stress and depression, the following should be noted: Chronic pathologies of the nervous and cardiovascular systems, as well as oncological diseases. Social and psychological factors related to human living conditions and contacts with surrounding members of society and the external environment. Excessive and prolonged intake of various pharmaceutical preparations, as well as toxic compounds from the external environment, the number of which increases with the deterioration of the overall environmental situation in the world. Alzheimer’s disease (AD) is the most common cause of dementia, which is estimated to account for 60% to 80% of cases [ 6 ]. AD is considered one of the main causes of morbidity and mortality among the elderly [ 7 ]. The prevalence of AD in Europe is estimated at 5.0%, which is 3.3% in men and 7.1% in women [ 8 ]. AD is a slowly progressive brain pathology that begins many years before the onset of symptoms. Clinical symptoms in early stages of AD include difficulty remembering recent conversations, names, or events, which are often accompanied by apathy and depression. In later stages of AD, symptoms include impaired communication, disorientation, behavioral changes, and ultimately difficulty speaking, swallowing, and walking [ 6 ]. AD is characterized by the accumulation of beta-amyloid peptide (Aß) (amyloid plaques) in brain tissues and a destabilization of the cytoskeleton of neurons caused by hyperphosphorylation of microtubule-associated Tau-protein. However, the poor correlation between cognitive decline and amyloid plaques raises the question of whether Aß accumulation actually causes neurodegeneration in AD. The formation of neurofibrillary tangles of Tau correlates better with neurodegeneration and clinical symptoms, and although Aß can initiate a cascade of events leading to neurodegeneration, Tau hyperphosphorylation is assumed to be key in neurodegeneration in AD [ 9 ]. The vast majority of cases of AD belong to a sporadic form (usually, late onset of symptoms). The sporadic form of AD is associated with the interaction of genetic and environmental factors, and aging is the main risk factor. Two main genetic risk factors for sporadic AD have been identified. Firstly, the presence of the APOE4 allele encoding one of the three isoforms of Apolipoprotein E (apoE2, apoE3 and apoE4), the main transporter of cholesterol in the brain, w

Depression is a complex neuropsychiatric disorder characterized by various neuropathological and physiological symptoms [ 19 ]. The pathogenesis of depression remains unclear, and several hypotheses have been proposed for the mechanisms of development of this mental disorder [ 37 ]. In particular, the development of depression is explained by a deficiency of serotonin or norepinephrine in the synaptic cleft (monoamine hypothesis), a deficiency of neurotrophic factors in certain parts of the brain (neurotrophic hypothesis), an impairment of neurogenesis in the adult hippocampus (in particular, associated with a deficiency of neurotrophic factors and neuroinflammation), inflammatory processes (inflammatory/cytokine hypothesis) and the deregulation of the HPAA, leading to a long-term abnormalities in the levels of stress hormones. The hypothesis of the deregulation of the HPAA as a basis for the development of depression is supported by clinical data on stressful events as a risk factor for depression, on HPAA hyperactivation in patients with depression and normalization of the HPAA with antidepressant therapy [ 19 , 37 , 38 , 39 ]. It is assumed that prolonged hyperactivation of the HPAA due to chronic stress, inflammatory processes and/or genetic predisposition may lead to an impairment of self-regulation of HPAA activity and a chronic increase in peripheral and central levels of CRF (corticoliberin), ACTH (adenocorticotropic hormone) and glucocorticoids ( Figure 3 ). In turn, long-term elevated central and circulating levels of stress hormones cause changes in the functioning of the brain, leading to the development of depression. In particular, glucocorticoids effectively cross the blood–brain barrier (BBB), having a significant impact on the functioning of the brain [ 40 ]. Severe depression is often associated with hypercortisolemia, and glucocorticoid therapy can cause symptoms of depression [ 41 , 42 ]. Patients with depression have elevated levels of CRF in blood plasma, and increased expression of CRF was observed in postmorten samples of patients with a long history of affective disorders [ 39 ]. Experimental data indicate that an increase in circulating glucocorticoids and in brain CRF causes depressive-like behavior in rodents [ 39 , 43 ]. CRF-producing neurons of the paraventricular nucleus of the hypothalamus have projections not only into the median eminence of the hypothalamus, from where CRF is released into the bloodstream, connecting the hypothalamus with the pituitary gland, but also into various extrahypothalamic regions of the brain. It is assumed that elevated levels of CRF, which plays the role of a neurotransmitter or neuromodulator, are responsible for the development of depressive states in chronic HPAA hyperactivation [ 42 ]. In addition, in stress, CRF levels increase in the amygdala, and it is assumed that this leads to an increase in CRF levels in the hippocampus; both the hippocampus and the amygdala also contain significant populations of neurons synthesizing CRF [ 44 ]. In stress, CRF-producing amygdala neurons activate noradrenergic neurons of the locus coeruleus (LC), which is the main source of noradrenaline in the brain. In turn, noradrenergic neurons of the LC activate CRF-producing neurons of the paraventricular nucleus of the hypothalamus, which leads to HPAA activation. Glucocorticoids inhibit the expression of tyrosine hydroxylase in the LC [ 45 ]. Thus, chronic stress can lead to dysfunction of LC neurons, resulting in persistent deregulation of the HPAA, changes in brain norepinephrine and glucocorticoid levels and the development of depression [ 46 ]. There are accumulated data that allow us to consider the neurotrophin and stress hypotheses of depression as two sides of the same coin. Excessive levels of glucocorticoids or stress inhibit BDNF-mediated neuroplasticity in areas of the brain that process emotional experiences, that is, in the hippocampus and prefrontal cortex (PFC), while BDNF-mediated neuroplasticity increases in the nucleus accumbens (Nac), amygdala and ventral tegmental area (VTA) involved in the management of reward pathways [ 47 , 48 ]. Stress reduces the expression of BDNF and/or its high affinity receptor TrkB in the hippocampus and PFC, while BDNF increases in the Nac, amygdala and VTA. A decrease in BDNF-TrkB signaling causes the suppression of the functioning of neural networks in the hippocampus and PFC, including loss of synapses, while an increase in BDNF-TrkB signaling increases the activity of networks in the Nac, amygdala and VTA. The hippocampus and PFC are key regulators of HPA activity, providing negative feedback by activating glucocorticoid receptors (GR) expressed in these regions in large amounts [ 49 ]. It is assumed that depression is associated, in particular, with the reduced expression or activity of GR due to chronic stress, which leads to the deregulation of HPAA activity [ 38 , 48 ]. GR activity is regulated by BDNF-Tr

Since astrocytes play a crucial role in the normal functioning of neurons, the question arises whether astrocytes undergo changes during depression and AD and the consequences of chronic stress. Astrocytes express receptors for all stress hormones and therefore are a target for them both in normal stress response and in conditions of impaired self-regulation of HPAA activity. Neuroimaging data show that in depression there are noticeable structural changes in a number of brain regions—first of all, a significant decrease in the volume of the closely interconnected medial PFC and hippocampus [ 14 , 16 , 80 ]. Apparently, such changes accompany a long-term and serious pathology, since a decrease in the volume of the hippocampus is observed only in cases of depression lasting longer than 2 years or in repeated episodes of the disease [ 81 ]. The interaction between the medial PFC and hippocampus integrates motivation, attention, memory and the results of past actions as the relevant circumstances change, which ensures adaptive behavior and mental health [ 59 ]. Atrophy of these brain regions is considered as a significant predictor of the development of clinical dementia [ 45 , 82 , 83 ]. Consequently, atrophy of the medial PFC and hippocampus may be associated with pre-existing depression and/or with early stages of dementia development. A decrease in the volume of these brain regions in patients with depression may be associated with both a decrease in astrocyte density [ 84 , 85 ], and neuronal atrophy [ 85 , 86 ] and a decrease in the number and functioning of synapses [ 80 , 87 ]. This brings depression closer to neurodegenerative diseases. Data on the presence of changes in the density of astrocytes and their sizes in various brain regions in depression are contradictory [ 88 ]. However, in particular, a decrease in astrocyte density in the hippocampus in patients with depression [ 84 , 89 ] or who received chronic treatment with glucocorticoids [ 89 ] has been shown. It should be noted that there is a significant problem in the identification of astrocytes, because astrocyte marker GFAP (glial fibrillary acidic protein) used in most studies can identify not all, and sometimes only a small part of astrocytes in tissues [ 90 ], which can lead to significant difficulties in assessing real differences in astrocyte density in brain tissues in various pathological conditions. In addition, a decrease in GFAP expression in astrocytes during depression can potentially lead to a decrease in the number of detected GFAP-positive cells [ 85 , 88 ]. A recent study shows that a decrease in hippocampal volume is associated with significant neurodegeneration in the CA1 region of the hippocampus at advanced stages of AD [ 91 ]. Additionally, there are indications of astrocyte atrophy in patients with advanced stages of AD [ 92 ]. Interestingly, astrocytes obtained from induced pluripotent stem cells (IPSC) of patients not only with familial, but also with a sporadic form of AD (the only patient studied), also showed features of atrophy in vitro in comparison with control astrocytes [ 93 ]. With regard to depression, a recent study using astrocytes differentiated from IPSC of healthy and depressed donors (treated with SSRIs) did not reveal morphological differences between cell lines, as well as significant differences in the expression of astrocytic markers and astrocytic glutamate transporter (EAAT2), and in the transcriptome profile [ 94 ]. There are still insufficient data on whether there is a correlation between the clinical manifestations of sporadic AD or depression and atrophy or other changes in astrocytes obtained from iPSC. However, it has already been shown that IPSC-derived human astrocytes carrying APOE ε4/ε4 genotype are less efficient than those with APOE ε3/ε3 in neuronal survival and synaptic integrity [ 95 ] and in the uptake and clearance of Aβ [ 96 ]. Research in this direction may shed light on how genetic risk factors for AD affect the functioning of astrocytes. Considering that the most significant genetic risk factor for sporadic AD is associated with the expression of isoforms of apolipoprotein E, the main source of which in the brain are astrocytes, this could help to identify the influence on the development of AD of the interaction of environmental factors, such as stress, with a genetic predisposition to AD.

Along with a decrease in astrocytes density, possibly occurring only in the most severe cases of depression, there are data indicating changes in the functioning of astrocytes in patients with mild depression. In the cortex, LC and hippocampus, transcriptomic analysis revealed a depression-associated decrease in the expression of astrocyte-specific glutamine synthetase and glutamate transporters, which are involved in glutamate-glutamine cycle between neurons and astrocytes and play an important role in the functioning of synapses [ 113 , 114 , 115 , 116 , 117 ]. Depression is also associated with the suppressed expression of astrocytic S100B [ 114 ], aquaporin 4 [ 85 ], as well as connexins 30 and 43 and the transcription factor sox-9 regulating the expression of connexins [ 118 , 119 , 120 ]. As for the expression of glutamate synthetase in AD, a decrease in the expression of this enzyme in the temporal cortex was found compared with controls [ 121 ]. Interestingly, in contrast to depression, increased protein expression of aquaporin 4 in the frontal cortex was found in AD [ 122 ], and hippocampal mRNA and protein expression of Cx43 (also known as GJA1) was found up-regulated in AD [ 123 , 124 , 125 ]. It is possible that such opposite differences in the expression of these astrocytic proteins between depression and AD are associated with different levels of astrocyte atrophy and/or astrocyte activation, more pronounced in AD, or reflect a compensatory mechanism in response to neurodegeneration in AD. The results obtained with patient tissues, indicating a change in the transcription of genes involved in the regulation of astrocyte functions, are of very high value for understanding the role of astrocytes in depression. However, it is obvious that such data have serious limitations due to the difficulty of establishing causal relationships between the detected changes and the disease and taking into account factors such as individual differences, age and concomitant pathologies of patients. In this regard, the use of experimental models plays an important role in the study of the mechanisms of astrocytes involvement in the development of depression and AD. In CUS, astrocyte dysfunctions were detected, such as the impairment of glutamate transport and metabolism in the rat cerebral cortex [ 126 ]. CUS results in a decrease in the expression of the glutamate transporter GLT-1 at mRNA and protein levels in the rat hippocampus [ 127 ]. However, chronic immobilization stress increases GLT-1 expression in the rat hippocampus [ 128 , 129 ], which may reflect the adaptation of animals to a repetitive (predictable for animals) stressor. Using the CUS model, a decrease in the expression of Cx43 and an impairment of astrocytic intercellular contacts were detected in the rat PFC. The administration of selective serotonin reuptake inhibitors (i.e., antidepressants) or a GR antagonist (mifepristone) prevented or reversed these changes [ 130 ]. The prolonged administration of corticosterone to mice also caused an antidepressant-reversible decrease in Cx43 expression in the hippocampus [ 131 ]. It is important to note that the administration of glucocorticoids to experimental animals does not fully reproduce the effect of stress, in particular, on the functioning of synapses, and it is assumed that, under stress, there is a synergistic effect of glucocorticoids and other stress hormones, such as CRF and norepinephrine [ 14 , 132 ]. In pathological brain conditions, astrocytes undergo a number of functional and morphological changes, passing into a state called reactive astrocytes. Such astrocytes have an increased thickness of processes and begin to express nestin, which is a marker of nerve progenitor cells, and overexpress two other protein components of the astrocyte cytoskeleton, GFAP and vimentin [ 133 ]. In AD, the transition of astrocytes to a reactive state is initiated by the activation of microglia and the formation of amyloid plaques and tau-tangles. Reactive astrocytes contribute to the development of neuroinflammation by releasing inflammatory cytokines, nitric oxide, ATP and reactive oxygen species (for review, see [ 134 ]). Obviously, the reactive state of astrocytes develops during the progression of AD. Moreover, if in the serious stages of AD, there is an increase in GFAP expression in hippocampal dentate gyrus; then, in the early stages, there is a decrease in GFAP expression [ 135 ]. Remarkably, a similar situation occurs in depression and chronic stress when there is a decrease in GFAP expression in the hippocampus and cortical areas [ 85 , 136 , 137 ], suggesting a decrease in the functioning of astrocytes both in depression and in the early stages of AD. The increased expression of GFAP in the AD brain is reflected in the elevated levels in the cerebrospinal fluid (CSF) [ 138 ]. In accordance with the involvement of reactive astrocytes in a wide spectrum of brain pathologies, increased GFAP CSF lev

Studies using cell cultures show that clinically applied antidepressants have significant effects on the functions of astrocytes. Selective serotonin reuptake inhibitors were found to stimulate glucose metabolism and expression of neurotrophic factors BDNF, VEGF and VGF in mouse cortical astrocytes, but not FGF-2, GDNF and IGF-1, while tricyclic antidepressants showed no effects [ 170 ]. However, the tricyclic antidepressant imipramine stimulated the expression of GDNF [ 171 ] and BDNF [ 172 ] in rat cortical astrocytes. Another tricyclic antidepressant, amitriptyline, also stimulated the expression of BDNF in rat cortical astrocytes [ 173 ]. Both selective serotonin reuptake inhibitors (fluoxetine and paroxetine) and tricyclic and tetracyclic antidepressants, but not haloperidol and diazepam, stimulated the expression and secretion of GDNF by rat astrocytoma cells C6 [ 174 ]. However, in another study, haloperidol, as well as atypical antipsychotics, also stimulated the secretion of GDNF by C6 cells [ 175 ]. The expression of a number of neurotrophic factors, including GDNF and BDNF, is under the control of the transcription factor CREB (cAMP responsive element binding protein). Antidepressants of various classes, but not haloperidol, increased GDNF secretion and activate CREB in C6 cells during acute and chronic (3 days) administration [ 176 ]. Amitriptyline, when administered acutely, increased the levels of CREB activation in C6 cells and astrocytes obtained from human embryos [ 177 ]. Interestingly, acute fluoxetine stimulated the release of ATP from rat and mouse hippocampal astrocytes in vitro and subsequently increased BDNF in astrocytes [ 178 ], while extracellular ATP transiently increased activation of CREB and expression of BDNF in cultured rat cortical astrocytes [ 179 ]. Thus, the administration of antidepressants stimulates the expression of key neurotrophic factors in cultured astrocytes and C6 cells. Data on the effect of antidepressants on connexins expression and the effectiveness of intercellular contacts are quite contradictory. Fluoxetine increased Cx43 mRNA and protein levels in human astrocytoma cells [ 180 ]. Amitriptyline stimulated the expression of Cx43 and increased the efficiency of intercellular communication in cultured rat cortical astrocytes [ 181 ]. However, a study of the effects of antidepressants of various classes on mouse cortical astrocytes revealed the absence of a stimulating effect on the expression of Cx43 and a decrease in the effectiveness of intercellular contacts, or the absence of an effect for all antidepressants except paroxetine [ 182 ]. In this regard, the question of the effects of antidepressants on the functionality of intercellular contacts in cultured astrocytes requires further investigation. Thus, cell culture models show that astrocytes are a target for antidepressants. Antidepressants regulate their functionality and, in particular, selectively stimulate the expression of neurotrophic factors. However, it is currently not clear what their effects are on astrocyte characteristics such as the functionality of neurotransmitter transporters and the expression of inflammatory mediators. It is also unknown how and in what ways antidepressants can modulate the effects of stress hormones on astrocytes.

It is well-known that BDNF is the most important neurotrophic factor involved in the differentiation of CNS cells and their further functioning, including the morphological maturation of neurons and glia, and the formation of active synaptic contacts [ 192 ]. There are two types of receptors for BDNF that are expressed on the cells of the nervous system, including astrocytes p75 and TrkB. The p75 receptor participates in the regulation of glial cell proliferation and their response to various injuries [ 193 ]. Interestingly, astrocytes predominantly express a truncated form of the receptor (TrkB-T1), devoid of tyrosine kinase activity ( Figure 4 ) [ 194 , 195 ]. It was shown that BDNF increased the viability of astrocytes by preventing their apoptosis caused by serum deprivation, and the anti-apoptotic effect of BDNF was prevented by a specific antagonist TrkB ANA-12 [ 195 ]. At the same time, BDNF induced activation of ERK, Akt and Src in astrocytes. Blocking of the ERK and Akt pathways canceled the protection of BDNF. In addition, BDNF protected astrocytes from death induced by of 3-nitropropionic acid (3-NP). This effect was also blocked by ANA-12, ERK and Src inhibitors. The conditioned medium of astrocytes treated with BDNF completely protected the neurons from 3-NP induced apoptosis [ 195 ]. This indicates that the neuroprotective effects of BDNF may involve indirect action via astrocytes. Holt et al. [ 194 ] established the important role of the BDNF-TrkB-T1 signaling in the morphological maturation of astrocytes. The authors showed that mouse astrocytes in vitro express high levels of the BDNF receptor, TrkB, with almost exclusive expression of its truncated isoform, TrkB-T1 (90%). Moreover, the maximum level of expression was observed during the morphological maturation of astrocytes. It has been demonstrated that the morphological complexity of astrocytes increases in the presence of BDNF and depends on its signaling via TrkB-T1. The inactivation of TrkB-T1 in mice led to the appearance of morphologically immature astrocytes with significantly reduced cell volume, as well as to the impaired expression of a number of genes (Kir4.1, Aqp4, Glt1) characteristic of mature astrocytes and related to their functioning. Moreover, astrocytes with TrkB-T1 knocked out did not support normal synaptogenesis in neurons. These data indicate a significant role of the BDNF-TrkB-T1 system in the morphological maturation of astrocytes, a critical process for the development of CNS. In a recently published paper [ 196 ], the authors demonstrated direct binding of typical antidepressants and ketamine (rapid antidepressant effect), with TRKB receptors, which led to their synaptic localization and effective activation. Mutations in the receptor site responsible for binding antidepressants eliminated both cellular and behavioral effects of antidepressants in vitro and in vivo. Thus, the authors proved that at least some effects of antidepressants can be carried out through direct regulation of activation of the BDNF-TRKB complex [ 196 ]. It is tempting to speculate that putative similar regulation of signaling through the truncated form of TrkB in astrocytes might be also involved in the clinical effects of antidepressants.

Clinical and experimental data suggest that stressful events are an important risk factor for depression. In turn, depression is not only a frequent symptom of AD, but can also be an important risk factor for AD. This suggests that stress response disorders may underlie both depression and AD. Clinical data indicate atrophic changes in the same areas of the brain, the hippocampus and PFC, in both pathologies. These same brain regions play a key role in regulating the stress response and are most vulnerable to the action of glucocorticoids due to the high expression of GR. The levels of GR in PFC astrocytes are critically important for the development of depression. There is clinical evidence that not only neurons, but also astrocytes undergo atrophy in depression and AD, although precise conclusions are difficult due to methodological problems in identifying astrocytes. Animal models demonstrate that chronic stress leads not only to atrophy, but also to the pyroptotic death of both neurons and astrocytes in the PFC. Clinical data indicate a decrease in the functioning of astrocytes in both in depression and AD. In animal and cell culture models, stress and glucocorticoids significantly alter the functions of astrocytes, including the expression of neurotrophic factors and glutamate transporters. Astrocytes are key players in stress-induced ATP-mediated neuroinflammation and depression. Clinical data show that AD is mainly associated with a decrease in BDNF levels in the hippocampus and a number of areas of the cerebral cortex. The BDNF-TrkB system not only plays a key role in depression and in normalizing the stress response, but also appears to be an important factor in the functioning of astrocytes. This suggests that compounds capable of stimulating the production of endogenous BDNF or the activation of TrkB-receptors in the PFC and hippocampus may be potentially useful for the treatment or prevention of depression and AD via protection of astrocytes. Conventional antidepressants and ketamine have this ability to stimulate the BDNF-TrkB system. It is not yet known whether ketamine can delay the development of AD. However, long-term treatment with SSRI antidepressants is associated with a reduced rate of AD and a delay in the onset of AD symptoms. This supports the idea that the overlapping neurobiological bases of depression and AD might provide common ways to treat or prevent these pathologies. Progress in understanding the development, progression and treatment of depression and AD, as well as resistance to stress factors, is possible with the identification and decoding of genetic and especially epigenetic changes in the human genome that contribute to their course [ 211 , 212 ].