Neuropathology of stress

Abstract

Environmental challenges are part of daily life for any individual. In fact, stress appears to be increasingly present in our modern, and demanding, industrialized society. Virtually every aspect of our body and brain can be influenced by stress and although its effects are partly mediated by powerful corticosteroid hormones that target the nervous system, relatively little is known about when, and how, the effects of stress shift from being beneficial and protective to becoming deleterious. Decades of stress research have provided valuable insights into whether stress can directly induce dysfunction and/or pathological alterations, which elements of stress exposure are responsible, and which structural substrates are involved. Using a broad definition of pathology, we here review the "neuropathology of stress" and focus on structural consequences of stress exposure for different regions of the rodent, primate and human brain. We discuss cytoarchitectural, neuropathological and structural plasticity measures as well as more recent neuroimaging techniques that allow direct monitoring of the spatiotemporal effects of stress and the role of different CNS structures in the regulation of the hypothalamic-pituitary-adrenal axis in human brain. We focus on the hypothalamus, hippocampus, amygdala, nucleus accumbens, prefrontal and orbitofrontal cortex, key brain regions that not only modulate emotions and cognition but also the response to stress itself, and discuss disorders like depression, post-traumatic stress disorder, Cushing syndrome and dementia.

Fig. 1

Identification of the brain regions that show activational changes during acute stress. Data are based on averages of 21 healthy subjects who were exposed to the Montreal Imaging Stress Task and showed subsequent cortisol increases, for details see [159]. Activation is prominent in a variety of limbic system and frontal lobe structures, including hippocampus, amygdala and the anterior cingulate cortex

Fig. 2

Exogenous or synthetic glucocorticoid treatment affects the human brain and reduces the numbers of CRH-immunoreactive cells in the hypothalamic PVN (a) but has no effect on oxytocin immunoreactivity in the PVN of corticosteroid-exposed subjects (d; CST). Reproduced, with permission, from [59]

Fig. 3

Quantification of glutamic acid decarboxylase (GAD)65/67-immunoreactivity (GAD-ir) (a), the total number of corticotropin-releasing hormone (CRH-ir) neurons in the hypothalamic paraventricular nucleus (PVN) (b) of depressed patients and controls. Con controls, Dep depression, MDD major depressive disorder, BD bipolar disorder. Significant increases are found in total numbers of CRH-ir in the depressed, major depressed and bipolar groups (b, see also [166]), whereas GAD65/67-ir is significantly reduced in the major depressed group (a). From [65], with permission

Fig. 4

Changes in proliferation in the human brain of depressed and anti-depressant-treated patients. a Cells immunopositive for the cell cycle marker minichromosome maintenance protein 2 (MCM2) that is involved in the control of DNA replication. In the hippocampus, many MCM2-immunopositive cells and doublets (arrows) are observed in cortical tissue of a 2-year-old subject that served as positive control. b MCM2-ir cell numbers are strongly reduced to very low numbers (arrow) in a 69-year-old control subject. c MCM2-ir doublet of 2 still closely attached cells that appear to be about to separate in the hippocampus of a depressed patient (arrow), cresyl violet counterstain. Cells are viewed under a 10× magnification (a, b) or at 63× (c). d Graphs depicting numbers of MCM2 and phosphorylated histone H3 (PH3) immunopositive cells (the latter marker reflecting late G2 and mitotic phase of cell division). PH3 immunoreactive cells in the subgranular zone and granular cell layer of the dentate gyrus, normalized to the surface area of the GCL and expressed per square micrometer. A significant reduction is found for MCM2, but not PH3, in a cohort of 10 elderly (average age of 68 years) depressed patients compared to 10 controls. e Neural progenitor and f dividing cells (Nestin and Ki-67 as respective immunocytochemical markers) are increased in the dentate gyrus of a younger cohort of (average ages of 40 and 54 years) of patients with major depressive disorder (MDD) who were treated with antidepressants compared to untreated MDDs and control subjects. Progenitor numbers (e, Nestin-ir) were higher in MDD patients treated with tryciclics (TCA) or with selective serotonin reuptake inhibitors (SSRI), compared to untreated MDD and Control cases whereas the numbers of dividing cells (f, Ki-67-ir) were higher only in the TCA but not SSRI-treated group (n = 5–7 cases per subgroup). Reproduced, with permission, from [15, 121]. [121] was used for (a–d) and [15] was used for (e, f)

Fig. 5

Schematic representation of the connections between the prefrontal cortex, the stress response and the immune system and the changed interactions during conditions of chronic stress. At basal conditions (left panel), the right medial prefrontal cortex (mPFC) is mainly under tonic inhibition from its left counterpart. Modulatory inputs from the mPFC, amygdala and hippocampus to the PVN relay on the bed nucleus of the stria terminalis (BNST). Furthermore, whereas activation of the infralimbic cortex (IL) and amygdala increases PVN activity, activation of the cingulate (Cg) and prelimbic (PL) parts of the PFC and from the hippocampus decreases it. In basal conditions the parasympathetic tone of the autonomic nervous system predominates. After chronic stress (right panel), which elevates glucocorticoid levels, changes are induced in the brain that include a decreased volume and dendritic retraction in the mPFC and hippocampus, but opposite changes in the BNST and amygdala (see Fig. 4). Damage to the hippocampus may decrease the influence of this brain structure on the mPFC and BNST (dotted lines); as a result, a reduced activity of the mPFC (especially in the left hemisphere) may occur, but an overactivation of the amygdala and over the neuroendocrine and autonomic control centers (BNST/hypothalamus). This may trigger HPA axis dysfunction, increase corticosteroid levels and activate the sympathetic nervous system, which, together, may induce immune dysregulation and contribute to behavioral dysfunction. Reproduced from [26], with permission

Fig. 6

Scheme representing the contrasting effects of chronic stress on dendritic spine numbers in the prefrontal cortex (PFC) and amygdala. a A decrease in the number of spines occurs in pyramidal neurons of the infralimbic cortex of rats after repeated restraint stress (21 days). b By contrast, chronic immobilization stress (10 days) triggers an increase in the number of spines in basolateral amygdala spiny neurons in rats. Amy amygdala, mPFC medial prefrontal cortex. Reproduced, with permission, from [26]

Fig. 7

Photomicrographs of Nissl stained sections from the hippocampus of a depressed patient (a, b), a steroid-treated patient (c, d) and a control subject (e, f). b, d, f Shows the CA3 area of the same patients at higher magnification. Although some rare apoptotic cells (brown TUNEL-positive cells indicated by arrows in g and h, compared to intact, non stained neuronal nuclei nearby (arrowhead)) were seen outside of subregions predicted to be at risk for glucocorticoid overexposure like the dentate gyrus or entorhinal cortex, no morphological evidence for neuronal damage or massive cell loss was observed in any of the groups. Bar indicates 710 μm in (a) and 45 μm in (b). Reproduced, with permission, from [117, 142]

Fig. 8

Immunohistochemical staining for synaptophysin (a, c, e) and for the neuronal growth-related phosphoprotein B-50 (b, d, f) in the hippocampus of a, b a depressed patient, c, d a steroid-treated patient and e, f a control subject. No marked qualitative difference is observed between the overall immunohistochemical staining patterns of the groups in the hippocampal subarea CA3 and the molecular layer of the dentate gyrus. Bar in e represents 50 μm, bar in f 115 μm. Reproduced, with permission, from [117, 142]

Fig. 9

Chronic stress inhibits neurogenesis in the adult hippocampal dentate gyrus. Representative confocal images of newborn neurons in the hippocampus of adult mice with low (a) and high magnification (b). A mixture of retroviruses expressing green and red fluorescent protein (CAG-IRES-GFP and CAG-IRES-RFP) was injected into the dentate gyrus of adult mice to label the newly born cells. Double-transduced cells are in yellow and DAPI is in blue. B. Czéh, D. Refojo, and D.C. Lie unpublished observations. Scale bars a 50 μm, b 20 μm. c Chronic stress inhibits both the proliferation rate and the survival rate of the newly generated cells in the hippocampal dentate gyrus of adult rats as it was shown with BrdU-labeling in a chronic social defeat stress model (modified fom [39]). Data are mean ± SEM, group sizes n = 6 rats/group, ***p < 0.001. Similar changes are seen in depressed individuals [16, 121] and treatment with some, but not all, antidepressants can normalize these changes [16, 37, 120]