06+Stress

[|Source]12.1 STRESS
[| Go to:]
 * [|Top▲]

12.1.1 Concept of Stress
Before dealing with the different topics related to stress, aging, and memory interactions, important issues in relation to the concept of stress and stress physiology must be introduced. Stress is considered to imply any challenge to the homeostasis of an individual that requires an adaptive response of that individual.[|8] Since life is a cumulative exposure to changing and challenging situations, virtually all living organisms experience stress, more or less frequently, during their life spans. Although stress is a loose concept that historically has meant different things for different authors, today we recognize the importance of distinguishing three components that together define every stress experience.

12.1.1.1 Stressors
Stressors are stimuli, generally aversive and potentially harmful, that exert impacts on individuals. Stressors can be classified as either exteroceptive (extreme temperatures, electric shocks, social situations) or interoceptive (ranging from health problems such as gastric disturbances to psychogenic problems such as unjustified fear).

12.1.1.2 Evaluation of Situation
The way an individual interprets a potentially stressful situation is a critical step to determine whether a specific stimulus acts as a stressor. A sudden noise can be judged as dangerous by one individual and experienced as harmless by another. Their respective reactions can depend on many factors such as previous experiences with similar noises, or may be based on the expectations that each individual generates about the potential consequences derived from that particular noise. Various psychological processes are important, with controllability or the ability to cope with the situation serving as a very important factor in determining how stressful situations are experienced.[|9–11]

12.1.1.3 Response of Individual
Response includes both physiological and behavioral reactions to a stressful situation. The physiological stress reaction typically comprises central (sensory, emotional, and cognitive processing of stimuli by the central nervous system) and peripheral (activation of the sympathetic nervous system and the hypothalamus–pituitary–adrenal axis) responses (see below). The behavioral reactions include both direct responses to the specific stressors and adaptive responses that are addressed to optimize survival.[|8]

12.1.2 Physiological Stress Response
The stress response involves a complex reaction in the organism that, in addition to the activation of peripheral stress systems, includes the activation of specific circuits in the brain. Most of these neural circuits have the capacity not only of processing information, but also eventually affect the degree and direction of activation of peripheral physiological systems.[|12] As to the peripheral responses, the two major systems activated during stress are the sympathetic (SNS) branch of the autonomic nervous system (ANS) and the neuroendocrine system consisting of the hypothalamus–pituitary–adrenocortical (HPA) axis.

12.1.2.1 Sympathetic Nervous System
Unlike the parasympathetic branch of the ANS that mediates calm vegetative functions such as growth, digestion, and relaxing responses of the organism, the SNS is stimulated by activating and stressful situations. This system comprises a number of projections that connect with virtually every organ in the body where they secrete norepinephrine. An important projection of the SNS is its input to the medulla of the adrenal glands, where adrenaline and noradrenaline hormones are secreted into the bloodstream. Many well-known responses to stress are caused by activation of the SNS, including increased heart rate and blood pressure, increased glucose levels, increased muscle tension, and increased sweating. In parallel, activation of the SNS delays functions that are not directly required to survive at that particular moment; typical examples are the lessening or suspension of digestion and reproduction.

12.1.2.2 Hypothalamus–Pituitary–Adrenal Axis
Most of the work examining the deleterious effects of stress on memory function has focused on the HPA axis ([|Figure 12.1]). This neuroendocrine system involves the sequential activation of messenger molecules produced by the hypothalamus, the pituitary, and the adrenal cortex. The main hypothalamic HPA messengers, corticotrophin releasing hormone (CRH) and vasopressin (AVP), are synthesized in the paraventricular nucleus. Upon the appropriate stimulus, these peptides are released and, through the portal vein system, get access to the anterior pituitary where they stimulate the production and release of the adrenocorticotropic hormone (ACTH) into the bloodstream. Eventually, ACTH reaches the adrenal cortex where it stimulates the secretion and production of glucocorticoids (cortisol in humans; corticosterone in a variety of animals including rodents). ====[|FIGURE 12.1]==== The hypothalamus–pituitary–adrenal axis. Activation of the hypothalamus results in a chain of events that eventually result in the release of glucocorticoids. Once in the bloodstream, these steroid hormones exert negative feedback at the [|(more...)] The hypothalamus–pituitary–adrenal axis. Activation of the hypothalamus results in a chain of events that eventually result in the release of glucocorticoids. Once in the bloodstream, these steroid hormones exert negative feedback at the different stations of this neuroendocrine axis. Importantly, glucocorticoids can also penetrate into the brain rapidly and affect different aspects of behavior and cognition. Glucocorticoids are steroid hormones that produce extensive effects on virtually all physiological systems. Among their many roles, they exert essential feedback actions at a variety of levels (prefrontal cortex, hippocampus, hypothalamus, and pituitary) to inhibit the activity of the axis. Such negative feedback is crucial to suppress excessive levels of these steroids, whose brief action can be highly adaptive, but their maintenance at high levels for prolonged periods can be highly detrimental to an organism. Due to their lipophilic nature, glucocorticoids can achieve rapid access to the brain. In addition to rapid nongenomic actions through membrane receptors, glucocorticoids affect the brain by acting through two classical intracellular corticosteroid receptors that exert genomic effects.[|13] They are the mineralocorticoid receptor (MR) and the glucocorticoid receptor (GR). Corticosterone binds with a 10-fold higher affinity to MRs than to GRs and therefore it is not surprising that many stress effects are mediated through GRs. The hippocampus shows the highest density of corticosteroid receptors, with some amygdala nuclei and the prefrontal cortex also showing moderate to high levels of GRs.[|1]

12.1.3 Stress and Memory Function
As noted above, //stress// covers a wide spectrum of circumstances that can eventually have differential effects in the acquisition, consolidation, and retrieval of information. Based on the importance for mental health of the negative effects that highly stressful circumstances can impinge on cognitive function, the focus of this chapter is on the detrimental effects of stress on memory processes.[|3],[|4],[|6],[|14] For reviews on the positive aspects of moderate brief stress periods on memory formation, see References [|5], [|7], and [|15]. A number of important factors related to stress and cognition must be taken into account when trying to understand how stress affects cognitive function. The following parameters are particularly important.

12.1.3.1 Stress Magnitude and Intensity
Intuitively, the impacts of extreme stressors (such as a real life threat, for example, a strong earthquake) on cognitive function are expected to differ greatly from those impinged by moderate stressors (such as exposure to novelty), and experimental evidence largely supports this view.[|16],[|17] In any case, it is important to note the drastic individual differences in stress reactivity existing among conspecifics. Therefore, when evaluating the impact of stress intensity, it is advisable to take into account both the specific characteristics of the stressor and measure individual behavioral and physiological responses in order to determine the actual stress magnitude experienced by each experimental subject.

12.1.3.2 Stressor Timing
The time when stress is experienced with regard to the cognitive function under evaluation seems to be a crucial factor of both the types of effects observed and the mechanisms implicated.[|17] Depending on whether stress is experienced before, during, or after the cognitive challenge, different processes (acquisition, consolidation, and/or retrieval of information) can be affected. Consistent evidence indicates that acute stress exerts different effects on consolidation (frequently facilitating) and retrieval (frequently impairing) of information.[|5–7],[|17]

12.1.3.3 Stressor Duration
As we will see in the following sections, taking into account the duration of the stressor is essential, particularly when the question under analysis is related to the mechanisms whereby stress impairs cognitive function. We will review the experimental work that illustrates impaired memory function following either acute or chronic stress (see below).

12.1.3.4 Stressor Controllability
Substantial work in humans and animals indicates that an individual’s perception of his or her ability to cope with a stressful experience has profound consequences on the degree of cognitive alteration induced by stress. Uncontrollable stressors generally provoke more behavioral impairments than controllable stressors, and many neurochemical changes ordinarily elicited by uncontrollable stressors are not observed when control is possible.[|10],[|11],[|18] However, recent evidence suggests that the different impacts produced by controllable and uncontrollable stressors in the brain may not be due simply to the contribution of uncontrollability, but may in fact be affected by the ability to control. By using a triadic design (see [|Figure 12.2]), the medial prefrontal cortex (mPFC) was proposed to inhibit stress-induced neural activity in brainstem nuclei (notably, the dorsal raphe nucleus) in individuals who exerted control over stress in contrast to the prior view that such brainstem activity was induced by the lack of control.[|19] In addition to the memory phase under study (see Section 12.1.3.2, “Stressor Timing”), other factors are also important to take into account with regard to the cognitive function under study.

12.1.3.5 Factors Related to Memory Processes
As mentioned above, not all phases in the information process related to memory function are equally susceptible to disruption by stress. It is, therefore, very important to design experiments that allow one to establish which memory phase (acquisition, consolidation, retrieval, or even reconsolidation) is affected by the stress procedure under study ([|Figure 12.3]). ====[|FIGURE 12.3]==== Depiction of the importance of the timing when stress is experienced with regard to the phase of information processing under study. If stress is given before acquisition of information (1), it can potentially affect all cognitive phases involved in memory [|(more...)] Depiction of the importance of the timing when stress is experienced with regard to the phase of information processing under study. If stress is given before acquisition of information (1), it can potentially affect all cognitive phases involved in memory function (learning, memory consolidation, retrieval). If stress is experienced after acquisition (2), any effect observed in retention may be due to an impact of stress on either consolidation or retrieval and any effects on acquisition can be discarded. If stress is delivered before the retention test (3), it should normally affect the retrieval processes. However, a note of caution should be mentioned, depending on how soon the retention test is applied with regard to training since consolidation mechanisms are increasingly recognized to last longer than previously hypothesized and therefore this type of manipulation may influence both consolidation and retrieval. Research should take into account this complexity and apply the necessary controls to ascertain which phases and mechanisms of information processing are affected by the stress procedure under study. Another particularly important factor is the type of learning process evaluated (i.e., implicit/procedural explicit/declarative, nonassociative learning, etc). Notably, implicit memory processes have been shown to be positively influenced by stress.[|20],[|21] Both acute[|20–22] and chronic[|23–25] stress experiences were reported to potentiate associative types of learning such as eyeblink conditioning and fear conditioning in male (but not female[|20]) rodents. On the contrary, explicit/declarative/relational types of memories are much more vulnerable to interference by stress. Since this chapter covers the detrimental effects of stress on memory, we will mainly focus on these latter types of memory that have been shown to be particularly vulnerable to alteration by stress.

12.1.4 Acute Stress and Memory Impairment
Experiencing an acute highly stressful situation can interfere with subsequent information processing. This holds true particularly for those circumstances in which a stressed individual is required to retrieve previously stored information while the acquisition of new information is shown to be particularly resistant to disruption in experimental animals. In fact, most rodent studies in which acute stress has been applied before animals were confronted to learn a hippocampus-dependent task failed to find alterations in the acquisition rate.[|26],[|27] If any consistent effects were observed, in most cases they were not evident in the performance of animals during the training (or learning) phase, but appeared in subsequent retention tests. For example, Baker and Kim[|28] showed that exposure to uncontrollable stress can affect a nonspatial task, the object-recognition memory. In their study, rats given inescapable restraint and tail-shock stress just before exposure to a novel object recognition task showed normal memory when tested 5 min after first exposure to objects, but were impaired when tested 3 hours later. Control rats displayed a preference for a novel object (over a familiar one) when they were tested at different time delays (5 min and 3 hours). Unlike the unstressed controls, at the 3-hour posttraining test, stressed animals spent comparable time exploring novel and familiar objects. When the impact of stress on the retrieval of previously acquired information was directly assessed, similar detrimental results on retention were reported. De Quervain et al.[|29] found that exposing rats to either stress or glucocorticoids 30 min before testing impaired retention performance in the spatial task Morris water maze. Convincing evidence indicates that the level of difficulty of the task (memory load) is a critical factor in observing the detrimental effects of stress on retrieval processes. Using the radial arm water maze (RAWM; a modified Morris water maze that contains four or six arms, with a hidden platform located at the end of one of them), Diamond et al.[|30] showed that exposure to a cat during a 30-min delay period between training and testing for the platform location (the platform was located in the same arm on each trial within a day and was in a different arm across days) had no effect on memory in the easiest RAWM, but stress impaired memory in more difficult versions of the RAWM. By lesioning the hippocampus, the authors also confirmed that the RAWM is a hippocampal-dependent task. In addition to the importance of memory load (difficulty or memory demand of the task), it seems that flexible forms of memory are particularly susceptible to show disrupted retrieval by stress, as opposed to more stable ones that remain largely unaffected.[|31] In humans, stress or pharmacological glucocorticoid treatments given just before retrieval have also been found to impair the recovery of information.[|7],[|32–35] As in animals, memory load is also an important factor for stress-induced retrieval impairments in humans.[|33] Interestingly, the effect of stress in memory retrieval seems to be related to the emotional content of the information. For example, psychosocial laboratory stress (as induced by the Trier Social Stress Test) was shown to particularly impair recall of emotionally arousing but not of neutral words.[|36] Therefore, emotionally arousing material appears to be especially sensitive to the impairing effects of stress in retrieval.

12.1.5 Neurobiological Mechanisms Involved in Acute Effects of Stress on Memory
Cognitive and neurobiological studies have provided converging evidence that the hippocampus is critically involved in long-term memory formation[|37–39] and also a primary central nervous system target of stress hormones.[|2],[|6] The great sensitivity of the hippocampus to stress is revealed by the profound suppression of hippocampal synaptic plasticity after acute exposure to stressors[|40–44] or increased glucocorticoids.[|44] Moreover, adrenergic activation in the basolateral amygdala and hippocampus was shown to be critical for the impairing effects of glucocorticoids on delayed memory retrieval in spatial water maze tasks.[|45] A crucial role for the medial temporal lobe (and the hippocampus in particular) in mediating these stress-induced retrieval impairments is also supported by human neuroimaging studies.[|33] Specifically, de Quervain et al.[|33] used positron emission tomography (PET) to investigate the effects of pharmacologically increased glucocorticoid levels in regional cerebral blood flow during declarative memory retrieval in healthy male humans. A single stress-level dose of cortisone (25 mg) given 1 hour before testing impaired cued recall of word pairs learned 24 hours earlier but did not significantly affect performance in other tasks such as verbal recognition, semantic generation, and categorization. Simultaneously, this treatment resulted in a large decrease in regional cerebral blood flow in a number of brain areas including the right posterior medial temporal lobe, left visual cortex, and cerebellum. The decrease in the right posterior medial temporal lobe was maximal in the parahippocampal gyrus, a region associated with successful verbal memory retrieval. In addition to the hippocampus, evidence indicates that acute stress-induced memory impairing effects can also be mediated by activation of dopaminergic[|46],[|47] and noradrenergic[|48] transmission in other structures known to be involved in high-order (including working memory and executive function) processing such as the prefrontal cortex (PFC). Only a few studies have been reported on potential molecular mechanisms whereby stress could lead to less effective functioning of neural networks during retrieval. Recently, the potential role of the neural cell adhesion molecule (NCAM) was investigated in a rat model of stress-induced retrieval deficits in the RAWM by cat stress.[|49] NCAM (see [|Figure 12.4]) is a part of a family of cell surface glycoproteins that play key roles in neural development and in synaptic plasticity in the adult brain.[|50–52] Encoded by a single gene, the three main isoforms derived by alternative splicing are NCAM-120, NCAM-140, and NCAM-180 according to their approximate molecular weights. In addition to playing roles in cell–cell recognition and synapse stabilization, NCAM also participates in neurite outgrowth, activation of signal transduction cascades, and synapse formation and elimination.[|53],[|54] Moreover, NCAM has been implicated in the induction of hippocampal long-term potentiation (LTP; a physiological model of memory) and in memory formation.[|50–52] Finally, these molecules have been shown to be sensitive to stress.[|4] ====[|FIGURE 12.4]==== Cell adhesion molecules of the immunoglobulin superfamily NCAM and L1. Left: the molecular structures of these molecules are represented. NCAM has three major isoforms that differ in molecular weight and type of attachment to cell membranes. Right: two [|(more...)] Cell adhesion molecules of the immunoglobulin superfamily NCAM and L1. Left: the molecular structures of these molecules are represented. NCAM has three major isoforms that differ in molecular weight and type of attachment to cell membranes. Right: two of the main mechanisms whereby these molecules regulate synaptic function and plasticity are illustrated: (1) cell–cell adhesion, important for the formation and maintenance of synapses and circuits and (2) synaptic de-adhesion, a process in which NCAM polysialylation by polysialic acid (PSA) plays a key role in allowing plasticity to remodel synapses and circuits. In the cat stress study,[|49] rats were trained to locate a hidden platform and then during a 30-min delay period they were either left undisturbed or exposed to a cat, after which all animals were given retention trials and brain samples [hippocampus, basolateral amygdala (BLA), PFC, and cerebellum] were extracted immediately afterward to assess for NCAM levels in synaptosomal preparations. Two other control groups were included: a group of undisturbed rats submitted only to handling and a swim control group that was exposed to the maze but not to spatial learning. The platform location changed from trial to trial. NCAM expression in the hippocampus was not altered in animals with intact spatial memories that were not stressed. However, predator exposure impaired spatial memory and dramatically reduced NCAM levels in the hippocampus (particularly the NCAM-180 isoform) and PFC (although specificity of the PFC effect is questioned since reduced NCAM levels were also found in trained but unstressed animals and in the swim control group). No significant changes in NCAM levels were observed in the amygdala or cerebellum. These observations of drastic reductions of NCAM in stressed memory-impaired rats is consistent with an increasing body of data indicating that hippocampal NCAM is important for long-term memory formation.[|55–58] The drastic suppression of hippocampal NCAM levels found in the hippocampus after rat stress may also contribute to impaired long-term consolidation and/or retrieval processes of spatial memory.

12.1.6 Impairing Effects of Chronic Stress on Cognitive Function
Prolonged exposure to stress is now recognized as a condition that can induce deleterious effects on brain structure and cognition[|2],[|59],[|60] and increase the risks of developing neuropsychiatric disorders.[|61],[|62] Since most of the pioneer work in the field focused on the hippocampus as a primary target of stress actions,[|2],[|63],[|64] the possibility that chronic stress affects hippocampal-dependent learning has been extensively tested. Chronically stressed male rats were shown to exhibit learning and memory deficits in a variety of spatial tasks including the radial-arm maze,[|65] the Y maze,[|66] the radial-arm water maze,[|67] and the Morris water maze.[|57],[|68] Evidence from the animal and human literature supports the existence of considerable variability in the vulnerability to stress among conspecific individuals.[|69–72] Numerous studies have reported important individual differences in the impact exerted by stress in learning, memory, and retrieval processes.[|70–73] While some individuals are particularly vulnerable, others may be resistant to the effects of stress. These differences may be due to predisposing factors, to previous life experiences or, more likely, to both. Given the devastating consequences that stress impinges in susceptible individuals, developing tools able to predict which individuals are in particular danger would be of great value for developing more effective strategies to prevent and/or reverse the effects of adverse life periods. Three types of factors have been identified as particularly important in influencing an individual’s susceptibility to develop cognitive alterations under chronic stress: (1) certain personality traits; (2) gender; and (3) age.

12.1.6.1 Personality Traits
The level of locomotor activity displayed by rats in a novel environment has been identified as an accurate index to categorize individuals with relevant psychobiological profiles.[|74],[|75] By exposing rats to novelty, it is possible to classify them in groups, one comprising those that exhibit a high locomotor activity (high-responding or HR) and another including those that present low levels of activity (low-responding or LR). See [|Figure 12.5]. This behavioral trait of novelty reactivity in rats has been proposed to resemble some of the features of high-sensation seekers in humans.[|76] ====[|FIGURE 12.5]==== One of the classic experimental procedures used to characterize animals on the behavioral trait of novelty reactivity. Animals were exposed to a novel environment (open field is depicted in the figure; circular corridors are classically used) and their [|(more...)] One of the classic experimental procedures used to characterize animals on the behavioral trait of novelty reactivity. Animals were exposed to a novel environment (open field is depicted in the figure; circular corridors are classically used) and their locomotor activity during a particular period (ranging from 5 min to 1 hour or longer) was monitored. The animals were then classified as highly reactive (HR) or low reactive (LR) by comparing their performance in relation to the sample distribution. Individual differences in reactivity to novelty in adult male rats have been related to differences in susceptibility to develop cognitive alterations after exposure to chronic stress.[|77] Specifically, when 4-month old LR and HR Wistar male rats were submitted to psychosocial stress for 21 days (daily cohabitation of each young adult rat with a new middle-aged rat), HR, but not LR, rats subsequently showed marked deficits in spatial learning in the water maze. Anxiety trait is a well-known risk factor for the development of stress-related neuropsychiatric disorders, like depression in humans,[|78],[|79] and it has been associated with degrees of cognitive impairment following chronic stress in rodents. Specifically, peripubertal anxiety levels of male rats (as evaluated using open field and elevated plus mazes at 43 days of age) were shown to be predictive of the detrimental effects of chronic restraint stress (21 days) on hippocampal-dependent spatial memory as assessed in young adulthood (75 days). Memory was tested on the spatial Y-maze using two inter-trial interval levels of difficulty (1 min or 4 hours). No differences among groups were observed in the less difficult 1-min version of the Y-maze. However, in the 4-hour version of the Y-maze, chronically stressed high anxiety rats — but not the other groups — showed impaired spatial memory. Moreover, a month after the chronic stress ended, high anxiety rats had significantly higher basal corticosterone levels than low anxiety rats (control and stress). In fact, anxiety trait in rats was also found to predict impaired spatial learning performance in the stressful water maze task under acute conditions[|80] that highlight anxious individuals as particularly prone to show cognitive deficits under stressful conditions.

12.1.6.2 Gender
The importance of gender on the effects of stress in cognition remained elusive until recently due to the routine use of only male rodents in behavioral studies. However, intensive work over the past few years involving female rodents shows that gender is indeed a critical factor in an individual’s susceptibility to chronic stress.[|81],[|82] When both male and female rats were submitted to chronic stress procedures (such as 21 days of chronic restraint stress), males were impaired in all tasks in which they were tested (novel object recognition and two spatial memory tasks: object placement and radial arm maze), while females were either enhanced (spatial memory tasks) or not impaired (nonspatial memory tasks).[|83] As indicated below, age seems to be an important factor in the modulatory role of gender in stress and memory interactions.

12.1.6.3 Age
Age has been identified as a critical factor in the interactions between stress and cognition from two main perspectives. One is related to a differential susceptibility to stress that is manifested by some individuals at different times across the life span. The second relates to the impact that experiencing stress at a particular life period might have in future cognitive functioning. With regard to the first perspective, aging has been clearly identified in male rats as a risk factor for developing stress-induced cognitive impairments. A pioneer study by Bodnoff et al.[|84] in young adult and mid-aged male rats showed that mid-aged rats were more vulnerable to the effects of chronic corticosterone administration. Three months of steroid treatment at doses sufficient to mimic the elevated hormone levels observed following exposure to mild stress induced learning impairments in the mid-aged but not young, rats in the Morris water maze. Mid-aged rats exposed for 6 months to high social stress were also pronouncedly impaired in spatial learning. This effect was prevented by adrenalectomy. This and related findings[|85],[|86] (see below) highlight midlife as a time of particular sensitivity to the effects of chronic stress and corticosteroid hormones. Interestingly, the effects of aging seem to be gender-dependent. A recent review of the literature[|82] pointed out that whereas the impairing effects of stress on male rodents are observed across the whole lifespan, females show more variable responses to stress. Stress-induced facilitations observed in females in young adulthood were not further observed following stress exposure at old age.[|82] The second perspective implies that exposure to stress at a particular life period may have long-term and/or delayed consequences in memory function. Early life experiences are known to exert profound influences in stress reactivity in adulthood and cognitive aging.[|62],[|87] Much work has been done with early postnatal stress manipulations[|88–91] which, in addition to affecting other behavioral and physiological aspects in later adulthood, consistently resulted in learning and memory deficits in hippocampus-dependent tasks such as the water maze. Interestingly, a recent study has presented evidence that some consequences of early-life stressful experiences may not be manifested during young adulthood, then become apparent later during midlife. Brunson et al.[|92] explored whether psychological early-life stress in rats caused an enduring deterioration of hippocampal function that worsened from young adulthood to middle age. To induce stress, environment and maternal behavior were altered by placing pups and dams in cages with limited nesting and bedding material on postnatal days 2 to 9. This resulted in abnormal nurturing behavior in the dams including reduced and fragmented nursing and grooming of pups. The selection of such procedure was based on its ability to produce substantial neuroendocrine changes early in life[|93] that become fully normalized by adulthood.[|92] Although the offspring showed virtually normal cognitive function during young adulthood (4 to 5 months of age), they were severely impaired in hippocampus-dependent tasks (spatial learning in the water maze and novel object recognition) at mid-age (12 months). The authors suggested that stress during periods of hippocampal development may permanently influence hippocampal systems that are particularly vulnerable during these periods.[|89],[|94] Substantial work indicates that lifetime exposure to stress can also affect cognitive function at aging. This is discussed in detail below, after the next section that provides a general overview on the main neurobiological mechanisms that have been implicated in the deleterious actions of chronic stress on memory function.

12.1.7 Neurobiological Mechanisms Involved in Deleterious Effects of Chronic Stress on Brain and Behavior
First, it is important to note that the mechanisms whereby chronic stress impairs cognitive function are not necessarily the same as the ones mediating acute stress effects. While neural alterations involved in acute stress effects seem to be mainly mediated by dynamic functional alterations among cellular and molecular interactions, chronic stress is now known to have a major impact on both functional aspects and neuronal structures. In this section, the main structural and functional effects of chronic stress on specific neural circuits will be discussed, followed by an overview of the molecular processes reported to contribute to such effects.

12.1.7.1 Structural Effects of Chronic Stress
Because many examples in the literature indicating impairing effects of chronic stress in memory processes were obtained in hippocampus-dependent tasks, the hippocampus is the brain region that has received the most attention. However, intensive work during the past few years is providing increasing evidence for a more integral impact of chronic stress throughout the brain that, as illustrated below, is now documented to a certain extent at the level of the prefrontal cortex and amygdala.

Hippocampus
The hippocampus plays a central role in memory processes,[|39],[|95] particularly in spatial learning which is generally affected by stress manipulations.[|96] In humans, neuroimaging studies have reported hippocampal atrophy in association with stress- and glucocorticoid-related cognitive and neuropsychiatric alterations.[|97–99] In rodents, the CA3 subregion appears to be particularly vulnerable to the effects of chronic stress. In rats subjected to stress for 3 to 4 weeks, CA3 has been reported to experience the following structural alterations: ====[|FIGURE 12.6]==== Structural effects of chronic stress in the hippocampus and amygdala. Different hippocampal subregions can be markedly affected by exposure to chronic stress. The upper part of the figure represents the stress-induced atrophy of apical dendrites in CA3 [|(more...)] Structural effects of chronic stress in the hippocampus and amygdala. Different hippocampal subregions can be markedly affected by exposure to chronic stress. The upper part of the figure represents the stress-induced atrophy of apical dendrites in CA3 and the inhibition of neurogenesis in the dentate gyrus (DG). The lower part shows the increased dendritic arborization described in the basolateral amygdala (BLA). Although stress-induced alterations in CA1 morphology are not as drastic as those occurring in CA3, some changes have also been reported in this hippocampal subregion (particularly in excitatory axospinous synaptic connectivity in rat CA1 stratum lacunosum moleculare) after stressing rats for 3 to 4 weeks. These changes include: In addition, stress and high glucocorticoid levels can suppress neurogenesis in the dentate gyrus[|106] ([|Figure 12.6]). Furthermore, stress can compromise cell survival and eventually lead to overt neuronal loss by exacerbating the neurotoxicity induced by other hippocampal insults.[|107]
 * Dendritic atrophy of apical CA3 pyramidal neurons[|100],[|101] ([|Figure 12.6])
 * A striking reorganization within mossy fiber terminals[|102]
 * Synaptic loss of excitatory glutamatergic synapses[|57],[|103]
 * A reduction in the surface area of postsynaptic densities[|103]
 * A marked retraction of thorny excrescences[|104]
 * Alterations in the lengths of the terminal dendritic segments of pyramidal cells in rat CA1[|103]
 * Increases in postsynaptic density surface area and volume in CA1 stratum lacunosum moleculare[|105]
 * An overall reduction of the dorsal anterior CA1 area volume[|105]

Prefrontal cortex
The prefrontal cortex (PFC), particularly its medial part (mPFC), is critically involved in higher cognitive processes and in the integration of cognitive and emotionally relevant information.[|108–110] Moreover, the PFC contains high levels of glucocorticoid receptors[|111],[|112] and is also involved in the regulation of stress-induced hypothalamic–pituitary–adrenal (HPA) activity.[|113] Clinical evidence highlights mPFC as a core alteration in a wide variety of neuropsychiatric disorders.[|114],[|115] Rodent studies have provided evidence that major neuronal remodeling occurs in the mPFC as a consequence of repeated exposure to chronic stress or repeated glucocorticoid treatment. Chronic stress also results in major changes in layer II/III of the PFC following 21 days of repeated stress: Glucocorticoids seem to be major players in the remodeling induced by stress in the mPFC. Rats chronically treated (4 weeks) with either corticosterone (25 mg/kg) or dexamethasone, a synthetic glucocorticoid (300 μg/kg), showed neuronal loss and atrophy of layer II of the infralimbic, prelimbic, and cingulate cortices.[|119] Moreover, morphological studies have established that chronic daily corticosterone injections (3 weeks) in rats resulted in dendritic reorganization in pyramidal neurons in layer II-III of the mPFC,[|120] with major changes observed in apical arbors consisting of increased dendritic material proximal to the soma and decreased dendritic material distal to the soma.
 * Dendritic atrophy: decrease of total length[|116],[|117] and number[|116] of apical dendrites from pyramidal neurons.
 * Spine loss: a decrease in apical dendritic spine density. It is estimated that nearly one-third of all axospinous synapses on apical dendrites of pyramidal neurons in medial PFC are lost following repeated stress.[|118]

Amygdala
The amygdala plays key role in emotional behavior and especially in fear.[|121] It is not yet clear whether this structure is involved in the deleterious effects of stress in memory function since amygdala-dependent memories such as fear conditioning are potentiated by chronic stress.[|23],[|24] Strikingly, the structural alterations that have been observed in the amygdala contrast with the dendritic atrophy observed in the hippocampus or PFC. Repeated exposure of rats to restraint stress (10 days) induced enhanced dendritic branching of pyramidal and stellate neurons in the BLA[|122] ([|Figure 12.6]). This effect was dependent on the stressor used, since no changes were observed in these neuronal types following a chronic unpredictable stress procedure that, instead, induced atrophy only in BLA bipolar neurons.[|122] Moreover, the restraint procedure also resulted in increased spine density across primary and secondary branches of spiny neurons in the BLA.[|123] Further studies are needed to confirm whether sensitization of amygdala activation occurring as a consequence of sustained stress exposure may also be an important component of the reported memory impairments in more explicit types of memories.

12.1.7.2 Effects of Chronic Stress on Synaptic Plasticity
Electrophysiological experiments have consistently shown impaired synaptic plasticity following chronic stress, indicative of functional consequences on neural circuits of the structural alterations described above. Thus, long-term potentiation (LTP) is impaired in different hippocampal areas including CA1,[|124],[|125] the commissural/associational (but not mossy fiber) input to CA3,[|126] and the dentate gyrus.[|124] Likewise, treating rats chronically with corticosterone was found to impair hippocampal synaptic potentiation.[|84],[|127] Moreover, evidence indicates stress-inducing changes in LTP in the mPFC–amygdala pathway.[|128] Interestingly, early-life stress can also result in late-onset hippocampal dysfunction. Early-life stress in rats causes a decline in a number of measures of synaptic function and plasticity (LTP in CA1 and CA3 hippocampal subregions) when evaluated at mid-age (12 months).[|92]

12.1.7.3 Molecular Alterations Induced by Chronic Stress
A number of molecular mechanisms seem to participate in the deleterious effects induced by stress in brain structure and cognitive function. Certain neurotransmitters, signal transduction pathways, neurotrophic factors, and adhesion molecules have been implicated in the effects of chronic stress on the brain.[|4],[|5],[|59],[|107],[|129]

Excitatory amino acids
Alterations in glutamatergic transmission have been proposed to result in an excitotoxic cascade of mechanisms finally leading to neuronal endangerment and/or neurotoxicity.[|107] In line with evidence that stress and glucocorticoids increase glutamate levels in the hippocampus and other brain regions,[|130–132] glutamate has been involved in the deleterious effects of stress and corticosterone on hippocampal structure.[|100],[|101] Furthermore, increased NMDA and decreased AMPA receptor density have been reported in the hippocampus after exposure to stress.[|133–135] In parallel, NMDA-mediated synaptic responses were found to be increased after chronic stress.[|136]

Neurotrophic factors
Changes in neurotrophin levels have been hypothesized to play a key role in stress-induced neuronal damage. Hippocampal BDNF is reduced both by stress and glucocorticoid[|137] treatments. Conversely, fibroblast growth factor-2 (FGF-2) expression was shown to be increased after both stress and glucocorticoid treatments, which might represent a neuroprotective mechanism to preserve neuronal viability in challenging situations.[|129] Moreover, stress can influence intracellular transduction pathways involved in neurotrophin receptor signaling as shown for Ras-MAP kinase cascades[|138],[|139] that play critical roles in synaptic plasticity and neuronal survival. Chronically stressed rats also showed severe and lasting hyperphosphorylation of the extracellular signal-regulated kinases ERK1 and ERK2 involved in the Ras–MAP kinase pathway, along with a decrease in phospho-CREB expression in a number of areas including the hippocampus.[|138],[|140] Interestingly, phosphorylated CREB modulates the transcription of several genes that code for molecules involved in neuronal plasticity including tyrosine hydroxylase, BDNF, and NCAM.

Cell adhesion molecules
Chronic stress can markedly affect the expression of cell adhesion molecules in the hippocampus. Exposure of rats to chronic stress for 21 days has been reported to result in: Early postnatal stress was also reported to cause a profound reduction of NCAM expression in the hippocampus and cortex when the rats reached adulthood.[|91]
 * Reduced mRNA and protein expression NCAM in the hippocampus.[|24],[|68] Although the expression of the mRNA coding for the NCAM-180 isoform was not altered,[|68] chronic stress specifically reduced NCAM-140 protein expression.[|77],[|141] Moreover, a milder but widespread decrease in NCAM mRNA levels was observed across other brain areas.[|68]
 * Post-translational modification of NCAM with -2,8-linked polysialic acid (PSA) is also profoundly affected by chronic stress that increases its hippocampal expression[|24] in the dentate gyrus.[|142] In addition to its role in cell–cell de-adhesion, PSA-NCAM has been associated with newly generated cells[|143] since this post-translational modification of PSA-NCAM contributes to the migration of new progenitors and neurons. However, because chronic stress actually decreases cell proliferation in the dentate gyrus, the PSA-NCAM increase induced by stress cannot be attributed to a secondary effect on neurogenesis.[|141] Interestingly, the effects of stress on NCAM polysialylation are not restricted to the hippocampus. Chronic stress was also reported to enhance PSA-NCAM expression in the piriform cortex[|144] and reduce it in several amygdala nuclei.[|145]
 * Increased L1 mRNA and protein expression in the hippocampus[|24],[|68]. Like NCAM, L1 is another cell adhesion molecule of the immunoglobulin superfamily that has been largely implicated in synaptic plasticity and memory formation.[|50] Based on the neuroprotective effects of this molecule, a neuroprotective role has been hypothesized for the stress-induced increases of L1.[|4]

12.1.8 Stress and Aging
Aging is a period during which individual differences in cognitive abilities become larger, both in humans[|146–149] and rodents.[|150–152] Lifetime exposure to stress and the corresponding increases in glucocorticoid hormones have been proposed to be critical factors contributing to variability in the aging process.[|60],[|153–156] In particular, exposure to stress or high levels of glucocorticoids has been implicated in the acceleration and/or exacerbation of cognitive deficits in elderly subjects.[|14],[|59],[|60],[|154],[|157–159] Therefore, in addition to enhancing the magnitude of cognitive disturbances observed in aged individuals, stress may also accelerate their appearance. Aging is associated with higher basal cortisol levels[|160] and reduced feedback sensitivity of the HPA axis to pharmacological challenges.[|161],[|162] A role for stress and stress hormones in cognitive deficits at aging is also supported by the finding that rats classified as inferior (as opposed to good) learners when aged over 22 months showed both impaired memory and increased corticosterone levels.[|157],[|163],[|164] Moreover, hippocampal corticosteroid receptors have been also implicated in aging-associated increased glucocorticoid levels and the accompanying alterations on negative feedback regulation of the HPA axis.[|1],[|165] In most rat strains, aging has been linked to decreased MR binding and/or expression, with alterations in GR function being normally mild or nonexistent.[|166–169] In addition to the hippocampus, differences in GR expression were found in aged rats (24 months), depending on their capability to learn the water maze task.[|170] Specifically, old rats classified as superior learners had lower expression of GR mRNA in the parvocellular paraventricular nucleus of the hypothalamus than aged inferior learners. In parallel, aged inferior learners showed exaggerated stress-induced ACTH responses.[|170] As stated above, middle age seems to be a relevant time for stress and neuroendocrine interactions with the subsequent aging processes. Middle-aged rats (10 to 12 months old) were shown to be more vulnerable than younger rats to stress- or glucocorticoid-induced cognitive disturbances.[|84],[|85] Also, interfering with age-associated increases in corticosterone levels by submitting rats to adrenalectomies at 12 months was found to prevent age-related cognitive impairment (in reversal learning) as well as certain alterations in hippocampal structure.[|158] The importance of individual differences in the impact of stress experienced at mid-age on accelerating cognitive decline is illustrated in a recent study.[|86] Male rats were classified according to their locomotor reactivities to novelty as either highly reactive (HR) or low reactive (LR) as young adults and submitted to chronic stress (1 month) during mid-age (12 months). At early aging (18 months), their learning abilities were tested in the water maze and a number of neuroendocrine (plasma corticosterone, hippocampal corticosteroid receptors) and neurobiological (hippocampal expression of neuronal cell adhesion molecules) parameters were evaluated. Impaired learning was observed in stressed HR rats. Increased hippocampal mineralocorticoid receptors were found in stressed LR rats when compared with stressed HR and control LR groups. Moreover, mid-life stress induced an increased corticosterone response and a reduction in NCAM-180 isoform and L1 regardless of the behavioral trait of novelty reactivity. These findings support the view that stress experienced throughout life can contribute to cognitive impairment occurring during the early aging period. Likewise, evidence in aged humans also supports such a link among increasing glucocorticoid levels, memory deficits, and hippocampal atrophy.[|159] In particular, aged humans with significant prolonged cortisol elevations were found to display reduced hippocampal volumes and deficits in hippocampus-dependent memory tasks as compared to normal-cortisol controls.[|159] More recently, Wolf et al.[|171] reported that individuals who complain about memory impairments (in the absence of measurable impairments) have enhanced HPA axis activity as indicated by both higher basal cortisol levels and higher cortisol levels after dexamethasone.