Stress memory hippocampus

Behavioural Brain Research 127 (2001) 137– 158 www.elsevier.com/locate/bbr

Stress, memory, and the hippocampus: can’t live with it, can’t live without it Sonia J. Lupien a,b,*, Martin Lepage a a

b

Research Center, Douglas Hospital Research Center, Department of Psychiatry, McGill Uni!ersity, Verdun, Quebec, Canada H4H 1R3 Laboratory of Human Psychoneuroendocrine Research, Douglas Hospital Research Center, Montreal Geriatric Institute, 6875 Bld. Lasalle, Verdun, Quebec, Canada H4H 1R3 Received 30 April 2001; received in revised form 19 June 2001; accepted 25 July 2001

Abstract Since the 1968s discovery of receptors for stress hormones (corticosteroids) in the rodent hippocampus, a tremendous amount of data has been gathered on the specific and somewhat isolated role of the hippocampus in stress reactivity. The hippocampal sensitivity to stress has also been extended in order to explain the negative impact of stress and related stress hormones on animal and human cognitive function. As a consequence, a majority of studies now uses the stress– hippocampus link as a working hypothesis in setting up experimental protocols. However, in the last decade, new data were gathered showing that stress impacts on many cortical and subcortical brain structures other than the hippocampus. The goal of this paper is to summarize the four major arguments previously used in order to confirm the stress– hippocampus link, and to describe new data showing the implication of other brain regions for each of these previously used arguments. The conclusion of this analysis will be that scientists should gain from extending the impact of stress hormones to other brain regions, since hormonal functions on the brain are best explained by their modulatory role on various brain structures, rather than by their unique impact on one particular brain region. © 2001 Elsevier Science B.V. All rights reserved.

Keywords: Glucocorticoids; Hippocampus; Memory; Brain imaging; Human; Receptors

1. Introduction For many years, endocrinologists and neuroscientists thought that hormones, which are biological products secreted by peripheral glands, did not access the brain and acted mainly at the level of the peripheral nervous system. However, in the early 1960s, the discovery of neuropeptides as substances having not only classical endocrine effects, but also affected brain and behavior, significantly extended our view of hormones and opened the door to new possibilities of hormonal actions on the brain (for a complete historical background, see [38]). The idea of a central action of hormones was supported by previous studies showing that long-term therapy with anti-inflammatory drugs (which are synthetic hormones) led to significant mental * Corresponding author. Tel.: +1-514-761-6131x3359; fax: +1514-888-4064.

E-mail address: lupson@douglas.mcgill.ca (S.J. Lupien).

and cognitive deficits, named ‘steroid psychosis’ [25].

The presence of a mental disorder induced by exposure to high levels of a hormone strongly suggested that these substances could, in some way, access the brain and impact on affect and behavior.

The search for brain receptors able to recognize peripheral hormones was then opened. In 1968, it culminated with Bruce McEwen’s seminal Nature paper ([120]; see also [63]) showing that the rodent brain was indeed able to recognize hormones, particularly corticosteroids, which are hormones involved in the endocrine response to stress. The story then took a very important detour when McEwen and collaborators reported that the brain region showing the highest density of receptors for corticosteroids was the hippocampus, a brain region significantly involved in learning and memory [183]. As pointed out by de Kloet [38] in a recent review of the historical background of stress and the hippocampus, ‘‘[…] the important stress hormone retained in the hippocampus, an area with a critical

0166-4328/01/$ - see front matter © 2001 Elsevier Science B.V. All rights reserved.

PII: S 0 1 6 6 - 4 3 2 8 ( 0 1 ) 0 0 3 6 1 - 8

138

S.J. Lupien, M. Lepage / Beha!ioural Brain Research 127 (2001) 137–158

Fig. 1. Schematic representation of the HPA axis. Following the perception of a stressor, the hypothalamus releases CRF, which activates the pituitary and leads to secretion of ACTH. The levels of ACTH are detected by the adrenal cortex which then secretes corticosteroids. Corticosteroids will enter the blood circulation and act peripherally and centrally. At the level of the brain, both the frontal lobes and the hippocampus contain corticosteroid receptors, although the hippocampus contains both MRs and GRs, while frontal lobes contains mostly GRs. Both cortical structures inhibit HPA axis.

Along with inhibitory actions of the frontal and hippocampal region, circulating corticosteroids will act at the level of the hypothalamus to inhibit further secretion of CRF, and at the level of the pituitary to inhibit the secretion of ACTH. Circulating levels of ACTH also act at the level of the hypothalamus to inhibit secretion of CRF. The consequence of these negative feedback loops will be a significant decrease of circulating levels of corticosteroids after the end of the stressor.

This is thus a very exciting time for the field of psychoneuroendocrinology, which measures the impact of hormones on animal and brain function. Most studies performed today uses the stress–hippocampus link as the first premise in the establishment of a working hypothesis regarding the impact of stress and related stress hormones on animal and human behavior. Although this is surely a positive aspect of the research performed to this day, this fact can, however, have for consequence to bias the study design in such a way that only one brain region (i.e. the hippocampus) will be studied at the detriment of other brain regions that could also be involved in corticosteroid actions. Indeed, new data obtained within the last decade confirmed that corticosteroids impact on various brain regions and cognitive functions in both animal and human populations. By concentrating the totality of our efforts on the stress–hippocampus link, we as scientists might miss a great opportunity to explain the real actions of stress hormones on the brain.

As with most scientific data, one needs solid arguments in order to prove or disprove a point. The main goal of this review will thus be to use each of the four arguments for the stress–hippocampus link and report new data showing the involvement of regions other than the hippocampus for each of these arguments. The ultimate goal here will be to convince the reader that, contrary to many researchers working in the field of psychoneuroendocrinology, stress hormones may not necessarily have a preference for the hippocampus.

2. Stress mediators

function in cognition and affect. That could not be a coincidence!’’.

At this point, the stress– hippocampus link was born and it has been kept alive for the last three decades by a variety of research findings confirming the significant impact of stress hormones on hippocampal structure and/or function, and on animal and human learning and memory (for reviews, see [106,124– 126]). In general, four major arguments are used in order to confirm the stress–hippocampus link. The first one relates to the presence of glucocorticoid receptor (GR) in the animal and human hippocampus. The second one concerns the fact that high levels of stress hormones are associated with significant impairments in declarative memory function, a type of memory thought to be dependent upon adequate hippocampal function. The third one relates to the fact that chronic exposure to high levels of stress hormones is associated with atrophy of the hippocampus, and the fourth one concerns that fact that stress (and its related stress hormones) can impair neurogenesis in the hippocampus.

Stress is often cited as the cause of many psychological and physical problems. Many of us have experienced stress at one point or another of our life, and noted that stress can have important effects on our memory. One can have forgotten an important meeting or anniversary due to work overload, or else, one can have a vivid recollection of a car accident or any other stressful experience. Because of their impact on our lives, we have a tendency to pay more attention to the negative effects of stress on our memory, and to forget that under certain conditions, stress can also have a positive impact on physical and mental health (for a review, see [127]). In general, the stimuli that induce a stress response involve novelty and unpredictability, a lack of sense of control and threats to the sense of self [118].

The subjective experience that individuals describe as ‘stress’ does not always predict increased activity of the hormonal systems that are linked to stress, namely, the sympathetic nervous system and the hypothalamo–pi-

S.J. Lupien, M. Lepage / Beha!ioural Brain Research 127 (2001) 137–158

tuitary– adrenal (HPA) axis (see Fig. 1). The HPA axis, along with the sympathoadrenal system governs metabolic responses to the slings and arrows of everyday life, as well as to the beleaguering demands that prevail under conditions of chronic, severe stress. In response to a stress, hypothalamic neurons release corticotropin-releasing factors (CRF) and this triggers subsequent secretion and release of ACTH from the pituitary. In response to ACTH stimulation, corticosteroids are secreted and released from the adrenal glands. The responsivity of the HPA axis to stress is, in part, determined by the ability of corticosteroids to regulate ACTH and CRF release. Circulating corticosteroids feedback onto the pituitary and hypothalamus to inhibit the secretion of CRF and ACTH. In addition to pituitary and hypothalamic sites, there is now considerable evidence for the importance of the limbic system, particularly the hippocampus, as well as cortical areas, particularly the prefrontal cortex, in the regulation of HPA activity. This regulation is possible through an inhibitory role of these regions on HPA activity [54,55].

In most studies, the physiological assessment of stress in animal and human populations involved measurements of corticosteroids (cortisol in humans, and corticosterone in animals), and ACTH levels in blood or cortisol in saliva or urine. Activity of the sympathetic nervous system is generally assessed by measures of catecholamines or monitoring of heart rate and blood pressure. When one is measuring reactive hormonal actions in response to stress, it is important to consider the time of day of this assessment, given the fact that cortisol follows a circadian rhythm. In humans, under basal conditions, cortisol secretion exhibits a 24-h circadian profile in which cortisol concentrations exhibit a morning maximum, declining levels during the late morning and afternoon, a nocturnal period of low concentrations, and an abrupt elevation after the first few hours of sleep [208].

3. Arguments for a stress – hippocampus link

3.1. Glucocorticoid receptors in the hippocampus Today, there is abundant evidence for the importance of the hippocampus in feedback regulation of HPA function [82,123,169]. Following the discovery of corticosteroid receptors in the hippocampus of the rat in 1968 [120], a plethora of studies characterized the role of the hippocampus in HPA feedback regulation. In general, these studies have shown that lesions to the hippocampus result in elevated corticosterone levels under basal and post-stress conditions [57,170,218], whereas corticosteroid implants into the hippocampus normalize ACTH levels in adrenalectomized rats [14].

139

Hippocampectomized animals show reduced suppression of HPA activity following exogenous corticosteroid administration [55]. Finally, hippocampectomy results in increased CRH mRNA levels in the parvocellular region of the PVN (a region which contains neurons whose axons terminate in the median eminence; [80]), and increased portal concentrations of ACTH secretagogues [172].

However, one cannot appreciate the entire array of functions of corticosteroids in the brain without knowing more about the historical background surrounding the discovery of corticosteroid receptors in the rodent hippocampus. Here, history shows us that the majority of physiological and behavioral effects previously attributed to the impact of corticosteroids on the hippocampus might be actually sustained by a receptor type that is poorly distributed in the rodent hippocampus and even less present in the primate and human brain.

3.1.1. One hippocampus, one receptor type Before the mid-60s, the endocrine knowledge surrounding the HPA axis unanimously predicted that the hypothalamus should be the major site of the central action of corticosteroids, so that receptors for this hormone were expected to be solely localized in the CRF producing cells of the hypothalamus (for a complete review, see [38]). However, in their important Nature paper published in 1968 [120], McEwen and collaborators described the retention of corticosterone — a naturally occurring corticosteroids in the rodent brain—in the adrenalectomized rat brain. The rats were first adrenalectomized in order to deplete the rat’s system of any endogenous circulating corticosteroids, and then corticosterone was injected and retention of this naturally occurring corticosteroid was assessed. Using this method, McEwen and collaborators [120] showed that corticosterone was highly retained by the hippocampus, an area critically involved in learning and memory [183]. In a second set of experiments, McEwen and collaborators [122] compared the brain regional distribution of cell nuclear corticosteroids of adrenalectomized rats with corticosterone replacement, with that of intact rats with normal levels of the circulating hormones and tested either at 08:30 h, or 15 min after ether stress. The results revealed a pattern of distribution of cell nuclear corticosteroids in intact rats of both groups very similar to that seen in adrenalectomized rats with corticosterone replacement. In both cases, the highest concentrations of corticosteroid receptors were observed in the hippocampus.

In 1974, de Kloet and collaborators [35], assuming that a synthetic corticosteroid called dexamethasone would be even better retained by the hippocampus than the naturally occurring corticosterone, duplicated

140

S.J. Lupien, M. Lepage / Beha!ioural Brain Research 127 (2001) 137–158

McEwen’s study using dexamethasone and found out that this assumption was in fact incorrect. Indeed, dexamethasone was very poorly retained by the hippocampus of the rodent brain, and this, irrespective of the route of administration (peripheral vs. intracerebroventricular), suggesting that this compound does not readily access the brain. Later, these authors in collaboration with Bruce McEwen, showed that the small amounts of dexamethasone that penetrated the brain was retained in a regional pattern that was distinctively different from corticosterone [35,121]. Moreover, another steroid (cortisol) that is not a natural corticosteroid in the rat (although it is in humans) was very poorly retained in the rodent brain [121]. On the contrary, corticosterone as well as aldosterone (another naturally occurring corticosteroid in the rodent) were highly retained by the rodent hippocampus and surrounding limbic structures.

The different modes of action of dexamethasone and corticosterone on the rodent brain gave indications that these two compounds might bind to different types of corticosteroid receptors. This idea was confirmed 6 years later by Veldhuis and collaborators [210].

3.1.2. One hippocampus, two receptor types The discovery that two types of corticosteroid receptors were present in the rodent brain was made possible by the development of potent selective corticosteroid agonists and antagonists by Roussel-Uclaf (France) in the early 1980s. Using these new selective compounds, Veldhuis et al. [210] observed the presence of mineralocorticoid (MR) and glucocorticoid receptors in the rodent hippocampus. In 1985, Reul and De Kloet [158] showed that the tracer amounts of corticosterone that were previously retained so abundantly by the rodent hippocampus were actually bound to MRs, and not to GRs. In fact, these authors showed that affinity of hippocampal GRs for corticosterone in the rat brain was actually too low for any signal to be detected.

In summary, before 1985, the majority of corticosteroid effects in the hippocampus were entirely ascribed to GRs, while after 1985, the field had to re-establish its focus on both MRs and GRs in order to explain the impact of stress hormones on the rodent brain.

3.1.3. Distribution and affinity of MRs and GRs: ethological considerations The story of MR/GR distribution in the brain continued with the report by Sutanto et al. [198] showing that the retention of corticosteroids in the hippocampus depended on the species studied. In the rat, which is a corticosterone secreting animal, naturally occurring corticosterone will be highly distributed in the hippocampus (by binding to MRs and GRs), while other synthetic corticosteroids will not bind with high affinity to hippocampal MRs. However, in the hamster,

which is a cortisol secreting species just as humans, hippocampal MRs will display a higher affinity for cortisol, when compared with corticosterone, although the same compound (cortisol) will very poorly penetrate the rodent brain [199]. In the rodent brain, MRs are more heterogeneously expressed than GRs, with highest levels of expression in the hippocampus and surrounding limbic system, and certain brainstem motor nuclei.

GRs are widely expressed in most brain regions and are detected, among other brain regions, in the hippocampus, the paraventricular nucleus and other hypothalamic nuclei, the limbic system, the cerebral cortex, and most brainstem monoaminergic nuclei (for a review, see [158]).

Whereas the anatomic distribution of rat hippocampal MRs and GRs has been extensively described, very little is known about MRs and GRs distribution in the primate brain. However, in the mid-1980s, the cDNAs for human MRs and GRs was cloned, and it has been shown that each corticosteroid receptor exists in two isoforms [4,51,81,224], allowing for their structural and functional characterization [1,165]. Recent studies mapping both MRs and GRs distribution in the primate brain strongly suggest that extrapolation from rat brain to primate brain may be misleading when discussing the impact of MRs and/or GRs on the hippocampus.

Indeed, a recent study performed by Sanchez and collaborators [167] reported that, in contrast to its well established distribution in the rat brain, GR mRNA is only weakly detected in the dentate gyrus and Cornu Ammonis of the macaque hippocampus. In contrast, GR mRNA is strongly detected in the pituitary, cerebellum, hypothalamic paraventricular nucleus and prefrontal cortices. When studying the distribution of MRs in the macaque’s brain, Sanchez et al. [167] reported that MR mRNA was abundantly observed within the dentate gyrus and Cornu Ammonis of the hippocampus. Although previous studies have reported the presence of GRs in the hippocampus of humans [102,183–185,201,216], and non-human primates [85], these results were obtained by means of in situ hybridization. However, two recent abstracts summarizing studies using immunohistochemistry report that the primate’s hippocampus is not the major site of GR expression [99,150]. Although these results are interesting in line with previous data obtained by in situ hybridization, immunohistochemistry only reveals the presence of the GR proteins at the time of the experiment. It is thus possible that GR protein expression may change following a stressor. Also, it is important to note that immunohistochemistry is at best a semiquantitative method and absolute densities of GR receptors cannot be estimated with this technique.

Indirect results obtained from human populations also suggest that GR expression might not be as important as previously thought in the human hippocampus. For

S.J. Lupien, M. Lepage / Beha!ioural Brain Research 127 (2001) 137–158

example, the down-regulation of hippocampal GRs that should be observed in patients showing chronic hypercortisolemia is not reported [185], and an up-regulation of GRs is actually reported in Alzheimer’s patients with high corticosteroid levels [216].

3.1.4. Corticosteroid receptors in the hippocampus— conclusion The results summarized above with regard to the search of corticosteroid receptors in the brain of various species show that although such receptors have been originally found in the rodent hippocampus, the characterization of the different distribution and affinity of two corticosteroid receptor types involves a more complex network of corticosteroid actions on the brain. First, when studying the impact of stress hormones on the brain, one has to consider the type of corticosteroid receptors that might be involved, in line with its preferential distribution and affinity in the brain. Second, when studying the effects of corticosteroids on the brain, one has to take into account the important species differences reported for the distribution of corticosteroid receptors in the brain. In this sense, extrapolation of rodent data to human populations might be misleading.

More importantly, the observation that MRs are widely distributed in the rodent, primate, and human brain while GRs are weakly distributed in the primate and human brain has another serious implication for the stress– hippocampus link theory. Indeed, as will be summarized in the next section, most of the negative effects of stress (and related corticosteroids) on animal and human cognitive function have been related to GR actions. The relative absence of such receptor type in the hippocampus either suggests that the negative effects of stress on cognitive function are related to hippocampal MRs, or that these negative effects are related to the isolated and/or combined actions of MRs and GRs on various brain regions. These two issues will now be discussed.

3.2. Corticosteroid effects on cognition Since the seminal work performed by Scoville and Milner [183] with amnesic patients having undergone a bilateral ablation of the hippocampus, it has been known that this structure serves a critical role in memory formation. Given the presence of corticosteroid receptors in the rodent and human hippocampus, it has been suggested that the corticosteroid modulation of hippocampal activity may underlie some aspects of the acute and/or chronic effects of corticosteroids observed in animal learning and memory processes. The effects of adrenal steroids on animal cognition and its neural substrate have been studied using, for the most part, three types of models that tap into hippocampal func-

141

tion. The first approach is to examine the neuroendocrine modulation of a physiological model of neuronal excitability that is relevant to memory.

Hippocampal long-term potentiation (LTP) describes a long-lasting enhancement in synaptic efficacy that occurs in response to high-frequency electrical stimulation [110,205]. LTP shares many characteristics in common with memory, the most important being its rapid induction and its long duration. The second approach is through the measure of associative learning, as defined by various aspects of conditioning behaviors. The third approach is through the study of spatial memory [84,149].

In contrast, the effects of corticosteroids on human cognition have been studied using, for the most part, measures of declarative and non-declarative memory.

The logic for the inclusion of this amnesic dissociation is due to the fact that studies report that the hippocampus is essential for declarative memory, while it is not essential for non-declarative memory [191]. Declarative memory refers to conscious or voluntary recollection of previous information, whereas non-declarative memory refers to the fact that experience changes the facility for recollection of previous information without affording conscious access to it (priming). Thus, this somewhat specialized role of the hippocampus serves as the basis for specific hypotheses regarding the relationship between increased cortisol secretion and impaired cognitive function in humans.

3.2.1. Animal studies A number of studies have reported that the induction of LTP in the hippocampus is blocked by the administration of corticosterone [47,56]. The role of corticosteroids in hippocampal LTP have further been confirmed by studies showing that the acute administration of corticosteroids in the dentate gyrus of the hippocampus produces LTP [56,152]. In 1991, Bennett and collaborators [8] reported the existence of a negative correlation between the magnitude of LTP in the CA1 population spike in the hippocampus and the level of circulating corticosteroids, thus suggesting a dose-dependent relationship between corticosteroids and their detrimental effects on LTP. One year later, Diamond and collaborators [42] reported the presence of an inverted-U shape relationship between the level of circulating corticosteroids, and LTP. They described a positive correlation between corticosterone and primed burst potentiation (PBP; which is a low threshold form of LTP; [8]) at low levels of corticosteroids and a negative correlation between corticosterone and PBP at high levels of corticosteroids. These results provided a strong support for the hypothesis that corticosteroids exert a concentration-dependent biphasic influence on LTP.

142

S.J. Lupien, M. Lepage / Beha!ioural Brain Research 127 (2001) 137–158

Direct effects of corticosteroids on associative learning paradigms have also been described and the majority of them revealed modulatory effects of corticosteroids on animal cognition. For example, acute administration of either corticosterone or dexamethasone accelerates the rate of extinction of a shock avoidance response [10,11,70,162]. Similarly to studies performed on LTP, it has been shown that the effects of corticosteroids on animal cognition follow an invertedU shape relationship [93]. In 1976, Kovacs and collaborators [93] reported that low doses of corticosterone facilitate extinction of an avoidance response, while high doses of corticosterone delay the rate of extinction of the conditioned response. Finally, biphasic modulatory effects of corticosteroids were also reported using spatial memory paradigms (see [105]). Altogether, these results (obtained mainly in the rodent population) confirmed the important role of the hippocampus in explaining corticosteroid-induced cognitive impairments.

3.2.2. Human studies In humans, the effects of corticosteroids on cognitive function have been measured using mainly cognitive tasks assessing declarative memory, which is a type of memory thought to be sustained by the hippocampus.

In general, declarative memory function can be assessed using tasks involving a conscious recollection of previously learned information, as in free or cued recall of material learned before.

In general, the majority of human studies that have measured the impact of corticosteroids on cognitive function report impaired declarative memory function after acute and/or chronic administration of synthetic corticosteroids (for a complete review, see [105]). Interestingly, the first study performed on the acute effects of corticosteroids on human memory process was a dose –response study. In 1986, Beckwith and collaborators [7] showed that the effects of hydrocortisone on human memory performance depend upon the dose administered. Only the highest doses of hydrocortisone (40 mg) enhanced recall when subjects were presented with more lists of words. In 1993, Fehm-Wolfsdorf and collaborators [53] reported that hydrocortisone administration in the morning (at the time of cortisol peak) impaired declarative memory function, while it had no effect on cognitive performance when administered at night. In 1996, Kirschbaum and collaborators [89] showed that the oral administration of 10 mg of hydrocortisone leads to a significant decrease in memory performance as tested 60 min after hydrocortisone intake. The results showed that subjects who received hydrocortisone treatment presented an impaired performance in the declarative memory task but not in the non-declarative memory task, thus suggesting that cortisol interacts with hippocampal neurons to induce cognitive deficits.

Besides acute actions, delayed effects of corticosteroids were reported in memory studies on human subjects. In 1990, Wolkowitz and collaborators [219,220] observed impaired memory performance in normal adults following 5 days administration of high doses of prednisone (80 mg p.o. daily), but normal memory performance in another group of subjects following a more acute administration of 1 mg of dexamethasone. In 1994, Newcomer and collaborators [143], using a 4 days administration procedure with 0.5, 1, 1, 1 mg per day of dexamethasone in normal controls, reported impaired declarative memory performance (acquisition and recall) on the fourth day of treatment only. No immediate or delayed effects of dexamethasone were observed on non-declarative memory, and selective attention performance. These results were in accordance with a hippocampal involvement in corticosteroid-related cognitive deficits and argued against a non-specific effect of the steroid on attention and arousal. The same year, our lab reported that 4-year exposure to high endogenous levels of cortisol in elderly humans led to significant declarative memory impairments, without impairing non-declarative memory function [104].

In summary, the majority of studies performed in human populations tend to confirm the rodent literature reporting negative effects of corticosteroids on hippocampal-dependent forms of memory. Altogether, the rodent and human data strengthened the view that stress hormones have a specific and isolated impact on the hippocampus.

3.2.3. Problem c 1: memory does not equal hippocampus in humans Even if both animal and human studies have confirmed the important impact of corticosteroid hormones on cognitive function, there are still many flaws to the conclusions that have been reached to this point. First, one has to consider the fact that tasks that are ethologically relevant to animals might not be as relevant in human individuals (for a complete review, see [200]).

Indeed, although hippocampal-dependent tasks in rodents have been confirmed using lesion studies, one has to keep in mind the important phylogenetic differences between development of the rodent and human brain.

Subcortical structures tend to have a more important role to play in cognitive function in lower species such as the rodent, while in primate and human populations, development of cortical areas leads to a stronger involvement of these regions in cognitive function. Also, studies have shown that the more neuroethologically relevant the spatial tasks are to the particular animal (here, the rat), the more the hippocampus can be shown to be prominent and to play an important role in these tasks [200]. Given this later fact, it is thus possible that a conclusion reached in the rodent population with

S.J. Lupien, M. Lepage / Beha!ioural Brain Research 127 (2001) 137–158

regard to the impact of corticosteroid on hippocampaldependent memory task might not be confirmed in human individuals, when submitted to so-called ‘declarative memory tasks’.

The second point relates directly to the measure of memory function in both animals and humans. In all species, memory lends itself to study through its retrieval, whether it is evaluated by the behavior of a rat in a swimming pool, or by a verbal report in a human subject. As summarized by William James in 1890 [83], ‘‘the only proof of there being retention is that recall actually takes place’’. In many psychoneuroendocrine studies performed in rodents and humans, a deficit in recall is often related to a hippocampal-dependent form of memory impairment. By doing so, researchers consider memory as a single entity that is mainly localized in the hippocampus. However, new studies performed in human populations and using in vivo imaging of brain function (functional neuroimaging technology) under various memory tasks now consistently report that memory is not a single entity. The success of the functional neuroimaging approach is partly imputable to the extensive use of concepts borrowed from the cognitive psychology of human memory. Because researchers have examined memory not as a single entity, but instead as involving multiple and separable cognitive processes such as encoding and retrieval just to name a few, this has favored the observation of consistent patterns of activation. A direct consequence is that we now know that when a person encodes or retrieve from memory a specific information, in addition to the hippocampus, several brain regions exhibit increased activity and the location of such increases is often dependent on the memory process examined. In order to extend our understanding of the impact of corticosteroids on human cognitive function, we will describe recent human data obtained in brain imaging studies which specify the role of the hippocampus and other brain structures during memory encoding and retrieval.

This is not intended as an exhaustive review of the literature as several have been published recently (see [30,96,180]).

3.2.3.1. Memory encoding Item encoding. Several PET and fMRI studies have reported activations in or near the hippocampal region during the encoding of information in memory. Memory encoding is a process that can be further divided into different subprocesses, some of which have also been linked to hippocampal activity. One of the first role ascribed to the hippocampus had to do with ‘item’ encoding. For example, PET and fMRI studies have shown that the hippocampus and adjacent cortex participate in encoding of information about faces but not about its retrieval [73]. The hippocampal region also encodes meaningful actions [34] relative to meaningless

143

ones. Other studies have reported hippocampal region activity during the encoding of visual objects [160,217], visual patterns [217], and deeply processed words [207].

No!elty encoding. In a similar vein, other early functional neuroimaging studies of memory suggested a role of the hippocampal region during novelty encoding/assessment [72,199]. This hypothesis states that a fundamental component of memory encoding is discriminating whether stimuli are novel or familiar, with the increased likelihood of novel stimuli to be further encoded. Several researchers have examined the involvement of the hippocampal formation during novelty assessment. For example, Tulving and collaborators [202,203] showed greater hippocampal activity during a recognition task in which subjects had to decide whether a presented item was ‘new’ (never presented before) as opposed to ‘old’ (previously studied item). Similarly, Stern et al. [197] observed greater hippocampal activation when subjects encoded new pictures than when they encoded a single picture repeatedly. Dolan and Fletcher [44] found greater hippocampal activations when both members of category-exemplar word pairs were new than when either the category or exemplar or both had been previously presented to subjects. Kopelman et al. [92] also observed greater hippocampal and parahippocampal activation, among other regions, during encoding of novel words when compared with encoding of familiar words and similar findings were observed using visual scenes [131]. Finally, Martin et al. [117] found greater right hippocampal activation when a task was being performed for the first time (novel) than when it was being performed for the second time (familiar), even though the stimuli differed between the first and second trial of the task. This possible involvement of the hippocampus in detection of novelty is to be put in relation with previous data showing that novelty is a potent inducer of stress reactivity in both animals and humans [118].

Associati!e encoding. Other functional neuroimaging studies have specifically investigated the neural correlates of associative encoding of verbal [77,87,98,139, 159,161,207], and non-verbal stimuli [76,138]. For instance, Henke et al. [76,77] contrasted neural activity induced by an associative semantic encoding task to that of a single-item encoding task for similar materials and reported greater anterior hippocampal region activity during the former. These studies have collectively identified brain regions that appear to make a significant contribution to processes involved in association formation. A common observation across these different studies is the activation of the hippocampal region during associative encoding [30]. Evidence in animal neuroscience also suggests a significant role of the hippocampal region in associative or relational processing [29,48].

144

S.J. Lupien, M. Lepage / Beha!ioural Brain Research 127 (2001) 137–158

Successful encoding. Two recent fMRI studies [18,212] have presented evidence for a role of hippocampal region in successful memory encoding.

These two studies examined neural activity during a memory encoding task consisting of the visual presentation of words [212] or pictures of landscapes [18].

Following the encoding task, subjects were administered a memory recognition test composed of items previously studied of never studied before. Based on the behavioral performance during this recognition task, an analysis of the fMRI data examined the differential activity for encoding items that were subsequently successfully recognized versus those items that were subsequently forgotten. This analysis identified brain regions where neural activity was greater during ‘successful’ encoding. Both studies reported greater activity for successfully encoded items relative to forgotten items in the parahippocampal region. In addition, in the Wagner et al., study which examined verbal material, this difference was observed in the left hippocampal region (parahippocampal gyrus) whereas in the Brewer et al.

[18] study which examined figural information, this difference was observed in the hippocampal region (parahippocampal gyrus) bilaterally.

While functional neuroimaging techniques can provide information about the contribution of particular brain region such as the hippocampus to a specific memory process, it can also provide valuable information concerning changes in neural activity in all areas of the brain. Perhaps we can understand better the role of the hippocampal region during a specific memory task/ process if we are to examine systematically what ‘happens’ elsewhere in the brain. McIntosh [128,129] has termed such an approach ‘neural context’ which simply put, suggests that a given brain region may be engaged in a specific process depending on activity in other brain regions. In other words, a mental set is not localizable to a given brain region but is instead represented in the activity of networks of brain regions.

Therefore, it is conceivable that a brain structure such as the hippocampus may contribute to different memory processes depending on the synchronous activity in other brain regions.

As we have seen earlier, functional neuroimaging studies that have examined associative encoding or relational processing have produced some of the most exciting and reliable results in brain imaging research of memory to date. From these PET and fMRI studies, one common finding is the observation of hippocampal activation in an experimental condition involving associative encoding compared with a baseline condition.

This has been taken as a strong evidence for a role of this brain region in relational processing [30].

In a recent PET study, Lepage and collaborators [98] examined associative encoding and retrieval for verbal material. In this study, three associative encoding con-

ditions differing in the number of words (0, 1, or 2) semantically related to a third word representing the name of a semantic category were examined. For example, the category could be ‘FRUITS’ and the exemplars ‘BANANA’ and ‘ORANGE’ A recall task consisting in the presentation of the category names (e.g. ‘FRUITS’) as cues for retrieving the other two members of the triads followed each encoding condition. A multivariate analysis (partial least squares) of the PET data identified task-related patterns of activity distinguishing the three associative encoding conditions from the three cued-recall conditions. During associative encoding, relative activation was observed in the anterior hippocampal region bilaterally, in the inferotemporal gyrus bilaterally, left prefrontal cortex and in the fusiform gyrus bilaterally.

The observation of hippocampal activation was certainly consistent with existing data but these authors were interested to know whether brain regions other than the hippocampus would exhibit the same consistent and reliable activation during associative encoding across the three conditions. They used these regions just described as a starting point to review the functional neuroimaging literature on association encoding. They found eight other functional neuroimaging studies of associative encoding [76,77,87,138,139,159,161,207].

Activation in four of these regions has been consistently reported in previous studies of associative encoding using PET, fMRI, and SPECT. These included the hippocampal region, but also the left fusiform gyrus, the left prefrontal cortex and the inferior temporal gyrus bilaterally. These studies have compared neural activity during an associative encoding task to all kinds of control tasks and have used different kinds of materials including words, faces, houses, landscapes and line drawings of objects.

The general picture that emerges is that hippocampal region and left fusiform gyrus activations constitute the most robust findings, observed in eight out of nine studies including our PET study. Left inferior prefrontal cortex and inferior temporal cortex have been reported in seven out of nine studies. The consistent observation of increased activation in these four regions is taken as strong evidence of their participation in associative encoding. This finding suggests that in addition to the hippocampal region, other brain regions make a unique contribution to the formation of association in memory. Perhaps these different regions play a different role in associative encoding. Studies that manipulate specific aspect or processes involved in associative encoding, such as the ease of association formation or inter-modal associations will surely provide interesting data on this problem.

3.2.3.2. Memory retrie!al. While many studies have proposed a role for the hippocampus during the encod-

S.J. Lupien, M. Lepage / Beha!ioural Brain Research 127 (2001) 137–158

ing of information in memory, several other experiments have produced data suggestive of a role of this region during memory retrieval. As with encoding studies, the first wave of functional neuroimaging studies of memory retrieval did not systematically manipulate specific processes. Instead, neural activity observed during an episodic (declarative) memory retrieval task was contrasted to that of a task not involving the retrieval of information from memory but otherwise, similar in terms of perceptual processing and response generation.

For instance, Squire et al. [192] contrasted a stem cued recall task (a declarative retrieval task) to a stem completion task (a non-declarative retrieval task), and observed greater hippocampal activity in the former.

Rugg et al. [164] reported greater left hippocampus activation during the retrieval of deeply encoded words relative to the retrieval of shallowly encoded words.

Kapur et al. [86] observed greater hippocampal activity during the recognition of previously studied faces compared with a resting condition, whereas Schacter et al.

[179] reported similar hippocampal activation during the recognition of drawings of three-dimensional objects.

These reports of hippocampal activation during memory retrieval are certainly interesting, but they say little about the underlying processes subserved by the hippocampus during recall and recognition tasks.

Schacter et al. [178] compared two memory recall conditions, one in which recall performance was high and a second one in which performance was low. Greater hippocampal activity was observed during the high performance condition relative to the low performance one. This finding is particularly interesting since both conditions involve retrieval attempts but one condition is more ‘successful’ than the other one. These findings prompted Schacter and colleagues to suggest that the hippocampal region is involved in the actual conscious recollection of previously studied events but not in the effort to do so. Another PET study based on this experimental design yielded similar results [75] whereas other studies that have looked at different levels of performance between cued-recall tasks have failed to reproduce these findings [163].

A recent fMRI study also suggests a role for the hippocampal region in the subjective (conscious) experience associated with the act of retrieving information from memory [49]. In this fMRI study, subjects indicated during a recognition task whether for each item previously categorized as ‘old’, there was a feeling of recollection (remember judgments) or a feeling of familiarity (know judgments). Relative to new items and ‘know’ judgments, ‘remember’ judgments led to a significant increase in activity in the hippocampal region among other regions. Similarly, Henson et al. [78] using an even-related fMRI design observed greater posterior hippocampal region activation during the recognition

145

of previously studied words associated with a ‘remember’ judgment relative to the presentation of new (never studied before) words.

Although hippocampal activations have been reported during memory retrieval, brain pattern of activity during recognition is quite different from that observed during memory encoding [20,41,145,166,204].

In a recent paper using a meta-analysis of various studies measuring brain activation during memory retrieval, Lepage et al., 2000 [97] have provided evidence for the involvement of frontal lobes in memory retrieval. In order to do so, these authors combined data from four previous PET studies [45,86,146,147] whose designs included the requisite conditions. The purpose was to get robust and stable results since the analysis would be done on data from 53 subjects, something that one rarely sees in functional neuroimaging studies.

The findings pointed to six specific cortical regions at which recognition testing of old items exhibited as much differential activation as did testing of new items compared with the control condition. One of these sites was in the anterior cingulate. The other five were all in prefrontal cortex, and included two homotopic sites situated bilaterally and symmetrically, one in anterior prefrontal cortex (BA 10), the other near the opercular region (BA 47/45). The fifth one was a right dorsal prefrontal site near BA 8/9. These activations were stronger and had a greater spatial extent in the right hemisphere. From these results, it was concluded that these six retrieval mode sites are involved in the establishment and maintenance of a brain state that subserves the cognitive set of episodic memory retrieval mode. It is interesting to note that such retrieval mode activity was observed only in the frontal lobes and not in posterior brain regions.

Moreover, other functional neuroimaging studies have shown that under certain conditions, old and new items can lead to differential activation with left prefrontal regions exhibiting greater activity for old (familiar) items [72,91]. We refer to such findings as item-related activity because they are attributable to the nature of the item while the task requirements are kept constant. In addition, other item-related retrieval processes have been proposed and include post-retrieval processing [163], monitoring [79], and retrieval effort [178]. A recent study by McDermott et al. [119] has reported fMRI data in which activity for old items were associated with greater activity in anterior prefrontal and right frontal opercular region.

Taken together, these studies make it increasingly clear that recognition memory relies heavily on prefrontal cortex, especially in the right hemisphere. It has been hypothesized that the hippocampal region is related to conscious recollection during recognition memory [49]. Since prefrontal cortex seems also involved in memory retrieval, with greater activity in several re-

146

S.J. Lupien, M. Lepage / Beha!ioural Brain Research 127 (2001) 137–158

gions greater for remembered items than for new items that cannot induce remembering, it seems difficult to disentangle the contribution of prefrontal and hippocampal regions to such retrieval process. At the same time, these results suggest that memory retrieval involves several brain regions, some of which appears to make a task-related contribution and others an item-related contribution to memory retrieval. Therefore, if a region of prefrontal cortex is from an hemodynamic point of view ‘behaving’ in the same way as a region of the hippocampal region then it cannot be ruled out that they are contributing to the same process.

In summary, recent contributions from brain imaging studies revealed that although the human hippocampus is certainly involved in some facets of memory encoding and retrieval, it is certainly not the only brain region involved in these processes. Multiple cortical and subcortical regions show increased activity during memory encoding and retrieval tasks and these findings are not limited to humans. For instance, a recent functional brain imaging study using 2-deoxyglucose autoradiography in mice has reported increased activity in multiple neocortical areas, notably in the frontal cortex, during a spatial memory retrieval task [13]. In consequence, taking recall performance as the sole measure of ‘hippocampal-dependent memory function’ in human populations when measuring the impact of corticosteroids might lead to significant errors in interpretation of cognitive results. Such discrepancies have been reported in human studies in which exogenous or endogenous increases in corticosteroids has been shown to impair either encoding [89,106,142,144] or retrieval [39]. Knowing the important involvement of the frontal lobes in retrieval of information (see previous section), one can ask whether the corticosteroid-induced memory retrieval impairments observed in human populations could in fact be better explained by the impact of corticosteroid receptors activation on the frontal lobes. The next section describing the differential effects of MRs and GRs activation on animal and human cognitive function will shed light on this question.

3.2.4. Problem c 2: differential in!ol!ement of MRs and GRs in corticosteroid-induced memory changes As we have summarized above, even if the majority of human psychoneuroendocrine studies reported corticosteroid-induced declarative memory impairments, recent brain imaging studies now show that one cannot interpret a deficit in declarative recall performance as being solely related to hippocampal function since (a) the hippocampus has been shown to be involved in memory encoding and (b) other brain regions actively participate in both memory encoding and retrieval, sometimes to a larger extent than the hippocampus per see.

The second problem that emerges with regard to interpretations of corticosteroid-induced declarative memory impairments relates to the differential distribution and affinity of MRs and GRs within the rodent and human brain. Although both receptor types have been implicated in mediating corticosteroid feedback effects (see [158]), there are two major differences between MRs and GRs that have to be taken into account when discussing the effects of corticosteroids on human cognition. First, MRs bind corticosteroids with an affinity that is about six to ten times higher than that of GRs. This differential affinity results in a striking difference in occupation of the two receptor types under different conditions and time of day. Low, basal corticosteroids levels observed during non-stressed periods or PM phase in humans serve to activate largely MRs, whereas the elevated corticosteroids levels characteristic of periods of stress or AM phase in humans activate both MRs and GRs [158]. The second major difference between these two receptor types is related to their distribution throughout the brain. MRs are present exclusively in the limbic system, with a preferential distribution in the hippocampus, parahippocampal gyrus, entorhinal and insular cortices. On the contrary, GRs are present in both subcortical (paraventricular nucleus and other hypothalamic nuclei, the hippocampus and parahippocampal gyrus) and cortical structures, with a preferential distribution in the prefrontal cortex [43,120,123,130,176,177]. Because of the high affinity of hippocampal MRs for corticosteroids, these receptors are thought to mainly control low basal circadian levels of circulating corticosteroids, and to enhance adrenocortical secretion following stress (tonic influence; [158]). In contrast, the low affinity GRs terminate high stress-induced levels of corticosteroids (dynamic influence; [158]).

To this day, the only consistent finding that has emerged from the literature concerned with the acute impact of corticosteroids on animal and human cognitive function is that an inverted-U shape relationship exists between the dose of corticosteroids acutely administered, and its impact on cognitive performance (for a review, see [105]). There have been two explanations for the existence of this inverted-U shape function. The first one relates to the exclusive activation of GRs [31,153], while the second one relates to a balance of MR/GR activation [37].

3.2.4.1. In!erted-U shape function and GR acti!ation.

This hypothesis of corticosteroid actions is mainly based on human studies measuring the effects of supraphysiological increases (stress and/or exogenous administration) of corticosteroids on cognitive function. In the majority of these studies, corticosteroids [39,58,143,144,182] and stress [89,106] have been shown to impair cognitive function, although this has not been

S.J. Lupien, M. Lepage / Beha!ioural Brain Research 127 (2001) 137–158

confirmed in other studies [193,209]. In most of these studies, the implicit assumption made about the impact of corticosteroids on human memory is that saturation of GRs, by acute elevations of corticosteroids, impairs hippocampal-dependent forms of memory.

Although this is certainly a very interesting hypothesis that has been validated using various protocols, there are still two main problems with the hypothesis as it is stated now. The first one relates to the differential distribution of MRs and GRs in the primate and human brain, as opposed to the rodent brain. The second relates to the type of neuroendocrine protocols used in animal versus human studies, and the different conclusions that are reached using these different protocols.

In primates and humans, as opposed to rodent, GRs have a preferential distribution in cortical areas, particularly in the prefrontal cortex [167,176], while the hippocampus contains high levels of MRs, with low levels of GRs. Today, there is evidence for a high density of corticosteroid receptors in the prefrontal cortex of both the rat [123,130], and human [177].

Binding studies in rats reveal a high retention of 3HCORT in the cortex, particularly in the medial prefrontal regions [43]. Further studies in rats [5,55,137], and humans [141,189] show that the prefrontal cortex is a significant target for the negative-feedback actions of circulating corticosteroids [55,137], which suggests that this area could play a significant role in the acute effects of corticosteroids on cognitive function.

Thus, if one attributes the negative impact of corticosteroids on human memory as being related to activation of GRs, then one has to take into account this preferential distribution of GRs in the prefrontal cortex and postulate that in humans, cognitive function preferentially sustained by prefrontal regions should be more sensitive to acute increases of corticosteroids, when compared with cognitive function preferentially sustained by the hippocampus.

In primates and humans, prefrontal cortex has been shown to be a key structure (without being the sole region involved) in working memory (WM) function [64,65,151,154] while the hippocampus is a key structure (again, without being the sole region involved) for declarative memory function [183,191]. As we have previously discussed, declarative memory refers to the conscious and voluntary recollection of previously learned information, while WM serves to store and manipulate information within a short period of time until it is transferred into long-term memory.

Two recent studies performed in humans report that WM is more sensitive than declarative memory to acute and short-term (10 days) increases in corticosteroids.

Young and collaborators [223] administered 20 mg hydrocortisone for 10 days to young normal male volunteers and measured various cognitive functions in

147

a randomized, placebo control, crossover, within-subject design. They showed that this regimen of hydrocortisone led to deficits in cognitive function sensitive to frontal lobe dysfunction, while it did not impact on cognitive function sensitive to hippocampal damage.

Similar results were obtained by our group [108] using an acute dose–response neuroendocrine protocol. By administering three different doses of hydrocortisone or placebo in the morning (AM Phase; saturation of MRs, partial occupancy of GRs), we were able to measure the curve function relating GRs occupancy and cognitive performance in humans. In our study, 40 young subjects were infused for 100 min with either hydrocortisone or placebo and memory function was tested during the infusion period. The results revealed that performance on the WM task decreased significantly whereas performance on the declarative memory task remained the same following an acute elevation of corticosteroids. Curve fit estimations revealed the existence of a significant quadratic function (U-shape curve) between performance on the WM task and changes in cortisol levels after hydrocortisone infusion.

The results of these two studies suggest that in young individuals, WM is more sensitive than declarative memory to an acute elevation of corticosteroids. They also suggested that this effect is significantly related to GRs occupancy/sensitivity within the prefrontal cortex.

The presence of a GR-induced cognitive impairment for prefrontal but not hippocampal function might at first sight seem puzzling since the hippocampus also contains GRs. However, animal studies have shown that the presence of MRs in the hippocampus acts by creating a physiological balance of both types of receptors for their action on the HPA axis (called the ‘binary hormone response system’ by Evans and Arriza [52] and the ‘MR/GR balance hypothesis’ by Oitz et al.

[148] and reviewed by De Kloet et al. [37]). This suggests that the presence of MRs within a structure acts by decreasing GR responsivity to corticosteroids because of the tonic influence of MRs on the HPA axis [148]. This suggestion implies that the absence of the tonic influence of MRs (e.g. in prefrontal regions) would increase GRs sensitivity and lead to increased sensitivity of prefrontal regions to acute changes in corticosteroids levels, when compared with the hippocampus.

3.2.4.2. In!erted-U shape function and MR/GR balance.

Although the majority of studies performed in human populations still report negative effects of corticosteroids on cognitive function (for a complete review, see [105]), various studies performed in rodents report a positive impact of corticosteroids on learning and memory [12,36,132,133,135,211]. One possible reason for such a discrepancy between rodent and human studies was recently discussed by De Kloet and collaborators

148

S.J. Lupien, M. Lepage / Beha!ioural Brain Research 127 (2001) 137–158

[37] as being related to the fact that every human study performed to this day has measured the impact of supra-physiological increases of corticosteroids on learning and memory, while rodent studies have used hormone replacement protocols.

In a hormone-replacement protocol, the behavior resulting from the absence of the hormone of interest is first measured, and then, baseline hormonal levels are restored to normal values and the same behavior is measured once again. It is postulated that if the hormone of interest has a real impact on the behavior tested, then this behavior should be restored to normal value after hormonal replacement (see [19]). In studies of the impact of corticosteroids on cognitive function, the reason why hormone replacement protocols might permit to assess the positive impact of corticosteroids on cognitive function relates directly to the differential involvement of MRs and GRs in corticosteroid-induced cognitive changes.

Remember that given their differential affinity for corticosteroids, MRs will always have to be totally saturated (100% activated) before GRs can start to be activated. During periods of stress or high levels of corticosteroids, both MRs and GRs will be saturated.

Many studies performed in rodents have shown that the ratio of MR/GR occupation is the major determinant of the direction of GR-induced cognitive changes (for a review, see [37]). For example, LTP has been reported to be at optimal level when corticosteroids levels are middly elevated, i.e. when the ratio of MR/GR occupation is high (i.e. MRs are saturated (100%) while GRs are only partially occupied (50%). In contrast, significant decreases in LTP are observed after adrenalectomy, when the ratio of MR/GR is very low (absence of MRs or GRs occupancy). The same negative impact of corticosteroids on LTP is reported after exogenous administration of supra-physiological doses of synthetic corticosteroids, which saturate both MRs and GRs, leading again to a low MR/GR ratio. It has thus been suggested that the negative view of corticosteroids actions on human cognitive function can be partly explained by limitations in previous human experimental designs, which did not allow to differentially manipulate MR and GR levels, as can be done in rodent hormone replacement protocols.

In order to test this suggestion, we have performed a hormone replacement study in a population of young normal controls [109]. In this protocol, we used a within-subject double-blind experimental protocol in which we first induced a chemical lowering of corticosteroids levels by administration of metyrapone, a potent inhibitor of corticosteroids synthesis, and then restored baseline circulating corticosteroid levels with subsequent infusion of hydrocortisone. Memory performance of participants under each of these conditions was compared with that measured on a placebo day.

The results showed that, when compared with placebo, the pharmacological decrease of circulating levels of corticosteroids induced by metyrapone significantly impaired delayed memory performance. Most importantly, we showed that this impairment was completely restored after hydrocortisone replacement. These results are the first obtained in a human population showing that corticosteroids can modulate memory function. We have suggested that this modulation can happen through a differential activation of MRs and GRs. Indeed, during the metyrapone condition, both MRs and GRs were unoccupied, resulting in a low MR/GR ratio and impaired memory performance. On the contrary, during the hydrocortisone replacement condition, cortisol levels were restored to the levels typical of those measured in the AM phase, i.e. leading to a saturation of MRs, with partial occupancy of GRs. This differential occupation thus led to an increased MR/GR ratio, and a restoration of baseline cognitive performance. In addition to confirming the positive impact of corticosteroids in human cognitive function, these results provided preliminary evidence that a threshold for corticosteroid-induced cognitive impairments might exist, and might be related to the balance in MR/GR occupancy.

As one can see from the above discussion of the differential impact of MRs and GRs on cognitive function, it is difficult at this point of research to disentangle the impact of MR/GR balance versus GR saturation on human cognitive function. However, it is becoming clear from this set of data that the actions of corticosteroids on the brain are less specific and isolated to the hippocampus than was previously thought. Hormones such as corticosteroids are usually thought to be modulators of behaviors, rather than inducers of particular sets of behaviors. Also, hormonal changes occur in response to environmental challenges, and such environmental challenges have to be interpreted as such by the brain in order to induce hormonal variations. Thus, although corticosteroids might modulate cognitive function, cognitive processing can in turn modulate corticosteroid secretion and enter a closed-loop system of modulatory actions. In this sense, the static unidirectional view of corticosteroid actions on the brain should be re-analyzed if one wishes to understand the multiple subtleties of the relationship between stress, memory and the brain. The impact of stress hormones on hippocampal atrophy and/or neurogenesis might help shed light on these complex relationships.

3.3. Corticosteroids and atrophy of the hippocampus The third argument that has been consistently used in order to confirm the stress–hippocampus link is that chronic exposure to high levels of corticosteroids leads to hippocampal atrophy in both animals [94,95], and humans [101,107,194].

S.J. Lupien, M. Lepage / Beha!ioural Brain Research 127 (2001) 137–158

In the rat and monkey, chronic hypercortisolemia and/or stress has been shown to increase hippocampal neuronal vulnerability to subsequent insults, and in some case, neuronal loss has been reported [3,24,32,94,95,136,168,171,206], although not in all experimental paradigms [6,9]. Interestingly, in the rat, less severe forms of chronic stress paradigms have been shown to produce a different pattern of hippocampal morphological changes reflected in neuronal dendritic atrophy. This type of atrophy is characterized by an atrophy of CA3 pyramidal neurons and includes a less complex branching pattern and decreased total length of apical dendrites. It appears generally after 21 days of corticosterone treatment of 6 h/day of restraint stress [62,111–113,140,214,215,221]. Interestingly, neurons exhibiting dendritic atrophy are not necessarily committed to stress-induced neuronal death [167], which could explain recent results showing that chronic stress- and/ or corticosterone-induced hippocampal atrophy in the rat is reversible [114]. However, given that no other brain regions have been studied in relation to chronic exposure to stress or high levels of corticosteroids, it is difficult at this point to ascertain that the impact of corticosteroids on brain atrophy is specific to the hippocampus.

Due to obvious ethical reasons, most of the human studies that assess the impact of hypercortisolemia on hippocampal atrophy are correlational, although some of them have prospectively measured the impact of treatment on hippocampal volume [195]. In general, four types of pathologies associated with changes in circulating levels of corticosteroids have been studied, namely Cushing’s disease [194,195], depression [17,103,157,187], post-traumatic stress disorder [15– 17,71,196,213], and unhealthy aging [33,107].

Patients with Cushing’s syndrome represent a powerful model in which to assess the association of hypercortisolemia and hippocampal atrophy in humans since these patients experience high levels of endogenous cortisol for periods of months or years. Using magnetic resonance imaging (MRI) of the brain, Starkman et al.

[194] showed that Cushing’s patients have decreased hippocampal volume when compared with controls. In the same study, hippocampal volume was negatively correlated with plasma cortisol concentrations, and positively correlated with scores on verbal learning and recall tasks. More recently, the same group [195] reported that therapeutic decrease in cortisol levels in Cushing’s patients increased by 10% the hippocampal formation of treated patients, revealing some forms of reversibility of hippocampal atrophy in human populations.

A similar atrophy of the hippocampus has been reported in depressed patients by Sheline and collaborators [187], and more recently by Bremner and collaborators [17]. However, and in contrast with studies

149

performed in Cushing’s patients, the hippocampal atrophy reported to occur in depressed populations was observed in the absence of increased hypercortisolemia [187], or absence of cortisol measurement [17] in patient populations. In this context, it is particularly difficult to associate the observed hippocampal atrophy to the presence of hypercortisolemia in the patient population.

Also, depressive disorder has been related to frontal lobe dysfunction, a suggestion confirmed by neuroimaging studies showing a 7% reduction of the total volume of the frontal lobe in major depressive disorder [28], and a substantial 39–48% decreases in the subgenual prefrontal gray matter in patients with major depressive disorder and bipolar depression [46]. The frontal lobe atrophy associated with depression has recently been confirmed by morphometric studies of human postmortem brain tissues [157], and a new study

by Lucassen and collaborators [103] report that hippocampal apoptosis in major depression is a minor event and is absent from areas at risk for corticosteroid exposure in human postmortem brain tissue. In conclusion, even if some form of hippocampal atrophy has been reported in depressive disorders, this atrophy is not isolated to the hippocampal formation, and has not been significantly correlated with increased secretion of corticosteroids.

However, the combined implication of frontal and hippocampal pathologies in depressed population is an interesting finding, in line with previously summarized data revealing the presence of MRs and GRs in both frontal and hippocampal human structures. The differential involvement of MRs and GRs in the acute versus chronic effect of corticosteroids on the brain could explain recent results obtained by Sheline and collaborators [188] showing that depression duration but not age predicts hippocampal volume in depressed women.

Given the possible protective effects of MR/GR balance in the hippocampus, one could suggest that hippocampal dysfunction and/or atrophy could appear after multiple exposure to high levels of corticosteroids (tonic protection from MRs), while frontal dysfunction and/or atrophy should appear more rapidly (no tonic protection from MRs) in response to stress-induced corticosteroid increase. Such a pattern could explain the impact of depression duration on hippocampal volume in depressed populations [187]. However, it is clear that only a longitudinal follow-up of depressed patients could confirm such a suggestion.

The same frontal/hippocampal rationale could also apply to patients with post-traumatic stress disorder, in whom atrophy of the hippocampus has been reported [15–17,71,196,213]. Again, there is discrepancy in this finding given data showing that PTSD patients present hypocortisolemia, instead of hypercortisolemia (for a review, see [222]), and other reports showing prefrontal system dysfunction in PTSD populations [90]. Given

150

S.J. Lupien, M. Lepage / Beha!ioural Brain Research 127 (2001) 137–158

these results, it is difficult at this point to relate hippocampal atrophy to basal cortisol levels in PTSD populations, or relate hippocampal atrophy to prefrontal cognitive impairments. However, in most PTSD populations studied at this point, trauma occurred many years before the actual measurement of hippocampal volume, so that it might be possible that similarly to depressed patients, duration of stress-induced symptoms is the important factor explaining development of hippocampal atrophy in this population.

Finally, using a 4-year longitudinal study performed with elderly humans, we previously showed that approximately 30% of this aged cohort presents a significant endogenous increase of cortisol levels with years that is related to a 14% atrophy of the hippocampus when compared with elderlies with normal secretion of cortisol over the 4-year period [107]. Interestingly, we also reported that this group of aged humans presented memory impairments [104] as well as reported more frequent feelings of stress, fatigue and depression when compared with the other elderly individuals [107]. Although in this study, the atrophy was shown to be specific to the hippocampus when compared with other temporal lobe structures, there was no MRI measurements performed on frontal lobe regions, thus preventing any definite conclusion as to the specificity of corticosteroid-induced effects on the aged human brain.

A retrospective analysis of these brain images is actually being performed by our group in order to assess whether long-term exposure to high levels of corticosteroids had any significant impact on frontal regions in this elderly population.

Altogether, these data tend to support the idea that chronic exposure to high levels of corticosteroids induce an atrophy of the hippocampus. However, most studies performed in humans are somewhat weakened by uncontrolled factors that could explain the observed hippocampal atrophy. Also, one has to come to the conclusion that a unidirectional point of view of negative corticosteroid actions on the hippocampus is usually taken by scientists in the field. However, the observation that some of the reported hippocampal atrophy is reversible would tend to suggest that high corticosteroid levels could be a consequence rather than a cause of hippocampal atrophy. Indeed, comparative studies of different species show that hippocampal volume is increased in mammalian and avian species that depend critically on spatial memory for survival. Examples of such critical memory skills involve home-range navigation, migration, brood parasitism, and memorybased cache recovery in birds that hide food [200].

This association between performance on spatial tasks and hippocampal size could be explained in terms of the impact of environmental demands on hippocampal volume, rather than the converse. For example,

Clayton and collaborators [26,27] reported that the hippocampus of titmice and chickadees increases in volume in association with the experience of storing and recovering food cache. This result shows that in animals, there can be an experience-dependent hippocampal growth that occurs at a relatively late stage in development. Similar results of experiencebased hippocampal volume were recently reported by Maguire and collaborators [116] in humans. These authors tested the hypothesis that the ability of taxi cab drivers to navigate correlates with hippocampal volume. They showed that compared with age-matched controls, the taxi drivers had larger posterior hippocampi. Although these results surely do not confirm that the experience of driving a taxi cab in the complex streets of London has a significant effect on hippocampal volume, they nonetheless raise the intriguing possibility of hippocampal plasticity in response to environmental demands.

This later hypothesis is to be put into perspective with the previously summarized data of hippocampal atrophy in PTSD, depressed, and unhealthy elderlies. In each of these states, stress has been shown to be a significant factor in inducing or perpetuating the negative symptoms. It is known that the HPA axis is a system, which monitors and responds to the environment throughout life. Given the combined implication of stress and the HPA axis in these various disorders related to hippocampal atrophy, it could be suggested that stress will act by decreasing the capacity of the individual to respond to environmental demands. Given the relationship observed between environmental demands and hippocampal volume, one could thus postulate that the hippocampal atrophy observed in these populations is the results of a stress-induced decrease in response to environmental demands, which then leads to atrophy of the hippocampus. Although this is surely a hypothesis that will need confirmation before being seriously proposed, it nonetheless shows that the stress –hippocampal atrophy link can be studied using other important variables.

3.4. Corticosteroids and hippocampal neurogenesis Although for a long time it was thought that the brain of adult mammals do not generate new nerve cells, more recent evidence showed that neurons are born in the adult mammalian brain [2]. Interestingly, in the adult brain, the generation of new neurons (called neurogenesis) occurs in only two regions. The first regions is the subventricular zone in the wall of the lateral ventricle where new interneurons are generated for the olfactory bulb, and the second region is the subgranular zone of the dentate gyrus of the hippocampus, which gives rise to the granule cells. Moreover, recent evidence show that adult neurogenesis in the

S.J. Lupien, M. Lepage / Beha!ioural Brain Research 127 (2001) 137–158

dentate gyrus of the hippocampus is a feature of all mammalian species, occurring in rats, mice, tree shrews, marmosets, macaques, and humans [23,50,67– 69,88].

Although the functional significance of hippocampal neurogenesis has been questioned in the past, a new study published in the March 2001 issue of Nature reported that neurogenesis in the adult is involved in the formation of trace memories, suggesting that the new neurons actually contribute to the function of the adult brain [190].

Corticosteroid effects on hippocampal neurogenesis have been extensively studied and it has been shown that throughout post-natal life, corticosteroid exert suppressive effects on cell proliferation in the dentate gyrus [165,176]. Interestingly, the stress hyporesponsive period observed shortly after birth in the rat coincides with a period of maximal granule cell production in the dentate gyrus [181], which suggests an inhibitory action of corticosteroids on post-natal cell proliferation.

Similar detrimental effects of corticosteroids on adult hippocampal neurogenesis have been reported. Treatment of adult rats with corticosterone decreases the proliferation of granule cell precursors [21], while removal of adrenal steroids stimulates the proliferation of granule cell precursors during adulthood [22]. Given the suppressive actions of corticosteroids on hippocampal neurogenesis, acute stressful experiences (which activate secretion of corticosteroids) have been suggested to inhibit cell proliferation in adulthood, a hypothesis that has been demonstrated. An acute stressful experience decreases the number of adult-generated neurons in the dentate gyrus in various species, including the rat [61,74,186], tree shrew [67], and marmoset [68]. In a similar vein, repeated stress also produces prolonged suppression of cell proliferation in the dentate gyrus of the adult tree shrew [59,60]. Although it has not been demonstrated, stress induced decrease in dentate gyrus cell proliferation could also contribute, in line with atrophy of the pyramidal cells of the hippocampus, to changes in hippocampal volume observed after chronic exposure to high levels of corticosteroids (see previous section).

Interestingly, it has been shown that experience can also induce cell proliferation in the dentate gyrus.

Gould and collaborators [69] reported that training on a task that requires the hippocampal formation results in an increase in the number of adult-generated granule cells. Importantly, these authors showed that in untrained adult laboratory animals, the majority of generated cells degenerate within 2 weeks of production, while training on hippocampal tasks rescued a significant proportion of these cells [69]. In line with experiments showing experience-dependent hippocampal volume changes, studies performed in adult neurogenesis show that learning can enhance the number of granule neurons. If this is the case, any change in the

151

learning experience (due to social isolation, disease, depression, aging etc.) could lead to significant changes in hippocampal neurogenesis and contribute to both the memory impairments and hippocampal atrophy reported to occur with chronic exposure to high levels of corticosteroids.

The mere presence of neurogenesis in the hippocampal regions (along with neurogenesis in the olfactory bulb), and the important effects of stress hormones on hippocampal neurogenesis have been taken as good evidence that stress has a specific impact on the hippocampus. However, a recent study reported the induction of neurogenesis in the neocortex of adult mice [115]. These authors destroyed a subset of pyramidal neurons that project from the neocortex to the thalamus, a technique that resulted in the slow death by apoptosis of the targeted cells only, without affecting the surrounding cortical tissue. Although the number of new cells formed in the 2 weeks after the lesion was similar in control and experimental mice, about 1–2% of the newly formed cells in the damaged neocortex expressed neuronal markers in the experimental mice only. Interestingly, the new neuronal cells occurred only in the cortical layer undergoing degeneration, and some of these cells had the morphological characteristics of pyramidal neurons. Although the neurogenic response observed by these authors was very limited, it nonetheless raises the possibility of neurogenesis elsewhere in the brain. Given the presence of MRs and/or GRs in the cortex of primates and humans, it is clear that further developments in this field of research will help extend our understanding of the impact of stress and related stress hormones on brain neurogenesis.

4. Conclusion In this review of the literature, we summarized the four arguments for a stress– hippocampus link used in animal and human psychoneuroendocrine research, and reported new evidence showing that this link, although it surely exists, might not be the only one relating stress to the brain. Indeed, two types of corticosteroid receptors are distributed differentially in the rodent and primate brain, which could explain some discrepancies in results obtained with both types of populations.

Also, what is called ‘learning and memory’ and conceptualized as a single entity in some studies is actually a composite of various cognitive processing components that are also distributed in different regions of the brain, including the hippocampus but also cortical areas such as the frontal and prefrontal cortices. Third, although chronic exposure to high levels of corticosteroids certainly contribute to atrophy of the hippocampus and inhibition of dentate gyrus cell proliferation, these two processes are also very sensitive to

152

S.J. Lupien, M. Lepage / Beha!ioural Brain Research 127 (2001) 137–158

environmental demands, which can themselves be modified by stress.

All these evidences point to a modulatory role of corticosteroids on the rodent and human brain and the necessity to integrate various levels of analysis in our search for corticosteroid actions on the brain. Three new levels of analysis should be seriously considered by scientists interested in studying the impact of corticosteroids on human cognitive function. The first one relates to a methodological refinement of our neuroendocrine protocols and to a tighter control of time of cognitive measurements. It has been known for a while that baseline cognitive function is not the same in the AM versus PM phase in humans [53]. In 1996, FehmWorlsdorf and collaborators [53] postulated that this natural diurnal variation of cognitive function could be related to circadian variations in cortisol levels. In their study, they measured the effect of an oral administration of 50 mg of hydrocortisone on a free recall task in young normal controls in the morning, when endogenous cortisol levels are at their peak, and at night, when they are at their lowest concentration. The results revealed that hydrocortisone administration suppressed the increased cognitive performance in the morning, while it had no effect on cognitive performance when administered at night. We recently extended this result in young normal controls in which we showed that administration of 35 mg hydrocortisone in the evening (PM phase) significantly increased cognitive efficiency when compared with placebo treated subjects [109].

Altogether, these results show that because of the differential involvement of MRs and GRs on the magnitude and/or direction of corticosteroid-induced cognitive changes, time of testing of endogenous or exogenous increases of corticosteroids can be a crucial factor in determining both the magnitude and direction of cognitive changes induced by corticosteroids.

The second level of analysis that should be taken into account concerns the type of population studied when assessing the effects of corticosteroids on human cognitive function. Knowledge of the differential impact of MRs and GRs on cognitive function might help disentangle some of the effects of hypo- versus hypercortisolemia on cognitive function. Given that absence of MR and GR activation (induced by surgical or chemical adrenalectomy) can induce cognitive impairments as important as saturation of MRs and GRs (for a review, see [37,105]), one has to take into account the fact that cognitive impairments in hypocortisolemic populations such as PTSD or burn-out patients might have a different cause than those observed in hypercortisolemic populations such as Cushing or depressed patients.

Also, and based on the MR/GR balance theory, one should postulate that endogenous or exogenous administration of corticosteroids should have a different impact on these populations, based on their baseline levels of cortisol before treatment.

Finally, our analysis of new human brain imaging data shows that memory function cannot be envisioned as a single entity process and each component of learning and memory (encoding, consolidation, retrieval) involves the combined activation of various brain regions. Although corticosteroid effects have been reported for encoding [89,143,144], consolidation [105,106], and retrieval [39], one has to take into account recent studies [142] of old concepts [40,66,100,134,156,173 –175] revealing the phenomenon of ‘reconsolidation’ and showing that reactivated memory are more sensitive to various amnestic treatments than newly acquired memories (for a recent review, see [174]). In summary, these studies show that treatments that are ineffective at impairing memory when administered right before recall can in fact do so if memory for the event is reactivated just before treatment. These results show that memories exist in an active state during which they are labile and susceptible to disruption by amnestic agents, and in an inactive state during which they are resistant to amnestic treatments.

Interestingly, a new study by Przybyslawski and collaborators [155] showed that although systemic injections of propanolol (a !-adrenergic antagonist) are ineffective in inducing memory impairments when given right before retrieval, the same treatment induces significant memory impairments when administered right after reactivation by a simple retention test applied 48 h after training. The demonstration of the vulnerability of memory when it is in an active state reinforces the idea that memories, reorganized as a function of new experiences, undergo a reconsolidation process that can be modulated by various treatments and can be sustained by various brain regions at different times of the memory formation.

In line with these new and exciting data, it is becoming clear that a conception of stress, memory and the entire brain will emerge in which the effects of stress on memory will be assessed using a theoretical framework taking into account the fact that the brain is not a spectator but rather an active participant in its response to the environment, and particularly environmental stress.

Acknowledgements S.J. Lupien research summarized in this paper was funded by a Scientist Research Award from Fonds de la recherche en sante´ du Que´ bec (FRSQ), by an operating grant from Canadian Institute of Health Research (CIHR), and by a Research Scholar Award from EJLB Foundation. The Douglas Hospital Longitudinal Study of Normal and Pathological Aging is funded by a grant from the Alzheimer Society of Canada. Part of M. Lepage research summarized in this paper was sup-

S.J. Lupien, M. Lepage / Beha!ioural Brain Research 127 (2001) 137–158

ported by a foundation of Anne and Max Tanenbaum in support of research in cognitive neuroscience.

References [1] Alnemri ES, Maksymowych AB, Robertson NM, Litwack G. Overexpression and characterization of the human mineralocorticoid receptor. J Biol Chem 1991;266:18072 –81.

[2] Altman J, Das GD. Autoradiographic and histological evidence of postnatal hippocampal neurogenesis in rats. J Comp Neurol 1965;124:319 – 35.

[3] Arbel I, Kadar T, Silbermann M, Levy A. The effects of long-term corticosterone administration on hippocampal morphology and cognitive performance of middle-aged rats. Brain Res 1994;657:227 –35.

[4] Arriza JL, Weinberger C, Cerelli G, Glaser TM, Handelin BL, Housmann DE, Evans RM. Cloning of human mineralocorticoid receptor cDNA: structural and functional kinship with the glucocorticoid receptor. Science 1987;237:268 –75.

[5] Bagley J, Moghaddam B. Temporal dynamics of glutamate efflux in the prefrontal cortex and in the hippocampus following repeated stress: effects of pretreatment with saline or diazepam. Neuroscience 1997;77:65 –73.

[6] Bardgett ME, Taylor GT, Csernansky JG, Newcomer JW, Nock B. Chronic corticosterone treatment impairs spontaneous alternation behavior in rats. Behav Neural Biol 1994;61:186 – 90.

[7] Beckwith BE, Petros TV, Scaglione C, Nelson J. Dose-dependent effects of hydrocortisone on memory in human males.

Physiol Behav 1986;36:283 –6.

[8] Bennett MC, Diamond DM, Fleshner M, Rose GM. Serum corticosterone level predicts the magnitude of hippocampal primed burst potentiation and depression in urethane-anesthetized rats. Psychobiology 1991;19:301 – 7.

[9] Bodnoff SR, Humphreys AG, Lehman JC, Diamond DM, Rose GM, Meaney MJ. Enduring effects of chronic corticosterone treatment on spatial learning, synaptic plasticity, and hippocampal neuropathology in young and middle-aged rats. J Neurosci 1995;15:61 –9.

[10] Bohus B. Central nervous structures and the effect of ACTH and corticosteroids on avoidance behaviour: a study with intracerebral implantation of corticosteroids in the rat. Prog Brain Res 1970;132:171 –84.

[11] Bohus B, Lissak K. Adrenocortical hormones and avoidance behaviour of rats. Int J Neuropharmacol 1968;7:301 – 6.

[12] Bohus B, De Kloet ER, Veldhuis HD. In: Ganten D, Pfaff DW, editors. Adrenal Steroids and Behavioral Adaptation: Relationship to Brain Corticoid Receptors. New York: Springer, 1983:107 –48.

[13] Bontempi B, Laurent-Demir C, Destrade C, Jaffard R. Timedependent reorganization of brain circuitry underlying longterm memory storage. Nature 1999;400:671 – 5.

[14] Bradbury M, Dallman MF. Effects of type 1 and type 2 glucocorticoid receptor antagonists on ACTH levels in the PM. Soc Neurosci Abstr 1989;19:716.

[15] Bremner JD, Randall P, Scott T, Bronen R, Seibyl J. MRIbased measurement of hippocampal volume in patients with combat-related posttraumatic stress disorder. Am J Psychiatry 1995;152:973 – 80.

[16] Bremner JD, Randall P, Vermetten E, Staib L, Bronen R. Magnetic resonance imaging-based measurement of hippocampal volume in posttraumatic stress disorder related to childhood physical and sexual abuse — a preliminary report. Biol Psychiatry 1997;41:23 –7.

153

[17] Bremner JD, Narayan M, Anderson ER, Staib LH, Miller HL, Charney DS. Hippocampal volume reduction in major depression. Am J Psychiatry 2000;157:115 – 7.

[18] Brewer JB, Zhao Z, Desmond JE, Glover GH, Gabrieli JDE. Making memories: brain activity that predicts how well visual experience will be remembered. Science 1998;281:1185 –7.

[19] Brown N. An Introduction to Neuroendocrinology. Cambridge: Cambridge University Press, 1998:408.

[20] Cabeza R, Nyberg L. Imaging cognition II: an empirical review of 275 PET and fMRI studies. J Cogn Neurosci 2000;12:1 –47.

[21] Cameron HA, Gould E. Adult neurogenesis is regulated by adrenal steroids in the dentate gyrus. Neuroscience 1994;61:203 – 9.

[22] Cameron HA, McKay R. Restoring production of hippocampal neurons in old age. Nat Neurosci 1999;2:894 – 7.

[23] Cameron HA, Woolley CS, McEwen BS, Gould E. Differentiation of newly born neurons and glia in the dentate gyrus of the adult rat. Neuroscience 1993;56:337 – 44.

[24] Clark AS, Mitre MC, Brinck-Johnsen T. Anabolic – androgenic steroid and adrenal steroid effects on hippocampal plasticity.

Brain Res 1995;679:64 – 71.

[25] Clark LD, Bauer W, Cobb S. Preliminary observations on mental disturbances occurring in patients under therapy with cortisone and ACTH. New Engl J Med 1952;246:205 –16.

[26] Clayton NS, Krebs JR. Hippocampal growth and attrition in birds affected by experience. PNAS 1994;91:7410 – 4.

[27] Clayton NS, Soha J. Memory in avian food-storing and song learning: a general mechanisms or different processesE´ . Adv Study Behav 1999;28:115 – 74.

[28] Coffey CE, Wilkinson WI, Weiner RD, Parashos IA, Djang WT, Webb MC. Quantitative cerebral anatomy in depression.

Arch Gen Psychiatry 1993;50:7 – 16.

[29] Cohen NJ, Eichenbaum H. Memory, Amnesia, and the Hippocampal System. Cambridge: MIT Press, 1993:330.

[30] Cohen NJ, Ryan J, Hunt C, Romine L, Wszalek T, Nash C. Hippocampal system and declarative (relational) memory: summarizing the data from functional neuroimaging studies.

Hippocampus 1999;9:83 – 98.

[31] Conrad CD, Lupien SJ, McEwen BS. Support for a bimodal role for Type II adrenal steroid receptors in hippocampus-dependent spatial recognition memory. Neurobiol Learn Mem 1999;72:39 – 46.

[32] Dachir S, Kadar T, Robinzon B, Levy A. Cognitive deficits induced in young rats by long-term corticosterone administration. Behav Neural 1993;60:103 –9.

[33] Davis KL, Davis BM, Greenwald BS, Mohs RC, Mathe´ AA, Johns CA, Horvath TB. Cortisol and Alzheimer’s disease, I: basal studies. Am J Psychiatry 1986;143:300 – 5.

[34] Decety J, Grazes J, Costes N, Perani D, Jeannerod M, Procyk E, Grassi F, Fazio F. Brain activity during observation of actions. Influence of action content and subject’s strategy. Brain 1997;120:1763 – 77.

[35] De Kloet ER, Wallach G, McEwen BS. Differences in corticosterone and dexamethasone binding to rat brain and pituitary.

Endocrinology 1975;96:598 – 609.

[36] De Kloet ER, De Kock S, Schild V, Veldhuis HD. Antiglucocorticoid RU 38486 attenuates retention of a behaviour and disinhibits the hypothalamic –pituitary adrenal axis at different brain sites. Neuroendocrinology 1988;47:109 –15.

[37] De Kloet ER, Oitzl MS, Joels M. Stress and cognition: are corticosteroids good or bad guys. Trends Neurosci 1999;22:422 – 6.

[38] De Kloet ER. Stress in the brain. Eur J Pharmacol 2000;405:187 – 98.

[39] De Quervain DJF, Roozendaal B, Nitsch RM, McGaugh JL, Hock C. Acute cortisone administration impairs retrieval of long-term declarative memory in humans. Nat Neurosci 2000;3:313 – 4.

154

S.J. Lupien, M. Lepage / Beha!ioural Brain Research 127 (2001) 137–158

[40] DeVietti TL, Hopfer TM. Complete amnesia induced by ECS and complete recovery of memory following reinstatement treatment. Physiol Behav 1974;12:599 – 603.

[41] Desgranges B, Baron J, Eustache F. The functional neuroanatomy of episodic memory: the role of the frontal lobes, the hippocampal formation, and other areas. Neuroimage 1998;8:198 – 213.

[42] Diamond DM, Bennett MC, Fleshner M, Rose GM. InvertedU relationship between the level of peripheral corticosterone and the magnitude of hippocampal primed burst potentiation.

Hippocampus 1992;2:421 – 30.

[43] Diorio D, Viau V, Meaney MJ. The role of the medial prefrontal cortex (cingulate gyrus) in the regulation of hypothalamic – adrenal responses to stress. J Neurosci 1993;13:3839 –47.

[44] Dolan RJ, Fletcher PC. Dissociating prefrontal and hippocampal function in episodic memory encoding. Nature 1997;388:582 – 5.

[45] Dizel E, Cabeza R, Picton TW, Yonelinas AP, Scheich H, Heinze H, Tulving E. Task-related an item-related brain processes of memory retrieval. Proc Natl Acad Sci USA 1999;96:1794 – 9.

[46] Drevets W, Price J, Simpson JRJ, Todd R, Reich T, Vannier M. Subgenual prefrontal cortex abnormalities in mood disorders. Nature 1997;386:824 – 7.

[47] Dubrovsky BO, Liquornik MS, Noble P, Gijsbers K. Effects of 5"-dihydrocorticosterone on evoked responses and long-term potentiation. Brain Res Bull 1987;19:635 – 8.

[48] Eichenbaum H, Otto T, Cohen NJ. Two functional components of the hippocampal memory system. Behav Brain Sci 1994;17:449– 518.

[49] Eldridge LL, Knowlton BJ, Furmanski CS, Bookheimer SY, Engel SA. Remembering episodes: a selective role for the hippocampus during retrieval. Nat Neurosci 2000;3:1149 – 52.

[50] Eriksson PS, Perfilieva E, Bjork-Eriksson T, Alborn AM, Nordborg C, Peterson DA. Neurogenesis in the adult human hippocampus. Nat Med 1998;4:1313 – 7.

[51] Evans RM. Primary structure and expression of a functional human glucocorticoid receptor cDNA. Nature 1985;318:635 – 41.

[52] Evans RM, Arriza JL. A molecular framework for the actions of glucocorticoid hormones in the nervous system. Neuron 1989;2:1105– 12.

[53] Fehm-Wolfsdorf G, Reutter D, Zenz H, Born J, Lorenz-Fehm H. Are circadian variations in taste thresholds cortisol-dependent. J Psychophysiol 1993;7:65 –72.

[54] Feldman S, Conforti N. Participation of the dorsal hippocampus in the glucocorticoid negative-feedback effect on adrenocortical activity. Neuroendocrinology 1980;30:52 –5.

[55] Feldman S, Conforti N. Modifications of adrenocortical responses following frontal cortex stimulation in rats with hypothalamic differentiations and medial forebrain bundle lesions.

Neuroscience 1985;15:1045 –7.

[56] Filipini D, Gijsbers K, Birmingham MK, Dubrovsky B. Effects of adrenal steroids and their reduced metabolites on hippocampal long-term potentiation. J Steroid Biochem Mol Biol 1991;40:87– 92.

[57] Fischette CT, Komisurak BR, Ediner HM, Feder HH, Siegal A. Differential fornix ablations and the circadian rhythmicity of adrenal corticosterone secretion. Brain Res 1980;195:373 – 80.

[58] Forget H, Lacroix A, Somma M, Cohen H. Cognitive decline in patients with Cushing’s syndrome. J Int Neuropsychol Soc 2000;6:20– 9.

[59] Fuchs E, Uno H, Flugge G. Chronic psychosocial stress induces morphological alterations in hippocampal pyramidal neurons of the tress shrew. Brain Res 1995;673:275 –82.

[60] Fuchs E, Flugge G, McEwen BS, Tanapat P, Gould E. Chronic subordination stress inhibits neurogenesis and decreases the volume of the granule cell layer. Soc Neurosci 1997;23:317.

[61] Galea LAM, Tanapat P, Gould E. Exposure to predator odor suppresses cell proliferation in the dentate gyrus of adult rats via a cholinergic mechanism. Soc Neurosci Abstr 1996;22:314.

[62] Galea LAM, McEwen BS, Tanapat P, Deak T, Spencer RL, Dhabhar FS. Sex differences in dendritic atrophy of CA3 neurons in response to chronic restraint stress. Neuroscience 1997;81:689 – 97.

[63] Gerlach JL, McEwen BS. Rat brain binds adrenal steroid hormone: ratioautography of hippocampus with corticosterone.

Science 1980;175:1133 – 6.

[64] Goldman-Rakic PS. Circuitry of primate prefrontal cortex and regulation of behavior by representational memory. In: Plum F, editor. Handbook of Physiology. Section 1: The Nervous System. Higher Functions of the Brain, vol. 5. Bethesda: American Psychological Society, 1987:373 – 417.

[65] Goldman-Rakic PS. Cellular basis of working memory. Neuron 1995;14:477 – 85.

[66] Gordon WC, Mowrer RR. An extinction trial as a reminder treatment following electroconvulsive shock. Anim Learn Behav 1980;8:363 – 7.

[67] Gould E, McEwen BS, Tanapat P, Galea LAM, Fuchs E. Neurogenesis in the dentate gyrus of the adult tree shrew is regulated by psychosocial stress and NMDA receptor activation. J Neurosci 1997;17:2492 – 8.

[68] Gould E, Tanapat P, McEwen BS, Flugge G, Fuchs E. Proliferation of granule cell precursors in the dentate gyrus of adult monkeys is diminished by stress. PNAS 1998;95:3168 – 71.

[69] Gould E, Reeves AJ, Fallah M, Tanapat P, Fuchs E. Hippocampal neurogenesis in adult Old Worked primates.

PNAS 1999;96:5263 – 7.

[70] Greidanus TJB. Effects of steroids on extinction of an avoidance response in rats. A structure-activity relationship study.

Prog Brain Res 1970;32:185 – 91.

[71] Gurvits T, Shenton M, Hokama H, Ohta H, Lasko N, Pitman R. Magnetic resonance imaging study of hippocampal volume in chronic, combat-related PTSD. Biol Psychiatry 1996;40:1091 – 9.

[72] Habib R, Lepage M. Novelty assessment in the brain. In: Tulving E, editor. Memory, Consciousness, and the Brain: The Tallinn Conference. Philadelphia: Psychology Press, 2000:265 – 77.

[73] Haxby JV, Ungerleider LG, Horwitz B, Maisog JM, Rapoport SI, Grady CL. Face encoding and recognition in the human brain. Proc Natl Acad Sci USA 1996;93:922 – 7.

[74] Heale VR, Vanderwolf CH, Kavaliers M. Components of weasel and fox odors elicit fast wave bursts in the dentate gyrus of rats. Behav Brain Res 1994;63:159 – 65.

[75] Heckers S, Rauch SL, Goff D, Savage CR, Schacter DL, Fischman AJ, Alpert NM. Impaired recruitment of the hippocampus during conscious recollection in schizophrenia.

Nat Neurosci 1998;1:318 –23.

[76] Henke K, Buck A, Weber B, Wieser HG. Human hippocampus establishes associations in memory. Hippocampus 1997;7:249 – 56.

[77] Henke K, Weber B, Kneifel S, Wieser HG, Buck A. Human hippocampus associates information in memory. Proc Natl Acad Sci USA 1999;96:5884 – 9.

[78] Henson RNA, Rugg MD, Shallice T, Josephs O, Dolan RJ. Recollection and familiarity in recognition memory: an eventrelated functional magnetic resonance imaging study. J Neurosci 1999;19:3962 –72.

[79] Henson RNA, Shallice T, Dolan RJ. Right prefrontal cortex and episodic memory retrieval: a functional MRI test of the monitoring hypothesis. Brain 1999;122:1367 –81.

[80] Herman JP, Schafer MK-H, Young EA, Thompson R, Douglass J, Akil H, Watson SJ. Evidence for hippocampal regulation of neuroendocrine neurons of the

S.J. Lupien, M. Lepage / Beha!ioural Brain Research 127 (2001) 137–158

[81]

[82]

[83] [84]

[85]

[86]

[87]

[88]

[89]

[90]

[91]

[92] [93]

[94]

[95]

[96]

[97]

[98]

[99]

[100] [101]

hypothalamo–pituitary –adrenocortical axis. J Neurosci 1989;9:3072 – 82.

Hollenberg SM, Weinberger C, Ong ES, Cerelli G, Oro A, Lebo R, Thompson EB, Rosenfeld MG, Evans RM. Primary structure and expression of a functional human glucocorticoid receptor cDNA. Nature 1985;318:635 – 41.

Jacobson L, Sapolsky RM. The role of the hippocampus in feedback regulation of the hypothalamic – pituitary –adrenal axis. Endocr Rev 1991;12:118 –34.

James W. The Principles of Psychology. New York: Henry Holt, 1890:202.

Jarrard LE, Okaichi H, Goldschmidt R, Steward O. On the role of the hippocampal connections in the performance of place and cue tasks: comparisons with damage to the hippocampus.

Behav Neurosci 1984;98:946 – 54.

Johnson EO, Brady L, Gold PW, Chrousos GP. Distribution of hippocampal mineralocorticoid and glucocorticoid receptor mRNA in a glucocorticoid resistant nonhuman primate.

Steroids 1996;61:69 –73.

Kapur S, Craik FIM, Jones C, Brown GM, Houle S, Tulving E. Functional role of the prefrontal cortex in retrieval of memories: a PET study. Neuroreport 1995;6:1880 – 4.

Kapur S, Tulving E, Cabeza R, McIntosh AR, Houle S, Craik FIM. Neural correlates of intentional learning of verbal materials: a PET study in humans. Cogn Brain Res 1996;4:243 –9.

Kempermann G, Kuhn HG, Gage FH. More hippocampal neurons in adult mice living in an enriched environment. Nature 1997;386:493 – 5.

Kirschbaum C, Wolf OT, May M, Wippich W, Hellhammer DH. Stress and drug-induced elevation of cortisol levels impair explicit memory in healthy adults. Life Sci 1996;58:1475 – 83.

Koenen KC, Driver KL, Oscar-Berman M, Wolfe J, Folsom S, Huang MT, Schlesinger L. Measures of prefrontal system dysfunction in posttraumatic stress disorder. Brain Cogn 2001;45:64 – 78.

Konishi S, Wheeler ME, Donaldson DI, Buckner RL. Neural correlates of episodic retrieval success. Neuroimage 2000;12:276 – 86.

Kopelman MD, Stevens TG, Foli S, Grasby P. PET activation of the medial temporal lobe in learning. Brain 1998;121:875 – 87.

Kovacs GL, Telegdy G, Lissak K. 5-Hydroxytryptamine and the mediation of pituitary – adrenocortical hormones in the extinction of active avoidance behavior. Neuroendocrinology 1976;1:219 – 30.

Landfield P, Waymire J, Lynch G. Hippocampal aging and adrenocorticoids: a quantitative correlation.

Science 1978;202:1098 – 102.

Landfield PW, Baskin RW, Pitler TA. Brain –aging correlates: retardation by hormonal –pharmacological treatments. Science 1981;214:581–4.

Lepage M, Habib R, Tulving E. Hippocampal PET activations of memory encoding and retrieval: the HIPER model.

Hippocampus 1998;8:313 – 22.

Lepage M, Ghaffar O, Nyberg L, Tulving E. Prefrontal cortex and episodic memory retrieval mode. Proc Natl Acad Sci USA 2000;97:506– 11.

Lepage M, Habib R, Cormier H, Houle S, McIntosh AR. Neural correlates of semantic associative encoding in episodic memory. Brain Res Cogn Brain Res 2000;9:271 –80.

Leverenz JB, Wilkinson CW, Raskind MA, Peskind ER. Immunohistochemical localization of glucocorticoid and mineralocorticoid receptors in the primate hippocampus and amygdala.

Soc Neurosci Abstr 1999;25:708.

Lewis D, Miller RR, Misanin JR. Control of retrograde amnesia. J Comp Physiol Psychol 1968;66:48 – 53.

Ling M, Perry P, Tsuang M. Side effects of corticosteroid therapy. Arch Gen Psychiatry 1981;38:471 –7.

155

[102] Lopez JF, Chalmers DT, Little KY, Watson SJ, Bennet AE. Regulation of serotonin 1A, glucocorticoid, and mineralocorticoid receptor in rat and human hippocampus: implications for the neurobiology of depression. Biol Psychiatry 1998;43:547 – 73.

[103] Lucassen PJ, Muller MB, Holsboer F, Bauer J, Holtrop A, Wouda J, Hoogendijk WJG, De Kloet ER, Swaab DJ. Hippocampal apoptosis in major depression is a minor event and absent from subareas at risk for glucocorticoid overexposure. Am J Pathol 2001;158:453 – 68.

[104] Lupien SJ, Lecours AR, Lussier I, Schwartz G, Nair NPV, Meaney MJ. Basal cortisol levels and cognitive deficits in human aging. J Neurosci 1994;14:2893 – 903.

[105] Lupien SJ, McEwen BS. The acute effects of corticosteroids on cognition: integration of animal and human model studies.

Brain Res Rev 1997;24:1 – 27.

[106] Lupien SJ, Gaudreau S, Tchiteya BM, Maheu F, Sharma S, Nair NPV, Hauger RL, McEwen BS, Meaney MJ. Stress-induced declarative memory impairments in healthy elderly subjects: relationship with cortisol reactivity. J Clin Endcrinol Metab 1997;82:2070 – 5.

[107] Lupien SJ, DeLeon M, DeSanti S, Convit A, Tarshish C, Nair NPV, McEwen BS, Hauger RL, Meaney MJ. Longitudinal increase in cortisol during human aging predicts hippocampal atrophy and memory deficits. Nat Neurosci 1998;1:69 –73.

[108] Lupien SJ, Gillin C, Hauger RL. Working memory is more sensitive than declarative memory to the acute effects of corticosteroids: a dose – response study. Behav Neurosci 1999;113:1 – 11.

[109] Lupien SJ, Wilkinson CW, Brie`re S, Me´ nard C, Kin NY, Nair NPV. The positive effects of corticosteroids on cognition: studies in young human populations. Psychoneuroendocrinology, in press.

[110] Lynch G, Muller D, Seubert P, Larson J. Long-term potentiation: persisting problems and recent results. Brain Res Bull 1988;21:363 – 72.

[111] Magarinos AM, McEwen BS. Stress-induced atrophy of apical dendrites of hippocampal CA3c neurons: involvement of glucocorticoid secretion and excitatory amino acid receptors. Neuroscience 1995;69:89 – 98.

[112] Magarinos AM, McEwen BS, Flugge G, Fuchs E. Chronic psychosocial stress causes apical dendritic atrophy of hippocampal CA3 pyramidal neurons in subordinate tree shrews. J Neurosci 1996;16:3534 – 40.

[113] Magarinos AM, Orchinik M, McEwen BS. Morphological changes in the hippocampal CA3 region induced by non-invasive glucocorticoid administration: a paradox. Brain Res 1998;809:314 – 8.

[114] Magarinos AM, Deslandes A, McEwen BS. Effects of antidepressants and benzodiazepine treatments on the dendritic structure of CA3 pyramidal neurons after chronic stress. Eur J Pharmacol 1999;371:113 – 22.

[115] Magavi SS, Leavitt BR, Macklis JD. Induction of neurogenesis in the neocortex of adult mice. Nature 2000;405:951 – 5.

[116] Maguire EA, Gadian DG, Johnsrude IS, Good CD, Ashburner J, Frackowiad RS, Frith CD. Navigation-related structural changes in the hippocampus of taxi-drivers. PNAS 2000;98:4398 – 403.

[117] Martin A, Wiggs CL, Weisberg J. Modulation of human medial temporal lobe activity by form meaning, and experience.

Hippocampus 1997;7:587 – 93.

[118] Mason JW. A review of psychoendocrine research on the pituitary – adrenal cortical system.

Psychosom Med 1968;30:576 – 607.

[119] McDermott KB, Jones TC, Petersen SE, Lageman SK, Roediger HL. Retrieval success is accompanied by enhanced activation in anterior prefrontal cortex during recognition memory: an event-related fMRI study. J Cogn Neurosci 2000;12:965 –76.

156

S.J. Lupien, M. Lepage / Beha!ioural Brain Research 127 (2001) 137–158

[120] McEwen BS, Weiss JM, Schwartz LS. Selective retention of corticosterone by limbic structure in rat brain. Nature 1968;220:911 – 2.

[121] McEwen BS, Wallach G, de Kloet ER. Interaction in vivo and in vitro of corticoids and progesterone with cell nuclei and soluble macromolecules from rat brain regions and pituitary.

Brain Res 1976;105:129 –36.

[122] McEwen BS, Stephenson BS, Krey LC. Radioimmunoassay of brain tissue and cell nuclear corticosterone. J Neurosci Methods 1980;3:57 – 65.

[123] McEwen BS, DeKloet ER, Rostene WH. Adrenal steroid receptors and actions in the nervous system. Physiol Rev 1986;66:1121 – 50.

[124] McEwen BS. Protective and damaging effects of stress mediators. New Engl J Med 1998;338:171 –9.

[125] McEwen BS. Stress and hippocampal plasticity. Annu Rev Neurosci 1999;22:105 – 22.

[126] McEwen B. Allostasis and allostatic load: implications for neuropsychopharmacology.

Neuropsychopharmacology 2000;22:108 – 24.

[127] McEwen BS, Lupien SJ. Stress: hormonal and neural aspects.

In: Levy N editor. Encyclopedia of the human brain, in press.

[128] McIntosh AR. Mapping cognition to the brain through neural interactions. Memory 1999;7:523 –48.

[129] McIntosh AR. From location to integration: how neural interactions form the basis for human cognition. In: Tulving E, editor. Memory, Consciousness, and the Brain: The Tallinn Conference. Philadelphia: Psychology Press, 2000:346 – 62.

[130] Meaney MJ, Aitken DH. [3H]dexamethasone binding in rat frontal cortex. Brain Res 1985;328:176 –80.

[131] Menon V, White CD, Eliez S, Glover GH, Reiss AL. Analysis of a distributed neural system involved in spatial information, novelty, and memory processing. Hum Brain Mapp 2000;11:117– 29.

[132] Micco DJ, McEwen BS, Shein W. Modulation of behavioral inhibition in appetitive extinction following manipulation of adrenal steroids in rats: implications for involvement of the hippocampus. J Comp Physiol Psychol 1979;93:323 –9.

[133] Micco DJ, McEwen BS. Glucocorticoids, the hippocampus, and behavior: interactive relation between task activation and steroid hormone binding specificity. J Comp Physiol Psychol 1980;94:624 – 33.

[134] Miller RR, Springer AD. Induced recovery of memory in rats following electroconvulsive shock. Physiol Behav 1972;8:645 – 51.

[135] Mitchell JB, Meaney MJ. Effects of corticosterone on response consolidation and retrieval in the forced swim test. Behav Neurosci 1991;105:798 –803.

[136] Mizoguchi K, Kunishita T, Chui DH, Tabira T. Stress induces neuronal death in the hippocampus of castrated rats. Neurosci Lett 1992;138:157 –60.

[137] Moghaddam B, Bolinao ML, Stein-Behrens B, Sapolsky R. Glucocorticoids mediate the stress-induced extracellular accumulation of glutamate. Brain Res 1994;655:251 –4.

[138] Montaldi D, Mayes AR, Barnes A, Pirie H, Hadley DM, Patterson J, Wyper DJ. Associative encoding of pictures activates the medial temporal lobes. Hum Brain Mapp 1998;6:85 – 104.

[139] Mottaghy FM, Shah NJ, Krause BJ, Schmidt D, Halsband U, Jdncke L, Miller-Gardtner HW. Neuronal correlates of encoding and retrieval in episodic memory during a paired-word association learning task: a functional magnetic resonance imaging study. Exp Brain Res 1999;128:332 –42.

[140] Muddanna S, Rao S, Raju TR. Effect of chronic restraint stress on dendritic spines and excrescences of hippocampal CA3 pyramidal neurons: a quantitative study. Brain Res 1995;694:312 – 7.

[141] Murros K, Fogelholm R, Kettunen S, Vuorela AL. Serum cortisol and outcome of ischemic brain infarctions. J Neurol Sci 1993;116:12 – 7.

[142] Nader K, Schafe GE, LeDoux JE. Fear memories require protein synthesis in the amygdala for reconsolidation after retrieval. Nature 2000;406:722 –5.

[143] Newcomer JW, Craft S, Hershey T, Haskins K, Bardgett ME. Glucocorticoid-induced impairment in declarative memory performance in adult humans. J Neurosci 1994;14:2047 – 53.

[144] Newcomer JW, Selke G, Melson AK, Hershey T, Craft S, Richards K, Alderson AL. Decreased memory performance in healthy humans induced by stress-level cortisol treatment. Arch Gen Psychiatry 1999;56:527 – 33.

[145] Nyberg L, Cabeza R, Tulving E. PET studies of encoding and retrieval: the HERA model. Psychonom Bull Rev 1996;3:135 – 48.

[146] Nyberg L, Persson J, Habib R, Tulving E, McIntosh AR, Cabeza R, Houle S. Large scale neurocognitive networks underlying episodic memory. J Cogn Neurosci 2000;12:163 –73.

[147] Nyberg L, Tulving E, Habib R, Nilsson L, Kapur S, Houle S, Cabeza R, McIntosh AR. Functional brain maps of retrieval mode and recovery of episodic information. Neuroreport 1995;7:249– 52.

[148] Oitzl MS, vanHaarst AD, Sutanto W, de Kloet ER. Corticosterone, brain mineralocorticoid receptors (MRs) and the activity of the hypothalamic – pituitary – adrenal (HPA) axis: the Lewis rat as an example of increased central MR capacity and a hyporesponsive HPA axis. Psychoneuroendocrinology 1995;6:655– 75.

[149] Olton DS, Becker JT, Handelmann GE. Hippocampus, space and memory. Behav Brain Sci 1979;2:313 – 65.

[150] Ongur D, Price JL. Distribution of glucocorticoid receptors in the macaque central nervous system. Soc Neurosci Abstr 1997;23:1494.

[151] Owen AM, Downes JJ, Sahakian VJ, Polkey CE, Robbins TW. Planning and spatial working memory following frontal lobe lesions in man. Neuropsychologia 1990;28:1201 – 34.

[152] Pavlides C, Watanabe Y, McEwen BS. Effects of glucocorticoids on hippocampal long-term potentiation. Hippocampus 1993;3:183– 92.

[153] Pavlides C, Kimura A, Magarinos AM, McEwen BS. Hippocampal homosynaptic long-term depression/depotentiation induced by adrenal steroids. Neuroreport 1994;5:2673 – 7.

[154] Petrides M, Milner B. Deficits on subject-ordered tasks after frontal- and temporal-lobe lesions in man. Neuropsychologia 1982;31:1 –15.

[155] Przybyslawski JP, Roullet P, Sara SJ. Attenuation of emotional and nonemotional memories after their reactivation: role of !-adrenergic receptors. J Neurosci 1999;19:6623 – 8.

[156] Quatermain D, McEwen BS, Azmita E. Recovery of memory following amnesia in rats and mouse. J Comp Physiol Psychol 1972;79:360 – 79.

[157] Rajkowska G, Miguel-Hidalgo JJ, Wei J, Dilley G, Pittman SD, Meltzer HY, Overholser JC, Roth BS, Stockmeier CA. Morphometric evidence for neuronal and glial prefrontal cell pathology in major depression. Biol Psychiatry 1999;45:1085 – 98.

[158] Reul JMHM, De Kloet ER. Two receptor systems for corticosterone in rat brain: microdistribution and differential occupation. Endocrinology 1985;117:2505 – 12.

[159] Ricci PT, Zelkowicz BJ, Nebes RD, Meltzer CC, Mintun MA, Becker JT. Functional neuroanatomy of semantic memory: recognition of semantic associations. Neuroimage 1999;9:88 – 96.

[160] Roland PE, Guly B. Visual memory, visual imagery, and visual recognition of large field patterns by the human brain: functional anatomy by positron emission tomography. Cereb Cortex 1995;5:79– 93.

S.J. Lupien, M. Lepage / Beha!ioural Brain Research 127 (2001) 137–158 [161] Rombouts S, Machielsen WCM, Witter MP, Barkhof F, Lindeboom J, Scheltens P. Visual association encoding activates the medial temporal lobe: a functional magnetic resonance imaging study. Hippocampus 1997;7:594 – 601.

[162] Roozendaal B, McGaugh JL. The memory –modulatory effects of glucocorticoids depend on an intact stria terminalis. Brain Res 1996;709:243 – 50.

[163] Rugg MD, Fletcher PC, Allan K, Frith CD, Frackowiak RSJ, Dolan RJ. Neural correlates of memory retrieval during recognition memory and cued recall. Neuroimage 1998;8:262 –73.

[164] Rugg MD, Fletcher PC, Frith CD, Frackowiak RSJ, Dolan RJ. Brain regions supporting intentional and incidental memory: a PET study. Neuroreport 1997;8:1283 –7.

[165] Rupprecht R, Reul JM, van Steensel B, Spengler D, Soder M, Berning B, Holsboer F, Damm K. Pharmacological and functional characterization of human mineralocorticoid and glucocorticoid receptor ligands. Eur J Pharmacol 1993;247:145 – 54.

[166] Ryan L, Nadel L, Keil K, Putnam K, Schnyer D, Trouard T, Moscovitch M. The hippocampal complex and retrieval of recent and very remote autobiographical memories: evidence from functional magnetic resonance imaging in neurologically intact people. Hippocampus, in press.

[167] Sanchez MM, Young LJ, Plotsky PM, Insel TR. Distribution of corticosteroid receptors in the Rhesus brain: relative absence of glucocorticoid receptors in the hippocampal formation. J Neurosci 2000;20:4657 –68.

[168] Sapolsky R. Stress, the Aging Brain, and the Mechanisms of Neuron Death. Cambridge: MIT Press, 1992:423.

[169] Sapolsky R, Meaney MJ. Maturation of the adrenocortical stress response: neuroendocrine control mechanisms and the stress hyporesponsive period. Brain Res 1986;396:64 –76.

[170] Sapolsky RM, Krey LC, McEwen BS. Glucocorticoid-sensitive hippocampal neurons are involved in terminating the adrenocortical stress response. Proc Natl Acad Sci USA 1984;81:6174 – 7.

[171] Sapolsky RM, Krey LC, McEwen BS. The neuroendocrinology of stress and aging: the glucocorticoid cascade hypothesis.

Endocr Rev 1986;7:284 –301.

[172] Sapolsky RM, Armanini MP, Packan DR, Sutton SW, Plotsky PM. Glucocorticoid feedback inhibition of adrenocorticotropic hormone secretagogue release: relationship to corticosteroid receptor occupancy in various limbic sites. Neuroendocrinology 1990;51:328 – 36.

[173] Sara SJ. Recovery from hypoxia and ECS induced amnesia after a single exposure to the training environment. Physiol Behav 1973;9:85 – 9.

[174] Sara SJ. Retrieval and reconsolidation: toward a neurobiology of remembering. Learn Mem 2000;7:73 – 84.

[175] Sara SJ, David-Remacle M, Lefevre D. Passive avoidance behavior in rats after electroconvulsive shock: facilitative effect of response retardation. J Comp Physiol Psychol 1975;89:489 – 97.

[176] Sarrieau AS, Dussaillant M, Agid F, Moguilewsky M, Philibert D, Agid Y, Rostene WH. Radioautographic localization of glucocorticosteroid and progesterone ginding sites in the human post-mortem brain. J Steroid Biochem 1986;25:717 –21.

[177] Sarrieau A, Dussaillant M, Sapolsky RM, Aitken DH, Olivier A, Lal S, Rostene WH, Quirion R, Meaney MJ. Glucocorticoid binding sites in human temporal cortex. Brain Res 1988;442:159 – 63.

[178] Schacter DL, Alpert NM, Savage CR, Rauch SL, Albert MS. Conscious recollection and the human hippocampal formation: evidence from positron emission tomography. Proc Natl Acad Sci USA 1996;93:321 –5.

[179] Schacter DL, Reiman E, Uecker A, Polster MR, Yun LS, Cooper LA. Brain regions associated with retrieval of structurally coherent visual information. Nature 1995;376:587 – 90.

157

[180] Schacter DL, Wagner AD. Medial temporal lobe activations in fMRI and PET studies of episodic encoding and retrieval.

Hippocampus 1999;9:7 –24.

[181] Schlessinger AR, Cowan WM, Gottlieb DI. An autoradiographic study of the time of origin and the pattern of granule cell migration in the dentate gyrus of the rat. J Comp Neurol 1975;159:149 – 75.

[182] Schmidt LA, Fox NA, Goldbert MC, Smith CC, Shulkin J. Effects of acute prednisone administration on memory, attention and emotion in healthy human subjects. Psychoneuroendocrinology 1999;24:461 – 83.

[183] Scoville WB, Milner B. Loss of recent memory after bilateral hippocampal lesions. J Neurol Neurosurg Psychiatry 1957;20:11 – 21.

[184] Seckl JR, Dickson KL, Yates C, Fink G. Distribution of glucocorticoid and mineralocorticoid receptor messenger RNA expression in human postmortem hippocampus. Brain Res 1991;561:332 – 7.

[185] Seckl JR, French KL, O’Donnell D, Meaney MJ, Nair NP, Yates CM, Fink G. Glucocorticoid receptor gene expression is unaltered in hippocampal neurons in Alzheimer’s disease. Brain Res Mol Brain Res 1993;18:239 –45.

[186] Sgoifo A, de Boer SF, Haller J, Koolhaas JM. Individual differences in plasma catecholamines and corticosterone stress responses of wild-type rats: relationship with aggression. Physiol Behav 1996;60:1403 –7.

[187] Sheline Y, Wang P, Gado M, Csernansky J, Vannier M. Hippocampal atrophy in recurrent major depression. PNAS 1996;93:3908 – 4003.

[188] Sheline YI, Sanghavi M, Mintun MA, Gado MH. Depression duration but not age predicts hippocampal volume loss in medically healthy women with recurrent major depression. J Neurosci 1999;19:5034 – 43.

[189] Shimizu E, Kodama K, Sakamoto T, Komatsu N, Yamanouchi N, Okada S, Sato T. Recovery from neuroendocrinological abnormalities and frontal hypoperfusion after remission in a case with rapid cycling bipolar disorder. Psychol Clin Neurosci 1997;51:207 – 12.

[190] Shors TJ, Miesegaes G, Beylin A, Zhao M, Rydel T, Gould E. Neurogenesis in the adult is involved in the formation of trace memories. Nature 2001;410:372 – 5.

[191] Squire LR. Memory and the hippocampus: a synthesis from findings with rats, monkeys, and humans. Psychol Rev 1992;99:195 – 231.

[192] Squire LR, Ojemann JG, Miezin FM, Petersen SE, Videen TO, Raichle ME. Activation of the hippocampus in normal humans: a functional anatomical study of memory. Proc Natl Acad Sci USA 1992;89:1837 – 41.

[193] Stansbury K, Haley D, Koeneker A. Higher cortisol values facilitate spatial memory in toddlers. Ann New York Acad Sci 2000;911:456 – 8.

[194] Starkman MN, Gebarski SS, Berent S, Schteingart DE. Hippocampal formation volume, memory dysfunction, and cortisol levels in patients with Cushing’s syndrome. Biol Psychiatry 1991;32:756 – 65.

[195] Starkman MN, Giordani B, Gebarski SS, Berent S, Schork MA, Schteingart DE. Decrease in cortisol reverses human hippocampal atrophy following treatment of Cushing’s disease.

Biol Psychiatry 1999;46:1595 – 602.

[196] Stein M, Koverola C, Hanna C, Torchia M, McClarty B. Hippocampal volume in women victimized by childhood sexual abuse. Psychol Med 1997;27:951 – 61.

[197] Stern CE, Corkin S, Gonzalez RG, Guimaraes AR, Baker JR, Jennings PJ, Carr CA, Sugiura RB, Vedantham V, Rosen BR. The hippocampal formation participates in novel picture encoding: evidence from functional magnetic resonance imaging. Proc Natl Acad Sci USA 1996;93:8660 – 5.

158

S.J. Lupien, M. Lepage / Beha!ioural Brain Research 127 (2001) 137–158

[198] Sutanto W, Van Eekelen JAM, Reul JMHM, de Kloet ER. Species-specific topography of corticosteroid receptor types in rat and hamster brain. Neuroendocrinology 1988;47:398 –404.

[199] Sutanto W, de Kloet ER. Species specificity of corticosteroid receptors in rat and hamster brain. Endocrinology 1987;121:1405 – 11.

[200] Suzuki WA, Clayton NS. The hippocampus and memory: a comparative and ethological perspective. Curr Opin Neurobiol 2000;10:768 – 73.

[201] Tohgi H, Utsugisawa K, Yamagata M, Yoshimura M. Effects of age on messenger RNA expression of glucocorticoid, thyroid hormone, androgen, and estrogen receptors in postmortem human hippocampus. Brain Res 1995;700:245 –53.

[202] Tulving E, Kroll N. Novelty assessment in the brain and long-term memory encoding. Psychonom Bull Rev 1995;2:387 – 90.

[203] Tulving E, Markowitsch HJ, Craik FIM, Habib R, Houle S. Novelty and familiarity activations in PET studies of memory encoding and retrieval. Cereb Cortex 1996;6:71 –9.

[204] Tulving E, Markowitsch HJ, Kapur S, Habib R, Houle S. Novelty encoding networks in the human brain: positron emission tomography data. Neuroreport 1994;5:2525 –8.

[205] Tyler TJ, Discenna P. Long-term potentiation. Ann Rev Neurosci 1987;10:131 – 61.

[206] Uno H, Tarara R, Else JG, Suleman MA, Sapolsky RM. Hippocampal damage associated with prolonged and fatal stress in primates. J Neurosci 1989;9:1705 –11.

[207] Vandenberghe R, Price CJ, Wise R, Josephs O, Frackowiak RSJ. Functional anatomy of a common semantic system for words and pictures. Nature 1996;383:254 – 6.

[208] Van Cauter E, Turek FW. Endocrine and other biological rhythms. In: DeGroot LJ, editor. Endocrinology. Philadelphia: Saunders, 1994:2487 –548.

[209] Vedhara K, Hyde J, Gilchrist ID, Tytherleigh M, Plummer S. Acute stress, memory, attention and cortisol. Psychoneuroendocrinology 2000;25:535 –49.

[210] Veldhuis HD, Van Koppen C, Van Ittersum M, de Kloet ER. Specificity of the adrenal steroid receptor system in rat hippocampus. Endocrinology 1982;110:2044 – 51.

[211] Veldhuis HD, De Korte CCMM, De Kloet ER. Glucocorticoids facilitate the retention of acquired immobility during forced swimming. Eur J Harm 1985;115:211 –7.

[212] Wagner AD, Schacter DL, Rotte M, Koutstaal W, Maril A, Dale AM, Rosen BR, Buckner RL. Building memories: remembering and forgetting of verbal experiences as predicted by brain activity. Science 1998;281:1188 – 91.

[213] Walker D, Barnes D, Zornetzer S, Hunter B, Kubanis P. Neuronal loss in hippocampus induced by prolonged ethanol consumption in rats. Science 1980;209:711 – 4.

[214] Watanabe Y, Gould E, McEwen BS. Stress induces atrophy of apical dendrites of hippocampal CA3 pyramidal neurons. Brain Res 1992;588:341 – 5.

[215] Watanabe Y, McKittrick CR, Blanchard DC, Lanchard RJ, McEwen BS, Sakai RR. Effects of chronic social stress on tyrosine hydroxylase mRNA and protein levels. Mol Brain Res 1995;32:176 – 80.

[216] Wetzel DM, Bohn MC, Kazee AM, Hamill RW. Glucocorticoid receptor mRNA in Alzheimer’s diseased hippocampus.

Brain Res 1995;679:72 – 81.

[217] Wiggs CL, Weisberg J, Martin A. Neural correlates of semantic and episodic memory retrieval. Neuropsychologia 1999;37:103 – 18.

[218] Wilson M, Greer M, Roberts L. Hippocampal inhibition of pituitary – adrenocortical function in female rats. Brain Res 1980;197:344 – 51.

[219] Wolkowitz OM, Reus VI, Weingartner H, Thompson K, Breier A, Doran A, Rubinow D, Pickar D. Cognitive effects of corticosteroids. Am J Psychiatry 1990;147:1297 – 303.

[220] Wolkowitz OM, Weingartner H, Rubinow DR, Jimerson D, Kling M, Berretini W, Thompson K, Breier A, Doran A, Reus VI, Pickar D. Steroid modulation of human memory: biochemical correlates. Biol Psychiatry 1993;33:744 – 6.

[221] Woolley CS, Gould E, McEwen BS. Exposure to excess glucocorticoids alters dendritic morphology of adult hippocmpal pyramidal neurons. Brain Res 1990;531:225 – 31.

[222] Yehuda R. Biology of posttraumatic stress disorder. J Clin Psychiatry 2000;61:14 – 21.

[223] Young AH, Sahakian BJ, Robbins TW, Cowen PJ. The effects of chronic administration of hydrocortisone on cognitive function in normal male volunteers. Psychopharmacology 1999;145:260 – 6.

[224] Zennaro MC, Keightley MC, Kotelvtsev Y, Conway GS, Soubrier F, Fuller PJ. Human mineralocorticoid receptor genomic structure and identification of expressed isoforms. J Biol Chem 1995;270:21016 – 20.