Neurochemical Research, Vol. 25, Nos. 9/10, 2000, pp. 1219–1231

Allostasis, , and the Aging : Role of Excitatory Amino Acids and Excitotoxicity*

Bruce S. McEwen1

(Accepted March 15, 2000)

The adaptive responses of the body to challenges, often known as “stressors”, consists of active re- sponses that maintain . This process of adaptation is known as “”, meaning “achieving stability through change”. Many systems of the body show allostasis, including the au- tonomic nervous system and hypothalamo-pituitary-adrenal (HPA) axis and they help to re-estab- lish or maintain homeostasis through adaptation. The brain also shows allostasis, involving the ac- tivation of nerve cell activity and the release of . When the individual is challenged repeatedly or when the allostatic systems remain turned on when no longer needed, the mediators of allostasis can produce a wear and tear on the body that has been termed “allostatic load”. Ex- amples of allostatic load include the accumulation of abdominal fat, the loss of bone minerals and the atrophy of nerve cells in the hippocampus. Circulating hormones play a key role, and, in the hippocampus, excitatory amino acids and NMDA receptors are important mediators of neuronal atrophy. The aging brain seems to be more vulnerable to such effects, although there are consider- able individual differences in vulnerability that can be developmentally determined. Yet, at the same time, excitatory amino acids and NMDA receptors mediate important types of plasticity in the hip- pocampus. Moreover, the brain retains considerable resilience in the face of stress, and estrogens appear to play a role in this resilience. This review discusses the current status of work on under- lying mechanisms for these effects.

KEY WORDS: Allostasis; allostatic load; aging brain; excitatory amino acid; excitotoxicity.

INTRODUCTION hormones. In the nervous system, for example, neuro- transmitters are released by neuronal activity, and they When the body is challenged by unexpected or threat- produce effects locally to either propagate or inhibit fur- ening events, it reacts physiologically in an adaptive ther neural activity. Neurotransmitters and hormones are manner in order to maintain homeostasis. This process is usually released during a discrete period of activation called “allostasis”, literally “maintaining stability through and then are shut off, and the mediators themselves are change” (1) and it involves the production and/or release removed from the intracellular space by reuptake or me- of physiological mediators such as adrenalin from the ad- tabolism in order not to prolong their effects. When the renal medulla and glucocorticoids from the adrenal cor- shut off or removal of the mediator does not occur, ef- tex. However, allostasis also applies to organs and tis- fects of the mediators on target cells are prolonged, lead- sues of the body, as well as the production of systemic ing to other consequences that may include receptor de- sensitization and tissue damage. This process has been 1 Harold and Margaret Milliken Hatch Laboratory of Neuroendocrinol- named “allostatic load” (2,3), and it refers to the price the ogy, Rockefeller University, 1230 York Avenue, New York, N.Y. 10021. Fax: 212 327 8634; E-mail: [email protected] tissue or organ pays for an overactive or inefficiently * Special issue dedicated to the 25th anniversary of Neurochemical Re- managed allostatic response. Therefore, allostatic load search. refers to the “cost” of adaptation. 1219 0364-3190/00/09/1000–1219$18.00/0 © 2000 Plenum Publishing Corporation 1220 McEwen

The processes of allostasis and allostatic load have cium channels increases in hippocampal CA1 pyramidal been described and measured for metabolic and cardio- neurons of aging rats and results in an increased after- vascular changes that are associated with obesity, Type 2 hyperpolarization (18). Some of this can be mimicked in diabetes and cardiovascular (4). However, ele- a cell culture system. In embryonic hippocampal neurons vated and prolonged secretion of glucocorticoids during that are maintained for 28d in cell culture, there is en- aging has also been associated with impairment of cog- hanced calcium channel activity and increased after- nitive function in rodents (5–7) and in humans (8–10). hyperpolarization that are accompanied by decreased Moreover, the endogenous excitatory amino acid neuro- neuronal survival; blocking L-type calcium channels in- transmitters appear to play a major role in these changes creased neuronal survival (19). It is interesting to note (7) even though they are also an essential part of normal that the increased after-hyperpolarization is associated synaptic neurotransmission and plasticity. Their actions with alterations of two important neurophysiologic re- lead to the formation of free radicals that can damage sponses in CA1 pyramidal neurons of the hippocampus, nerve cells, leading to the search for agents that can in- namely, enhanced induction of long-term depression terfere with free radical production or enhance free rad- (LTD) and an impaired induction of long-term potentia- ical quenching. tion (LTP)(20). Thus, insofar as LTP and LTD may be In spite of its vulnerability, the brain retains consid- related to synaptic plasticity during learning (21), these erable resilience in the face of challenges to adapt through age-related changes suggest a possible basis for cogni- allostasis. Studies on the hippocampus reveal a number of tive impairment in aging rats (20). types of structural plasticity, ranging from neurogenesis Glucocorticoids enhance calcium channel activity in the dentate gyrus to remodelling of dendrites to the for- and after-hyperpolarization (18;22), and hippocampal mation and replacment of synapses. These changes, along glucocorticoid receptor expression shows a progressive with compensatory neurochemical and neuroendocrine failure of negative feedback regulation in old versus responses, provide the brain with a considerable amount young rats. In young rats, repeated stress causes a down- of resilience. This has led to a search for agents that help regulation of glucocorticoid receptor levels, thus de- the brain maintain its resilience as it ages. This article dis- creasing glucocorticoid efficacy on various target genes, cusses allostasis and allostatic load in the brain in relation whereas this down-regulation is lost with increasing age, to the aging process and a number of brain disorders in thus potentiating glucocorticoid actions, some of which which there is overactivity of stress mediators that causes may be destructive to brain cells (23). Therefore, there is brain dysfunction. It also discusses the topic of neuropro- a natural mechanism in the young hippocampus for re- tection and the potential value of estrogens and flavonoids silience in the face of repeated stress that acts to reduce as anti-oxidants in promoting allostasis and enhancing re- the magnitude of the glucocorticoid feedback signal and silience and countering the allostatic load promoted by thus reduce the impact of glucocorticoids on calcium excitatory amino acids and other free radical generators channel activity, among other effects. This may be pro- such as the beta amyloid protein. tective, insofar as increased calcium channel activity Age-Related Shifts of Calcium Homeostasis and Its contributes to free radical generation and other processes Consequences. The hippocampus is a brain region that is that may damage neurons (24,25). With the loss of very important for declarative and spatial learning and stress-induced down-regulation of glucocorticoid recep- memory, and yet is a particularly vulnerable and sensitive tors, older rats appear to lose this protective device and region of the brain that expresses high levels of receptors may be more vulnerable to increased levels of glucocor- for adrenal steroid “stress” hormones (11,12). Hippocam- ticoids, particularly in cognitively-impaired rats (23). pal neurons are vulnerable to seizures, strokes and head It is still unclear whether outright neuronal loss is a trauma, as well as responding to stressful experiences major event in the aging hippocampus of cognitively- (12–14). At the same time these neurons show remarkable impaired rats ((26,27); see (28) for review). Neverthe- and paradoxical plasticity, involving long-term synaptic less, there are indications that gene products associated potentiation and depression, dendritic remodeling, synap- with neurodegeneration and damage are differentially tic turnover and neurogenesis in the case of the dentate regulated in the aging-impaired brain compared to un- gyrus (15–17). This will be discussed further below. impaired aging rats and young rats, although the inter- Studies in animal models have shown that the hip- pretation of the results is very complex (29). In aging, pocampus undergoes progressive changes with age in cognitively-impaired rats, the levels of mRNA for the calcium homeostasis, in the plasticity of response to glu- 695 amino acid form of the beta amyloid precursor pro- cocorticoids, and in the expression of markers related to tein (betaAPP) and for the magnesium-dependent super- neuroprotection and damage. The activity of L-type cal- oxide dismutase (Mg-SOD) were both elevated through- Allostasis, Allostatic Load, and the Aging Nervous System 1221 out the hippocampus compared with young rats; at the production and/or removal of a mediator of neuronal ac- same time the levels of the betaAPP protein and Mg-SOD tivation. See Fig. 2. protein were both depressed (29). Levels of mRNA for Free radical formation is a by-product of excitatory glial fibrillary acidic protein (GFAP), a marker of astro- amino acid release and a consequence of the activation cytes which increases with damage, were elevated in the of second messenger systems (24;36). A key factor in hippocampus of aging, cognitively impaired rats, although regulating production of free radicals is the maintenance the level of the GFAP protein was not elevated (29). Since of homeostasis of calcium ions (37). When excitatory betaAPP gives rise to both a toxic beta amyloid protein amino acid neurotransmitters are released, calcium ions and a protective secreted form, the reduced levels of be- are mobilized via activation of NMDA and AMPA re- taAPP expression in aging, cognitively impaired rats is ceptors, and second messenger systems are activated difficult to interpret without a separate measurement of the leading to a cascade of effects including the long-term two forms of the protein. On the other hand, lower Mg- potentiation and long-term depression that are believed SOD protein is consistent with a lower capacity for free- to be related to information storage mechanisms (38,39). radical scavenging and an increased risk for free-radical The reuptake and rebuffering of calcium ions is an induced neural damage (30). active process (37;40). If the calcium ions are not re- Next, although the role of glucocorticoids in pro- moved and put back into intracellular stores rapidly and moting these changes is still under investigation, it is im- efficiently, the cascade of events is potentiated and can portant to consider how progressive changes with age in result in the increased accumulation of free radicals and these indices of damage fit into a broader view of the role free-radical induced by-products of lipid peroxidation of the hypothalamo-pituitary-adrenal (HPA) axis in indi- that can produce an allostatic load on brain and cardio- vidual differences in the aging process. vascular cells (24,37).A scheme depicting some of these Allostasis and Allostatic Load in the Brain. Allosta- events and the role of glucocorticoids is presented in sis is the process of adaptation to challenge that maintains Fig. 3 (36). There is a link between stressful events and stability, or homeostasis, through an active process (1), the production of the free radicals, namely, that acute and allostatic load is the wear and tear produced by the re- stress increases the production of free radicals in the peated activation of allostatic, or adaptive, mechanisms brain and other organs (25). However, the role of gluco- (2;3). Four types of allostatic load have been identified and corticoids in this process is not known. Glucocorticoids are summarized in Fig 1. These consist of: 1) Repeated may play a role in the process by facilitating the activity challenges; 2) Failure to habituate with repeated chal- of NMDA receptors (41,42), by impairing glucose up- lenges; 3) Failure to shut off the response after the chal- take and reducing intracellular energy supplies (7;36) lenge is past; 4) Failure to mount an adequate response. In and by increasing calcium currents (see above), and in- the hippocampus, we can recognize at least two types of dividual differences in glucocorticoid secretion over the allostatic load involving excitatory amino acid release, life-course may thus make a contribution. What are these namely, 1) the potential to cause damage with repeated individual differences in glucocorticoid secretion and stressful challenges and 2) the failure in aging rats to shut how may they come about? off glutamate release after stress. See Fig. 2. Developmental Determinants of Individual Differ- Under restraint stress, rats show increased extracel- ences in Allostatic Load. The vulnerability of many sys- lular levels of glutamate in hippocampus, as determined tems of the body to stress is influenced by experiences by microdialysis, and adrenalectomy markedly attenu- early in life. In animal models, unpredictable prenatal ates this elevation (31). Glucocorticoids appear to be in- stress causes increased emotionality and increased reac- volved in potentiating the increased extracellular levels tivity of the HPA axis and of excitatory amino acids under stress (32). Similar re- and these effects last throughout the lifespan (43). Post- sults have been reported using lactography, a method natal handling in rats, a mild stress involving brief daily that is based upon the stimulation of glucose metabolism separation from the mother, counteracts the effects of by increased neuronal activity (33;34). The conse- prenatal stress and results in reduced emotionality and quences of this increased level of glutamate, extracellu- reduced reactivity of the HPA axis and autonomic nerv- larly, will be discussed below in terms of hippocampal ous system (44–46). dendritic remodelling, which is an example of type 1 al- For prenatal stress and postnatal handling, once the lostatic load. In aging rats, hippocampal release of exci- emotionality and the reactivity of the adrenocortical sys- tatory amino acids during restraint stress is markedly po- tem are established by events early in life, it is the sub- tentiated (35), and this constitutes an example of type 3 sequent actions of the hypothalamo-pituitary-adrenal allostatic load in the brain, i.e., the failure to shut off the (HPA) axis in adult life, as discussed above, that are 1222 McEwen

Fig. 1. The top panel illustrates the normal allostatic response, in which a response is initiated by a stressor, sustained for an appropriate interval, and then turned off. The remaining panels illustrate four conditions that lead to allostatic load: l) Repeated “hits” from multiple novel stressors; 2) Lack of adaptation; 3) Prolonged response due to delayed shut down; and 4) inadequate response that leads to compensatory hyperactivity of other mediators: e.g., inadequate secretion of glucocorticoid, resulting in increased levels of cytokines that are normally counter-regulated by glucocorticoids). Figure drawn by Dr. Firdaus Dhabhar, Rockefeller University. Reprinted from (3) by permission. likely to contribute to the rate of brain and body aging. significant and progressive increase in levels, dur- Rats with increased HPA reactivity show early decline of ing yearly exams, over the 4 years, and had high basal cognitive functions associated with the hippocampus cortisol levels in year 4, showed deficits on tasks meas- (47), as well as increased propensity to self-administer uring explicit memory as well as selective attention, com- drugs such as amphetamine and cocaine (48,49). In con- pared to subjects with either decreasing cortisol levels trast, rats with a lower HPA reactivity as a result of over four years or subjects with increasing basal cortisol neonatal handling have a slower rate of cognitive aging but moderate current cortisol levels (8). Using MRI, they and a reduced loss of hippocampal function (50–52). also showed a hippocampus that was 14% smaller than Thus, life-long patterns of adrenocortical function, deter- age-matched controls who did not show progressive cor- mined by early experience, contribute to rates of brain tisol increases and were not cognitively impaired (10). aging, at least in experimental animals. Adaptive Plasticity—Another Role for Excitatory Evidence for a human counterpart to the story of in- Amino Acids and Hormones. The hippocampus is not dividual differences in rat HPA activity and hippocampal only a vulnerable brain structure to damage but is also a aging is very limited. Individual differences in human very plastic region of the brain and expresses high levels brain aging that are correlated with cortisol levels have of adrenal steroid receptors. Adrenal steroids, which, as been recognized in otherwise healthy individuals that are we have noted, have a bad reputation as far as their role followed over a number of years (8–10). In the most ex- in exacerbating these forms of damage (7), are also in- tensive investigation, healthy elderly subjects were fol- volved in three types of adaptive plasticity in the hip- lowed over a four year period, and those who showed a pocampal formation. Allostasis, Allostatic Load, and the Aging Nervous System 1223

whereas acute non-painful novelty stress inhibits primed- burst potentiation and spatial memory (60;63) and post- training shock stress inhibits recall of a spatial memory task that depends on the hippocampus (64). Second, adrenal steroids participate along with ex- citatory amino acids in regulating neurogenesis of den- tate gyrus granule neurons (15), in which acute stressful experiences can suppress the ongoing neurogenesis(65). These effects may be involved in fear-related learning and memory, because of the anatomical and functional connections between the dentate gyrus and the amygdala (66), a brain area important in memory of aversive and fear-producing experiences (67). Third, adrenal steroids participate along with exci- tatory amino acids in a reversible stress-induced atrophy, or remodeling, of dendrites in the CA3 region of hip- pocampus of male rats and tree shrews (16), a process that affects only the apical dendrites and results in cog- nitive impairment in the learning of spatial and short- term memory tasks (16). Although this type of plasticity does impair cognitive function at least temporarily, it may be beneficial to the brain in the long run if the re- modeling of dendrites reduces the impact of excitatory amino acids and glucocorticoids in causing more perma- nent damage. This is a hypothesis that remains to be rig- orously tested. Besides what stress does to change hippocampal structure, there are other forms of plasticity in the hip- pocampus, including reversible synaptogenesis that is reg- ulated by ovarian steroids in female rats and occurs in the CA1 region (68) and a very rapid and reversible atrophy of dendrites of CA3 neurons during hibernation in ground squirrels and hamsters (69,70). The estrogen-regulated CA1 synaptic plasticity is also a rapid event, occurring during the female rats’ 5 d estrous cycle, with the synap- ses taking several days to be induced under the influence of estrogens and endogenous glutamic acid, and then dis- appearing within 12h under the influence of the proestrus Fig. 2. Effect of 1h immobilization stress on extracellular levels of surge of progesterone (71;72). excitatory amino acids in the hippocampus (A) and medial prefrontal cortex (B) of young (3–4 month old) adult and aging (22–24 month In view of the discussion above, concerning excita- old) rats. Note that both young and old rats released glutamate during tory amino acids and NMDA receptors, it is important to stress, but that during the 2h post stress period, old rats continued to note that the above-mentioned hormone effects on mor- release glutamate whereas young rats did not. From (35) by permission. phology and function of the hippocampus do not occur alone but rather in the context of ongoing neuronal ac- First, they reversibly and biphasically modulate ex- tivity. In particular, excitatory amino acids and NMDA citability of hippocampal neurons and influence the mag- receptors play an important role in the adaptive functional nitude of long-term potentiation, as well as producing and structural changes produced in the hippocampal for- long-term depression (12;53 – 56). These effects may be mation by steroid hormones. This includes not only the involved in biphasic effects of adrenal secretion on ex- estradiol-induced synaptogenesis (72) but also the effects citability and cognitive function and memory during the of adrenal steroids to produce atrophy of CA3 pyramidal diurnal rhythm and after stress (57 – 60). In particular, ad- neurons (16), as well as the actions of adrenal steroids to renal steroids facilitate fear-motivated learning (61;62), contain dentate gyrus neurogenesis (73). Blocking NMDA 1224 McEwen

Fig. 3. Potential mechanisms for glucocorticoid initiation of hippocampal neuronal death. During ischemia and excitotoxicity, and possibly as a result of the aging process (1) Glucocorticoids (GC’s) damage particular hippocampal neuronal populations by inhibiting glucose utilization (b), which decreases the activity of energy-dependent excitatory amino acid (EAA) transporters (c). Consequently, glutamate concentrations in the synapse rise (d), leading to prolonged activation of NMDA receptors (e), and concomitant increases in intracellular Ca++ levels (f). Increased Ca++ concentrations activates Ca++-dependent endonuclease (g), in addition to other Ca++-dependent enzymes (i). Mitochondria attempt to sequester greater concentrations of Ca++ (h), disrupting mitochondrial homeostasis. Mitochondrial dysfunction, as well as sustained activation of Ca++ dependent enzymes, increases the production of reactive free radicals (j). The combined actions of free radicals and endonucleases contribute to DNA strand breakage (k). When DNA damage is too great, p53 promotes neuronal apoptosis (m). Neurons do possess survival mechanisms, including DNA repair enzymes such as PARP (l). However, the actions of other apoptosis-related genes, such as ICE and CPP32, inhibit PARP activity (n). Oxygen radical Allostasis, Allostatic Load, and the Aging Nervous System 1225 receptors prevents atrophy as well as estrogen-induced ovarian hormones that indicate that estrogens may have synaptogenesis (74,75), and NMDA receptors are induced a neuroprotective role in relation to stress and glucocor- by estrogens on CA1 neurons (76,77) and by glucocorti- ticoid secretion and actions in the brain. For example, es- coids throughout the hippocampus (41). At the same time, trogens stimulate neurogenesis in the female dentate as we have already noted, excitatory amino acids and gyrus (84). Moreover, female rats appear to be resistant NMDA receptors are involved in free radical generation to the stress-induced atrophy of hippocampal dendrites leading to neural damage, and one of the challenges for seen in male rats (85). Because there are developmen- future research is to understand what triggers the transi- tally programmed structural and functional sex differ- tion from adaptive plasticity to permanent damage. ences in the hippocampus (86–89), it is not clear whether Adaptive Plasticity and the Concept of Resilience. the estrogen-stimulation of neurogenesis or the lack of We have noted that the young brain is resilient and able atrophy in the hippocampus after stress are processes that to withstand challenges and adapt, and the structural plas- would occur in males if they were castrated and given es- ticity noted above is an example of this resilience and trogens as adults, or if these effects reflect programming adaptability. The term allostasis means adaptation and earlier in life during sexual differentiation. coping and implies resilience. Allostasis operates most Extrapolation of these findings to humans is difficult efficiently when the body is doing its best to maintain because of a lack of comparable studies on the human homeostasis without doing harm. As noted and illustrated brain. However, there is evidence for increased vulnera- above, allostatic load is the cost of adaptation, reflecting bility of postmenopausal women to declines in hippocam- both the overuse of the system by repeated stressors as pal dependent cognitive function that is correlated with well as the inefficient management of allostasis–failure to elevated urinary cortisol (9). Moreover, HPA activity in shut off or habituate; failure to turn on when needed. Here women tends to increase postmenopausally as increased we consider how structural plasticity is related to resilience levels across the diurnal cycle and a flattening of the and how this plasticity may be lost as the brain ages. rhythm, although there were considerable individual dif- Neurogenesis in the dentate gyrus provides an excel- ferences (90). Although in that particular study it was not lent example of resilience in the adult brain (78). Production clear which women, if any, were receiving estrogen re- of new cells is increased by voluntary exercise (79), by an placement therapy, a recent study indicates that ERT does enriched environment (80) and by estrogens, as noted reduce both HPA reactivity and sympathetic nervous sys- above. Dendritic remodeling by repeated stress provides an- tem reactivity (91), both measures indicating that estrogens other example of resilience, since it is a reversible process may reduce allostatic load that can exacerbate cardiovas- that may protect nerve cells from permanent damage cular disease, hypertension and abdominal obesity (3). (16;81). Down-regulation of glucocorticoid receptors in This is a new area of study and much more needs to response to repeated stress (82) is another example of a pro- be done, but the hypothesis that arises from the available tective response, since glucocorticoids exacerbate per- data is that estrogens work to contain the HPA axis and manent damage to hippocampal nerve cells (7). to counteract some of the potentially damaging actions of Although we do not have examples for each type of glucocorticoids on nerve cells. Recent support for this structural plasticity described above for the aging brain, latter notion comes from studies showing that estrogens a number of examples exist to show that plasticity is fre- reduce excitotoxic damage and that glucocorticoids in- quently lost in the aging brain. This is the case for neu- crease it (92). In vivo studies have shown that estrogens rogenesis (83). It is also the case for the down-regulation reduce damage produced by ischemia in an animal model of glucocorticoid receptors in the hippocampus (18). We of stroke, in which excitotoxicity is involved (93;94) and do not know yet if dendritic remodeling is lost with age. in which it is known that glucocorticoids exacerbate Nor do we know whether estrogen-induced synapse for- ischemic damage (95). mation persists in aging animals. These are important re- Resilience of the Brain in the Face of Stress and search questions for the future. Allostatic Load. We have seen that stress and glucocor- Stress and Estrogen Interactions on Brain Func- ticoids act in concert with excitatory amino acids to mod- tion. There are a number of points of interaction with ulate the branching of dendrites in the hippocampus of

scavengers such as SOD also rescue neurons from apoptosis (2); bcl-2 promotes neuronal survival by increasing mitochondrial Ca2+ buffering capacity (3). In this manner, bcl-2 promotes neuronal survival as an antioxidant by maintaining mitochondrial homeostasis, thereby increasing decreasing reactive free radical production. Such a cascade is unlikely for adrenalectomy (ADX)-induced granule cell loss in the dentate gyrus (DG); however, increased p53 expression is associated with apoptotic granule neurons (1). Therefore, p53 expression may represent a common final pathway by which GC-facilitated and ADX-mediated cell death converge. Reprinted from (36) by permission. 1226 McEwen

Table I. Actions of Estrogens Related to Excitability and Cell Membrane Events

Membrane binding sites - identified but not well-characterized in pituitary, liver and endometrium, but not in brain. Some membrane sites may be related to intracellular ER. Genomic effects on membrane events, e.g.: Induction of the MINK potassium channel in pituitary via genomic mechanism. Calcium channel expression in pituitary and hippocampus. Apparent non-genomic actions, e.g.: Rapid excitation of electrical activity in cerebellum, hippocampus, striatum and cerebral cortex. Effects occur within seconds and are unlikely to involve a transcriptional activation. Second messenger activation CREB phosphorylation: genomic vs. non-genomic mechanism unclear. MAP kinase activation: possible novel receptor pathway or involvement of classical ER in a novel signalling pathway. Calcium homeostasis Rapid actions: 17b E is more potent, but tamoxifen is an agonist on Ca ++ currents. Rapid actions: 17a E is as potent as 17b E on calcium entry. Possible genomic actions: delayed and sustained increase in Ca channel activity. Neuroprotection Rapid actions: 17a Estradiol is as potent as 17b estradiol vs. oxidative damage. Genomic actions: 17b estradiol is more potent; anti-estrogen blockade.

Examples are provided for each topic. For detailed summary, see (17). Note that these estrogen actions are not mutually exclusive but may repre- sent different endpoints of interacting intracellular signalling.

Reprinted from (126) experimental animals and the replacement of neurons in activated protein kinase (MAP kinase) pathways, effects the dentate gyrus (16). Atrophy of dendrites and inhibi- on calcium channels and calcium ion entry, and protec- tion of neurogenesis caused by stress compromises cog- tion of neurons from damage by excitotoxins and free nitive functions that depend on the hippocampus, such as radicals (17). See Table I and Fig. 4. These estrogen ac- spatial, declarative and contextual memory. However, tions occur through at least two types of intracellular re- these effects are reversible, along with the morphological ceptors, ER alpha and ER beta, as well as a handful of changes, as long as the stress is terminated after a num- other mechanisms for which receptor sites are not clearly ber of weeks - much longer periods of stress may cause identified (17). Indeed, for a number of processes, there permanent damage to the hippocampus (96). Thus the are conflicting reports, based upon estrogen structure- brain is resilient and capable of adaptive plasticity, and activity studies and the actions of estrogen antagonists, as we must consider the role of endogenous mediators of re- to whether intracellular receptors are involved. Thus, for silience. Such mediators include estrogens and other sub- estrogen actions on some aspects of calcium homeo- stances that can reduce the allostatic load generated by stasis, certain aspects of second messenger systems and excitatory amino acids and they also include genes that some features of neuroprotection, a novel receptor mech- afford neuroprotection. anism is implicated, in which stereospecificity for 17β Estrogens, Flavonoids and Neuroprotection: Be- over 17α estradiol is replaced by a broader specificity for sides the protective effects of estrogen treatment on is- the 3 hydroxyl group on the A ring (for review, see (17)). chemic damage, noted above, there is increasing evi- Some of these actions of estrogens appear to reduce the dence that estrogens reduce the risk for Alzheimer’s production of or actions of free radicals in causing cell disease (97–99). Thus the search for neuroprotective damage and promoting cell death through apoptosis mechanisms has intensified, and multiple mechanisms (92,100). have been uncovered for estrogen action in the brain. Flavonoids are potentially useful exogenous agents The variety of estrogen effects has been expanded to in- in protecting the aging brain and other organs and tissue clude rapid actions on excitability of neuronal and pitu- of the body against free-radical induced damage (101). itary cells, the activation of cyclic AMP and mitogen- Flavonoids include substances that are estrogenic (102) Allostasis, Allostatic Load, and the Aging Nervous System 1227

with the continuing investigation of neuroprotection by estrogen-replacement therapy towards Alzheimer’s disease (98,99,108–112). Indeed, flavonoids, along with estradiol and other antioxidants, may be useful agents to protect the brain without preventing the normal plastic- ity that the same systems, involving NMDA receptors, calcium ions and circulating glucocorticoids, are medi- ating. Other strategies, involving direct interference with NMDA receptors or calcium channels, or glucocorticoid secretion or action, may have the effect of disrupting the normal processes and impairing cognitive and other im- portant functions even if they are effective in retarding permanent damage. One major weakness with the story of flavonoids and other antioxidants is that there is very little data col- lected in vivo on animal models or human subjects. However, recent data on a rat aging model lends support to the efficacy of plant antioxidant compounds (113). In this particular study, treatment of 6 month old rats for 8 months with a number of fruit or vegetable extracts or Vitamin E conferred some protection against age-related decline of cognitive function and a number of neuro- chemical processes (113). Flavonoids are known to be among the antioxidant substances in such extracts. More- over, a recent report on human subjects indicates that an extract of Gingko biloba, which contains flavonoids, sta- bilized and even improved the cognitive performance Fig. 4. Schematic diagram of intracellular estrogen action via ERα and ERβ, as well as possible cell surface effects of a putative membrane and social functioning of demented patients for 6 months estrogen receptors that produce neuro-protection (Top) or affect to one year (114). intracellular signalling (Bottom) via the cyclic AMP and MAP kinase Nevertheless, caution is in order since flavonoids pathways. Top Panel: Estradiol exerts its effects intracellularly via two principal receptor types, ERα and ERβ, and these are characterized by may have potentially deleterious actions that may be re- a distinct specificity for estradiol 17α over estradiol 17β. Estrogens lated to their partial agonist/antagonist profile, re: estro- also exert neuroprotective effects in part via a mechanism in which gen receptors and their demonstrated ability to exert op- estradiol 17α has equal or greater potency compared to estradiol 17β. Bottom panel: Estradiol acts either via cell surface receptors or an posite effects to that of estradiol(115). Another recent intracellular estrogen receptor to activate two different second report indicates that they lower circulating levels of messenger pathways, one involving the MAP kinase cascade and the estradiol in women(116). other involving cyclic AMP. Both pathways result in activation of gene transcription via at least 3 possible response elements: CRE, Genes and Vulnerability: In spite of the need to de- SRE and AP-1. Note that in the case of intracellular second velop exogenous neuroprotective strategies to treat brain messengers, there is some uncertainty concerning the involvement of damage associated with aging, it is important to note that ERα and ERβ in the signalling process versus the role of other, as yet uncharacterized, receptors (see text). Abbreviations: AC, adenylate the brain is normally resilient in the face of acute and re- cyclase; cAMP, cyclic-3′,5′-AMP; CREB-P, phosphorylated form of peated stress, indicating that there are protective factors CREB; ras, ras oncogene; MAPK, mitogen activated protein kinase; that promote resilience of brain cells over the life-span MAPKK, mitogen activated protein kinase kinase; fos-jun, fos-jun heterodimer. This figure republished from(17) by permission. Please and in the face of stressful challenges. How can these see this reference for details. protective factors be identified and studied, and how can a resilient brain be made vulnerable to allostatic load? This is an important question for study of animal models, and estrogenic substances have neuroprotective effects and there are a number of approaches. For example, against free-radical damage which may occur via another transgenic mice are beginning to be used in identifying mechanism, in addition to the traditional intracellular es- factors that promote vulnerability or protection. Among trogen receptors, based upon the structure-activity profile these are the p53 tumor suppressor gene that triggers (103–107). Thus, the further study of flavonoids and apoptosis and death of cells when DNA damage is large; their neuroprotective effects may have some overlap knock out of the p53 gene protects brain cells from 1228 McEwen epileptic and other damage, both in vivo and in vitro REFERENCES (117,118). On the other hand, deletion of the superoxide dismutase gene increases vulnerability of the hippocam- 1. Sterling, P. and Eyer, J. 1988. Allostasis: A New Paradigm to pus to ischemia-induced damage (30;119). Another im- Explain Arousal Pathology. In Fisher, S., and J. Reason, eds. Handbook of Life Stress, Cognition and Health. New York, portant gene is bcl-2 which plays a key role in maintain- John Wiley & Sons., 629–649. ing mitochrondrial calcium homeostasis (see (36) for 2. McEwen, B. S. and Stellar, E. 1993. Stress and the Individual: review). Thus, one strategy in studying protective factors Mechanisms leading to disease. Archives of Internal Medicine 153:2093–2101. is to manipulate genes that are likely to provide protec- 3. McEwen, B. S. 1998. Protective and Damaging Effects of Stress tion, such as the neurotrophins or superoxide dismutase; Mediators. New England J. Med. 338:171–179. mice with deficiencies in these genes should be more 4. Seeman, T. E., Singer, B. H., Rowe, J. W., Horwitz, R. I., and McEwen, B. S. 1997. Price of adaptation—allostatic load and its vulnerable to age- and stress-induced damage of hippo- health consequences: MacArthur studies of successful aging. campal structure and function, and studies are underway Arch. Intern. Med. 157:2259–2268. to test the validity of this strategy. 5. Landfield, P., Waymire, J., and Lynch, G. 1978. Hippocampal aging and adrenocorticoids: quantitative correlation. Science Another approach is to manipulate metabolic fac- 202:1098–1101. tors, e.g., by making rats diabetic or by stressing animals 6. Landfield, P. 1987. Modulation of brain aging correlates by that have genetic risk for either Type I or Type II dia- long-term alterations of adrenal steroids and neurally-active peptides. Prob. Brain Res. 72:279–300. betes. Some initial results indicate that diabetes may ac- 7. Sapolsky, R. 1992. Stress, the Aging Brain and the Mechanisms celerate stress-induced dendritic atrophy in hippocampus of Neuron Death. Cambridge MIT Press 1:423. and promote stress-induced neuronal damage (118,120) 8. Lupien, S., Lecours, A. R., Lussier, I., Schwartz, G., Nair, N. P. V., and Meaney, M. J. 1994. Basal cortisol levels and cog- A third approach is to use hippocampal cell culture nitive deficits in human aging. J. Neurosci. 14:2893–2903. models and study the interaction of androgens, estrogens, 9. Seeman, T. E., McEwen, B. S., Singer, B. H., Albert, M. S., and glucocorticoids and excitatory amino acids in producing Rowe, J. W. 1997. Increase in Urinary Cortisol Excretion and Memory Declines: MacArthur Studies of Successful Aging. excitotoxic damage (7,19,121,122). As noted above, the J. Clin. Endocr. Metab. 82:2458–2465. vulnerability to excitotoxicity in hippocampal neurons 10. Lupien, S. J., DeLeon, M. J., De Santi, S., Convit, A., Tarshish, has been related to increased calcium channel activity C., Nair, N. P. V., Thakur, M., McEwen, B. S., Hauger, R. L., and Meaney, M. J. 1998. Cortisol levels during human aging that develops with increasing age in culture (19). The predict hippocampal atrophy and memory deficits. Nature Neu- cell culture approach has been extended recently to dem- roscience 1:69–73. onstrate the protective effects of another steroid that de- 11. Eichenbaum, H. 1997. How does the brain organize memories? Science 277:330–332. clines with age in humans, namely, dehydroepiandros- 12. DeKloet, E. R., Vreugdenhil, E., Oitzl, M. S., and Joels, M. terone (DHEA), towards NMDA-induced neurotoxicity 1998. Brain corticosteroid receptor balance in health and dis- (123,124) A fourth strategy is to use targeted delivery of ease. Endocr. Rev. 19:269–301. 13. Sapolsky, R., Krey, L., and McEwen, B. S. 1986. The neuroen- genetic material in viral vectors in order to overcome the docrinology of stress and aging: the glucocorticoid cascade hy- restrictions of energy supply in the face of excitotoxic pothesis. Endocr. Rev. 7:284–301. challenge using local enhancement of glucose transporter 14. McEwen, B. S., Albeck, D., Cameron, H., Chao, H. M., Gould, E., Hastings, N., Kuroda, Y., Luine, V., Magarinos, A. M., activity (125). McKittrick, C. R., Orchinik, M., Pavlides, C., Vaher, P., Watan- abe, Y., and Weiland, N. 1995. Stress and the Brain: A Para- doxical Role for Adrenal Steroids. In Litwack, G. D., ed. Vitamins CONCLUSION And Hormones. Academic Press, Inc., 371–402. 15. Cameron, H. A. and Gould, E. 1996. The Control of Neuronal Birth and Survival. In Shaw, C. A., ed. Receptor Dynamics in In conclusion, the wear and tear (allostatic load) that the Neural Development. New York, CRC Press, 141–157. body experiences as a result of stressful life experiences 16. McEwen, B. S. 1999. Stress and hippocampal plasticity. Annu. Rev. Neurosci. 22:105–122. is a highly individual matter, dependent on genotype, 17. McEwen, B. S. and Alves, S. H. 1999. Estrogen Actions in the early experience and the types of experiences throughout Central Nervous System. Endocr. Rev. 20:279–307. life. We strongly suspect that stress hormones play a role 18. Landfield, P. W. and Eldridge, J. C. 1994. Evolving aspects of the glucocorticoid hypothesis of brain aging: Hormonal modulation in determining the rate of brain and body aging and that of neuronal calcium homeostasis. Neurobiol. Aging 15:579–588. they do so in part by exacerbating processes involving the 19. Porter, N. M., Thibault, O., Thibault, V., Chen, K.-C., and Land- generation of free radicals that cause damage to tissues field, P. W. 1997. Calcium channel density and hippocampal cell death with age in long-term culture. J. Neurosci. 17:5629–5639. and organs, including cardiac smooth muscle cells and 20. Norris, C. M., Halpain, S., and Foster, T. C. 1998. Reversal of brain cells. Yet, at the same time, there are natural age-related alterations in synaptic plasticity by blockade of L- processes and agents such as estrogens that have type Ca2+ channels. J. Neurosci 18:3171–3179. 21. Maren, S. 1995. Properties and mechanisms of long-term synap- neuroprotective effects and enhance allostasis while tic plasticity in the mammalian brain: relationships to learning minimizing allostatic load. and memory. Neurobiol. Learning & Mem. 63:1–18. Allostasis, Allostatic Load, and the Aging Nervous System 1229

22. Joels, M. 1997. Steroid hormones and excitability in the mam- 43. Weinstock, M., Poltyrev, T., Schorer-Apelbaum, D., Men, D., malian brain. Frontiers in Neuroendocrinol. 18:2–48. and McCarty, R. 1998. Effect of prenatal stress on plasma corti- 23. Kerr, S., Campbell, L., Applegate, M., Brodish, A., and Land- costerone and catecholamines in response to footshock in rats. field, P. 1991. Chronic stress-induced acceleration of electro- Physiol. & Behav. 64:439–444. physiologic and morphometric biomarkers of hippocampal aging. 44. Ader, R. 1968. Effects of early experiences on emotional and J. Neurosci. 11:1316–1324. physiological reactivity in the rat. J. Comp. Physiol. Psych. 66: 24. McCord, J. 1985. Oxygen-derived free radicals in postischemic 264–268. tissue injury. New Engl. J. Med. 312:159–163. 45. Hess, J. L., Denenberg, V. H., Zarrow, M. X., and Pfeifer, 25. Liu, J., Fischer, A., Amiri, N., Timiras, P. S., and Ames, B. N. W. D. 1968. Modification of the corticosterone response 1998. Stress-induced oxidative damage in the brain: stress hor- curve as a function of handling in infancy. Physiol. & Behav. 4: mones enhance, but an antistress hormone (DHEA), inhibits, ox- 109–111. idative damage and beta-amyloid production in cell culture. Oxy- 46. Levine, S., Haltmeyer, G., Kara, G., and Denenberg, V. 1967. gen Society Meeting Poster. Physiological and behavioral effects of infantile stimulation. 26. Rasmussen, T., Schliemann, T., Sorensen, J. C., Zimmer, J., and Physiol. Behav. 2:55–59. West, M. J. 1996. Memory Impaired Aged Rats: No loss of Prin- 47. Dellu, F., Mayo, W., Vallee, M., LeMoal, M., and Simon, H. cipal Hippocampal and Subicular Neurons. Neurobiol. Aging 1994. Reactivity to novelty during youth as a predictive factor of 14:143–147. cognitive impairment in the elderly: a longitudinal study in rats. 27. Rapp, P. R. and Gallagher, M. 1996. Preserved neuron number in Brain Res. 653:51–56. the hippocampus of aged rats with spatial learning deficits. Proc. 48. Piazza, P. V., Marinelli, M., Jodogne, C., Deroche, V., Rouge- Natl. Acad. Sci. USA 93:9926–9930. Pont, F., Maccari, S., Le Moal, M., and Simon, H. 1994. Inhibi- 28. McEwen, B. S. 1999. Stress and the Aging Hippocampus. Fron- tion of corticosterone synthesis by Metrapone decreases cocaine- tiers in Neuroendocrinol. 20:49–70. induced locomotion and relapse of cocaine self-administration. 29. Sugaya, K., Chouinard, M., Greene, R., Robbins, M., Personett, Brain Res. 658:259–264. D., Kent, C., Gallagher, M., and McKinney, M. 1996. Molecular 49. Deroche, V., Piazza, P. V., LeMoal, M., and Simon, H. 1993. In- Indices of Neuronal and Glial Plasticity in the Hippocampal For- dividual differences in the psychomotor effects of morphine are mation in a Rodent Model of Age-Induced Spatial Learning Im- predicted by reactivity to novelty and influenced by corticos- pairment. J. Neurosci. 16:3427–3443. terone secretion. Brain Res. 623:341–344. 30. Chan, P. H. 1996. Role of oxidants in ischemic brain damage. 50. Meaney, M., Aitken, D., Berkel, H., Bhatnager, S., and Sapol- Stroke 27:1124–1129. sky, R. 1988. Effect of neonatal handling of age-related impair- 31. Lowy, M. T., Gault, L., and Yamamoto, B. K. 1993. Adrenalec- ments associated with the hippocampus. Science 239:766–768. tomy Attenuates Stress-Induced Elevations in Extracellular Glu- 51. Catalani, A., Marinelli, M., Scaccianoce, S., Nicolai, R., Mus- tamate Concentrations in the Hippocampus. J. Neurochem. colo, L. A. A., Porcu, A., Koranyi, L., Piazza, P. V., and An- 61:1957–1960. gelucci, L. 1993. Progeny of mothers drinking corticosterone 32. Moghaddam, B., Boliano, M. L., Stein-Behrens, B., and Sapol- during lactation has lower stress-induced corticosterone secre- sky, R. 1994. Glucocorticoids mediate the stress-induced extra- tion and better cognitive performance. Brain Res. 624:209–215. cellular accumulation of glutamate. Brain Res. 655:251–254. 52. Meaney, M. J., Tannenbaum, B., Francis, D., Bhatnagar, S., 33. Schaasfoort, E., DeBrin, L., and Korf, J. 1988. Mild stress stim- Shanks, N., Viau, V., O’Donnell, D., and Plotsky, P. M. 1994. ulates rat hippocampal glucose utilization transiently via NMDA Early environmental programming hypothalamic-pituitary-adrenal receptors as assessed by lactography. Brain Res. 575:58–63. responses to stress. Seminars in Neurosci. 6:247–259. 34. De Bruin, L. A., Schasfoort, M. C., Stefens, A. B., and Korf, J. 53. Pavlides, C., Kimura, A., Magarinos, A. M., and McEwen, 1994. Effects of stress and exercise on rat hippocampus and stria- B. S. 1995. Hippocampal homosynaptic long-term depression/ tum extracellular lactate. Am. J. Physiol. 259:R773–R779. depotentiation induced by adrenal steroids. Neuroscience 35. Lowy, M. T., Wittenberg, L., and Yamamoto, B. K. 1995. Effect 68:379–385. of Acute Stress on Hippocampal Glutamate Levels and Spectrin 54. Pavlides, C., Watanabe, Y., Magarinos, A. M., and McEwen, Proteolysis in Young and Aged Rats. J. Neurochem. 65:268– 274. B. S. 1995. Opposing role of adrenal steroid Type I and Type 36. Reagan, L. P., and McEwen, B. S. 1997. Controversies sur- II receptors in hippocampal long-term potentiation. Neuro- rounding glucocorticoid-mediated cell death in the hippocampus. science 68:387–394. J. Chem. Neuroanat. 13:149–167. 55. Pavlides, C., Kimura, A., Magarinos, A. M., and McEwen, 37. Keller, J. N. and Mattson, M. P. 1998. Roles of lipid peroxida- B. S. 1994. Type I adrenal steroid receptors prolong hippocam- tion in modulation of cellular signaling pathways, cell dysfunc- pal long-term potentiation. NeuroReport 5:2673–2677. tion, and death in the nervous system. Reviews in Neurosci. 56. Kerr, D. S., Huggett, A. M., and Abraham, W. C. 1994. Modu- 9:105–116. lation of hippocampal long-term potentiation and long-term de- 38. Maren, S. 1999. Long-term potentiation in the amygdala: a mech- pression by corticosteroid receptor activation. Psychobiology anism for emotional learning and memory. TINS 22:561–567. 22:123–133. 39. Bliss, T. V. P. and Collingridge, G. L. 1993. A synaptic model of 57. Barnes, C., McNaughton, B., Goddard, G., Douglas, R., and memory: long-term potentiation in the hippocampus. Nature Adamec, R. 1977. Circadian rhythm of synaptic excitability in 361:31–39. rat and monkey central nervous system. Science 197:91–92. 40. Choi, D. 1988. Calcium-mediated neurotoxicity: relationship to 58. Dana, R. C. and Martinez, J. L. 1984. Effect of adrenalectomy specific channel types and role in ischemic damage. TINS 11: on the circadian rhythm of LTP. Brain Res. 308:392–395. 465–469. 59. Diamond, D. M., Bennett, M. C., Fleshner, M., and Rose, G. M. 41. Weiland, N. G., Orchinik, M., and Tanapat, P. 1997. Chronic cor- 1992. Inverted-U relationship between the level of peripheral ticosterone treatment induces parallel changes in N-methyl-D- corticosterone and the magnitude of hippocampal primed burst aspartate receptor subunit messenger RNA levels and antagonist potentiation. Hippocampus 2:421–430. binding sites in the hippocampus. Neuroscience 78:653– 662. 60. Diamond, D. M., Fleshner, M., Ingersoll, N., and Rose, G. M. 42. Bartanusz, V., Aubry, J. M., Pagliusi, S., Jezova, D., Baffi, J., 1996. impairs spatial working memory: rele- and Kiss, J. Z. 1995. Stress-Induced Changes in Messenger RNA vance to electrophysiological studies of hippocampal function. Levels of N-Methyl-D-Aspartate and Ampa Receptor Subunits Behav. Neurosci. 110:661–672. in Selected Regions of the Rat Hippocampus and Hypothalamus. 61. Roozendaal, B. and McGaugh, J. L. 1997. Glucocorticoid receptor Neuroscience 66:247–252. agonist and antagonist administration into the basolateral but not 1230 McEwen

central amygdala modulates memory storage. Neurobiol. Learning independently of causing hippocampal CA3 dendritic atrophy. & Mem. 67:176–179. Behav. Neurosci. 113:1–12. 62. Corodimas, K. P., LeDoux, J. E., Gold, P. W., and Schulkin, J. 82. Eldridge, J. C., Brodish, A., Kute, T. E., and Landfield, P. W. 1994. Corticosterone potentiation of learned fear. Annals of the 1989. Apparent age-related resistance of type II hippocampal New York Acad. Sci. 746:392. corticosteroid receptors to down-regulation during chronic es- 63. Diamond, D. M., Fleshner, M., and Rose, G. M. 1994. Psycho- cape training. J. Neurosci. 9:3237–3242. logical stress repeatedly blocks hippocampal primed burst potenti- 83. Cameron, H. A. and McKay, D. G. 1999. Restoring production ation in behaving rats. Behav. Brain Research 62:1–9. of hippocampal neurons in old age. Nature Neurosci. 2:894– 64. deQuervain, D. J. F., Roozendaal, B., and McGaugh, J. L. 1998. 858. Stress and glucocorticoids impair retrieval of long-term spatial 84. Tanapat, P., Hastings, N. B., Reeves, A. J., and Gould, E. 1999. memory. Nature 394:787–790. Estrogen stimulates a transient increase in the number of new 65. Galea, L. A. M., Tanapat, P., and Gould, E. 1996. Exposure to neurons in the dentate gyrus of the adult female rat. J. Neurosci. predator odor suppresses cell proliferation in the dentate gyrus of 19:5792–5801. adult rats via a cholinergic mechanism. Abstract, Soc. Neurosci. 85. Galea, L. A. M., McEwen, B. S., Tanapat, P., Deak, T., Spencer, 22:#474.8, p.1196. R. L., and Dhabhar, F. S. 1997. Sex differences in dendritic at- 66. Ikegaya, Y., Saito, H., and Abe, K. 1997. The basomedial and ba- rophy of CA3 pyramidal neurons in response to chronic restraint solateral amygdaloid nuclei contribute to the induction of long- stress. Neuroscience 81:689–697. term potentiation in the dentate gyrus in vivo. Eur. J. Neurosci. 86. Gould, E., Westlind-Danielsson, A., Frankfurt, M., and 8:1833–1839. McEwen, B. S. 1990. Sex differences and thyroid hormone 67. LeDoux, J. E. 1995. In search of an emotional system in the brain: sensitivity of hippocampal pyramidal neurons. J. Neurosci. 10: leaping from fear to emotion and consciousness. In Gazzaniga, M., 996–1003. ed. The Cognitive Neurosciences. Cambridge, MIT Press, 87. Williams, C. L. and Meck, W. H. 1991. The organizational ef- 1049–1061. fects of gonadal steroids on sexually dimorphic spatial ability. 68. Woolley, C., Gould, E., Frankfurt, M., and McEwen, B. S. 1990. Psychoneuroendocrinology 16:155–176. Naturally occurring fluctuation in dendritic spine density on adult 88. Juraska, J. M. 1991. Sex differences in “cognitive” regions of the hippocampal pyramidal neurons. J. Neurosci. 10:4035–4039. rat brain. Psychoneuroendocrinology 16:105–119. 69. Popov, V. I. and Bocharova, L. S. 1992. Hibernation-induced struc- 89. Roof, R. L. 1993. The dentate gyrus is sexually dimorphic in tural changes in synaptic contacts between mossy fibres and hip- prepubescent rats: testosterone plays a significant role. Brain pocampal pyramidal neurons. Neuroscience 48:53–62. Res. 610:148–151. 70. Popov, V. I., Bocharova, L. S., and Bragin, A. G. 1992. Repeated 90. Van Cauter, E., Leproult, R., and Kupfer, D. J. 1996. Effects of changes of dendritic morphology in the hippocampus of ground Gender and Age on the Levels and Circadian Rhythmicity of squirrels in the course of hibernation. Neuroscience 48:45–51. Plasma Cortisol. J. Clin. Endocrinol. & Metabol. 81:2468– 71. McEwen, B. S. and Woolley, C. S. 1994. Estradiol and proges- 2473. terone regulate neuronal structure and synaptic connectivity in 91. Komesaroff, P. A., Esler, M. D., and Sudhir, K. 1999. Estrogen adult as well as developing brain. Exp. Geront. 29:431–436. supplementation attenuates glucocorticoid and catecholamine re- 72. McEwen, B. S., Gould, E., Orchinik, M., Weiland, N. G., and sponses to mental stress in perimenopausal women. J. Clin. En- Woolley, C. S. 1995. Oestrogens and the structural and functional docrinol. & Metab. 84:606–610 plasticity of neurons: implications for memory, aging and neurode- 92. Goodman, Y., Bruce, A. J., Cheng, B., and Mattson, M. P. 1996. generative processes. In Goode, J., ed. Ciba Foundation Sympo- Estrogens attenuate and corticosterone exacerbates excitotoxic- sium #191 The Non-reproductive Actions of Sex Steroids. London, ity, oxidative injury, and amyloid β-peptide toxicity in hip- CIBA Foundation, 52–73. pocampal neurons. J. Neurochem. 66:1836–1844. 73. Cameron, H. A. and Gould, E. 1994. Adult neurogenesis is regu- 93. Hawk, T., Zhang, Y.-Q., Rajakumar, G., Day, A. L., and Simp- lated by adrenal steroids in the dentate gyrus. Neuroscience kins, J. W. 1998. Testosterone increases and estradiol decreases 61:203–209. middle cerebral artery occlusion lesion size in male rats. Brain 74. Woolley, C. and McEwen, B. S. 1994. Estradiol regulates hip- Res. 796:296–298. pocampal dendritic spine density via an N-methyl-D-aspartate 94. Dubal, D. B., Shughrue, P. J., Wilson, M. E., Merchenthaler, I., receptor dependent mechanism. J. Neurosci. 14:7680–7687. and Wise, P. M. 1999. Estradiol modulates bcl-2 in cerebral is- 75. Magarinos, A. M. and McEwen, B. S. 1995. Stress-induced at- chemia: A potential role for estrogen receptors. J. Neurosci. rophy of apical dendrites of hippocampal CA3c neurons: in- 19:6385–6393. volvement of glucocorticoid secretion and excitatory amino acid 95. Sapolsky, R. and Pulsinelli, W. 1985. Glucocorticoids potentiate receptors. Neuroscience 69:89–98. ischemic injury to neurons: therapeutic implications. Science 76. Weiland, N. G. 1992. Estradiol Selectively Regulates Agonist 229:1397–1399. Binding Sites on the N-Methyl-D-Aspartate Receptor Complex in 96. Uno, H., Ross, T., Else, J., Suleman, M., and Sapolsky, R. 1989. the CA1 Region of the Hippocampus. Endocrinology 131: Hippocampal damage associated with prolonged and fatal stress 662– 668. in primates. J. Neurosci 9:1709–1711. 77. Gazzaley, A. H., Benson, D. L., Huntley, G. W., and Morrison, 97. Henderson, V. W. and Paganini-Hill, A. 1994. Oestrogens and J. H. 1997. Differential subcellular regulation of NMDAR1 pro- Alzheimer’s disease. Ann. Prog. Rep. Med. 2:1–21. tein and mRNA in dendrites of dentate gyrus granule cells after 98. Tang, M. X., Jacobs, D., Stern, Y., Marder, K., Schofield, P., perforant path transection. J. Neurosci. 17:2006–2017. Gurland, B., Andrews, H., and Mayeux, R. 1996. Effect of oe- 78. Gould, E. 1999. Serotonin and hippocampal neurogenesis. Neu- strogen during menopause on risk and age at onset of Alzhei- ropsychopharmacology 21:46S–51S. mer’s disease. Lancet 348:429–432. 79. van Praag, H., Kempermann, G., and Gage, F. H. 1999. Running 99. Kawas, C., Resnick, S., Morrison, A., Brookmeyer, R., Corrada, increases cell proliferation and neurogenesis in the adult mouse M., Zonderman, A., Bacal, C., Lingle, D. D., and Metter, E. dentate gyrus. Nature Neurosci. 2:266–270 1997. A prospective study of estrogen replacement therapy and 80. Kempermann, G., Kuhn, H. G., and Gage, F. H. 1997. More hip- the risk of developing Alzheimer’s disease: the Baltimore Lon- pocampal neurons in adult mice living in an enriched environ- gitudinal Study of Aging. Neurology 48:1517–1521. ment. Nature 586:493–495. 100. Green, P. S., Gridley, K. E., and Simpkins, J. W. 1996. Estradiol 81. Conrad, C. D., Magarinos, A. M., LeDoux, J. E., and McEwen, protects against β-amyloid (25–35)-induced toxicity in SK-N- B. S. 1999. Repeated restraint stress facilitates fear conditioning SH human neuroblastoma cells. Neurosci. Lett. 218:165–168. Allostasis, Allostatic Load, and the Aging Nervous System 1231

101. Blaylock, R. L. 1999. Neurodegeneration and aging of the cen- blind, randomized trial of an extract of Ginko biloba for demen- tral nervous system: Prevention and treatment by phytochemi- tia. JAMA 278:1327–1332. cals and metabolic nutrients. Integrat. Med. 1:117–133 115. Patisaul, H. B., Whitten, P. L., and Young, L. J. 1999. Regulation 102. Kuiper, G. G. J. M., Lemmen, J. G., Carlsson, B., Corton, J. C., of estrogen receptor beta mRNA in the brain: opp osite effects of Safe, S. H., van der Saag, P. T., van der Burg, B., and Gustafsson, 17β-estradiol and the phytoestrogen, coumestrol. Mol. Brain Res. J.-A. 1998. Interaction of estrogenic chemicals and phytoestrogens 67:165–171. with estrogen receptor beta. Endocrinology 139:4252–4263. 116. Nagata, C., Takatsuka, N., Inaba, S., Kawakami, N., and Shimizu, 103. Bishop, J. and Simpkins, J. W. 1994. Estradiol treatment in- H. 1998. Effect of soymilk consumption on serum estrogen con- creases viability of glioma and neuroblastoma cells in vitro. Mol. centrations in premenopausal Japanese women. J. Natl. Cancer & Cell. Neurosci. 5:303–308. Inst. 90:1830–1835. 104. Green, P. S., Bishop, J., and Simpkins, J. W. 1997. 17α-Estra- 117. Morrison, R. S., Wenzel, H. J., Kinoshita, Y., Robbins, C. A., diol exerts neuroprotective effects on SK-N-SH cells. J. Neu- Donehower, L. A., and Schwartzkroin, P. A. 1996. Loss of the rosci. 17:511–515. p53 tumor suppressor gene protects neurons from kainate-in- 105. Mooradian, A. D. 1993. Antioxidant properties of steroids. J. duced cell death. J. Neurosci. 16:1337–1345. Steroid Biochem. Molec. Biol. 45:509–511. 118. Xiang, H., Hochman, D. W., Saya, H., Fujiwara, T., 106. Behl, C., Skutella, T., Lezoualc’h, F., Post, A., Widmann, M., Schwartzkroin, P. A., and Morrison, R. S. 1996. Evidence for Newton, C. J., and Holsboer, F. 1997. Neuroprotection against p53-mediated modulation of neuronal viability. J. Neurosci. oxidative stress by estrogens: structure-activity relationship. 16:6753–6765. Mole. Pharm. 51:535–541. 119. Kondo, T., Reaume, A. G., Huang, T.-T., Carlson, E., Murakami, 107. Behl, C., Widmann, M., Trapp, T., and Holsboer, F. 1998. 17 beta K., Chen, S. F., Hoffman, E. K., Scott, R. W., Epstein, C. J., and estradiol protects neurons from oxidative stress-induced cell Chan, P. H. 1997. Reduction of CuZn-superoxide dismutase ac- death in vitro. Biochem. Biophys. Res. Comm. 216:473– 482. tivity exacerbates neuronal cell injury and edema formation after 108. Birge, S. J. 1994. The Role of Estrogen Deficiency in the Aging transient focal cerebral ischemia. J. Neurosci. 17:4180–4189. Central Nervous System. In Lobo, R. A., ed. Treatment of the 120. Reagan, L. P., Magarinos, A. M., and McEwen, B. S. 1998. Mol- Postmenopausal Woman: Basic and Clinical Aspects. New ecular changes induced by stress in streptozotocin (STZ) diabetic York, Raven Press, Ltd., 153–157. Notes: Chapter Num: 14. rats. Society for Neuroscience Abstract. 109. Henderson, V. W., Paganini-Hill, A., Emanuel, C. K., Dunn, 121. Pouliot, W. A., Handa, R. J., and Beck, S. G. 1996. Androgen M. E., and Buckwalter, J. G. 1994. Estrogen Replacement Ther- Modulates N-Methyl-D-Aspartate-mediated Depolarization in apy in Older Women: Comparisons between Alzheimer’s Dis- CA1 Hippocampal Pyramidal Cells. Synapse 23:10–19. ease Cases and Nondemented Control Subjects. Archives of 122. Grindley, K. E., Green, P. S., and Simpkins, J. W. 1997. Low Neurology 51:896–900. concentrations of estradiol reduce β-amyloid (25–35) induced 110. Paganini-Hill, A. and Henderson, V. W. 1994. Estrogen Defi- toxicity, lipid peroxidation and glucose utilization in human SK- ciency and Risk of Alzheimer’s Disease in Women. American N-SH neuroblastoma cells. Brain Res. 778:158–165. Journal of Epidemilogy 3:3–16. 123. Kimonides, V. G., Khatibi, N. H., Sofroniew, M. V., and Herbert, 111. Henderson, V. W., Watt, L., and Buckwalter, J. G. 1996. Cognitive J. 1998. Dehydroepiandrosterone (DHEA) and DHEA-sulfate Skills Associated With Estrogen Replacement In Women With (DHEAS) protect hippocampal neurons against excitatory amino Alzheimer’s Disease. Psychoneuroendocrinology 12:421–430. acid-induced neurotoxicity. Proc. Nat. Acad. Sci. USA 95:1852– 112. Yaffe, K., Sawaya, G., Lieberburg, I., and Grady, D. 1998. Es- 1857. trogen therapy in postmenopausal women: Effects on cognitive 124. Bastianetto, S., Ramassamy, C., Poirier, J., and Quirion, R. 1999. function and dementia. JAMA 279:688–695. Dehydroepiandrosterone (DHEA) protects hippocampal cells 113. Joseph, J. A., Shukitt-Hale, B., Denisova, N. A., Prior, R. L., from oxidative stress-induced damage. Mol. Brain Res. 66:35–41 Cao, G., Martin, A., Taglialatela, G., and Bickford, P. C. 1998. 125. Ho, D. Y., Saydam, T. C., Fink, S. L., Lawrence, M. S., and Long-term dietary strawberry, spinach, or vitamin E supple- Sapolsky, R. M. 1995. Defective herpes simplex virus vectors ex- mentation retards the onset of age-related neuronal signal- pressing the rat brain glucose transporter protect cultured neurons transduction and cognitive behavioral deficits. J. Neurosci. 18: from necrotic insults. J. Neurochem. 65:842–850. 8047–8055. 126. McEwen, B. S. 1999. The molecular and neuroanatomical basis 114. Le Bars, P. L., Katz, M. M., Berman, N., Itil, T. M., Freedman, for estrogen effects in the central nervous system. J. Clin. En- A. M., and Schatzberg, A. F. 1997. A placebo-controlled, double- docrinol. Metab. 84:1790–1797.