Journal of Neuroendocrinology 19, 743751

REVIEW ARTICLE

2007 The Authors. Journal Compilation 2007 Blackwell Publishing Ltd

Chronic Stress and Neural Function: Accounting for Sex and Age
V. N. Luine,* K. D. Beck, R. E. Bowman, M. Frankfurt and N. J. MacLusky**
*Department of Psychology, Hunter College of CUNY, New York, NY, USA. VA NJ Health Care System, East Orange, NJ, USA. Department of Neuroscience, UMDNJ, Newark, NJ, USA. Department of Psychology, Sacred Heart University, Faireld, CT, USA. Department of Physiology Pharmacology, CUNY Medical School, New York, NY, USA. **Department of Biochemistry, Ontario Veterinary College, Guelph, ON, Canada

Journal of Neuroendocrinology

Correspondence to: Dr Victoria Luine Department of Psychology, Hunter College of CUNY, 695 Park Ave, New York, NY 10021, USA (e-mail: vluine@hunter.cuny.edu).

Cognitive responses to stress follow the temporally dependent pattern originally established by Selye (1) wherein short-term stressors elicit adaptive responses whereas continued stress (chronic) results in maladaptive changes deleterious effects on physiological systems and impaired cognition. However, this pattern for cognitive effects appears to apply to only half the population (males) and, more specically, to young, adult males. Females show different cognitive responses to stress. In contrast to impaired cognition in males after chronic stress, female rodents show enhanced performance on the same memory tasks after the same stress. Not only cognition, but anxiety, shows sex-dependent changes following chronic stress stress is anxiolytic in males and anxiogenic in females. Moreover, behavioral responses to chronic stress are different in developing as well as aging subjects (both sexes) as compared to adults. In aged rats, chronic stress enhances recognition memory in both sexes, does not alter spatial memory, and anxiety effects are opposite to young adults. When pregnant dams are exposed to chronic stress, at adulthood the offspring display yet different consequences of stress on anxiety and cognition, and, in contrast to adulthood when the behavioral effects of stress are reversible, prenatal stress effects appear enduring. Changing levels of estradiol in the sexes over the lifespan appear to contribute to the differences in response to stress. Thus, theories of stress dependent moduations in CNS function developed solely in male models, focused on peripheral physiological processes and tested in adults may require revision when applied to a more diverse population (age- and sex-wise) at least in relation to the neural functions of codnition and anxiety. Moreover, these results suggest that other stressors and neural functions should be investigated to determine whether age, sex and gonadal hormones also have an impact. Key words: anxiety, memory, oestradiol, sex differences, stress. doi: 10.1111/j.1365-2826.2007.01594.x

Beginning with Selyes seminal studies (1), stress-dependent responses have been considered to follow a temporal pattern of response in which short-term stressors evoke adaptive changes by the organism whereas long-term stressors (chronic stress) cause maladaptive changes. Recent studies report that neural functions such as cognition and anxiety also follow this temporal pattern of response in rats. However, this review highlights other recent studies from our own and other laboratories examining the effects of chronic stress in female as well as male subjects during development (prenatally), adulthood and old age. These studies show a

more complex pattern of responses to chronic stress for cognition and anxiety and suggest that neural stress responses are dependent on the sex, age and gonadal hormone status of subjects.

Chronic stress effects on cognitive function are different in the sexes

Similar to its deleterious effects on several physiological systems, chronic stress also exerts damaging effects in the central nervous system. A broad range of studies show that neurones are adversely

744

V. N. Luine et al.

affected in the frontal cortex and hippocampus (28) and show altered dendritic arbours (generally a pruning) and structural remodelling of synaptic connectivity. Activities of monoaminergic systems in both cell bodies in the hindbrain and terminals throughout the brain are changed (915). Receptors for glucocorticoids are generally down regulated in hippocampus (16), which leads to greater release of corticosterone and an escalation in the adverse changes as long as the stressor is present. Functional indices of neurones are also affected, and include impairments in cognitive function and changes in anxiety level. In adult male rats, restraint (not immobilisation) in small plastic or mesh containers for 6 h day for 21 days impairs performance on a variety of spatial memory tasks including radial arm maze (17), object placement (10, 11), Y-maze (18) and water maze (16) and a nonspatial, recognition memory test (10, 11). For example, Fig. 1(A) shows that unstressed males made an average of less than two errors in completing radial arm maze trials whereas chronically stressed males made more than two errors. In object recognition trials, a recognition memory task (Fig. 1B), control males signicantly discriminated between recently explored objects and new objects (spending over 50% of the exploration time with the new object) whereas stressed males did not. At shorter intervals of restraint (12 weeks), performance of the tasks that have been tested, namely radial arm maze (19) and object placement (20), is enhanced. In addition, chronic stressdependent changes in memory (17, 19) and dendritic arbors (4, 6) are not permanent and return to prestress levels 710 days following cessation of stress. Thus, stress effects on central nervous system (CNS) function are consistent with the notion that the brain responds like the peripheral physiological systems in that short-term stress elicits adaptive responses but continued stress exposure results in maladaptive responses. It should be noted, however, that these studies investigated only one half of the population (i.e males). Quite surprisingly, recent studies using female rodents have arrived at different conclusions concerning the impact of stress, both acute and chronic, on CNS function. When exposed to the same stress paradigm that impairs performance of cognitive tasks, females show enhanced performance (12, 21, 22). Thus, 21 days of restraint stress for 6 h day given to female rats is associated with better performance of spatial memory tasks including radial arm maze (23, 24), object placement (11), Y-maze (7, 25) and water maze (16). Performance of another memory task, object recognition, which is also impaired by stress in males, is not affected in female rats following chronic stress (11, 26). As shown in Fig. 1(A) for the radial arm maze, control females made four errors to complete the task whereas, following chronic stress, subjects made only two errors and, in object recognition (Fig. 1B), both control and stressed females spend more than 50% of the exploration time with the new object. In the few studies where female brains have been examined, a different pattern of stress effects is seen than in males with respect to morphology, where females show dendritic pruning in basal, not apical, hippocampal CA3 pyramidal neurones (26) and an increase in dendritic spine heads in CA1 pyramidal neurones (7). Glucocorticoid receptors are increased in CA1 of females (16) as well as levels of some monoamines (11, 12). These results suggest that stress may

(A)

Radial arm maze

Number of errors

3 * 2 **

0 MCON MSTR FCON FSTR

(B)

Object recognition 100 *

% Time with new object

75

50

25

MCON

MSTR

FCON

FSTR

Fig. 1. Effect of chronic restraint stress on performance of memory tasks. (A) Radial arm maze: effect of 21 days of daily restraint stress on performance of control males (MCON, solid bar), stressed males (MSTR, open bar), control females (FCON, grey bar) and stressed females (FSTR, shaded bar). Entries are the mean SEM for ve trials in males and eight trials in females with eight to ten subjects per group. Data from separate experiments in males (17) and females (23) are shown and were analysed by separate repeated measures ANOVA (treatment trials) in each sex (*P < 0.05 and **P < 0.01 for treatment effect). (B) Object recognition: effect of 21 days of daily restraint stress on performance of control males (MCON, solid bar), stressed males (MSTR, open bar), control females (FCON, grey bar) and stressed females (FSTR, shaded bar). The intertrial delay between the sample and recognition trial was 4 h. Entries are the mean SEM of the percentage time spent exploring the new object. Dashed line at 50% indicates chance performance of task (same amount of time spent exploring the old and new object). Experiments in the sexes were performed separately: male data are from (10); female data are unpublished data from Beck and Luine, for published female data, see (11). Data are analysed by a paired t-test within each group. *P < 0.05 indicates that subjects spent more time exploring the new object than the old object.

affect neural function in the sexes differently, a notion not previously considered or investigated in most studies.

Chronic stress effects on anxiety are different in the sexes

The observation that females cognitive responses to stress are different than males is novel and unexpected. It could be argued that

2007 The Authors. Journal Compilation 2007 Blackwell Publishing Ltd, Journal of Neuroendocrinology, 19, 743751

Chronic stress and neural function

745

stress responses were an anomaly in females and related to the tasks themselves (i.e. females are less procient at spatial memory tasks) [e.g. control males make fewer errors than control females, two versus four, respectively, in completing the eight arm radial maze task (Fig. 1); for further information, see 11, 12, 16, 21, 22, 25, 27]. Thus, females performance may already be so minimal that stress cannot decrease performance any further. On the other hand, chronic stress did not impair object recognition performance in the females. Nonetheless, another behaviour in males and females, anxiety, was examined in response to chronic stress in the same subjects as received cognitive testing. The latency to enter an open eld following the opening of a guillotine door was assessed (11, 28) (Fig. 2A). As in performance of some memory tasks, anxiety was sexually differentiated: control male rats were more anxious than females entering 120 s after the door was opened whereas females entered in approximately 30 s. Moreover, stress decreased entry latencies in the males but increased latencies in the females. Thus, stress elicited opposite effects on the sexes and resulted in the elimination of sex differences in anxiety in the stressed subjects. Thus, for two behaviours, chronic stress exerts different effects in rats, dependent on their sex. Moreover, it does not appear that changes in anxiety account for stress effects on cognition. Greater fear responses (in females) to the eld and objects within the eld would most likely lead to less exploration of objects, especially new objects, during the object memory tests. However, this is not the case (11) (Fig. 2A). These data show that sexually differentiated responses to stress are present across a number of neural domains; nonetheless, the number of studies is quite limited. Further research should consider whether the type of stressor impacts the response (i.e. perhaps the stress of predator exposure, which is more relevant to life than restraint, might result in similar responses in the sexes) (29).
Fig. 2. Effect of chronic restraint stress on latency to enter an open eld. (A) Entry latencies in young adult rats: effect of 21 days of daily restraint stress in 3-month-old, control males (MCON, solid bar), stressed males (MSTR, open bar), control females (FCON, grey bar) and stressed females (FSTR, shaded bar). Entries are the mean SEM for eight subjects per group. Data were analysed by two-way ANOVA (sex treatment) and a main effect of sex (P < 0.005) and an interaction of sex treatment (P < 0.02) was present. Post-hoc testing by the Bonferroni MC test showed a signicant difference between controls (**P < 0.01) but stress groups did not differ from each other and fell between the control means. Data are adapted from (11). (B) Entry latencies in aged adult rats: effect of 21 days of daily restraint stress in 21-month-old, control males (MCON, solid bar, n 7), stressed males (MSTR, open bar, n 10), control females (FCON, grey bar, n 8) and stressed females (FSTR, shaded bar, n 11). Entries are the mean SEM. Data were analysed by two-way ANOVA (sex treatment) and no main effects were present but a trend in sex treatment (P 0.1) was present. Data are from (33). (C) Entry latencies in prenatally stressed subjects measured at adulthood: effect of daily restraint stress during prenatal days 1421. Behaviour was assessed at 2 months of age in control males (MCON, solid bar, n 16), stressed males (MSTR, open bar, n 17), control females (FCON, grey bar, n 14) and stressed females (FSTR, shaded bar, n 15). Data were analysed by two-way ANOVA (sex treatment) and there was a sextreatment interaction (P < 0.03). Post-hoc testing by Fishers LSD showed that MCON, MSTR and FSTR did not differ but that FCON and FST were signicantly different from each other (*P < 0.05). Reproduced from (27) by permission of the Endocrine Society.

Furthermore, whether longer stress durations are associated with maladaptive responses in females should be determined. We investigated the effects of 28 days of daily restraint stress on radial arm maze performance, and females were neither enhanced, nor impaired in performance (23). Thus, it is possible that the response of the sexes to chronic stress is ultimately the same impaired function, but that females are innately less sensitive to restraint stress and require longer periods of stress for the appearance of maladaptive changes (12, 21, 22). Anxiety responses to stress should also be examined

(A)
200

Open field entry latency Stress in adult **

Time (s)

100

MCON

MSTR

FCON

FSTR

(B)
200

Open field entry latency Stress in aged

Time (s)

100

MCON

MSTR

FCON

FSTR

(C)

Open field entry latency Stress in prenatal, test in adult * 125 100

Time (s)

75 50 25 0

MCON

MSTR

FCON

FSTR

2007 The Authors. Journal Compilation 2007 Blackwell Publishing Ltd, Journal of Neuroendocrinology, 19, 743751

746

V. N. Luine et al.

% Time at new object

using the elevated plus maze to determine whether opposite responses to stress are obtained in the sexes with this task. Some studies have utilised the plus maze (30, 31), but most use only one sex, a different stress paradigm, or have assessed effects following prenatal stress. Thus, sex comparisons of the pattern of stress effects on the open eld and plus maze await further testing.

(A)
75

Object recognition Aged rats *

50

Sex differences in chronic stress: corticosteroid levels

Our laboratories and others have shown that females have generally higher basal circulating corticosterone levels, up to four-fold higher, compared to males,. Females also release more corticosterone following a stressor than males, two to four-fold higher levels at 13 h after initiating a restraint stress (26, 27, 32). Although not extensively investigated, these sex differences in stress-dependent corticosterone release also appear to be maintained into old age (33, 34). Because corticosterone is believed to be responsible for the majority of stress-dependent CNS changes, it is surprising that stressed males are more cognitively impaired than stressed females because males have lower levels of corticosterone than females. However, we found that free corticosterone levels, measured over 21 days of restraint, appear to habituate faster in females than in males and may thus contribute to females stress resistance (23). By contrast, total corticosterone (free + corticosterone bound to corticosterone binding protein) in males has been shown to exhibit greater habituation across 21 days of restraint stress than females (26). Thus, the possible contributions of stress-induced levels of corticosterone to sex differences in neural and behavioural responses require further investigation. A complicating factor in the analysis of corticosterone contributions to female neural function is that corticosterone levels are not constant but vary over the oestrous cycle and are highest on pro-oestrous (32, 35). In addition, the magnitude of stress-induced corticosterone release is highest on pro-oestrous (35), and rats in pro-oestrous show a delayed shut-off of corticosterone release compared to rats in other stages of the oestrous cycle (32). Restraint stress for as long as 21 days does not appear to interfere with oestrous cyclicity (7, 22, 23); thus, daily changes in hypothalamic-pituitary-adrenal responses during a chronic stress regimen would be present in females and might contribute to sex differences in the responses.

25

MCON

MSTR

FCON

FSTR

(B )
75 % Time at new location

Object placement Aged rats

50

25

MCON

MSTR

FCON

FSTR

Chronic stress effects on cognition and anxiety are different in aged as compared to adult rats
Few studies have considered the effects of ageing on responses to stress. Because aged rats, like humans, show general declines debilitations in physiological systems, it can be hypothesised that chronic stress might exert greater impairing effects in aged subjects than in young subjects or that the change from adaptive to maladaptive responses might be faster. In addition, aged male rats show prolonged elevations in corticosterone following stress (36) that might contribute to impairments (females have not been investigated). We recently examined the effects of 21 days of daily (6 h day) restraint on aged (21 months old) male and female rats (33). Surprisingly,

Fig. 3. Effect of chronic restraint stress in aged rats on performance of memory tasks. (A) Object recognition: effect of 21 days of daily restraint stress in 21.5-month-old, control male (MCON, solid bar, n 7), stressed male (MSTR, open bar, n 9), control female (FCON, grey bar, n 6) and stressed female (FSTR, shaded bar, n 10) rats. The intertrial delay between the sample and recognition trial was 1 h. Entries are the mean SEM of the percentage time spent with the new object. Dashed line at 50% indicates chance performance of task (same amount of time spent exploring the old and new object). Two-way ANOVA of time (sex treatment) showed a signicant effect of stress (*P < 0.046). Data are from (33). (B) Object placement; effect of 21 days of daily restraint stress in 21.5-month-old, control male (MCON, solid bar, n 8), stressed male (MSTR, open bar, n 10), control female (FCON, grey bar, n 7) and stressed female (FSTR, shaded bar, n 11) rats. The intertrial delay between the sample and recognition trial was 1 h. Entries are the mean SEM of the percentage time spent with the object at the new location. Dashed line at 50% indicates chance performance of task (same amount of time spent exploring at old and new location). Two-way ANOVA of time (sex treatment) showed no signicant effects. Data are from (33).

stress effects on memory tasks were different compared to young adults and no sex differences were evident in the aged rats. For object recognition, both male and female rats that had been stressed performed better than control rats (Fig. 3A). By contrast, when young rats were stressed, males were impaired on object recognition whereas the performance of females was not affected (Fig. 1B). In the spatial memory task (i.e. object placement), sex differences in performance between unstressed, control males and females were no longer present, and stress did not alter performance in either sex (Fig. 3B). By contrast, young males are better at this task and radial

2007 The Authors. Journal Compilation 2007 Blackwell Publishing Ltd, Journal of Neuroendocrinology, 19, 743751

Chronic stress and neural function

747

arm maze than females, and 21 days of stress impairs young male performance and enhances female performance of both spatial memory tasks (11, 17, 23, 24) (Fig. 1A). Thus, in contrast to the maladaptive effects of chronic stress in young male rats, stress is associated with some adaptive changes in aged males. Aged females, on the other hand, have lost adaptive effects of stress on spatial memory, but show some adaptive changes in object recognition performance. It should be emphasised that the performance of the aged rats in both tasks was not as good as that of young rats, often below chance discrimination, and suggests perseverative behaviour. Perseveration is the tendency to maintain old behaviours and not learn new ones (e.g. re-exploring old objects, not new ones). Perseveration by old subjects is noted in many cognitive tasks and complicates analyses. This poor baseline performance may also inuence stress effects on performance. However, aged subjects could signicantly discriminate between objects at a shorter intertrial delay (10 min) and neither sex, nor stress affected performance (33). When anxiety is measured on an open eld, chronic restraint stress elicits different effects in aged compared to young rats, and sex differences in responses are also altered. Anxiety, measured by latency to enter an open eld, is not different between the sexes in aged rats whereas young males were more anxious than young females (Figs 2A,B). Following chronic stress, a trend for increased anxiety in males and a decreased anxiety in females (P < 0.1) was found, which is in contrast to the result for changes in young adults. Thus, chronic stress effects on anxiety are different in young compared to aged rats. Thus, in two markers of neural function, memory and anxiety tests, chronic stress elicited different effects in young versus aged rats, and the effects were also sex dependent. In general, chronic stress was benecial or adaptive to the aged rats because they performed some tasks better. The age-dependent changes are especially notable in males because young males exhibit maladaptive changes to chronic stress. However, these results in aged rats were obtained in a single study (33) and require further conrmation.

task was not present in prenatally stressed subjects (27). For anxiety (latency to enter open eld), as previously found in young adults, control males were more anxious than females and, remarkably, prenatally stressed females had similar entry latencies as control males (Fig. 2C). Anxiety in the males was not signicantly altered by prenatal stress. As is evident in these results, the pattern of effects was different than in adulthood where stress to males was associated with decreased anxiety but, more importantly, these effects were found 3 months following the stress period. Thus, prenatal stress was associated with sex-dependent, permanent alterations in cognition and anxiety and appeared to have feminised the males and masculinised the females. Masculinisation was also seen in corticosterone release in prenatally stressed females (27). When stressed, control females released a higher amount of corticosterone, but the prenatally stressed females released the same amount as control males. Thus, prenatal stress was associated with long-term changes in memory, anxiety, and corticosterone release. Moreover, these stress-dependent responses during the prenatal period were different than during adulthood and old age.

Oestradiol contributes to sex and age dependent effects of stress

The bases for sex and age differences in response to chronic stress are mainly unknown. Because oestrogen levels are different between the sexes, change over the lifespan, and oestrogen has neuroprotective properties in a number of in vitro neural systems and in vivo (38), the interactive effects of this gonadal hormone with stress responses might be important. First, if higher circulating levels of oestradiol in adult females versus males render the females less sensitive to the impairing effects of stress on memory function and alter anxiety responses, then ovariectomised (OVX) females might be expected to be more sensitive to or respond differently to chronic stress. Consistent with this hypothesis, OVX females receiving 21 days of daily restraint stress no longer showed better radial arm maze performance than nonstressed females (24); however, OVX females were not impaired by the stress as in males. Whether oestrous cycle day, with associated changes in gonadal hormone levels, inuences responses to chronic stress has received little attention, but corticosterone release (35), c-fos activation in the cortex and hippocampus (32) and cognitive function (39) appear to be sensitive to which day of the oestrous cycle that stress is administered. In addition, oestradiol treatment to aged male rats attenuates the prolonged rise in corticosterone following stress and increases glucocorticoid receptors in hippocampal areas (39; females were not analysed). Thus, gonadal hormones, most likely oestradiol, may contribute to adult female rats resistance to chronic stress. With ageing, female rats undergo decreases in gonadal hormones, including oestradiol, and thus, responses to stress might be altered if oestradiol contributes to sex differences in the stress response. Levels of oestradiol and testosterone are shown in 3- and 21-month-old Fischer 344 female and male rats (Fig. 4A). As expected, oestradiol was lower in the aged females but, surprisingly,

Chronic stress effects on cognition and anxiety are different in pre-natal as compared to adult rats
The effects of chronic stress at the beginning of the lifespan (i.e. on foetuses in utero) have also been examined. Pregnant dams were given three sessions of 45 min of restraint stress daily from days 1421 of pregnancy, a procedure which has been shown to attenuate and shorten the prenatal rise of testosterone in male foetuses and results in less robust male sexual behaviour and the display of some female sexual behaviour (37). Thus, we hypothesised that prenatally stressed males might display a more femalelike phenotype for spatial memory and anxiety whereas females should not be affected by this paradigm because they have no testes to release testosterone. As adults, the prenatally stressed rats were tested for object recognition, radial arm maze and anxiety performance (27). As expected for the radial arm maze, prenatally stressed males performed less well than unstressed males and, surprisingly, prenatally stressed females performed better than unstressed females so that the usual male advantage in this spatial

2007 The Authors. Journal Compilation 2007 Blackwell Publishing Ltd, Journal of Neuroendocrinology, 19, 743751

748

V. N. Luine et al.

(A)
20.0 17.5 Concentration (pg/ml) 15.0 12.5 10.0 7.5 5.0 2.5 0.0 Males a

Serum oestradiol b Young Old Number of inner visits 7.5

Open field inner visits Adult females ** CON STR

5.0

2.5

Females Serum testosterone

0.0 OVX Young OVX + E

(B)
4 a

Concentration (ng/ml)

Old 3 b 2

1 c 0 Males Females c

Fig. 5. Effect of oestradiol on open eld inner visits in control or chronically stressed, ovariectomised rats. Rats were ovariectomised and received 1-cm long cholesterol or 90% cholesterol)10% oestradiol silastic capsules. Ten days later, daily restraint for 6 h day was given for 21 days. Open eld was tested 1 day after the termination of stress. Inner visits are the mean SEM for eight rats per group. Data were collected for the rst 3 min and second 3 min of a 6-min trial, and the second 3 min is shown. A three-factor mixed ANOVA (stress treatment time) with post-hoc Fishers LSD test was used to test for differences. There was no stress effect, but a signicant treatment (P < 0.01), time (P < 0.001) and treatmenttime interaction (P < 0.04). Fishers LSD (**P < 0.01) showed that both oestradiol treated groups made more inner visits in the second 3 min than the cholesterol treated group. Data are from (24).

Fig. 4. Levels of gonadal hormones in young adult and aged, male and female rats. (A) Serum oestradiol: levels of oestradiol in young, 2.5-monthold, male and female (solid bars, n 6) and aged, 21.5-month-old, male (n 7) and female (n 8) rats. Entries are the mean SEM. Data analysed by two-way ANOVA, sex age, and showed a signicant interaction (P < 0.01); bars with different superscripts (a, b) are signicantly different from each other (P < 0.05) by post-hoc Fishers LSD test. Young data are from unpublished data from Luine and MacLusky, and aged data are from (33). (B) Serum testosterone: levels of testosterone in young, 2.5-month-old, male and female (solid bars, n 6) and aged, 21.5-month-old, male (n 7) and female (n 8) rats. Entries are the mean SEM. Data were analysed by two-way ANOVA, sex age, and showed a signicant sex effect (P < 0.00001) and an interaction trend (P < 0.08); bars with different superscripts (a, b, c) are signicantly different from each other (P < 0.05) by post-hoc Fishers LSD test. Young data are unpublished data from Luine and MacLusky, and aged data are from (33).

oestradiol was higher in the aged males compared to the young males, which resulted in aged males having approximately two-fold higher circulating oestradiol levels than aged females. Testosterone, on the other hand, showed a modest decline with ageing in the males (30%), and did not differ between the old and young females, with both female groups having very low levels compared to males (0.03 ng ml; Fig. 4B). These changes in circulating oestradiol levels may have contributed to the different responses to stress of aged compared to young rats (Figs 13). As shown in Fig. 3(B), aged males are no longer impaired in object placement following stress. Like OVX females, aged females with lowered oestradiol did not show enhanced object placement performance following

chronic stress (Fig. 3B). For object recognition, both males and females performed better than control rats after the chronic stress. Higher oestradiol levels may account for the enhancement in males, but mechanisms for the enhanced performance of object recognition in the aged, stressed females is unclear. Finally, oestrogens may act as anxiolytic agents. Young males were more anxious than young females to enter the open eld (Fig. 2A), possibly due to lower oestradiol levels. Consistent with this hypothesis, young, OVX females were more anxious than OVXoestrogen replaced females on the open eld (Fig. 5). Oestradiol treated subjects, control or stressed, made more entries into the centre of the open eld than OVX subjects, control or stressed. Ovariectomised females treated with oestradiol also showed more entries into open arms of a plus maze, suggesting an anxiolytic effect (40). On the open eld, oestradiol treatment did not signicantly interact with stress on entry latencies (Fig. 5), whereas acute stress attenuated the oestrogen-dependent anxiolytic effect in females on the plus maze (40). It is also notable that levels of anxiety between the sexes changed with ageing and that aged males became less anxious and entered the open eld with the same latency as females (Fig. 2B). Using the plus maze, aged male Wistar rats also showed less anxiety than young males (41; females were not analysed). Thus, the age-related changes in anxiety in males may be related to increased oestradiol levels in aged males compared to aged females (Fig. 4A). It is also notable that stress effects on anxiety were lower in the aged rats of both sexes and were opposite in direction to the effects in young adult rats.

2007 The Authors. Journal Compilation 2007 Blackwell Publishing Ltd, Journal of Neuroendocrinology, 19, 743751

Chronic stress and neural function

749

Neural mechanisms for stress effects

The mechanisms whereby sex and age inuence stress effects are largely unknown, but several hypothesis have received attention. Likewise, how differences in circulating oestradiol, which contributes to behavioural differences (see above), alter downstream modulators of neural function is also unknown. A likely candidate for regulation by both stress and gonadal hormones is brain derived neurotrophic factor (BDNF) because this neurotrophin positively affects many aspects of neural function and is present in high amounts in the hippocampus (42). Recently, 3 weeks of restraint was shown to reversibly decrease hippocampal BDNF levels in male rats (43) and, when BDNF was infused into the hippocampus, it protected against stress-induced impairments in the spatial memory task, water maze (44). Thus, lowering of BDNF in males may contribute to stress-dependent impairments in spatial memory. Effects in females have not been examined, and some results (45) are not consistent with these results. Moreover, oestradiol is known to increase hippocampal BDNF levels (46, 47), which may account for the different cognitive responses to stress in young adult females (enhancements in spatial memory) compared to males. In the prefrontal cortex, chronic stress was associated with reduced phosphoCREB (42) which is important for memory function in general and BDNF signalling specically (47, 48). The polypeptide hormone oxytocin is also a potential target for modulation of the stress response by oestradiol because oxytocin is higher in females, is important for social recognition (49) and may enhance spatial memory as well (50). Monoaminergic neural systems were among the rst to be identied as stress sensitive, both to acute and chronic stress. For brain areas contributing to memory, decreased norepinephrine, dopamine and serotonin levels have been demonstrated in the hippocampus of chronically stressed males (1014, 21) whereas we have reported an opposite change in response to stress in females; serotonin and norepinephrine levels increase following chronic stress in females (1012, 21, 24). In addition, dopaminergic activity in the medial prefrontal cortex is decreased in males and increased in females following chronic stress (11, 12, 21). Whether stress-dependent changes in monoamines result from direct effects of corticosterone or indirect effects via the neural factors discussed above is unknown, but a recent study in serotonin transporter (SERT) BDNF knockout mice (1 BDNF allele) showed that males had decreased serotonin and dopamine levels whereas females had smaller reductions in serotonin and no reductions in dopamine (15). These results suggest a complex relationship between sex sex hormones, neurotrophins and amines. Finally, stress-dependent alterations in glutamatergic systems are prominent in the hippocampus, but little information is available in female rats for comparison.

but, when stress continues on a chronic basis, these responses become maladaptive. In terms of brain function, robust sex differences in cognitive and anxiolytic responses to stress are evident both qualitatively and quantitatively, and females show adaptive responses to chronic stress and maintain the adaptive neural responses longer than males; however, because most studies of stress have been completed in male subjects, it not clear whether the CNS is unique compared to peripheral systems in regard to these sex differences in the observed response. CNS responses to stress are also different depending upon when during the lifespan stress is encountered. In aged rats, stress enhances performance of some cognitive tasks and alters anxiety differently than in young rats. In foetuses, chronic stress exerts lifelong changes in cognition and anxiety, and the effects are also dependent on the sex of the fetus. Some evidence shows that sex and age-dependent differences in stress responses may be dependent on changes differences in circulating oestrogens. However, these age and sexdependent differences in the response to stress are based on only a few studies, applied mainly restraint stress and investigated a limited number of behaviours; therefore, these results require further investigation and validation. Moreover, little information is available concerning the mechanisms responsible for these stress effects. Nonetheless, the current results highlight that sex, age and gonadal hormone state should be considered in stress research because these variables can impact the results and may have important consequences for understanding stress related disorders and diseases.

Acknowledgements
This review was presented, in part, at the Sixth International Congress of Neuroendocrinology in Pittsburgh, PA (USA) in June 2006, and the authors thank the meeting organisers and this journals editors for the opportunity to present the results. This research was supported by NIH Grants GM60654, DA121136 to V.N.L.; GM60665, RR03037 to Hunter College; NRSA MH12515 to R.E.B.; and VA Merit Review Funds to K.D.B.

Received: 2 September 2006, revised 25 May 2007, accepted 22 June 2007

References
1 Selye H. The Stress of Life. New York, NY: McGraw-Hill, 1976. 2 Wellman L. Dendritic reorganization in pyramidal neurons in medial prefrontal cortex after chronic corticosterone administration. J Neurobiol 2001; 49: 245253. 3 Radley JJ, Sisti HM, Hao J, Rocher AB, McCall T, Hof PR, McEwen BS, Morrison JH. Chronic behavioral stress induces apical dendritic reorganization in pyramidal neurons of the medial prefrontal cortex. Neuroscience 2004; 125: 16. 4 Radley JJ, Rjocher AB, Miller M, Janssen WG, Listen C, Hof PR, McEwen BS, Morrison JH. Repeated stress induces dendritic spine loss in the rat medial prefrontal cortex. Cereb Cortex 2006; 16: 313320. 5 Watanabe Y, Gould E, McEwen BS. Stress induces atrophy of apical dendrites of hippocampal CA3 pyramidal neurons. Brain Res 1992; 588: 341345.

Conclusions
The accumulating data demonstrate that stress effects on neural function may be different than on physiological systems outside of the brain. In the periphery, acute stress initiates adaptive responses,

2007 The Authors. Journal Compilation 2007 Blackwell Publishing Ltd, Journal of Neuroendocrinology, 19, 743751

750

V. N. Luine et al.

6 Magarinos AM, Verdugo JM, McEwen BS. Chronic stress alters synaptic terminal structure in hippocampus. Proc Natl Acad Sci USA 1997; 94: 1400214008. 7 McLaughlin KF, Baran SE, Wright RL, Conrad CD. Chronic stress enhances spatial memory in ovariectomized female rats despite CA3 dendritic retraction. Possible involvement of CA1 neurons. Neuroscience 2005; 135: 10451054. 8 Donohue HS, Gabbott PL, Davies HA, Rodriques JJ, Cordero MI, Sandi C, Medvedev NI, Popov VI, Colyer FM, Peddie CJ, Stewart MG. Chronic restraint stress induces changes in synapse morphology in stratum lacuosum-moleculare CA1 rat hippocampus: a stereological and threedimensional ultrastructural study. Neuroscience 2006; 140: 597606. 9 Luine V, Spencer RL, McEwen BS. Effects of chronic corticosterone ingestion on spatial memory performance and hippocampal serotonergic function. Brain Res 1993; 616: 6570. 10 Beck KD, Luine VN. Food deprivation modulates chronic stress effects on object recognition in male rats: role of monoamines and amino acids. Brain Res 1999; 830: 5671. 11 Beck KD, Luine VN. Sex differences in behavioral and neurochemical proles after chronic stress: role of housing conditions. Physiol Behav 2002; 75: 661673. 12 Luine VN. Sex differences in chronic stress effects on memory in rats. Stress 2002; 5: 205216. 13 Sunanda RBS, Raju TR. Restraint stress-induced alterations in the levels of biogenic amines, amino acids, and AchE activity in the hippocampus. Neurochem Res 2000; 25: 15471552. 14 Torres IL, Gamaro GD, Vasconcellos AP, Silveira R, Dalmaz C. Effects of chronic restraint stress on feeding behavior and on monoamine levels in different brain structures in rats. Neurochem Res 2002; 27: 519525. 15 Ren-Patterson RF, Cochran LW, Holmes A, Lesch KP, Lu B, Murphy DL. Gender-dependent modulation of brain monoamines and anxiety-like behaviors in mice with genetic serotonin transporter and BDNF deciencies. Cell Mol Neurobiol 2006; 26: 755780. 16 Kitraki E, Kremmyda O, Youlatos D, Alexis MN, Kittas C. Gender-dependent alterations in corticosteroid receptor status and spatial performance following 21 days of restraint stress. Neuroscience 2004; 125: 4755. 17 Luine V, Villegas M, Martinez C, McEwen BS. Repeated stress causes reversible impairments of spatial memory performance. Brain Res 1994; 639: 167170. 18 Conrad CD, Galea LA, Kuroda Y, McEwen BS. Chronic stress impairs rat spatial memory on the Y maze, and this effect is blocked by tianeptine pretreatment. Behav Neurosci 1996; 110: 13211334. 19 Luine V, Martinez C, Villegas M, Magarinos AM, McEwen BS. Restraint stress reversibly enhances spatial memory performance. Physiol Behav 1996; 59: 2732. 20 Bowman RE, Micik R, Luine VN. Sex-dependent effects of short-term restraint stress on anxiety and cognitive functioning are task specic. Soc Neurosci Abstract 2006; 5: 778.1. 21 Bowman RE, Beck KD, Luine VN. Chronic stress effects on memory. Sex differences in performance and monoamines. Horm Behav 2003; 43: 4859. 22 Bowman RE. Stress-induced changes in spatial memory are sexually differentiated and vary across the lifespan. J Neuroendocrinol 2005; 17: 526535. 23 Bowman RE, Zrull MC, Luine VN. Chronic restraint stress enhances radial arm maze performance in female rats. Brain Res 2001; 904: 279289. 24 Bowman RE, Ferguson D, Luine VN. Effects of chronic restraint stress and estradiol on open eld activity, spatial memory, and monoaminergic neurotransmitters in ovariectomized rats. Neuroscience 2002; 113: 401410.

25 Conrad CD, Grote KD, Hobbs RJ, Ferayorni A. Sex differences in spatial and non-spatial Y-maze performance after chronic stress. Neurobiol Learn Mem 2003; 79: 3240. 26 Galea LA, McEwen BS, Tanapat P, Deak T, Spencer RL, Dhabha FS. Sex differences in dendritic atrophy of CA3 pyramidal neurons in response to chronic restraint stress. Neuroscience 1997; 81: 689697. 27 Bowman R, MacLusky NJ, Sarmiento Y, Frankfurt M, Gordon M, Luine VN. Sexually dimorphic effects of prenatal stress on cognition, hormonal responses and central neurotransmitters. Endocrinology 2004; 145: 37783787. 28 Bisagno V, Grillo CA, Piroli GG, Giraldo P, McEwen BS, Luine VN. Chronic stress alters amphetamine effects on behavior and synaptophysin levels in female rats. Pharmac Biochem Behav 2004; 78: 541550. 29 Diamond DM, Campbell AM, Park CR, Woodson JC, Conrad CD, Bachsteter AD, Mervis RF. Inuence of predator stress on the consolidation versus retrieval of long-term spatial memory and hippocampal spinogenesis. Hippocampus 2006; 16: 571576. 30 Mitra R, Vyas A, Chatterjee G, Chattarji S. Chronic-stress induced modulation of different states of anxiety-like behavior in female rats. Neurosci Lett 2005; 383: 278283. 31 Gotz AA, Stefanski V. Psychosocial maternal stress during pregnancy affects serum corticosterone, blood immune parameters and anxiety behaviour in adult male rat offspring. Physiol Behav 2007; 90: 108115. 32 Figueiredo HF, Dolgas CM, Herman JP. Stress activation of cortex and hippocampus is modulated by sex and stage of estrus. Endocrinology 2002; 143: 25342540. 33 Bowman R, Maclusky NJ, Zrull MJ, Diaz SE, Luine VN. Aged rats: Sex differences and responses to chronic stress. Brain Res 2006; 1126: 156166. 34 Di Nicola AF, Saravia FE, Beauquis J, Pietranera L, Ferrini MG. Estrogens and neuroendocrine hypothalamic-pituitary-adrenal axis function. Front Horm Res 2006; 35: 157168. 35 Viau V, Meany MJ. Variations in the hypothalamic-pituitary-adrenal response to stress during the estrous cycle in the rat. Endocrinology 1991; 129: 25032511. 36 McEwen BS. The neurobiology of stress: from serendipity to clinical relevance. Brain Res 2000; 886: 172189. 37 Ward IL, Ward OB. Sexual behavior differentiation: effects of prenatal manipulations in rats. In: Adler N, Pfaff D, Goy RW, eds. Handbook of Behavioral Neurobiology. New York, NY: Plenum, 1985: 301309. 38 Garcia-Segura LM, Azcortia I, DonCarlos LC. Neuroprotection by estrogen. Prog Neurobiol 2001; 63: 2960. 39 Conrad CD, Jackson JL, Wieczorek L, Baran SE, Harman JS, Wright RL, Korol DL. Actue stress impairs spatial memory in male but not female rats: inuence of estrous cycle. Pharmacol Biochem Behav 2004; 78: 569579. 40 Walf AA, Frye CA. Antianxiety and antidepressive behavior produced by physiological estradiol regimen may be modulated by hypothalamic-pituitary-adrenal axis activity. Neuropsychopharmacology 2005; 30: 1288 1201. 41 Torras-Garcia M, Costa-Miserachs D, Coli-Andreu M, Portell-Cortes I. Decreased anxiety levels related to aging. Exp Brain Res 2005; 164: 177184. 42 Scharfman HE, Maclusky NJ. Similarities between actions of estrogen and BDNF in the hippocampus: coincidence or clue? Trends Neurosci 2005; 28: 7985. 43 Luo C, Xu H, Li XM. Post-stress changes in BDNF and Bcl-2 immunoreactivities in hippocampal neurons: effect of chronic administration of olanzapine. Brain Res 2004; 1025: 194202. 44 Radecki DT, Brown LM, Martinez J, Teyler TJ. BDNF protects against stress-induced impairments in spatial learning and memory and LTP. Hippocampus 2005; 15: 246253.

2007 The Authors. Journal Compilation 2007 Blackwell Publishing Ltd, Journal of Neuroendocrinology, 19, 743751

Chronic stress and neural function

751

45 Kuroda Y, McEwen BS. Effect of chronic restraint stress and tianeptine on growth factors, growth-associated protein-43 and microtubule-associated protein 2 mRNA expression in the rat hippocampus. Brain Res Mol Brain Res 1998; 59: 3539. 46 Blurton-Jones M, Kuan PN, Tuszynski MH. Anatomical evidence for transsynaptic inuences of estrogen on brain-derived neurotrophic factor expression. J Comp Neurol 2004; 468: 347360. 47 Zhou J, Zhang H, Cohen RS, Pandey SC. Effects of estrogen treatment on expression of brain-derived neurotrophic factor and cAMP response element-binding protein expression and phosphorylation in rat amygdaloid and hippocampal structures. Neuroendocrinology 2005; 81: 294310.

48 Kuipers SD, Trentani A, Den Boer JA, Ter Horst GJ. Molecular correlates of impaired prefrontal plasticity in response to chronic stress. J Neurochem 2003; 85: 13121323. 49 Belsky IF, Young LJ. Oxytocin, vasopressin and social recognition in mammals. Peptides 2004; 25: 15651574. 50 Tomizawa K, Iga N, Lu YG, Moriwaki A, Matsushita M, Li ST, Miyamoto O, Itano T, Matsui H. Oxytocin improves long-lasting spatial memory during motherhood through MAP kinase cascade. Nat Neurosci 2003; 6: 384390.

2007 The Authors. Journal Compilation 2007 Blackwell Publishing Ltd, Journal of Neuroendocrinology, 19, 743751