Stability of Neuroendocrine and Behavioral Responsiveness in Aging Fischer 344/Brown-Norway Hybrid Rats
http://www.100md.com
内分泌学杂志 2005年第7期
Psychiatry Service (J.W.K.), Cincinnati Veterans Affairs Medical Center, Cincinnati, Ohio 45220; and Departments of Psychiatry (J.W.K., T.M.S., C.X., A.R.F., N.K.E., M.M.O., J.P.H.) and Surgery (C.X.), University of Cincinnati, Cincinnati, Ohio 45221
Address all correspondence and requests for reprints to: James P. Herman, Genomic Research Institute, 2170 East Galbraith Road, Building E, Room 216, Cincinnati, Ohio 45237. E-mail: hermanjs@ucmail.uc.edu.
Abstract
Aging in rodents and primates is accompanied by changes in hypothalamic-pituitary-adrenal (HPA) activity. We examined behavioral and neuroendocrine responses in 3, 15-, and 30-month-old F344/Brown-Norway rats. Basal corticosterone and ACTH levels did not differ with age, although ACTH responses, but not corticosterone responses to restraint stress, were significantly lower in the 30-month-old group relative to 3- and 15-month-old rats. Induction of c-fos mRNA in the paraventricular nucleus from restraint was not affected by age. Furthermore, there was an enhanced sensitivity to dexamethasone suppression in aged animals as evidenced by lesser ACTH and corticosterone release after dexamethasone administration. Evaluation of emotional behaviors in the forced swim test revealed no differences between the age groups. With fear conditioning, aged rats had decreased freeze times relative to middle-aged or young rats. Regression analysis revealed no significant correlations between the behavioral and HPA axis data in any group. Overall, the data suggest that an apparent decrease in pituitary drive is compensated for at the level of the adrenal, resulting in stable patterns of glucocorticoid secretion. The lack of a correlation between HPA axis measures and emotional as well as fear conditioning-related behaviors indicates that corticosteroid dysfunction may not predict age-related behavioral deficits in this aging model.
Introduction
THE AGING PROCESS is accompanied by homeostatic changes in hypothalamo-pituitary-adrenocortical (HPA) responses to stress. This has been confirmed in rodents, dogs, and humans (1, 2, 3). Studies in multiple rat strains associate aging with enhanced responsiveness of the HPA axis to acute and chronic stress (3, 4, 5, 6, 7, 8). Age-related HPA axis hyperresponsiveness is associated with decreased hippocampal glucocorticoid and/or mineralocorticoid receptor expression (3, 4, 5, 6, 7, 8, 9), suggesting a connection between glucocorticoid hyperresponsiveness and altered glucocorticoid signaling in suprahypothalamic stress-regulatory circuitry.
Age-related changes in HPA axis function are correlated with cognitive decline and altered hippocampal glucocorticoid signaling. Previous work indicates that aged Fischer 344 (F344) rats exhibiting impaired spatial memory show decreased hippocampal glucocorticoid receptor binding and prolonged corticosterone responses to restraint (6). Similarly, glucocorticoid receptor mRNA expression is significantly reduced in aged, memory-impaired Long-Evans rats, compared with young or aged unimpaired rats; corticosterone responses were also prolonged in the aged-impaired group (4). These studies indicate an important relationship between age-related changes in HPA function and cognitive decline, and indicate that there are individual differences in susceptibility to the deleterious effects of aging.
In humans and other primates, there are reports suggesting that there are minimal changes associated with HPA axis function and aging. For instance, in studies by Kudielka et al. (10), humans exhibited no age-related differences in basal HPA activity or ACTH, plasma cortisol, and salivary cortisol responses to the Trier Social Stress Test. Kudielka et al. (11) later reanalyzed data from five independent studies in humans to further investigate the impact of age and gender on HPA responses to an acute psychosocial test. They determined that ACTH responses to stress were higher in younger adults, compared with older adults, particularly in the males. However, for total plasma cortisol, the pattern of reactivity did not differ with regard to age, suggesting decreased adrenal sensitivity to ACTH in aging.
Gust et al. (12) examined HPA responses to dexamethasone challenge with regard to aging in female rhesus monkeys; they demonstrated that the older monkeys exhibited significant elevations in cortisol 19 h after dexamethasone administration, but before this time point at both 10 and 15 h, there were no age-related effects. The authors concluded that the data are consistent with a diminished sensitivity to glucocorticoid-negative feedback with aging.
The manner in which aging affects the relationship of HPA neuroendocrine indices with emotional and cognitive behaviors has not yet been explored in detail. There is some indication of diminished feedback efficacy in aged F344 rats (13), but how this relates to behavioral changes has not yet been evaluated. Therefore, the current study sought to examine the connection between age-related HPA responsiveness, dexamethasone suppression of ACTH release, behaviors, and cognitive function using the F344/Brown-Norway (BN) F1 hybrid aging model. The F344/BN strain is free of systematic age-related pathologies associated with other aging strains (e.g. renal disease and splenomegaly in F344s) and has been put forth as a model of choice for aging studies by the National Institute of Aging (14, 15).
Materials and Methods
Subjects
Male F344/BN F1 hybrid rats aged 3, 15, and 30 months (n = 12 for 3 and 15 months and n = 10 for 30 months) were acquired from the NIA colony maintained at Harlan Labs (Indianapolis, IN). The NIA colony was maintained under specific pathogen-free conditions. All animals were housed three per cage at the University of Cincinnati, in a constant temperature- and humidity-controlled environment on a 12-h light, 12-h dark cycle with the lights on at 0600 h.
Animal protocols
All procedures were approved by the University of Cincinnati Institutional Animal Care and Use Committee. Young, middle-aged, and aged rats were subjected to a battery of neuroendocrine and behavioral tests over the course of several weeks (sequence summarized in Figure 1). In all cases, animals received at least 7 d between each paradigm to allow recovery from the effects of prior testing. All of the procedures were performed in the summer months and in parallel over several days. We made sure that there were equal numbers of rats represented from each age group on each day of testing. Testing was performed at the same time of day. At the end of the testing protocols, animals were exposed to an additional 60-min restraint and killed by rapid decapitation immediately after removal from restrainers. The time of the killing was the same each day. One of the purposes of the final restraint was to verify that any observed HPA axis changes seen in the first session carried through the duration of the study.
FIG. 1. Time sequence of experiments. The lower numbers indicate day after arrival.
Restraint stress
Animals were initially subjected to a restraint stress test. Given the large size of middle-aged and aged animals, adjustable restraint cages were fashioned from Plexiglas; plastic inserts were used to ensure that all rats had a similar snug fit in the apparati. Tail blood was sampled by tail nick immediately on placement in the restrainers and represent basal hormone measures (0 min time point). After the initial 30-min restraint, blood was again collected by tail nick, and the subjects were returned to their home cages. Additional samples were attained 60 and 120 min post stress. Approximately 250–300 μl were collected each time blood was obtained, enough to allow dual- or triple-point determinations of ACTH and corticosterone (CORT). As mentioned above, a final 60-min restraint stress was performed at the end of all of the experiments (after fear conditioning) as per the same restraint procedures described above.
Forced swim test
Based on previous methods (16), subjects were placed in a round Plexiglas cylinder (46 cm high x 21 cm wide) filled with 21–25 C water at a depth of 30 cm for a 15-min pretest on d 1. On d 2, subjects were again placed into the same cylinders for 5 min, and their behavior was videotaped for later visual analysis. Subjects were dried with a towel and returned to their home cage after each exposure to the test. Scorers (who were blind to the group assignments) classified the subjects’ behavior as immobile, swimming, climbing, or diving every 5 sec and tallied the occurrence of each behavior for the entire 5 min on the test day.
Dexamethasone suppression test
The dexamethasone suppression test was conducted as noted by Mizoguchi et al. (17), with the intent of prohibiting the normal circadian rise in corticosterone. Briefly, this test assesses the ability of dexamethasone to block circadian rises in CORT in unstressed animals. Animals received an ip injection of dexamethasone (30 μg/kg) at 0900 h and were then returned to their home cages. At 1300, 1500, and 1700 h, blood was sampled via tail nick.
Fear conditioning
Subjects were tested based on methods of Moita et al. (18), with minor modifications. They were placed in the right chamber of a two-chamber Gemini avoidance system (San Diego Instruments, San Diego, CA). The floor was composed of parallel stainless steel rods that could be electrified to various intensities. Subjects could not escape the shock and were confined to the right side for the entire 5 min. On d 1 of testing, subjects were placed into the chamber with the house lights on for 1 min. After 1 min, they were given a 1 mA foot shock for 1 sec and then remained in the cage for an additional 5 min. (It should be noted that the constant-current source adjusts for the resistance of the contact area and thus any observed differences are unlikely to be related to age-related changes in skin resistance.) After testing, subjects were returned to their home cage. The chamber was sanitized between subjects to eliminate any odor. On the second day, subjects were simply placed into the chamber for 5 min without shock and their behavior was videotaped for later examination by scorers blind to treatment. The time spent freezing was recorded.
In situ hybridization
Brains were sectioned at 16 μm using a Microm cryostat (Kalamazoo, MI) mounted on Gold Seal slides (BD Biosciences, Portsmouth, NH) and stored at –20 C. For in situ hybridization, sections were fixed in 4% phosphate-buffered paraformaldehyde for 10 min and rinsed twice in 5 mM potassium PBS (KPBS), twice in 5 mM KPBS with 0.2% glycine, and two more times in KPBS for 5 min. Sections were then acetylated in 0.25% acetic anhydride [suspended in 0.1 M triethanolamine (pH 8)] for 10 min, rinsed twice in 2x standard saline citrate (SSC) for 5 min, and dehydrated through graded alcohols.
Antisense rat c-fos probe was generated by in vitro transcription using [35S]-uridine 5-triphosphate (UTP) as a label. The c-fos DNA construct was a 398-bp fragment of a pGEM4Z full-length construct provided by Dr. T. Curran. The specificity of this probe has been validated in previous studies (19). Briefly, plasmid was linearized with HindIII and then transcribed with T7 polymerase. The transcription reaction consisted of 10x transcription buffer; 125 μCi [35S]-UTP; 200 μM ATP, CTP, and GTP; 10 μM unlabeled UTP; 100 mM dithiothreitol; and 40 U/μl SP6 RNA polymerase. The mixture was incubated for 90 min at 37 C after which the template DNA was digested with ribonuclease-free deoxyribonuclease and probe was separated from free nucleotides by ammonium acetate precipitation.
Radiolabeled c-fos probes were diluted in hybridization buffer [50% formamide, 20 mM Tris-HCl (pH 7.5), 1 mM EDTA, 335 mM NaCl, 1x tRNA, 20 mM dithiothreitol, and 10% dextran sulfate] to yield 1 million cpm per 50 μl buffer. Diluted aliquots were applied to each slide whereupon the slides were coverslipped and incubated overnight at 50 C in humidified chambers containing 50% formamide. The next day coverslips were removed in 2x SSC, and slides were incubated in 100 μg/ml ribonuclease A for 30 min at 37 C. Slides were briefly rinsed in 2x SSC, washed three times in 0.2x SSC (65 C), dehydrated, and exposed to Bio-MAX x-ray film (Kodak, Rochester, NY). Image analysis was performed using National Institutes of Health Image 1.62 software. In all cases, the paraventricular nucleus (PVN) region was sampled and PVN gray level values corrected for background (determined from a hybridization-negative region from the same section). Average PVN corrected gray level was calculated from two to six PVN images from each animal, with the mean value from each animal used for statistical analysis.
Hormone assays
Plasma samples were collected and stored at –20 C. Plasma CORT was assessed by RIA using a kit (MP Biomedicals, Inc., Costa Mesa, CA). Plasma ACTH concentrations were determined by RIA using a specific antiserum generously donated by Dr. William Engeland (University of Minnesota, Minneapolis, MN) at a dilution of 1:210,000 and [125I] ACTH (Amersham Biosciences, Piscataway, NJ) as labeled tracer (20). Plasma samples for each experiment were processed at the same time.
Data analysis
Hormone time-course data were analyzed by repeated-measures ANOVA. Total ACTH and CORT secretion was determined by measuring area under the curve and along with c-fos mRNA expression and the behavioral end points, they were subjected to analysis using one-way ANOVA. In all cases, Newman-Keuls post hoc testing was used to identify differences among the experimental groups.
Previous work has observed that HPA reactivity in aged rats is related to behavioral deficits, suggesting that age-related cognitive impairments are correlated with HPA dysfunction. To test this hypothesis, we performed regression analyses of our behavioral and neuroendocrine data both between and within age groups, using immobility in the forced swim test, freezing time in the fear conditioning test, and the area under the curve as a value for total ACTH and CORT responses to restraint stress and in the dexamethasone suppression test. In addition, we ran the equality of variance F test of homoscedasticity to determine whether the variance within the three age groups differed significantly on any of these tests.
Results
Effects of restraint stress
We examined the effects of restraint on ACTH and corticosterone secretion in 3-, 15-, and 30-month-old F344/BN rats. These changes are depicted in Fig. 2, A–D. There were significant effects of time and age on ACTH levels (for age: F2,124 = 17.8, P < 0.01, time: F3,124 = 69.5, P < 0.01; interaction of time x age: F6,124 = 2.48, P < 0.05) based on two-way repeated-measures ANOVA. Post hoc analysis using Newman-Keuls test revealed that aged animals secreted significantly less ACTH than either the young or middle-aged groups (at 30 min: 30 vs. 15 months, P < 0.01; 3 vs. 30 and 15 months, P < 0.05; at 60 min: 30 vs. 15 and 3 months, P < 0.01; at 120 min: 30 vs. 3 and 15 months, P < 0.05). Repeated-measures two-way ANOVA revealed a significant effect of time on plasma CORT levels (F3,124 = 91.8, P < 0.01); however, CORT responses to restraint did not differ across age groups (P > 0.05). Aged animals also exhibited lower cumulative ACTH secretion over the testing period in comparison with young and middle-aged animals after expressing the data as area under the curve (Fig. 2B; P < 0.01 based on one-way ANOVA and Neuman-Keuls). However, there were no differences in total CORT levels based on area under the curve determinations (Fig. 2D; P > 0.05 based on one-way ANOVA).
FIG. 2. A and C, Respectively, depict the effects of the initial restraint on ACTH and CORT secretion in 3-, 15-, and 30-month-old F344/BN hybrid rats. The ordinate represents concentration of hormone (picograms per milliliter for ACTH and nanograms per milliliter for CORT). The abscissa represents time after restraint. As explained in Results, there were significant effects of time and age on ACTH levels based on two-way repeated-measures ANOVA (P < 0.01). In addition, there was a significant interaction effect at P < 0.05. Post hoc analysis using Newman-Keuls test revealed that aged animals secreted significantly less ACTH than either the young or middle-aged groups (, P < 0.01 at 30 min, 30 vs. 15 months; **, P < 0.05, 3 vs. 30 and 15 months; , P < 0.01 at 60 min, 30 vs. 15 and 3 months; **, P < 0.05, at 120 min, 30 vs. 3 and 15 months). Plasma CORT responses to restraint were not significantly different among the different age groups of rats (P > 0.05). B and D, Respectively, represent total ACTH and CORT secretion in 3-, 15-, and 30-month-old F344/BN hybrid rats resulting from the effects of the initial restraint. The ACTH responses in the older rats were significantly different from both 3- and 15-month-old rats (, P < 0.01 based on one-way ANOVA and Neuman-Keuls). Plasma CORT responses to restraint were not significantly different among the different age groups of rats. In all groups, n = 12 for 3- and 15-month-old rats and n = 10 for 30-month-old rats.
Forced swim test
Figure 3 depicts the effects of the forced swim test (FST) on each age group with respect to immobility, swimming time, climbing, or diving. There were no significant differences among the three age groups with regard to any of these four parameters (P > 0.05 based on one-way ANOVA).
FIG. 3. The figure depicts data for the FST. The ordinate depicts time of immobility as a function of age. The differences among groups of rats were not significantly different from each other based on one-way ANOVA (P > 0.05). In all groups, n = 12 for 3- and 15-month-old rats and n = 10 for 30-month-old rats.
Effects of dexamethasone challenge
We next examined the effects of dexamethasone challenge on ACTH and CORT secretion in 3-, 15-, and 30-month-old F344/BN rats. These changes are depicted in Fig. 4. Repeated-measures two-way ANOVA revealed significant effects of age and time: ACTH [age (F2,93 = 6.0, P < 0.01); time (F2,93 = 3.4, P < 0.05)] and CORT [age (F2,93 = 3.99, P < 0.05); time (F2,93 = 25.3, P < 0.01)]. Post hoc analysis did not reveal any significant difference among any of the age groups at each time point for ACTH secretion (P > 0.05). However, at 8 h, the CORT response in the 30-month-old rats was significantly different from that of the 3-month-old rats (based on Neuman-Keuls; P < 0.01). In addition, there was not a significant interaction effect between time and age for ACTH or CORT (P > 0.05).
FIG. 4. A and C, Respectively, effects of dexamethasone administration on ACTH and CORT secretion in 3-, 15-, and 30-month-old F344/BN hybrid rats. Repeated-measures two-way ANOVA revealed significant effects of age and time: ACTH (age: P < 0.01; time: P < 0.05) and CORT (age: P < 0.05; time: P < 0.01). (C), CORT response at 8 h in the 30-month-old rats was significantly different from that of the 3-month-old rats (based on two-way repeated-measures ANOVA and Neuman-Keuls; , P < 0.01). B and D, Respectively, total ACTH and CORT secretion in 3-, 15-, and 30-month-old F344/BN rats after dexamethasone administration. One-way ANOVA revealed a nonsignificant trend toward lower total ACTH levels in the 30-month-old rats (P = 0.059). Furthermore, for total CORT secretion, there was no difference among age groups (P > 0.05). In all groups, n = 12 for 3- and 15-month-old rats and n = 10 for 30-month-old rats.
For total ACTH secretion, expressed as area under the curve, one-way ANOVA revealed a trend for lower ACTH levels in the oldest group of rats (P = 0.059; see Fig 4B). Furthermore, for total CORT secretion, there was no difference between age groups (P > 0.05).
Fear conditioning
Rats were subjected to a fear conditioning paradigm (Fig. 5). One-way ANOVA revealed a significant effect of age on freezing time as measured 24 h after shock (F2,30 = 4.92, P < 0.05). Surprisingly, freezing time was significantly enhanced only in the middle-aged group as determined by Newman-Keuls test (P < 0.05 for 15 vs. 3 and 15 vs. 30 months old); no differences were observed between 30- and 3-month-old animals (P > 0.05).
FIG. 5. This figure depicts freezing time in the three age groups. One-way ANOVA revealed a significant effect of age on freezing time as measured 24 h after shock (P < 0.05). Freezing time was significantly enhanced only in the middle-aged group as determined by Newman-Keuls test (**, P < 0.05 for 15 vs. 3 and 15 months vs. 30 months old); no differences were observed between 30- and 3-month-old animals (P > 0.05). In all groups, n = 12 for 3- and 15-month-old rats and n = 10 for 30-month-old rats.
Final restraint test
Figure 6 depicts the changes in ACTH (Fig. 6A) and CORT (Fig. 6B) 60 min after the last restraint in 3-, 15-, and 30-month-old rats. Consistent with the original restraint, there were significant effects of restraint on ACTH levels (F2,31 = 5.32, P < 0.01); post hoc Newman-Keuls test revealed that the 30-month-old animals secreted significantly less ACTH than both 3- and 15-month-old groups (P < 0.05).
FIG. 6. A and B, Respectively, total ACTH and CORT secretion resulting from the effects of the final 60-min restraint on in 3-, 15-, and 30-month-old F344/BN rats. There were significant effects of restraint on ACTH levels (P < 0.01); Newman-Keuls post hoc analysis revealed that the 30-month-old animals secreted significantly less ACTH than both 3- and 15-month-old groups (**, P < 0.05). In all groups, n = 12 for 3- and 15-month-old rats and n = 10 for 30-month-old rats.
Stress-induced PVN c-fos mRNA
We sought to examine central induction of c-fos mRNA to assess activation of medial parvocellular PVN neurons after stress. Semiquantitative analysis of c-fos mRNA expression after 60 min restraint revealed that c-fos levels did not differ among the three age groups (P > 0.05; Fig. 7).
FIG. 7. Semiquantitative analysis of c-fos mRNA expression in the PVN. Semiquantitative analysis of c-fos mRNA expression after 60 min restraint revealed that c-fos levels did not differ among the three age groups (P > 0.05 as per one-way ANOVA). In all groups, n = 4 for each age group.
Regression analyses: neuroendocrine function and behavior
Regression analyses were run to determine whether behavioral end points in the forced swim (immobility) and fear conditioning (freeze time) tests correlated with endocrine end points from the stress tests. There was a positive correlation between ACTH secretion in the stress test and dexamethasone suppression test (P < 0.05; r = 0.43; 3 and 15 months: n = 12; 30 months: n = 10; Fig. 8A). No significant correlations were observed between the endocrine measures and behavioral end points (P > 0.05; Fig. 8, B and C). In addition, the equality of variance F test for homoscedasticity, which tests for differences in variances among the different age groups, did not reveal enhancement of variance in the aged group on any endocrine or behavioral measure.
FIG. 8. Regression analyses relating the area under the curve of the ACTH response to restraint stress with ACTH secretion in the dexamethasone suppression test (A), freezing time in the fear conditioning test (B), and immobility time in the FST (C). Note that there was a weak but significant correlation between ACTH levels in the stress test with ACTH levels seen after dexamethasone administration (*, P < 0.05; r = 0.43); however, there were no significant correlations between stress-induced ACTH release and either of the behavioral measures (P > 0.05). In all groups, n = 12 for 3- and 15-month-old rats and n = 10 for 30-month-old rats. AUC, Area under the curve.
Discussion
The current study demonstrates that aged F344/BN F1 hybrid rats exhibit reduced ACTH responses to stress and enhanced sensitivity to dexamethasone suppression of ACTH release. Decreased stress-induced ACTH release appears to be a stable trait in aged rats, as was evident at the initial restraint test at 30 months of age and the final test at 32 months. In the case of the stress response, decreased ACTH release is not accompanied by reduced corticosterone levels, suggesting compensation at the level of the adrenal. The decrements in HPA axis responsiveness were unrelated to performance of behavioral tasks associated with depression or contextual memory.
Several systematic studies of HPA axis function have been conducted in a variety of rodent aging models. Many of these studies have been inconsistent within strains as well as across strains. For instance, at the level of the HPA axis, glucocorticoid hypersecretion (21) and hyposecretion (22, 23) have been reported in aged (F344) rats. Decrements in adrenal responsiveness have also been reported with aging in BN and Sprague Dawley rats (24, 25), whereas our findings in the F344/BN rat and those of Sonntag (F344 strain) (26) are consistent with adrenal hyperresponsiveness.
Of course, strain differences provide one likely explanation for the contradictory data on neuroendocrine responsiveness with aging. One limitation with all of these studies also was that the entire circadian rhythm was not measured in every trial but instead single or only several point measures of ACTH and corticosterone were made. The same limitation noted in these studies also applies to the data reported in this current study. Individual strains of rats are susceptible to tissue- or organ-specific pathologies that can impact HPA axis parameters. For example, the F344 strain is prone to development of pituitary tumors, renal disease, and splenomegaly (14), all of which can directly (e.g. ACTH-secreting adenomas) or indirectly (e.g. space-occupying lesions in the sella turcica/basal hypothalamus, altered fluid-electrolyte status, immune stimulation) impact HPA axis functioning. Indeed, in a previous study, our group found increased spleen size in 75% of aged F344 rats (26). For this reason, our group has more recently employed the F344/BN cross in our aging studies. This strain shows lower incidence of systematic pathologies and has a substantially longer life span than the F344 rat (up to 36 months) (14).
The current data also indicate that aged F344/BN rats exhibit alterations in peripheral limbs of the HPA axis and enhanced responsiveness to dexamethasone suppression. More specifically, the neuroendocrine findings in this report are consistent with an age-related enhancement in the sensitivity of the adrenal gland to ACTH in conjunction with reduced pituitary drive. Previous reports from our group revealed that aged F344/BN rats have reduced pituitary CRH-R1 receptor and proopiomelanocortin mRNA expression (5). Thus, the greater adrenal responsiveness noted here in F344/BN rats appears to be a potential compensatory adaptation to the loss of CRH receptivity and ACTH synthesis. The enhanced responsiveness appears not to be related to adrenal size because we compared total adrenal weight expressed relative to total body weight and found no difference in this variable among the three groups. However, the responsiveness of adrenocortical cells to trophic actions of ACTH could be enhanced in the older group of rats as a result of other factors (27).
Recent human data demonstrated that there were no age-related differences in basal HPA activity or ACTH, plasma cortisol and salivary cortisol responses to the Trier Social Stress Test (10). In addition, a reanalysis of five independent human studies by Kudielka et al. (11) determined that ACTH responses to stress were higher in younger adults, compared with older adults, but total plasma cortisol responses did not differ with regard to age. Thus, these studies are consistent with our data and suggest that age-related neuroendocrine responsiveness in the BN/F344 strain appear to model that observable in humans.
Our group has also previously shown that the older F344/BN rats are hypersensitive to chronic mild stress exposure (5) as reflected in enhanced PVN CRH immunoreactivity, increased pituitary proopiomelanocortin expression, and enhanced forebrain GAD65 up-regulation. However, in our current experiments, we have demonstrated no change across age in F344/BN rats in the FST. The reasons for this are not known.
In addition to the above neuroendocrine and behavioral paradigms, we also assessed fear conditioning. This paradigm has been used as a test of hippocampal-dependent memory (28). Other rat strains show age-related deficits in fear conditioning. Oler and Markus (29) demonstrated that older F344 rats exposed to a fear conditioning paradigm exhibit a weaker ability to retain the conditioning context relative to younger animals at a time at which consolidation should have been completed. Houston et al. (30) also demonstrated in their fear conditioning paradigm that older (27 months old) F344 rats exhibited less memory incubation, compared with younger rats. In contrast, our data indicate that aged and young rats exhibited equivalent freezing times the day after shock exposure. This suggests that deficient performance of behavioral tasks in senescence may be a strain-dependent characteristic. The responsiveness of the middle-aged group in the fear conditioning paradigm was markedly greater than either the young or old groups. The decrease in freeze time observed in aged rats could be interpreted as a contextual memory deficit, in that a response decrement is seen in old age. In contrast, enhanced freezing time in middle-aged rats may reflect enhanced stimulus salience or pain sensitivity in this group, which is effectively reversed in old age.
The emotional component of fear conditioning, however, cannot be ignored. Fear conditioning clearly also has an amygdalar basis. Knierim (31) states that it is the amygdalar-ventral hippocampal pathways and/or amygdalar-perforant path input through the entorhinal cortex, which serve as likely anatomical substrate(s) for the emotional and cognitive processes accounting for fear conditioning. In taking into account this emotional component, it is possible that amygdalar changes with aging could also account for the lower freeze time in the older rats. For instance, our group has demonstrated decreased anxiety-like behaviors in older F344 rats, which is associated with decreased amygdalar CRH expression (32).
Regardless of the interpretation, it appears that the pattern of fear conditioning observed with aging in this rat strain obeys an inverted-U-shaped function, with the most optimal mnemonic processing occurring in middle age. In addition, given the stability of the HPA axis over time in this strain and the absence of a correlation between HPA measures and freeze times, it is unlikely that the observed alterations in contextual memory are due to age-related changes in glucocorticoid secretion.
Previous work in F344 and Long-Evans strains has been able to distinguish subgroups of animals that differ in performance of spatial memory tasks (33, 34). For instance, in the Long-Evans aged rats, poor Morris water maze performance is accompanied by glucocorticoid hypersecretion (35). Despite the relatively large number of F344/BN rats used in our study, we were not able to differentiate any subgroups on tests of fear conditioning, FST behaviors, or neuroendocrine function. Indeed, the stability of these traits occurs at a time when 40–50% of the F344/BN population has died (36), verifying that the population we sampled from is indeed aged.
In summary, it appears that retention of normal glucocorticoid secretory patterns may be a characteristic of successful neuroendocrine aging and perhaps even the aging process in general. It is clear from our work that glucocorticoid hypersecretion and loss of negative feedback are not an immutable part of the aging process in this rodent strain. Indeed, the long-lived F344/BN strain appears to exhibit an active compensation within the axis to maintain normal glucocorticoid tone. It remains to be determined whether age-related HPA dysfunction and accelerated cognitive decline or even other processes are related to the dysregulation of such homeostatic processes in other aging strains (4).
Acknowledgments
We thank Mark Dolgas, Ben Packard, and Yvonne Ulrich-Lai for their assistance with this project.
References
Born J, Ditschuneit I, Schreiber M, Dodt C, Fehm HL 1995 Effects of age and gender on pituitary-adrenocortical responsiveness in humans. Eur J Endocrinol 132:705–711
Reul JM, Rothuizen J, de Kloet ER 1991 Age-related changes in the dog hypothalamic-pituitary-adrenocortical system: neuroendocrine activity and corticosteroid receptors. J Steroid Biochem Mol Biol 40:63–69
Sapolsky RM, McEwen BS, Rainbow TC 1983 Quantitative autoradiography of [3H]corticosterone receptors in rat brain. Brain Res 271:331–334
Bizon JL, Helm KA, Han JS, Chun HJ, Pucilowska J, Lund PK, Gallagher M 2001 Hypothalamic-pituitary-adrenal axis function and corticosterone receptor expression in behaviourally characterized young and aged Long-Evans rats. Eur J Biochem 14:1739–1751
Herman JP, Larson BR, Speert DB, Seasholtz AF 2001 Hypothalamo-pituitary-adrenocortical dysregulation in aging F344/Brown-Norway F1 hybrid rats. Neurobiol Aging 22:323–332
Issa AM, Rowe W, Gauthier S, Meaney MJ 1990 Hypothalamic-pituitary-adrenal activity in aged, cognitively impaired and cognitively unimpaired rats. J Neurosci 10:3247–3254
Morano MI, Vazquez DM, Akil H 1994 The role of the hippocampal mineralocorticoid and glucocorticoid receptors in the hypothalamo-pituitary-adrenal axis of the aged Fisher rat. Mol Cell Neurosci 5:400–412
Van Eekelen JA, De Kloet ER 1992 Co-localization of brain corticosteroid receptors in the rat hippocampus. Prog Histochem Cytochem 26:250–258
Murphy EK, Spencer RL, Sipe KJ, Herman JP 2002 Decrements in nuclear glucocorticoid receptor (GR) protein levels and DNA binding in aged rat hippocampus. Endocrinology 143:1362–1370
Kudielka BM, Schmidt-Reinwald AK, Hellhammer DH, Kirschbaum C 1999 Psychological and endocrine responses to psychosocial stress and dexamethasone/corticotropin-releasing hormone in healthy postmenopausal women and young controls: the impact of age and a two week estradiol treatment. Neuroendocrinology 70:422–430
Kudielka BM, Buske-Kirschbaum A, Hellhammer DH, Kirschbaum C 2004 HPA responses to laboratory psychosocial stress in healthy elderly adults, younger adults, and children: impact of age and gender. Psychoneuroendocrinology 70:83–98
Gust DA, Wilson ME, Stocker T, Conrad S, Plotsky PM, Gordon TP 2000 Activity of the hypothalamic-pituitary-adrenal axis is altered by aging and exposure to social stress in female rhesus monkeys. J Clin Endocrinol Metab 85:2556–2563
Sapolsky R, Meaney MJ 1986 Maturation of the adrenocortical stress response: neuroendocrine control mechanisms and the stress hyporesponsive period. Brain Res Rev 11:65–76
Lipman RD, Chrisp CE, Hazzard DG, Bronson RT 1996 Pathologic characterization of brown Norway, brown Norway x Fischer 344, and Fischer 344 x brown Norway rats with relation to age. J Gerontol A Biol Sci Med Sci 51:B54–B59
Sprout RL 1991 Development of animal models of aging at the National Institute on Aging. Neurobiol Aging 12:635–638
Porsolt RD, Bertin A, Jalfre M 1977 Behavioral despair in mice: a primary screening test for antidepressants. Arch Int Pharmacodyn Ther 229:327–336
Mizoguchi K, Ishige A, Aburada M, Tabira T 2003 Chronic stress attenuates glucocorticoid negative feedback: involvement of the prefrontal cortex and hippocampus. Neuroscience 119:887–897
Moita MA, Rosis S, Zhou Y, LeDoux JE, Blair HT 2004 Putting fear in its place: remapping of hippocampal place cells during fear conditioning. J Neurosci 24:7015–7023
Cullinan WE, Herman JP, Battaglia DF, Akil H, Watson SJ 1995 Pattern and time course of immediate early gene expression in rat brain following acute stress. Neuroscience 64:477–505
Jasper MS, Engeland WC 1991 Synchronous ultraradian rhythms in adrenocortical secretion detected by microdialysis in awake rats. Am J Physiol 261:R1257–R1268
Hauger RL, Thrivikraman KV, Plotsky PM 1994 Age-related alterations of hypothalamic-pituitary-adrenal axis function in male Fischer 344 rats. Endocrinology 134:1528–1536
Cizza G, Calogero AE, Brady LS, Bagdy G, Bergamini E, Blackman MR, Chrousos GP, Gold PW 1994 Male Fischer 344/N rats show a progressive central impairment of the hypothalamic-pituitary-adrenal axis with advancing age. Endocrinology 134:1611–1620
Kasckow JW, Regmi A, Mulchahey JJ, Plotsky PM, Hauger RL 1999 Changes in brain corticotropin-releasing factor messenger RNA expression in aged Fischer 344 rats. Brain Res 822:228–230
Carnes M, Goodman BM, Lent SJ, Vo H, Jaeckels R 1994 Coincident plasma ACTH and corticosterone time series: comparisons between young and old rats. Exp Gerontol 29:625–643
Van Eekelen JA, Oitzl MS, de Kloet ER 1995 Adrenocortical hyporesponsiveness and glucocorticoid feedback resistance in old male brown Norway rats. J Gerontol A Biol Sci Med Sci 50:B83–B89
Herman JP, Chen KC, Booze R, Landfield PW 1998 Up-regulation of 1D Ca2+ channel subunit mRNA expression in the hippocampus of aged F344 rats. Neurobiol Aging 19:581–587
Sonntag WE, Goliszek AG, Brodish A, Eldridge JC 1987 Diminished diurnal secretion of adrenocorticotropin (ACTH), but not corticosterone, in old male rats: possible relation to increased adrenal sensitivity to ACTH in vivo. Endocrinology 120:2308–2315
Rudy JW, Huff NC, Matus-Amat P 2004 Understanding contextual fear conditioning: insights from a two process model. Neurosci Biobehav Rev 28:675–685
Oler JA, Markus EJ 1998 Age-related deficits on the radial maze and in fear conditioning: hippocampal processing and consolidation. Hippocampus 8:402–415
Houston FP, Stevenson GD, McNaughton BL, Barnes CA 1999 Effects of age on the generalization and incubation of memory in the F344 rat. Learn Mem 6:111–119
Knierim JJ 2003 Hippocampus and memory. Can we have our place and fear it too? Neuron 37:372–374
Pisarska M, Mulchahey JJ, Welge J, Geracioti TD, Kasckow JW 2000 Age-related alterations in emotional behaviors and amygdalar corticotropin-releasing factor (CRF) and CRF-binding protein expression in aged Fischer 344 rats. Brain Res 877:959–966
Brightwell JJ, Gallagher M, Colombo PJ 2004 Hippocampal CREBI but not CREB2 is decreased in aged rats with spatial memory impairments. Neurobiol Learn Mem. 81:19–26
Shukitt-Hale B, Mouzakis G, Joseph JA 1998 Psychomotor and spatial memory performance in aging male Fischer 344 rats. Exp Gerontol 33:615–624
Issa BG, Lewis BM, Ham J, Peters JR, Scanlon MF 1998 Glutamate pathways mediate somatostatin responses to glucose in normal and diabetic rat hypothalamus. J Neuroendocrinol 10:377–381
Turturro A, Witt WW, Lewis S, Hass BS, Lipman RD, Hart RW 1999 Growth curves and survival characteristics of the animals used in the Biomarkers of Aging Program. J Steroid Biochem Mol Biol 54:B492–B501(John W. Kasckow, Tracy M.)
Address all correspondence and requests for reprints to: James P. Herman, Genomic Research Institute, 2170 East Galbraith Road, Building E, Room 216, Cincinnati, Ohio 45237. E-mail: hermanjs@ucmail.uc.edu.
Abstract
Aging in rodents and primates is accompanied by changes in hypothalamic-pituitary-adrenal (HPA) activity. We examined behavioral and neuroendocrine responses in 3, 15-, and 30-month-old F344/Brown-Norway rats. Basal corticosterone and ACTH levels did not differ with age, although ACTH responses, but not corticosterone responses to restraint stress, were significantly lower in the 30-month-old group relative to 3- and 15-month-old rats. Induction of c-fos mRNA in the paraventricular nucleus from restraint was not affected by age. Furthermore, there was an enhanced sensitivity to dexamethasone suppression in aged animals as evidenced by lesser ACTH and corticosterone release after dexamethasone administration. Evaluation of emotional behaviors in the forced swim test revealed no differences between the age groups. With fear conditioning, aged rats had decreased freeze times relative to middle-aged or young rats. Regression analysis revealed no significant correlations between the behavioral and HPA axis data in any group. Overall, the data suggest that an apparent decrease in pituitary drive is compensated for at the level of the adrenal, resulting in stable patterns of glucocorticoid secretion. The lack of a correlation between HPA axis measures and emotional as well as fear conditioning-related behaviors indicates that corticosteroid dysfunction may not predict age-related behavioral deficits in this aging model.
Introduction
THE AGING PROCESS is accompanied by homeostatic changes in hypothalamo-pituitary-adrenocortical (HPA) responses to stress. This has been confirmed in rodents, dogs, and humans (1, 2, 3). Studies in multiple rat strains associate aging with enhanced responsiveness of the HPA axis to acute and chronic stress (3, 4, 5, 6, 7, 8). Age-related HPA axis hyperresponsiveness is associated with decreased hippocampal glucocorticoid and/or mineralocorticoid receptor expression (3, 4, 5, 6, 7, 8, 9), suggesting a connection between glucocorticoid hyperresponsiveness and altered glucocorticoid signaling in suprahypothalamic stress-regulatory circuitry.
Age-related changes in HPA axis function are correlated with cognitive decline and altered hippocampal glucocorticoid signaling. Previous work indicates that aged Fischer 344 (F344) rats exhibiting impaired spatial memory show decreased hippocampal glucocorticoid receptor binding and prolonged corticosterone responses to restraint (6). Similarly, glucocorticoid receptor mRNA expression is significantly reduced in aged, memory-impaired Long-Evans rats, compared with young or aged unimpaired rats; corticosterone responses were also prolonged in the aged-impaired group (4). These studies indicate an important relationship between age-related changes in HPA function and cognitive decline, and indicate that there are individual differences in susceptibility to the deleterious effects of aging.
In humans and other primates, there are reports suggesting that there are minimal changes associated with HPA axis function and aging. For instance, in studies by Kudielka et al. (10), humans exhibited no age-related differences in basal HPA activity or ACTH, plasma cortisol, and salivary cortisol responses to the Trier Social Stress Test. Kudielka et al. (11) later reanalyzed data from five independent studies in humans to further investigate the impact of age and gender on HPA responses to an acute psychosocial test. They determined that ACTH responses to stress were higher in younger adults, compared with older adults, particularly in the males. However, for total plasma cortisol, the pattern of reactivity did not differ with regard to age, suggesting decreased adrenal sensitivity to ACTH in aging.
Gust et al. (12) examined HPA responses to dexamethasone challenge with regard to aging in female rhesus monkeys; they demonstrated that the older monkeys exhibited significant elevations in cortisol 19 h after dexamethasone administration, but before this time point at both 10 and 15 h, there were no age-related effects. The authors concluded that the data are consistent with a diminished sensitivity to glucocorticoid-negative feedback with aging.
The manner in which aging affects the relationship of HPA neuroendocrine indices with emotional and cognitive behaviors has not yet been explored in detail. There is some indication of diminished feedback efficacy in aged F344 rats (13), but how this relates to behavioral changes has not yet been evaluated. Therefore, the current study sought to examine the connection between age-related HPA responsiveness, dexamethasone suppression of ACTH release, behaviors, and cognitive function using the F344/Brown-Norway (BN) F1 hybrid aging model. The F344/BN strain is free of systematic age-related pathologies associated with other aging strains (e.g. renal disease and splenomegaly in F344s) and has been put forth as a model of choice for aging studies by the National Institute of Aging (14, 15).
Materials and Methods
Subjects
Male F344/BN F1 hybrid rats aged 3, 15, and 30 months (n = 12 for 3 and 15 months and n = 10 for 30 months) were acquired from the NIA colony maintained at Harlan Labs (Indianapolis, IN). The NIA colony was maintained under specific pathogen-free conditions. All animals were housed three per cage at the University of Cincinnati, in a constant temperature- and humidity-controlled environment on a 12-h light, 12-h dark cycle with the lights on at 0600 h.
Animal protocols
All procedures were approved by the University of Cincinnati Institutional Animal Care and Use Committee. Young, middle-aged, and aged rats were subjected to a battery of neuroendocrine and behavioral tests over the course of several weeks (sequence summarized in Figure 1). In all cases, animals received at least 7 d between each paradigm to allow recovery from the effects of prior testing. All of the procedures were performed in the summer months and in parallel over several days. We made sure that there were equal numbers of rats represented from each age group on each day of testing. Testing was performed at the same time of day. At the end of the testing protocols, animals were exposed to an additional 60-min restraint and killed by rapid decapitation immediately after removal from restrainers. The time of the killing was the same each day. One of the purposes of the final restraint was to verify that any observed HPA axis changes seen in the first session carried through the duration of the study.
FIG. 1. Time sequence of experiments. The lower numbers indicate day after arrival.
Restraint stress
Animals were initially subjected to a restraint stress test. Given the large size of middle-aged and aged animals, adjustable restraint cages were fashioned from Plexiglas; plastic inserts were used to ensure that all rats had a similar snug fit in the apparati. Tail blood was sampled by tail nick immediately on placement in the restrainers and represent basal hormone measures (0 min time point). After the initial 30-min restraint, blood was again collected by tail nick, and the subjects were returned to their home cages. Additional samples were attained 60 and 120 min post stress. Approximately 250–300 μl were collected each time blood was obtained, enough to allow dual- or triple-point determinations of ACTH and corticosterone (CORT). As mentioned above, a final 60-min restraint stress was performed at the end of all of the experiments (after fear conditioning) as per the same restraint procedures described above.
Forced swim test
Based on previous methods (16), subjects were placed in a round Plexiglas cylinder (46 cm high x 21 cm wide) filled with 21–25 C water at a depth of 30 cm for a 15-min pretest on d 1. On d 2, subjects were again placed into the same cylinders for 5 min, and their behavior was videotaped for later visual analysis. Subjects were dried with a towel and returned to their home cage after each exposure to the test. Scorers (who were blind to the group assignments) classified the subjects’ behavior as immobile, swimming, climbing, or diving every 5 sec and tallied the occurrence of each behavior for the entire 5 min on the test day.
Dexamethasone suppression test
The dexamethasone suppression test was conducted as noted by Mizoguchi et al. (17), with the intent of prohibiting the normal circadian rise in corticosterone. Briefly, this test assesses the ability of dexamethasone to block circadian rises in CORT in unstressed animals. Animals received an ip injection of dexamethasone (30 μg/kg) at 0900 h and were then returned to their home cages. At 1300, 1500, and 1700 h, blood was sampled via tail nick.
Fear conditioning
Subjects were tested based on methods of Moita et al. (18), with minor modifications. They were placed in the right chamber of a two-chamber Gemini avoidance system (San Diego Instruments, San Diego, CA). The floor was composed of parallel stainless steel rods that could be electrified to various intensities. Subjects could not escape the shock and were confined to the right side for the entire 5 min. On d 1 of testing, subjects were placed into the chamber with the house lights on for 1 min. After 1 min, they were given a 1 mA foot shock for 1 sec and then remained in the cage for an additional 5 min. (It should be noted that the constant-current source adjusts for the resistance of the contact area and thus any observed differences are unlikely to be related to age-related changes in skin resistance.) After testing, subjects were returned to their home cage. The chamber was sanitized between subjects to eliminate any odor. On the second day, subjects were simply placed into the chamber for 5 min without shock and their behavior was videotaped for later examination by scorers blind to treatment. The time spent freezing was recorded.
In situ hybridization
Brains were sectioned at 16 μm using a Microm cryostat (Kalamazoo, MI) mounted on Gold Seal slides (BD Biosciences, Portsmouth, NH) and stored at –20 C. For in situ hybridization, sections were fixed in 4% phosphate-buffered paraformaldehyde for 10 min and rinsed twice in 5 mM potassium PBS (KPBS), twice in 5 mM KPBS with 0.2% glycine, and two more times in KPBS for 5 min. Sections were then acetylated in 0.25% acetic anhydride [suspended in 0.1 M triethanolamine (pH 8)] for 10 min, rinsed twice in 2x standard saline citrate (SSC) for 5 min, and dehydrated through graded alcohols.
Antisense rat c-fos probe was generated by in vitro transcription using [35S]-uridine 5-triphosphate (UTP) as a label. The c-fos DNA construct was a 398-bp fragment of a pGEM4Z full-length construct provided by Dr. T. Curran. The specificity of this probe has been validated in previous studies (19). Briefly, plasmid was linearized with HindIII and then transcribed with T7 polymerase. The transcription reaction consisted of 10x transcription buffer; 125 μCi [35S]-UTP; 200 μM ATP, CTP, and GTP; 10 μM unlabeled UTP; 100 mM dithiothreitol; and 40 U/μl SP6 RNA polymerase. The mixture was incubated for 90 min at 37 C after which the template DNA was digested with ribonuclease-free deoxyribonuclease and probe was separated from free nucleotides by ammonium acetate precipitation.
Radiolabeled c-fos probes were diluted in hybridization buffer [50% formamide, 20 mM Tris-HCl (pH 7.5), 1 mM EDTA, 335 mM NaCl, 1x tRNA, 20 mM dithiothreitol, and 10% dextran sulfate] to yield 1 million cpm per 50 μl buffer. Diluted aliquots were applied to each slide whereupon the slides were coverslipped and incubated overnight at 50 C in humidified chambers containing 50% formamide. The next day coverslips were removed in 2x SSC, and slides were incubated in 100 μg/ml ribonuclease A for 30 min at 37 C. Slides were briefly rinsed in 2x SSC, washed three times in 0.2x SSC (65 C), dehydrated, and exposed to Bio-MAX x-ray film (Kodak, Rochester, NY). Image analysis was performed using National Institutes of Health Image 1.62 software. In all cases, the paraventricular nucleus (PVN) region was sampled and PVN gray level values corrected for background (determined from a hybridization-negative region from the same section). Average PVN corrected gray level was calculated from two to six PVN images from each animal, with the mean value from each animal used for statistical analysis.
Hormone assays
Plasma samples were collected and stored at –20 C. Plasma CORT was assessed by RIA using a kit (MP Biomedicals, Inc., Costa Mesa, CA). Plasma ACTH concentrations were determined by RIA using a specific antiserum generously donated by Dr. William Engeland (University of Minnesota, Minneapolis, MN) at a dilution of 1:210,000 and [125I] ACTH (Amersham Biosciences, Piscataway, NJ) as labeled tracer (20). Plasma samples for each experiment were processed at the same time.
Data analysis
Hormone time-course data were analyzed by repeated-measures ANOVA. Total ACTH and CORT secretion was determined by measuring area under the curve and along with c-fos mRNA expression and the behavioral end points, they were subjected to analysis using one-way ANOVA. In all cases, Newman-Keuls post hoc testing was used to identify differences among the experimental groups.
Previous work has observed that HPA reactivity in aged rats is related to behavioral deficits, suggesting that age-related cognitive impairments are correlated with HPA dysfunction. To test this hypothesis, we performed regression analyses of our behavioral and neuroendocrine data both between and within age groups, using immobility in the forced swim test, freezing time in the fear conditioning test, and the area under the curve as a value for total ACTH and CORT responses to restraint stress and in the dexamethasone suppression test. In addition, we ran the equality of variance F test of homoscedasticity to determine whether the variance within the three age groups differed significantly on any of these tests.
Results
Effects of restraint stress
We examined the effects of restraint on ACTH and corticosterone secretion in 3-, 15-, and 30-month-old F344/BN rats. These changes are depicted in Fig. 2, A–D. There were significant effects of time and age on ACTH levels (for age: F2,124 = 17.8, P < 0.01, time: F3,124 = 69.5, P < 0.01; interaction of time x age: F6,124 = 2.48, P < 0.05) based on two-way repeated-measures ANOVA. Post hoc analysis using Newman-Keuls test revealed that aged animals secreted significantly less ACTH than either the young or middle-aged groups (at 30 min: 30 vs. 15 months, P < 0.01; 3 vs. 30 and 15 months, P < 0.05; at 60 min: 30 vs. 15 and 3 months, P < 0.01; at 120 min: 30 vs. 3 and 15 months, P < 0.05). Repeated-measures two-way ANOVA revealed a significant effect of time on plasma CORT levels (F3,124 = 91.8, P < 0.01); however, CORT responses to restraint did not differ across age groups (P > 0.05). Aged animals also exhibited lower cumulative ACTH secretion over the testing period in comparison with young and middle-aged animals after expressing the data as area under the curve (Fig. 2B; P < 0.01 based on one-way ANOVA and Neuman-Keuls). However, there were no differences in total CORT levels based on area under the curve determinations (Fig. 2D; P > 0.05 based on one-way ANOVA).
FIG. 2. A and C, Respectively, depict the effects of the initial restraint on ACTH and CORT secretion in 3-, 15-, and 30-month-old F344/BN hybrid rats. The ordinate represents concentration of hormone (picograms per milliliter for ACTH and nanograms per milliliter for CORT). The abscissa represents time after restraint. As explained in Results, there were significant effects of time and age on ACTH levels based on two-way repeated-measures ANOVA (P < 0.01). In addition, there was a significant interaction effect at P < 0.05. Post hoc analysis using Newman-Keuls test revealed that aged animals secreted significantly less ACTH than either the young or middle-aged groups (, P < 0.01 at 30 min, 30 vs. 15 months; **, P < 0.05, 3 vs. 30 and 15 months; , P < 0.01 at 60 min, 30 vs. 15 and 3 months; **, P < 0.05, at 120 min, 30 vs. 3 and 15 months). Plasma CORT responses to restraint were not significantly different among the different age groups of rats (P > 0.05). B and D, Respectively, represent total ACTH and CORT secretion in 3-, 15-, and 30-month-old F344/BN hybrid rats resulting from the effects of the initial restraint. The ACTH responses in the older rats were significantly different from both 3- and 15-month-old rats (, P < 0.01 based on one-way ANOVA and Neuman-Keuls). Plasma CORT responses to restraint were not significantly different among the different age groups of rats. In all groups, n = 12 for 3- and 15-month-old rats and n = 10 for 30-month-old rats.
Forced swim test
Figure 3 depicts the effects of the forced swim test (FST) on each age group with respect to immobility, swimming time, climbing, or diving. There were no significant differences among the three age groups with regard to any of these four parameters (P > 0.05 based on one-way ANOVA).
FIG. 3. The figure depicts data for the FST. The ordinate depicts time of immobility as a function of age. The differences among groups of rats were not significantly different from each other based on one-way ANOVA (P > 0.05). In all groups, n = 12 for 3- and 15-month-old rats and n = 10 for 30-month-old rats.
Effects of dexamethasone challenge
We next examined the effects of dexamethasone challenge on ACTH and CORT secretion in 3-, 15-, and 30-month-old F344/BN rats. These changes are depicted in Fig. 4. Repeated-measures two-way ANOVA revealed significant effects of age and time: ACTH [age (F2,93 = 6.0, P < 0.01); time (F2,93 = 3.4, P < 0.05)] and CORT [age (F2,93 = 3.99, P < 0.05); time (F2,93 = 25.3, P < 0.01)]. Post hoc analysis did not reveal any significant difference among any of the age groups at each time point for ACTH secretion (P > 0.05). However, at 8 h, the CORT response in the 30-month-old rats was significantly different from that of the 3-month-old rats (based on Neuman-Keuls; P < 0.01). In addition, there was not a significant interaction effect between time and age for ACTH or CORT (P > 0.05).
FIG. 4. A and C, Respectively, effects of dexamethasone administration on ACTH and CORT secretion in 3-, 15-, and 30-month-old F344/BN hybrid rats. Repeated-measures two-way ANOVA revealed significant effects of age and time: ACTH (age: P < 0.01; time: P < 0.05) and CORT (age: P < 0.05; time: P < 0.01). (C), CORT response at 8 h in the 30-month-old rats was significantly different from that of the 3-month-old rats (based on two-way repeated-measures ANOVA and Neuman-Keuls; , P < 0.01). B and D, Respectively, total ACTH and CORT secretion in 3-, 15-, and 30-month-old F344/BN rats after dexamethasone administration. One-way ANOVA revealed a nonsignificant trend toward lower total ACTH levels in the 30-month-old rats (P = 0.059). Furthermore, for total CORT secretion, there was no difference among age groups (P > 0.05). In all groups, n = 12 for 3- and 15-month-old rats and n = 10 for 30-month-old rats.
For total ACTH secretion, expressed as area under the curve, one-way ANOVA revealed a trend for lower ACTH levels in the oldest group of rats (P = 0.059; see Fig 4B). Furthermore, for total CORT secretion, there was no difference between age groups (P > 0.05).
Fear conditioning
Rats were subjected to a fear conditioning paradigm (Fig. 5). One-way ANOVA revealed a significant effect of age on freezing time as measured 24 h after shock (F2,30 = 4.92, P < 0.05). Surprisingly, freezing time was significantly enhanced only in the middle-aged group as determined by Newman-Keuls test (P < 0.05 for 15 vs. 3 and 15 vs. 30 months old); no differences were observed between 30- and 3-month-old animals (P > 0.05).
FIG. 5. This figure depicts freezing time in the three age groups. One-way ANOVA revealed a significant effect of age on freezing time as measured 24 h after shock (P < 0.05). Freezing time was significantly enhanced only in the middle-aged group as determined by Newman-Keuls test (**, P < 0.05 for 15 vs. 3 and 15 months vs. 30 months old); no differences were observed between 30- and 3-month-old animals (P > 0.05). In all groups, n = 12 for 3- and 15-month-old rats and n = 10 for 30-month-old rats.
Final restraint test
Figure 6 depicts the changes in ACTH (Fig. 6A) and CORT (Fig. 6B) 60 min after the last restraint in 3-, 15-, and 30-month-old rats. Consistent with the original restraint, there were significant effects of restraint on ACTH levels (F2,31 = 5.32, P < 0.01); post hoc Newman-Keuls test revealed that the 30-month-old animals secreted significantly less ACTH than both 3- and 15-month-old groups (P < 0.05).
FIG. 6. A and B, Respectively, total ACTH and CORT secretion resulting from the effects of the final 60-min restraint on in 3-, 15-, and 30-month-old F344/BN rats. There were significant effects of restraint on ACTH levels (P < 0.01); Newman-Keuls post hoc analysis revealed that the 30-month-old animals secreted significantly less ACTH than both 3- and 15-month-old groups (**, P < 0.05). In all groups, n = 12 for 3- and 15-month-old rats and n = 10 for 30-month-old rats.
Stress-induced PVN c-fos mRNA
We sought to examine central induction of c-fos mRNA to assess activation of medial parvocellular PVN neurons after stress. Semiquantitative analysis of c-fos mRNA expression after 60 min restraint revealed that c-fos levels did not differ among the three age groups (P > 0.05; Fig. 7).
FIG. 7. Semiquantitative analysis of c-fos mRNA expression in the PVN. Semiquantitative analysis of c-fos mRNA expression after 60 min restraint revealed that c-fos levels did not differ among the three age groups (P > 0.05 as per one-way ANOVA). In all groups, n = 4 for each age group.
Regression analyses: neuroendocrine function and behavior
Regression analyses were run to determine whether behavioral end points in the forced swim (immobility) and fear conditioning (freeze time) tests correlated with endocrine end points from the stress tests. There was a positive correlation between ACTH secretion in the stress test and dexamethasone suppression test (P < 0.05; r = 0.43; 3 and 15 months: n = 12; 30 months: n = 10; Fig. 8A). No significant correlations were observed between the endocrine measures and behavioral end points (P > 0.05; Fig. 8, B and C). In addition, the equality of variance F test for homoscedasticity, which tests for differences in variances among the different age groups, did not reveal enhancement of variance in the aged group on any endocrine or behavioral measure.
FIG. 8. Regression analyses relating the area under the curve of the ACTH response to restraint stress with ACTH secretion in the dexamethasone suppression test (A), freezing time in the fear conditioning test (B), and immobility time in the FST (C). Note that there was a weak but significant correlation between ACTH levels in the stress test with ACTH levels seen after dexamethasone administration (*, P < 0.05; r = 0.43); however, there were no significant correlations between stress-induced ACTH release and either of the behavioral measures (P > 0.05). In all groups, n = 12 for 3- and 15-month-old rats and n = 10 for 30-month-old rats. AUC, Area under the curve.
Discussion
The current study demonstrates that aged F344/BN F1 hybrid rats exhibit reduced ACTH responses to stress and enhanced sensitivity to dexamethasone suppression of ACTH release. Decreased stress-induced ACTH release appears to be a stable trait in aged rats, as was evident at the initial restraint test at 30 months of age and the final test at 32 months. In the case of the stress response, decreased ACTH release is not accompanied by reduced corticosterone levels, suggesting compensation at the level of the adrenal. The decrements in HPA axis responsiveness were unrelated to performance of behavioral tasks associated with depression or contextual memory.
Several systematic studies of HPA axis function have been conducted in a variety of rodent aging models. Many of these studies have been inconsistent within strains as well as across strains. For instance, at the level of the HPA axis, glucocorticoid hypersecretion (21) and hyposecretion (22, 23) have been reported in aged (F344) rats. Decrements in adrenal responsiveness have also been reported with aging in BN and Sprague Dawley rats (24, 25), whereas our findings in the F344/BN rat and those of Sonntag (F344 strain) (26) are consistent with adrenal hyperresponsiveness.
Of course, strain differences provide one likely explanation for the contradictory data on neuroendocrine responsiveness with aging. One limitation with all of these studies also was that the entire circadian rhythm was not measured in every trial but instead single or only several point measures of ACTH and corticosterone were made. The same limitation noted in these studies also applies to the data reported in this current study. Individual strains of rats are susceptible to tissue- or organ-specific pathologies that can impact HPA axis parameters. For example, the F344 strain is prone to development of pituitary tumors, renal disease, and splenomegaly (14), all of which can directly (e.g. ACTH-secreting adenomas) or indirectly (e.g. space-occupying lesions in the sella turcica/basal hypothalamus, altered fluid-electrolyte status, immune stimulation) impact HPA axis functioning. Indeed, in a previous study, our group found increased spleen size in 75% of aged F344 rats (26). For this reason, our group has more recently employed the F344/BN cross in our aging studies. This strain shows lower incidence of systematic pathologies and has a substantially longer life span than the F344 rat (up to 36 months) (14).
The current data also indicate that aged F344/BN rats exhibit alterations in peripheral limbs of the HPA axis and enhanced responsiveness to dexamethasone suppression. More specifically, the neuroendocrine findings in this report are consistent with an age-related enhancement in the sensitivity of the adrenal gland to ACTH in conjunction with reduced pituitary drive. Previous reports from our group revealed that aged F344/BN rats have reduced pituitary CRH-R1 receptor and proopiomelanocortin mRNA expression (5). Thus, the greater adrenal responsiveness noted here in F344/BN rats appears to be a potential compensatory adaptation to the loss of CRH receptivity and ACTH synthesis. The enhanced responsiveness appears not to be related to adrenal size because we compared total adrenal weight expressed relative to total body weight and found no difference in this variable among the three groups. However, the responsiveness of adrenocortical cells to trophic actions of ACTH could be enhanced in the older group of rats as a result of other factors (27).
Recent human data demonstrated that there were no age-related differences in basal HPA activity or ACTH, plasma cortisol and salivary cortisol responses to the Trier Social Stress Test (10). In addition, a reanalysis of five independent human studies by Kudielka et al. (11) determined that ACTH responses to stress were higher in younger adults, compared with older adults, but total plasma cortisol responses did not differ with regard to age. Thus, these studies are consistent with our data and suggest that age-related neuroendocrine responsiveness in the BN/F344 strain appear to model that observable in humans.
Our group has also previously shown that the older F344/BN rats are hypersensitive to chronic mild stress exposure (5) as reflected in enhanced PVN CRH immunoreactivity, increased pituitary proopiomelanocortin expression, and enhanced forebrain GAD65 up-regulation. However, in our current experiments, we have demonstrated no change across age in F344/BN rats in the FST. The reasons for this are not known.
In addition to the above neuroendocrine and behavioral paradigms, we also assessed fear conditioning. This paradigm has been used as a test of hippocampal-dependent memory (28). Other rat strains show age-related deficits in fear conditioning. Oler and Markus (29) demonstrated that older F344 rats exposed to a fear conditioning paradigm exhibit a weaker ability to retain the conditioning context relative to younger animals at a time at which consolidation should have been completed. Houston et al. (30) also demonstrated in their fear conditioning paradigm that older (27 months old) F344 rats exhibited less memory incubation, compared with younger rats. In contrast, our data indicate that aged and young rats exhibited equivalent freezing times the day after shock exposure. This suggests that deficient performance of behavioral tasks in senescence may be a strain-dependent characteristic. The responsiveness of the middle-aged group in the fear conditioning paradigm was markedly greater than either the young or old groups. The decrease in freeze time observed in aged rats could be interpreted as a contextual memory deficit, in that a response decrement is seen in old age. In contrast, enhanced freezing time in middle-aged rats may reflect enhanced stimulus salience or pain sensitivity in this group, which is effectively reversed in old age.
The emotional component of fear conditioning, however, cannot be ignored. Fear conditioning clearly also has an amygdalar basis. Knierim (31) states that it is the amygdalar-ventral hippocampal pathways and/or amygdalar-perforant path input through the entorhinal cortex, which serve as likely anatomical substrate(s) for the emotional and cognitive processes accounting for fear conditioning. In taking into account this emotional component, it is possible that amygdalar changes with aging could also account for the lower freeze time in the older rats. For instance, our group has demonstrated decreased anxiety-like behaviors in older F344 rats, which is associated with decreased amygdalar CRH expression (32).
Regardless of the interpretation, it appears that the pattern of fear conditioning observed with aging in this rat strain obeys an inverted-U-shaped function, with the most optimal mnemonic processing occurring in middle age. In addition, given the stability of the HPA axis over time in this strain and the absence of a correlation between HPA measures and freeze times, it is unlikely that the observed alterations in contextual memory are due to age-related changes in glucocorticoid secretion.
Previous work in F344 and Long-Evans strains has been able to distinguish subgroups of animals that differ in performance of spatial memory tasks (33, 34). For instance, in the Long-Evans aged rats, poor Morris water maze performance is accompanied by glucocorticoid hypersecretion (35). Despite the relatively large number of F344/BN rats used in our study, we were not able to differentiate any subgroups on tests of fear conditioning, FST behaviors, or neuroendocrine function. Indeed, the stability of these traits occurs at a time when 40–50% of the F344/BN population has died (36), verifying that the population we sampled from is indeed aged.
In summary, it appears that retention of normal glucocorticoid secretory patterns may be a characteristic of successful neuroendocrine aging and perhaps even the aging process in general. It is clear from our work that glucocorticoid hypersecretion and loss of negative feedback are not an immutable part of the aging process in this rodent strain. Indeed, the long-lived F344/BN strain appears to exhibit an active compensation within the axis to maintain normal glucocorticoid tone. It remains to be determined whether age-related HPA dysfunction and accelerated cognitive decline or even other processes are related to the dysregulation of such homeostatic processes in other aging strains (4).
Acknowledgments
We thank Mark Dolgas, Ben Packard, and Yvonne Ulrich-Lai for their assistance with this project.
References
Born J, Ditschuneit I, Schreiber M, Dodt C, Fehm HL 1995 Effects of age and gender on pituitary-adrenocortical responsiveness in humans. Eur J Endocrinol 132:705–711
Reul JM, Rothuizen J, de Kloet ER 1991 Age-related changes in the dog hypothalamic-pituitary-adrenocortical system: neuroendocrine activity and corticosteroid receptors. J Steroid Biochem Mol Biol 40:63–69
Sapolsky RM, McEwen BS, Rainbow TC 1983 Quantitative autoradiography of [3H]corticosterone receptors in rat brain. Brain Res 271:331–334
Bizon JL, Helm KA, Han JS, Chun HJ, Pucilowska J, Lund PK, Gallagher M 2001 Hypothalamic-pituitary-adrenal axis function and corticosterone receptor expression in behaviourally characterized young and aged Long-Evans rats. Eur J Biochem 14:1739–1751
Herman JP, Larson BR, Speert DB, Seasholtz AF 2001 Hypothalamo-pituitary-adrenocortical dysregulation in aging F344/Brown-Norway F1 hybrid rats. Neurobiol Aging 22:323–332
Issa AM, Rowe W, Gauthier S, Meaney MJ 1990 Hypothalamic-pituitary-adrenal activity in aged, cognitively impaired and cognitively unimpaired rats. J Neurosci 10:3247–3254
Morano MI, Vazquez DM, Akil H 1994 The role of the hippocampal mineralocorticoid and glucocorticoid receptors in the hypothalamo-pituitary-adrenal axis of the aged Fisher rat. Mol Cell Neurosci 5:400–412
Van Eekelen JA, De Kloet ER 1992 Co-localization of brain corticosteroid receptors in the rat hippocampus. Prog Histochem Cytochem 26:250–258
Murphy EK, Spencer RL, Sipe KJ, Herman JP 2002 Decrements in nuclear glucocorticoid receptor (GR) protein levels and DNA binding in aged rat hippocampus. Endocrinology 143:1362–1370
Kudielka BM, Schmidt-Reinwald AK, Hellhammer DH, Kirschbaum C 1999 Psychological and endocrine responses to psychosocial stress and dexamethasone/corticotropin-releasing hormone in healthy postmenopausal women and young controls: the impact of age and a two week estradiol treatment. Neuroendocrinology 70:422–430
Kudielka BM, Buske-Kirschbaum A, Hellhammer DH, Kirschbaum C 2004 HPA responses to laboratory psychosocial stress in healthy elderly adults, younger adults, and children: impact of age and gender. Psychoneuroendocrinology 70:83–98
Gust DA, Wilson ME, Stocker T, Conrad S, Plotsky PM, Gordon TP 2000 Activity of the hypothalamic-pituitary-adrenal axis is altered by aging and exposure to social stress in female rhesus monkeys. J Clin Endocrinol Metab 85:2556–2563
Sapolsky R, Meaney MJ 1986 Maturation of the adrenocortical stress response: neuroendocrine control mechanisms and the stress hyporesponsive period. Brain Res Rev 11:65–76
Lipman RD, Chrisp CE, Hazzard DG, Bronson RT 1996 Pathologic characterization of brown Norway, brown Norway x Fischer 344, and Fischer 344 x brown Norway rats with relation to age. J Gerontol A Biol Sci Med Sci 51:B54–B59
Sprout RL 1991 Development of animal models of aging at the National Institute on Aging. Neurobiol Aging 12:635–638
Porsolt RD, Bertin A, Jalfre M 1977 Behavioral despair in mice: a primary screening test for antidepressants. Arch Int Pharmacodyn Ther 229:327–336
Mizoguchi K, Ishige A, Aburada M, Tabira T 2003 Chronic stress attenuates glucocorticoid negative feedback: involvement of the prefrontal cortex and hippocampus. Neuroscience 119:887–897
Moita MA, Rosis S, Zhou Y, LeDoux JE, Blair HT 2004 Putting fear in its place: remapping of hippocampal place cells during fear conditioning. J Neurosci 24:7015–7023
Cullinan WE, Herman JP, Battaglia DF, Akil H, Watson SJ 1995 Pattern and time course of immediate early gene expression in rat brain following acute stress. Neuroscience 64:477–505
Jasper MS, Engeland WC 1991 Synchronous ultraradian rhythms in adrenocortical secretion detected by microdialysis in awake rats. Am J Physiol 261:R1257–R1268
Hauger RL, Thrivikraman KV, Plotsky PM 1994 Age-related alterations of hypothalamic-pituitary-adrenal axis function in male Fischer 344 rats. Endocrinology 134:1528–1536
Cizza G, Calogero AE, Brady LS, Bagdy G, Bergamini E, Blackman MR, Chrousos GP, Gold PW 1994 Male Fischer 344/N rats show a progressive central impairment of the hypothalamic-pituitary-adrenal axis with advancing age. Endocrinology 134:1611–1620
Kasckow JW, Regmi A, Mulchahey JJ, Plotsky PM, Hauger RL 1999 Changes in brain corticotropin-releasing factor messenger RNA expression in aged Fischer 344 rats. Brain Res 822:228–230
Carnes M, Goodman BM, Lent SJ, Vo H, Jaeckels R 1994 Coincident plasma ACTH and corticosterone time series: comparisons between young and old rats. Exp Gerontol 29:625–643
Van Eekelen JA, Oitzl MS, de Kloet ER 1995 Adrenocortical hyporesponsiveness and glucocorticoid feedback resistance in old male brown Norway rats. J Gerontol A Biol Sci Med Sci 50:B83–B89
Herman JP, Chen KC, Booze R, Landfield PW 1998 Up-regulation of 1D Ca2+ channel subunit mRNA expression in the hippocampus of aged F344 rats. Neurobiol Aging 19:581–587
Sonntag WE, Goliszek AG, Brodish A, Eldridge JC 1987 Diminished diurnal secretion of adrenocorticotropin (ACTH), but not corticosterone, in old male rats: possible relation to increased adrenal sensitivity to ACTH in vivo. Endocrinology 120:2308–2315
Rudy JW, Huff NC, Matus-Amat P 2004 Understanding contextual fear conditioning: insights from a two process model. Neurosci Biobehav Rev 28:675–685
Oler JA, Markus EJ 1998 Age-related deficits on the radial maze and in fear conditioning: hippocampal processing and consolidation. Hippocampus 8:402–415
Houston FP, Stevenson GD, McNaughton BL, Barnes CA 1999 Effects of age on the generalization and incubation of memory in the F344 rat. Learn Mem 6:111–119
Knierim JJ 2003 Hippocampus and memory. Can we have our place and fear it too? Neuron 37:372–374
Pisarska M, Mulchahey JJ, Welge J, Geracioti TD, Kasckow JW 2000 Age-related alterations in emotional behaviors and amygdalar corticotropin-releasing factor (CRF) and CRF-binding protein expression in aged Fischer 344 rats. Brain Res 877:959–966
Brightwell JJ, Gallagher M, Colombo PJ 2004 Hippocampal CREBI but not CREB2 is decreased in aged rats with spatial memory impairments. Neurobiol Learn Mem. 81:19–26
Shukitt-Hale B, Mouzakis G, Joseph JA 1998 Psychomotor and spatial memory performance in aging male Fischer 344 rats. Exp Gerontol 33:615–624
Issa BG, Lewis BM, Ham J, Peters JR, Scanlon MF 1998 Glutamate pathways mediate somatostatin responses to glucose in normal and diabetic rat hypothalamus. J Neuroendocrinol 10:377–381
Turturro A, Witt WW, Lewis S, Hass BS, Lipman RD, Hart RW 1999 Growth curves and survival characteristics of the animals used in the Biomarkers of Aging Program. J Steroid Biochem Mol Biol 54:B492–B501(John W. Kasckow, Tracy M.)