- Email: [email protected]
Corticosterone receptors decline in a site-specific manner in the aged rat brain
Brain Research, 289 (1983) 235-240 Elsevier
235
Corticosterone Receptors Decline in a Site-Specific Manner in the Aged Rat Brain ROBERT M. SAPOLSKY, LEWIS C. KREY* and BRUCE S. McEWEN The Rockefeller University, 1230 York Avenue, New York, NY 10021 (U.S.A.)
(Accepted April 19th, 1983) Key words: aging - - hippocampus - - amygdala - - glucocorticoids - - receptors - - negative feedback
Putative glucocorticoid receptors were measured in the brain and pituitary glands of young and aged rats in vivo and in vitro. Adrenalectomized rats were injected with a set dose of radiolabeled corticosterone plus unlabeled eortieosterone; 2 h later [3H]cortieosterone levels were measured in purified nuclear pellets from pituitary and 5 brain regions. Substantial decreases were seen in aged subjects in the maximal number of corticosterone binding sites in hippocampus and amygdala; all other regions showed no age-related changes. In contrast, there were no declines in the nuclear uptake of [3H]dexamethasone (DEX) in the aged rat brains. Since DEX interacts selectively with non-neuronal receptors in vivo, the deficit in glucocorticoid binding is selective for neurons. Subsequent studies assessing glucocorticoid receptor levels in cytosol preparations in vitro revealed significant declines in hippocampus and amygdala quantitatively comparable to the decline observed in nuclear binding in these loci. This suggests that the primary deficit leading to nuclear depletion may be the reduction of cytosolic receptor number, rather than other possible factors such as the reduction in receptor affinity or translocation of steroid-receptor complex. This decline may play a role in a number of limbic functions which are influenced by glucocorticoids and show deficits with age.
INTRODUCTION
METHODS
Aging has been characterized as a loss of the capacity to adapt to environmental demands20. As such, interest has focused on the stress-response of aged organisms, in particular the secretion of glucocorticoids by the adrenocortical axis. In the rat, several functional changes appear in this system with aging: elevated basal corticosterone titers, delayed adaptation to a mild stressor and delayed recovery from stress 19. Roth is has argued that age-related deficits in endocrine function are often mediated by a decreased number of receptors for hormones in target tissues. Since the glucocorticoid hormones have been hypothesized to influence activity of the adrenocortical axis via interactions with high affinity binding sites in pituitary cells and neurons in the hippocampus, amygdala, hypothalamus and other distinct brain regions12, we have measured cytosol receptor levels and cell nuclear uptake of corticosterone in these regions in young and aged rats.
Male Fisher 344 rats, 3 and 24 months of age, were provided by the Charles River Breeding Laboratory through a grant from the National Institute on Aging. Subjects arrived from a pathogen-free colony and were housed one/cage for at least two weeks before use. Subjects were maintained in a 12:12 h light:dark schedule and were given food and water ad libitum. Mean lifespan is over 27 months in our colony. Occasionally we observed adrenal tumors or tumors of the brain or pituitary in the aged rats which were then excluded from study. Rats were adrenalectomized under ether anesthesia. Aged subjects were maintained on subcutaneous injections of corticosterone and intraperitoneal injections of glucose-saline solution for one day after surgery to minimize post-operative mortality. Subjects were used in uptake experiments 5-7 days after withdrawal from corticosterone, an interval that allows the maximal post-adrenalectomy increase in binding sites~L
* Irma T. Hirschl Career Scientist. 0006-8993/83/$03.00 (~ 1983 Elsevier Science Publishers B.V.
236 Rats were injected with radiolabeled corticosterone ([1,2,6,7-3H]corticosterone, Amersham; spec. act. 82.1 Ci/mmol) or dexamethasone ([1,2,4-3H]dexamethasone, Amersham; spec. act. 46 Ci/mmol) along with varying amounts of unlabeled competitive corticosterone or dexamethasone, respectively. Unlabeled corticosterone (Sigma, St. Louis, MO) was injected s.c. in quantities of 0, 1, 2, 2.5, 3, 5, 8.5, 10, 15, 20, 40, 50 and 60/~g along with 0.15 ml ethanol in 0.3 ml saline. Unlabeled dexamethasone (Sigma, St. Louis, MO) was injected in quantities of 10 or 20 pg along with ethanol/saline. Twenty minutes after injection of unlabeled material, 100 pCi of radiolabeled corticosterone or dexamethasone, evaporated under N 2 and resuspended in 0.15 ml ethanol and 0.3 ml saline, was injected s.c. Subjects were killed 2 h later by decapitation. Brains were dissected as previously described8 and the following regions quickly placed on a chilled plate: midbrain, frontal cortex, septum, hypothalamus (including POA), amygdala and hippocampus (including dorsal, ventral and subiculum). In addition, the pituitary and liver were removed. Purified nuclear pellets were prepared by centrifugation in 2 M sucrose according to the method of McEwen and Zigmond9. The nuclear pellets were extracted overnight in 1 ml ethanol, centrifuged at 16,000 gay for 10 min and the supernatant saved. The pellets were then resuspended in 1 ml ethanol and respun. The second supernatant was added to the first, the pool then being dried down and counted in Linquiscint (National Diagnostics, Somerville, N J) at an efficiency of 35%. Following ethanol extraction, the pellets were air dried for 4 days and DNA measured 4. Results are expressed as fmol glucocorticoid/tissue and fmol glucocorticoid/mg DNA/tissue. A theoretically maximal number of binding sites (Bmax) was calculated using a double reciprocal plotl0 Lines derived from double reciprocal analysis of young and old data were compared with analysis of covariance. Levels of significance of differences in slope and intercepts of these lines were determined. Curves for bound versus dose were drawn from the double reciproca ! lines. All other statistical comparisons were made with Student's t-test for independent variables. For cytosol receptor determinations, pairs of young and aged rats were adrenalectomized and decapitated 12 h later. This allows ample time for clear-
ance of endogenous hormone 10. The hippocampus and amygdala were dissected and homogenized in 5 mM Tris buffer (pH 7.4) containing 1 mM EDTA, 10 mM sodium molybdate, 10% glycerol and 1 mM dithiothreitoi. The homogenate was centrifuged at 1 °C for 30 min at 105,000 ga~.- Aliquots of cytosol of 250 pl were added to previously evaporated [3H]dexamethasone solutions (concentration range 5-40 nM, as determined by direct counting of aliquots). Dexamethasone binds well in vitro to glucocorticoid receptors in neuronal and non-neuronal (e.g. glial) cellsl4. Non-specific binding was determined in parallel incubations which also contained a 500-fold excess of corticosterone. Macromolecular bound steroid was separated form free steroid using LH-20 columns that contained the homogenate buffer with molybdate and dithiothreitol. Duplicate 250pl samples of the incubated cytosol were applied to the columns at 4 °C and washed into the columns with 150 pl buffer. After 30 min the columns were eluted with 800/A buffer and the eluate collected and counted in Liquiscint, as described above. Cytosol protein concentrations were determined 7 and results standardized as fmol [3H]dexamethasone bound/mg protein. Bma x and Kd were derived from double reciprocal plots and statistical comparisons between age groups were made with paired t-tests. RESULTS
In order to assess cytosolic corticosterone receptor TABLE I Estimated K a and maximal binding (B,nax) for [3H]dexamethasone to cytosolic receptors in vitro in young and aged rats
Bmax (fmol/mg protein) and dissociation constant (mol/t x 109) derived from double reciprocal plots. Mean + S.E. Correlation coefficients for linear regression analyses of reciprocal plots, n = 5 pairs of subjects of each age group for hippocampus, 3 pairs for amygdala. Old subjects had significantly lower Bmax for both hippocampus (t = 2.87, df = 4, P < 0.05, two-tailed paired t-test) and amygdala (t = 6.98, df = 2, P < 0.025). Tissue
B max
Kd
r
Hippocampus Young Old
160 + 31 105 + 16
2.7 + 0.8 2.3 + 0.5
0.96 + 0:02 0.91 + 0.03
Amygdala Young Old
106 + 11 86 + 10
1.6 + 0.3 1.0 + 0.2
0.97 _+ 0.01 0.93 -+ 0.02
237 80
80-
A~IYGDALA
HI PPOCAMPUS
~
X
Kd
Bmax
Kd
x
2.7
133
4.0
211 103
.6
o
.8
75
X
C
~40
~.~ ~
~-
Bmax
40
O
L
I0
x ~
0
I O0
200
50
I00
150
3H-DEXANETHASONE BOUND
Fig. 1. Representative Scatchard plot of in vitro [3H]dexamethasone binding in cytosol of hippocampus and amygdala in young and aged rats. K d, dissociation constant (nM); Bmax, maximal binding capacity (fmol/mg cytosol protein), x, young rats; C), aged rats.
number and affinity, we carried out an in vitro dose response analysis, incubating the synthetic glucocorticoid dexamethasone with cytosol from brain tissues of young and aged rats. Significant decreases in receptor number were observed in both hippocampus and amygdala of aged rats (Table I). K d , however, was not affected in either tissue. Fig. 1 shows Scatchard plots for a representative experiment. Maximal binding capacity in cell nuclei was then measured with an in vivo dose response analysis with radiolabeled corticosterone. The age-related cytosolic receptor declines in hippocampus and amygdala were accompanied by comparable declines in binding capacity in cell nuclei from these tissues (Table II). The aged hippocampus showed a significant (P< 0.001) depletion of 48% (Fig. 2) while the amygdala showed a significant (P < 0.02) depletion of 40%. A number of other tissues were examined in this experiment. The pituitary showed a 32% reduction which approached, but did not reach, the P = .05 level of significance, while no age-related changes were observed in midbrain, cortex and hypothalamus. Values from young animals were comparable to previously published levels from this laboratoryt2. Since glia also contain putative glucocorticoid receptors 14. we next investigated whether the depletion
of glucocorticoid receptor capacity in cytosol and cell nuclei was predominantly neuronal or glial in origin. We measured in vivo uptake, by cell nuclei, of radiolabeled dexamethasone. When administered in vivo, dexamethasone preferentially interacts with glial glucocorticoid receptors~4,16. Choosing two specific activities of labeled corticosterone at which the decrease in uptake in hippocampus and amygdala were TABLE II Uptake o f radiolabeled corticosterone by cell nuclei o f pituitary and brain regions in y o u n g and aged male rats
Bmax is maximal binding (pg/mg DNA) from y intercepts of reciprocal plots, r is correlation coefficients for linear regression analysis of reciprocal plots. Hypothalamus includes pre-optic area; hippocampus includes dorsal, plus ventral portion and subiculum. T~sue
Hippocampus* Amygdala* Pituitary Hypothalamus Cortex Midbrain
3 Month (n = 36)
24 Month (n = 18-20)
B max
r
n ,,~
r
770 262 106 61 130 67
0.95 0.89 0.89 0.75 0.74 0.62
410 159 72 80 117 62
0.82 0.96 0.56 0.65 0.81 0.69
* Indicates a statistically significant age difference in Bmax, determined with analysis of covariance (see Methods).
238 .006
800
X
~400
X
"1:3 c-
=.004
~
I
o
20
4c
;
~g. corticosterone administered
~ /
/
-.6
/
/
I
/
I
-,4
J
_
/
~
× / ,, X ..qb."x
, .-
,
/
•
.002
.
f
~/.'/
/
m
0
/
/
./
X
/
•
/
Ja
t
I
-,2
I
I
.
0
I /
I
.2
I
|
.4
Dose
Fig. 2. Dose-response analysis of binding capacity of nuclei of hippocampus of young (O---Q) and aged ( x - - x ) rats, presented as bound versus dose (inset) and 1/bound versus 1/dose. Each point represents two pooled subjects. The dose of administered unlabeled corticosterone was injected as competition with 100/tCi of radiolabeled corticosterone. Note that the figure in the inset is extended only to the highest dose where counts were detectable in both age groups.
quite clear (100/,tCi radiolabeled corticosterone + 10 /~g or 20 #g unlabeled corticosterone), we administered the same dosages of labeled plus unlabeled dexamethasone to 3- and 24-month-old subjects. Whereas the depletion of binding sites in hippocampus and amygdala was noted with corticosterone at these two different specific activities, no such depletion, and in fact, a non-significant increase, was seen with dexamethasone (Fig. 3). This trend towards an age-related increase was observed in all neural areas examined (data not shown). The decreased hippocampal binding of dexamethasone relative to corticosterone and the more general distribution of dexamethasone binding throughout the brain relative to the marked site-specific differences in binding of corticosterone observed in this study are in good agreement with previously published studies from our laboratoryS. DISCUSSION
In the present study, we have found age-related
HIPPOCAMPUS <
800-
AMYGDALA I
3 month
-
400
g 200
-
Onmol:
29
58
Cort
28
53
De×
29 58 Cort
28 53 De
K
Fig. 3. Binding of two different specific activities of radiolabeled corticosterone and dexamethasone by cell nuclei of hippocampus and amygdala in young and aged rats. Corticosterone was administered as 100/tCi-radiolabeled material plus t0 or 20/~g (29 or 58 nmol) of unlabeled. Dexamethasone was administered as 100/~Ci radiolabeted material plus 10 or 20#g (28 or 53 nmol) unlabeled. Two determinations were made for each dose and age group for dexamethasone, with each determination from a pair of pooled subjects; mean - S.E. Corticosterone determinations were made from a single pooled pair of subjects for each dose and age group as part of the dose response study in Table II. No significant age differences were observed in dexamethasone binding (independent t-test ).
239 decreases in the nuclear uptake of corticosterone in the brain, an endpoint which is relevent to understanding glucocorticoid-genomic interactions. The key finding is the regional specificity of the diminution of cell nuclear corticosterone receptor capacity in the aging brain. On the average, receptor capacity per cell (i.e. per mg DNA) declines with age in both the hippocampus and amygdala. The pituitary shows a non-significant trend towards a decline; the tremendous variance in pituitary determinations in the aged probably reflects microadenomas typical in that age-group6. The rest of the brain shows no change in binding with age. Our further investigation of cytosolic binding of dexamethasone demonstrated significant declines of hippocampal and amygdaloid binding comparable to declines in nuclear binding in those loci. Dexamethasone, when used in an in vitro cytosol assay, labels both glial and neuronal receptors, as does corticosterone when used in an in vivo cell nuclear uptake assay. The comparable declines revealed by both methods suggest that the reduced nuclear binding is due to reductions in cytosolic receptor number, rather than to decreases in receptor affinity or nuclear translocation of the steroid-receptor complex per se. The decline in hippocampal cytosolic binding with age is in agreement with previous reports 1. We next considered whether the declines were predominantly in the neuronal or non-neuronal (including glial) receptor populations. Selective labeling of non-neuronal nuclear receptors is evident by autoradiographic evidence after in vivo [3H]dexamethasone administration13,17, 21. Further evidence for dexamethasone preferentially interacting with glial and not neuronal receptors in vivo comes from its capacity to induce a glial-specific enzyme and its failure to induce a neuron-specific proteinl4,16. We have examined nuclear binding of dexamethasone throughout the brain and pituitary and find that the observed site-specific decreases in binding of corticosterone with age are not evident when dexamethasone is the ligand. A tendency towards increased binding of dexamethasone in aged subjects is in fact seen, a trend repeated in the pituitary and all brain regions assayed. In light of the present data, we conclude that the receptor decline revealed by in vivo administra-
tion of corticosterone but not by dexamethasone indicates an age-related deficit specific to neuronal populations. Glial hypertrophy appears to be a consistent correlate of aging in the non-pathologic brain, accounting for at least some of the increased DNA levels throughout the aged rat brain that we and others have observed (ref. 3 and unpublished observations). Such an increase might account for the trend towards increased in vivo dexamethasone binding seen throughout the aged brain in this study. In any case, since corticosterone interacts with both neuronal and glial receptors14,16, the inferred decline in hippocampal and amygdaloid neuronal receptors may even be more pronounced than reported, as the deficits may be somewhat masked by the binding due to glial receptors. The pituitary, hippocampus and amygdala are all thought to play roles in negative-feedback control of glucocorticoid secretion 11, and in the rat aging brings an impairment of such negative-feedback regulation19. Further, the adrenocortical axis exerts influences on learning and memory via the hippocampus 2, and the aged rat shows impairments of cognitive function 6. The possible relationships between the observed receptor declines and the age-related changes in neuroendocrine function are being investigated. Note. We have recently completed a detailed anatomic analysis of this corticosterone binding decline, using a method (Sapolsky, McEwen and Rainbow, in press, Brain Research) for quantitative steroid autoradiography. We find no loss of binding in the subiculum and dentate regions in the aged hippocampus, extensive loss in the CA3 cell field and lesser declines at other hippocampal sites, (Sapolsky, Rainbow, McEwen, in preparation). ACKNOWLEDGEMENTS We wish to thank Dr. B. Parsons for technical assistance during this study and Mrs. Oksana Wengerchuk and Dr. Amy Mandelker for manuscript assistance. Research support was provided by the National Institute on Aging via a predoctoral grant to R.M.S.
240 REFERENCES 1 Angelucci, L., Valeri, P., Grossi, E., Veldhuis, H., Bohus. B. and De Kloet, E., Involvement of hippocampal corticosterone receptors in behavioral phenomena. In F. Brambilla, G. Racagni and D. de Wied (Eds.), Progress in Psychoneuroendocrinology, Elsevier, Biomedical, Amsterdam, 1980, pp. 177-185. 2 Bohus, B. and De Kloet, E., Behavioral effects of neuropeptides related to LPH and ACTH (endorphins, enkephalin, ACTH fragments) and corticosteroids. In M. Jones, B. Gilham, M. Dallman and S. Chattopadhyay (Eds.), Interaction within the Brain-Pituitary-Adrenocortical System, Academic Press, New York, 1979, pp. 7-16. 3 Brizzee, K., Aging changes in relation to diseases of the nervous system. In M. Ordy and K. Brizzee (Eds.), Neurobiology of Aging, Plenum, New York, 1975, pp. 545-573. 4 Burton, K., A study of the conditions and mechanisms of the diphenylamine reaction for the colorimetric estimation of DNA, Biochem. J., 62 (1956) 315-323. 5 De Kloet, R., Wallach, G. and McEwen, B., Differences in corticosterone and dexamethasone binding to rat brain and pituitary, Endocrinology, 96 (1975) 598-609. 6 Finch, C. and Hayflick, L., Handbook of the Biology of Aging, Van Nostrand, 1977, 770 pp. 7 Lowry, O., Rosebrough, N., Farr, A. and Randall, R.. Protein measurement with the Folin phenol reagent. J. biol. Chem., 193 (1953) 265-275. 8 McEwen, B. and Pfaff, D., Factors infuencing sex hormone uptake by rat brain regions. I. Effects of neonatal treatment hypophysectomy, and competing steroid on estradiol uptake, Brain Research, 21 (1970) 1-16. 9 McEwen, B. and Zigmond, R., Isolation of brain cell nuclei. In N. Marks and R. Rodnight (Eds.), Research Methods in Neurochemistry, Vol. 1, Plenum Press, New York. 1972, pp. 140-161. 10 McEwen, B. Wallach, G. and Magnus, C., Corticosterone binding to hippocampus: immediate and delayed influences
11 12
13
14
15
16
17
18 19
20
21
of the absence of adrenal secretion, Brain Research, 7~J (1974) 321-324. McEwen, B., Adrenal steroid feedback on neuroendocrinc tissues, Ann. N. • Acad. Sci., 297 (1977) 568--579. McEwen, B., Stephenson. B. and Krey, L., Radioimmunoassay of brain tissue and cell nuclear corticosterone, J. Neurosci. Methods, 3 (1980) 57-65. McEwen, B., Glucocorticoids and hippocampus: receptors in seach of a function. In D. Ganten and D. Pfaff (Eds.). Current Topics in Neuroendocrinology, Vol. 2, Springer, Berlin, 1982, pp. 23-43. Meyer. J. and McEwen, B., Evidence for glucocorticoid target cells in the rat optic nerve; physicochemical characterization of cytosolic binding sites, J. Neurochem., 39 (1982) 435-442. Meyer, J., Micco, D., Stephenson, B., Krey, L. and McEwen, B., Subcutaneous implantation method for chronic glucocorticoid replacement therapy, Physiol. Behay., 22 (1970) 867-870. Nestler, E., Rainbow, T., MeEwen, B. and Greengard, P., Effect of steroid hormones on the level of protein in rat brain. In K. Fuxe (Ed.), Steroid Hormone Regulation of the Brain, Pergamon Press, Oxford, 1981, pp. 205-216. Rees, H., Stumpf, W. and Sar, M., Autoradiographic studies with -~H-dexamethasone in the rat brain and pituitary. In W. Stumpf and L. Grant (Eds.), Anatomical Neuroendocrinology, Karger Basel, 1966, pp. 262-269. Roth, G., Hormone action during aging; alteration and mechanisms, Mech. Aging. Develop., 9 (1979) 49%514. Sapolsky, R., Krey, L. and McEwen, B., The adrenocortical stress-response in the aged male rat; impairment of recovery from stress, Exp. Gerontol., 16 (1983) 55-64. Selye, H. and Tuchweber, B., Stress in relation to aging and disease. In A. Everitt and J. Burgess (Eds.). Hypothalamus, Pituitao,. and Aging, Charles Thomas, Springfield IL, 1976, pp. 557-573. Warembourg, M., Radioautographic study of the rat brain and pituitary after injection of 3H dexamethasone, Cell Tiss. Res., 16t (1975) 183-191.
235
Corticosterone Receptors Decline in a Site-Specific Manner in the Aged Rat Brain ROBERT M. SAPOLSKY, LEWIS C. KREY* and BRUCE S. McEWEN The Rockefeller University, 1230 York Avenue, New York, NY 10021 (U.S.A.)
(Accepted April 19th, 1983) Key words: aging - - hippocampus - - amygdala - - glucocorticoids - - receptors - - negative feedback
Putative glucocorticoid receptors were measured in the brain and pituitary glands of young and aged rats in vivo and in vitro. Adrenalectomized rats were injected with a set dose of radiolabeled corticosterone plus unlabeled eortieosterone; 2 h later [3H]cortieosterone levels were measured in purified nuclear pellets from pituitary and 5 brain regions. Substantial decreases were seen in aged subjects in the maximal number of corticosterone binding sites in hippocampus and amygdala; all other regions showed no age-related changes. In contrast, there were no declines in the nuclear uptake of [3H]dexamethasone (DEX) in the aged rat brains. Since DEX interacts selectively with non-neuronal receptors in vivo, the deficit in glucocorticoid binding is selective for neurons. Subsequent studies assessing glucocorticoid receptor levels in cytosol preparations in vitro revealed significant declines in hippocampus and amygdala quantitatively comparable to the decline observed in nuclear binding in these loci. This suggests that the primary deficit leading to nuclear depletion may be the reduction of cytosolic receptor number, rather than other possible factors such as the reduction in receptor affinity or translocation of steroid-receptor complex. This decline may play a role in a number of limbic functions which are influenced by glucocorticoids and show deficits with age.
INTRODUCTION
METHODS
Aging has been characterized as a loss of the capacity to adapt to environmental demands20. As such, interest has focused on the stress-response of aged organisms, in particular the secretion of glucocorticoids by the adrenocortical axis. In the rat, several functional changes appear in this system with aging: elevated basal corticosterone titers, delayed adaptation to a mild stressor and delayed recovery from stress 19. Roth is has argued that age-related deficits in endocrine function are often mediated by a decreased number of receptors for hormones in target tissues. Since the glucocorticoid hormones have been hypothesized to influence activity of the adrenocortical axis via interactions with high affinity binding sites in pituitary cells and neurons in the hippocampus, amygdala, hypothalamus and other distinct brain regions12, we have measured cytosol receptor levels and cell nuclear uptake of corticosterone in these regions in young and aged rats.
Male Fisher 344 rats, 3 and 24 months of age, were provided by the Charles River Breeding Laboratory through a grant from the National Institute on Aging. Subjects arrived from a pathogen-free colony and were housed one/cage for at least two weeks before use. Subjects were maintained in a 12:12 h light:dark schedule and were given food and water ad libitum. Mean lifespan is over 27 months in our colony. Occasionally we observed adrenal tumors or tumors of the brain or pituitary in the aged rats which were then excluded from study. Rats were adrenalectomized under ether anesthesia. Aged subjects were maintained on subcutaneous injections of corticosterone and intraperitoneal injections of glucose-saline solution for one day after surgery to minimize post-operative mortality. Subjects were used in uptake experiments 5-7 days after withdrawal from corticosterone, an interval that allows the maximal post-adrenalectomy increase in binding sites~L
* Irma T. Hirschl Career Scientist. 0006-8993/83/$03.00 (~ 1983 Elsevier Science Publishers B.V.
236 Rats were injected with radiolabeled corticosterone ([1,2,6,7-3H]corticosterone, Amersham; spec. act. 82.1 Ci/mmol) or dexamethasone ([1,2,4-3H]dexamethasone, Amersham; spec. act. 46 Ci/mmol) along with varying amounts of unlabeled competitive corticosterone or dexamethasone, respectively. Unlabeled corticosterone (Sigma, St. Louis, MO) was injected s.c. in quantities of 0, 1, 2, 2.5, 3, 5, 8.5, 10, 15, 20, 40, 50 and 60/~g along with 0.15 ml ethanol in 0.3 ml saline. Unlabeled dexamethasone (Sigma, St. Louis, MO) was injected in quantities of 10 or 20 pg along with ethanol/saline. Twenty minutes after injection of unlabeled material, 100 pCi of radiolabeled corticosterone or dexamethasone, evaporated under N 2 and resuspended in 0.15 ml ethanol and 0.3 ml saline, was injected s.c. Subjects were killed 2 h later by decapitation. Brains were dissected as previously described8 and the following regions quickly placed on a chilled plate: midbrain, frontal cortex, septum, hypothalamus (including POA), amygdala and hippocampus (including dorsal, ventral and subiculum). In addition, the pituitary and liver were removed. Purified nuclear pellets were prepared by centrifugation in 2 M sucrose according to the method of McEwen and Zigmond9. The nuclear pellets were extracted overnight in 1 ml ethanol, centrifuged at 16,000 gay for 10 min and the supernatant saved. The pellets were then resuspended in 1 ml ethanol and respun. The second supernatant was added to the first, the pool then being dried down and counted in Linquiscint (National Diagnostics, Somerville, N J) at an efficiency of 35%. Following ethanol extraction, the pellets were air dried for 4 days and DNA measured 4. Results are expressed as fmol glucocorticoid/tissue and fmol glucocorticoid/mg DNA/tissue. A theoretically maximal number of binding sites (Bmax) was calculated using a double reciprocal plotl0 Lines derived from double reciprocal analysis of young and old data were compared with analysis of covariance. Levels of significance of differences in slope and intercepts of these lines were determined. Curves for bound versus dose were drawn from the double reciproca ! lines. All other statistical comparisons were made with Student's t-test for independent variables. For cytosol receptor determinations, pairs of young and aged rats were adrenalectomized and decapitated 12 h later. This allows ample time for clear-
ance of endogenous hormone 10. The hippocampus and amygdala were dissected and homogenized in 5 mM Tris buffer (pH 7.4) containing 1 mM EDTA, 10 mM sodium molybdate, 10% glycerol and 1 mM dithiothreitoi. The homogenate was centrifuged at 1 °C for 30 min at 105,000 ga~.- Aliquots of cytosol of 250 pl were added to previously evaporated [3H]dexamethasone solutions (concentration range 5-40 nM, as determined by direct counting of aliquots). Dexamethasone binds well in vitro to glucocorticoid receptors in neuronal and non-neuronal (e.g. glial) cellsl4. Non-specific binding was determined in parallel incubations which also contained a 500-fold excess of corticosterone. Macromolecular bound steroid was separated form free steroid using LH-20 columns that contained the homogenate buffer with molybdate and dithiothreitol. Duplicate 250pl samples of the incubated cytosol were applied to the columns at 4 °C and washed into the columns with 150 pl buffer. After 30 min the columns were eluted with 800/A buffer and the eluate collected and counted in Liquiscint, as described above. Cytosol protein concentrations were determined 7 and results standardized as fmol [3H]dexamethasone bound/mg protein. Bma x and Kd were derived from double reciprocal plots and statistical comparisons between age groups were made with paired t-tests. RESULTS
In order to assess cytosolic corticosterone receptor TABLE I Estimated K a and maximal binding (B,nax) for [3H]dexamethasone to cytosolic receptors in vitro in young and aged rats
Bmax (fmol/mg protein) and dissociation constant (mol/t x 109) derived from double reciprocal plots. Mean + S.E. Correlation coefficients for linear regression analyses of reciprocal plots, n = 5 pairs of subjects of each age group for hippocampus, 3 pairs for amygdala. Old subjects had significantly lower Bmax for both hippocampus (t = 2.87, df = 4, P < 0.05, two-tailed paired t-test) and amygdala (t = 6.98, df = 2, P < 0.025). Tissue
B max
Kd
r
Hippocampus Young Old
160 + 31 105 + 16
2.7 + 0.8 2.3 + 0.5
0.96 + 0:02 0.91 + 0.03
Amygdala Young Old
106 + 11 86 + 10
1.6 + 0.3 1.0 + 0.2
0.97 _+ 0.01 0.93 -+ 0.02
237 80
80-
A~IYGDALA
HI PPOCAMPUS
~
X
Kd
Bmax
Kd
x
2.7
133
4.0
211 103
.6
o
.8
75
X
C
~40
~.~ ~
~-
Bmax
40
O
L
I0
x ~
0
I O0
200
50
I00
150
3H-DEXANETHASONE BOUND
Fig. 1. Representative Scatchard plot of in vitro [3H]dexamethasone binding in cytosol of hippocampus and amygdala in young and aged rats. K d, dissociation constant (nM); Bmax, maximal binding capacity (fmol/mg cytosol protein), x, young rats; C), aged rats.
number and affinity, we carried out an in vitro dose response analysis, incubating the synthetic glucocorticoid dexamethasone with cytosol from brain tissues of young and aged rats. Significant decreases in receptor number were observed in both hippocampus and amygdala of aged rats (Table I). K d , however, was not affected in either tissue. Fig. 1 shows Scatchard plots for a representative experiment. Maximal binding capacity in cell nuclei was then measured with an in vivo dose response analysis with radiolabeled corticosterone. The age-related cytosolic receptor declines in hippocampus and amygdala were accompanied by comparable declines in binding capacity in cell nuclei from these tissues (Table II). The aged hippocampus showed a significant (P< 0.001) depletion of 48% (Fig. 2) while the amygdala showed a significant (P < 0.02) depletion of 40%. A number of other tissues were examined in this experiment. The pituitary showed a 32% reduction which approached, but did not reach, the P = .05 level of significance, while no age-related changes were observed in midbrain, cortex and hypothalamus. Values from young animals were comparable to previously published levels from this laboratoryt2. Since glia also contain putative glucocorticoid receptors 14. we next investigated whether the depletion
of glucocorticoid receptor capacity in cytosol and cell nuclei was predominantly neuronal or glial in origin. We measured in vivo uptake, by cell nuclei, of radiolabeled dexamethasone. When administered in vivo, dexamethasone preferentially interacts with glial glucocorticoid receptors~4,16. Choosing two specific activities of labeled corticosterone at which the decrease in uptake in hippocampus and amygdala were TABLE II Uptake o f radiolabeled corticosterone by cell nuclei o f pituitary and brain regions in y o u n g and aged male rats
Bmax is maximal binding (pg/mg DNA) from y intercepts of reciprocal plots, r is correlation coefficients for linear regression analysis of reciprocal plots. Hypothalamus includes pre-optic area; hippocampus includes dorsal, plus ventral portion and subiculum. T~sue
Hippocampus* Amygdala* Pituitary Hypothalamus Cortex Midbrain
3 Month (n = 36)
24 Month (n = 18-20)
B max
r
n ,,~
r
770 262 106 61 130 67
0.95 0.89 0.89 0.75 0.74 0.62
410 159 72 80 117 62
0.82 0.96 0.56 0.65 0.81 0.69
* Indicates a statistically significant age difference in Bmax, determined with analysis of covariance (see Methods).
238 .006
800
X
~400
X
"1:3 c-
=.004
~
I
o
20
4c
;
~g. corticosterone administered
~ /
/
-.6
/
/
I
/
I
-,4
J
_
/
~
× / ,, X ..qb."x
, .-
,
/
•
.002
.
f
~/.'/
/
m
0
/
/
./
X
/
•
/
Ja
t
I
-,2
I
I
.
0
I /
I
.2
I
|
.4
Dose
Fig. 2. Dose-response analysis of binding capacity of nuclei of hippocampus of young (O---Q) and aged ( x - - x ) rats, presented as bound versus dose (inset) and 1/bound versus 1/dose. Each point represents two pooled subjects. The dose of administered unlabeled corticosterone was injected as competition with 100/tCi of radiolabeled corticosterone. Note that the figure in the inset is extended only to the highest dose where counts were detectable in both age groups.
quite clear (100/,tCi radiolabeled corticosterone + 10 /~g or 20 #g unlabeled corticosterone), we administered the same dosages of labeled plus unlabeled dexamethasone to 3- and 24-month-old subjects. Whereas the depletion of binding sites in hippocampus and amygdala was noted with corticosterone at these two different specific activities, no such depletion, and in fact, a non-significant increase, was seen with dexamethasone (Fig. 3). This trend towards an age-related increase was observed in all neural areas examined (data not shown). The decreased hippocampal binding of dexamethasone relative to corticosterone and the more general distribution of dexamethasone binding throughout the brain relative to the marked site-specific differences in binding of corticosterone observed in this study are in good agreement with previously published studies from our laboratoryS. DISCUSSION
In the present study, we have found age-related
HIPPOCAMPUS <
800-
AMYGDALA I
3 month
-
400
g 200
-
Onmol:
29
58
Cort
28
53
De×
29 58 Cort
28 53 De
K
Fig. 3. Binding of two different specific activities of radiolabeled corticosterone and dexamethasone by cell nuclei of hippocampus and amygdala in young and aged rats. Corticosterone was administered as 100/tCi-radiolabeled material plus t0 or 20/~g (29 or 58 nmol) of unlabeled. Dexamethasone was administered as 100/~Ci radiolabeted material plus 10 or 20#g (28 or 53 nmol) unlabeled. Two determinations were made for each dose and age group for dexamethasone, with each determination from a pair of pooled subjects; mean - S.E. Corticosterone determinations were made from a single pooled pair of subjects for each dose and age group as part of the dose response study in Table II. No significant age differences were observed in dexamethasone binding (independent t-test ).
239 decreases in the nuclear uptake of corticosterone in the brain, an endpoint which is relevent to understanding glucocorticoid-genomic interactions. The key finding is the regional specificity of the diminution of cell nuclear corticosterone receptor capacity in the aging brain. On the average, receptor capacity per cell (i.e. per mg DNA) declines with age in both the hippocampus and amygdala. The pituitary shows a non-significant trend towards a decline; the tremendous variance in pituitary determinations in the aged probably reflects microadenomas typical in that age-group6. The rest of the brain shows no change in binding with age. Our further investigation of cytosolic binding of dexamethasone demonstrated significant declines of hippocampal and amygdaloid binding comparable to declines in nuclear binding in those loci. Dexamethasone, when used in an in vitro cytosol assay, labels both glial and neuronal receptors, as does corticosterone when used in an in vivo cell nuclear uptake assay. The comparable declines revealed by both methods suggest that the reduced nuclear binding is due to reductions in cytosolic receptor number, rather than to decreases in receptor affinity or nuclear translocation of the steroid-receptor complex per se. The decline in hippocampal cytosolic binding with age is in agreement with previous reports 1. We next considered whether the declines were predominantly in the neuronal or non-neuronal (including glial) receptor populations. Selective labeling of non-neuronal nuclear receptors is evident by autoradiographic evidence after in vivo [3H]dexamethasone administration13,17, 21. Further evidence for dexamethasone preferentially interacting with glial and not neuronal receptors in vivo comes from its capacity to induce a glial-specific enzyme and its failure to induce a neuron-specific proteinl4,16. We have examined nuclear binding of dexamethasone throughout the brain and pituitary and find that the observed site-specific decreases in binding of corticosterone with age are not evident when dexamethasone is the ligand. A tendency towards increased binding of dexamethasone in aged subjects is in fact seen, a trend repeated in the pituitary and all brain regions assayed. In light of the present data, we conclude that the receptor decline revealed by in vivo administra-
tion of corticosterone but not by dexamethasone indicates an age-related deficit specific to neuronal populations. Glial hypertrophy appears to be a consistent correlate of aging in the non-pathologic brain, accounting for at least some of the increased DNA levels throughout the aged rat brain that we and others have observed (ref. 3 and unpublished observations). Such an increase might account for the trend towards increased in vivo dexamethasone binding seen throughout the aged brain in this study. In any case, since corticosterone interacts with both neuronal and glial receptors14,16, the inferred decline in hippocampal and amygdaloid neuronal receptors may even be more pronounced than reported, as the deficits may be somewhat masked by the binding due to glial receptors. The pituitary, hippocampus and amygdala are all thought to play roles in negative-feedback control of glucocorticoid secretion 11, and in the rat aging brings an impairment of such negative-feedback regulation19. Further, the adrenocortical axis exerts influences on learning and memory via the hippocampus 2, and the aged rat shows impairments of cognitive function 6. The possible relationships between the observed receptor declines and the age-related changes in neuroendocrine function are being investigated. Note. We have recently completed a detailed anatomic analysis of this corticosterone binding decline, using a method (Sapolsky, McEwen and Rainbow, in press, Brain Research) for quantitative steroid autoradiography. We find no loss of binding in the subiculum and dentate regions in the aged hippocampus, extensive loss in the CA3 cell field and lesser declines at other hippocampal sites, (Sapolsky, Rainbow, McEwen, in preparation). ACKNOWLEDGEMENTS We wish to thank Dr. B. Parsons for technical assistance during this study and Mrs. Oksana Wengerchuk and Dr. Amy Mandelker for manuscript assistance. Research support was provided by the National Institute on Aging via a predoctoral grant to R.M.S.
240 REFERENCES 1 Angelucci, L., Valeri, P., Grossi, E., Veldhuis, H., Bohus. B. and De Kloet, E., Involvement of hippocampal corticosterone receptors in behavioral phenomena. In F. Brambilla, G. Racagni and D. de Wied (Eds.), Progress in Psychoneuroendocrinology, Elsevier, Biomedical, Amsterdam, 1980, pp. 177-185. 2 Bohus, B. and De Kloet, E., Behavioral effects of neuropeptides related to LPH and ACTH (endorphins, enkephalin, ACTH fragments) and corticosteroids. In M. Jones, B. Gilham, M. Dallman and S. Chattopadhyay (Eds.), Interaction within the Brain-Pituitary-Adrenocortical System, Academic Press, New York, 1979, pp. 7-16. 3 Brizzee, K., Aging changes in relation to diseases of the nervous system. In M. Ordy and K. Brizzee (Eds.), Neurobiology of Aging, Plenum, New York, 1975, pp. 545-573. 4 Burton, K., A study of the conditions and mechanisms of the diphenylamine reaction for the colorimetric estimation of DNA, Biochem. J., 62 (1956) 315-323. 5 De Kloet, R., Wallach, G. and McEwen, B., Differences in corticosterone and dexamethasone binding to rat brain and pituitary, Endocrinology, 96 (1975) 598-609. 6 Finch, C. and Hayflick, L., Handbook of the Biology of Aging, Van Nostrand, 1977, 770 pp. 7 Lowry, O., Rosebrough, N., Farr, A. and Randall, R.. Protein measurement with the Folin phenol reagent. J. biol. Chem., 193 (1953) 265-275. 8 McEwen, B. and Pfaff, D., Factors infuencing sex hormone uptake by rat brain regions. I. Effects of neonatal treatment hypophysectomy, and competing steroid on estradiol uptake, Brain Research, 21 (1970) 1-16. 9 McEwen, B. and Zigmond, R., Isolation of brain cell nuclei. In N. Marks and R. Rodnight (Eds.), Research Methods in Neurochemistry, Vol. 1, Plenum Press, New York. 1972, pp. 140-161. 10 McEwen, B. Wallach, G. and Magnus, C., Corticosterone binding to hippocampus: immediate and delayed influences
11 12
13
14
15
16
17
18 19
20
21
of the absence of adrenal secretion, Brain Research, 7~J (1974) 321-324. McEwen, B., Adrenal steroid feedback on neuroendocrinc tissues, Ann. N. • Acad. Sci., 297 (1977) 568--579. McEwen, B., Stephenson. B. and Krey, L., Radioimmunoassay of brain tissue and cell nuclear corticosterone, J. Neurosci. Methods, 3 (1980) 57-65. McEwen, B., Glucocorticoids and hippocampus: receptors in seach of a function. In D. Ganten and D. Pfaff (Eds.). Current Topics in Neuroendocrinology, Vol. 2, Springer, Berlin, 1982, pp. 23-43. Meyer. J. and McEwen, B., Evidence for glucocorticoid target cells in the rat optic nerve; physicochemical characterization of cytosolic binding sites, J. Neurochem., 39 (1982) 435-442. Meyer, J., Micco, D., Stephenson, B., Krey, L. and McEwen, B., Subcutaneous implantation method for chronic glucocorticoid replacement therapy, Physiol. Behay., 22 (1970) 867-870. Nestler, E., Rainbow, T., MeEwen, B. and Greengard, P., Effect of steroid hormones on the level of protein in rat brain. In K. Fuxe (Ed.), Steroid Hormone Regulation of the Brain, Pergamon Press, Oxford, 1981, pp. 205-216. Rees, H., Stumpf, W. and Sar, M., Autoradiographic studies with -~H-dexamethasone in the rat brain and pituitary. In W. Stumpf and L. Grant (Eds.), Anatomical Neuroendocrinology, Karger Basel, 1966, pp. 262-269. Roth, G., Hormone action during aging; alteration and mechanisms, Mech. Aging. Develop., 9 (1979) 49%514. Sapolsky, R., Krey, L. and McEwen, B., The adrenocortical stress-response in the aged male rat; impairment of recovery from stress, Exp. Gerontol., 16 (1983) 55-64. Selye, H. and Tuchweber, B., Stress in relation to aging and disease. In A. Everitt and J. Burgess (Eds.). Hypothalamus, Pituitao,. and Aging, Charles Thomas, Springfield IL, 1976, pp. 557-573. Warembourg, M., Radioautographic study of the rat brain and pituitary after injection of 3H dexamethasone, Cell Tiss. Res., 16t (1975) 183-191.