Neuroscience Letters, 138 (1992) 157 160
157
© 1992 Elsevier Scientific Publishers Ireland Ltd. All rights reserved 0304-3940/92/$ 05.00
NSL 08548
Stress induces neuronal death in the hippocampus of castrated rats Kazushige Mizoguchi, Tatsuhide Kunishita, D e - H u a Chui and Takeshi Tabira National Institute of Neuroscience, NCNP, Tokyo (Japan) (Received 11 December 1991; Revised version received 20 January 1992; Accepted 22 January 1992)
Key words: Stress; Neuronal degeneration; Hippocampus; Aging; Steroid; Sex hormone Whereas loss of CA3 neurons in the hippocampus of monkeys which died of stress ulcers suggests that some structural changes may occur, there is no direct evidence that shows stress-induced irreversible changes of neurons. When rats were orchidectomized (castrated) and stressed by restraint and immersion in water for 15 min/day for 30 days, significant loss of hippocampal CA3 and CA4 neurons was observed. Furthermore, primary cultured hippocampal neurons survived shorter when treated with corticosterone. This neuronal loss was prevented by simultaneous administration of testosterone in vivo and in vitro. These findings indicate that stress can contribute to neuronal degeneration associated with hypogonadal conditions such as aging.
A variety of studies have indicated that stress can induce neuronal damage via the pituitary-adrenal axis in the aged [15]. For example, reduction of CA3 neurons was observed in the hippocampus of aged primates which died of stress ulcers [21]. Chronic administration of high-dose corticosterone induces neuronal loss in the hippocampus of 8-month-old normal rats [14], and adrenalectomy prevents hippocampal neuronal loss in aged normal rats [7]. In addition, stress increases basal levels of plasma corticosterone in rats and levels remain elevated for longer periods in aged rats compared to young rats [13]. The mechanism by which aged rats are more susceptible to corticosterone-induced neuronal damage is unknown. Since hypofunction of the reproductive system is a characteristic feature of senescence, we hypothesized that gonadal hypofunction may be involved in the neuronal loss in aged rats. To test this hypothesis, the testes or ovaries of adult rats were removed (orchidectomy or ovariectomy) and the effect of stress on the number of hippocampal neurons was investigated. In all experiments 9-month-old Wistar rats (Japan Charles River) were used, maintained under a specific pathogen-free environment, and all animal experiments were done in line with our institutional guide lines after permission of the Animal Research Committee. Orchidectomy (ORX) or ovariectomy (OVX) was done under anesthesia with ether, and control animals received sham Correspondence." T. Tabira, National Institute of Neuroscience, NCNP, 4-1-10gawa-higashi, Kodaira, Tokyo 187, Japan.
operation (no-ORX or no-OVX). One week after surgery, animals were stressed by restraint and immersion in water at room temperature for 15 min/day for 30 days (stress). This method of stress was used for the study of gastroduodenal ulcers [10, 19]. Then, animals were sacrificed by cardiac perfusion with 10% formalin under anesthesia, brains were removed and processed to paraffin, sectioned at a thickness of 5 ¢tm, and stained with hematoxylin and eosin. For quantitative studies, slides containing brain sections were coded, hippocampal neuronal cells were counted in 3 sections in each brain, and expressed as number/mm2. For counting, similar regions were selected in each group. Significance was determined by Student's t-test. Compared to control rats, the number of pyramidal neurons was decreased significantly in hippocampal CA3 and CA4 regions in orchidectomized and stressed rats (Figs. 1A-F and 2A). The number of neurons in the CA1, CA2 and dentate gyrus was not changed. Neurons that remained in CA3 and CA4 regions in the orchidectomized and stressed rats tended to be atrophic, and glial cells were increased. To see if this neuronal loss is induced by corticosterone alone and if testosterone can prevent the stress-induced neuronal loss, corticosterone (CORT, Sigma) and testosterone propionate (TEST, Nakarai Tesque) were dissolved in sesame oil and 20 mg/ kg of CORT or 6 mg/kg of TEST was injected subcutaneously every day for 30 days. The loss of neurons was also significant in orchidectomized rats that received daily injections of corticosterone (Figs. 1G,H and 2B),
158
Fig. 1. Representativephotomicrographs showing the effect of stress, corticosterone and testosterone on hippocampal neurons in orchidectomized or ovariectomizedrats. Animals received orchidectomy (ORX), ovariectomy (OVX) or sham operation (no-ORX or no-OVX) and stress (Stress). Someanimals receivedcorticosterone(CORT), or were stressed and receivedtestosterone (TEST) simultaneously. All animals were sacrificed 30 days after the first stress. Hippocampal regions are shown. A: CAl/no-ORX/stress; B: CAl/ORX/stress: C: CA3/no-ORX/stress; D: CA3/ORX/stress; E: CA4/no-ORX/stress, F: CA4/ORX/stress; G: CA3/ORX/vehicle; H: CA3/ORX/CORT: I: CA3/no-ORX/no-stress: J: CA3/ORX/stress; K: CA3/ ORX/stress/TEST: L: CA3/OVX/stress. Magnification, all 90x.
but neuronal loss was not found in orchidectomized and stressed rats which received simultaneous injections of testosterone (Figs, 11-K and 2C). The neuronal loss was not significant in ovariectomized and stressed rats, although atrophic and dark pyramidal neurons were observed in some cases (Figs. 1L and 2D). These results indicate that stress induces neuronal damage in the hippocampus and that sex hormones prevent such damage. The reason why males are more susceptible to stressinduced neuronal damage than females is unclear, but similar findings have been reported previously [14, 21]. Although the underlying mechanism of stress-induced neuronal loss is unknown, it is likely that stress-induced corticosterone from the adrenal cortex is primarily involved. We next tested if similar effects could be demonstrated in vitro. Primary cultures were established from the hippocampal region of Wistar rats on embryonic day 18 as previously described [6]. Various steroids were added, and the cells were cultured for another 20 h. Neuronal survival was measured by M T T (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide) cell growth assay using a commercial kit (Chemicon, Temecula, CA), which shows rnitochondrial activities in living cells [9]. Briefly, 10/~1 of 5 mg/ml M T T was added to the culture medium and incubated for 1 h at 37°C.
Brown color produced was extracted with 0.01 M HC1 in isopropanol and the absorbance value at 630 nm was subtracted from that of 570 nm. Statistical analysis was done by Student's t-test. As shown in Fig. 3, corticosterone and RU28362 but not aldosterone (data not shown) decreased survival of primary cultured hippocampal neurons. This decrease in neuronal survival was prevented by simultaneous treatment with estradiol or testosterone, but not with progesterone or dihydroepiandrosterone. Corticosterone-induced neuronal cell death was described previousy in in vitro studies [17]. These results suggest that sex hormones reactive on the glucocorticoid type II receptors can prevent the hippocampal neuronal damage. The hippocampus is a center for memory and a major site for pathology associated with Alzheimer's disease. Glucocorticoid type lI receptors are widely distributed in the central nervous system (CNS) [3, 4, 8, 12, 16], and stress- or glucocorticoid-induced neuronal damage has been described in the retina [11]. Ischemia-induced neuronal damage can be enhanced by glucocorticoid not only in the hippocampus but also in the neocortex and striatum [18]. The possibility exists, therefore, that stressinduced alterations of CNS neurons could affect multiple neuronal systems.
159
B
A
~00 3°c Q C~
"6 15C
i
150
,,,,,
E z
Z CA1
CA2
CA3
CA1
CA4
CA2
CA3
CA4
cI C~
IP
.a
lsb
.O
E
E
=3 Z
Z
CA3
CA3
CA4
CA4
Fig. 2. Effect of stress, corticosterone and testosterone on the number of hippocampal neurons in orchidectomized or ovariectomized rats. Experiments were done as in Fig. 1 and hippocampal neurons were counted in each region. Each column shows the mean and 1 S.D. of data from 4 rats. A: no-ORX/no-stress (solid), no-ORX/stress (open), ORX/stress (ladder); B: ORX/vehicle (solid), ORX/CORT (open); C: no-ORX/no-stress (solid), ORX/stress (open), ORX/stress/TEST (ladder); D: no-OVX/no-stress (solid), no-OVX/stress (open), OVX/stress (ladder); *P
157
© 1992 Elsevier Scientific Publishers Ireland Ltd. All rights reserved 0304-3940/92/$ 05.00
NSL 08548
Stress induces neuronal death in the hippocampus of castrated rats Kazushige Mizoguchi, Tatsuhide Kunishita, D e - H u a Chui and Takeshi Tabira National Institute of Neuroscience, NCNP, Tokyo (Japan) (Received 11 December 1991; Revised version received 20 January 1992; Accepted 22 January 1992)
Key words: Stress; Neuronal degeneration; Hippocampus; Aging; Steroid; Sex hormone Whereas loss of CA3 neurons in the hippocampus of monkeys which died of stress ulcers suggests that some structural changes may occur, there is no direct evidence that shows stress-induced irreversible changes of neurons. When rats were orchidectomized (castrated) and stressed by restraint and immersion in water for 15 min/day for 30 days, significant loss of hippocampal CA3 and CA4 neurons was observed. Furthermore, primary cultured hippocampal neurons survived shorter when treated with corticosterone. This neuronal loss was prevented by simultaneous administration of testosterone in vivo and in vitro. These findings indicate that stress can contribute to neuronal degeneration associated with hypogonadal conditions such as aging.
A variety of studies have indicated that stress can induce neuronal damage via the pituitary-adrenal axis in the aged [15]. For example, reduction of CA3 neurons was observed in the hippocampus of aged primates which died of stress ulcers [21]. Chronic administration of high-dose corticosterone induces neuronal loss in the hippocampus of 8-month-old normal rats [14], and adrenalectomy prevents hippocampal neuronal loss in aged normal rats [7]. In addition, stress increases basal levels of plasma corticosterone in rats and levels remain elevated for longer periods in aged rats compared to young rats [13]. The mechanism by which aged rats are more susceptible to corticosterone-induced neuronal damage is unknown. Since hypofunction of the reproductive system is a characteristic feature of senescence, we hypothesized that gonadal hypofunction may be involved in the neuronal loss in aged rats. To test this hypothesis, the testes or ovaries of adult rats were removed (orchidectomy or ovariectomy) and the effect of stress on the number of hippocampal neurons was investigated. In all experiments 9-month-old Wistar rats (Japan Charles River) were used, maintained under a specific pathogen-free environment, and all animal experiments were done in line with our institutional guide lines after permission of the Animal Research Committee. Orchidectomy (ORX) or ovariectomy (OVX) was done under anesthesia with ether, and control animals received sham Correspondence." T. Tabira, National Institute of Neuroscience, NCNP, 4-1-10gawa-higashi, Kodaira, Tokyo 187, Japan.
operation (no-ORX or no-OVX). One week after surgery, animals were stressed by restraint and immersion in water at room temperature for 15 min/day for 30 days (stress). This method of stress was used for the study of gastroduodenal ulcers [10, 19]. Then, animals were sacrificed by cardiac perfusion with 10% formalin under anesthesia, brains were removed and processed to paraffin, sectioned at a thickness of 5 ¢tm, and stained with hematoxylin and eosin. For quantitative studies, slides containing brain sections were coded, hippocampal neuronal cells were counted in 3 sections in each brain, and expressed as number/mm2. For counting, similar regions were selected in each group. Significance was determined by Student's t-test. Compared to control rats, the number of pyramidal neurons was decreased significantly in hippocampal CA3 and CA4 regions in orchidectomized and stressed rats (Figs. 1A-F and 2A). The number of neurons in the CA1, CA2 and dentate gyrus was not changed. Neurons that remained in CA3 and CA4 regions in the orchidectomized and stressed rats tended to be atrophic, and glial cells were increased. To see if this neuronal loss is induced by corticosterone alone and if testosterone can prevent the stress-induced neuronal loss, corticosterone (CORT, Sigma) and testosterone propionate (TEST, Nakarai Tesque) were dissolved in sesame oil and 20 mg/ kg of CORT or 6 mg/kg of TEST was injected subcutaneously every day for 30 days. The loss of neurons was also significant in orchidectomized rats that received daily injections of corticosterone (Figs. 1G,H and 2B),
158
Fig. 1. Representativephotomicrographs showing the effect of stress, corticosterone and testosterone on hippocampal neurons in orchidectomized or ovariectomizedrats. Animals received orchidectomy (ORX), ovariectomy (OVX) or sham operation (no-ORX or no-OVX) and stress (Stress). Someanimals receivedcorticosterone(CORT), or were stressed and receivedtestosterone (TEST) simultaneously. All animals were sacrificed 30 days after the first stress. Hippocampal regions are shown. A: CAl/no-ORX/stress; B: CAl/ORX/stress: C: CA3/no-ORX/stress; D: CA3/ORX/stress; E: CA4/no-ORX/stress, F: CA4/ORX/stress; G: CA3/ORX/vehicle; H: CA3/ORX/CORT: I: CA3/no-ORX/no-stress: J: CA3/ORX/stress; K: CA3/ ORX/stress/TEST: L: CA3/OVX/stress. Magnification, all 90x.
but neuronal loss was not found in orchidectomized and stressed rats which received simultaneous injections of testosterone (Figs, 11-K and 2C). The neuronal loss was not significant in ovariectomized and stressed rats, although atrophic and dark pyramidal neurons were observed in some cases (Figs. 1L and 2D). These results indicate that stress induces neuronal damage in the hippocampus and that sex hormones prevent such damage. The reason why males are more susceptible to stressinduced neuronal damage than females is unclear, but similar findings have been reported previously [14, 21]. Although the underlying mechanism of stress-induced neuronal loss is unknown, it is likely that stress-induced corticosterone from the adrenal cortex is primarily involved. We next tested if similar effects could be demonstrated in vitro. Primary cultures were established from the hippocampal region of Wistar rats on embryonic day 18 as previously described [6]. Various steroids were added, and the cells were cultured for another 20 h. Neuronal survival was measured by M T T (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide) cell growth assay using a commercial kit (Chemicon, Temecula, CA), which shows rnitochondrial activities in living cells [9]. Briefly, 10/~1 of 5 mg/ml M T T was added to the culture medium and incubated for 1 h at 37°C.
Brown color produced was extracted with 0.01 M HC1 in isopropanol and the absorbance value at 630 nm was subtracted from that of 570 nm. Statistical analysis was done by Student's t-test. As shown in Fig. 3, corticosterone and RU28362 but not aldosterone (data not shown) decreased survival of primary cultured hippocampal neurons. This decrease in neuronal survival was prevented by simultaneous treatment with estradiol or testosterone, but not with progesterone or dihydroepiandrosterone. Corticosterone-induced neuronal cell death was described previousy in in vitro studies [17]. These results suggest that sex hormones reactive on the glucocorticoid type II receptors can prevent the hippocampal neuronal damage. The hippocampus is a center for memory and a major site for pathology associated with Alzheimer's disease. Glucocorticoid type lI receptors are widely distributed in the central nervous system (CNS) [3, 4, 8, 12, 16], and stress- or glucocorticoid-induced neuronal damage has been described in the retina [11]. Ischemia-induced neuronal damage can be enhanced by glucocorticoid not only in the hippocampus but also in the neocortex and striatum [18]. The possibility exists, therefore, that stressinduced alterations of CNS neurons could affect multiple neuronal systems.
159
B
A
~00 3°c Q C~
"6 15C
i
150
,,,,,
E z
Z CA1
CA2
CA3
CA1
CA4
CA2
CA3
CA4
cI C~
IP
.a
lsb
.O
E
E
=3 Z
Z
CA3
CA3
CA4
CA4
Fig. 2. Effect of stress, corticosterone and testosterone on the number of hippocampal neurons in orchidectomized or ovariectomized rats. Experiments were done as in Fig. 1 and hippocampal neurons were counted in each region. Each column shows the mean and 1 S.D. of data from 4 rats. A: no-ORX/no-stress (solid), no-ORX/stress (open), ORX/stress (ladder); B: ORX/vehicle (solid), ORX/CORT (open); C: no-ORX/no-stress (solid), ORX/stress (open), ORX/stress/TEST (ladder); D: no-OVX/no-stress (solid), no-OVX/stress (open), OVX/stress (ladder); *P
