Brain Research, 585 (1992) 161-168
161
© 1992 Elsevier Science Publishers B.V. All rights reserved 0006-~993/92/$05.00
BRES 17823
Adrenal hormone effects on hippocampal excitatory amino acid binding A n n S. Clark and Carl W. C o t m a n Department of Psychobiology, University of California, lrvine, CA 92717 (USA) (Accepted 28 January 1992)
Key words: Corticosterone; Adrenalectomy; N-MethyI-D-aspartate receptor; Kainate; AMPA; Hippocampus
The influence of short-term adrenalectomy or corticosterone treatment on the binding of glutamate receptor subtypes in the rat hippocampus was explored using the technique of in vitro autoradiograpby. Analysis of NMDA, kainate and AMPA binding in the hippocampus was conducted on the brains of control, adrenalectomized, and adrenalectomized animals given corticosterone treatment. In addition, serum corticosterone levels were determined by RIA. No striking effects of acute adrenalectomy on the distribution or density of any glutamate receptor subtype were observed in the hippocampus. Adrenalectomy had a small but significant effect on kainate binding in the stratum lucidum and stratum radiatum of CA3 in the first experiment, but no effect in follow-up experiments. Short-term treatment with stress levels of corticosterone had no effect on the binding of NMDA or kainate in any hippocampal subfield. However, a small effect of high doses of corticosterone (CORT) was observed on AMPA binding in one subregion. Although the hippocampus is a target for glucocorticoids and uses excitatow amino acids as a primary neurotransmitter, transient manipulation of adrenal hormone levels did not directly modulate excitatory amino acid receptor binding.
INTRODUCTION The hippocampus is a critical structure for certain types of learning and memory and hippocampal damage can produce cognitive dysfunction 17'1s. Excitatory synaptic transmission in the hippocampus is primarily mediated by excitatory amino acids (EAA). Glutamate receptors in the brain can be classified into at least 3 different subtypes according to their sensitivity to exogenous compounds such as NMDA, quisqualic acid or kainate (reviewed in ref. 4). Among these receptor subtypes, the NMDA receptor has received considerable attention for its role in the neurophysiological processes underlying synaptic plasticity, in particular the induction of long-term potentiation (LTP)4'n'3z. In addition, the NMDA receptor has been associated with pathological processes such as the loss of neurons accompanying hypoxia-ischemia and the early pathophysiology which occurs in Alzheimer's disease 5'1°. The hippocampus is also a principal neural target site for adrenal steroids. Both type 1 and type 2 glucocortieoid receptors are present at high concentrations in the hippocampus s'tg. Our study of the interactions
between adrenal steroids and EAA binding in the hippocampus was prompted by research which analyzed the influence of adrenal steroids secreted in response to stress, such as corticosterone (CORT), on a variety of behavioral and physiological endpoints. For example, it has been shown that acute stress, using a combination of restraint plus tail shock, impairs learning of a behavioral task as well as the induction of LTP in the hippocampus 7'27'2s. Further evidence indicates that adrenalectomy reduces the threshold for primed burst potentiation in the hippocampus ~'. Acute changes in circulating adrenal steroids could rapidly modify hippocampal EAA systems, possibly changing EAA binding parameters and impacting on the processes required for learning and memory. A second point of convergence of action between adrenal steroids and EAA involves excitotoxicity. Exposure to CORT in vivo or in vitro makes hippocampal neurons more vulnerable to a variety of neurological insults 21-24. The compromised neuronal integrity most likely reflects the disruption of cellular energy metabolism systems by C O R T 13'16'21. It has been proposed that the synergy between CORT and neurologi-
Correspondence: A.S. Clark, (present address) Department of Psychology, Dartmouth College, 6207 Gerry Hall, Hanover, NH 03755-3459, USA. Fax: (1) 603-646-2810.
°
IM'~
DG-Ventml/o"
DG-Ventnd/l'~
DG-DmuVo
oo--
~
p_J
o
o
o
~M,le~mS~otem o
N
.
°~
8
fmol~
8
protein
•
DG-Venlzul/o"
D~rVentndfl-
DC,-Dmml/o£
°-
'
i
|
:N
iml
fmoles/msprotein
|
t~
163
cal insults involves activation of the NMDA cascade 2'23. In support of this hypothesis, administration of an NMDA receptor antagonist can block the synergistic actions of CORT and a non-NMDA type toxin 2. Furthermore, short-term treatment of adult rats with high levels of CORT significantly decreased glutamate binding in the CA3 and CA2 regions of hippocampus ~. The previous study examined glutamate binding in the presence or absence of chloride using L-[3H]-glutamate as the iigand. The goal of the present study was to extend this finding to determine the specific hippocampal glutamate receptor subtype(s) (i.e. NMDA, kainate or AMPA) modified by CORT treatment. Our experiment measured NMDA, kainate and AMPA receptors in the hippocampus of control, short-term adrenalectomized, and corticosterone-treated rats using the technique of in vitro autoradiography. MATERIALS AND METHODS
Animals Adult male Sprague-Dawley rats (Charles River) were used for these studies at approximately 9{) days of age. Five to ten rats were assigned to each treatment group. Adrenalectomized (ADX) rats were bilaterally adrenalectomized under sodium methohexital (Brevital, 50 mg/kg) anesthesia. Control rats were anesthetized and the adrenal gland exposed but not removed. All animals were housed in a temperature-controlled room with a 12 h light/dark cycle and had access to Purina lab chow and {}.9% saline drinking water. Hormone replacement was initiated on the day of surgery. In the first experiment, a subset of ADX rats received CORT replacement in the 0.9% saline drinkinl~ water at a concentration sufficient to maintain basal CORT levels (CORT-H 20, 25 #g/ml). This protocol allows the animals to drink o1,, a circadian rhythm, and the circulating CORT levels follow a diurnal cycle that mimics pattern in the intact animal ~'). in Exp. 2, a subset of the ADX rats received daily subcutaneous injections of 10 mg CORT in oil (CORT-inject) given during the first 2 h of the light portion of the L / D cycle. This regimen pn)duces transient stress levels of CORT in the supraphysiological range ~1,2,~. Receptor autoradiography In vitro autoradiographic measurements of excitatory amino acid binding in the hippocampus were assessed in control, ADX, ADX + CORT-treated rats in two separate experiments. Five to seven days after surgery, animals were sacrificed by decapitation and their brains rapidly removed and placed in powdered dry ice. Trunk blood was collected for radioimmunoassay of corticosterone. Once frozen, the brains were stored at -70°C until use (within I month). For autoradiographic procedures, the brains were placed into a cryostat cooled to -20°C. Six./~m coronal sections were taken through the
level of the hippocampus and thaw-mounted onto cooled acid-washed chrom-alum subbed slides. The slides were stored overnight at - 20°C prior to in vitro autoradiographic assay. Sections from animals of each treatment group were processed simultaneously for receptor binding so that subtle interassay variations would not confound the interpretation of the autoradiograms. The assay conditions are described below. NMDA. Briefly, for NMDA-sensitive L-glutamate binding, slides were warmed to room temperature prior to incubation in Tris-acetate buffer (pH 7.0) for 30 rain at 0°C. The slides were transferred to the same buffer for 2× 10 rain washes at 30°C to remove endogenous ligands. Slides were then incubated at 0°C with 100 nM L[3H]glutamate (spec. act. 63.5 Ci/mmol, NEN) for 10 rain in the presence of 1 ,¢M kainate, 5 p.M AMPA and 100 #M SITS (4acetmamido-4'-isothiocyanatostilbene-2,2'disulfonic acid, an anion channel blocker) to inhibit binding to non-NMDA receptors. Four rinses for a total of 30 s in ice-cold assay buffer followed, and the slides were rapidly dried under an air stream. Non-specific binding was determined in the presence of 200 ~M unlabeled NMDA. Kainate. To examine kainate receptors, slide-mounted tissue sections were preincubated in 50 mM Tris-citrate buffer (pH 7.0) for 30 min at 0°C and for 10 min at 30°C. The slides were then incubated for 30 min at 0°C with 50 nM [3H]kainate (spec. act. 58 Ci/mmol, NEN). At the completion of the incubation period, the slides were rapidly rinsed in 4 washes of ice-cold Tris buffer (for a total rinse time of 30 s) and dried rapidly under a stream of air. Non-specific binding was determined in the presence of 100 ,¢M unlabeled kainate. AMPA. For measurement of AMPA/quisqualate binding sites, slide-mounted tissue sections were preincubated for 30 min at 0°C in 50 mM Tris-acetate (pH 7.2) followed by a 10 rain incubation at 30°C. The slides were then incubated with 50 nM ['~H]AMPA ((R,S)a-amino-3-hydro~-5 methyi-4-isoxazolepropionic acid, spec. act. 27.6 Ci/mmol, NEN) in the presence of 100 mM KSCN (potassium thiocyanate) for 10 min. Slides were rinsed for 30 s total rinse time in 4 washes of Tris-acetate+KSCN. Non-specific binding was determined in the presence of 100 ~M unlabeled AMPA. This protocol allows measurement of the ionotropic AMPA-sensitive site, but does not show binding to the metabotropic, inositol phosphate-linked quisqualate site '~. Each group of slides was apposed to tritium.sensitive film (Amersham) in X-ray cassettes for 18-21 days along with a set of micro-scale standards (Amersham). The films were developed in Kodak D-19, fixed with rapid fix and allowed to dry.
Serunl hormone k,l.'els CORT levels were determined in the serum according to the method of Abraham et al?. CORT standards were purchased from Sigma and ['~H]CORT was obtained from Amersham. Antiserum to CORT was purchased from ICN Biochemicals. This antiserum dis. plays 100% cross-reactivity with CORT, 6% cross.reactivity with deoxycorticosterone and < 1% cross-reactivity with all other steroids. The intra- and inter-assay coefficients of variation were 6% and 10%, respectively. Data analysis Quantitative analysis of the autoladiograms was performed using tlle MCID system (Imaging Research, St. Catherine's, Canada). Slides were coded so that the experimenter making the densitometric measurements was blind to the treatment group of the animal. Eight
Fig. I. EAA binding sites (fmol/mg protein) in hippocampal subregions of control, ADX and ADX + CORT-H 20 rats. Brain regions: CAl-or, CAl-stratum oriens; CAl-rad, CAl-stratum radiatum; CA3-or, CA3-stratum oriens; CA3-rad, CA3-stratum radiatum; CA3-1uc, CA3.stratum lucidum; granule, dentate gyrus, granule cell layer; DG-dorsal/o, dentate gyrus-dorsal blade, outer molecular layers; DG.dorsal/i, inner molecular layers; DG-ventral/i, dentate gyrus-ventral blade, inner molecular layers; DG-ventral/o, dentate gyrus-ventral blade, outer molecular layers. A: short-term adrenal hormone manipulations failed to modify NMDA binding in any hippocampal region measured (F2.ts--0.59, P > 0.05). B: ADX rats had significantly increased kainate binding in the stratum lucidum of CA3 relative to controls, while kainate binding in the stratum radiatum of CA3 was significantly lower than controls (Newman-Keuls). Kainate levels in other hippocampal subregions did not differ between treatment groups. C: short-term manipulations of adrenal hormones had no effect on AMPA binding in any hippocampal subregion (F2.ts -- 0.23, P < 0.05). * P < 0.05 vs. control.
164 4OOO ~3000
1000
0
4000
"•
Kainate Binding
3000
2000
1000'
0,4
~]
40¢"
~ ~
30q
201
]
10(
1I
I j I i
165 to twelve measures were taken from each subregion of the hippocampus, and 3-5 sections were analyzed from each animal for each ligand. Specific binding was determined by subtracting the non-specific binding ( < 10%) from the total binding levels. Binding data were analyzed using a two-way analysis of variance with repeated measures. Post-hoc comparisons were made using the Newman-Keuls test. RESULTS
Short-term adrenalectomy and CORT-H20 NMDA. N M D A binding in the hippocampus was measured in control rats, adrenalectomized rats and adrenalectomized rats given corticosterone replacement in the drinking water for 5-7 days. The distribution of NMDA binding in the adult rat hippocampus is shown in Fig. 1A, The highest levels of NMDA binding were measured in CAl-stratum radiatum and the molecular layers of the dentate gyrus ( ,-, 2,600 fmol/mg protein). No main effect of hormone treatment (ADX or CORT-H 2O) was observed upon NMDA binding in the hippocampus (F:,aa ffi 0.59, P > 0.05). There was a significant main effect of brain region on NMDA binding, reflecting the nonuniform distribution of NMDA receptors throughout the subregions of the hippocampus (Fto, tao ffi 263, P < 0.05). No significant interaction between treatment and brain region was observed (F,o, ta0 ffi 0.46, P > 0.05). Kainate. The subregion of the hippocampus with the highest level of kainate binding in all treatment groups was the stratum lucidum of CA3 (Fig. IB, 2,500 fmol/mg protein). The two-way ANOVA with repeated measures revealed no significant overall main effect of hormone treatment on kainate binding in the rat hippocampus (F2,tH- 0.17, P > 0.05). As expected, based on the relatively low level of kainate binding in most of the hippocampal subregions studied, a main effect of brain region was observed (Fro, t, 0 ffi 2023, P < 0.05). In addition, there was a significant interaction between treatment and brain region (F2.ta 0 = 4.21, P < 0.05). The presence of this interaction warranted further analysis which revealed a small but significant increase and decrease in kainate binding in ADX rats relative to controls in CA3 stratum radiatum and stratum lucidum, respectively (Newman-Keuls, P < 0.05). AMPA. AMPA binding was highest in the stratum radiatum of CA1 and the molecular layers of the dentate gyrus (Fig. IC, 3,000 fmol/mg protein). Short-
4-Fig.
term manipulation of adrenal hormone levels had no significant main effect on AMPA binding in the rat hippvcampus (F2.~s=0.23, P>0.05). A significant main effect of brain region on AMPA binding was present (F~0,tso = 247, P < 0.05). There was also a significant interaction between treatment and brain region (F2oas0 = 1.87, P < 0.05). Further analysis of this interaction revealed no significant differences in AMPA binding between the treatment groups in any hippocampal subregion (Ncwman-Keuls, P > 0.05).
Glutamate binding following CORT-injections The second experiment examined the effects of short-term treatment with high doses of CORT (10 rag/day for 5-7 days) on hippocampal glutamate receptor binding. This injection protocol was selected on the basis of previous reports which had shown this regimen to produce stress levels of serum CORT in rats 11'24. Several animals treated with CORT-inject had serum CORT levels not different from controls and the data from these 4 animals was not included in the autoradiographic analysis. NMDA. No significant main effect of hormone treatment (ADX or CORT-inject) upon NMDA binding in the rat hippocampus was detected (F2,1s ffi 0.77, P > 0.05; Fig. 2A). Once again, a significant main effect of brain region on NMDA binding was present (Fto, lso = 696, P < 0.05). There was no significant interaction between brain region and treatment condition (F2oAxo = 0.78, P > 0.05). Kainate. No significant main effect of hormone treatment on binding of kainate in the hippocampus was observed in this experiment (F2.t, - 0.29, P > 0.05; Fig. 2B). However, a significant main effect of brain region was observed (Fi0,t,0 = 2278, P < 0.05). No significant interaction between treatment and brain area was observed (F20.1s0 ffi 1.47, P ffi 0.09). AMPA. Treatment had no significant overall main effect on AMPA binding in the hippocampus (F2.t~ = 0.73, P > 0.05; Fig. 2C). A significant main effect of brain region on AMPA binding was observed (Fit.20,~ = 167, P < 0.05). Furthermore, a significant treatment x brain region interaction was present (F22,2o,~= 2.15, P < 0.05). Post-hoc analysis of this interaction with Newman-Keuls test revealed that the level of AMPA binding in the outer molecular layer of the ventral blade of
2. EAA binding (fmol/mg protein) was measured in the hippocampusof control, ADX and ADX + CORT-injected (10 mg/day) animals.A: no effect of ADX or CORT-inject were observed on NMDA binding in any hippocampal subregion (F2.ts = 0.23, P > 0.05). B: no differences in kainate binding were detected between treatment groups (F2.ts= 0.29, P > 0.05). C: the level of AMPA binding in the DG.ventrai/o in CORT-inject rats was significantlyhigherthan in ADX rats (Newman-Keuls). * P < 0.05 vs. ADX.
166
the dentate gyrus was significantly higher in CURT-injected rats than in ADX rats (Newman-Keuls, P < 0.05).
Senmz CORT leceis Serum CURT levels were determined in the animals from each treatment group. As expected CORT-inject animals had serum CORT levels approximately 2-fold higher than control (intact) rats (control = 64 + 9 ng/ml vs. ADX + CORT-inject = 138 + 11 ng/ml). These values are in the range of levels observed after major stress -'4. Serum CURT levels in animals receiving CURT-HzU were significantly lower than control levels but significantly above the levels measured in ADX rats (ADX = 13 + 1 ng/mi vs. ADX + CURTH,O = 19 + 1 ng/ml; Student's t-test. P < 0.05). The low level of serum CORT in ADX + CORT-H,O animals reflects the rapid metabolism ot CURT administered in the drinking water TM. DISCUSSION
Short.term CORT and ghaamate hireling in the present study, adrenalectomy or treatment with corticosterone had minimal effects on excitatory amino acid binding in the hippocampu:;. Using the technique of quantitative in vitro autoradiography, the first experiment examined the impact of sl~ort-term adrenalectomy (5=7 days) and corticosteronc replacement on EAA receptor binding. No effect of adrenalectomy alone was observed upon the density of NMDA or AMPA binding in any brain region, Unly in the stratum radiatum and stratum lueidum of CA3 did we observe an effect of adrenalectomy on kainate binding and even then, this effect was in opposite directions in the two different brain regions. This change in binding in response to ADX was not reversed by corticosterone treatment and furthermore, additional attempts to replicate the effect of adrenalectomy on kainate binding were not successful (Exp. 2 and data not shown) leading us to consider this result somewhat cautiously, The marginal effect of adrenalectomy on EAA binding is in agreement with the report by Haipain and McEwen t~ of a small change in hippocampal glutamate binding in the CA2 and CA3 subfield in ADX rats. in a second experiment we examined the influence of short-term exposure to high levels of corticosterone on EAA receptor subtype binding, A CORT injection regimen which had previously been reported to produce high (stress) levels in rats, and had also been shown to modify hippocampal glutamate binding was
utilized 1=.24. Although serum CORT levels were elevated above control levels, the effects of elevated CORT on hippocampal EAA binding were minimal. No changes in NMDA or kainate binding in response to high CURT were detected in any hippocampal subfield. The only change in EAA receptor binding observed in response to high CURT was a small, but significant increase in AMPA binding in the outer molecular layer of the ventral blade of the dentate gyrus relative to ADX levels. There was no suggestion in our results of a tendency for CURT to decrease glutamate binding in any hippocampai subfield as had been previously reported !~.
M~thodological issues The observation of minimal effects of high doses of CORT on glutamate receptor binding was unexpected. We anticipated that the binding of at least one of the glutamate receptor subtypes under study in the hippocampus would be significantly reduced by CORT treatment. This result would be predicted based on the observation of a reduction in overall glutamate binding in response to CORT as reported by Halpain and McEwen ~. in a separate set of experiments we attempted to reproduce the exact assay conditions utilized by Halpain and McEwen (data not shown). These conditions call for two 5 rain rinses in assay buffer after incubation with the radioligand, in our hands such extensive rinsing results in a significant reduction in specific binding compared to the rinse time of 30 s routinely used in our laboratory fix, 2,100 I'mol/mg protein in the present study vs. 340 fmol/mg protein for the Haipain protocol, data not shown). We chose to use the more current assay conditions for our autoradiographic procedures because, in combination with the newly available ligands (['~H]kainate and ['~H]AMPA), and specific blockers (i.e. NMDA) these methods would produce the best, and least ambiguous, results, in addition, in our study we analyzed the individual layers (stratum oriens, stratum radiatum etc.) of hippocampal subregions, which significantly reduced the variability of our measurements. Finally, while the present experiment analyzed 3 glutamate receptor subtypes, additional EAA binding sites have recently been characterized 4"1~'2", The possibility that CORT could modify other hippocampal EAA binding sites remains to be determined. A second methodological issue pertains to the interpretation of the present findings in light of the study by Sloviter describing a loss of dentate granule cells in the adrenalectomized rat. While Sloviter 29 and others2s have reported a substantial loss of dentate granule cells in animals adrenalectomized for 3 months, one
167
group 9 has observed a significant increase in degenerating granule cells in the dentate as early as 7 days after adrenalectomy. The prcsent study failed to detect a significant reduction in the binding of any glutamate receptor subtype in the granule cells or molecular layer of the dentate gyrus following short-term adrenalectomy. It would appear that the magnitude of granule cell loss at this early stage does not translate into measurable changes in glutamate binding. However, in studies of the long-term adrenalectomized rat we have observed a dramatic reduction in glutamate receptor binding in the hippocampus of those animals with histologically confirmed granule cell loss (Clark, Cotman and Sloviter, unpublished results).
EAA and CORT interactions and hippocampal plasticity How can we integrate the present results into our
understanding of EAA and CORT actions in the hippocampus and their role in cognition? The most recent evidence argues against a direct role for corticosterone in modulating LTP. Although adrenalectomy has been shown to disrupt the induction of LTP 7'27, new evidence indicates LTP is not restored in ADX animals given CORT replacement, suggesting that a substance other than CORT is involved as. Furthermore, although it has been proposed that the induction of LTP and primed burst potentiation are influenced by circulating levels of adrenal steroids ~s. Warren et al. reported that inescapable shock had no effect on water maze training or retention "~t, a task known to be dependent on the integrity of the hippocampus ts. This finding raises new issues about attributing the cognitive effects of inescapable shock to impaired hippocampal functioning.~t. Two recent brief reports examining the influence of acute stress on glutamate receptor binding raise additional concerns regarding the mechanism of stress-induced changes in memory processes26''~°.These reports indicate that acute stress, i.e., a combination of restraint p!us tailshock for 60 min, significantlyincreased [ 3 H ] A M P A binding in some hippocampal subfields26'30. Surprisingly, but not inconsistent with our results,this effect could not be reproduced by administering high doses of corticosterone26. Thus, it appears that stress may modify hippocampal E A A binding vis a vis mechanisms which have not yet been determined. Future analyses of adrenal hormone effectson the plasticityof hippocampal E A A systems will address the impact of C O R T on other candidate endpoints for modulation, including C O R T modulation of calcium currents and second messenger systems. Acknowledgements. Portions of this work were presented at the Society for Neuroscienee meeting in October, 1990. We acknowledge
the technical assistance of Ms. Andrea Walencewiez. Serum eorticos-
terone levels were determined courtesy of Dr. AnthcJny L. Giordano, Washington University School of Medicine. Supported by a grant from the MacArthur Foundation to C.W.C. REFERENCES 1 Abraham, G.E., Manliemos, F.S. and Garza, R., Radioimmunoassay of steroids. In G.E. Abraham (Ed.), Handbook of Radioimmunoassays, Marcel Dekker, New York, 1977. 2 Armanini, M., Hutcbins, C., Stein, B. and Sapolslqt, R., Glucocorticoid endangerment of hippocampal neurons is NMDA-receptor dependent, Brain Res., 532 (1990) 7-12. 3 Cha, J., Makowiee, R., Penney, J. and Young, A., L-[3H]Giutamate labels the metabotropic excitatory amino acid receptor in rodent brain, Neurosci. Lett., 13 (1990) 78-83. 4 Cotman, C., Bridges, R., Taube, J., Clark, A.S., Geddes, J. and Monaghan, D., The role of the NMDA receptor in central nervous system plasticity and pathology, J. NIH Res,, 1 (1989) 65-74. 5 Cotman, C.W., Geddes, J., Bridges, R. and Monaghan, D., NMDA receptors and Alzheimer's disease, Neurobiol. Aging, 10 (1989) 603-605. 6 Diamond, D., Bennett, M., Engstrom, D., Fleshner, M. and Rose, G., Adrenalectomy reduced the threshold for hippocampal primed-burst potentiation in the anesthetized rat, Brain Res., 492 (1989) 356-360. 7 Foy, M., Stanton, M., Levine, S. and Thompson, R., Behavioral stress impairs long-term potentiation in rodent hippocampus, Behat'. Neural Biol., 48 (1987) 138-149. 8 Gerlach, J. and McEwen, B., Rat brain binds adrenal steroid hormone: radioautography of hippocampus with corticosterone, Science 175 (1972) 1133-1136. 9 Gould, E., Wooley, C. and McEwen, B.S., Short-term glucocorticoid manipulations affect neuronal morphology and survival in the adult dentate gyrus, Neuroscience 37 (1990) 367-375. 10 Greenamyre, J. and Young, A., Excitatory amino acids and Alzheimers disease, Neurobiol. Aging, 10 (1989) 593-602. II Halpain, S. and McEwen, B.S., Cor:!costerone decreases 31-1glutamate binding in rat hippocampal formation, Neuroen. docrinology, 48 (1988) 235-241. 12 Harris, E., Ganong, A. and Cotman, C., Long-term potentiation in the hippocampus involves activation of NMDA receptors, Brain Res., 323 (1984) 132-137. 13 Horner, H., Packan, D. and Sapolsky, R., Oiucocorticoids inhibit glucose transport in cultures of hippocampal neurons and gila, Neuroendocrinology, 52 (1990) 57-64. 14 McDonald, J., Penney, J., Johnston, M. and Young, A., Charac. terization and regional distribution of strychnine-insensitive [3H]glycine binding sites in rat brain by quantitative receptor autoradiography, Neuroscience, 35 (1990) 653-668. 15 Morris, R., Synaptic plasticity and learning: selective impairment of learning in rats and blockade of long-term potentiation in rive by the NMDA receptor antagonist AP5, J. Neurosci., 9 (1989) 3040-3057. 16 Munck, A., Guyre, P. and Holbrook, N., Physiological functions of glucocorticoids in stress and their relation to pharmacological actions, Endocr. Roy. 5 (1984)25-44. 17 Olton, D., Becker, L and Handelmann, G., Hippocampus, space and memory, Behae. Brain Sci. 2 (1979) 313-365. 18 O'Keefe, J. and Nadel, L., The Hippocampus as a Cognitive Map, Clarendon, Oxford University Press, New York, 1978, 19 Reul, J. and de Kloet, E., Two receptor systems for corticosterone in the rat brain: microdistribution and differential occupation, Endocrinology, 117 (1985) 2505-2511. 20 Sakurai, S., Cha, J., Penney, J. and Young, A., Regional distribution and properties of [3H] MK-801 binding sites determined by quantitative autoradiography in rat brain, Neuroscience, 40 (1991) 533-543. 21 Sapolsky, R., Packan, D. and Vale, W., Glucocorticoid toxicity in the hippocampus: in vitro demonstration, Brain Res., 453 (1988) 367-371.
168 22 Sapolsk'y, R. and Pulsinelli, W., Glucocorticoids potentiate ischemic injury to neurons: therapeutic implications. Science, 229 (1985) 1397-1400. 23 Sapolsky, R., Glucocorticoids, hippocampal damage and the glutamatergic synapse, Ping. Brain Res., 86 (19c;0) 13-23. 24 Sapolsky, R., Krey, L. and McEwen, B., Prolonged glucocorticoid exposure reduces hippcgampal number: implications for aging, J. Neurosci., 5 (1985) 1222-1227. 25 SaDolsky, R., Stein-Behrens, B. and Armanini, M., Long-term adrenalectomy causes loss of dentate gyrus and pyramidal neurons in the adult hippocampus, Exp. Neurol., 114 (1991) 246-249. 26 Shots, T., Tocco, G , Patel K., Baudry, M. and Thompson, R., Acute stress increases 3H-AMPA binding to the AMPA/quisqualate receptor in the hippocampus and the increase is not gluco.rurticoid-dependent, Soc. Neurosci. Abstr., 17 (1991) 915. 27 Shors T., Seib, T., Levine, S. and Thomps.an, R., Inescapable versus escapable shock modulates long-term potentiation in the rat hil,pocampus, Science, 244 (1989) 224-226.
28 Shors, T.J., Levine, S. and Thompson, R., Effect of adrenalectomy and demedullation on the stress-induced impairment of long-term potentiation, Neuroendocrinology, 51 (1990) 70-75. 29 $1oviter, R., Valiquette, G., Abrams, G., Ronk, E., Sollas, A., Faul, L. and Neubort, S., Selective loss of hippocampal granule cells in the mature rat brain after adrenalectomy, Science, 243 (1989) 535-538. 30 Tocco, G., Shors, T., Baudry, M. and Thompson, R., Selective increase in AMPA binding to the AMPA/quisqualate receptor in the hippocampus in response to acute stress, Brain Res., 559 (1991) 168-171. 31 Warren, D., Castro, C., Rudy, J. and Maier, S., No spatial learning impairment following exposure to inescapable shock, Psychobiology, 19 (1991) 127-134. 32 Watkins, J. and Evans, R. Excitatory amino acid transmitters, Atom. Ret'. Pharmacol. ToxicoL 21 (1981) 165-204.
161
© 1992 Elsevier Science Publishers B.V. All rights reserved 0006-~993/92/$05.00
BRES 17823
Adrenal hormone effects on hippocampal excitatory amino acid binding A n n S. Clark and Carl W. C o t m a n Department of Psychobiology, University of California, lrvine, CA 92717 (USA) (Accepted 28 January 1992)
Key words: Corticosterone; Adrenalectomy; N-MethyI-D-aspartate receptor; Kainate; AMPA; Hippocampus
The influence of short-term adrenalectomy or corticosterone treatment on the binding of glutamate receptor subtypes in the rat hippocampus was explored using the technique of in vitro autoradiograpby. Analysis of NMDA, kainate and AMPA binding in the hippocampus was conducted on the brains of control, adrenalectomized, and adrenalectomized animals given corticosterone treatment. In addition, serum corticosterone levels were determined by RIA. No striking effects of acute adrenalectomy on the distribution or density of any glutamate receptor subtype were observed in the hippocampus. Adrenalectomy had a small but significant effect on kainate binding in the stratum lucidum and stratum radiatum of CA3 in the first experiment, but no effect in follow-up experiments. Short-term treatment with stress levels of corticosterone had no effect on the binding of NMDA or kainate in any hippocampal subfield. However, a small effect of high doses of corticosterone (CORT) was observed on AMPA binding in one subregion. Although the hippocampus is a target for glucocorticoids and uses excitatow amino acids as a primary neurotransmitter, transient manipulation of adrenal hormone levels did not directly modulate excitatory amino acid receptor binding.
INTRODUCTION The hippocampus is a critical structure for certain types of learning and memory and hippocampal damage can produce cognitive dysfunction 17'1s. Excitatory synaptic transmission in the hippocampus is primarily mediated by excitatory amino acids (EAA). Glutamate receptors in the brain can be classified into at least 3 different subtypes according to their sensitivity to exogenous compounds such as NMDA, quisqualic acid or kainate (reviewed in ref. 4). Among these receptor subtypes, the NMDA receptor has received considerable attention for its role in the neurophysiological processes underlying synaptic plasticity, in particular the induction of long-term potentiation (LTP)4'n'3z. In addition, the NMDA receptor has been associated with pathological processes such as the loss of neurons accompanying hypoxia-ischemia and the early pathophysiology which occurs in Alzheimer's disease 5'1°. The hippocampus is also a principal neural target site for adrenal steroids. Both type 1 and type 2 glucocortieoid receptors are present at high concentrations in the hippocampus s'tg. Our study of the interactions
between adrenal steroids and EAA binding in the hippocampus was prompted by research which analyzed the influence of adrenal steroids secreted in response to stress, such as corticosterone (CORT), on a variety of behavioral and physiological endpoints. For example, it has been shown that acute stress, using a combination of restraint plus tail shock, impairs learning of a behavioral task as well as the induction of LTP in the hippocampus 7'27'2s. Further evidence indicates that adrenalectomy reduces the threshold for primed burst potentiation in the hippocampus ~'. Acute changes in circulating adrenal steroids could rapidly modify hippocampal EAA systems, possibly changing EAA binding parameters and impacting on the processes required for learning and memory. A second point of convergence of action between adrenal steroids and EAA involves excitotoxicity. Exposure to CORT in vivo or in vitro makes hippocampal neurons more vulnerable to a variety of neurological insults 21-24. The compromised neuronal integrity most likely reflects the disruption of cellular energy metabolism systems by C O R T 13'16'21. It has been proposed that the synergy between CORT and neurologi-
Correspondence: A.S. Clark, (present address) Department of Psychology, Dartmouth College, 6207 Gerry Hall, Hanover, NH 03755-3459, USA. Fax: (1) 603-646-2810.
°
IM'~
DG-Ventml/o"
DG-Ventnd/l'~
DG-DmuVo
oo--
~
p_J
o
o
o
~M,le~mS~otem o
N
.
°~
8
fmol~
8
protein
•
DG-Venlzul/o"
D~rVentndfl-
DC,-Dmml/o£
°-
'
i
|
:N
iml
fmoles/msprotein
|
t~
163
cal insults involves activation of the NMDA cascade 2'23. In support of this hypothesis, administration of an NMDA receptor antagonist can block the synergistic actions of CORT and a non-NMDA type toxin 2. Furthermore, short-term treatment of adult rats with high levels of CORT significantly decreased glutamate binding in the CA3 and CA2 regions of hippocampus ~. The previous study examined glutamate binding in the presence or absence of chloride using L-[3H]-glutamate as the iigand. The goal of the present study was to extend this finding to determine the specific hippocampal glutamate receptor subtype(s) (i.e. NMDA, kainate or AMPA) modified by CORT treatment. Our experiment measured NMDA, kainate and AMPA receptors in the hippocampus of control, short-term adrenalectomized, and corticosterone-treated rats using the technique of in vitro autoradiography. MATERIALS AND METHODS
Animals Adult male Sprague-Dawley rats (Charles River) were used for these studies at approximately 9{) days of age. Five to ten rats were assigned to each treatment group. Adrenalectomized (ADX) rats were bilaterally adrenalectomized under sodium methohexital (Brevital, 50 mg/kg) anesthesia. Control rats were anesthetized and the adrenal gland exposed but not removed. All animals were housed in a temperature-controlled room with a 12 h light/dark cycle and had access to Purina lab chow and {}.9% saline drinking water. Hormone replacement was initiated on the day of surgery. In the first experiment, a subset of ADX rats received CORT replacement in the 0.9% saline drinkinl~ water at a concentration sufficient to maintain basal CORT levels (CORT-H 20, 25 #g/ml). This protocol allows the animals to drink o1,, a circadian rhythm, and the circulating CORT levels follow a diurnal cycle that mimics pattern in the intact animal ~'). in Exp. 2, a subset of the ADX rats received daily subcutaneous injections of 10 mg CORT in oil (CORT-inject) given during the first 2 h of the light portion of the L / D cycle. This regimen pn)duces transient stress levels of CORT in the supraphysiological range ~1,2,~. Receptor autoradiography In vitro autoradiographic measurements of excitatory amino acid binding in the hippocampus were assessed in control, ADX, ADX + CORT-treated rats in two separate experiments. Five to seven days after surgery, animals were sacrificed by decapitation and their brains rapidly removed and placed in powdered dry ice. Trunk blood was collected for radioimmunoassay of corticosterone. Once frozen, the brains were stored at -70°C until use (within I month). For autoradiographic procedures, the brains were placed into a cryostat cooled to -20°C. Six./~m coronal sections were taken through the
level of the hippocampus and thaw-mounted onto cooled acid-washed chrom-alum subbed slides. The slides were stored overnight at - 20°C prior to in vitro autoradiographic assay. Sections from animals of each treatment group were processed simultaneously for receptor binding so that subtle interassay variations would not confound the interpretation of the autoradiograms. The assay conditions are described below. NMDA. Briefly, for NMDA-sensitive L-glutamate binding, slides were warmed to room temperature prior to incubation in Tris-acetate buffer (pH 7.0) for 30 rain at 0°C. The slides were transferred to the same buffer for 2× 10 rain washes at 30°C to remove endogenous ligands. Slides were then incubated at 0°C with 100 nM L[3H]glutamate (spec. act. 63.5 Ci/mmol, NEN) for 10 rain in the presence of 1 ,¢M kainate, 5 p.M AMPA and 100 #M SITS (4acetmamido-4'-isothiocyanatostilbene-2,2'disulfonic acid, an anion channel blocker) to inhibit binding to non-NMDA receptors. Four rinses for a total of 30 s in ice-cold assay buffer followed, and the slides were rapidly dried under an air stream. Non-specific binding was determined in the presence of 200 ~M unlabeled NMDA. Kainate. To examine kainate receptors, slide-mounted tissue sections were preincubated in 50 mM Tris-citrate buffer (pH 7.0) for 30 min at 0°C and for 10 min at 30°C. The slides were then incubated for 30 min at 0°C with 50 nM [3H]kainate (spec. act. 58 Ci/mmol, NEN). At the completion of the incubation period, the slides were rapidly rinsed in 4 washes of ice-cold Tris buffer (for a total rinse time of 30 s) and dried rapidly under a stream of air. Non-specific binding was determined in the presence of 100 ,¢M unlabeled kainate. AMPA. For measurement of AMPA/quisqualate binding sites, slide-mounted tissue sections were preincubated for 30 min at 0°C in 50 mM Tris-acetate (pH 7.2) followed by a 10 rain incubation at 30°C. The slides were then incubated with 50 nM ['~H]AMPA ((R,S)a-amino-3-hydro~-5 methyi-4-isoxazolepropionic acid, spec. act. 27.6 Ci/mmol, NEN) in the presence of 100 mM KSCN (potassium thiocyanate) for 10 min. Slides were rinsed for 30 s total rinse time in 4 washes of Tris-acetate+KSCN. Non-specific binding was determined in the presence of 100 ~M unlabeled AMPA. This protocol allows measurement of the ionotropic AMPA-sensitive site, but does not show binding to the metabotropic, inositol phosphate-linked quisqualate site '~. Each group of slides was apposed to tritium.sensitive film (Amersham) in X-ray cassettes for 18-21 days along with a set of micro-scale standards (Amersham). The films were developed in Kodak D-19, fixed with rapid fix and allowed to dry.
Serunl hormone k,l.'els CORT levels were determined in the serum according to the method of Abraham et al?. CORT standards were purchased from Sigma and ['~H]CORT was obtained from Amersham. Antiserum to CORT was purchased from ICN Biochemicals. This antiserum dis. plays 100% cross-reactivity with CORT, 6% cross.reactivity with deoxycorticosterone and < 1% cross-reactivity with all other steroids. The intra- and inter-assay coefficients of variation were 6% and 10%, respectively. Data analysis Quantitative analysis of the autoladiograms was performed using tlle MCID system (Imaging Research, St. Catherine's, Canada). Slides were coded so that the experimenter making the densitometric measurements was blind to the treatment group of the animal. Eight
Fig. I. EAA binding sites (fmol/mg protein) in hippocampal subregions of control, ADX and ADX + CORT-H 20 rats. Brain regions: CAl-or, CAl-stratum oriens; CAl-rad, CAl-stratum radiatum; CA3-or, CA3-stratum oriens; CA3-rad, CA3-stratum radiatum; CA3-1uc, CA3.stratum lucidum; granule, dentate gyrus, granule cell layer; DG-dorsal/o, dentate gyrus-dorsal blade, outer molecular layers; DG.dorsal/i, inner molecular layers; DG-ventral/i, dentate gyrus-ventral blade, inner molecular layers; DG-ventral/o, dentate gyrus-ventral blade, outer molecular layers. A: short-term adrenal hormone manipulations failed to modify NMDA binding in any hippocampal region measured (F2.ts--0.59, P > 0.05). B: ADX rats had significantly increased kainate binding in the stratum lucidum of CA3 relative to controls, while kainate binding in the stratum radiatum of CA3 was significantly lower than controls (Newman-Keuls). Kainate levels in other hippocampal subregions did not differ between treatment groups. C: short-term manipulations of adrenal hormones had no effect on AMPA binding in any hippocampal subregion (F2.ts -- 0.23, P < 0.05). * P < 0.05 vs. control.
164 4OOO ~3000
1000
0
4000
"•
Kainate Binding
3000
2000
1000'
0,4
~]
40¢"
~ ~
30q
201
]
10(
1I
I j I i
165 to twelve measures were taken from each subregion of the hippocampus, and 3-5 sections were analyzed from each animal for each ligand. Specific binding was determined by subtracting the non-specific binding ( < 10%) from the total binding levels. Binding data were analyzed using a two-way analysis of variance with repeated measures. Post-hoc comparisons were made using the Newman-Keuls test. RESULTS
Short-term adrenalectomy and CORT-H20 NMDA. N M D A binding in the hippocampus was measured in control rats, adrenalectomized rats and adrenalectomized rats given corticosterone replacement in the drinking water for 5-7 days. The distribution of NMDA binding in the adult rat hippocampus is shown in Fig. 1A, The highest levels of NMDA binding were measured in CAl-stratum radiatum and the molecular layers of the dentate gyrus ( ,-, 2,600 fmol/mg protein). No main effect of hormone treatment (ADX or CORT-H 2O) was observed upon NMDA binding in the hippocampus (F:,aa ffi 0.59, P > 0.05). There was a significant main effect of brain region on NMDA binding, reflecting the nonuniform distribution of NMDA receptors throughout the subregions of the hippocampus (Fto, tao ffi 263, P < 0.05). No significant interaction between treatment and brain region was observed (F,o, ta0 ffi 0.46, P > 0.05). Kainate. The subregion of the hippocampus with the highest level of kainate binding in all treatment groups was the stratum lucidum of CA3 (Fig. IB, 2,500 fmol/mg protein). The two-way ANOVA with repeated measures revealed no significant overall main effect of hormone treatment on kainate binding in the rat hippocampus (F2,tH- 0.17, P > 0.05). As expected, based on the relatively low level of kainate binding in most of the hippocampal subregions studied, a main effect of brain region was observed (Fro, t, 0 ffi 2023, P < 0.05). In addition, there was a significant interaction between treatment and brain region (F2.ta 0 = 4.21, P < 0.05). The presence of this interaction warranted further analysis which revealed a small but significant increase and decrease in kainate binding in ADX rats relative to controls in CA3 stratum radiatum and stratum lucidum, respectively (Newman-Keuls, P < 0.05). AMPA. AMPA binding was highest in the stratum radiatum of CA1 and the molecular layers of the dentate gyrus (Fig. IC, 3,000 fmol/mg protein). Short-
4-Fig.
term manipulation of adrenal hormone levels had no significant main effect on AMPA binding in the rat hippvcampus (F2.~s=0.23, P>0.05). A significant main effect of brain region on AMPA binding was present (F~0,tso = 247, P < 0.05). There was also a significant interaction between treatment and brain region (F2oas0 = 1.87, P < 0.05). Further analysis of this interaction revealed no significant differences in AMPA binding between the treatment groups in any hippocampal subregion (Ncwman-Keuls, P > 0.05).
Glutamate binding following CORT-injections The second experiment examined the effects of short-term treatment with high doses of CORT (10 rag/day for 5-7 days) on hippocampal glutamate receptor binding. This injection protocol was selected on the basis of previous reports which had shown this regimen to produce stress levels of serum CORT in rats 11'24. Several animals treated with CORT-inject had serum CORT levels not different from controls and the data from these 4 animals was not included in the autoradiographic analysis. NMDA. No significant main effect of hormone treatment (ADX or CORT-inject) upon NMDA binding in the rat hippocampus was detected (F2,1s ffi 0.77, P > 0.05; Fig. 2A). Once again, a significant main effect of brain region on NMDA binding was present (Fto, lso = 696, P < 0.05). There was no significant interaction between brain region and treatment condition (F2oAxo = 0.78, P > 0.05). Kainate. No significant main effect of hormone treatment on binding of kainate in the hippocampus was observed in this experiment (F2.t, - 0.29, P > 0.05; Fig. 2B). However, a significant main effect of brain region was observed (Fi0,t,0 = 2278, P < 0.05). No significant interaction between treatment and brain area was observed (F20.1s0 ffi 1.47, P ffi 0.09). AMPA. Treatment had no significant overall main effect on AMPA binding in the hippocampus (F2.t~ = 0.73, P > 0.05; Fig. 2C). A significant main effect of brain region on AMPA binding was observed (Fit.20,~ = 167, P < 0.05). Furthermore, a significant treatment x brain region interaction was present (F22,2o,~= 2.15, P < 0.05). Post-hoc analysis of this interaction with Newman-Keuls test revealed that the level of AMPA binding in the outer molecular layer of the ventral blade of
2. EAA binding (fmol/mg protein) was measured in the hippocampusof control, ADX and ADX + CORT-injected (10 mg/day) animals.A: no effect of ADX or CORT-inject were observed on NMDA binding in any hippocampal subregion (F2.ts = 0.23, P > 0.05). B: no differences in kainate binding were detected between treatment groups (F2.ts= 0.29, P > 0.05). C: the level of AMPA binding in the DG.ventrai/o in CORT-inject rats was significantlyhigherthan in ADX rats (Newman-Keuls). * P < 0.05 vs. ADX.
166
the dentate gyrus was significantly higher in CURT-injected rats than in ADX rats (Newman-Keuls, P < 0.05).
Senmz CORT leceis Serum CURT levels were determined in the animals from each treatment group. As expected CORT-inject animals had serum CORT levels approximately 2-fold higher than control (intact) rats (control = 64 + 9 ng/ml vs. ADX + CORT-inject = 138 + 11 ng/ml). These values are in the range of levels observed after major stress -'4. Serum CURT levels in animals receiving CURT-HzU were significantly lower than control levels but significantly above the levels measured in ADX rats (ADX = 13 + 1 ng/mi vs. ADX + CURTH,O = 19 + 1 ng/ml; Student's t-test. P < 0.05). The low level of serum CORT in ADX + CORT-H,O animals reflects the rapid metabolism ot CURT administered in the drinking water TM. DISCUSSION
Short.term CORT and ghaamate hireling in the present study, adrenalectomy or treatment with corticosterone had minimal effects on excitatory amino acid binding in the hippocampu:;. Using the technique of quantitative in vitro autoradiography, the first experiment examined the impact of sl~ort-term adrenalectomy (5=7 days) and corticosteronc replacement on EAA receptor binding. No effect of adrenalectomy alone was observed upon the density of NMDA or AMPA binding in any brain region, Unly in the stratum radiatum and stratum lueidum of CA3 did we observe an effect of adrenalectomy on kainate binding and even then, this effect was in opposite directions in the two different brain regions. This change in binding in response to ADX was not reversed by corticosterone treatment and furthermore, additional attempts to replicate the effect of adrenalectomy on kainate binding were not successful (Exp. 2 and data not shown) leading us to consider this result somewhat cautiously, The marginal effect of adrenalectomy on EAA binding is in agreement with the report by Haipain and McEwen t~ of a small change in hippocampal glutamate binding in the CA2 and CA3 subfield in ADX rats. in a second experiment we examined the influence of short-term exposure to high levels of corticosterone on EAA receptor subtype binding, A CORT injection regimen which had previously been reported to produce high (stress) levels in rats, and had also been shown to modify hippocampal glutamate binding was
utilized 1=.24. Although serum CORT levels were elevated above control levels, the effects of elevated CORT on hippocampal EAA binding were minimal. No changes in NMDA or kainate binding in response to high CURT were detected in any hippocampal subfield. The only change in EAA receptor binding observed in response to high CURT was a small, but significant increase in AMPA binding in the outer molecular layer of the ventral blade of the dentate gyrus relative to ADX levels. There was no suggestion in our results of a tendency for CURT to decrease glutamate binding in any hippocampai subfield as had been previously reported !~.
M~thodological issues The observation of minimal effects of high doses of CORT on glutamate receptor binding was unexpected. We anticipated that the binding of at least one of the glutamate receptor subtypes under study in the hippocampus would be significantly reduced by CORT treatment. This result would be predicted based on the observation of a reduction in overall glutamate binding in response to CORT as reported by Halpain and McEwen ~. in a separate set of experiments we attempted to reproduce the exact assay conditions utilized by Halpain and McEwen (data not shown). These conditions call for two 5 rain rinses in assay buffer after incubation with the radioligand, in our hands such extensive rinsing results in a significant reduction in specific binding compared to the rinse time of 30 s routinely used in our laboratory fix, 2,100 I'mol/mg protein in the present study vs. 340 fmol/mg protein for the Haipain protocol, data not shown). We chose to use the more current assay conditions for our autoradiographic procedures because, in combination with the newly available ligands (['~H]kainate and ['~H]AMPA), and specific blockers (i.e. NMDA) these methods would produce the best, and least ambiguous, results, in addition, in our study we analyzed the individual layers (stratum oriens, stratum radiatum etc.) of hippocampal subregions, which significantly reduced the variability of our measurements. Finally, while the present experiment analyzed 3 glutamate receptor subtypes, additional EAA binding sites have recently been characterized 4"1~'2", The possibility that CORT could modify other hippocampal EAA binding sites remains to be determined. A second methodological issue pertains to the interpretation of the present findings in light of the study by Sloviter describing a loss of dentate granule cells in the adrenalectomized rat. While Sloviter 29 and others2s have reported a substantial loss of dentate granule cells in animals adrenalectomized for 3 months, one
167
group 9 has observed a significant increase in degenerating granule cells in the dentate as early as 7 days after adrenalectomy. The prcsent study failed to detect a significant reduction in the binding of any glutamate receptor subtype in the granule cells or molecular layer of the dentate gyrus following short-term adrenalectomy. It would appear that the magnitude of granule cell loss at this early stage does not translate into measurable changes in glutamate binding. However, in studies of the long-term adrenalectomized rat we have observed a dramatic reduction in glutamate receptor binding in the hippocampus of those animals with histologically confirmed granule cell loss (Clark, Cotman and Sloviter, unpublished results).
EAA and CORT interactions and hippocampal plasticity How can we integrate the present results into our
understanding of EAA and CORT actions in the hippocampus and their role in cognition? The most recent evidence argues against a direct role for corticosterone in modulating LTP. Although adrenalectomy has been shown to disrupt the induction of LTP 7'27, new evidence indicates LTP is not restored in ADX animals given CORT replacement, suggesting that a substance other than CORT is involved as. Furthermore, although it has been proposed that the induction of LTP and primed burst potentiation are influenced by circulating levels of adrenal steroids ~s. Warren et al. reported that inescapable shock had no effect on water maze training or retention "~t, a task known to be dependent on the integrity of the hippocampus ts. This finding raises new issues about attributing the cognitive effects of inescapable shock to impaired hippocampal functioning.~t. Two recent brief reports examining the influence of acute stress on glutamate receptor binding raise additional concerns regarding the mechanism of stress-induced changes in memory processes26''~°.These reports indicate that acute stress, i.e., a combination of restraint p!us tailshock for 60 min, significantlyincreased [ 3 H ] A M P A binding in some hippocampal subfields26'30. Surprisingly, but not inconsistent with our results,this effect could not be reproduced by administering high doses of corticosterone26. Thus, it appears that stress may modify hippocampal E A A binding vis a vis mechanisms which have not yet been determined. Future analyses of adrenal hormone effectson the plasticityof hippocampal E A A systems will address the impact of C O R T on other candidate endpoints for modulation, including C O R T modulation of calcium currents and second messenger systems. Acknowledgements. Portions of this work were presented at the Society for Neuroscienee meeting in October, 1990. We acknowledge
the technical assistance of Ms. Andrea Walencewiez. Serum eorticos-
terone levels were determined courtesy of Dr. AnthcJny L. Giordano, Washington University School of Medicine. Supported by a grant from the MacArthur Foundation to C.W.C. REFERENCES 1 Abraham, G.E., Manliemos, F.S. and Garza, R., Radioimmunoassay of steroids. In G.E. Abraham (Ed.), Handbook of Radioimmunoassays, Marcel Dekker, New York, 1977. 2 Armanini, M., Hutcbins, C., Stein, B. and Sapolslqt, R., Glucocorticoid endangerment of hippocampal neurons is NMDA-receptor dependent, Brain Res., 532 (1990) 7-12. 3 Cha, J., Makowiee, R., Penney, J. and Young, A., L-[3H]Giutamate labels the metabotropic excitatory amino acid receptor in rodent brain, Neurosci. Lett., 13 (1990) 78-83. 4 Cotman, C., Bridges, R., Taube, J., Clark, A.S., Geddes, J. and Monaghan, D., The role of the NMDA receptor in central nervous system plasticity and pathology, J. NIH Res,, 1 (1989) 65-74. 5 Cotman, C.W., Geddes, J., Bridges, R. and Monaghan, D., NMDA receptors and Alzheimer's disease, Neurobiol. Aging, 10 (1989) 603-605. 6 Diamond, D., Bennett, M., Engstrom, D., Fleshner, M. and Rose, G., Adrenalectomy reduced the threshold for hippocampal primed-burst potentiation in the anesthetized rat, Brain Res., 492 (1989) 356-360. 7 Foy, M., Stanton, M., Levine, S. and Thompson, R., Behavioral stress impairs long-term potentiation in rodent hippocampus, Behat'. Neural Biol., 48 (1987) 138-149. 8 Gerlach, J. and McEwen, B., Rat brain binds adrenal steroid hormone: radioautography of hippocampus with corticosterone, Science 175 (1972) 1133-1136. 9 Gould, E., Wooley, C. and McEwen, B.S., Short-term glucocorticoid manipulations affect neuronal morphology and survival in the adult dentate gyrus, Neuroscience 37 (1990) 367-375. 10 Greenamyre, J. and Young, A., Excitatory amino acids and Alzheimers disease, Neurobiol. Aging, 10 (1989) 593-602. II Halpain, S. and McEwen, B.S., Cor:!costerone decreases 31-1glutamate binding in rat hippocampal formation, Neuroen. docrinology, 48 (1988) 235-241. 12 Harris, E., Ganong, A. and Cotman, C., Long-term potentiation in the hippocampus involves activation of NMDA receptors, Brain Res., 323 (1984) 132-137. 13 Horner, H., Packan, D. and Sapolsky, R., Oiucocorticoids inhibit glucose transport in cultures of hippocampal neurons and gila, Neuroendocrinology, 52 (1990) 57-64. 14 McDonald, J., Penney, J., Johnston, M. and Young, A., Charac. terization and regional distribution of strychnine-insensitive [3H]glycine binding sites in rat brain by quantitative receptor autoradiography, Neuroscience, 35 (1990) 653-668. 15 Morris, R., Synaptic plasticity and learning: selective impairment of learning in rats and blockade of long-term potentiation in rive by the NMDA receptor antagonist AP5, J. Neurosci., 9 (1989) 3040-3057. 16 Munck, A., Guyre, P. and Holbrook, N., Physiological functions of glucocorticoids in stress and their relation to pharmacological actions, Endocr. Roy. 5 (1984)25-44. 17 Olton, D., Becker, L and Handelmann, G., Hippocampus, space and memory, Behae. Brain Sci. 2 (1979) 313-365. 18 O'Keefe, J. and Nadel, L., The Hippocampus as a Cognitive Map, Clarendon, Oxford University Press, New York, 1978, 19 Reul, J. and de Kloet, E., Two receptor systems for corticosterone in the rat brain: microdistribution and differential occupation, Endocrinology, 117 (1985) 2505-2511. 20 Sakurai, S., Cha, J., Penney, J. and Young, A., Regional distribution and properties of [3H] MK-801 binding sites determined by quantitative autoradiography in rat brain, Neuroscience, 40 (1991) 533-543. 21 Sapolsky, R., Packan, D. and Vale, W., Glucocorticoid toxicity in the hippocampus: in vitro demonstration, Brain Res., 453 (1988) 367-371.
168 22 Sapolsk'y, R. and Pulsinelli, W., Glucocorticoids potentiate ischemic injury to neurons: therapeutic implications. Science, 229 (1985) 1397-1400. 23 Sapolsky, R., Glucocorticoids, hippocampal damage and the glutamatergic synapse, Ping. Brain Res., 86 (19c;0) 13-23. 24 Sapolsky, R., Krey, L. and McEwen, B., Prolonged glucocorticoid exposure reduces hippcgampal number: implications for aging, J. Neurosci., 5 (1985) 1222-1227. 25 SaDolsky, R., Stein-Behrens, B. and Armanini, M., Long-term adrenalectomy causes loss of dentate gyrus and pyramidal neurons in the adult hippocampus, Exp. Neurol., 114 (1991) 246-249. 26 Shots, T., Tocco, G , Patel K., Baudry, M. and Thompson, R., Acute stress increases 3H-AMPA binding to the AMPA/quisqualate receptor in the hippocampus and the increase is not gluco.rurticoid-dependent, Soc. Neurosci. Abstr., 17 (1991) 915. 27 Shors T., Seib, T., Levine, S. and Thomps.an, R., Inescapable versus escapable shock modulates long-term potentiation in the rat hil,pocampus, Science, 244 (1989) 224-226.
28 Shors, T.J., Levine, S. and Thompson, R., Effect of adrenalectomy and demedullation on the stress-induced impairment of long-term potentiation, Neuroendocrinology, 51 (1990) 70-75. 29 $1oviter, R., Valiquette, G., Abrams, G., Ronk, E., Sollas, A., Faul, L. and Neubort, S., Selective loss of hippocampal granule cells in the mature rat brain after adrenalectomy, Science, 243 (1989) 535-538. 30 Tocco, G., Shors, T., Baudry, M. and Thompson, R., Selective increase in AMPA binding to the AMPA/quisqualate receptor in the hippocampus in response to acute stress, Brain Res., 559 (1991) 168-171. 31 Warren, D., Castro, C., Rudy, J. and Maier, S., No spatial learning impairment following exposure to inescapable shock, Psychobiology, 19 (1991) 127-134. 32 Watkins, J. and Evans, R. Excitatory amino acid transmitters, Atom. Ret'. Pharmacol. ToxicoL 21 (1981) 165-204.
