by corticosteroid receptor activation

Transcription

1 Psyhobiology 1994, 22 (2), Modulation of hippoampal long-term potentiation and long-term depression by ortiosteroid reeptor ativation D. STEVEN KERR, ANN M. HUGGETT, and WICKLIFFE C. ABRAHAM University of Otago, Dunedin, New Zealand Long-term potentiation (LTP) and long-term depression (LTD) are lasting hanges in synapti effiay whih may underlie memory and learning proesses. Hippoampal LTP has been shown to vary inversely with the degree of stress or adrenal ativity, and adrenal ortiosteroids (CORT) have been strongly impliated in this proess. The present in vitro study was undertaken to further examine the role of gluoortioids in LTP, and to extend this analysis to heterosynapti LTD. Animals were injeted twie daily (s..) for 2 days prior to sarifie with either the ortiosteroid biosynthesis inhibitor metyrapone (200 mg/kg/day) or CORT (20 mg/kg/day). Noninjeted and saline-injeted ontrols were also examined. CAl population spike amplitudes and field EPSP slopes were assessed in response to ativation of two separate stratum radiatum inputs. One input reeived tetani stimulation and both inputs were monitored for 30 min following tetanization to determine long-term hanges in synapti effiay (homosynapti LTP of tetanized synapses; heterosynapti LTD of nontetanized synapses). Animals exhibiting low and high serum CORT titers exhibited signifiantly less LTP and LTD than did animals exhibiting moderate levels of CORT. In addition, in vitro perfusion of RU-28362, a Type II gluoortioid reeptor agonist, markedly redued both LTP and LTD. There was no orrelation between serum CORT and serum gluose titers, nor was there any orrelation between serum gluose and either LTP or LTD. These results are onsistent with reent reports of an inverted-u funtion for CORT levels versus primedburst LTP (Bennett, Diamond, Fleshner, & Rose, 1991; Diamond, Bennett, Fleshner, & Rose, 1992) and further demonstrate that manipulation of ortiosterone in adrenal-intat animals affets LTP and LTD in a similar fashion. It is well established that the hippoampus and related brain strutures are rih in speifi ortiosteroid reeptors (CSRs) (Fuxe et al., 1985; MEwen, Wallah, & Magnus, 1974; MEwen, Weiss, & Shwartz, 1969; Reul & de Kloet, 1986). At least two types of ytosoli CSRs have been differentiated on the basis of their steroid binding harateristis: Type I (MR, mineraloortioid/ortiosteroid) and Type IT (GR, gluoortioid) reeptors. In the rat, Type I reeptors exhibit the higher affinity for ortiosterone (CORT, the naturally ourring endogenous gluoortioid in the rodent) and are probably tonially ativated even under onditions of low plasma titers of CORT. In ontrast, Type IT reeptors exhibit a muh lower affinity for CORT and, as suh, appear to be ativated mostly under onditions of high irulating CORT (e.g., diurnal peaks, stress; Dallman et al., 1987; de Kloet, We gratefully aknowledge the exellent tehnial assistane of B. Dingwall, R. Phillips, and W. van der Vliet. We thank B. Christie for reviewing this manusript. We also thank Roussel-Ulaf, Romainville, Frane for kindly providing us with RU This study was supported by the New Zealand Health Researh Counil and the Human Frontier Siene Program. D. S. K. was supported by a New Zealand Health Researh Counil Postdotoral Fellowship. Correspondene should be sent to D. S. Kerr, Department of Psyhology, University of Otago, P. O. Box 56, Dunedin, New Zealand. Ratka, Reul, Sutanto, & van Eekelen, 1987; Reul, van den Bosh, & de Kloet, 1987). It appears that the Type I and Type IT reeptors are omponents of two funtionally distint systems that mediate differing metaboli and physiologial responses within the hippoampus (see 10els & de Kloet, 1992a, 1992b; Shlatter & Dokas, 1989). Beause of the widely varying ations of ortiosteroids on hippoampal funtion, and beause of the prominent role of the hippoampus in memory and learning proesses, a number of studies have been arried out in reent years to address the question of modulation of longterm potentiation (L TP) under varying onditions of stress or adrenal ativity. Early studies demonstrated that LTP in the dentate gyrus is signifiantly diminished following adrenaletomy but an be restored by in vivo or in vitro administration ofcort (Dana et al., 1982; Dana & Martinez, 1984; Nowiky, Vardaris, & Teyler, 1983). In ontrast, a number of investigators have demonstrated a signifiant impairment of LTP among animals exposed to CORT (Pavlides, Watanabe, & MEwen, 1993) or to a variety of stressful onditions, in omparison with unstressed ontrols (Diamond, Bennett, Stevens, Wilson, & Rose, 1990; Foy, Stanton, Levine, & Thompson, 1987; Shors, Seib, Levine, & Thompson, 1989). Similarly, several studies have indiated that surgial adrenaletomy (ADX) an prevent stress-related redutions in LTP (Di- 123 Copyright 1994 Psyhonomi Soiety, In.

2 124 KERR, HUGGETT, AND ABRAHAM amond, Bennett, Engstrom, Fleshner, & Rose, 1989; Shors, Levine, & Thompson, 1988). Reently, in vivo studies of hippoampal region CA 1 (Bennett, Diamond, Fleshner, & Rose, 1991; Diamond, Bennet, Fleshner, & Rose, 1992) have shown that the magnitude of primed-burst long-term potentiation (PB L TP) varies with serum CORT levels aording to an inverted-u shaped funtion, suh that very low and very high CORT levels are assoiated with diminished PB LTP, whereas midrange serum CORT titers (presumably suffiient to ativate the Type I reeptor only) are assoiated with a robust expression of PB-LTP. A reent report by Pavlides, Watanabe, Margarinos, and MEwen (1992) has indiated a strikingly similar pattern of LTP modulation by CORT in the dentate gyrus. In the latter study, surgially adrenaletomized (ADX) rats exhibited robust LTP in vivo following the administration of the Type I CSR agonist aldosterone, while exhibiting diminished LTP following administration of the Type IT CSR agonist RU or vehile (Pavlides et al., 1992). Although LTP has been reported to vary with serum CORT titers, there is still little atual evidene that gluoortioids (alone) are involved in the various reported effets. That is, many of the studies have utilized surgial ADX or stress (or both) to produe variations in serum CORT, without providing suffiient ontrol of ADX- and stress-indued variations in adrenal ateholamines, adrenal opioids, and ACTH. The latter fators have also been impliated in the modulation of LTP (Gold, Delanoy, & Merrin, 1984; Shors, Levine, & Thompson, 199Oa, 199Ob) and memory and learning proesses (de Wied, 1989; Gold & van Buskirk, 1975, 1976), and they should be onsidered as possible experimental onfounds (or ofators; f. Axelrod & Reisine, 1984) in at least some of the gluoortioid studies ited above. In addition, most of the onlusions to date regarding LTP have been based on extraellular measures ofpopulation spike amplitude only, and more diret measures of altered synapti effiay by assessment of the field EPSP have been largely negleted. Moreover, although a homosynapti primed-burst long-term depression (PB-LTD) was reently observed in very high CORT animals (Diamond, Bennett, et al., 1992), neither this nor other forms of synapti plastiity suh as heterosynapti LTD have been systematially examined. Thus, a number of questions regarding the modulation of LTP and LTD by stress or ADX remain unresolved. In the present experiments, adrenal-intat adult rats were administered either the ortiosteroid biosynthesis inhibitor metyrapone or ortiosterone as a means of diretly manipulating serum CORT levels in vivo. LTP and heterosynapti LTD were then assessed in vitro, utilizing measures of both the population spike and the field EPSP to assess hanges in synapti effiay and gross hippoampal funtion (preliminary results of this study have appeared in abstrat form: Kerr, Huggett, & Abraham, 1993). Consistent with reent reports (Bennett et al., 1991; Diamond, Bennett, et al., 1992; Pavlides et ai., 1992), our results indiate a strong inverted-u shaped relationship between serum CORT level and the magnitude of LTP. In addition, modulation of heterosynapti LTD also onforms to an inverted-u shaped funtion. For both LTP and LTD, alterations in the slope of the field EPSP lend support to the hypothesis that the hanges observed are due to diret effets on synapti effiay, as opposed to modulation of loal inhibition or neuronal exitability. METHOD Hippoampal slies (400-J.Im transverse setions) were prepared from 3- to 4-month-old, male Sprague-Dawley rats and were maintained fully submerged in a standard brain slie hamber. The slies were onstantly perfused with an artifiial erebrospinal fluid onsisting of mm NaCI, 3.2 mm KCI, 1.25 mm NaH,PO., 26.0 mm NaHC03, 2.0 mm CaC!" 2.0 mm MgC!" 10.0 mm gluose, and 0.5 I'M pirotoxin (33-34 C, saturated with 95%0,/ 5%CO,). Pirotoxin has been shown to enhane the indution of LTD in the hippoampus (Abraham & Wikens, 1991; Kerr & Abraham, 1993) and was inluded here for that reason. In order to redue potential hyperexitability and the spread of seizure-like ativity during tetanization, region CA3 was removed by a manual knife ut during hippoampal dissetion (Figure 1). This proedure, previously employed by Abraham and Wikens (1991) among others, does not adversely affet the evoked responses normally observed in vitro. CAl field potentials, reorded from stratum pyramidale and stratum radiatum, were assessed in response to monosynapti ativation of separate Shaffer ollateral/ommissural inputs. Monopolar eletrial mirostimulation with 25-J.Im teflon-oated steel wires, applied on opposite sides of the reording site (see Figure 1), yielded robust synapti responses. These mirostimulation proedures have been shown to reliably yield greater than 95 % separation of the ativated onverging afferent inputs, as evidened by summation tests of the individual slopes of the fiber volleys (Kerr & Abraham, 1993). One input served as the onditioning pathway and reeived "theta-burst" tetani stimulation onsisting of eight trains (2-se duration) of brief, l00-hz bursts (five biphasi pulses per burst, 200-J.Ise per pulse) delivered 200 rnse apart, suffiient to produe homosynapti LTP; the other (test) pathway reeived no onurrent stimulation during the onditioning proedure. Baseline potentials were monitored for 15 min prior to onditioning stimulation, and reordings ontinued for at least 30 min following stimulation proedures. Initial stimulation intensities were adjusted to yield moderate (1- to 1.5-mY population spike) responses (Figure 3). Population spike amplitudes (in millivolts, defined as the height of the spike from peak negativity to a point of intersetion with a tangent line along the rising fae of the positive wave; see Figure 1 legend) and field EPSP initial slopes (millivolts/milliseond) were measured offline, using ASYST-based software written in our lab, and were expressed as perent hange from baseline values. Slies not exhibiting stable baseline potentials for at least 10 min prior to onditioning were disarded from the study. Slies were not disarded for reasons of failure to exhibit L TP following onditioning. Student t tests and two-way analyses of variane (ANOY As) were performed with repeated measures designs for the 10-, 15-, and 30-min posttetanization time points in order to assess group differenes following the various treatments. Earlier time points were exluded from these analyses in order to avoid on-

3 CORTICOSTERONE, LTP, AND LTD 125 ( :s 0.6 % ). Under our assay onditions, parallelism tests indiated no interferene by metyrapone or any other endogenous serum omponents, and although not assessed during the present study, intraand interassay oeffiients of variation for this assay generally vary between 4.0%-4.3% and 4.8%-5.8%, respetively. Beause CORT an affet peripheral gluose titers, and beause brain gluose levels have been shown to orrelate with ertain types of learning (Gold, 1986; Lee, Graham, & Gold, 1988), we also assessed serum gluose levels in all animals used in Experiment 1 of the present study. Gluose titers were determined with the use of an automati analyzer and a ommerially prepared diagnosti kit (Gluose HK kit, Rohe In.) utilizing a standard two-step enzymati reation (hexokinase and gluose-6-phosphate dehydrogenase) yielding NADH quantified by measurement of absorbane at 340 nm. Spik (mv) EPSP (mv/mse) Figure 1. Top: Shemati diagram of bippoampal slie. Area CAl was removed during dissetion to redue hyperexitability. ExtraeUuIar reording eletrodes were positioned in stratum pyramidale and stratum radiatum of area CAl; insets show typial evoked responses observed at these sites. Stimulation sites are indiated by the blak dots superimposed over the SdIaIrer oiiaterak:ommi fibers in stratum radiatum, on opposite sides of the reording sites. Bottom: Representative waveforms (population spike ahove, field EPSP below) illustrating methods of waveform analysis. Spike amplitudes (millivolts) were defined as the height of a line extending from the peak negativity to a tangent aross the rising fae of the positive wave (vertial dashed line). Field EPSP slopes (millivoltsl milliseond) were def"med as the rate of hange along the most linear portion of the negative-going fae of the EPSP. tamination of the long-term hanges with the more dramati shortterm depression normally arising after strong tetanization. Statistial signifiane was determined at a onfidene level of p <.05. Blood Colletion, CORT, and Gluose Assays Following anesthetization (ether vapor) and rapid deapitation, trunk bloods were olleted from eah animal. Samples were maintained on ie and allowed to lot for approximately 3 h; then they were entrifuged at 800 x G for 10 min to yield whole-ell and platelet-free serum. Serum ortiosterone was assayed with the use of a ommerially prepared radioimmunoassay (RIA) highly speifi for rat ortiosterone (Coat-a-Count RIA Kit; Diagnosti Produts Corp., Los Angeles, CAl. Minimum detetable levels for this assay are less than 1 JLg/dl, and nonspeifi binding is negligible Experiment 1 Twenty-four rats (average body weight, 328 g; range, g) were divided into four groups: noninjeted intat ontrols (lnt, n = 8), saline-injeted ontrols (SAL, n = 6), and two groups onsisting of animals injeted with either the ortiosteroid biosynthesis inhibitor metyrapone (Sigma Chemial Co., St. Louis; MET, 200 mg/kg/day, s.., n = 5) or ortiosterone (Sigma Chemial Co.; CORT, 20 mglkglday, s.., n = 5). All animals were weighed and rehoused in isolation. Experimental subjets and saline-injeted ontrols were eah injeted twie daily (two half-doses), one at 9:00 a.m. and again at 4:00 p.m. for 2 days prior to sarifie, and one again on the morning of Day 3, approximately 2 h prior to sarifie. This method was adopted to ensure slow release, yielding relatively onstant levels of metyrapone and CORT throughout the treatment period and to ensure that any CORT-related genomi events were at equilibrium well before the in vitro studies. These dosages are within ranges previously reported to produe either relatively CORT -free, pharmaologially adrenaletomized animals (f. Murison, Overmeier, Hellhammer, & Carmona, 1989; Stein & Sapolsky, 1988), or stress-equivalent high-cort animals (f. Sapolsky, Krey, & MEwen, 1985) respetively. Noninjeted and saline-injeted animals were inluded to ontrol for stress effets due to handling and rehousing. Experiment 2 To assess possible diret effets of metyrapone itself (i.e., effets on synapti plastiity not related to inhibition of CORT synthesis), an additional group of animals was injeted with both metyrapone and CORT (MET + CORT, n = 6, dosages as above). In addition, to diretly assess the role of the Type IT gluoortioid reeptor in the modulation of LTP and LTD, a highly speifi Type IT reeptor agonist, RU-28362, was administered to slies by onstant in vitro perfusion (10 I'M in artifiial erebrospinal fluid). These experiments were arried out in metyrapone-injeted animals (MET + RU, n = 2) as well as noninjeted ontrols (INT + RU, n = 4). RESULTS Serum CORT and Gluose Titers Radioimmunoassay of serum CORT revealed that it differed for the four groups assessed in Experiment 1 (Figure 2A; final ns indiated for eah group below.). Salineinjeted and noninjeted ontrols exhibited low to moderate levels of serum CORT, respetively (mean j.tg/dl ± SEM: SAL, 3.8±O.7, n = 6; INT, 7.1±3.8, n = 5). CORT-injeted animals exhibited elevated serum CORT titers ( , n = 4) onsistent with those values observed in autely or hronially stressed animals (Eldridge, Brodish, Klute, & Landfield, 1989; Foy et al.,

4 126 KERR, HUGGETT, AND ABRAHAM b2 o u E ;:3 l V C/) ---- "0 E v (/) o 10 u ;:3 o E 5 ;:3 l- V C/) A B sal int met ort sal int met ort Figure 2. (A) Serum ortiosterone (mean I-'Wdl ± SEM) and (B) senun gluose (mean mmoiesil ± SEM) levels for the four groups in Experiment 1. Final ns for eah group are as follows: SAL, n = 6; INT, n = 5; MET, n = 5; CORT, n = ; Odio & Brodish, 1989). In addition, three MET + CORT animals (Experiment 2) for whih bloods were available exhibited even larger inreases in serum CORT (40.0±6.8). Unexpetedly, metyrapone-treated animals did not exhibit diminished serum CORT levels relative to ontrols. Instead, MET animals in Experiment 1 exhibited a mean CORT level of Itg/dl (n = 5), and bloods taken from the 2 MET + RU animals used in Experiment 2 plus an additional metyrapone-injeted animal not used in this study exhibited similar serum CORT levels ( ). The range of values observed aross all 8 metyraponeinjeted animals was large ( Itg/dl), and it is lear that, overall, metyrapone treatment was not an effetive means of pharmaologial adrenaletomy in these animals. Student t tests revealed signifiant differenes in CORT titers between the saline- and metyrapone-injeted groups, and between the saline- and CORT-injeted groups (p <.05). Serum gluose levels did not differ aross groups and are summarized as follows (gluose, mean millimoles/l ± SEM: SAL, 1O.65±0.34, n = 6; INT, 9.68±0.20, n = 5; MET, 1O.36±0.1O, n = 5; CORT, 11.18±1.17, n = 4). These data are shown in Figure 2B. Regression analyses indiated no signifiant orrelation between serum CORT and serum gluose levels. LTP and LTD: Experiment 1 LTP and LTD were assessed in several slies from eah animal, and group means were based on the total number of stable slies per group (total slies: SAL, n = 16; INT, n = 17; MET, n = 14; CORT, n = 15). All data summarized below are uniformly expressed as mean perent hange from the baseline level (±SEM), at 30 min posttetanization. Signifiant differenes in both LTP and LTD were observed between onditions (Figure 3). Saline-injeted (low serum CORT) and CORT -injeted (high serum CORT) animals exhibited only modest inreases in population spike amplitude homosynaptially (SAL, 58.6%±6.8%; CORT, 57.1%±13.8%) and dereases heterosynaptially (SAL, -24.9%±4.6%; CORT, % ± 9.4 %) following homosynapti onditioning. Intat (noninjeted) and metyrapone-injeted animals (midrange serum CORT levels) exhibited greater LTP and LTD of the spike following tetanization (lnt, LTP, 91.7%±1O.6%, LTD, -31.7%±8.3%; MET, LTP, 102.1%±11.4%, LTD, -55.0%±6.3%). An analysis of variane (ANOV A) indiated a signifiant main effet of pharmaologial treatment (LTP: F = 5.68,p <.001; LTD: F = 3.97,p <.002), and post ho ttests revealed signifiant differenes between the MET and SAL groups (p ::s.01, both LTP and LTD) and the MET and CORT groups (p ::s.025, both LTP and LTD). EPSP slope measurements onfirmed that the pattern of hanges observed for population spikes were related to hanges in synapti effiay. Saline-injeted and CORTinjeted animals exhibited small perent hanges in EPSP slope (SAL, LTP, 5.8% ± 3.0%; LTD, -1.8% ± 1.9%; CORT, LTP, 1.4%±5.5%, LTD, -2.4%±13.8%) whih were not signifiantly different between groups. Noninjeted and metyrapone-injeted animals generally exhibited onsiderably larger EPSP slope hanges both homosynaptially and heterosynaptially than the salineand CORT-injeted groups (INT, LTP, 4.4%±2.7%; LTD, -7.3%±2.6%; MET, LTP, 17.1 %±2.3%, LTD, -8.1 %±2.5%). An ANOVA indiated a signifiant main effet aross onditions for LTP of the EPSP (F = 2.56, p <.03), but not for heterosynapti LTD of the EPSP (F = 0.94, p >.05). Post ho analyses revealed signifiant differenes between the MET and SAL groups (p =.01) and the MET and CORT groups (p <.03) for LTP of the EPSP. Linear and polynomial regressions were performed to ompare mean level of LTP or LTD (both spike and EPSP) for eah animal versus serum CORT and serum gluose. Seond-order polynomial regressions revealed strong orrelations between serum CORT titer and magnitude ofltp or LTD for the homosynapti spike, heterosynapti spike, and heterosynapti EPSP.

5 CORTICOSTERONE, LTP, AND LTD 127 CORT METYRAPONE -r V -\7= l- -V - baseline LTP 30min baseline LTP 30min 200 LTP 50 LTD,-..,-.. Q,j Q,j CIJ CIJ = :.:.: \,/ \,/ 0 Q,j 100 Q,j 'tl B 'tl ::.":: -25 :a S S -50 -< -< Q,j Q,j s..s Time (minutes post-tetanization) Time (minutes post-tetanization) Figure 3. Top: Population spikes and field EPSPs reorded from representative slies obtained from CORT -treated and metyrapone-treated animals. Eah waveform represents the average of four to five individual, onseutive waveforms during the 1 min prior to tetanization (baseline) or at 30 min posttetanization Oong-term potentiation, LTP). Double-spiking following tetanization is routinely observed in pirotoxin (see Method setion). Calibration bars apply to all waveforms shown. Bottom: LTP and long-term depression (LTD) for the four groups in Experiment 1. Values shown are population spike amplitude (mean perent hange from baseline ± SEM). Conditioning stimulation (see Method setion) was administered at time = zero (arrow). Dashed line represents baseline levels measured prior to onditioning stimulation. Filled irles, CORT -treated animals; open irles, intat (noninjeted) animals; rilled triangles, metyrapone-treated animals; open triangles, saiine-injeted animals. In addition, regression analyses were performed to establish urves of best fit desribing group mean population spike and EPSP slope hanges versus group mean serum CORT titers (Figures 4 and 5). Again, seond-order polynomial funtions were found to provide good desriptions of plastiity versus hormone level for eah of the four onditions (spike LTP, r = 0.987; spike LTD, r = 0.960; EPSP LTP, r = 0.850; EPSP LTD, r = 0.956). No signifiant orrelations were observed between any of the measures of synapti plastiity and serum gluose levels. As shown in Figures 4 and 5, the CORT dependeny of both LTP and LTD appears to be well desribed by inverted-u shaped funtions. Aordingly, maximal synapti plastiity, evidened by L TP or LTD of either the population spike or EPSP slope, was evident at serum CORT levels of approximately Itg/dl. At lower or higher levels of serum CORT, plastiity (as assessed by any of these four measures) was markedly redued. LTP and LTD: Experiment 2 Additional experiments were onduted in order to assess the possible role of side effets of metyrapone on the modulation of synapti plastiity seen in Experiment 1, and to determine whether plastiity ould also be modulated by in vitro perfusion with the Type II gluoortioid reeptor agonist RU Animals injeted with both metyrapone and CORT (n = 17 slies) exhibited redutions in L TP and LTD omparable to those levels seen in animals injeted with CORT alone (spike LTP, 47.0% ±

6 128 KERR. HUGGETT. AND ABRAHAM LTP Spike ISO e 25.. '" e 20 <"J OJ '1:l '" i:!? 15 == 75 OJ Q. 0.. C ,-: -;;; OJ rjl =-.:.: 'Q. 25 to.; =- Vl (I 10 5 LTP EPSP 0 mel me' int <al inl mel orl met mel inl or! RU nu <nl int met or! ort R RU ISO 25 ell C<I. 125 u e 20 t,!i' '" u.. IS "0 == 75 OJ Q E tij <.. OJ rjl 's.. Vl III IS ort (ng/dl) to.; f ort (ng/dl) Figure 4. Long-term potentiation (LTP) of the spike and field EPSP for au groups (both Experiments 1 and 2). Bar graphs (top) indiate mean perent hange from the baseline level ± SEM; line graphs (below) illustrate mean perent hange from baseline ± SEM for the four groups in Experiment 1 versus mean serum CORT titer for eah group. Seond-order regressions (lines of best fit) have been superimposed, illustrating inverted-u shaped modulation funtions for CORT versus LTP. 10.5%; spike LTD, -22.5% ±4.0%; EPSP LTP, 2.7% ± 2.1 %; EPSP LTD, -4.7% ± 1.7%). Comparison of the MET + CORT group versus animals injeted with metyrapone alone revealed signifiant differenes (unpaired t tests, p :S.001) for all measures exept LTD of the EPSP. Slies taken from MET -injeted animals and subsequently exposed to RU (n = 7 slies) exhibited signifiant redutions in LTP (p <.01) and LTD (p <. 001) of the spike in omparison with MET-injeted animals alone. Similarly, noninjeted ontrols exhibited marked redutions in synapti plastiity following in vitro perfusion with RU (n = 8 slies). Student t tests revealed signifiant differenes (p <.01) between this group and the MET group on all measures exept LTD of the EPSP. Posttetani Potentiation and Homosynapti Depression In addition to CORT-dependent differenes in the indution of L TP and LTD, signifiant differenes were also apparent between groups from 1 to 5 min posttetanization (Figure 3). Homosynapti posttetani potentiation (PTP) and heterosynapti posttetani depression (PTD) were redued in both the SAL and the CORT groups relative to MET-injeted animals. An ANOVA indiated a signifiant treatment effet for both LTP and LTD of the spike, as well as LTP of the EPSP, and post ho omparisons revealed signifiant differenes between the SAL and MET (p <.05) and CORT and MET (p <.05) onditions. Again, the pattern of hange aross the four groups onformed to an inverted-u shaped funtion for CORT level versus degree of synapti plastiity. In both Experiments 1 and 2, a number of slies from the low CORT, high CORT, and RU onditions not only failed to exhibit spike LTP homosynaptia1ly following tetanization, but in some ases even exhibited a degree of lasting homosynapti depression. No slies from the intermediate CORT onditions (lnt and MET) failed to exhibit LTP (Table 1). EPSP slope hanges, while relatively small and more variable, showed the same trend. DISCUSSION The results of the present study onfirm a prominent role of ortiosteroids in the modulation of synapti plastiity. Using adrenal-intat animals, and thereby avoiding many of the experimental onfounds assoiated with

7 CORTICOSTERONE, LTP, AND LTD 129 LTD Spike LTD EPSP IIlet luet illt sal illt Illt COI"t COI-t RU RU <; v e -20 v "0-40. Q. E -«-60 v.:.: 'Q. C/l <;. <::. v -5 <; 0.. v -20 Ql "'0. Q. -40 E -«-60 Ql.:.: 'Q. C/l 0 <;. ('I.. v -5 " Q. 0 <ii -10 Q., C/l Q., ;.;.J o. 1 (ug/d\) orl (ug/d\) Figure 5. Long-term depression (LTD) of the spike and field EPSP for all groups (both Experiments 1 and 2). All data and figure elements are as in Figure 5. stress or surgial ADX, we have shown that diret in vivo manipulations of serum CORT an affet both long-term potentiation as well as heterosynapti long-term depression. Animals with either low or high serum CORT levels exhibited signifiantly less LTP and LTD than did animals with midrange CORT titers. A onsistent inverted U shaped funtion for CORT versus LTP or LTD was observed in vitro for both the population spike and the field EPSP. The fat that EPSP hanges mirrored those ourring for the population spike suggests that ortiosteroids modulate plastiity diretly at the level of the synapse. The modulation of synapti plastiity by our in vivo CORT manipulations most likely ourred as a diret result of CSR ativation, as opposed to indiret CORT effets. First, seletive ativation of the Type n gluoortioid reeptor by in vitro appliation of RU produed signifiant redutions in LTP and LTD. Seond, although serum gluose levels have been shown to orrelate with memory proesses in an inverted-u relationship (Gold & van Buskirk, 1975; Lee et al., 1988), in the present study serum gluose levels did not vary between groups and no relationship was observed between serum CORT and gluose, or betweeen serum gluose and either LTP or LTD. These findings are similar to those of reent eletrophysiologial studies onduted by Bennett et al. (1991) and Diamond, Bennett, et al. (1992) in hippoampal region CAl, and by Pavlides et al. (1992) in the dentate gyrus, in whih inverted-u shaped funtions were observed for CORT (or speifi CSR ativation) versus LTP. In the Table 1 Perent of Slies Either Failing to Exhibit Long-Term Potentiation or Exhibiting Homosynapti Depression of the Population Spike MET MET INT SAL INT MET CORT CORT RU RU No LTP ( %) Homosynapti LTD «100%) Note-LTP, long-term potentiation; LTD, long-term depression; SAL, saiine-injeted ontrols; INT, noninjeted intat ontrols; MET, injeted with metyrapone; CORT, injeted with ortiosterone; RU, Type II reeptor agonist RU-28362, administered in MET or INT animals,

8 130 KERR, HUGGETT, AND ABRAHAM studies by Bennett et al. (1991) and Diamond, Bennett, et al. (1992) using urethane-anesthetized ADX and intat rats (exhibiting varying levels of urethane-indued CORT seretion), primed burst potentiation (PB-LTP) was redued in low CORT and high CORT animals relative to animals exhibiting midrange (9-20 p.g/dl) serum CORT titers. In addition, primed burst homosynapti depression (PB-LTD) was routinely observed in animals manifesting very high CORT levels (50-90 p.g/dl). Similarly, in a preliminary report by Pavlides et al. (1992), LTP of perforant path-dentate granule ell responses was dependent upon speifi CSR ativation. In these studies, dentate LTP was redued in surgially ADX animals and in ADX animals injeted with the Type II gluoortioid reeptor agonist RU-28362, whereas ADX animals administered the Type I reeptor agonist aldosterone exhibited signifiantly greater LTP than did animals in either the ADX or the RU onditions (Pavlides et al., 1992). Our findings onfirm and extend previous reports by showing that, in addition to homosynapti LTP, heterosynapti LTD in area CAl also varies with serum CORT level and, presumably, the relative degree of CSR ativation. In the present studies, metyrapone-injeted animals (midrange CORT levels) onsistently exhibited signifiantly more heterosynapti LTD in vitro than did either low- or high-cort subjets. These findings may be of partiular interest in light of the fat that many urrent models of Hebbian-type plastiity now inlude synapti depression as an important theoretial omponent of memory and learning proesses (see, e.g., Brown, Kairiss, & Keenan, 1990). Sine LTD has been shown to be a major aspet of funtional remodeling of other brain areas, suh as the erebellum (Ekerot & Kano, 1985; Sakurai, 1990) and visual ortex (Aroniadou & Teyler, 1991; Artola, Broher, & Singer, 1990), it is possible that ortiosteroid modulation of synapti plastiity may aount for stressrelated impairments of basi signal proessing in these and other brain regions (Daw, Sato, Fox, Carmihael, & Gingerih, 1991; Fehm-Wolfsdorf, Nagel, & Fehm, 1991). While metyrapone has been widely employed as an effetive means of reduing serum CORT levels autely in animals and hronially in humans suffering ardrenoortial hyperseretion (e.g., Cushing's syndrome; see Haynes, 1990), the blokade ofcort biosynthesis was not omplete in the present studies. Metyrapone-injeted animals atually exhibited higher CORT titers than did either the saline-injeted or the uninjeted ontrols. This result, while unexpeted, proved fortuitous in that the robust expression of LTP and LTD assoiated with midrange CORT titers may have gone unobserved had the drug fully redued serum CORT levels in this group of animals. The fat that the drug treatment was ineffetive remains without explanation. We administered metyrapone in suffiient doses (Murison et al., 1989; Stein & Sapolsky, 1988) subutaneously to ensure a steady release and presumably onstant inhibition of CORT synthesis for the 2 days prior to sarifie. However, hroni metyrapone treatment has been assoiated with lasting elevations of adrenoortiotropi hormone (ACTH) apparently beause of redued negative feedbak by CORT on the hypothalamipituitary axis (Plotsky & Sawhenko, 1987; Rose, Turner, Ray, & Rawashdeh, 1988), and it may be that inreased ACTH stimulation of the adrenal ortex is suffiient to ounterat the inhibition of CORT synthesis by the drug. Thus, although metyrapone may be linially effetive in ases of adrenal hyperseretion, it does not appear to be fully effetive in animals with normal (i.e., unperturbed) hypothalarni-pituitary-adrenal axes (see, e.g., Freo, Holloway, Kalogeras, Rapoport, & Sonrant, 1992; Murison et al., 1989). These findings may be of relevane in future investigations of pharmaologially adrenaletomized animals. It is possible that group differenes in ACTH levels aount for the relative differenes in LTP and LTD observed in this study. Although serum ACTH was not assessed in these studies, given the fat that in vitro administration of RU to slies taken from METinjeted animals and noninjeted ontrols resulted in marked redutions in LTP and LTD, we believe that the modulation of synapti plastiity observed here is due to ativation of Type I and/or Type II gluoortioid reeptors by CORT (or RU-28362) alone as opposed to any modulatory influenes of ACTH. The effet of RU observed here is onsistent with preliminary reports demonstrating (1) a dose-dependent suppression of PB-LTP by the Type II reeptor agonist dexamethasone in urethaneanesthetized ADX rats (presumably exhibiting high ACTH) in vivo (Diamond, Branh, Bennett, Fleshner, & Rose, 1992), and (2) a Type II reeptor dependent suppression of LTP in the dentate gyrus of ADX rats (Pavlides et al., 1992). We also onsidered the possibility that the robust expression ofltp and LTD seen in the METinjeted group was due to a side effet of the metyrapone itself. However, animals injeted with both MET and CORT exhibited signifiant redutions in both forms of synapti plastiity, suggesting either that MET does not exert non-cort related modulatory effets on LTP/LTD, or that the o-administration of CORT is suffiient to override or reverse this modulation. Cortiosteroid reeptor up- and down-regulation have been demonstrated following various long-term manipulations (Reul et al., 1987), but it does not seem likely that the group differenes seen in this study are due to individual differenes in reeptor apaity. In general, CSR up-regulation in young adult rats does not beome apparent before 3-7 days of ADX (Eldridge, Fleenor, Kerr, Campbell, & Landfield, 1989), and signifiant downregulation of CSRs has not been observed earlier than 1 week or longer following exogenous CORT administration (Meaney, Sapolsky, & MEwen, 1985) or hroni stress (Eldridge, Brodish, et al., 1989). Our manipulations took plae over a period of 2 days (sarifie on the morning of the 3rd day), and presumably suffiient time had not elapsed for MR or GR autoregulation. The finding that RU produed redutions in LTP/LTD suggests that Type I reeptor oupany faili-

9 CORTICOSTERONE, LTP, AND LTD 131 tates synapti plastiity, but that Type IT reeptor oupany is inhibitory. This idea is onsistent with the findings of loels and de Kloet that Type I and Type IT reeptor ativation produes opposite effets on ell exitability. Type I reeptor ativation produes inreases in ell exitability, as is evidened by measures of spike aomodation and suppression of 5-HT indued hyperpolarizations (Joels & de Kloet, 1990; loels, Hesen, & de Kloet, 1991), whereas Type IT ativation produes onsistent dereases in ell exitability, whih are apparently due to prolongation of the Ca 1 + -dependent afterhyperpolarization (AHPa; loels & de Kloet, 1989, 1990; Kerr, Campbell, Hao, & Landfield, 1989). Speifi mehanisms underlying the modulation of LTP/LTD by CORT have not been identified. However, beause of the fat that the Type I and Type IT reeptors exert opposing ations on CAl AHPs, Bennett et al. (1991) and Diamond, Bennett, et al. (1992) have suggested that the group differenes in PB-LTP that they observed may be related to differenes in the effetiveness of the primed burst onditioning. That is, CORT enhanement of the AHPa that follows the priming pulse may have shunted or offset the depolarization normally produed during the high-frequeny burst delivered afterward. The same reasoning may also apply in our use of theta-burst onditioning in the present study. Thus, by differentially regulating ell exitability, CORT ould be indiretly affeting the voltage-dependent ativation of NMDA reeptor/hannels, whih are known to be important for both LTP and heterosynapti LTD (Abraham & Wikens, 1991; Christie & Abraham, 1992). As seletive Type IT CSR ativation has been shown to modulate Cal+ influx through both N- and L-type Cal+ hannels (Kerr, Campbell, Thibault, & Landfield, 1992), the hypothesis that ortiosteroids modulate L TP and LTD through a Ca 1 + -dependent proess seems attrative and ertainly merits further attention. For example, if that is true, then it is possible that the appliation of speifi Ca 1 + hannel blokers might prevent CORT-related redutions in synapti plastiity. Regardless of the preise mehanisms underlying these effets, the present findings support the ontention that gluoortioids alone an modulate hippoampal synapti plastiity, and that this modulation may depend on the degree of Type I or Type IT CSR ativation. Our data are onsistent with reent reports demonstrating an inverted-u shaped funtion for CORT versus LTP in the dentate gyrus (Pavlides et al., 1992) and area CAl (Bennett et al., 1991; Diamond, Bennett, et al., 1992), and they further demonstrate a similar pattern of ortiosteroid modulation of heterosynapti LTD. Colletively, these findings may be relevant to previous reports of LTP variation during the iradian rhythm (Dana & Martinez, 1984) or following ADX versus aute stress (Shors et al., 199Oa), and may further relate to reent behavioral evidene of differential roles of the hippoampal MR and GR reeptor systems in memory and learning (Bodnoff, Humphreys, Rose, & Meaney, 1992; Oitzl & de Kloet, 1992a, 1992b). When onsidered together with the inverted-u shaped learning funtions previously demonstrated for peripheral epinephrine (Gold & van Buskirk, 1975) and entral gluose (Lee et al., 1988), the relatively slow and lingering ations of ortiosteroids may well at in onert with the more immediate adrenal ateholamines to failitate the aquisition and onsolidation of information in a stressful environment. REFERENCES ABRAHAM, W. C., & WICKENS, J. R. (1991). Heterosynapti long-term depression is failitated by blokade of inhibition in area CA 1 of the hippoampus. Brain Researh, 546, ARONIADOU, V. A., & TEYLER, T. J. (1991). The role ofnmda reeptors in long-term potentiation (L TP) and depression (LTD) in rat visual ortex. Brain Researh, 562, ARTOLA, A., BROCHER, S., & SINGER, W. (1990). Different voltagedependent thresholds for induing long-term depression and long-term potentiation in slies of rat visual ortex. Nature, 347, AXELROD, J., & REISINE, T. D. (1984). Stress hormones: The interation and regulation. Siene, 224, BENNETT, M. C., DIAMOND, D. M., FLESHNER, M., & ROSE, G. M. (1991). Serum ortiosterone level predits the magnitude of hippoampal primed burst potentiation and depression in urethaneanesthetized rats. Psyhobiology, 19, BODNOFF, S. R., HUMPHREYS, A., ROSE, G. M., & MEANEY, M. J. (1992). Effets of ortiosterone and stress on spatial learning, hippoampal plastiity, and neuron loss. Soiety for Neurosiene Abstrats, 18,534. BROWN, T. H., KAIRISS, E. W., & KEENAN, C. L. (1990). Hebbian synapses: Biophysial mehanisms and algorithms. Annual Review of Neurosiene, 13, CHRISTIE, B. R., & ABRAHAM, W. C. (1992). NMDA-dependentheterosynapti long-term depression in the dentate gyrus of anaesthetized rats. Synapse, 10, 1-6. DALLMAN, M. F., AKANA, S. F., JACOBSON, L., LEVIN, N., CASCIO, C. S., & SHINSAKO, J. (1987). Charaterization of ortiosterone feedbak regulation of ACTH seretion. In W. F. Ganong, M. F. Dallman, J. L. Roberts (Eds.), The hypothalami-pituitary-adrenal axis revisited (Annals of the New York Aademy of Sienes, Vol. 512, pp ). New York: New York Aademy of Sienes. DANA, R. C., GERREN, A., STERNBERG, D. B., MARTINEZ, J. L., HALL, J., STANSBURY, N. A., & WEINBERGER, N. M. (1982). Long-term potentiation is impaired by adrenaletomy and restored by ortiosterone. Soiety for Neurosiene Abstrats, 8, 316. DANA, R. C., & MARTINEZ, J. L. (1984). Effet of adrenaletomy on the iradian rhythm of LTP. Brain Researh, 308, DAW, N. W., SATO, H., Fox, K., CARMICHAEL, T., & GINGERICH, R. (1991). Cortisol redues plastiity in the kitten visual ortex. Journal of Neurobiology, 22, DE KWET, E. R., RATKA, A., REUL, J. M. H. M., SUTANTO, W., & VAN EEKELEN, J. A. M. (1987). Cortiosteroid reeptor types in brain: regulation and putative funtion. In In W. F. Ganong, M. F. Dallman, J. L. Roberts (Eds.), The hypothalami-pituitary-adrenal axis revisited (Annals of the New York Aademy of Sienes, Vol. 512, pp ). New York: New York Aademy of Sienes. DE WIED, D. (1989). Neuroendorine aspets oflearning and memory proesses. News in Physiologial Sienes, 4, DIAMOND, D. M., BENNETT, M. C., ENGSTROM, D. A., FLESHNER, M., & ROSE, G. M. (1989). Adrenaletomy redues the threshold for hippoampal primed burst potentiation in the anesthetized rat. Brain Researh, 492, DIAMOND, D. M., BENNETT, M. C., FLESHNER, M., & ROSE, G. M. (1992). Inverted-U relationship between the level of peripheral orti-

10 132 KERR, HUGGETT, AND ABRAHAM osterone and the magnitude of hippoampal primed burst potentiation. Hippoampus, 2, DIAMOND, D. M., BENNETT, M.., STEVENS, K. E., WILSON, R. L., & ROSE, G. M. (1990). Exposure to a novel environment interferes with the indution of hippoampal primed burst potentiation in the behaving rat. Psyhobiology, 18, DIAMOND, D. M., BRANCH, B., BENNETT, M. C., FLESHNER, M., & ROSE, G. M. (1992). Dexamethasone bloks the indution of hippoampal primed burst potentiation. Soiety for Neurosiene Abstrats, 18, EKEROT, C.-F., & KANO, M. (1985). Long-term depression of parallel fibre synapses following stimulation of limbing fibres. Brain Researh, 342, ELDRIDGE, J. C., BRODISH, A., KLUTE, T. E., & LANDFIELD, P. W. (1989). Apparent age-related resistane of type IT hippoampal ortiosteroid reeptors to down-regulation during hroni esape training. Journal of Neurosiene, 9, ELDRIDGE, J. C., FLEENOR, D. G., KERR, D. S., CAMPBELL, L. W., & LANDFIELD, P. W. (1989). Impaired up-regulation of type IT ortiosteroid reeptors in hippoampus of aged rats. Brain Researh, 478, FEHM-WOLFSDORF, G., NAGEL, D., & FEHM, H. L. (1991). Stress indued ortisol seretion influenes information proessing in man. Soiety for Neurosiene Abstrats, 17, Foy, M. R., STANTON, M. E., LEVINE, S., & THOMPSON, R. F. (1987). Behavioral stress impairs long-term potentiation in rodent hippoampus. Behavioral & Neural Biology, 48, FREO, U., HOLLOWAY, H. W., KALOGERAS, K., RAPOPORT, S. I., & SONCRANT, T. T.(I992). Adrenaletomy or metyrapone-pretreatment abolishes erebral metaboli responses to the serotonin agonist 1-(2,5- dimethoxy-4-iodophenyl)-2-arninopropane (DOl) in the hippoampus. Brain Researh, 586, FUXE, K., WIKSTROM, A.-C., OKRET, S., AGNATI, L. F., HARF STRAND, A., Yu, Z. Y., GRANHOLM, L., ZOLI, M., VALE, W., & GUSTAFSSON, J.-A. (1985). Mapping of gluoortioid reeptor immunoreative neurons in the rat tel- and dienephalon using a monolonal antibody against rat liver gluoortioid reeptor. Endorinology, 117, GoLD, P. E. (1986). Gluose modulation of memory storage. Behavioral & Neural Biology, 45, GOLD, P. E., DELANOY, R. L., & MERRIN, J. (1984). Modulation of long-term potentiation by peripherally administered amphetamine and epinephrine. Brain Researh, 305, GOLD, P. E., & VAN BUSKIRK, R. (1975). Failitation of time -dependent memory proesses with posttrial epinephrine injetions. Behavioral Biology, 13, GOLD, P. E., & VAN BUSKIRK, R. (1976). Enhanement and impairment of memory proesses with post-trial injetions of adrenoortiotropi hormone. Behavioral Biology, 16, HAYNES, R. C. (1990). Adrenoortiotropi hormone: Adrenoortial steroids and their syntheti analogs; inhibitors of the synthesis and ations of adrenoortial hormones. In A. G. Goodman, T. W. Rail, A. S. Nies, & P. Taylor (Eds.), Goodman and Gilman's The pharmaologial basis of therapeutis (8th ed., pp ). Elmsford, NY: Pergamon Press. JOELS, M., & DE KLOET, E. R. (1989). Effets of gluoortioids and norepinephrine in the hippoampus. Siene, 245, JOELS, M., & DE KLOET, E. R. (1990). Mineraloortioid reeptormediated hanges in membrane properties of rat CAl pyramidal neurons in vitro. Proeedings of the National Aademy of Sienes, 87, JOELS, M., & DE KLOET, E. R. (1992a). Control of neuronal exitability by ortiosteroid hormones. Trends in Neurosiene, 15, JOELS, M., & DE KLOET, E. R. (1992b). Coordinative mineraloortioid and gluoortioid reeptor-mediated ontrol of responses to serotonin in rat hippoampus. Neuroendorinology, 55, JOELS, M., HESEN, W., & DE KLOET, E. R. (1991). Mineraloortioid hormones suppress serotonin-indued hyperpolarization of rat hippoampal CAl neurons. Journal of Neurosiene, 11, KERR, D. S., & ABRAHAM, W. C. (1993). A omparison of assoiative and non-assoiative onditioning proedures in the indution of LTD in CAl of the hippoampus. Synapse, 14, KERR, D. S., CAMPBELL, L. W., HAO, S-Y., & LANDFIELD, P. W. (1989). Cortiosteroid modulation of hippoampal potentials: inreased effet with aging. Siene, 245, KERR, D. S., CAMPBELL, L. W., THIBAULT, 0., & LANDFIELD, P. W. (1992). Hippoampal gluoortioid reeptor ativation enhanes voltage-dependent alium ondutane: Relevane to brain aging. Proeedings of the National Aademy of Sienes, 89, KERR, D. S., HUGGETT, A. M., & ABRAHAM, W. C. (1993). Gluoortioid modulation of synapti plastiity in area CA 1 of the hippoampus. International Journal of Neurosiene, 71, LEE, M. K., GRAHAM, S. N., & GOLD, P. E. (1988). Memory enhanement with posttraining intraventriular gluose injetions in rats. Behavioral Neurosiene, 102, MEANEY, M. J., SAPOLSKY, R. M., & MEWEN, B. S. (1985). The development of the gluoortioid reeptor system in the rat limbi brain: I. Ontogeny and autoregulation. Developmental Brain Researh, 18, MEWEN, B. S., WALLACH, G., & MAGNUS, C. (1974). Cortiosterone binding to hippoampus: Immediate and delayed influene of the absene of adrenal seretion. Brain Researh, 70, MEWEN, B. S., WEISS, J. M., & SHWARTZ, C. S. (1969). Uptake of ortiosterone by rat brain and its onentration by ertain limbi strutures. Brain Researh, 16, MURISON, R., OVERMEIER, J. B., HELLHAMMER, D. H., & CAR MONA, M. (1989). Hypothalamo-pituitary-adrenal manipulations and stress ulerations in rats. Psyhoneuroendorinology, 14, NOWICKY, A. V., VARDARIS, R. M., & TEYLER, T. J. (1983). Cortiosterone and long-term potentiation in the hippoampus. Soiety for Neurosiene Abstrats, 9, ODIO, M., & BRODISH, A. (1989). Age-related adaptation of pituitaryadrenoortial responses to stress. Neuroendorinology, 49, OITZL, M. S., & DE KLOET, E. R. (1992a). Cortiosterone modulates exploration via entral mineraloortioid reeptors. Soiety for Neurosiene Abstrats, 18, 718. OITZL, M. S., & DE KLOET, E. R. (1992b). Seletive ortiosteroid antagonists modulate speifi aspets of spatial orientation learning. Behavioral Neurosiene, 106, PAVLIDES, C., WATANABE, Y., MARGARINOS, A. M., & MEWEN, B. S. (1992). Effets of gluoortioid Type I and Type IT agonists on hippoampal long-term potentiation. Soiety for Neurosiene Abstrats, 18, 343. PAVLIDES, C., WATANABE, Y., & MEWEN, B. S. (1993). Effets of gluoortioids on hippoampal long-term potentiation. Hippoampus, 3, PLOTSKY, P. M., & SAWCHENKO, P. E. (1987). Hypophysial-portal plasma levels, median eminene ontent, and immunohistohemial staining of ortiotropin-releasing fator, arginine vasopressin, and oxytoin after pharmaologial adrenaletomy. Endorinology, 120, REUL, J. M. H. M., & DE KLOET, E. R. (1986). Anatomial resolution of two types of ortiosterone reeptor sites in rat brain with in vitro autoradiography and omputerized image analysis. Journal of Steroid Biohemistry, 24, REUL, J. M. H. M., VAN DEN BoSCH, F. R., & DE KLOET, E. R. (1987). Differential response of Type I and Type IT ortiosteroid reeptors to hanges in plasma steroid level and iradian rhythmiity. Neuroendorinology, 45, ROSE, J. C., TURNER, C. S., RAY, D., & RAWASHDEH, N. (1988). Evidene that ortisol inhibits basal adrenoortiotropin seretion in the sheep fetus by 0.70 gestation. Endorinology, 123, SAKURAI, M. (1990). Calium is an intraellular mediator of the limbing fiber in indution of erebellar long-term depression. Proeedings of the National Aademy of Sienes, 87, SAPOLSKY, R. M., KREY, L. C., & MEWEN, B. S. (1985). Prolonged gluoortioid exposure redues hippoampal neuron number: Impliations for aging. Journal of Neurosiene,S,

11 CORTICOSTERONE, LTP, AND LTD 133 SCHLATTER, L. K., & DOKAS, L. A. (1989). Reeptor speifiity of a gluoortioid- and stress-indued hippoampal protein. Journal of Neurosiene, 9, l40. SHORS, T. J., LEVINE, S., & THOMPSON, R. F. (1988). Effet of stress and adrenaletomy on long-term potentiation in rat hippoampus. Soiety for Neurosiene Abstrats, 14, 443. SHORS, T. J., LEVINE, S., & THOMPSON, R. F. (l990a). Effet of adrenaletomy and demedullation on the stress-indued impairment of long-term potentiation. Neuroendorinology, 51, SHORS, T. J., LEVINE, S., & THOMPSON, R. F. (199Ob). Opioidantagonist eliminates the stress-indued impairment of long-term potentiation (LTP). Brain Researh, 506, SHORS, T. J., SEIB, T. B., LEVINE, S., & THOMPSON, R. F. (1989). Inesapable versus esapable shok modulates long-term potentiation in the rat hippoampus. Siene, 244, STEIN, B. A., & SAPOLSKY, R. M. (1988). Chemial adrenaletomy redues hippoampal damage indued by kaini aid. Brain Researh, 473, (Manusript reeived May 3, 1993; revision aepted for publiation November 5, 1993.)