- Email: [email protected]
Altered regulation of lesion-induced synaptogenesis by adrenalectomy and corticosterone in young adult rats
EXPERIMENTAL
NEUROLOGY
93,456-470
(1986)
Altered Regulation of Lesion-Induced Synaptogenesis by Adrenalectomy and Cotticosterone in Young Adult Rats STEPHENW. SCHEPP, STEVEN F. HOFF, AND KEVIN J. ANDERSON’ Departments ofAnatomy and Neurology, Sanders-Brown Research Center on Aging, University of Kentucky, Lexington, Kentucky 40536, and Department of Pharmacology, Chicago Medical School, University Health Sciences, North Chicago, Illinois 60064 Received September 30, 1985; revision received April 22, 1986 Quantitative electron microscopy was used to examine the effect of circulating glucocorticoids on the removal of degenerating synapses and the replacement of lost synaptic contacts in young adult rats that follow partial denervation of the hippocampal dentate gyrus. Subjects were adrenalectomized prior to subcutaneous implantation of pellets containing a specified concentration of corticosterone and subsequent unilateral ablation of the entorhinal cortex. Animals maintained at high circulating concentrations of glucocorticoids were significantly retarded in the early phase of degenerating synapse removal and in the rate of synaptic replacement. Subjects maintained at extremely low concentrations of glucocorticoids were also significantly retarded in the early stagesof synapse removal but showed an early replacement of lost synaptic contacts followed by a dramatic decrease in the rate of replacement. By 60 days atIer the lesion both groups of animals showed synapse replacement equivalent to young adult controls while significant amounts of degenerating synapses still remained in the denervated neuropil. The results demonstrate that circulating glucocorticoids can exert a marked influence on lesion-induced synaptic replacement in the hippoCampal dentate gyrus. 8 1986 AC&~~C PIB, 1~.
Abbreviations: ADX-adrenalectomy, ADXCORT-adrenalectomy and corticosterone, CORT-cmticosterone, YA-young adult control. ’ We thank C. W. Cotman for his support and encouragement in these studies. This research was supported by grants from the National Institute of Health NS1698 I, MH19691 and the John D. MacArthur Research Program on Successful Aging. Dr. Anderson’s present address is Department of Psychobiology, University of California, Irvine, CA 927 17. Please address reprint requests to Dr. S. W. Scheff, Sanders-Brown Research Center, Univ. of Kentucky, Lexington, KY 40536. 456 0014-4886186 $3.00 Copyriebt Q 1986 by Academic I’m& Inc. AII rights of mpmdwtion in any form resew&
CORTICOSTERONE
ALTERS
REACTIVE
SYNAPTOGENESIS
457
INTRODUCTION Selective lesions in the central nervous systemare capable of stimulating axonal growth and new synapseformation in the partially denervated regions (7, 8). Several findings suggest that this growth process in the hippocampus might be influenced by glucocorticoids. Neurons in the hippocampus have been shown to possess binding sites for glucocorticoids ( 17, 34,62); corticosteroids can modify the metabolism of hippocampal cells (67); and corticosteroids have been shown to alter the electrical activity of the hippocampus (39,49, 56, 58). Moreover, in certain cells in tissue culture, glucocorticoids can modulate process outgrowth (57). Previous work has shown that peripherally administered glucocorticoids can modify this axonal growth and significantly decreases the response of certain reactive afferent fibers in the hippocampal dentate gyrus of young adult rats ( 11,46-48). Animals maintained at elevated levels of circulating corticosterone (CORT) show significantly less reactive fiber growth than control animals or subjectsmaintained at low circulating levels (47). High levels of circulating CORT have been also shown to disrupt growth of sympathetic fibers from the superior cervical ganglion after a transection of the fimbria ( 1 1), suggesting that steroids influenced not only central neurons but peripheral as well. Finally, glucocorticoids have been found to suppress reactive fiber growth in a dose-dependent fashion. Dexamethasone as well as hydrocortisone has been shown to alter axon sprouting and the suppressive effect can be directly related to the anti-inflammatory properties of the hormones (48). In the present study, as one step in clarifying the effects of glucocorticoids on axodendritic interactions in the hippocampus, quantitative studies were carried out on the replacement of lost synaptic contacts in the denervated zone. At the light microscopic level only changes in axon collateral growth can be monitored; in many cases only one afferent system is analyzed. Our aim was to critically analyze electron microscopically a denervated neuropil in subjects treated with glucocorticoids and ascertain whether or not new synapse formation and removal of degenerating contacts are affected. Because the time course of the removal of degenerating synapses and the replacement of synaptic contacts is well characterized in the hippocampus (25, 26,32,33) and the limbic system has been shown to be sensitive to glucocorticoids (38,39,56,58), this area ofthe central nervous system was considered ideal to test the influence of these steroids on reactive synapse replacement. METHODS Young adult male Sprague-Dawley rats, 90 days of age (3 10 to 340 g), were housed three to a cage and had food and water ad libitum. The experimental animals were subjected to a bilateral adrenalectomy at least 10 days prior
458
SCHEFF, HOFF, AND ANDERSON
to a unilateral entorhinal lesion using methods identical to those described elsewhere (45). At the time of the lesion, each experimental animal was randomly assigned to one of two different groups: group I (ADX) consisted of animals implanted subcutaneously with cholesterol pellets in the posterior cervical region of the neck (N = 17); these animals were subsequently maintained on 0.9% sodium chloride. Group II (ADXCORT) consisted of corticosterone-maintained animals (N = 20). These subjects were subcutaneously implanted with CORT pellets in the posterior cervical region of the neck identical to the procedure used for the ADX subjects and maintained on saline. Animals from both groups were killed at 4 (ADX N = 7; ADXCORT N = 6), 10 (ADX N = 4; ADXCORT N = 6), 30 (ADX N = 3; ADX N = 5), or 60 (ADX N = 3; ADXCORT N = 3) days postlesion. Group III (YA) consisted of young adult control animals 90 days of age (N = 14). These animals received entorhinal cortex lesions identical to those of the other two groups. Animals in this group were killed at 4 (N = 4), 10 (N = 3), 30 (N = 3), or 60 (N = 4) days postlesion, Pellets were formed from molten CORT according to the procedure of Meyer et al. (37). This method produced stable blood and brain CORT concentrations in the range of 28 pg/ 100 ml serum within a very short time and maintained these concentrations during the course of the study. Pellets of cholesterol were formed in a similar fashion and did not alter the concentrations of CORT (less than 0.5 &lOO ml serum) in the ADX animals. At the designated postlesion survival times, the animal was anesthetized with sodium pentobarbital(50 mg/kg) and blood samples were obtained by cardiac puncture. Plasma was then collected and processed in triplicate by radioimmunoassay to determine circulating concentrations of CORT as described elsewhere (11). Subsequently the brain was fixed by vascular perfusion with a solution of 4% paraformaldehyde, 0.5% glutaraldehyde, 0.54% D-ghCOSe, in 0.1 Msodium phosphate buffer (pH 7.4). After a l-h perfusion, the entire animal was placed in a plastic bag containing 10 ml perfusate and refrigerated 24 h. The hippocampi were then carefully dissected using a procedure identical to that described elsewhere to study synaptic density in this brain region (25,26). The hippocampi were sectioned in a coronal plane (175 to 200 pm) with a Sorvall tissue chopper. Isolated portions of the anterior hippocampus (levels 3290 to 4380 of K&rig and Klippel(30), were postfixed in 1% osmium tetroxide in 0.1 M sodium phosphate buffer and stained en bloc with 0.5% many1 acetate (28). The samples were dehydrated in a graded ethanol series and propylene oxide and embedded in Epon-Araldite. Blocks were coded at this time so that all subsequent processing and analyses were carried out blind with respect to treatment group.
CORTICOSTERONE
ALTERS
REACTIVE
SYNAFTOGENESIS
459
One-micrometer-thick sections were cut and stained with thionin for light microscopic study and to facilitate trimming for ultrathin sectioning. These thick sections were used for measurements of the width of the dentate neuropil. Camera lucida drawings were obtained for several sections from each animal and depicted the zone from the granule cell layer to the hippocampal fissure of the dorsal blade of the dentate gyrus. After trimming of the block, ultrathin sections were cut and collected on Formvar-coated slot grids (1 X 2-mm slot size). The sections were stained in an aqueous solution of uranyl acetate for 20 min, rinsed, and stained with Reynolds lead citrate (4 1) for 4 to 5 min. Under the electron microscope the granule cell layer was ascertained in the ultrathin sections. For each animal, a montage, one photographic frame wide, was made of the entire dentate gyrus molecular layer at right angles to the granule cells of the dentate gyrus dorsal leaf. Micrographs were taken at an initial magnification of X7500 and enlarged photographically to X 15,000. The average tissue area analyzed for each montage of the outer two-thirds of the dentate gyrus molecular layer was 1150 pm*. In this study the criteria for identification of a normal synapse were the presence of synaptic vesicles and presynaptic membrane in association with a postsynaptic density, identical to that used previously (24-26,32, 33). Every recognizable synapse in the outer dentate gyrus molecular layer was counted and the mean density of synapses per 100 pm* was determined for each animal. If a presynaptic specialization made a discontinuous connection with the same postsynaptic specialization, it was considered a single contact area. However, if a presynaptic specialization made contact with more than one postsynaptic specialization it was considered a multiple synaptic contact. Degenerating synapses were characterized by either an electron-dense or electron-lucent profile connected with a synaptic junction or an abnormally translucent profile containing a watery cytoplasm connected with a postsynaptic density (9,24,35). Anatomical reconstruction, to assesscompleteness of the lesion, was performed on all animals by sectioning the posterior portion of the brain in the horizontal plane and staining frozen sections with cresyl violet. Sections through the region normally occupied by the entorhinal cortex were drawn from these stained sections to assess extent of the lesions. This procedure confirmed that in all animals reported in this study, the surgical procedures removed both the medial and lateral aspects of the entorhinal cortex. Verification of adrenalectomy was accomplished by autopsy at the time of perfusion. RESULTS
Serum Corticosterone Concentrations. Plasma concentrations in adrenalectomized
animals were determined
of CORT by radioimmunoassay. The
460
SCHEFF, HOFF, AND ANDERSON
mean serum concentration of CORT found in animals implanted with pellets of cholesterol (N = 17) used in this study was 0.2 (kO.2) pg/ 100 ml serum. The mean serum concentration of CORT found in animals implanted with pellets of CORT (N = 20) used in this study was 27.9 (k4.8) &lo0 ml serum. An ANOVA (treatment by days) (66) revealed no significant days effect (F(3,29) = 1.03; P > 0.1) indicating that for each experimental group CORT concentrations remained constant. These values are identical to those reported earlier using a similar technique (11). CORT was not determined for young adult control animals because the process of anesthetizing the animals for blood sampling and EM perfusion would artificially elevate the circulating CORT due to stress (16,40). Experimental animals were not affected by this procedure because their adrenals had been removed. Shrinkage of Dentate Gyrus Molecular Layer. One-micrometer-thick sections stained with thionin revealed no significant changes in molecular layer width as a function of treatment (F(2,39) = 1.28; P > 0.1) or days after entorhinal lesion (F(3,39) = 1.89; P > 0.1). Thus, differences found in this study could not be accounted for by differential shrinkage of the molecular layer. Degeneration The dynamics of degenerating synapse removal from the denervated ipsilateral outer two-thirds of the molecular layer were assessed as a function of days postlesion. Two aspects of the degenerating synapse clearance were analyzed. First, do each of the groups show a significant clearance of degenerating contacts and second, are there any differences in the time course of the clearance? Mean density of degenerating synaptic contacts were subjected to a two-way ANOVA (treatment by days) (66) with subsequent individual group comparisons assessed with the least significant difference test (29). These analyses revealed a significant clearing response (removal of degenerating synapses) with time postlesion for all groups (F(3,39) = 27.85, P < 0.00 1) and a significant difference in the clearing response between groups (F(2,39) = 15.69, P < O.OOl), indicating that circulating concentrations of hormones had an effect. Although all groups showed significant declines in density of degenerating synapses during the survival period studied, the process was not uniform between groups (Fig. 1). The differences in the clearing response can be best appreciated by analyzing total amount of degenerating synapses at various days postlesion. The two experimental groups, ADXCORT (N = 6) and ADX (N = 7) animals, showed a significantly greater accumulation of degenerating synapses (P < 0.0 1) at 4 days postlesion compared with young adult controls (YA, N = 4) at the same survival time. This difference was still apparent by 60 days postlesion (P < 0.05 for ADX (N = 3) and P < 0.01 for
CORTICOSTERONE
ALTERS
REACTIVE
I
7)
(6)
SYNAPTOGENESIS
461
q YAControl
(3)
(4)
(6)
(3)
(3) (5)
(4)
(3) (3)
10 60 4 30 Days Post Entorhinal Lesion FOG. 1. Bar graph showing time course of changes in the density of degenerating contacts in the outer two-thirds of the dentate gyrus molecular layer following ipsilateral removal of the entorhinal cortex. Animals maintained at high concentrations of circulating glucocorticoids (ADXCORT) showed the greatest accumulation of degenerating synaptic contacts throughout the study compared with animals maintained at extremely low concentrations of glucocorticoids (ADX) and young adult controls (YA). The bars represent group means + SD, numbers in parentheses below bars represent number of subjects in that group. * = P < 0.05t = P < 0.01.
ADXCORT (N = 3)). By 10 days postlesion ADX animals (N = 4) had significantly cleared the degenerating contacts from the outer two-thirds of the molecular layer and were no longer reliably different from YA (N = 3) (P > 0.1). ADXCORT subjects (N = 6) however, failed to demonstrate an equivalent clearance response within the same postlesion interval: this group showed a significantly greater accumulation of degenerating synapses (P < 0.05). By 30 days postlesion all groups removed the synapses equally well with the ADXCORT subjects showing the greatest difference between 10 and 30 days survival. ADX and ADXCORT animals, however, failed to show any additional clearing between 30 and 60 days postlesion; this was in contrast to the response observed in YA subjects. Moreover, both experimental groups were significantly different from controls (N = 4) in total amount of remaining degenerating contacts at 60 days postlesion (ADX P < 0.05 and ADXCORT P < 0.0 1). Reinnervation Changes in the density of synaptic contacts in the outer two-thirds of the denervated molecular layer of the dentate gyrus were analyzed using a two-
462
SCHEFF. HOFF, AND ANDERSON
q VAControl 0
ADX
q ADXCORT
(4)
(6)
(3) (2
(6)
(4)
30 10 60 4Days Post Entorhinal Lesion FIG. 2. Bar graph showing time course of changes in the density of normal synaptic contacts in the outer two-thirds of the dentate gyrus molecular layer following ipsilateral removal of the entorhinal cortex. Animals maintained at either extremely high (ADXCORT) or extremely low (ADX) concentrations of circulating glucocorticoids were markedly influenced in synapse replacement compared with young adult controls (YA). The bars represent group means + SD, numbers in parentheses below bars represent number of subjects in that group. * = P < 0.05 *=p
way ANOVA (treatment by days) (66). Subsequent individual group comparisons when appropriate were made utilizing the least significant difference test of significance (29). Those analyses revealed significant differences in the density of synaptic contacts within groups (F(3,39) = 92.2, P < O.OOl), indicating change with time, and between groups (F(2,39) = 4.09, P -c 0.05), indicating that concentrations of hormone had an effect. At 4 days postlesion there were no significant differences between groups (P > 0.1) (see Fig. 2). All groups showed a significant (P < 0.0 1) replacement of lost synaptic contacts by 10 days postlesion compared with the 4-day survival time. This increase in density was not uniform between groups. The ADX animals showed the most dramatic increase in density of intact synapses and were significantly greater compared with both YA (P < 0.01) and ADXCORT subjects (P < 0.0 I), with the latter two groups not differing from each other (P > 0.1). By 30 days postlesion, the ADX animals (N = 3) failed to show any additional significant increase in synaptic density relative to similarly treated animals examined at 10 days (P > 0.1). In fact, the ADX subjects exhibited significantly less synaptic replacement than the YA controls (P < 0.0 1) at 30 days, indicating a subsequent delay in the replacement process. The ADXCORT animals showed a significant increase in synaptic
CORTICOSTERONE
ALTERS
REACTIVE
SYNAPTOGENESIS
463
density at 30 days postlesion compared with similarly treated rats at 10days postlesion (P < O.Ol), but were found to be significantly less dense than YA animals. By 60 days postlesion all groups showed a significant increase in synaptic density and no significant differences were observed between the experimental and control groups (P > 0.1).
Correlation of Synapse Replacement to Removal of Degenerating Contacts. The relationship between the removal of degenerating synapses and the replacement of synapses was assessed by determining the correlation coefficient. The YA controls showed a significant correlation (r = 0.894, d.f. 12, P < 0.00 1) indicating that as dying synapses were removed new synapses were formed. The ADX and ADXCORT animals also showed a significant correlation (r = 0.903, d.f. 15, P < 0.001 and r = 0.782, d.f. 18, P< O.OOl), respectively, in the replacement of synaptic density as a function of degeneration clearance. DISCUSSION We reported elsewhere that the hippocampal formation responds to changes in circulating levels of glucocorticoids by retarding the growth of collateral sprouting following selective damage (11,46-48). The present experiment is an extension of that work and carefully examines the influence of glucocorticoids on the time-dependent replacement of synaptic contacts in the outer two-thirds of the molecular layer of the hippocampal dentate gyrus. This area is massively denervated (85%) by an ipsilateral lesion of the entorhinal cortex (25,32,45). Two different yet related aspects of the replacement process were studied: (a) removal of degenerating synapses and (b) the return of normal synaptic contacts. Degeneration. All groups showed removal of degenerating synapses throughout the duration of the experiment, but the time course of this removal was different depending on circulating hormone concentrations. The most striking differences occur on postlesion days 4 and 10 (Fig. 1). Both ADX and ADXCORT animals showed a significantly greater density of degenerating synapses compared with controls at 4 days after entorhinal cortex removal. Previous experiments reported significant accumulation of degenerating synaptic contacts at 2 days after entorhinal cortex removal in young adult animals, with the 2-day values similar to those observed at 4 days in the ADX and ADXCORT groups (26,32). Even in these experiments there appears to be a rapid removal of dying synapses by 2 days for the total number of dying synapses and intact synapses does not equal the prelesion synaptic density. Thus the clearing response is initiated very early. Either the terminals in the ADX and ADXCORT groups are protected from immediate degradation as suggested previously (2, 13, 21-23) or these animals lack the
464
SCHEFF, HOFF, AND ANDERSON
ability to initiate the early rapid phase of the clearing response which is seen in young adult controls (25). Previous studies with hydrocortisone and corticosterone demonstrated a significant increase in astrocytes in animals maintained with high concentrations of steroid (46,47). That finding is suggestive of a glial influence which may affect the removal of degenerative debris and presumably alter the replacement of lost synaptic contacts. The same type of delay of the clearing response was observed in aged animals which are known to have elevated concentrations of glucocorticoids (26) and normally show hypertrophied astrocytes in the hippocampal formation (3 1, 45). It is possible that glial cell mechanisms involved in phagocytosis are regulated by glucocorticoids. Glucocorticoids have been reported to stabilize lysosomal-related events in both neuronal and nonneuronal systems (3, 14, 19,27, 36, 55, 63-65). The present experimental results suggest that the removal of dying synapses is directly or indirectly slowed by the presence of altered glucocorticoid concentrations. By 10 days after the lesion the ADX animals have begun to show a normal clearing response whereas the ADXCORT animals continue to lag significantly behind. The latter animals continue to show delayed clearing compared with the control group (YA) throughout the remainder of the experiment. Thus, the removal of dying contacts in the elevated CORT animals closely parallels the response observed in aged animals following the same lesion paradigm (26). The failure to detect further reduction in amounts of these contacts after 30 days postlesion in the ADXCORT group might possibly indicate a combination of an abortive protection of synaptic contacts and a clearing response which can keep pace only with low levels of newly deteriorating synapses. These results emphasize the fact that glucocorticoid concentrations can markedly influence the clearing response, regardless of whether they are elevated or extremely low. Reinnervation. The replacement of lost synaptic contacts is also disrupted by changes in glucocorticoid concentrations as shown in Fig. 2. At 4 days postlesion the replacement response for the two experimental groups reflects their delay in degenerating synapse clearance observed at this same time point. The most interesting finding in the early reinnervation phase is the dramatic replacement by the ADX group at 10 days postlesion. This group’s reinnervation response is significantly ahead of the controls and ADXCORT animals. The accelerated replacement at 10 days is curious but similar to findings by others (50). In other experiments, however, adrenalectomy without steroid replacement did not significantly alter axon sprouting at early time points compared with controls (11, 47, 68). The accelerated replacement of synaptic density by the ADX animals is apparently time-dependent for these animals experienced no further increase by 30 days postlesion; both YA and ADXCORT animals had a significant gain between 10 and 30 days.
CORTICOSTERONE
ALTERS
REACTIVE
SYNAPTOGENESIS
465
This failure to show a significant increase in synaptic density at 30 days (compared with controls) is in agreement with a previous study (68). As high concentrations of glucocorticoids did not prevent significant replacement during this period, it can be assumed that a certain concentration of glucocorticoid may be necessary for the later phase of the recovery process. In contrast to earlier findings directed toward axon sprouting, high concentrations of glucocorticoids did not retard early replacement of synaptic density. Glucocorticoid-treated animals in this study showed a continuous replacement of synaptic density. The high concentrations of CORT did result in some retardation of synapse replacement as evidenced by the fact that ADXCORT animals exhibited significantly lower synaptic densities relative to controls throughout most of the reinnervation time course. By 60 days postlesion, the longest period used in this study, the groups were no longer significantly different from each other. Thus, although adrenalectomy alone may have retarded the late phase of the reinnervation process, the animals were able to attain control values. Numerous studies have shown that after partial denervation the number of degenerating synaptic contacts plus the number of intact synapses does not equal the preoperative normal synaptic density, suggesting that degenerating contacts need to be removed before reactive synaptogenesis can continue (18, 25,26, 32, 33, 54). Elevated levels of degenerating synapses may actually retard synapse formation. Our findings support the contention that the removal of degenerating contacts may not be the limiting factor in the replacement of lost synaptic contacts. Although all groups show a significant correlation between removal of dying contacts and replacement of synaptic density, clearly there are instances when levels of degeneration are equivalent between groups yet replacement of lost contacts are significantly different. A comparison of Figs. 1 and 2 reveals that whereas YA and ADX animals have equivalent density of degeneration at 10 days postlesion, their synaptic replacement is significantly different. Moreover, although the ADXCORT animals are delayed in clearance of degenerating contacts, their replacement level is identical to that of controls. Finally at 60 days postlesion, although all groups show no significant differences in synapse replacement there are significant differences in density of degenerating synapses. In this study, the overall reinnervation of a denervated zone was not significantly impaired with elevated concentrations of CORT. These same CORT concentrations have been shown to markedly impair collateral growth into this area (47). This might reflect differences in the type of afferent fibers which give rise to the newly formed synapses. The reactive growth of the commissural-associational system from the inner molecular layer into the middle molecular layer after an entorhinal lesion has been shown to be influenced by steroids administered within the 1st 15 days postlesion (46-
466
SCHEFF, HOFF, AND ANDERSON
48); the pyramidal cells which give rise to these sprouting afferent fibers have receptors for glucocorticoids. The cells which give rise to the crossed entorhinal-dentate afferent fibers (5 l-53) and which are known to play a role in the reinnervation of the outer molecular layer, have not been shown to bind glucocorticoids. Thus, of the neurons that give rise to new synapses, those that have receptors for glucocorticoids may be influenced by the elevated hormonal state and their sprouting efforts suppressed. On the other hand, the difference may relate to the type of growth required for reinnervation, homotypic (entorhinal, septal) vs. heterotypic (commissural, associational). These different types of collateral sprouts have been shown to respond differently to glucocorticoids (46,47, 68) and have very different lesion-induced growth properties ( 1, 15). Glucocorticoids may well regulate some aspects of the lesion-induced growth response in the hippocampal dentate gyrus, because this area possessesconsiderable binding sites for glucocorticoids ( 10, 17,36,42,6 1,62). In addition, glucocorticoids have been shown to alter brain development (46, 12, 20, 43, 44, 59, 60). If developmental synaptogenesis can be altered it is not unlikely that reactive synaptogenesis would also be affected. The reinnervation selectivity probably has functional consequences, as it would differentially alter the proportional distribution of neuronal input in a previously denervated zone. Under conditions of elevated glucocorticoids (ADXCORT), sprouting of the commissural-associational system is impaired, thereby limiting its contribution to the new input and allowing other afferent fibers to occupy the available sites for synaptic contact. Thus the percentage of total input to the molecular layer for each afferent would vary between YA and ADXCORT animals. Behavioral and electrophysiological studies of such animals would be necessary to aid in differentiation of such alterations in relative synaptic populations. It appears from our experimental data that the neuroendocrine axis can selectively act on the removal of degenerations and the replacement of lost synaptic contacts in the mammalian central nervous system. REFERENCES 1. /iZMITIA, E. C., AND F. C. ZHOU. 1986. Induced homotypic collateral sprouting of hippocampal serotonergic afferent fibers. Pages 129-141 in G. GILAD, Ed., Processes ofRecoveryfrom Neuronal Trauma. Elsevier, Amsterdam. 2. BAKER, T., A. B. DRAKONTIDES, AND W. F. RIKER. 1982. Prevention of the organosphosphorus neuropathy by glucocorticoids. Exp. Neural. 78: 397-408. 3. BIRD, J. W., T. BERG, AND J. H. LEATHEM. 1968. Cathepsin activity of liver and muscle fractions of adrenalectomized rats. Proc. Sot. Exp. Biol. Med. 127: 182-l 88. 4. BOHN, M. C. 1980. Granule cell genesis in the hippocampus of rats treated neonatally with hydrocortisone. Neuroscience5: 2003-2012.
CORTICOSTERONE
ALTERS
REACTIVE
SYNAPTOGENESIS
467
5. BOHN, M. C., AND J. M. LAUDER, 1978. The effects of neonatal hydrccortisone on rat cerebeIIar development: an autoradiographic and light microscopic study. Dev. Neurosci. 1: 250-264. 6. BOHN, M. C., AND J. M. LAUDER. 1979. CerebeIIar granule cell genesis in the hydrocortisane treated rat. Dev. Neurosci. 3: 8 l-89. 7. COTMAN, C. W., AND J. V. NADLER. 1978. Reactive synaptogenesis in the hippocampus. Pages 227-27 1 in C. W. COTMAN, Ed., Neuronal Plasticity. Raven Press, New York. 8. COTMAN, C. W., M. NIETO-SAMPEDRO, AND E. HARRIS. 198 1. Synapse replacement in the nervous system of adult vertebrates. Physiol. Rev. 61: 684-784. 9. CURCIO, C. A., AND J. W. HINDS. 1983. Stability of synaptic density and spine volume in dentate gyrus of aged rats. Neurobiol. Aging 4: 77-87. 10. DEKLOET, R., G. WALLACH, AND B. S. MCEWEN. 1975. Differences in corticosterone and dexamethasone binding to rat brain and pituitary. Endocrinology 96: 598-609. 11. DEKOSKY, S. T., S. W. SCHEFF,AND C. W. COTMAN. 1984. Elevated corticosterone levels: possible cause of reduced axon sprouting in aged animals. Neuroendocrinology 38: 3338. 12. DEKOSKY, S. T., A. J. NONNEMAN, AND S. W. !XXEFF. 1982. Morphologic and behavioral effects of perinatal glucocorticoid administration. Physiol. Behav. 29: 895-900. 13. DRAKONTIDES, A. B. 1982. The effects of glucocorticoid treatment on denervated rat hemidiaphragm. Brain Rex 239: 175- 189. 14. EPSTEIN, S. M., E. VERNEY, T. D. MIALE, AND T. SIDRANSKY. 1967. Studies on the pathogenesis of experimental pulmonary aspergillosis. Am. J. Pathol. 51: 769-788. 15. GAGE, F. H., A. BJ&KLUND, U. STENEVI, AND S. B. DUNNE~~. 1983. Functional correlates of compensatory collateral sprouting by aminergic and cholinetgic afferents in the hippocampal formation. Brain Res. 268: 39-47. 16. GALANT, S. 1979. Serum levels of corticosterone and 18-hydroxy- 1ldeoxycorticosterone in the female rat at the high and low points of the circadian rhythm. Steroids 33: I83185. 17. GERLACH, J. L., AND B. S. MCEWEN. 1972. Rat brain binds adrenal steroid hormone: radioautography of hippocampus with corticosterone. Science 175: 1133-l 136. 18. GOLDOWITZ, D., S. W. SCHEFF,AND C. W. COTMAN. 1979. The specificity of reactive synaptogenesis: a comparative study in adult rat hippocampal formation. Brain Res. 199: 21-38. 19. GORDIS, L., AND H. M. NITOWSKY. 1965. Lysosomes in human cell cultures: kinetics of enzyme release from injured particles. Exp. Cell Res. 38: 556-569. 20. GRANICH, M., AND P. S. TIMIRAS. 197 1. Mechanisms of action of cortisol in maturation of brain lipid patterns in embryonal and young chicks. Pages 2 13-2 18 in M. HAMBURGH AND E. J. BARRINGTON, Eds., Hormones in Development. Appleton-Century-Croft, New York. 21. HALL, E. D., W. F. RIKER, AND T. BAKER. 1977. Glucocorticoid effects on the edrophonium responsiveness of normal and degenerating mammalian motor nerve terminals. Ann. Neurol. 2: 404-408. 22. HALL, E. D., W. F. RIKER, AND T. BAKER. 1983. Beneficial action of glucocorticoid treatment on neuromuscular transmission during early motor nerve degeneration. Exp. Neut-01.79: 488-496. 23. HALL, E. D., T. BAKER, AND W. F. RAKER. 1977. Glucocorticoid preservation of motor nerve function during early degeneration. Ann. Neural. 1: 263-269. 24. HOFF, S. F., S. W. SCHEFF,W. Y. KWAN, AND C. W. COTMAN. 198 1. A new type of lesioninduced synaptogenesis: I. Synaptic turnover in nondenervated zones. Brain Res. 222: l-13.
468
SCHEFF, HOFF, AND ANDERSON
25. HOFF, S. F., S. W. SCHEFF, L. S. BENARW, AND C. W. COTMAN. 1982. Lesion-induced synaptogenesis in the dentate gyrus of aged rats. I. Loss and reacquisition of normal synaptic density. J. Comp. Neurol. 205: 246-252. 26. HOFF, S. F., S. W. SCHEFF,AND C. W. COTMAN. 1982. Lesion-induced synaptogenesis in the dentate gyrus of aged rats. II. Demonstration of an impaired degeneration clearing response. J. Comp. Neurol. 205: 253-259. 27. HOLTZMAN, E., AND A. B. NOYIKOFF. 1965. Lysosomes in the rat sciatic nerve following crush. J. Cell Biol. 27: 65 l-669. 28. KELLENBERGER, E., A. RYTER, AND J. SECHAUD. 1958. Electron microscopic study of DNA-containing plasma. II. Vegetative and mature phase DNA as compared with normal bacterial nucleotides in deifferent physiological states. J. Biophys. Biochemm. Cytol. 4: 67 l-678. 29. KIRK, P. E. 1968. Experimental Design: Procedures for the Behavioral Sciences. Pages 8788. Wadsworth, Belmont, California. 30. KOENIG,J. F. R., AND R. A. KLIPPEL. 1963. The Rat Brain. Williams & Wilkins, Baltimore. 3 1. LANDFIELD, P. W., G. ROSE, L. SANDLES, T. WOHLSTADTER, AND G. S. LYNCH. 1977. Patterns of astroglial hypertrophy and neuronal degeneration in the hippocampus of aged memory-deficient rats. J. Gerontol. 32: 3-12. 32. MATTHEWS, D. A., C. W. COTMAN, AND G. LYNCH. 1976. An electron microscopic study of lesion-induced synaptogenesis in the dentate gyms of the adult rat. I. Magnitude and time course of the degeneration. Brain Res. 115: l-2 1. 33. MATTHEWS, D. A., C. W. COTMAN, AND G. LYNCH. 1976. An electron microscopic study of lesion-induced synaptogenesis in the dentate gyms of the adult rat. II. Reappearance of morphologically normal synaptic contacts. Brain Res. 115: 22-4 1. 34. MCEWEN, B. S., J. M. WEISS, AND L. S. SCHWARTZ. 1969. Uptake of corticosterone by rat brain and its concentration by certain limbic structures. Brain Res. 16: 227-24 1. 35. MCWIUIAMS, J. R., AND G. LYNCH. 1984. Synaptic density and axonal sprouting in rat hippocampus: stability in adulthood and decline in late adulthood. Brain Res. 294: 152156. 36. MERKOW, L., M. PARDO, S. M. EPSTEIN, E. AND H. SIDRANSKY. 1968. Lysosomal stability during phagocytosis of Aspergillus flaws spores by aveolar macrophages of cortisoneteated mice. Science 160: 79-8 1. 37. MEYER, J. S., D. J. MICCO, B. S. STEPHENSEON,L. C. KREY, AND B. S. MCEWEN. 1979. Subcutaneous implantation method for chronic glucocorticoid replacement therapy. Physiol. Behav. 22: 867-870. 38. NESTLER, E. J., T. C. RAINBOW, B. S. MCEWEN, ANDP. GREENGARD. 1981. Corticosterone increases the amount of protein 1, a neuron-specific phosphoprotein, in rat hippocampus. Science212: 1162-l 164. 39. PFAFF, D. W., M. T. A. SILVA, AND J. M. WEISS. 197 1. Telemetered recording of hormone effectson hippocampal neurons. Science 172: 394-395. 40. R.-~P, J. P. 1970. Gas chromatographic measurement of peripheral plasma ll-hydroxydeoxycorticosterone in adrenal regeneration hypertension. Endocrinology 86: 668-677. 41. REYNOLDS, E. S. 1963. The use of lead citrate at high pH as an electron-opaque stain in electron microscopy. J. Cell Biol. 17,208. 42. RHEES, R. W., B. I. GROSSER,AND W. STEVENS. 1975. The autoradiographic localization of [“Hldexamethasone in the brain and pituitary of the rat. Brain Res. 100: 15 l-l 56. 43. SALAS, M., AND S. SCHAPIRO. 1970. Hormonal influences upon the maturation of the rat brain’s responsiveness to sensory stimuli. Physiol. Behav. 5: 7-I 1.
CORTICOSTERONE
ALTERS
REACTIVE
SYNAPTOCENESIS
469
44. ScHAPIRO, S. 1968. Some physiological, biochemical and behavioral consequences of neonatal hormone administration: cortisol and thyroxine. Gen. Camp. Endocrinol. 10: 2 14228. 45. SCXEFF, S. W., L. S. BENARW,
AND C. W. COTMAN. 1980. Decline of reactive fiber growth in the dentate gyms of aged rats compared to young adult rats following entorhinal cortex removal. Bruin Res. 199: 21-38. 46. SCHEFF, S. W., L. S. BENARDO, AND C. W. COTMAN. 1980. Hydrocortisone administration retards axon sprouting in the rat dentate gyms. Exp. Nezuol. 68: 195-20 1. 47. S~HEFF, S. W., AND C. W. COTMAN. 1982. Chronic glucocorticoid therapy alters axon sprouting in the hippocampal dentate gyrus. Exp. Neural. 76: 644-654. 48. SCXEFF, S. W., AND S. T. DEKOSKY. 1983, Steroid suppression of axon sprouting in the hippocampal dentate gyrus of the adult rat: dose-response relationship. Exp. Neural. 82: 183-191. 49. SEGAL, M. 1976. Interactions of ACTH and norepinephrine on the activity of rat hippocampal cells. Neuropharmacology 15: 329-333. 50. SILVERMAN, A. J., AND E. A. ZIMMERMAN. 1982. Adrenalectomy increases sprouting in a peptidergic neurosecretory system. Neuroscience 7: 2705-27 14. 5 1. STEWARD, 0. 1976. Reinnervation of dentate gyrus by homologous afferents following entorhinal cortex lesions in adult rats. Science 194: 426-428. 52. STEWARD, O., AND S. L. VINSANT. 1978. Identification of the cells of origin of a central pathway which sprouts following lesions in mature rats. Brain Res. 147: 223-243. 53. STEWARD, O., AND S. L. VINSANT. 1978. Collateral projections of cells in the surviving entorhinal area which reinnervates the dentate gyrus of the rat following unilateral entorhinal lesions. Brain Res. 149: 2 16-222. 54. STEWARD, O., AND S. L. VINSANT. 1983. The process of reinnervation in the dentate gyms of the adult rat: a quantitative electron microscopic analysis of terminal proliferation and reactive synaptogenesis. J. Comp. Neural. 214: 370-386. 55. SYMONS, A. M., D. A. LEWIS, AND R. J. ANCILL. 1969. Stabilizing action ofanti-intlammatory steroids on lysosomes. B&hem. Pharmacol. 1% 258 l-2582. 56. TAYLOR, A. N., G. K. MATHESON, AND N. DAFNY. 197 1. Modification of the responsiveness of components of the limibic-midbrain circuit by corticosteroids and ACTH. Pages 67-78 in C. H. SAWYER AND R. A. GORSIU, E&., Steroid Hormones and Brain Function. Univ. of California Press, Berkley. 57. UNSICKER, K., B. KRISCH, U. &TEN, AND H. THOENEN. 1978. Nerve growth factor-induced fiber outgrowth from isolated rat adrenal chromaffin cells: impairment by glucocorticoids. Proc. Natl. Acad. Sci. U.S.A. 753498-3502. 58. URBAN, I., AND D. DEWIED. 1976. Changes in excitability of the theta activity generating substrate by ACT~rO in the rat. Exp. Brain Res. 24: 325-334. 59. VICEDOMINI, J. P., A. J. NONNEMAN, S. T. DEKOSKY, AND S. W. SCHEFF. 1985. Perinatal glucocorticoids disrupt learning: a sexually dimorphic response. Physiol. Behav. 36: 145149. 60. VICEWMINI, J. P., A. J. NONNEMAN, S. T. DEKOSKY, AND S. W. SCHEFF. 1985. Perinatal glucocorticoids alter dentate gyrus electrophysiology. Bruin Res. Bull. 15: Ill- 116. 6 1. WAREMBOURG, M. 1975. Radioautographic study of the rat brain and pituitary after injection of r3H]dexamethasone. Cell Tissue Res. 161: 183- 19 1. 62. WAREMBOURG, M. 1975. Radioautographic study ofthe rat brain after injection of 1.2 [‘HI corticosterone. Brain Res. 89: 6 l-70. 63. WARR, B. W., J. S. DE OLMOS, AND L. HEIMER. 1981. Horseradish peroxidasez the basic procedure. Pages 207-262 in L. HEIMER AND M. J. ROBARDS, I?.&., Neuroanatomical Tract-TracingMethods. Plenum, New York.
470
SCHEFF, HOFF, AND ANDERSON
64. WEISSMANN, G. 1969. The effects of steroids and drugs on lysosomes. Pages 276-295 in J. T. DINGLE AND H. B. FEu, Fds., Lysosomes in SioZogy undPathology.North-Holland, Amsterdam. 65. WEISSMANN, G., AND J. T. DINGLE. 1961. Release of lysosomal protease by ultraviolet irradiation and inhibition by hydrocortisone. Exp.CellRes.25,201-210. 66. WINER, B. J. 1971. StatisticalPrinciples in Experimental Design.McGraw-Hill, New York. 67. WITHROW, C. D., AND D. M. WOODBURY. 1972. Some aspect of the pharmacology of adrenal steroids and the central nervous system. Pages 41-55 in H. J. REULEN AND K. SCHURMANN, Eds.,Steroids andBrainEdema.Springer-Verlag, New York. 68. ZHOU, F. C., AND E. C. &MITIA. 1985. The effect of adrenalectomy and corticosterone on homotypic collateral sprouting of serotonergic fibers in hippocampus. Neurosci. Left. 54: 111-116.
NEUROLOGY
93,456-470
(1986)
Altered Regulation of Lesion-Induced Synaptogenesis by Adrenalectomy and Cotticosterone in Young Adult Rats STEPHENW. SCHEPP, STEVEN F. HOFF, AND KEVIN J. ANDERSON’ Departments ofAnatomy and Neurology, Sanders-Brown Research Center on Aging, University of Kentucky, Lexington, Kentucky 40536, and Department of Pharmacology, Chicago Medical School, University Health Sciences, North Chicago, Illinois 60064 Received September 30, 1985; revision received April 22, 1986 Quantitative electron microscopy was used to examine the effect of circulating glucocorticoids on the removal of degenerating synapses and the replacement of lost synaptic contacts in young adult rats that follow partial denervation of the hippocampal dentate gyrus. Subjects were adrenalectomized prior to subcutaneous implantation of pellets containing a specified concentration of corticosterone and subsequent unilateral ablation of the entorhinal cortex. Animals maintained at high circulating concentrations of glucocorticoids were significantly retarded in the early phase of degenerating synapse removal and in the rate of synaptic replacement. Subjects maintained at extremely low concentrations of glucocorticoids were also significantly retarded in the early stagesof synapse removal but showed an early replacement of lost synaptic contacts followed by a dramatic decrease in the rate of replacement. By 60 days atIer the lesion both groups of animals showed synapse replacement equivalent to young adult controls while significant amounts of degenerating synapses still remained in the denervated neuropil. The results demonstrate that circulating glucocorticoids can exert a marked influence on lesion-induced synaptic replacement in the hippoCampal dentate gyrus. 8 1986 AC&~~C PIB, 1~.
Abbreviations: ADX-adrenalectomy, ADXCORT-adrenalectomy and corticosterone, CORT-cmticosterone, YA-young adult control. ’ We thank C. W. Cotman for his support and encouragement in these studies. This research was supported by grants from the National Institute of Health NS1698 I, MH19691 and the John D. MacArthur Research Program on Successful Aging. Dr. Anderson’s present address is Department of Psychobiology, University of California, Irvine, CA 927 17. Please address reprint requests to Dr. S. W. Scheff, Sanders-Brown Research Center, Univ. of Kentucky, Lexington, KY 40536. 456 0014-4886186 $3.00 Copyriebt Q 1986 by Academic I’m& Inc. AII rights of mpmdwtion in any form resew&
CORTICOSTERONE
ALTERS
REACTIVE
SYNAPTOGENESIS
457
INTRODUCTION Selective lesions in the central nervous systemare capable of stimulating axonal growth and new synapseformation in the partially denervated regions (7, 8). Several findings suggest that this growth process in the hippocampus might be influenced by glucocorticoids. Neurons in the hippocampus have been shown to possess binding sites for glucocorticoids ( 17, 34,62); corticosteroids can modify the metabolism of hippocampal cells (67); and corticosteroids have been shown to alter the electrical activity of the hippocampus (39,49, 56, 58). Moreover, in certain cells in tissue culture, glucocorticoids can modulate process outgrowth (57). Previous work has shown that peripherally administered glucocorticoids can modify this axonal growth and significantly decreases the response of certain reactive afferent fibers in the hippocampal dentate gyrus of young adult rats ( 11,46-48). Animals maintained at elevated levels of circulating corticosterone (CORT) show significantly less reactive fiber growth than control animals or subjectsmaintained at low circulating levels (47). High levels of circulating CORT have been also shown to disrupt growth of sympathetic fibers from the superior cervical ganglion after a transection of the fimbria ( 1 1), suggesting that steroids influenced not only central neurons but peripheral as well. Finally, glucocorticoids have been found to suppress reactive fiber growth in a dose-dependent fashion. Dexamethasone as well as hydrocortisone has been shown to alter axon sprouting and the suppressive effect can be directly related to the anti-inflammatory properties of the hormones (48). In the present study, as one step in clarifying the effects of glucocorticoids on axodendritic interactions in the hippocampus, quantitative studies were carried out on the replacement of lost synaptic contacts in the denervated zone. At the light microscopic level only changes in axon collateral growth can be monitored; in many cases only one afferent system is analyzed. Our aim was to critically analyze electron microscopically a denervated neuropil in subjects treated with glucocorticoids and ascertain whether or not new synapse formation and removal of degenerating contacts are affected. Because the time course of the removal of degenerating synapses and the replacement of synaptic contacts is well characterized in the hippocampus (25, 26,32,33) and the limbic system has been shown to be sensitive to glucocorticoids (38,39,56,58), this area ofthe central nervous system was considered ideal to test the influence of these steroids on reactive synapse replacement. METHODS Young adult male Sprague-Dawley rats, 90 days of age (3 10 to 340 g), were housed three to a cage and had food and water ad libitum. The experimental animals were subjected to a bilateral adrenalectomy at least 10 days prior
458
SCHEFF, HOFF, AND ANDERSON
to a unilateral entorhinal lesion using methods identical to those described elsewhere (45). At the time of the lesion, each experimental animal was randomly assigned to one of two different groups: group I (ADX) consisted of animals implanted subcutaneously with cholesterol pellets in the posterior cervical region of the neck (N = 17); these animals were subsequently maintained on 0.9% sodium chloride. Group II (ADXCORT) consisted of corticosterone-maintained animals (N = 20). These subjects were subcutaneously implanted with CORT pellets in the posterior cervical region of the neck identical to the procedure used for the ADX subjects and maintained on saline. Animals from both groups were killed at 4 (ADX N = 7; ADXCORT N = 6), 10 (ADX N = 4; ADXCORT N = 6), 30 (ADX N = 3; ADX N = 5), or 60 (ADX N = 3; ADXCORT N = 3) days postlesion. Group III (YA) consisted of young adult control animals 90 days of age (N = 14). These animals received entorhinal cortex lesions identical to those of the other two groups. Animals in this group were killed at 4 (N = 4), 10 (N = 3), 30 (N = 3), or 60 (N = 4) days postlesion, Pellets were formed from molten CORT according to the procedure of Meyer et al. (37). This method produced stable blood and brain CORT concentrations in the range of 28 pg/ 100 ml serum within a very short time and maintained these concentrations during the course of the study. Pellets of cholesterol were formed in a similar fashion and did not alter the concentrations of CORT (less than 0.5 &lOO ml serum) in the ADX animals. At the designated postlesion survival times, the animal was anesthetized with sodium pentobarbital(50 mg/kg) and blood samples were obtained by cardiac puncture. Plasma was then collected and processed in triplicate by radioimmunoassay to determine circulating concentrations of CORT as described elsewhere (11). Subsequently the brain was fixed by vascular perfusion with a solution of 4% paraformaldehyde, 0.5% glutaraldehyde, 0.54% D-ghCOSe, in 0.1 Msodium phosphate buffer (pH 7.4). After a l-h perfusion, the entire animal was placed in a plastic bag containing 10 ml perfusate and refrigerated 24 h. The hippocampi were then carefully dissected using a procedure identical to that described elsewhere to study synaptic density in this brain region (25,26). The hippocampi were sectioned in a coronal plane (175 to 200 pm) with a Sorvall tissue chopper. Isolated portions of the anterior hippocampus (levels 3290 to 4380 of K&rig and Klippel(30), were postfixed in 1% osmium tetroxide in 0.1 M sodium phosphate buffer and stained en bloc with 0.5% many1 acetate (28). The samples were dehydrated in a graded ethanol series and propylene oxide and embedded in Epon-Araldite. Blocks were coded at this time so that all subsequent processing and analyses were carried out blind with respect to treatment group.
CORTICOSTERONE
ALTERS
REACTIVE
SYNAFTOGENESIS
459
One-micrometer-thick sections were cut and stained with thionin for light microscopic study and to facilitate trimming for ultrathin sectioning. These thick sections were used for measurements of the width of the dentate neuropil. Camera lucida drawings were obtained for several sections from each animal and depicted the zone from the granule cell layer to the hippocampal fissure of the dorsal blade of the dentate gyrus. After trimming of the block, ultrathin sections were cut and collected on Formvar-coated slot grids (1 X 2-mm slot size). The sections were stained in an aqueous solution of uranyl acetate for 20 min, rinsed, and stained with Reynolds lead citrate (4 1) for 4 to 5 min. Under the electron microscope the granule cell layer was ascertained in the ultrathin sections. For each animal, a montage, one photographic frame wide, was made of the entire dentate gyrus molecular layer at right angles to the granule cells of the dentate gyrus dorsal leaf. Micrographs were taken at an initial magnification of X7500 and enlarged photographically to X 15,000. The average tissue area analyzed for each montage of the outer two-thirds of the dentate gyrus molecular layer was 1150 pm*. In this study the criteria for identification of a normal synapse were the presence of synaptic vesicles and presynaptic membrane in association with a postsynaptic density, identical to that used previously (24-26,32, 33). Every recognizable synapse in the outer dentate gyrus molecular layer was counted and the mean density of synapses per 100 pm* was determined for each animal. If a presynaptic specialization made a discontinuous connection with the same postsynaptic specialization, it was considered a single contact area. However, if a presynaptic specialization made contact with more than one postsynaptic specialization it was considered a multiple synaptic contact. Degenerating synapses were characterized by either an electron-dense or electron-lucent profile connected with a synaptic junction or an abnormally translucent profile containing a watery cytoplasm connected with a postsynaptic density (9,24,35). Anatomical reconstruction, to assesscompleteness of the lesion, was performed on all animals by sectioning the posterior portion of the brain in the horizontal plane and staining frozen sections with cresyl violet. Sections through the region normally occupied by the entorhinal cortex were drawn from these stained sections to assess extent of the lesions. This procedure confirmed that in all animals reported in this study, the surgical procedures removed both the medial and lateral aspects of the entorhinal cortex. Verification of adrenalectomy was accomplished by autopsy at the time of perfusion. RESULTS
Serum Corticosterone Concentrations. Plasma concentrations in adrenalectomized
animals were determined
of CORT by radioimmunoassay. The
460
SCHEFF, HOFF, AND ANDERSON
mean serum concentration of CORT found in animals implanted with pellets of cholesterol (N = 17) used in this study was 0.2 (kO.2) pg/ 100 ml serum. The mean serum concentration of CORT found in animals implanted with pellets of CORT (N = 20) used in this study was 27.9 (k4.8) &lo0 ml serum. An ANOVA (treatment by days) (66) revealed no significant days effect (F(3,29) = 1.03; P > 0.1) indicating that for each experimental group CORT concentrations remained constant. These values are identical to those reported earlier using a similar technique (11). CORT was not determined for young adult control animals because the process of anesthetizing the animals for blood sampling and EM perfusion would artificially elevate the circulating CORT due to stress (16,40). Experimental animals were not affected by this procedure because their adrenals had been removed. Shrinkage of Dentate Gyrus Molecular Layer. One-micrometer-thick sections stained with thionin revealed no significant changes in molecular layer width as a function of treatment (F(2,39) = 1.28; P > 0.1) or days after entorhinal lesion (F(3,39) = 1.89; P > 0.1). Thus, differences found in this study could not be accounted for by differential shrinkage of the molecular layer. Degeneration The dynamics of degenerating synapse removal from the denervated ipsilateral outer two-thirds of the molecular layer were assessed as a function of days postlesion. Two aspects of the degenerating synapse clearance were analyzed. First, do each of the groups show a significant clearance of degenerating contacts and second, are there any differences in the time course of the clearance? Mean density of degenerating synaptic contacts were subjected to a two-way ANOVA (treatment by days) (66) with subsequent individual group comparisons assessed with the least significant difference test (29). These analyses revealed a significant clearing response (removal of degenerating synapses) with time postlesion for all groups (F(3,39) = 27.85, P < 0.00 1) and a significant difference in the clearing response between groups (F(2,39) = 15.69, P < O.OOl), indicating that circulating concentrations of hormones had an effect. Although all groups showed significant declines in density of degenerating synapses during the survival period studied, the process was not uniform between groups (Fig. 1). The differences in the clearing response can be best appreciated by analyzing total amount of degenerating synapses at various days postlesion. The two experimental groups, ADXCORT (N = 6) and ADX (N = 7) animals, showed a significantly greater accumulation of degenerating synapses (P < 0.0 1) at 4 days postlesion compared with young adult controls (YA, N = 4) at the same survival time. This difference was still apparent by 60 days postlesion (P < 0.05 for ADX (N = 3) and P < 0.01 for
CORTICOSTERONE
ALTERS
REACTIVE
I
7)
(6)
SYNAPTOGENESIS
461
q YAControl
(3)
(4)
(6)
(3)
(3) (5)
(4)
(3) (3)
10 60 4 30 Days Post Entorhinal Lesion FOG. 1. Bar graph showing time course of changes in the density of degenerating contacts in the outer two-thirds of the dentate gyrus molecular layer following ipsilateral removal of the entorhinal cortex. Animals maintained at high concentrations of circulating glucocorticoids (ADXCORT) showed the greatest accumulation of degenerating synaptic contacts throughout the study compared with animals maintained at extremely low concentrations of glucocorticoids (ADX) and young adult controls (YA). The bars represent group means + SD, numbers in parentheses below bars represent number of subjects in that group. * = P < 0.05t = P < 0.01.
ADXCORT (N = 3)). By 10 days postlesion ADX animals (N = 4) had significantly cleared the degenerating contacts from the outer two-thirds of the molecular layer and were no longer reliably different from YA (N = 3) (P > 0.1). ADXCORT subjects (N = 6) however, failed to demonstrate an equivalent clearance response within the same postlesion interval: this group showed a significantly greater accumulation of degenerating synapses (P < 0.05). By 30 days postlesion all groups removed the synapses equally well with the ADXCORT subjects showing the greatest difference between 10 and 30 days survival. ADX and ADXCORT animals, however, failed to show any additional clearing between 30 and 60 days postlesion; this was in contrast to the response observed in YA subjects. Moreover, both experimental groups were significantly different from controls (N = 4) in total amount of remaining degenerating contacts at 60 days postlesion (ADX P < 0.05 and ADXCORT P < 0.0 1). Reinnervation Changes in the density of synaptic contacts in the outer two-thirds of the denervated molecular layer of the dentate gyrus were analyzed using a two-
462
SCHEFF. HOFF, AND ANDERSON
q VAControl 0
ADX
q ADXCORT
(4)
(6)
(3) (2
(6)
(4)
30 10 60 4Days Post Entorhinal Lesion FIG. 2. Bar graph showing time course of changes in the density of normal synaptic contacts in the outer two-thirds of the dentate gyrus molecular layer following ipsilateral removal of the entorhinal cortex. Animals maintained at either extremely high (ADXCORT) or extremely low (ADX) concentrations of circulating glucocorticoids were markedly influenced in synapse replacement compared with young adult controls (YA). The bars represent group means + SD, numbers in parentheses below bars represent number of subjects in that group. * = P < 0.05 *=p
way ANOVA (treatment by days) (66). Subsequent individual group comparisons when appropriate were made utilizing the least significant difference test of significance (29). Those analyses revealed significant differences in the density of synaptic contacts within groups (F(3,39) = 92.2, P < O.OOl), indicating change with time, and between groups (F(2,39) = 4.09, P -c 0.05), indicating that concentrations of hormone had an effect. At 4 days postlesion there were no significant differences between groups (P > 0.1) (see Fig. 2). All groups showed a significant (P < 0.0 1) replacement of lost synaptic contacts by 10 days postlesion compared with the 4-day survival time. This increase in density was not uniform between groups. The ADX animals showed the most dramatic increase in density of intact synapses and were significantly greater compared with both YA (P < 0.01) and ADXCORT subjects (P < 0.0 I), with the latter two groups not differing from each other (P > 0.1). By 30 days postlesion, the ADX animals (N = 3) failed to show any additional significant increase in synaptic density relative to similarly treated animals examined at 10 days (P > 0.1). In fact, the ADX subjects exhibited significantly less synaptic replacement than the YA controls (P < 0.0 1) at 30 days, indicating a subsequent delay in the replacement process. The ADXCORT animals showed a significant increase in synaptic
CORTICOSTERONE
ALTERS
REACTIVE
SYNAPTOGENESIS
463
density at 30 days postlesion compared with similarly treated rats at 10days postlesion (P < O.Ol), but were found to be significantly less dense than YA animals. By 60 days postlesion all groups showed a significant increase in synaptic density and no significant differences were observed between the experimental and control groups (P > 0.1).
Correlation of Synapse Replacement to Removal of Degenerating Contacts. The relationship between the removal of degenerating synapses and the replacement of synapses was assessed by determining the correlation coefficient. The YA controls showed a significant correlation (r = 0.894, d.f. 12, P < 0.00 1) indicating that as dying synapses were removed new synapses were formed. The ADX and ADXCORT animals also showed a significant correlation (r = 0.903, d.f. 15, P < 0.001 and r = 0.782, d.f. 18, P< O.OOl), respectively, in the replacement of synaptic density as a function of degeneration clearance. DISCUSSION We reported elsewhere that the hippocampal formation responds to changes in circulating levels of glucocorticoids by retarding the growth of collateral sprouting following selective damage (11,46-48). The present experiment is an extension of that work and carefully examines the influence of glucocorticoids on the time-dependent replacement of synaptic contacts in the outer two-thirds of the molecular layer of the hippocampal dentate gyrus. This area is massively denervated (85%) by an ipsilateral lesion of the entorhinal cortex (25,32,45). Two different yet related aspects of the replacement process were studied: (a) removal of degenerating synapses and (b) the return of normal synaptic contacts. Degeneration. All groups showed removal of degenerating synapses throughout the duration of the experiment, but the time course of this removal was different depending on circulating hormone concentrations. The most striking differences occur on postlesion days 4 and 10 (Fig. 1). Both ADX and ADXCORT animals showed a significantly greater density of degenerating synapses compared with controls at 4 days after entorhinal cortex removal. Previous experiments reported significant accumulation of degenerating synaptic contacts at 2 days after entorhinal cortex removal in young adult animals, with the 2-day values similar to those observed at 4 days in the ADX and ADXCORT groups (26,32). Even in these experiments there appears to be a rapid removal of dying synapses by 2 days for the total number of dying synapses and intact synapses does not equal the prelesion synaptic density. Thus the clearing response is initiated very early. Either the terminals in the ADX and ADXCORT groups are protected from immediate degradation as suggested previously (2, 13, 21-23) or these animals lack the
464
SCHEFF, HOFF, AND ANDERSON
ability to initiate the early rapid phase of the clearing response which is seen in young adult controls (25). Previous studies with hydrocortisone and corticosterone demonstrated a significant increase in astrocytes in animals maintained with high concentrations of steroid (46,47). That finding is suggestive of a glial influence which may affect the removal of degenerative debris and presumably alter the replacement of lost synaptic contacts. The same type of delay of the clearing response was observed in aged animals which are known to have elevated concentrations of glucocorticoids (26) and normally show hypertrophied astrocytes in the hippocampal formation (3 1, 45). It is possible that glial cell mechanisms involved in phagocytosis are regulated by glucocorticoids. Glucocorticoids have been reported to stabilize lysosomal-related events in both neuronal and nonneuronal systems (3, 14, 19,27, 36, 55, 63-65). The present experimental results suggest that the removal of dying synapses is directly or indirectly slowed by the presence of altered glucocorticoid concentrations. By 10 days after the lesion the ADX animals have begun to show a normal clearing response whereas the ADXCORT animals continue to lag significantly behind. The latter animals continue to show delayed clearing compared with the control group (YA) throughout the remainder of the experiment. Thus, the removal of dying contacts in the elevated CORT animals closely parallels the response observed in aged animals following the same lesion paradigm (26). The failure to detect further reduction in amounts of these contacts after 30 days postlesion in the ADXCORT group might possibly indicate a combination of an abortive protection of synaptic contacts and a clearing response which can keep pace only with low levels of newly deteriorating synapses. These results emphasize the fact that glucocorticoid concentrations can markedly influence the clearing response, regardless of whether they are elevated or extremely low. Reinnervation. The replacement of lost synaptic contacts is also disrupted by changes in glucocorticoid concentrations as shown in Fig. 2. At 4 days postlesion the replacement response for the two experimental groups reflects their delay in degenerating synapse clearance observed at this same time point. The most interesting finding in the early reinnervation phase is the dramatic replacement by the ADX group at 10 days postlesion. This group’s reinnervation response is significantly ahead of the controls and ADXCORT animals. The accelerated replacement at 10 days is curious but similar to findings by others (50). In other experiments, however, adrenalectomy without steroid replacement did not significantly alter axon sprouting at early time points compared with controls (11, 47, 68). The accelerated replacement of synaptic density by the ADX animals is apparently time-dependent for these animals experienced no further increase by 30 days postlesion; both YA and ADXCORT animals had a significant gain between 10 and 30 days.
CORTICOSTERONE
ALTERS
REACTIVE
SYNAPTOGENESIS
465
This failure to show a significant increase in synaptic density at 30 days (compared with controls) is in agreement with a previous study (68). As high concentrations of glucocorticoids did not prevent significant replacement during this period, it can be assumed that a certain concentration of glucocorticoid may be necessary for the later phase of the recovery process. In contrast to earlier findings directed toward axon sprouting, high concentrations of glucocorticoids did not retard early replacement of synaptic density. Glucocorticoid-treated animals in this study showed a continuous replacement of synaptic density. The high concentrations of CORT did result in some retardation of synapse replacement as evidenced by the fact that ADXCORT animals exhibited significantly lower synaptic densities relative to controls throughout most of the reinnervation time course. By 60 days postlesion, the longest period used in this study, the groups were no longer significantly different from each other. Thus, although adrenalectomy alone may have retarded the late phase of the reinnervation process, the animals were able to attain control values. Numerous studies have shown that after partial denervation the number of degenerating synaptic contacts plus the number of intact synapses does not equal the preoperative normal synaptic density, suggesting that degenerating contacts need to be removed before reactive synaptogenesis can continue (18, 25,26, 32, 33, 54). Elevated levels of degenerating synapses may actually retard synapse formation. Our findings support the contention that the removal of degenerating contacts may not be the limiting factor in the replacement of lost synaptic contacts. Although all groups show a significant correlation between removal of dying contacts and replacement of synaptic density, clearly there are instances when levels of degeneration are equivalent between groups yet replacement of lost contacts are significantly different. A comparison of Figs. 1 and 2 reveals that whereas YA and ADX animals have equivalent density of degeneration at 10 days postlesion, their synaptic replacement is significantly different. Moreover, although the ADXCORT animals are delayed in clearance of degenerating contacts, their replacement level is identical to that of controls. Finally at 60 days postlesion, although all groups show no significant differences in synapse replacement there are significant differences in density of degenerating synapses. In this study, the overall reinnervation of a denervated zone was not significantly impaired with elevated concentrations of CORT. These same CORT concentrations have been shown to markedly impair collateral growth into this area (47). This might reflect differences in the type of afferent fibers which give rise to the newly formed synapses. The reactive growth of the commissural-associational system from the inner molecular layer into the middle molecular layer after an entorhinal lesion has been shown to be influenced by steroids administered within the 1st 15 days postlesion (46-
466
SCHEFF, HOFF, AND ANDERSON
48); the pyramidal cells which give rise to these sprouting afferent fibers have receptors for glucocorticoids. The cells which give rise to the crossed entorhinal-dentate afferent fibers (5 l-53) and which are known to play a role in the reinnervation of the outer molecular layer, have not been shown to bind glucocorticoids. Thus, of the neurons that give rise to new synapses, those that have receptors for glucocorticoids may be influenced by the elevated hormonal state and their sprouting efforts suppressed. On the other hand, the difference may relate to the type of growth required for reinnervation, homotypic (entorhinal, septal) vs. heterotypic (commissural, associational). These different types of collateral sprouts have been shown to respond differently to glucocorticoids (46,47, 68) and have very different lesion-induced growth properties ( 1, 15). Glucocorticoids may well regulate some aspects of the lesion-induced growth response in the hippocampal dentate gyrus, because this area possessesconsiderable binding sites for glucocorticoids ( 10, 17,36,42,6 1,62). In addition, glucocorticoids have been shown to alter brain development (46, 12, 20, 43, 44, 59, 60). If developmental synaptogenesis can be altered it is not unlikely that reactive synaptogenesis would also be affected. The reinnervation selectivity probably has functional consequences, as it would differentially alter the proportional distribution of neuronal input in a previously denervated zone. Under conditions of elevated glucocorticoids (ADXCORT), sprouting of the commissural-associational system is impaired, thereby limiting its contribution to the new input and allowing other afferent fibers to occupy the available sites for synaptic contact. Thus the percentage of total input to the molecular layer for each afferent would vary between YA and ADXCORT animals. Behavioral and electrophysiological studies of such animals would be necessary to aid in differentiation of such alterations in relative synaptic populations. It appears from our experimental data that the neuroendocrine axis can selectively act on the removal of degenerations and the replacement of lost synaptic contacts in the mammalian central nervous system. REFERENCES 1. /iZMITIA, E. C., AND F. C. ZHOU. 1986. Induced homotypic collateral sprouting of hippocampal serotonergic afferent fibers. Pages 129-141 in G. GILAD, Ed., Processes ofRecoveryfrom Neuronal Trauma. Elsevier, Amsterdam. 2. BAKER, T., A. B. DRAKONTIDES, AND W. F. RIKER. 1982. Prevention of the organosphosphorus neuropathy by glucocorticoids. Exp. Neural. 78: 397-408. 3. BIRD, J. W., T. BERG, AND J. H. LEATHEM. 1968. Cathepsin activity of liver and muscle fractions of adrenalectomized rats. Proc. Sot. Exp. Biol. Med. 127: 182-l 88. 4. BOHN, M. C. 1980. Granule cell genesis in the hippocampus of rats treated neonatally with hydrocortisone. Neuroscience5: 2003-2012.
CORTICOSTERONE
ALTERS
REACTIVE
SYNAPTOGENESIS
467
5. BOHN, M. C., AND J. M. LAUDER, 1978. The effects of neonatal hydrccortisone on rat cerebeIIar development: an autoradiographic and light microscopic study. Dev. Neurosci. 1: 250-264. 6. BOHN, M. C., AND J. M. LAUDER. 1979. CerebeIIar granule cell genesis in the hydrocortisane treated rat. Dev. Neurosci. 3: 8 l-89. 7. COTMAN, C. W., AND J. V. NADLER. 1978. Reactive synaptogenesis in the hippocampus. Pages 227-27 1 in C. W. COTMAN, Ed., Neuronal Plasticity. Raven Press, New York. 8. COTMAN, C. W., M. NIETO-SAMPEDRO, AND E. HARRIS. 198 1. Synapse replacement in the nervous system of adult vertebrates. Physiol. Rev. 61: 684-784. 9. CURCIO, C. A., AND J. W. HINDS. 1983. Stability of synaptic density and spine volume in dentate gyrus of aged rats. Neurobiol. Aging 4: 77-87. 10. DEKLOET, R., G. WALLACH, AND B. S. MCEWEN. 1975. Differences in corticosterone and dexamethasone binding to rat brain and pituitary. Endocrinology 96: 598-609. 11. DEKOSKY, S. T., S. W. SCHEFF,AND C. W. COTMAN. 1984. Elevated corticosterone levels: possible cause of reduced axon sprouting in aged animals. Neuroendocrinology 38: 3338. 12. DEKOSKY, S. T., A. J. NONNEMAN, AND S. W. !XXEFF. 1982. Morphologic and behavioral effects of perinatal glucocorticoid administration. Physiol. Behav. 29: 895-900. 13. DRAKONTIDES, A. B. 1982. The effects of glucocorticoid treatment on denervated rat hemidiaphragm. Brain Rex 239: 175- 189. 14. EPSTEIN, S. M., E. VERNEY, T. D. MIALE, AND T. SIDRANSKY. 1967. Studies on the pathogenesis of experimental pulmonary aspergillosis. Am. J. Pathol. 51: 769-788. 15. GAGE, F. H., A. BJ&KLUND, U. STENEVI, AND S. B. DUNNE~~. 1983. Functional correlates of compensatory collateral sprouting by aminergic and cholinetgic afferents in the hippocampal formation. Brain Res. 268: 39-47. 16. GALANT, S. 1979. Serum levels of corticosterone and 18-hydroxy- 1ldeoxycorticosterone in the female rat at the high and low points of the circadian rhythm. Steroids 33: I83185. 17. GERLACH, J. L., AND B. S. MCEWEN. 1972. Rat brain binds adrenal steroid hormone: radioautography of hippocampus with corticosterone. Science 175: 1133-l 136. 18. GOLDOWITZ, D., S. W. SCHEFF,AND C. W. COTMAN. 1979. The specificity of reactive synaptogenesis: a comparative study in adult rat hippocampal formation. Brain Res. 199: 21-38. 19. GORDIS, L., AND H. M. NITOWSKY. 1965. Lysosomes in human cell cultures: kinetics of enzyme release from injured particles. Exp. Cell Res. 38: 556-569. 20. GRANICH, M., AND P. S. TIMIRAS. 197 1. Mechanisms of action of cortisol in maturation of brain lipid patterns in embryonal and young chicks. Pages 2 13-2 18 in M. HAMBURGH AND E. J. BARRINGTON, Eds., Hormones in Development. Appleton-Century-Croft, New York. 21. HALL, E. D., W. F. RIKER, AND T. BAKER. 1977. Glucocorticoid effects on the edrophonium responsiveness of normal and degenerating mammalian motor nerve terminals. Ann. Neurol. 2: 404-408. 22. HALL, E. D., W. F. RIKER, AND T. BAKER. 1983. Beneficial action of glucocorticoid treatment on neuromuscular transmission during early motor nerve degeneration. Exp. Neut-01.79: 488-496. 23. HALL, E. D., T. BAKER, AND W. F. RAKER. 1977. Glucocorticoid preservation of motor nerve function during early degeneration. Ann. Neural. 1: 263-269. 24. HOFF, S. F., S. W. SCHEFF,W. Y. KWAN, AND C. W. COTMAN. 198 1. A new type of lesioninduced synaptogenesis: I. Synaptic turnover in nondenervated zones. Brain Res. 222: l-13.
468
SCHEFF, HOFF, AND ANDERSON
25. HOFF, S. F., S. W. SCHEFF, L. S. BENARW, AND C. W. COTMAN. 1982. Lesion-induced synaptogenesis in the dentate gyrus of aged rats. I. Loss and reacquisition of normal synaptic density. J. Comp. Neurol. 205: 246-252. 26. HOFF, S. F., S. W. SCHEFF,AND C. W. COTMAN. 1982. Lesion-induced synaptogenesis in the dentate gyrus of aged rats. II. Demonstration of an impaired degeneration clearing response. J. Comp. Neurol. 205: 253-259. 27. HOLTZMAN, E., AND A. B. NOYIKOFF. 1965. Lysosomes in the rat sciatic nerve following crush. J. Cell Biol. 27: 65 l-669. 28. KELLENBERGER, E., A. RYTER, AND J. SECHAUD. 1958. Electron microscopic study of DNA-containing plasma. II. Vegetative and mature phase DNA as compared with normal bacterial nucleotides in deifferent physiological states. J. Biophys. Biochemm. Cytol. 4: 67 l-678. 29. KIRK, P. E. 1968. Experimental Design: Procedures for the Behavioral Sciences. Pages 8788. Wadsworth, Belmont, California. 30. KOENIG,J. F. R., AND R. A. KLIPPEL. 1963. The Rat Brain. Williams & Wilkins, Baltimore. 3 1. LANDFIELD, P. W., G. ROSE, L. SANDLES, T. WOHLSTADTER, AND G. S. LYNCH. 1977. Patterns of astroglial hypertrophy and neuronal degeneration in the hippocampus of aged memory-deficient rats. J. Gerontol. 32: 3-12. 32. MATTHEWS, D. A., C. W. COTMAN, AND G. LYNCH. 1976. An electron microscopic study of lesion-induced synaptogenesis in the dentate gyms of the adult rat. I. Magnitude and time course of the degeneration. Brain Res. 115: l-2 1. 33. MATTHEWS, D. A., C. W. COTMAN, AND G. LYNCH. 1976. An electron microscopic study of lesion-induced synaptogenesis in the dentate gyms of the adult rat. II. Reappearance of morphologically normal synaptic contacts. Brain Res. 115: 22-4 1. 34. MCEWEN, B. S., J. M. WEISS, AND L. S. SCHWARTZ. 1969. Uptake of corticosterone by rat brain and its concentration by certain limbic structures. Brain Res. 16: 227-24 1. 35. MCWIUIAMS, J. R., AND G. LYNCH. 1984. Synaptic density and axonal sprouting in rat hippocampus: stability in adulthood and decline in late adulthood. Brain Res. 294: 152156. 36. MERKOW, L., M. PARDO, S. M. EPSTEIN, E. AND H. SIDRANSKY. 1968. Lysosomal stability during phagocytosis of Aspergillus flaws spores by aveolar macrophages of cortisoneteated mice. Science 160: 79-8 1. 37. MEYER, J. S., D. J. MICCO, B. S. STEPHENSEON,L. C. KREY, AND B. S. MCEWEN. 1979. Subcutaneous implantation method for chronic glucocorticoid replacement therapy. Physiol. Behav. 22: 867-870. 38. NESTLER, E. J., T. C. RAINBOW, B. S. MCEWEN, ANDP. GREENGARD. 1981. Corticosterone increases the amount of protein 1, a neuron-specific phosphoprotein, in rat hippocampus. Science212: 1162-l 164. 39. PFAFF, D. W., M. T. A. SILVA, AND J. M. WEISS. 197 1. Telemetered recording of hormone effectson hippocampal neurons. Science 172: 394-395. 40. R.-~P, J. P. 1970. Gas chromatographic measurement of peripheral plasma ll-hydroxydeoxycorticosterone in adrenal regeneration hypertension. Endocrinology 86: 668-677. 41. REYNOLDS, E. S. 1963. The use of lead citrate at high pH as an electron-opaque stain in electron microscopy. J. Cell Biol. 17,208. 42. RHEES, R. W., B. I. GROSSER,AND W. STEVENS. 1975. The autoradiographic localization of [“Hldexamethasone in the brain and pituitary of the rat. Brain Res. 100: 15 l-l 56. 43. SALAS, M., AND S. SCHAPIRO. 1970. Hormonal influences upon the maturation of the rat brain’s responsiveness to sensory stimuli. Physiol. Behav. 5: 7-I 1.
CORTICOSTERONE
ALTERS
REACTIVE
SYNAPTOCENESIS
469
44. ScHAPIRO, S. 1968. Some physiological, biochemical and behavioral consequences of neonatal hormone administration: cortisol and thyroxine. Gen. Camp. Endocrinol. 10: 2 14228. 45. SCXEFF, S. W., L. S. BENARW,
AND C. W. COTMAN. 1980. Decline of reactive fiber growth in the dentate gyms of aged rats compared to young adult rats following entorhinal cortex removal. Bruin Res. 199: 21-38. 46. SCHEFF, S. W., L. S. BENARDO, AND C. W. COTMAN. 1980. Hydrocortisone administration retards axon sprouting in the rat dentate gyms. Exp. Nezuol. 68: 195-20 1. 47. S~HEFF, S. W., AND C. W. COTMAN. 1982. Chronic glucocorticoid therapy alters axon sprouting in the hippocampal dentate gyrus. Exp. Neural. 76: 644-654. 48. SCXEFF, S. W., AND S. T. DEKOSKY. 1983, Steroid suppression of axon sprouting in the hippocampal dentate gyrus of the adult rat: dose-response relationship. Exp. Neural. 82: 183-191. 49. SEGAL, M. 1976. Interactions of ACTH and norepinephrine on the activity of rat hippocampal cells. Neuropharmacology 15: 329-333. 50. SILVERMAN, A. J., AND E. A. ZIMMERMAN. 1982. Adrenalectomy increases sprouting in a peptidergic neurosecretory system. Neuroscience 7: 2705-27 14. 5 1. STEWARD, 0. 1976. Reinnervation of dentate gyrus by homologous afferents following entorhinal cortex lesions in adult rats. Science 194: 426-428. 52. STEWARD, O., AND S. L. VINSANT. 1978. Identification of the cells of origin of a central pathway which sprouts following lesions in mature rats. Brain Res. 147: 223-243. 53. STEWARD, O., AND S. L. VINSANT. 1978. Collateral projections of cells in the surviving entorhinal area which reinnervates the dentate gyrus of the rat following unilateral entorhinal lesions. Brain Res. 149: 2 16-222. 54. STEWARD, O., AND S. L. VINSANT. 1983. The process of reinnervation in the dentate gyms of the adult rat: a quantitative electron microscopic analysis of terminal proliferation and reactive synaptogenesis. J. Comp. Neural. 214: 370-386. 55. SYMONS, A. M., D. A. LEWIS, AND R. J. ANCILL. 1969. Stabilizing action ofanti-intlammatory steroids on lysosomes. B&hem. Pharmacol. 1% 258 l-2582. 56. TAYLOR, A. N., G. K. MATHESON, AND N. DAFNY. 197 1. Modification of the responsiveness of components of the limibic-midbrain circuit by corticosteroids and ACTH. Pages 67-78 in C. H. SAWYER AND R. A. GORSIU, E&., Steroid Hormones and Brain Function. Univ. of California Press, Berkley. 57. UNSICKER, K., B. KRISCH, U. &TEN, AND H. THOENEN. 1978. Nerve growth factor-induced fiber outgrowth from isolated rat adrenal chromaffin cells: impairment by glucocorticoids. Proc. Natl. Acad. Sci. U.S.A. 753498-3502. 58. URBAN, I., AND D. DEWIED. 1976. Changes in excitability of the theta activity generating substrate by ACT~rO in the rat. Exp. Brain Res. 24: 325-334. 59. VICEDOMINI, J. P., A. J. NONNEMAN, S. T. DEKOSKY, AND S. W. SCHEFF. 1985. Perinatal glucocorticoids disrupt learning: a sexually dimorphic response. Physiol. Behav. 36: 145149. 60. VICEWMINI, J. P., A. J. NONNEMAN, S. T. DEKOSKY, AND S. W. SCHEFF. 1985. Perinatal glucocorticoids alter dentate gyrus electrophysiology. Bruin Res. Bull. 15: Ill- 116. 6 1. WAREMBOURG, M. 1975. Radioautographic study of the rat brain and pituitary after injection of r3H]dexamethasone. Cell Tissue Res. 161: 183- 19 1. 62. WAREMBOURG, M. 1975. Radioautographic study ofthe rat brain after injection of 1.2 [‘HI corticosterone. Brain Res. 89: 6 l-70. 63. WARR, B. W., J. S. DE OLMOS, AND L. HEIMER. 1981. Horseradish peroxidasez the basic procedure. Pages 207-262 in L. HEIMER AND M. J. ROBARDS, I?.&., Neuroanatomical Tract-TracingMethods. Plenum, New York.
470
SCHEFF, HOFF, AND ANDERSON
64. WEISSMANN, G. 1969. The effects of steroids and drugs on lysosomes. Pages 276-295 in J. T. DINGLE AND H. B. FEu, Fds., Lysosomes in SioZogy undPathology.North-Holland, Amsterdam. 65. WEISSMANN, G., AND J. T. DINGLE. 1961. Release of lysosomal protease by ultraviolet irradiation and inhibition by hydrocortisone. Exp.CellRes.25,201-210. 66. WINER, B. J. 1971. StatisticalPrinciples in Experimental Design.McGraw-Hill, New York. 67. WITHROW, C. D., AND D. M. WOODBURY. 1972. Some aspect of the pharmacology of adrenal steroids and the central nervous system. Pages 41-55 in H. J. REULEN AND K. SCHURMANN, Eds.,Steroids andBrainEdema.Springer-Verlag, New York. 68. ZHOU, F. C., AND E. C. &MITIA. 1985. The effect of adrenalectomy and corticosterone on homotypic collateral sprouting of serotonergic fibers in hippocampus. Neurosci. Left. 54: 111-116.