EXPERIMENTAL

NEUROLOGY

115,271-281

(19%)

Alterations in [3H]Kainate and ~-methyl-o-Aspa~ate-Sensitive Glutamate Binding in the Rat Hippocampal Formation following Fimbria-Fornix Lesions JAMES W. GEDDES,*T~ LANCE BRUNNER,? CARL W. *Division

AND GYORGY BUZS~KI$

of~eurasurge~ and tDepartment of Psychobiology, University of Cal~for~iu, Irvine, California 92717; and $Center and ~e~uuioral ~earosc~e~ce, Rutgers university, 195 ~~~vers~ty AueR~e, cedars, New Jersey 07102

INTRODUCTION a lesion

of the ~mbria-fornix

(FE‘), interic-

tal spikes develop and susceptibility

to seizures in-

’ To whom correspondence should be addressed Center on Aging, 209 Sanders-Brown Building, tucky, Lexington, Kentucky 40536-0230.

at Sanders-Brown University of Ken-

for Molecular

creases (7). The electrophysiological and behavioral alterations take approximately 2 to 3 weeks to develop and persist for several months (6). The delayed induction of seizure vulnerability is similar to the time course for the sprouting of axon collaterals and reestablishment of synaptic connections following lesions (9). This suggests the possibility that the epileptiform activity may result from the sprouting and rearrangement of excitatory and inhibitory circuits within the deafferented hippocampus (6). Lesions of the FF transect. the cholinergic and GABAergic afferents from the septal area in addition to several afferent pathways from other subcortical nuclei (see (7, 34, 52)). The cholinergic septohippo~ampal fibers terminate on all types of hippocampal neurons, whereas the GABAergic fibers terminate on GABAergic interneurons in the hippocampus and dentate gyrus (18). The cholinergic fibers facilitate, and the GABAergic fibers inhibit, the excitatory loop initiated by afferent volleys arriving from the entorhinal cortex {see (7, 34)). In addition to the FF, the lesion severs the commissural connections between CA3, CAl, and hilar cells. Electrophysiological studies demonstrate that following FF lesions, the deafferented hippocampus responds in an exaggerated manner to perforant path activity. This is likely the result of a reduction in feedforward inhibition and strengthening of both feedback inhibition and excitation (6). To investigate possible alterations in excitatory intrahippocampal circuits following deafferentation induced by aspiration lesions of the FF, we used receptor autoradiography to examine the density and distribution of excitatory amino acid receptors. The results demonstrate that FF lesions results in timedependent alterations in the density and distribution of hippocampal [3H]kainic acid ([3H]KA) and N-methylD-aspartate (N~DA)-sensitive L-[3H]glutamate binding sites and that the KA and N&IDA receptors differ in their response to the deafferentation.

Following lesions of the fimbria-fornix, there is a time-dependent increase in interictal spikes and seizure susceptibility. This may result from sprouting of local excitatory and inhibitory circuits in response to the loss of subcortical and commissural innervation of the hippocampal formation. We used receptor autoradiography to examine the density of N-methyl-D-aspart&e (NMDA)-sensitive L-[‘HIglutamate and [3H]kainate (KA) binding sites in the hippocampal formation at 5 days, 3 months, and 1 year following bilateral aspiration lesions of the fimbria-fornix. At 5 days postlesion, the CA3 and CA1 strata radiatum and oriens displayed a decrease (20-42%, P < 0.01) in NMDA-sensitive L-[3H]glutamate binding. The initial decrease was followed by a moderate recovery at later time points but was still evident at 1 year postlesion. This may reflect a lesion-induced turnover of synaptic complexes, downregulation of postsynaptic receptors, or loss of presynaptic receptors. Five days following fimbria-fornix lesion there was also a decrease (13-15%, P < 0.05) in 13H]KA binding in CA3 strata radiatum and pyramidale. However, at 3 months postlesion KA receptor density was elevated by 29-33% (P < 0.01) in the outer molecular layer of the dentate gyrus with no significant change in binding to the inner molecular layer. By 1 year postlesion, the density of t3H]KA binding sites was not significantly different from that observed in control animals of the same age. The increase in KA receptor density in the outer molecular layer 3 months after fimbria-fornix lesion may reflect sprouting of the perforant path input or mossy fibers to this region and contribute to the increase in interictal spikes and seizure susceptibility. 8 1992 Academic Press, Inc.

Following

COTMAN,?

L-[~H]-

EXPERIMENTAL

PROCEDURES

These experiments used female Sprague-Dawley rats (225-300 g, n = 24). For FF lesions, 12 animals were 271 Ail

Copyright Q 1992 rights of reproduction

0014-4886192 $3.00 by Academic Press, Inc. in any form reserved.

272

GEDDES

anesthetized with a mixture (4 ml/kg) of ketamine (25 mg/ml), xylazine (1.3 mg/ml), and acepromazine (0.25 mg/ml). The FF lesion was made bilaterally by aspirating the medial portion of the parietal cortex and cingulate cortex, the cingulate bundle, the supracallosal stria, the corpus callosum, the dorsal fornix, the fimbria, and the ventral hippocampal commissure. Four rats were killed by decapitation at each of the following survival times: 5 days, 90 days, and 1 year. The brains were immediately removed, frozen in powdered dry ice, and stored at -70°C until required. Control groups, each consisting of four unoperated animals the same age as the lesion group, were included at each survival time. For autoradiography, 6-pm coronal brain sections were cut on a cryostat and thaw-mounted onto gelatin coated microscope slides. Twenty-micrometer sections were also obtained for cresyl violet staining and acetylcholinesterase histochemistry (43). The completeness of the lesions was confirmed by the absence of acetylcholinesterase-positive fibers in the dorsal hippocampus (Fig. 1). The 6-/*rn tissue sections were stored at -20°C overnight and then processed for autoradiography using previously published methods (25, 50, 51). Slides were thawed and placed into 50 mM Tris-acetate buffer, pH 7.2 (for NMDA receptors) or into 50 mM Tris-citrate, pH 7.0 (for KA receptors) for at least 1 h at 0-2°C. The tissue sections were then incubated for 10 min at 30°C in the same buffer to further remove endogenous glutamate and various ions. The tissue sections were then incubated under conditions designed to specifically label the two receptor subclasses. For binding to the KA subclass of excitatory amino acid receptors, tissue sections were incubated for 30 min at 0-4°C in 50 mM Tris-citrate containing 50 nM [3H] kainic acid (60 Ci/mmol, New England Nuclear, Boston, MA). Nonspecific binding was determined by including 100 PM KA in the incubation buffer. Saturation analysis of [3H]KA binding was evaluated using five ligand concentrations ranging between 25 and 400 nM. Ligand concentrations lower than 25 nM could not be accurately assessed due to the low resolution of the autoradiograms. To label the NMDA receptor, tissue sections were incubated for 10 min at 0-4°C in 50 mM Tris-acetate buffer, pH 7.2, containing 100 nM L-[3H]glutamate (50 Ci/mmol, ICN, Irvine, CA, or New England Nuclear, Boston, MA). Quisqualate (10 PM) and the chloride channel, blocker 4-acetamido-4’-isothiocyano-2,2’-disulphinic acid (SITS, 100 PM) were also included in the incubation buffer to minimize binding to non-NMDA sites. Nonspecific binding was determined by including 200 pM NMDA in the incubation buffer. Following incubation, unbound radioactivity was removed by rinsing the slides in a series of four Coplin jars with ice-cold buffer for a total time of 30 s. Sections

ET

AL.

were then dried under a stream of cool air and placed into X-ray cassettes with tritium-sensitive film (LKB instruments, Gaithersburg MD; or Amersham, Arlington Heights, IL). Films were exposed for 3-4 weeks at 4”C, followed by development in Kodak D-19 at 20°C. Receptor localization was evaluated by comparison with corresponding 20-pm sections stained with cresyl violet. Autoradiograms were analyzed by computer-assisted densitometry (Imaging Research, St. Catherines, Ontario, Canada). The tissue equivalent was calibrated using [3H]methacrylate tritium standards (Amersham, Arlington Heights, IL). For each hippocampal region investigated (dentate gyrus molecular layer; hilus; CA3 strata oriens, pyramidale, lucidum, and radiatum; CA1 strata oriens, pyramidale, and radiatum; subiculum, parahippocampal gyrus) measurements were taken throughout the entire extent of the layer to ensure a representative sample. At least six readings per single brain section were made for each region studied. For each individual animal, readings were obtained from three brain sections and averaged. Results are presented as the mean f SEM for the results obtained from four animals. Statistical analyses were performed using Student’s t test (unpaired, 2-tailed) and one-way analysis of variance (ANOVA) with differences between individual groups estimated using the Scheffe F test. The possible influence of shrinkage on receptor density was evaluated by determining the width of the molecular layer in three control and three FF lesion animals at each of the time points. Measurements were taken directly from the KA autoradiograms and three measurements were taken for each animal.

RESULTS [3H]Kuinic

Acid Binding

Sites

In control animals, the greatest density of [3H]KA binding sites in the rat hippocampal formation was in CA3 s. lucidum. Other regions with considerable binding included the hilus and the inner one-third of the molecular layer of the dentate gyrus (Figs. 2A and 3A). Low [3H]KA binding densities were observed throughout the remainder of the hippocampus. The density of KA receptors was relatively constant in the three control groups although in the l-year control group (15 months of age) slight decreases in binding (P < 0.05) were observed in areas CA3 s. oriens, CA1 strata oriens and radiatum, and the outer molecular layer of the dentate gyrus. [3H]KA binding in the hilus of the l-year control animals was also decreased relative to the 3month but not the 5-day group (P < 0.05). In the inner molecular layer of the dentate gyrus, the density of [3H]KA binding sites was slightly elevated in the 3-

13H]KA

AND

NMDA-SENSITIVE

L-[3H]GLUTAMATE

BINDING

273

FIG. 1. The effect of fimbria-fornix lesion on acetylcholinesterase (AChE) staining in the hippocampal formation. AChE staining is markedly reduced in the rat hippocampal formation 1 year following a lesion of the fimbria-fornix (B) as compared to the staining observed in a control animal of the same age (A). Hippocampal AChE staining was also absent 5 days and 3 months following the fimbria-fornix lesion (data not shown).

month as compared to the 5-day control group (Fig. 3A, P < 0.05). At 5 days following the FF lesions, the hipp~ampal density of [3H]KA binding sites was similar to that observed in control animals of the same age except for a significant decrease in CA3 s. oriens (13%, P < 0.05) and CA3 s. radiatum (15%, P < 0.05) (Figs. 2B and 4A). At 3 months postlesion, the density of [3H]KA binding sites was elevated by 33% in the outer two-thirds of the molecular layer of the ventral blade, and by 29% in the dorsal blade, of the dentate gyrus (P c 0.01) (Figs. 2C and 4A). No significant changes in binding density were observed in the deafferented inner molecular layer. A 15% decrease (P < 0.05) in [3H]KA binding density was observed in CA1 s. radiatum, whereas binding densities were similar t,o control values in CA1 s. oriens and other areas of the hippocampal formation. At 1 year postlesion, there was a trend toward increased densities of 13H]KA binding sites in the outer molecular layer of the

dentate gyrus and in CA1 s. oriens although the differences were not significant when compared to control animals {Figs. 2D and 4A). The density and affinity of 13H]KA binding in the ventral blade outer molecular layer was evaluated using saturation analysis in one control and one FF lesion animal at the 3-month time point. Analysis of saturation binding data was performed using the EBDA/LIGAND computer program (Elsevier-BIOSOFT, Cambridge, UK). A single-site model provided the best fit for the data. In the outer molecular layer of an unoperated control animal, the apparent dissociation constant was 7.5 nM, similar to values obtained previously in the rat (50) and human (10) brain. The maximal binding density was 0.45 pmollmg protein, also similar to the values reported previously (50). Three months following FF lesion the maximal binding density was elevated by 27% (0.57 pmol/mg protein), whereas the & value was similar (6.9 n&f) to that observed in the control animal.

GEDDE

FIG. 2. Distribution of f3H]KA (A-D) and i.#H]glutamate {E-H used, L-[3H]glutamate binding was mainly to NMDA-sensitive sites animals of the same age as the 5-day postlesion group. B and F repro correspond to the S-month postlesion, and D and H to the l-year posl layer of the dentate gyrus is most evident at 3 months and is indicat glutamate binding in CA1 strata radiatum and oriens is most apparel heads in F. Abbreviations: DC, dentate gyrus; H, hilus; g, granule cells; s. radiatum. p, s. pyramidale; or, s. oriens. Scale bar = 100 grn.

S ET

AL.

L) binding sites in the rat hippocampal formation. Under the conditions (see Experimental Procedures). A and E are from unoperated control ?sent binding observed 5 days following fimbria-fornix lesion, C and G ilesion animals. The increase in [“H]KA binding in the outer molecular ed by an arrowhead in the ventral blade in C. The decraase in L-[~H]It 5 days following the fimbria-fornix lesion as indicated by the arrowm, molecular layer of the dentate gyrus; lm, s. laeunosum-mole~ulare; r,

[3H]KA

AND

NMDA-SENSITIVE

L-[3H]GLUTAMATE

r

2500

275

BINDING

m

A

q q

5 day 3 month 1 year

1000

500

n

”

DC-outer

DG-inner

Hi/us

CA3

or

CA3

/UC

CA1

or

CA1

rad

Region 2500

2000

1500

1000

500

0 DG-outer

DC-inner

Hi/us

CA3

rad

CA3

or

CA1

rad

CA1

or

Region

FIG. 3. Density of hippocampal (A) [aH]KA and (B) L-[3H]glutamate binding in control animals. The solid black bars represent binding densities in the control group of the same age as the 5-day postlesion animals. The white bars with black stripes bars correspond to animals paired to the 3-month postlesion group, and the black bars with white stripes correspond to animals paired to the l-year postlesion group. Values represent specific binding and are the mean +- SEM, n = 4, for each group. The values shown for the inner and outer molecular layer of the dentate gyrus were obtained from the ventral blade. The values obtained from the dorsal blade were similar although slightly lower. Abbreviations: DG, molecular layer of the dentate gyrus with inner corresponding to the inner one-third and outer to the outer two-thirds; or, s. oriens; rad, s. radiatum; luc, s. lucidum. *P < 0.05, ** P < 0.01 with respect to values in corresponding hippocampal regions of the 5-day control group. $ P < 0.05 as compared to values of the 3-month postlesion control group (Student’s t test, 2-tailed).

276

GEDDES 1M: I-

ET

AL.

A 6.

121 i -

15&Y

q

3 month 1 year

T

T

I -

, -

I -

25 ,

aI-

DO-outer

DO-Inner

HI/US

CA3

or

CA3

luc

CA1

or

CA1

red

Region 1%

B

5 day 3 month 1 year

125

2!

DO-outer

DG-inner

Hilus

CA3

or

CA3

rad

CAI

or

CA1

rad

Region

FIG. 4. Density of (A) [3H]KA and lesion. Data (mean & SEM) are expressed black bars represent binding densities postlesion animals, and the black bars animals of the same age: * P c 0.05, **

(B) L-[3H]glutamate binding in the hippocampal formation at various times following fimbria-fomix as a percentage of specific binding relative to unoperated controls of the same age (n = 4). The solid 6 days following fimbria-fomix lesion. The white bars with black stripes correspond to the 3-month with white stripes correspond to the l-year postlesion group. Significantly different from the control P < 0.01 (Student’s t test, l-tailed).

[3H]KA

TABLE Width

of the Molecular

AND NMDA-SENSITIVE

1

Layer of the Dentate

Gyrus

Group

5 day

3 month

1 year

Control Lesion

220 f 10 pm 200 + 20 pm

200 * 10 pm 190 * 10 pm

190 + 20 Frn 180 * 30 firn

Note. Results represent an average of three measurements from the ventral blade of the dentate gyrus. The control animals were the same age as the corresponding lesion group (see Experimental Procedures). Values represent the mean *SD. n = 3.

N-Methyl-o-Aspartate-Sensitive Binding Sites

L-13HjGlutamate

In unoperated control animals, the highest density of binding sites was in s. radiatum and oriens of CA1 and the molecular layer of the dentate gyrus. Moderate binding densities were observed throughout the remainder of the hippocampus (Figs. 2E and 3B). This distribution was similar to that reported in previous studies (Z&41). In the 3-month as compared to the 5-day control group, binding densities were elevated in s. radiatum and oriens of both CA3 and CA1 and in the inner molecular layer of the dentate gyrus. In the 1 year control group, the density of NMDA-sensitive ~-t3H]glutamate binding sites remained elevated in these regions and in addition the binding density in the outer molecular layer of the dentate gyrus was significantly elevated (P < 0.05) compared to the 5-day animals (Fig. 3B). Following aspiration lesions of the FF, significant decreases in the density of NMDA-sensitive L-[~H]glutamate binding sites were observed. At 5 days postlesion, NMDA receptor density was decreased by approximately 20% in CA3 s. oriens and radiatum (P < 0.01) and by 40% in CA1 s. oriens and radiatum (P < 0.01) (Fig. 4B). The loss of NMDA binding was still evident in CA1 at both the 3-month and l-year postlesion intervals, whereas the alterations in CA3 were no longer significant by 3 months although a slight decrease in binding density in CA3 radiatum was evident at l-year postlesion (P < 0.05). Shrinkage

The width of the molecular layer was slightly less in the lesion group as compared to the control animals at each time point, although the differences were not significant (Table 1). Age-dependent atrophy of the molecular layer was also evident in both the lesion and control groups (Table 1) although again the changes were not significant. DISCUSSION Control Animals

The results demonstrate significant age-related differences in the density of f3H]KA and NMDA-sensitive

L-[3H]GLUTAMATE

277

BINDING

L-[3H]glutamate binding sites in the three control groups used in this study. The FF lesion results were therefore compared to those obtained in unoperated animals of the same age. The &day FF lesion and control group represents animals approximately 100 days of age and the two other groups of control and FF-lesioned animals were approximately 6 months and 15 months of age. The density of [3H]KA bin~ng sites was relatively stable in the three control groups although changes were observed in the molecular layer of the dentate gyms in the 3-month as compared to the 5-day control group, and a slight loss of binding sites was evident in CA3 s. oriens and CA1 s. oriens and radiatum in the l-year (age 15 month) control group (Fig. 3A). The effect of age on [‘H]KA binding has not been examined previously. L-13H]Glutamate binding to NMDA-sensitive sites exhibited an age-dependent increase in binding in several hippocampal regions including the dentate gyrus, CA3, and CA1 (Fig. 3B). This ligand labels an agonist-preferring population of NMDA receptors (42). Previous studies have demonstrated a preservation of antagonist-preferring NMDA receptors, labeled by t3H]CPP, during aging in the rat brain (33) and there are conflicting results regarding the loss (33, 39) or preservation (4) of strychnine-insensitive $H]glycine binding to the NMDA receptor complex during aging. Fimbria-Fornix

Lesions

The results from this study demonstrate that hippocampal NMDA and KA receptors differ in their response to bilateral aspiration lesions of the fimbria fornix. The alterations in ligand binding are thought to reflect the deafferentation and reactive synaptogenesis induced by the FF lesions. The lesion removes the septohippocampal input to the hippocampal formation and dentate gyrus (12) and also transects the ventral hippocampal commissure whose projections normally innervate the inner one-third of the dentate gyrus and CA1 and CA3 s. oriens and the middle portion of s. radiatum (26). Following FF lesion an increase in glial fibrillary acidic protein immunoreactivity, indicative of areas of afferent fiber degeneration, is observed in CA3 s. oriens and s. pyramidale as well as in the inner molecular layer of the dentate gyrus (20). NMDA-Sensitive

L-[3H]Glutamate

Binding

The initial loss of NMDA-sensitive L-[3H]glutamate binding in CA3 and CA1 strata oriens and radiatum may be the result of a lesion-induced turnover of synaptic complexes, downregulation of postsynaptic receptors, or possibly a loss of presynaptic’receptors as fibers degenerate. NMDA receptors are generally considered to be postsynaptic (2), although presynaptic NMDA receptors have also been h~othesized (16).

278

GEDDES ET AL.

Following unilateral lesions of the perforant path, an increase in NMDA receptor density is observed in both the inner and outer molecular layer of the ipsilateral dentate gyrus at 21 days postlesion, persisting until at least Postlesion Day 60 (51). In the entorhinal lesion paradigm, there is sprouting of both the septohippocampal and commissural/associational pathways into the deafferented molecular layer (35,53). In the present study, both the septohippocampal and commissural inputs are removed by the FF aspiration. The loss of L[3H]glutamate binding in the FF-lesioned animals, and the increase in binding to the entorhinal cortex-lesioned animals, could be explained by the presence of presynaptic NMDA receptors on, or regulation of postsynaptic NMDA receptors by, either the septohippocampal or associational pathways. Although there is some recovery, the density of L[3H]glutamate binding to NMDA-sensitive sites does not return to control values even 1 year following the FF lesions. The mechanism(s) underlying the partial recovery is uncertain, but one possibility is shrinkage of the hippocampal formation which would result in an apparent increase in the density of binding sites. The volume of the hippocampal formation was not quantitated in the present study but the hippocampus obtained from animals 1 year following FF lesion appeared smaller than in unoperated animals of the same age.

The initial decrease (5-day postlesion) in [3H]KA binding in the CAl/CA3 strata oriens and radiatum could also reflect a loss of presynaptic receptors, downregulation, or increased turnover of postsynaptic receptors. Although the specific pre- vs postsynaptic localization of KA receptors has not been established (40, 45), the lack of change in [3H]KA binding density in the deafferented inner molecular layer of the dentate gyrus at 5 days postlesion suggests that KA receptors are not presynaptically located on the septohippocampal or commissural fiber terminals and that they may be located postsynaptically. The increase in [3H]KA binding in the outer molecular layer of the dentate gyrus at 3 months postlesion parallels the incidence of interictal spikes (6) and the disappearance of long-term potentiation of the population spike in the dentate gyrus (5,13). It is unlikely that the alterations in [3H]KA binding are the result of the interictal spikes and increased vulnerability to seizures since a decrease in [3H]KA binding in the hippocampal formation has been observed in kindled (47) and in genetically epilepsy-prone rats (38). Limbic seizures also result in the decreased expression of a putative kainate receptor mRNA (21). Moreover, the restricted localization of the increase in [3H]KA binding to the outer molecular layer of the dentate gyrus argues against a gen-

era1 upregulation of KA receptor density resulting from the increase in interictal spikes. The specific increase in [3H]KA binding in the outer molecular layer of the dentate gyrus is thought to reflect sprouting, analogous to the increase in KA receptor density in the molecular layer of the dentate gyrus following lesions of the perforant path (25, 50). The delayed emergence of interictal spikes and the time-dependent increase in paired pulse suppression following the FF lesion is also suggestive of sprouting (6). Sprouting has been shown previously in human temporal lobe epilepsy and in the kindling and kainate lesion models. In most studies, however, this has been restricted to sprouting of mossy fibers or GABAergic neurons into the supragranular/inner molecular layer of the dentate gyrus (1, 14, 31, 49). The mossy fiber sprouting is induced by a loss of polymorphic hilar neurons, the source of the commissural/associational pathway which terminates in the inner molecular layer of the dentate gyrus and which represents one target of the dentate granule cells. The GABAergic sprouting may be the result of a loss of GABAergic interneurons in the animal models. In human temporal lobe epilepsy, glutamate-decarboxylase immunoreactive neurons are preserved (1) although there is a loss of somatostatin and neuropeptide Y immunoreactive interneurons accompanied by sprouting of remaining neuropeptide Y neurons into the inner molecular layer of the dentate gyrus (15). The status of the polymorphic hilar neurons in the FF lesion model is uncertain but there is a loss (40%) of parvalbumin-immunoreactive, presumably GABAergic, neurons in the hilus 2 months after the lesion (8). In addition, there is a partial loss of parvalbumin-immunoreactive neurons in the dentate gyrus and CA3. It is possible that sprouting induced by the loss of polymorphic or GABAergic hilar neurons indirectly induces alterations in [3H]KA binding in the outer molecular layer. However, we would then have to assume that sprouting mossy fibers or remaining GABAergic neurons avoid the deafferented inner molecular layer for reasons unknown at present. The increase in [3H]KA binding cannot result from sprouting of commissural fibers since these connections are severed by the bilateral FF lesions. Sprouting of the ipsilateral associational system is a possibility, although these fibers do not normally extend into the outer molecular layer following FF lesions (35, 53). Another possible explanation for the increase in [3H]KA binding density in the outer molecular layer of the dentate gyrus is sprouting of the perforant path input to this region. Previously we have observed an intensification of SNAP-25, a novel presynaptic marker, in the outer molecular layer of the dentate gyrus following deafferentation of the inner molecular layer by kainic acid-induced destruction of the commissurallassociational pathways (24). The FF lesion also partially

[aH]KA

AND NMDA-SENSITIVE

deafferents the inner molecular layer, transecting both the septohippocampal and commissural inputs to this region. This may indirectly induce sprouting of the perforant pathway which would result in increased excitability of the hippocampus and possibly allow hypersynchronous activation of a large population of neurons and reverberation of information in the entorhinal-hippocampal-entorhinal cortex circuitry. The increased KA receptors are unlikely to be provided by sprouting sympathetic fibers which originate in the superior cervical ganglion and are normally confined to extracerebral arteries. Following fimbrial transection sympathetic fibers innervate the supragranular and infragranular region of the dentate gyrus but do not ramify in the outer two-thirds of the molecular layer (11). Moreover, sympathetic fibers do not contribute to the hyperexcitability of the FF-lesioned hippocampus, since removal of the superior cervical ganglia does not alter the incidence or morphology of interictal spikes (6, 30). Intrinsic cholinergic neurons are also unlikely to participate since these neurons do not sprout following partial deafferentation of the hippocampus (3,17). The absence of acetylcholinesterase staining in the dorsal hippocampus (Fig. 1) also rules out the possibility that subcortical fibers innervating in the ventral hippocampus may sprout into the dorsal hippocampus. These fibers originate in the ventral limb of the diagonal band and project via the ventral amygdalafugal pathway to the ventral hippocampus. Despite sprouting of these fibers they do not reach the septal one-third of the hippocampus even after FF lesion (19). The large increase in E3H]KA binding 3 months after the FF lesions is surprising given the relatively minor contribution of the septohippocampal pathway to synaptic density in the outer molecular layer. Although the number of synapses contributed by this pathway has not been quantitated, the perforant path and crossed perforant path contribute approximately 90% of the synapses in the outer molecular layer (29, 36, 48). The remaining 10% of the synapses include those occupied by the septophippocampal pathway, other subcortical inputs, and interneurons. Thus, the increase in [3H]KA binding exceeds the magnitude of the deafferentation of the outer molecular layer. This may reflect hyperinnervation, as has been reported in the middle molecular layer following entorhinal cortex lesions (37), or structural plasticity in addition to the sprouting of axon collaterals (44). The return to control values of [3H]KA binding density in the outer molecular layer of the dentate gyrus at the l-year postlesion time point may result from a correction of the hyperinnervation, additional degeneration (Buzsaki and Freund, unpublished observations), or other unknown mechanisms. An increase in KA receptor density in the molecular layer of the dentate gyrus has been observed in patholog-

L-[3H]GLUTAMATE

BINDING

279

ical cases that involve the human temporal lobe, including Alzheimer’s disease (22, 25) and temporal lobe epilepsy (23). The subcortical input to the hippocampus is not known to be compromised in temporal lobe epilepsy, although a loss of acetylcholinesterase staining in the inner molecular layer of the dentate gyrus has been described (27). In Alzheimer’s disease, an increase in KA receptor density in the outer molecular layer of the dentate gyrus has also been observed in some in~viduals (22,25). There can be extensive neuropathology in several subcortical nuclei in Alzheimer’s disease, particularly the basal forebrain (32). It is of interest that advanced Alzheimer’s disease patients are at risk for developing seizures (46). The results of the present study would predict that dementia patients with severe subcortical pathology may have a higher incidence of interictal spikes and be more vulnerable to seizures than other Alzheimer’s disease patients. To summarize, our findings indicate that partial deafferentation of the hippocampus resulting from aspiration lesions of the FF result in a substantial and continuous reorganization of the intrinsic circuitry. The time course of the physiological and anatomical changes appear similar, although the causal relationship between abnormal physiological patterns and the changes and/or NMDA-sensitive L-[~H]in [3H]kainate glutamate binding has yet to be determined. ACKNOWLEDGMENTS We thank Dr. Jolanta Ulas for helpful discussions, Dr. Charles Ribak for critical reviews of the manuscript, and Suzanne Cooper for excellent technical assistance. This work was supported by NINDS (NS27058, NS 28121), the ADRDA (G.B.), and the American Epilepsy Society (J.W.G.). REFERENCES 1. BABB, T. L., J. K. PRETORIUS, W. R. KUPFER, AND P. H. C&ANDALL. 1989. Glutamate decarboxylase-immunoreactive neurons are preserved in human epileptic hippocampus. J. Neurosci. 9: 2562-2574. 2. BEKENSTEIN, J. W., J. P. J. BENNETT, G. F. WOOTEN, AND E. W. LOTHMAN. 1990. Autora~o~aphi~ evidence that NMDA receptor-coupled channels are located postsynaptically and not presynaptically in the perforant path-dentate granule cell system of the rat hippocampal formation. Brain Res. 514: 334-42. 3. BLAKER, S. N., D. M. ARMSTRONG, AND F. H. GAGE. 1988. Cholinergic neurons within the rat hippocampus: Response to fimbria-fornix transection. J. Camp. Neurot. 272: 127-138. 4. BONHAUS, D. W., W. B. PERRY, AND J. 0. McN~A~A. 1990. Decreased density, but not number, of N-methyl-D-aspartate, glycine and phencyclidine binding sites in hippocampus of senescent rats. Brain Res. 532: 82-86. 5. BUZSAKI, G., AND F. H. GAGE. 1989. Absence of long-term potentiation in the subcortically deafferented dentate gyms. Brain Res. 484: 94-101. 6. BUZSAKI, G., M. Hsu, C. SLAMKA, F. H. GAGE, AND 2. HORVATH.

280

7.

8.

9. 10.

11.

12.

13.

14.

15.

16.

17.

18. 19.

20.

21.

22.

23.

24.

GEDDES 1991. Emergence and propagation of interictal spikes in the subcortically denervated hippocampus. Hippocampus 2: 163-180. BUZSAKI, G., G. L. PONOMAREFF, F. BAYARDO, R. RUIZ, AND F. H. GAGE. 1989. Neuronal activity in the subcortically denervated hippocampus: A chronic model for epilepsy. Neuroscience 28: 527-538. CHEN, L. S., T. F. FREUND, G. BUZSAKI, K. D. BAIMBRIDGE, AND F. H. GAGE. 1989. Reduction of parvalbumin-immunoreactivity in the subcortically denervated hippocampus. Sot. Neurosci. Abstr. 16: 700. COTMAN, C. W., AND M. NIETO-SAMPEDRO. 1984. Cell biology of synaptic plasticity. Science 225: 1287-1293. COWBURN, R. F., J. A. HARDY, R. S. BRIGGS, AND P. J. ROBERTS. 1989. Characterisation, density, and distribution of kainate receptors in normal and Alzheimer’s diseased human brain. J. Neurochem. 52: 140-147. CRUTCHER, K. A. 1987. Sympathetic sprouting in the central nervous system: A model for studies of axonal growth in the mature mammalian brain. Bruin Res. Reu. 12: 203-233. CRUTCHER, K. A., R. MADISON, AND J. N. DAVIS. 1981. A study of the rat septohippocampal pathway using anterograde transport of horseradish peroxidase. Neuroscience 6: 1961-1973. CZEH, G., Z. HORVATH, AND J. CZOPF. 1990. Long-term potentiation in slices from subcortically deafferented hippocampi. Brain Res. 518:279-282. DAVENPORT, C. J., W. J. BROWN, AND T. L. BABB. 1990. Sprouting of GABAergic and mossy fiber axons in dentate gyrus following intrahippocampal kainate in the rat. Exp. Neurol. 109: 180190. DE LANEROLLE, N. C., J. H. KIM, R. J. ROBBINS, AND D. D. SPENCER. 1989. Hippocampal interneuron loss and plasticity in human temporal lobe epilepsy. Bruin Res. 1989: 387-395. FINK, K., H. BBNISCH, AND M. G~THERT. 1990. Presynaptic NMDA receptors stimulate noradrenaline release in the cerebral cortex. Eur. J. Pharmacol. 185: 115-117. FROTSCHER, M. 1988. Cholinergic neurons in the rat hippocampus do not compensate for the loss of septohippocampal cholinergic fibers. Neurosci Lett. 87: 18-22. FROTSCHER, M. 1991. Target cell specificity of synaptic connections in the hippocampus. Hippocampus 2: 123-130. GAGE, F. H., A. BJSRKLUND, AND U. STENEVI. 1983. Functional correlates of compensatory sprouting by biogenic amines and cholinergic afferents in the hippocampal formation. Bruin Res. 268: 27-37. GAGE, F. H., P. OLEXNICZAK, AND D. M. ARMSTRONG. 1988. Astrocytes are important for sprouting in the septohippocampal circuit. Exp. Neurol. 102: 2-13. GALL, C., K. SUMIKAWA, AND G. LYNCH. 1990. Levels of mRNA for a putative kainate receptor are affected by seizures. Proc. Nutl. Ad. Sci. USA 87: 7643-7647. GEDDES, J. W., L. C. BRUNNER, J. ULAS, AND C. W. COTMAN. 1990. Hippocampal excitatory amino acid receptors in Alzheimer’s disease. II. Non-NMDA receptors. Sot. Neurosci. Abst. 16: 611. GEDDES, J. W., L. D. CAHAN, S. M. COOPER, R. C. KIM, B. H. CHOI, AND C. W. COTMAN. 1990. Altered distribution of excitatory amino acid receptors in temporal lobe epilepsy. Exp. Neural. 108:214-220. GEDDES, J. W., E. J. HESS, R. A. HART, J. P. KESSLAK, C. W. COTMAN, AND M. C. WILSON. 1990. Lesions of hippocampal circuitry define synaptosomal-associated protein-25 (SNAP-25) as a novel presynaptic marker. Neuroscience 38: 515-525.

ET AL. 25. GEDDES, J. W., D. T. MONAGHAN, C. W. COTMAN, I. T. Lo=, R. C. KIM, AND H. C. CHUI. 1985. Plasticity of hippocampal circuitry in Alzheimer’s disease. Science 230: 1179-1181. 26. GOTTLIEB, D. I., AND W. M. COWAN. 1973. Autoradiographic studies of the commissural and ipsilateral association connection of the hippocampus and detentate gyrus of the rat. I. The commissural connections. J. Comp. Neurol. 149: 393-422. 27. GREEN, R. C., H. W. BLUME, S. B. KUPFERPCHMID, AND M.-M. MESULAM. 1989. Alterations of hippocampal acetylcholinesterase in human temporal lobe epilepsy. Ann Neural. 26: 347-351. 28. GREENAMYRE, J. T., J. M. M. OLSEN, J. B. PENNEY, AND A. B. YOUNG. 1985. Autoradiographic characterization of N-methylD-aspartate-, quisqualate-and kainate-sensitive glutamate binding sites. J. Phurmucol. Exp. Ther. 233: 254-263. 29. HOFF, S. F., S. W. SCHEFF, L. S. BERNARDO, AND C. W. COTMAN. 1982. Lesion-induced synaptogenesis in the dentate gyrus of aged rats. I. Loss and reaquisition of normal synaptic density. J. Comp. Neurol. 205: 246-252. 30. HORVATH, Z., S. BERGER, M. Hsu, F. H. GAGE, AND G. BUZSAKI. 1989. The subcortically deafferented hippocampus: epileptic patterns and plastic properties. Sot. Neurosci. Abstr. 15: 700. 31. HOUSER, C. R., J. E. MNASHIRO, B. E. SWARTZ, G. 0. WALSH, J. R. RICH, AND A. V. DELGADO-ESCUETA. 1990. Altered patterns of dynorphin immunoreactivity suggest mossy fiber reorganization in human hippocampal epilepsy. J. Neurosci. 10: 267-282. 32. JELLINGER, K. 1987. Quantitative changes in some subcortical nuclei in aging, Alzheimer’s disease and Parkinson’s disease. Neurobiol. Aging 8: 556-561. 1990. Influence of age on 33. KITO, S., R. MNOSHI, AND T. NOMOTO. NMDA receptor complex in rat brain studied by in vitro autoradiography. J. Histochem. Cytochem. 38: 1725-1731. 34. LOPES DA SILVA, F. H., M. P. WITTER, P. H. BOELJINGA, AND A. H. M. LOHMAN. 1990. Anatomic organization and physiology of the limbic cortex. Physiol. Reu. 70: 453-511. 35. LYNCH, G., C. GALL, G. ROSE, AND C. W. COTMAN. 1976. Changes in the distribution of the dentate gyrus associational system following unilateral or bilateral entorhinal lesion in the adult rat. Bruin Res. 110: 57-71. 36. MATHEWS, D. A., C. COTMAN, AND G. LYNCH. 1976. An electron microscopic study of lesion-induced synaptogenesis in the dentate gyrus of the adult rat. I. Magnitude and time course of degeneration. Bruin Res. 115: 1-21. 37. MATTHEWS, D. A., C. COTMAN, AND G. LYNCH. 1976. An electron microscopic study of lesion-induced synaptogenesis in the dentate gyrus of the adult rat. II. Reappearance of morphologically normal contacts. Bruin Res. 115: 23-41. 38. MILLS, S. A., S. RAZANI-BOROUJEF~DI,C. E. REIGEL, P. C. JOBE, AND D. D. SAVAGE. 1990. Decrease in hippocampal [3H]vinylidene kainic acid binding in genetically epilepsy-prone rats. Neuroscience 35: 519-524. 39. MNOSHI, R., S. KITO, N. DOUDOU, AND T. NOMOTO. 1990. Agerelated changes of strychnine-insensitive glycine receptors in rat brain as studied by in vitro autoradiography. Synapse 6: 338343. 40. MONAGHAN, D. T., AND C. W. COTMAN. 1982. The distribution of [3H]kainic acid binding sites in rat CNS as determined by autoradiography. Bruin Res. 262: 91-100. 41. MONAGHAN, D. T., AND C. W. COTMAN. 1985. Distribution of N-methyl-D-aspartate-sensitive L-[3H]glutamate-binding sites in rat brain. J. Neurosci. 5: 2909-2919. 42. MONAGHAN, D. T., H. J. OLVERMAN, L. NGUYEN, J. C. WATKINS, AND C. W. COTMAN. 1988. Two classes of N-methyl-D-aspartate

[3H]KA recognition sites: lation by glycine. 43.

44.

45.

46.

47.

48.

Differential Proc. Natl.

AND

NMDA-SENSITIVE

L-[3H]GLUTAMATE tron microscopic synaptogenesis.

distribution and differential reguAcad. Sci. USA 85: 9836-9840.

NAIK, N. T. 1963. Technical variations on Koelle’s histochemical method for demonstrating choline&erase activity. Q. J. Microse. sci. 104: 89-100. PETIT, T. L., J. D. LE BOIJTILLIER, E. J. MARKUS, AND N. W. MILGRAM. 1989. Synaptic structural plasticity following repetitive activation in the rat hippocampus. Exp. Neural. 105: 72-79. REPRF,SA, A., E. TRJZMBLAY, AND Y. BEN-ARI. 1987. Keinate binding sites in the hippoc~pal mossy fibers: Localization and plasticity. Neuroscience 20: 739-748. ROMANELLI, M. F., J. C. MORRIS, K. ASHKIN, AND L. A. COBEN. 1990. Advanced Alzheimer’s disease is a risk factor for late-onset seizures. Arch. Neurol. 47: 847-850. SAVAGE, D. D., J. V. NADLER, AND J. 0. MCNAMARA. 1984. Reduced kainic acid binding in rat hippocampal formation after limbic kindling. Brain Res. 323: 128-131. STEWARD, O., AND S. V. VINSANT. 1978. The process of reinnervation of the dentate gyms of the adult rat: A quantitative elec-

49.

50.

281

BINDING analysis J. Comp.

of terminal proliferation Neural. 214: 370-386.

and reactive

SUTUI,A, T., X. X. HE, J. CAVAZOS, AND G. SCOTT. 1988. Synaptic reorganization in the hippocampus induced by abnormal functional activity. Science 239: 1147-1150. ULAS, J., D. T. MONAGHAN, AND C. W. COTMAN. 1990. Kainate receptors in the rat hippocampus: A distribution and time course of changes in response to unilateral lesions of the entorhinal cortex. J. Neurosc~. 10: 2352-2362.

51.

ULAS, J., D. T. MONAGHAN, AND C. W. COTMAN. 1990. Plastic response of hippocampal excitatory amino acid receptors to deafferentation and reinnervation. Neuroscience 34: 9-17.

52.

WYSS, J. M., L. W. SWANSON, AND W. M. COWAN. 1979. A study of subcortical afferents to the hippocampal formation in the rat. Neuroscience 4: 463-476. ZIMMER, J. 1973. Extended commissural and ipsilateral projections in postnatally deentorhinated hippocampus and fascia dentata demonstrated in rats by silver impregnation. Brain Res. 64: 293-311.

53.