AMPA receptor alterations precede mossy fiber sprouting in young children with temporal lobe epilepsy

Neuroscience 126 (2004) 105–114

AMPA RECEPTOR ALTERATIONS PRECEDE MOSSY FIBER SPROUTING IN YOUNG CHILDREN WITH TEMPORAL LOBE EPILEPSY E. LYND-BALTA,a,b W. H. PILCHERa AND S. A. JOSEPHa*

seizure(s) cause cell loss has been the impetus for many studies, but remains unresolved (Fisher et al., 1998). The well-characterized feed-forward circuitry of the hippocampus transmits information via binding of the neurotransmitter glutamate to several types of glutamate receptors including the ionotropic receptors N-methyl-Daspartate (NMDA), ␣-amino-3-hydroxy-5-methyl-isoxazole4-proprionate (AMPA), and kainate, as well as the metabotropic glutamate receptors (Bettler and Mulle, 1995). During seizures, excessive amounts of glutamate are released which creates an excitotoxic milieu for cells bathed in neurotransmitter (Carlson et al., 1992; During and Spencer, 1993). Based on studies of human tissue from temporal lobe epileptics as well as animal models of epilepsy, a link has been made suggesting a causal relationship between permutations in excitatory circuitry and disturbances in brain function. In a seminal report, we showed dramatic morphologic changes in glutamatergic AMPA receptor subunit immunocytochemistry in the sclerotic hippocampus harvested from adult patients with intractable TLE (Lynd-Balta et al., 1996; Joseph and Lynd-Balta, 2001). In particular AMPA receptor profiles, specifically glutamate AMPA receptor subunit 1 (GluR1) and glutamate AMPA receptor subunit 2 and 3 (GluR2/3) were shown to be enhanced in density and extent throughout the inner and outer molecular layers of the dentate gyrus. Others have confirmed these findings and/or extended them by identifying the receptor subunits on dendritic elements with electron microscopy (Babb et al., 1996; de Lanerolle et al., 1998). In sclerotic TLE specimens, there is both an augmentation of glutamate receptor subunit expression on postsynaptic elements in the dentate gyrus inner and outer molecular layers and aberrant presynaptic elements confined to the dentate gyrus inner molecular layer emanating from mossy fiber collateral sprouting (Sutula et al., 1989; Babb et al., 1991). This novel aberrant projection of mossy fiber collaterals to the inner molecular layer has been associated with a loss of hilar mossy cells (Babb et al., 1989). The temporal progression of cell loss, glutamate receptor alterations and mossy fiber sprouting in TLE is not known. The present study was undertaken to identify agerelated patterns of glutamate receptor changes, mossy fiber sprouting, and cell loss in hippocampal specimens of children with TLE.

a Department of Neurosurgery, University of Rochester Medical Center, Rochester, NY 14642, USA b Department of Biology, St. John Fisher College, Rochester, NY 14618, USA

Abstract—Following neurological injury early in life numerous events, including excitotoxicity, neural degeneration, gliosis, neosynaptogenesis, and circuitry reorganization, may alone or in concert contribute to hyperexcitability and recurrent seizures in temporal lobe epilepsy. Our studies provide new evidence regarding the temporal sequence of key elements of hippocampal reorganization, mossy fiber sprouting and glutamate receptor subunit up-regulation, in a subset of young temporal lobe epileptic patients. Without evidence of mossy fiber sprouting, the youngest age group (3–10 years old) of mesial temporal lobe epileptic patients demonstrated enhanced glutamate receptor subunit profiles, suggesting that the dendritic change precedes axonal sprouting. However, sclerotic hippocampal specimens from epileptic patients ages 12–15 years old had the characteristic features of glutamate receptor up-regulation and mossy fiber sprouting first identified in the adult, indicating that reconstructed circuits appear early in the course of the disease. Non-sclerotic hippocampal specimens from lesion associated temporal lobe epileptic patients of all age groups showed minimal cell loss, sparse staining of glutamate receptor subunits in the dentate gyrus, and little or no mossy fiber sprouting. These compelling findings suggest a progressive sequence of events in the reorganization of the dentate gyrus of sclerotic hippocampal specimens. We suggest that cell loss and upregulation of glutamate receptor subunits appear early in temporal lobe epilepsy and contribute to the synaptic plasticity that may facilitate the subsequent sprouting of mossy fiber collaterals which compound an already precipitous state of decline. The combination of pre-synaptic and post-synaptic changes serves as a potential substrate for hyperexcitability. © 2004 IBRO. Published by Elsevier Ltd. All rights reserved. Key words: glutamate receptor, hippocampal sclerosis, synaptic reorganization, plasticity, hippocampus, excitotoxicity.

Medically intractable temporal lobe epilepsy (TLE) is commonly characterized by cell loss and gliosis in hippocampal subfields culminating in hippocampal sclerosis. The prevailing question of whether cell loss causes seizures or *Corresponding author. Tel: ⫹1-585-275-2579; fax: ⫹1-585-273-2960. E-mail address: [email protected] (S. A. Joseph). Abbreviations: AMPA, ␣-amino-3-hydroxy-5-methyl-isoxazole-4proprionate; GluR1, glutamate AMPA receptor subunit 1; GluR2/3, glutamate AMPA receptor subunits 2 and 3; LTLE, lesional temporal lobe epilepsy; MTLE, mesial temporal lobe epilepsy; NMDA, N-methyl-D-aspartate; TLE, temporal lobe epilepsy.

EXPERIMENTAL PROCEDURES Patients Hippocampal specimens were harvested under local or general anesthesia from patients of varying ages during the course of

0306-4522/04$30.00⫹0.00 © 2004 IBRO. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.neuroscience.2004.03.004

105

106

E. Lynd-Balta et al. / Neuroscience 126 (2004) 105–114

Table 1. Clinical histories of young TLE patients used in the etiological classification of MTLE and LTLE Age

Gender

Side

Early history/precipitating injury

MTLE 3.5

F

L

Encephalitis, status

2

1.5

7 10 10 10

F M F M

L R R R

Febrile seizure (1.5 yr), status (2 yr) Febrile seizure (1 yr) Multiple closed head injuries Head trauma (8 months) febrile seizure (1 yr) with status

2 6 8 3

5 4 1 7

F M

R R

Astrocytoma Oligodendroglioma

2 1.5

2 4.5

F

L

10

13 14

F F

R R

Prenatal injury (encephalomalacia) febrile seizure (15 mo) Closed head injury (1.5 yr)

14 14 15 LTLE 14 14 15 15

M F M

L L R

Early afebrile seizure Closed head injury (5 yr) Febrile seizure (1 yr)

5 10 2

F F F M

L L L R

Astrocytoma Cavernous hemangioma Astrocytoma Ganglioglioma

11.5 5.5 7 13

LTLE 4 6 MTLE 12

temporal lobe resections for intractable seizure disorders. Patients were selected for surgery following a comprehensive evaluation in our epilepsy center using a standard protocol (Pilcher et al., 1992). Informed consent was obtained from each patient or their legal guardian prior to conducting this analysis. Tissue used for these studies was collected into buffer in the operating room, followed by immersion in a 0.4% sodium sulfide 0.1 M phosphate buffer solution for 20 min and then fixed overnight at 4 °C in a solution of 0.1 M phosphate buffer containing 1% paraformaldehyde, 1.25% glutaraldehyde and 20% sucrose. Serial sections of 35 ␮m were cut on a freezing microtome and stored in cryoprotectant solution till used.

Timm histochemical stain Free floating sections were stained for the presence of heavy metal cations according to the Timm sulfide silver method (Ribak and Peterson, 1991). In a darkroom, free floating sections were incubated in a physical developer (60 ml of a 50% gum Arabic solution, 10 ml of an aqueous solution of 2.35 g of sodium citrate and 2.6 g of citric acid, 30 ml of an aqueous solution of 1.7 g hydroquinone, and 500 ␮l of a 17% silver nitrate solution added just prior to use) for 60 min. To terminate development, sections were rinsed in distilled water for 15 min, then placed in a 5% sodium thiosulfate solution for 12 min, followed by an additional distilled water rinse. Sections were then mounted on slides.

Immunocytochemistry and histochemistry Cryocut sections were immunostained for AMPA receptor subunits GluR1 and GluR2/3 (rabbit polyclonal; Chemicon International, Temecula, CA, USA) and dynorphin A (Peninsula Laboratories Inc., San Carlos, CA, USA) according to our previously published protocol (Lynd-Balta et al., 1996). Sections were placed directly in primary antisera at a dilution of 1:4000 with PBS with

Year of 1st afebrile

2.5 8

Years intractable

MRI hippocampal atrophy

⫹ Hemisphere atrophy ⫹ ⫹ ⫹ ⫹

2

⫹

5 6

⫹ ⫹ Hemisphere atrophy ⫹ ⫹ ⫹

8 4 13 3 3 8 2

0.4% Triton X-100, 1% bovine serum albumin, 4% normal goat serum for five nights. The avidin– biotin reaction (rabbit Vectastain ABC elite kit; Vector Laboratories, Burlingame, CA, USA) was used to process the tissue. Specificity controls consisted of deletion of primary antisera. Specimens from each group were run simultaneously so that the histochemical and immunocytochemical results could be interpreted for their endogenous differences and not due to a technical limitation. We stained every eighth section with Cresyl Violet to define nuclear boundaries and subdivisions and to examine patterns of cell loss.

Classification We have adopted an etiological classification of TLE patients (Engel, 1996). Table 1 summarizes the clinical histories of the young TLE patients. The first and most common category is mesial TLE (MTLE), which is characterized by hippocampal sclerosis identified with pre-operative qualitative MRI and postoperative microscopic analysis. The sclerotic hippocampus is a shrunken, hardened, cell poor, gliosed specimen. The histories of MTLE patients almost invariably reveal a neurological insult at an early age (i.e. febrile seizure, CNS infection, closed head injury) with the subsequent development of recurrent seizures. The second category used in this study is lesional associated TLE (LTLE). Lesions include tumors, vascular malformations, cicatrices, cysts, cortical dysgenesis and cortical heterotopia. LTLE patients show minimal cell loss and their hippocampal specimens do not appear sclerotic. Epileptic patients that did not fit into the above categories were eliminated from these studies. In this study, MTLE and LTLE patients were subdivided into one of three groups based on age: group 1: children 3–10 years old (seven specimens: five MTLE, two LTLE); group 2: children 12–15 years old (10 specimens: six MTLE, four LTLE); and group 3: adults 25 years and older.

E. Lynd-Balta et al. / Neuroscience 126 (2004) 105–114

107

In all LTLE hippocampal specimens, the cells in the CA fields and dentate gyrus appear preserved with minimal cell loss detected. The granule cell layer of the dentate gyrus appears as a dark band consisting of densely packed, darkly stained neurons. Neurons in the CA subfields, hilus, and subiculum appear cytoarchitectonically normal. No focal damage is evident. In contrast, MTLE hippocampal specimens from even the youngest patients appear sclerotic, having varying degrees of cell loss, with CA1 most compromised, followed by CA3 and hilus (Fig. 1A). All MTLE specimens from the youngest age group have clear signs of cell loss in CA1, with large portions of CA1 completely devoid of neurons (Fig. 1A). Degenerative changes are also apparent in CA2, CA3 and hilus and upon examination remaining cells appear pale with Nissl stain and edematous with some degree of vacuolization. The hippocampal specimens from MTLE patients ages 12 to15 years old appear shrunken with extensive neuron loss evident in all CA fields and the hilus. Typically CA1 is almost completely devoid of cells, with significant reductions of cells in CA3 and the hilus. A similar pattern of cell loss is found in MTLE specimens of adult patients. In addition, there is loss to the granule cell layer with dispersion evident in a subset of MTLE specimens from all age groups.

terneurons and pyramidal cells in the CA fields (Figs. 1B, 2B, D, F, 4A). In the hilus, there is a distinct profile of GluR1 immunoreactivity evident in sclerotic specimens regardless of age. Immunoreactive tufts of GluR1 often surround GluR1 immunoreactive somata and proximal processes (Fig. 4A). GluR2/3 immunostaining is also remarkably intense in surviving neurons of the hippocampal formation with cell bodies typically darkly stained along with the proximal processes (Figs. 1C, 3B, D, F, 4B). Like the pattern evident for GluR1 immunostaining, clusters of GluR2/3 immunoreactive neuropil surround GluR2/3 immunostained hilar neurons (Fig. 4B). The most striking difference in AMPA receptor subunit immunostaining when comparing MTLE specimens to LTLE specimens is the dense immunostaining of GluR1 and GluR2/3 detected throughout the dentate gyrus of MTLE specimens (Figs. 2, 3). Intense immunostaining of GluR1 and GluR2/3 in the dentate gyrus of MTLE specimens is observed, regardless of the age of the patient or the extent of cell loss (Figs. 2B, D, F, 3B, D, F). The granule cells are darkly stained and there is an extensive labeling of processes (presumably dendrites) throughout the inner and outer molecular layers giving an extended appearance to that dendritic zone (Figs. 2B, D, F, 3B, D, F). Staining for GluR1 is more intense than GluR2/3 in the molecular layers. In some cases, GluR1 immunoreactivity is denser in the outer portion of the molecular layer, with more moderate staining of the inner molecular layer.

GluR1and GluR2/3 immunoreactivity

Mossy fiber sprouting

In LTLE specimens of all age groups, GluR1 and GluR2/3 immunoreactivity is quite prevalent throughout the hippocampal formation with staining evident in all CA fields, hilus and subiculum. GluR1 immunoreactivity is confined to processes in the CA subfields with a notable lack of somal staining whereas GluR2/3 immunoreactivity is distributed in both the pyramidal cell bodies as well as their dendritic processes. Unlike the dense immunostaining of the CA fields, only sparse GluR1 and GluR2/3 immunostaining is found in the dentate gyrus (Figs. 2A, C, E and 3A, C, E). GluR1 immunostaining is faint in the molecular layers with few labeled somata in the granule cell layer (Fig. 2A, C, E). GluR2/3-positive granule cell somata show moderate staining with minimal process staining in the dentate gyrus molecular layers of LTLE specimens (Fig. 3A, C, E). In MTLE specimens of all age groups, GluR1 and GluR2/3 immunostaining is diminished or absent in areas of cell loss, and this is clearly evident in CA1 (Fig. 1B, C). Because the extent of cell loss is greater in the older children, there is a corresponding decrease in GluR1 and GluR2/3 immunoreactivity in the CA fields. However in sclerotic specimens from all age groups, surviving neurons have dense immunoreactivity for both GluR1 and GluR2/3 (Fig. 1B, C). Unlike LTLE, in MTLE specimens the subcellular location of GluR1 immunoreactivity is not limited to processes, rather many of the somata of surviving neurons have robust immunoreactivity for GluR1, including the granule cells in the dentate gyrus, hilar neurons, and in-

Using dynorphin immunocytochemistry and/or Timm histochemistry, the normal distribution of mossy fibers in the hilus and CA3 is clearly evident in LTLE specimens from all age groups. Intense staining fills the hilar region and the stratum lucidum of CA3. A dense band of Timm stain or dynorphin immunoreactivity is found subjacent to the granule cell layer of the dentate gyrus, with the typical heavy labeling of fibers within and traversing the hilus. There is little to no staining in the molecular layer of the dentate gyrus (Fig. 5A, C, E). In sclerotic MTLE specimens from all age groups, Timm staining continues to be evident in the hilus and CA3, although the intensity of staining varies. It is common to find tufts of Timm positive neuropil surrounding surviving neurons in the hilus. This patchy distribution of Timm stain is not seen in LTLE specimens. Silver impregnated mossy fibers continue on to terminate in the stratum lucidum of CA3 with some staining evident in the stratum pyramidale of CA3. Clearly evident in adult MTLE specimens is the novel staining of mossy fiber collaterals in the dentate gyrus inner molecular layer, representing the mossy fiber sprouting that is considered a hallmark feature of MTLE (Fig. 5F). Interestingly, there is no mossy fiber sprouting into the molecular layer of the dentate gyrus in four of the five specimens procured from MTLE patients under the age of 10 (Fig. 5B). Occasionally a mossy fiber can be seen traversing the granule cell layer in these four cases. Only one of the brains, that of a 10-year-old female, shows significant mossy fiber sprouting into the dentate gyrus

RESULTS Patterns of cell loss

108

E. Lynd-Balta et al. / Neuroscience 126 (2004) 105–114

Fig. 1. A composite of low power photomicrographs of a hippocampal specimen from a 10-year-old MTLE patient. (A) A photomontage depicting a Nissl stained section. Note the remarkable cell loss in CA1. (B) A photomontage of an alternate section stained for GluR1 immunoreactivity demonstrating an absence of GluR1 immunoreactivity in CA1 consistent with cell loss and dense immunostaining of GluR1 in surviving neurons of hippocampal formation. Note the dense immunoreactivity in the dentate gyrus (DG) of MTLE specimen. (C) A photomontage of an alternate section stained for GluR2/3 immunostaining. A lack of GluR2/3 immunostaining is evident in CA1 corresponding to cell loss. As shown with GluR1 immunoreactivity, dense GluR2/3 immunoreactivity is also found throughout the hippocampal formation including the dentate gyrus in MTLE specimens. Scale bar⫽1000 ␮m.

Fig. 2. Photomicrographs comparing GluR1 immunoreactivity in the dentate gyrus of LTLE and MTLE specimens. (A, C, E) GluR1 immunoreactivity in LTLE specimens demonstrating sparse immunostaining in putative dendritic processes in the dentate inner and outer molecular layers (IML, OML) in a 4-year-old patient (A), 15-year-old patient (C), and 30-year-old patient (E). Few if any granule cells are labeled in LTLE specimens. (B, D, F) Comparative photomicrographs depicting GluR1 immunoreactivity in MTLE specimens from patients ages 10years-old (B), 15-years-old (D), and 52-years-old (F). Note the robust GluR1 immunoreactivity in the soma and processes of granule cells (GCL). Staining spans the entire zone of the molecular layers. Scale bar⫽100 ␮m.

E. Lynd-Balta et al. / Neuroscience 126 (2004) 105–114

109

Fig. 4. High power photomicrographs of GluR1 and GluR2/3 immunoreactivity in young MTLE specimens. (A) GluR1 immunostaining is evident in cell bodies and processes. Note the dense GluR1 immunostaining localized to cell bodies which is not seen in LTLE specimens. In the hilus, GluR1 immunostaining is found in many neurons and in associated clusters of neuropil surrounding the proximal processes of the neurons (arrows). (B) GluR2/3 positive somal staining is accompanied by GluR2/3 immunoreactive clusters of neuropil that surround proximal processes (arrows) in the hilus of a young MTLE specimen. Scale bar⫽100 ␮m.

inner molecular layer, as evidenced by dynorphin immunocytochemistry. All six specimens from MTLE patients ages 12 to15 years old show evidence of mossy fiber sprouting into the dentate gyrus molecular layer, although to varying amounts (Fig. 5D).

DISCUSSION Fig. 3. Photomicrographs comparing GluR2/3 immunoreactivity in the dentate gyrus of LTLE and MTLE specimens. (A, C, E) LTLE specimens show moderate GluR2/3 immunostaining of dentate granule cell bodies and sparse immunostaining on dendritic processes in the inner and outer molecular layers of the dentate gyrus (IML, OML) in a 4-year-old LTLE patient (A), 15-year-old LTLE patient (C), and 30year-old LTLE patient (E). Few if any granule cells are labeled in LTLE specimens. (B, D, F) MTLE specimens from patients ages 10-yearsold (D), 12-years-old (E), and 52-years-old (F) have robust GluR2/3 immunostaining in the perikarya (GCL) and dendritic processes (IML, OML) of granule cells. Scale bar⫽100 ␮m.

Controversies exist surrounding the sequence of events responsible for the genesis and progression of recurrent seizures. Analysis of surgical specimens excised from patients with intractable TLE has revealed several hallmark features including hippocampal sclerosis, mossy fiber sprouting and glutamate receptor upregulation (Margerison and Corsellis, 1966; Sutula et al., 1989; Babb et al., 1991; Lynd-Balta et al., 1996). A history of neurological injury early in life is an etiological factor associated with TLE (Falconer, 1971; Mathern et al., 1995). We hypothesize that following injury subsequent neural degenerative

110

E. Lynd-Balta et al. / Neuroscience 126 (2004) 105–114

Fig. 5. Photomicrographs of mossy fiber staining in LTLE and MTLE hippocampal specimens. (A, C, E) Little or no mossy fiber sprouting is evident in the dentate gyrus of LTLE specimens from a 4-year-old patient (A), 15-year-old patient (C), and 30-year-old patient (E). (B) Notably little or no mossy fiber sprouting is evident in the dentate gyrus of an MTLE specimen from a 10-year-old patient. This impressive finding depicts the temporal progression of glutamate receptor upregulation and mossy fiber sprouting as an age-related phenomenon (compare with Figs. 2B and 3B). (D, F) Note that the aberrant mossy fiber sprouting is clearly evident in the dentate gyrus inner molecular layer of MTLE specimens from a 12-year-old patient (E) and a 52year-old patient (F). Scale bar⫽100 ␮m.

changes eventually cascade into a reverberating cycle of excitotoxicity and progressive neuronal loss with circuitry changes and synaptic reorganization that facilitate hyperexcitability and seizure propagation. Taking these factors into consideration, we have analyzed a subset of specimens from TLE patients with a history of early injury, recurrent epileptic seizures, and hippocampal surgery at a young age and compared them to adult TLE specimens. The study was done to determine if we could identify a temporal sequence of reorganization in the pre- and postsynaptic elements of the dentate gyrus molecular layers. Several important findings emerge from this study: (i) AMPA receptor subunit changes (GluR1 and GluR2/3) appear early, often before any evidence of mossy fiber sprouting in children ages 3–10 years old; (ii) MTLE specimens from children ages 12–15 years old have the characteristic features of hippocampal sclerosis, increased AMPA receptor subunit immunoreactivity, and mossy fiber sprouting previously documented in adult MTLE specimens; (iii) significant cell loss is apparent in CA1 in MTLE specimens, starting as early as age three; and (iv) loss to CA3 and hilus is more variable in children with MTLE, but is clearly evident in MTLE children ages 12–15 years old. Neuronal plasticity is a phenomenon associated with diverse brain functions including development, learning and memory, and response to injury and degeneration. The dentate gyrus plays a pivotal role in hippocampal connectivity and undergoes morphological, chemical, and physiological changes in MTLE. In normal hippocampal circuitry, layer II of the entorhinal cortex provides extrinsic glutamatergic input to the outer molecular layer of the dentate gyrus (Andersen et al., 1966; Hjorth-Simonsen, 1972; Hjorth-Simonsen and Jeune, 1972; Steward, 1976; Steward and Scoville, 1976), and hilar mossy cells provide intrinsic glutamatergic input to the inner molecular layer of the dentate gyrus (Swanson et al., 1978; Laurberg and Sorensen, 1981). In turn, granule cell axons (mossy fibers) send excitatory projections to mossy cells, other hilar neurons, and to CA3 neurons (Blackstad et al., 1970; Gaarskjaer, 1978, 1981; Claiborne et al., 1986). Early studies demonstrating pre-synaptic and post-synaptic alterations of granule cells in response to entorhinal lesions have provided a template for predicting the dynamics of circuit reorganization associated with TLE. Specifically in the entorhinal lesion model, there is an initial period of dendritic atrophy, followed by dramatic and robust changes in the post-synaptic element (Parnavelas et al., 1974; Caceres and Steward, 1983). Steward and colleagues identified a redistribution of polyribosomes around the base of dendritic spines and increased incorporation of protein precursors, which precedes axonal regeneration (Fass and Steward, 1983; Steward, 1983; Steward and Fass, 1983). These lesion studies suggest that post-synaptic changes may induce or permit new innervations by pre-synaptic elements (sprouting axons). We hypothesize a similar progressive scenario in MTLE (Fig. 6). MTLE patients typically have a history of an early injury (i.e. febrile seizure, status epilepticus, closed head injury, or CNS infection) and with variable latencies developed

E. Lynd-Balta et al. / Neuroscience 126 (2004) 105–114

111

Early Injury Seizure Activity Normal Dentate Gyrus A

Progressive Reorganization C

B

Aberrant Mossy Fiber Sprouting

Dendritic Response: atrophy then reconstruction OML IML

R

R

ERC

R

R

R

R

R R R

GC

hilar mossy cells

R R

ERC

Deafferentation

R

R

R

R

R

R

R

R

R

R

ERC

Deafferentation R

mossy fiber sprouting

hilar mossy cells

hilar mossy cells

Fig. 6. A schematic diagram depicting a possible scenario of reorganizational events that occur in the dentate gyrus of sclerotic, epileptic hippocampus. (A) Granule cells receive inputs from the entorhinal cortex and hilar mossy cells and express low levels of the AMPA receptor subunits GluR1 and GluR2/3 in the normal hippocampus. (B) Early-age injury produces a partial deafferentation of granule cells, possibly leading to a period of dendritic atrophy. Subsequent reorganizational events occur in a temporal progression. The dendrites express high levels of the AMPA receptor subunits GluR1 and GluR2/3 in the inner and outer molecular layers. (C) Subsequent to receptor changes, continued neural degeneration causes further deafferentation and mossy fiber collaterals sprout and terminate aberrantly in the dentate gyrus inner molecular layer.

spontaneous recurrent seizures. Cell loss was observed in the young patients of this study, indicating that significant damage may occur at the time of the early precipitating injury. Furthermore, the distinct profiles of cell death evident in our experimental groups suggest that recurrent seizure activity and synaptic reorganization may precipitate further neuronal damage (Mathern et al., 1995). The extent of hilar cell loss in TLE is variable and may depend on the initial insult; however, any loss of mossy cell input to the inner molecular layer of the dentate gyrus causes a partial deafferentation of the granule cells (Laurberg and Sorensen, 1981). Previous studies have demonstrated that anatomical remodeling of hippocampal cell dendrites can occur in response to deafferentation, changes in circulating hormone levels, seizure activity and other alterations in synaptic activity (Scheibel, 1974; Caceres and Steward, 1983; Woolley and McEwen, 1994; von Campe et al., 1997). A loss of dendritic spines on granule cell dendrites has been documented in the epileptic brain (Swann et al., 2000). In our previous studies, we showed a dramatic upregulation of AMPA receptor subunits in the inner and outer molecular layers of the dentate gyrus of adult MTLE specimens, suggesting a synaptically reconstructed dendritic profile (Lynd-Balta et al., 1996). Subsequent human studies using an intracellular labeling technique showed more branching and extensions of dendrites in the inner molecular layer of MTLE specimens compared with LTLE specimens (von Campe et al., 1997). Furthermore, Isokawa (1997, 2000) showed a proliferation of dendritic

spines on the same granule cells that exuded an aberrant axon collateral. Our study suggests a temporal progression in AMPA receptor subunit changes and mossy fiber sprouting. One could hypothesize that this putative chronology of events indicates a causal relationship with dendritic changes promoting axonal plasticity. However, there is a disparity to this concept since the up-regulation of AMPA receptor subunits is not confined to the zone of collateral sprouting in the inner molecular layer; rather AMPA receptor subunit staining is evident throughout the inner and outer molecular layers of the dentate gyrus. In MTLE, not only are the granule cells partially deafferented, but they also lose some portion of their target cells with the death of hilar and CA3 neurons. The combination of glutamate receptor changes, deafferentation, and target loss may be the impetus for collateral sprouting of mossy fibers and their aberrant route of termination into the inner molecular layer. To date, a focal point of studies on epileptic reorganization is the easily identifiable mossy fiber and its collateral sprouting (Timm stain) into the inner molecular layer. However, one cannot disregard the possibility that connectivity changes and synaptic reorganization occur in the outer molecular layer as well. Thus one challenge would be to identify a sprouting presynaptic element projecting to the outer molecular layer. Elucidating the timing of these phenomena contributes new knowledge to the pathogenesis associated with MTLE. In specimens from the youngest MTLE patients, glutamate receptor alterations are consistently identified in the absence of mossy fiber sprouting. However by age 12,

112

E. Lynd-Balta et al. / Neuroscience 126 (2004) 105–114

all specimens show concomitant glutamate receptor enhancement and mossy fiber sprouting which allows us to conclude a putative temporal sequence. Similar progressive changes have been delineated in animal models of TLE. An up-regulation of AMPA receptors and NMDA receptors has been shown to precede mossy fiber sprouting following intra-hippocampal injection of kainate (Babb et al., 1996; Mikuni et al., 1998). In a kindling model, Sutula and colleagues (1996) used an NMDA antagonist to demonstrate that glutamate receptors regulate the development of mossy fiber sprouting. An examination of hippocampal specimens from children ages 1–18 years old found different patterns of cell loss, mossy fiber sprouting, and neurochemical changes consistent with distinct and progressive synaptic reorganization in sclerotic and nonsclerotic specimens (Ying et al., 1999). A study of children with TLE, with etiological factors (cortical dysgenic lesions, gliomas) that did not meet the criteria of MTLE, showed variable amounts of mossy fiber sprouting that increased with age (Mathern et al., 1996). It is interesting to speculate that the heightened plasticity associated with the dentate gyrus may underlie some of the phenomena associated with injury early in life and the subsequent development of MTLE. For example, evidence indicates that granule cell neurogenesis can be altered by a variety of factors (Kempermann et al., 1997a,b; Gould et al., 1999; Gould and Tanapat, 1999; van Praag et al., 1999). Further, it has been shown in a rodent model of TLE that seizures promote neurogenesis and that some of the newly born cells form axon collaterals that terminate in the dentate gyrus molecular layer (Parent et al., 1997, 1998, 1999). Dendritic and axonal reorganization in MTLE may occur in response to neurotrophic factors which are expressed at high levels in the developing and immature brain. Granule cells are known to express several neurotrophic factors. Gall and her colleagues (1997) have identified increases in glial-derived or neural-derived neurotrophic factor expression following deafferentation, seizure activity, and other changes in neural activity. Changes in pre-synaptic and post-synaptic elements have the potential to generate a ‘new’ reverberating excitatory circuit. There is evidence that mossy fiber sprouting establishes both excitatory connections and inhibitory connections in the molecular layer creating a complex interplay between neighboring granule cells and between granule cells and interneurons (Frotscher and Zimmer, 1983; Dudek et al., 1994; Kotti et al., 1997). Electron microscopic studies have identified synaptic contacts between mossy fibers and granule cell dendrites and electrophysiological recordings from animal models of TLE have extended the evidence for excitatory circuits being formed via mossy fiber sprouting (Cronin et al., 1992; Dudek et al., 1994; Represa et al., 1995; Wuarin and Dudek, 1996; Okazaki et al., 1999). This novel excitatory monosynaptic feedback circuit could be a substrate for hippocampal hyperexcitability. Recurrent mossy fiber collaterals, using glutamate, reduce the threshold for synchronization of granule cells and

thereby make them more susceptible to seizure activity. However since seizure activity occurs without mossy fiber sprouting, as evidenced by LTLE patients and the youngest MTLE patients studied here, the functional significance of mossy fiber sprouting and its relationship to seizure activity needs to be explored further. To understand the complex mechanisms that initiate and sustain epileptic activity in the hippocampus it is critical to elucidate the chronology of cell loss, synaptic reorganization, and neural circuitry changes. This qualitative study of cell loss and synaptic reorganization is a first step in this direction. Acknowledgements—Excellent technical assistance was provided by Jeannie Padowski.

REFERENCES Andersen P, Holmqvist B, Voorhoeve PE (1966) Entorhinal activation of dentate granule cells. Acta Physiol Scand 66:448 –460. Babb TL, Kupfer WR, Pretorius JK, Crandall PH, Levesque MF (1991) Synaptic reorganization by mossy fibers in human epileptic fascia dentata. Neuroscience 42:351–363. Babb TL, Mathern GW, Leite JP, Pretorius JK, Yeoman KM, Kuhlman PA (1996) Glutamate AMPA receptors in the fascia dentata of human and kainate rat hippocampal epilepsy. Epilepsy Res 26: 193–205. Babb TL, Pretorius JK, Kupfer WR, Crandall PH (1989) Glutamate decarboxylase-immunoreactive neurons are preserved in human epileptic hippocampus. J Neurosci 9:2562–2574. Bettler B, Mulle C (1995) Review: neurotransmitter receptors: II. AMPA and kainate receptors. Neuropharmacology 34:123–139. Blackstad TW, Brink K, Hem J, Jeune B (1970) Distribution of hippocampal mossy fibers in the rat: an experimental study with silver impregnation methods. J Comp Neurol 138:433–449. Caceres A, Steward O (1983) Dendritic reorganization in the denervated dentate gyrus of the rat following entorhinal cortical lesions: a Golgi and electron microscopic analysis. J Comp Neurol 214:387– 403. Carlson H, Ronne-Engstrom E, Ungerstedt U, Hillered L (1992) Seizure related elevations of extracellular amino acids in human focal epilepsy. Neurosci Lett 140:30 –32. Claiborne BJ, Amaral DG, Cowan WM (1986) A light and electron microscopic analysis of the mossy fibers of the rat dentate gyrus. J Comp Neurol 246:435–458. Cronin J, Obenaus A, Houser CR, Dudek FE (1992) Electrophysiology of dentate granule cells after kainate-induced synaptic reorganization of the mossy fibers. Brain Res 573:305–310. de Lanerolle NC, Eid T, von Campe G, Kovacs I, Spencer DD, Brines M (1998) Glutamate receptor subunits GluR1 and GluR2/3 distribution shows reorganization in the human epileptogenic hippocampus. Eur J Neurosci 10:1687–1703. Dudek FE, Obenaus A, Schweitzer JS, Wuarin JP (1994) Functional significance of hippocampal plasticity in epileptic brain: electrophysiological changes of the dentate granule cells associated with mossy fiber sprouting. Hippocampus 4:259 –265. During MJ, Spencer DD (1993) Extracellular hippocampal glutamate and spontaneous seizure in the conscious human brain. Lancet 341:1607–1610. Engel J Jr (1996) Introduction to temporal lobe epilepsy. Epilepsy Res 26:141–150. Falconer MA (1971) Genetic and related aetiological factors in temporal lobe epilepsy: a review. Epilepsia 12:13–31. Fass B, Steward O (1983) Increases in protein-precursor incorporation in the denervated neuropil of the dentate gyrus during reinnervation. Neuroscience 9:653–664.

E. Lynd-Balta et al. / Neuroscience 126 (2004) 105–114 Fisher PD, Sperber EF, Moshe SL (1998) Hippocampal sclerosis revisited. Brain Dev 20:563–573. Frotscher M, Zimmer J (1983) Lesion-induced mossy fibers to the molecular layer of the rat fascia dentata: identification of postsynaptic granule cells by the Golgi-EM technique. J Comp Neurol 215:299 –311. Gaarskjaer FB (1978) Organization of the mossy fiber system of the rat studied in extended hippocampi: I. Terminal area related to number of granule and pyramidal cells. J Comp Neurol 178:49 –72. Gaarskjaer FB (1981) The hippocampal mossy fiber system of the rat studied with retrograde tracing techniques: correlation between topographic organization and neurogenetic gradients. J Comp Neurol 203:717–735. Gall CM, Lauterborn JC, Guthrie KM, Stinis CT (1997) Seizures and the regulation of neurotrophic factor expression: associations with structural plasticity in epilepsy. Adv Neurol 72:9 –24. Gould E, Beylin A, Tanapat P, Reeves A, Shors TJ (1999) Learning enhances adult neurogenesis in the hippocampal formation. Nat Neurosci 2:260 –265. Gould E, Tanapat P (1999) Stress and hippocampal neurogenesis. Biol Psychiatry 46:1472–1479. Hjorth-Simonsen A (1972) Projection of the lateral part of the entorhinal area to the hippocampus and fascia dentata. J Comp Neurol 146:219 –232. Hjorth-Simonsen A, Jeune B (1972) Origin and termination of the hippocampal perforant path in the rat studied by silver impregnation. J Comp Neurol 144:215–232. Isokawa M (1997) Preservation of dendrites with the presence of reorganized mossy fiber collaterals in hippocampal dentate granule cells in patients with temporal lobe epilepsy. Brain Res 744:339 – 343. Isokawa M (2000) Remodeling dendritic spines of dentate granule cells in temporal lobe epilepsy patients and the rat pilocarpine model. Epilepsia 41 (Suppl 6):S14 –17. Joseph SA, Lynd-Balta E (2001) Evidence for an anatomical substrate of hyperexcitability in human temporal lobe epilepsy: glutamate receptor alterations and reorganized circuitry. Funct Neurol 16: 347–365. Kempermann G, Kuhn HG, Gage FH (1997a) Genetic influence on neurogenesis in the dentate gyrus of adult mice. Proc Natl Acad Sci USA 94:10409 –10414. Kempermann G, Kuhn HG, Gage FH (1997b) More hippocampal neurons in adult mice living in an enriched environment. Nature 386:493–495. Kotti T, Riekkinen PJ Sr., Miettinen R (1997) Characterization of target cells for aberrant mossy fiber collaterals in the dentate gyrus of epileptic rat. Exp Neurol 146:323–330. Laurberg S, Sorensen KE (1981) Associational and commissural collaterals of neurons in the hippocampal formation (hilus fasciae dentatae and subfield CA3). Brain Res 212:287–300. Lynd-Balta E, Pilcher WH, Joseph SA (1996) Distribution of AMPA receptor subunits in the hippocampal formation of temporal lobe epilepsy patients. Neuroscience 72:15–29. Margerison JH, Corsellis JA (1966) Epilepsy and the temporal lobes: a clinical, electroencephalographic and neuropathological study of the brain in epilepsy, with particular reference to the temporal lobes. Brain 89:499 –530. Mathern GW, Leite JP, Pretorius JK, Quinn B, Peacock WJ, Babb TL (1996) Severe seizures in young children are associated with hippocampal neuron losses and aberrant mossy fiber sprouting during fascia dentata postnatal development. Epilepsy Res Suppl 12:33– 43. Mathern GW, Pretorius JK, Babb TL (1995) Influence of the type of initial precipitating injury and at what age it occurs on course and outcome in patients with temporal lobe seizures. J Neurosurg 82: 220 –227. Mikuni N, Babb TL, Chakravarty DN, Hadam JL, Penrod CE (1998)

113

NMDAR2 upregulation precedes mossy fiber sprouting in kainate rat hippocampal epilepsy. Neurosci Lett 255:25–28. Okazaki MM, Molnar P, Nadler JV (1999) Recurrent mossy fiber pathway in rat dentate gyrus: synaptic currents evoked in presence and absence of seizure-induced growth. J Neurophysiol 81:1645– 1660. Parent JM, Janumpalli S, McNamara JO, Lowenstein DH (1998) Increased dentate granule cell neurogenesis following amygdala kindling in the adult rat. Neurosci Lett 247:9 –12. Parent JM, Tada E, Fike JR, Lowenstein DH (1999) Inhibition of dentate granule cell neurogenesis with brain irradiation does not prevent seizure-induced mossy fiber synaptic reorganization in the rat. J Neurosci 19:4508 –4519. Parent JM, Yu TW, Leibowitz RT, Geschwind DH, Sloviter RS, Lowenstein DH (1997) Dentate granule cell neurogenesis is increased by seizures and contributes to aberrant network reorganization in the adult rat hippocampus. J Neurosci 17:3727–3738. Parnavelas JG, Lynch G, Brecha N, Cotman CW, Globus A (1974) Spine loss and regrowth in hippocampus following deafferentation. Nature 248:71–73. Pilcher WH, Locharernkul C, Primrose D, Ojemann LM, Ojemann GA (1992) Update in epilepsy: Part III: surgical therapy of intractable epilepsy. NY State J Med 92:92–96. Represa A, Niquet J, Pollard H, Ben-Ari Y (1995) Cell death, gliosis, synaptic remodeling in the hippocampus of epileptic rats. J Neurobiol 26:413–425. Ribak CE, Peterson GM (1991) Intragranular mossy fibers in rats and gerbils form synapses with the somata and proximal dendrites of basket cells in the dentate gyrus. Hippocampus 1:355–364. Scheibel ME, Crandall PH, Scheibel AB (1974) The hippocampaldentate complex in temporal lobe epilepsy: a Golgi study. Epilepsia 15:55–80. Steward O (1976) Topographic organization of the projections from the entorhinal area to the hippocampal formation of the rat. J Comp Neurol 167:285–314. Steward O (1983) Alterations in polyribosomes associated with dendritic spines during the reinnervation of the dentate gyrus of the adult rat. J Neurosci 3:177–188. Steward O, Fass B (1983) Polyribosomes associated with dendritic spines in the denervated dentate gyrus: evidence for local regulation of protein synthesis during reinnervation. Prog Brain Res 58: 131–136. Steward O, Scoville SA (1976) Cells of origin of entorhinal cortical afferents to the hippocampus and fascia dentata of the rat. J Comp Neurol 169:347–370. Sutula T, Cascino G, Cavazos J, Parada I, Ramirez L (1989) Mossy fiber synaptic reorganization in the epileptic human temporal lobe. Ann Neurol 26:321–330. Sutula T, Koch J, Golarai G, Watanabe Y, McNamara JO (1996) NMDA receptor dependence of kindling and mossy fiber sprouting: evidence that the NMDA receptor regulates patterning of hippocampal circuits in the adult brain. J Neurosci 16:7398 –7406. Swann JW, Al-Noori S, Jiang M, Lee CL (2000) Spine loss and other dendritic abnormalities in epilepsy. Hippocampus 10:617–625. Swanson LW, Wyss JM, Cowan WM (1978) An autoradiographic study of the organization of intrahippocampal association pathways in the rat. J Comp Neurol 181:681–715. van Praag H, Christie BR, Sejnowski TJ, Gage FH (1999) Running enhances neurogenesis, learning, and long-term potentiation in mice. Proc Natl Acad Sci USA 96:13427–13431. von Campe G, Spencer DD, de Lanerolle NC (1997) Morphology of dentate granule cells in the human epileptogenic hippocampus. Hippocampus 7:472–488. Woolley CS, McEwen BS (1994) Estradiol regulates hippocampal dendritic spine density via an N-methyl-D-aspartate receptordependent mechanism. J Neurosci 14:7680 –7687.

114

E. Lynd-Balta et al. / Neuroscience 126 (2004) 105–114

Wuarin JP, Dudek FE (1996) Electrographic seizures and new recurrent excitatory circuits in the dentate gyrus of hippocampal slices from kainate-treated epileptic rats. J Neurosci 16:4438 –4448. Ying Z, Babb TL, Hilbig A, Wyllie E, Mohamed A, Bingaman W,

Prayson R, Staugaitis S, Najm I, Luders HO (1999) Hippocampal chemical anatomy in pediatric and adolescent patients with hippocampal or extrahippocampal epilepsy. Dev Neurosci 21:236 –247.

(Accepted 7 March 2004)