Electrophysiology of dentate granule cells after kainate-induced synaptic reorganization of the mossy fibers

Brain Research, 573 (1992) 305-310 © 1992 Elsevier Science Publishers B.V. All rights reserved. 0006-8993/92/$05.00

BRES 25036

305

Short Communications

Electrophysiology of dentate granule cells after kainate-induced synaptic reorganization of the mossy fibers James Cronin a, Andre Obenaus b, Carolyn R. Houser c'd and E Edward Dudek b aDepartment of Psychology, Tulane University, New Orleans, LA 70118 (U.S.A.), bMental Retardation Research Center and Brain Research Institute, UCLA School of Medicine, Los Angeles, CA 90024 (U.S.A.), CNeurology and Research Services, Veterans Administration Medical Center, West Los Angeles, CA 90025 (U.S.A.) and aDepartment of Anatomy and Cell Biology and Brain Research Institute, UCLA School of Medicine, Los Angeles, CA 90024 (U.S.A.)

(Accepted 29 October 1991) Key words: Kainic acid; Dentate gyrus; Hippocampus; Mossy fiber; Sprouting; Recurrent excitation; Epilepsy; Seizure

Morphological data from humans with temporal lobe epilepsy and from animal models of epilepsy suggest that seizure-induced damage to dentate hilar neurons causes granule cells to sprout new axon collaterals that innervate other granule cells. This aberrant projection has been suggested to be an anatomical substrate for epileptogenesis. This hypothesis was tested in the present study with intra- and extracellular recordings from granule cells in hippocampal slices removed from rats 1-4 months after kainate treatment. In this animal model, hippocampal cell loss leads to sprouting of mossy fiber axons from the granule cells into the inner molecular layer of the dentate gyms. Unexpectedly, when shces with mossy fiber sprouting were examined in normal medium, extracellular stimulation of the hilus or perforant path evoked relatively normal responses. However, in the presence of the GABAA-receptor antagonist, bicucuUine, low-intensity hilar stimulation evoked delayed bursts of action potentials in about one-quarter of the slices. In one-third of the bicuculline-treated slices with mossy fiber sprouting, spontaneous bursts of synchronous spikes were superimposed on slow negative field potentials. Slices from normal rats or kainate- treated rats without mossy fiber sprouting never showed delayed bursts to weak hilar stimulation or spontaneous bursts in bicuculline. These data suggest that new local excitatory circuits may be suppressed normally, and then emerge functionally when synaptic inhibition is blocked. Therefore, after repeated seizures and excitotoxic damage in the hippocampus, synaptic reorganization of the mossy fibers is consistently associated with normal responses; however, in some preparations, the mossy fibers may form functional recurrent excitatory connections, but synaptic inhibition appears to mask these potentially epileptogenic alterations. Although much has been learned about the electrophysiological characteristics of cortical neurons and the mechanisms of seizure activity, there is little information on the cellular basis of chronic epilepsy. Treatment of rats with kainic acid induces a period of sustained seizure activity and degeneration of some hippocampal neurons, leading to a pattern of cell loss similar to that seen in human temporal lobe epilepsy 3'19. After kainateinduced hippocampal lesions in rats, dentate granule cell axons (i.e. mossy fibers) sprout aberrant collaterals into the inner molecular layer of the dentate gyrus 2°. Similar anatomic changes have been demonstrated in the kindling model of epilepsy 27. Recently, reorganization of granule cell axon collaterals has also been observed in surgically removed specimens from humans with severe temporal lobe epilepsy 2'7'14'26. Thus, determining the functional effects of this reorganization has assumed increasing importance. One hypothesis suggests that the reorganized connections functionally contact granule cell dendrites, forming a local excitatory circuit that leads to hyperexcitability and epileptiform events 3°. We tested

this hypothesis with simultaneous intra- and extracellular recordings followed by histochemical staining of the mossy fibers in hippocampal slices from kainate-treated rats. Male Sprague-Dawley rats (250-300 g) were injected subcutaneously with kainate (18 mgckg) and monitored behaviorally for seizure activity over the next several hours. After 1-4 months, 500-/zm-thick transverse slices were taken from the temporal pole of the hippocampus and maintained with conventional techniques. The slices were continuously perfused (32-34°C) with artificial cerebrospinal fluid (ACSF) containing (in raM): 124 NaCl, 3 KC1, 1.3 MgSO4, 1.4 NaHEPO4, 26 NaHCO3, 2.4 CaCl 2 and 11 glucose. Control preparations included rats that were not injected with kainate and kainate-injected rats without mossy fiber sprouting. All numerical data representing controls are from hippocampal slices of kainate-treated rats without mossy fiber sprouting. Electrophysiological data were included only if population spikes to hilar and perforant path stimulation were > 4 inV. The hilar stimulating electrode was positioned within the hilus

Correspondence: F. Edward Dudek, Mental Retardation Research Center, UCLA School of Medicine, 760 Westwood Plaza (NPI 58-258), Los Angeles, CA 900124, U.S.A. Fax: (1) (213) 206-5060.

306 on an imaginary line connecting the end blades of the dentate gyrus, and stimulus frequency was 0.016 Hz to 0.10 Hz. Intracellular recordings were accepted only if the resting potential was more negative than - 6 0 mV, the input resistance was >20 Mg2, and the action potentials were >70 mV. All traces were filtered at 3 kHz. Following the electrophysiological recordings, the slices (or adjacent slices) were prepared for evaluation of mossy fiber sprouting. The Timm's histochemical method 6, modified for non-perfused hippocampal slices, was used to label zinc-containing mossy fiber boutons. Only slices with consistent dark staining in the inner molecular layer (i.e. scores of 2 or 3 in the rating scale of Tauck and Nadler 3°) were considered to have dense mossy fiber sprouting. Of the 25 rats injected with kainate, 17 animals had hippocampal slices with dense mossy fiber sprouting. Intra- and/or extracellular recordings were obtained from 31 slices; 18 showed sprouting (i.e. 2 or 3 on Tauck and Nadler 3° scale), and 13 had little or no sprouting (0 or 1 rating). Nissl staining of adjacent sections consistently revealed that hilar cell loss was observed in slices with mossy fiber sprouting, but this was not examined quantitatively. We first investigated whether, in the normal ACSF used for electrophysiological studies of hippocampal slices, mossy fiber sprouting was associated with enhanced excitability in the dentate gyrus and with the presence of seizure-like electrical activity. Granule cells in slices with mossy fiber sprouting had resting potentials and input resistances that were similar to those of cells in slices without sprouting (sprouting: -77.9 + 2.4 mV, 61.4 + 9.4 Mr2, n =14 cells versus no sprouting: -76.1 + 3.7 mV, 83.8 + 28.6 Mfl, n =9 cells; mean + S.E.M.). In control slices and slices with sprouting, stimulation of granule cell axons within the hilus of the dentate gyrus usually evoked a single, short-latency extracellular population spike from the granule cells, and simultaneous intracellular recordings showed an antidromic action potential that was synchronous with the population spike (Fig. 1A1 and B1). In response to perforant path stimulation, which produces synaptic excitation of granule cell dendrites, slices with or without mossy fiber sprouting had a single population spike superimposed on an extracellular positive-going wave, while an action potential rising from an excitatory postsynaptic potential (EPSP) was recorded intracellularly (Fig. 1A2 and B2). Fast and slow inhibitory postsynaptic potentials (IPSPs) appeared normal, and no spontaneous activity in normal ACSF was noted in slices with or without mossy fiber sprouting. Thus, 1-4 months after kainate treatment, granule cells with mossy fiber sprouting appeared electrophysiologically normal. We then determined whether an acute decrease in in-

No Sprouting

A

1

Hilus

2

B Sprouting 1

i

Perforant Path

_ , 2 0 mY

2

20rnsec

Fig. 1. Normal responses in slices with mossy fiber sprouting. In normal medium 1-4 months after kainate treatment, hilar (left) and perforant path (right) stimulation elicited normal responses in hippocampal slices with mossy fiber sprouting when compared to controis. In this and subsequent figures, the top traces are intracellular records and the lower traces are extracellular records. Hilar stimulation in control slices (A1) and slices with mossy fiber sprouting (B1) evoked normal intracellular action potentials and extracellular population spikes. Stimulation of the perforant path also evoked normal intracellular action potentials and population spikes in control slices (A2) and slices with mossy fiber sprouting (B2).

hibition would lead to a greater propensity for epileptiform activity in slices with mossy fiber sprouting than in slices without sprouting. This was accomplished by treating slices with bicuculline methiodide (10-30/~M, Sigma) and then observing spontaneous activity and the responses to electrical stimulation of the hilus. We first tested whether hilar stimulation could evoke epileptiform bursts in bicuculline-treated slices. Low-intensity hilar stimulation in slices without mossy fiber sprouting evoked single, low-amplitude population spikes in 10-30 /~M bicuculline (Fig. 2A1). As the stimulus intensity was raised, the population response became larger and an antidromic action potential was recorded intracellularly (Fig. 2A2). In contrast, about one-quarter of the slices (5 of 18 slices; 5 of 17 animals) with mossy fiber sprouting responded to low-intensity hilar stimulation with a long-latency negative field potential in bicuculline, and occasionally small population spikes were superimposed on the negativity. Intracellular recordings showed that the granule cells h a d large EPSPs and/or depolarization shifts with action potential bursts during the extracellular negativity (Fig. 2B1). The delay between stimulus and response varied across stimulations and was reduced at higher intensities (Fig. 2B2). Therefore, although the responses to hilar stimulation were normal in control ACSF, 28% of the slices with mossy fiber sprouting showed delayed EPSPs and/or synchronous bursts of electrical activity in bicuculline. These delayed events

307

A

No Sprouting

~,.

1

-

~

•w

B

. . . .

20mY 5 mV

20 msec

Sprouting 3

"

20 mv 5 mV •

20 m s e c

Fig. 2. Response in bicuculline to hilar stimulation from a control slice (A) versus a slice with mossy fiber sprouting (B). A: in bicuculline, the normal response to hilar stimulation in slices without mossy fiber sprouting was a short-latency population spike. AI: at low intensities (10-50/~A), the population spike was very small and did not cause an antidromic action potential in the impaled granule cell. A2: at higher intensity (50-100/~A), hilar stimulation evoked a larger antidromic population spike and triggered an antidromic action potential in the recorded cell. A3: light micrograph of area around the granule cell layer (G) with combined Nissl (blue) and Timm's stain (dark brown) from a control slice without mossy fiber sprouting (bar = 50/~m). B: in some slices with mossy fiber sprouting, weak hilar stimuli evoked long-latency bursts. BI: at very low intensities of hilar stimulation (10-50/zA), bursts only occurred intermittently and their latencies were long and variable. B2: at a higher intensity (50-100/~A), the bursts had a shorter but still variable delay. B3: light micrograph of granule cell layer (G) as above A3, but from a slice with mossy fiber sprouting (arrows; Bar = 50/~m). In each panel of AI&2 and BI&2, the top extracellular trace was recorded simultaneously with the intracellular record, and the other extracellular traces show responses to subsequent stimuli at the same intensity.

never occurred in control slices without mossy fiber sprouting (0 of 13) 5. These findings are consistent with (but do not prove) the hypothesis that mossy fiber sprouting leads to new local excitatory circuits among dentate granule cells which can lead to synchronous burst discharges. Previous studies have shown that spontaneous epileptiform bursts do not normally occur in the d e n t a t e granule cells when inhibition is depressed 1°A2. Therefore, we tested whether slices with mossy fiber sprouting could generate spontaneous bursts of population spikes in bicuculline. Small-amplitude, positive-going field potentials occurred spontaneously in some bicuculline-treated slices without mossy fiber sprouting (Fig. 3A); population spikes were not o b s e r v e d and intracellular recordings indicated that the granule cells did not fire action potentials during these events. H o w e v e r , when slices

with mossy fiber sprouting were perfused with 10-30/~M bicuculline, spontaneous negative field potentials lasting 50-400 ms occurred in the d e n t a t e gyrus in one-third (6 of 18 slices; 6 of 17 animals) of the p r e p a r a t i o n s (Fig. 3B). Small population spikes were often s u p e r i m p o s e d on the negative field potentials. Intracellular recordings showed that the granule cells had large EPSPs and/or slow depolarization shifts with multiple action potentials during these events. Spontaneous negative-going field potentials and synchronous bursts of action potentials in bicuculline were only seen in those slices with mossy fiber sprouting and were never detected in the n o r m a l dentate gyrus or in preparations from kainate-treated rats without mossy fiber sprouting (0 of 9 slices). W h e n considered relative to earlier electrophysiological research by Tauck and N a d l e r 3° on h i p p o c a m p a l slices from chronic k a i n a t e - t r e a t e d rats, it was surprising

308 that the evoked responses of dentate granule cells in normal ACSF were normal in slices with sprouting, and only showed epileptiform activity when inhibition was depressed. Tauck and Nadler 3° studied granule cell responses with extracellular recordings only 12-21 days after kainate treatment; therefore, the bursts they observed to hilar stimulation in slices with sprouting could have been associated with decreased synaptic inhibition, which occurs immediately after kainate treatment ~'t1'21' 25 The discrepancies between our data and those of Tauck and Nadler 3° could reflect differences in methodology, and further experiments are required to resolve this issue. Computer modeling and experimental studies on CA3 pyramidal cells have indicated that neuronal populations with recurrent excitatory interactions are normally under tonic inhibitory control 4'8'9'15'16-18'29. When inhibition is reduced, however, activity in a few neurons can propagate to an increasing number of neurons as excitation cascades through local excitatory circuits. An electrical event can be recorded from the neuronal population after recruitment and synchronization progress to a suffi-

NO SPROUTING

SPROUTING

A

&.._ 1,/"

//' /.' /'

///

, , , " " " I 510mvmV 1.5 sec

i /; /,,;

1.5 s e e

~i

20 mV 1.5 mV

/

J lOmV 5 mV 100 msec

-

" ~ .... I

O

m

V . ~ i mV 100 m s e c

Fig. 3. In bicuculline, some slices with mossy fiber sprouting showed spontaneous bursts consisting of prolonged negative shifts in the extracellular field potential and synchronous action potentials. A: slices without sprouting had small intracellular depolarizations followed by hyperpolarizations and positive-going extracellular fields;

no action potentials or extracellular population spikes were observed. B: in slices with sprouting, large spontaneous depolarizations that evoked intracellular action potentials could occur synchronously with slow extracellular negative shifts and population spikes. Arrows show expansion of traces.

cient level. For example, spontaneous bursts occur in the CA3 region in bicuculline-treated slices, and low-intensity stimulation (i.e. that only activates a small fraction of the neurons) evokes population bursts with long and variable latencies in bicuculline 23'24'31. Detailed computer models have suggested that the population response latency varies across stimulations of equal intensity because synaptic activity spreads along different local excitatory pathways 2s'29. The delay is shorter with stimuli of higher intensity, because more cells are excited initially and thus less time is required to recruit the population. In our experiments, low-intensity hilar stimulation evoked population bursts with long and variable delays in the dentate gyrus of slices with mossy fiber sprouting bathed in bicuculline. The presence of new local excitatory circuits among granule cells in slices with mossy fiber sprouting would also be expected to generate spontaneous bursts of action potentials and compound synaptic potentials in bicuculline. Furthermore, new excitatory synapses close to the granule cell bodies would lead to a current sink and a slow negative field potential in this area, as we observed. There is ultrastructural evidence for synaptic contacts between mossy fiber terminals and the proximal dendrites of granule cells in animals with lesion-induced reorganization of mossy fibers 13 and such morphological findings are consistent with the present evidence for recurrent excitation in kainatetreated rats with mossy fiber sprouting. Light microscopic studies also suggest that some mossy fibers normally contact interneurons resembling GABAergic neurons in the granule cell and molecular layers of the dentate gyrus 22. Such putative connections with inhibitory neurons could contribute to the control of granule cell excitability until local inhibition is perturbed. Therefore, our data support the hypothesis that after kainateinduced loss of hilar neurons (which project to the granule cells), new local excitatory circuits are formed in the dentate gyrus and these circuits impart an enhanced epileptogenicity. The presence of synaptic inhibition, however, masks the ability of the synaptically reorganized dentate gyrus to generate seizure activity. Although these experiments provide evidence that new local excitatory circuits have formed in the dentate gyrus after mossy fiber sprouting, dual intracellular recordings are required to demonstrate conclusively that monosynaptic excitatory connections exist among granule cells. Furthermore, the effects on granule cell electrophysiology of other l~ainate-induced alterations associated with sprouting (e.g. hilar cell loss) also need to be determined. The two main findings in the present study are that the responses of granule cells to extracellular stimulation were normal in slices with kainate-induced mossy fiber

309 sprouting, but that in a proportion of the slices with mossy fiber sprouting, the granule cells showed abnormal epileptiform activity when bathed in bicuculline-containing solution (i.e. 8 of 18 (44%) slices showed delayed bursts to hilar stimulation and/or spontaneous negativegoing events). There are at least two possible reasons why these alterations were not observed in all of the slices with sprouting. First, both the delayed synchronous bursts of action potentials to hilar stimulation and the spontaneous bursts of synchronous activity presumably require the existence of sufficient local excitatory circuits to excite the population 4A6-1s'Es. Modeling studies have argued that if recurrent excitatory circuits are too limited in number, then burst synchronization will not occur 28'29. Thus, for example, if sprouted mossy fibers projected out of the plane of the slice to any substantial degree, it is quite likely that delayed bursts to hilar stimulation or spontaneous bursts of synchronous activity would not occur. A n o t h e r possible reason why some slices with mossy fiber sprouting did not show abnormal responses, even in bicuculline, is that some proportion of the sprouting may not have been to granule cells but instead to inhibitory neurons, and this could be variable across preparations. Sloviter 25 has observed loss of granule cell inhibition shortly after kainate treatment, and this recovers as mossy fiber sprouting occurs during the subsequent weeks; preliminary data from our laboratory 21 are consistent with his observations. The results of our experiments suggest that kainateinduced seizures and excitotoxic damage in the hippo-

1 Ashwood, T.J. and Wheal, H.V., Loss of inhibition in the CA1 region of the kainic acid lesioned hippocampus is not associated with changes in postsynaptic responses to GABA, Brain Res., 367 (1986) 390-394. 2 Babb, T.L., Kupfer, W.R., Pretorius, J.K., Crandall, P.H. and Levesque, M.E, Synaptic reorganization by mossy fibers in human epileptic fascia dentata, Neuroscience, 42 (1991) 351-363. 3 Ben-Ari, Y., Limbic seizure and brain damage produced by kainic acid: mechanisms and relevance to human temporal lobe epilepsy, Neuroscienee, 14 (1985) 375-403. 4 Christian, E.P. and Dudek, EE., Characteristics of local excitatory circuits studied with glutamate microapplication in the CA3 area of rat hippocampal slices, J. Neurophysiol., 59 (1988) 90-109. 5 Cronin, J. and Dudek, EE., Multiple population spikes in the dentate gyrus to hilar stimulation after blockade of G A B A A receptors, Soc. Neurosci. Abstr., 15 (1989) 700.

6 Danscher, G., Histochemical demonstration of heavy metals. A revised version of the sulphide silver method suitable for both light and electron microscopy, Histochemistry, 71 (1981) 1-16. 7 de Lanerolle, N.C., Kim, J.H., Robbins, R.J. and Spencer, D.D., Hippocampal interneuron loss and plasticity in human temporal lobe epilepsy, Brain Res., 495 (1989) 387-395. 8 Dichter, M. and Spencer, W.A., Penicillin-induced interictal discharges from the cat hippocampus. I. Characteristics and topographical features, J. Neurophysiol., 32 (1969) 649-662. 9 Dichter, M. and Spencer, W.A., Penicillin-induced interictal discharges from the cat hippocampus. II. Mechanisms underly-

campus 3'19, followed by formation of new local synaptic circuits in the dentate gyrus, may increase the susceptibility of granule cells to abnormal seizure-like activity. Although the responses of granule cells to extracellular stimulation appeared normal after mossy fiber sprouting had occurred, reductions in GABAA-mediated inhibition allowed expression in some preparations of new network properties, which have been shown in other hippocampal systems to be characteristic of local excitatory circuits 4'8'9'17As'29. Reduced inhibition occasionally unmasked a latent form of epileptogenicity, manifested here as spontaneous and synchronous bursts of epileptiform activity. Although other physiological changes (e.g. alterations in the number or properties of N M D A receptors) could have occurred in the dentate granule cells that underwent mossy fiber sprouting, our results suggest how synaptic reorganization could contribute to the sporadic nature of seizures in the epileptic brain. Transiently decreased levels of inhibition, caused by repetitive activity or increases in extracellular [K+], could temporarily enhance multisynaptic transmission through new local excitatory circuits and lead to epileptiform activity.

We thank Dr. R.S. Sloviter for advice on the kainate model, for sharing his unpublished data with us, and for comments on the manuscript. We are grateful to G. Allen for technical assistance and L. Auzins for word processing. We thank Drs. W. Dunlap, A. Gerall, C. Meier, G.J. Strecker, J.G. Tasker and J.-E Wuarin for comments on earlier versions of the paper. Supported by NIH Grants NS16683 (EE.D.) and NS21908 (C.R.H.).

ing origin and restriction, J. Neurophysiol., 32 (1969) 663-687. 10 Fournier, E. and Crepel, F., Electrophysiological properties of dentate granule cells in mouse hippocampal slices maintained in vitro, Brain Res., 311 (1984) 75-86. 11 Franck, J.E. and Schwartzkroin, P.A., Do kainate-lesioned hippocampi become epileptogenic, Brain Res., 329 (1985) 309-313. 12 Fricke, R.A. and Prince, D.A., Electrophysiology of the dentate gyrus granule cells, J. Neurophysiol., 51 (1984) 195-209. 13 Frotscher, M. and Zimmer, J., 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 (1983) 299-311. 14 Houser, C.R., Miyashiro, J.E., Swartz, B.E., Walsh, G.O., Rich, J.R. and Delgado-Escueta, A.V., Altered patterns of dynorphin immunoreactivity suggest mossy fiber reorganization in human hippocampal epilepsy, J. Neurosci., 10 (1990) 267282. 15 MacVicar, B.A. and Dudek, EE., Local synaptic circuits in rat hippocampus: interactions between pyramidal cells, Brain Res., 184 (1980) 220-223. 16 Miles, R. and Wong, R.K.S., Single neurones can initiate synchronized population discharge in the hippocampus, Nature, 306 (1983) 371-373. 17 Miles, R. and Wong, R.K.S., Excitatory synaptic interactions between CA3 neurones in the guinea-pig hippocampus, J. Physiol., 373 (1986) 397-418. 18 Miles, R. and Wong, R.K.S., Inhibitory control of local excitatory circuits in the guinea pig hippocampus, J. Physiol., 388

310 (1987) 611-629. 19 Nadler, J.V., Kainic acid as a tool for the study of temporal lobe epilepsy, Life Sci., 29 (1981) 2031-2042. 20 Nadler, J.V., Perry, B.W. and Cotman, C.W., Selective reinnervation of hippocampal area CA1 and the fascia dentata after destruction of CA3-CA4 afferents with kainic acid, Brain Res., 182 (1980) 1-9. 21 Obenaus, A., Cronin, J., Houser, C.R. and Dudek, EE., Shortand long-term changes in the electrophysiology of dentate granule cells of kainate-treated rats, Epilepsia, 31 (1990) 608. 22 Ribak, C.E. and Peterson, G.M., Intragranular mossy fibers in rats form synapses with the somata and proximal dendrites of basket cells, Soc. Neurosci. Abstr., 16 (1990) 123. 23 Schwartzkroin, P.A. and Prince, D.A., Cellular and field potential properties of epileptogenic hippocampal slices, Brain Res., 147 (1978) 117-130. 24 Schwartzkroin, P.A. and Prince, D.A., Changes in excitatory and inhibitory synaptic potentials leading to epileptogenic activity, Brain Res., 183 (1980) 61-76. 25 Sloviter, R.S., Reorganization hypothesis of epilepsy; Granule

26

27

28

29 30

31

cell sprouting in rats is associated with decreased, not increased, excitability, Epilepsia, 31 (1990) 675. Sutula, T., Cascino, G., Cavazos, J., Parada, I. and Ramirez, L., Mossy fiber synaptic reorganization in the epileptic human temporal lobe, Ann. Neurol., 26 (1989) 321-330. Sutula, T., Xiao-Xian, H., Cavazos, J. and Scott, G., Synaptic reorganization in the hippocampus induced by abnormal functional activity, Science, 239 (1988) 1147-1150. Traub, R.D. and Wong, R.K.S., Penicillin-induced epileptiform activity in the hippocampal slice: a model of synchronization of C A 3 pyramidal cell bursting, Neuroscience, 6 (1981) 223-230. Traub, R.D. and Wong, R.K.S., Cellular mechanism of neuronal synchronization in epilepsy, Science, 216 (1982) 745-747. Tauck, D.L. and Nadler, J.V., Evidence of functional mossy fiber sprouting in hippocampal formation of kainic acid-treated rats, J. Neurosci., 5 (1985) 1016-1022. Wong, R.K.S. and Traub, R.D., Synchronized burst discharge in disinhibited hippocampal slice. I. Initiation in CA2-CA3 region, J. Neurophysiol., 49 (1983) 442-458.