Neuron loss, mossy fiber sprouting, and interictal spikes after intrahippocampal kainate in developing rats

EPILEPSY RESEARCH ELSEVIER

Epilcps3, Re~carci~26 ( !9961 219 231

Neuron ioss, mossy fiber sprouting, and"mtei ..:,-,a!l spikes after intrahippocampal kainate in developing rats Jo~o P. Leite ~'~'*'~T h o m a s L. Babb ~ J a m e s K. Pretorius ~ Paula A. K u h l m a a ~ 5~

9

9

Krisfin M. Yeoman a Gary W. Mathern a.b.~ ~' The Brain Research lnstituh'. Uni~ ersitv q/'Cali/ornia, lx~s Angeie.~. CA. USA h Dclmrtment qf/Ve'urology. Uniler.~itv r~fCali/'ornia. Lo~ Angeles. CA. USA ¢ Dicision ofNeutwsurger3; Unh,ersio" of Califi,7~ia. Los Angeles. CA, USA Department of Netoology, Ribeirfio Preto SchorF of Medici~w, Unit'e:'xit3' of S~to Paulo, Ribeirfio Prelo. S~o Paulo. Brazil c DepaJwnents of Neuroscience and Neutoh~3, Ctet ~ehmd Clinic Foundation. Cleveland. OH. USA

Received 26 October 1995; accepted 12 March 1996

Abstrac~ This study determined neuron losses~ mossy fiber sprouting, and fllterictal spike frequencies in adult rats following intrahippocampal kainic acid (KA) injections during postnatal (PN) development. KA (0.4 p-g/0.2 p.k n = 64J was injected into one hippocampus and saline into the contralateral side between PN 7 to 30 days. Animals were sacrificed 28 to 256 days later, along with age-matched naive anflnals (controls: n = 20). Hippocampi were studied for: (!) Fascia dentata granule ceil, hilar, and CA3c neuron counts; (2) neo-Timm's stained supragranular mossy fiber sprouting; and (3) hippocampal and intracerebral interictal spike densities (n = 13). Mossy fiber sprouting was quantified as the gray va!ue differences between the inner and outer molecular layer. Statistically significant results ( p < 0.05) showed the following: (I) Compared to controls, CA3c and hilar neuron counts were reduced in KA-hippocampi with injections at PN 7-10 and PN 12-14 respectively and counts decreased with older PN injections. Granule cell densities on the KA-side and saline-injected hippocampi were not reduced compared to controls. (2) In adult rats, supragranular mossy fiber sprouting was observed in 2 of 7 PN 7 injected animals. Compared to controls, increased gray value differences, indicating mossy fiber sprouting, were found on the KA-side beginning with injuries at PN 12-14 and increasing with older PN injections. On the saline-side only PN 30 animals st',~wed minimal sprouting. (3) Mossy fiber sprouting progressively increased on the KA-side with longer survivals in rats injured after PN 15. Sprouting correlated positively with later PN injections and longer post-injection survival intervals and not with reduced hilar or CA3c neuron counts. (4) On the KA-side, mossy fiber gray value differences correlated positively with in vivo intrahippocanlpal interictal spike densities. These results indicate that during postnatal rat development intrahippocampal kainate excitotoxicity can occur as early as PN 7 and increases with older ages at injection. This rat model reproduces many of the pathologic, behavioral, and electrophysiologic features of human mesial temporal

* Corresponding author. Reed Neurological Research Center, UCLA Medical Center, Los Angeles, CA 90095-1769. Fax: + I (310) 206-8461. 0920-1211/96/$15.00 Copyright © 1996 Elsevier Science B.V. All rights reserved. Pll S0920- 1 21 I (96)00055- I

220

J.P. Leite et al. / Epilepsy Researrh 26 (1996) 219-23 l

lobe epilepsy, and supports the hypothesis that hippocampal sclerosis can be the consequence oJ" focal hljury during early postnatal development that progressively evolves into a pathologic and epileptic focus. Kevwords: Epilepsy; Synaptic reorganization;Ontogeny; NIH image

1. Introduction Hippocampal sclerosis (HS) is the most frequent pathological finding in patients with temporal lobe epilepsy [7]. In addition, recent human studies have shown that sclerosis is associated with reactive synaptogenesis of surviving axons [8,14,20,2729,45]. This has suggested to several authors that in temporal lobe epilepsy abnormal hippocampal axon circuitry may contribute to neuron hyperexcitability and seizure generation. Other authors have however suggested that hippocampal axon sprouting is not necessary for seizure generation and may be an epiphenomena related to the neuron damage [42]. Data from clinical-pathologic studies on patients with seizures supports both viewpoints. For example, in temporal lobe epilepsy patients there is an association between an early brain injury (e.g. prolonged febrile convulsion, trauma, etc.) and the finding of HS at surgery or autopsy supporting the notion that sclerosis and presumably axon sprouting was caused by the initial injury and eventually evolved into an epileptogenic focus [7,14,16,24,27,29,30,39]. However, other clinical epidemiological studies indicate that the risk of epilepsy following childhood seizures is very low [5,6,16,35,39]. For example, two recent studies of status epilepticus in children found a low risk for subsequent seizures in the absence of acute neurologic insults [15,31]. Such findings support the view that seizures during early human development do not cause excitotoxic hippocampal injury, adult hippocampal sclerosis, and temporal lobe epilepsy. Experimental studies have also shown differing amounts of hippocampal damage and axon sprouting depending on the experimental design and the age of the animals. For example, in paralyzed ventilated adult baboons prolonged sustained seizures damage hippocampal neurons, supporting that seizures can induce excitotoxic neuron loss [33]. Developmental studies however have not consistently shown hippocampal damage as a consequence of seizure activity supporting the notion that the immature hippocampus may be resistant to seizure associated

excitotoxicity (for review see [50]). One problem with many developing rat studies is that immature animals have a lower excitoto×ic threshold ~br reactive seizures and status epilepticus compared to adult animals. As a consequence these studies use lower doses of excitotoxins in order to optimize animal survival and the apparent neuropathoiogic resistance may depend on the lower amounts of excitotoxic injury [34,49]. For example, compared to adult rats, lower doses per body weight (mg/kg) of parenteral kainic acid (KA) evokes status epilepticus in neonates, and does not seem to generate visible hippocampal damage or mossy fiber synaptic reorganization before postnatal (PN) day 18 [9,10,36,43,44]. Yet, directly injected intrahippocampal KA, which is a more powerful excitotoxin, can induce hippocampal neuron losses and mossy fiber sprouting in young animals [11,12], and perforant path stimulation has also been associated with damage in the immature hippocampus [47]. Hence, the type of excitotoxicity may determine the amount of hippocampal damage during postnatal development, and it is therefore important to study hippocampal neuropathology using equal excitotoxic insults. In other words, in neonatal hippocampi it may be that reactive seizures using lower does of excitotoxins are not as strong as other excitotoxic insults in producing neuron damage and sprouting. This study was designed to determine in developing hippocampi whether seizure- versus chemicallyinduced excitotoxicity showed differences in neuron loss and mossy fiber sprouting and at what developmental age does comparable excitotoxic injury mimic the pathology of adult human hippocampal sclerosis. We tested the following hypotheses: (1) Mossy fiber sprouting and neuron losses would occur with intrahippocampal KA at young PN days and would increase with older PN injuries; (2) there would be minimal neuron loss and mossy fiber sprouting in saline-injected hippocampi that experienced KA-induced postnatal seizures but not the direct chemical excitotoxicity; (3) once damaged, mossy fiber sprouting would increase with longer survival periods; and

J.F. Leile el uL / E[)ihT~sr Re~curck 26 (1996)

(4) mossy fiber sprouting would correlate with increasing interictat EEG activity.

2. Materia~ a~d megheds 2.1. Surgeo, Sixty-four male Sprague-Dawley rats, ranging in age from 7 to 30 days were anesthetized with chloral hydrate (400 m g / k g , i.p.), given atropine sulfate (0.04 rag, i.m.), and kainic acid (KA; 0.4 rag/0.2 /al) was stereotacfically injected over 20 rain into the right posterior hippocampus using a microsyringe (anterior to lambda 2.1 mm for PN 7 to 2.5 for PN 30; lateral to midline 3.6 fi~r PN 7 to 4.5 for PN 30: and vertical from the cranium 4.0 for PN 7 to 5.0 for PN 30) [41]. Buffered saline (0.2 /al) was similarly injected into the homologous left hippoeampus. The brains were processed 28 to 256 days after KA in batches of four animals including a naive litter-mate matched for age at the time of sacrifice (controls). 2.2. h~terictal EEG recordings Four weeks prior to sacrifice a subset of 13 randomly selected rats sampling PN 7, 10, 12, 14, 18 and 30 KA-injected animals were implanted as adults with bilateral intracranial electrodes into both hippocampi and frontal cortices, according to a previous protocol [26]. Two depth electrodes were placed into each hippocampus and a single electrode into the right and left frontal neocortex. One week folBowing implantation and for the next 3 weeks each rat was EEG monitored between 21.5 to 99 h using bipolar chain-linked montages sampling both hippocampi and the neocortex simultaneously. EEG data were digitized into a computerized system (Biomedical Monitoring Systems Inc., Nicolet~), formatted with the video signal to produce a split-screen EEG and video image, and analyzed off-line. On-line spike and seizure detection programs quantified interictal spike densities and spontaneous chronic seizures [18]. 2.3. Tissue process and staining At sacrifice, animals were deeply anesthetized with pentobarbital and perfused through the heart

,.TM] 9.-.-. -. : , -'

22]

with the fctlowing: Buffered normal sa!ine %r 5 rain (100 ml), 0.1% sodium sulphide in buffer (pH 7.3) for 10 rain (150 ml), and 4% para~rmaldehyde for 10 rain {~50 ml). The brains were removed, blocked, kept in fixative for 24 h (4°C) and placed overnight in a R0% buffered sucrose solution. For each rat, t0 coronal sections covering the entire septo-temporal length of the hippocampus were dewtoped for neoTimm histochemistry (30 /am thicK), and Nissl stained for histopathologic review and cell counts (30 p,m and l0 lain respectivel>). The method of neo-Timm's processing and development are as previously described [8,37]. 2.4. Neuron counts Counts were performed ipsilateral and contralateral to the KA-injection at one site per hippocampus near the region of inju~'. The hippocampa! subfields were counted based on the Lorente de N6's classification [22], and included the fascia dentata granule cells, hilar neurons and the pyramidal cells within the two blades of the stratum granulosum. These later neurons between the granule eel] blades were collectively termed CA3c which incorporates distal CA3 and CA4 pyramids. Granule cell densities, measured as neurons per cubic millimeter, were estimated in 10 p~m slides at 400 × using grid morphometric techniques with Abercrombie's con'ection El]. CA3c pyramids (10 ,am sections) and hilar cells (30 /am) were counted within the fascia dentata blades and data presented as the number of neurons per section. 2.5. Mossy fiber sprouting The optical density of mossy fiber terminals in the Timm's sections adjacent to the cell counts were mzasured by a computer-based image analysis syste.n as a gray value (GV) between white (0) to black (255). Video signals were collected by a high resolution CCD monochrome camera attached to a Zeiss microscope and digitized into a Macintosh computer (Model 8100/80) using a frame grabber and image system software (public domain NIH Image program, 1.56). The operator imaged the fascia dentata molecular layer between the hippocampal fissure to the stratum granulosum, the inner molecular layer

222

J.P. Leite et al./ Epilepsy Research 26 ¢1996) 2t9-231

(IML) or outer molecular layer (O.ML) was outlined, and the average GV within the encircled region determined. The OML GV was used as an internal background reference to control for the potential bias of animals processed in different batches. The average (5: SEM) OML GV for K_A-injectedhippocampi were 64.57 + 1.5; saline-side 64.64 _ 1.3; and agematched controls 63.86 5- 2.8 (p = 0.85). IML mossy fiber sprouting was quantified as the calculated GV difference between the IML and OML. 2.6. Data analysis Data were entered into a database on a personal computer and analyzed using a statistical program (Super ANOVA, version 1.1, Abacus Concepts, Inc., Berkeley, CA). Differences between PN groups were statistically compared using analysis of v~_riance (ANOVA) and further compared between individual groups (at p < 0.05) using the Games-Howch test that controls for multiple comparisons of unequally sized samples and of unassumed variances. Other statistical tests included regression analyses and analysis of covariance (ANCOVA). Results were plotted with Deltagraph Professional (Delta Point, Inc., Monterey, CA), and were considered statistically significant at a confidence level of p < 0.05.

3. Results 3.1. General information There were no differences in the number of animals or the sacrifice age between the different PN categories. Of the 64 rats injected with intrahippocampal KA, there were 13 at PN 7-10, 13 at PN 12-14, 12 at PN 15-16, 16 at PN 17-18, and 10 at PN 30. The average (___SEM) number of days between injection and perfusion for all animals was 134 5- 9.5 and there were no differences between the PN categories ( p = 0.08). Likewise, the ~tverage sacrifice age for the KA-treated and naive control animals averaged 149 + 9.6 days and was not different between PN categories or between PN groups and controls ( p > 0.11). In other words, despite differences in the ages when KA was injected, there were

no differences t~etween the PN groups in fl~e post-injury days (PID) or the age at sacrifice including the naive control rats. 3.2. Acute behavioral seizures With intrahippocampal KA the acute generalized seizures differed with PN age as previously Described for parenteral KA injections [2,10,19~481. The youngest rats (PN 7-1a) exhibited scratching behavior with a few wet rat shakes and body tremors, Slightly older rats (PN !5-18) displayed more limbic-like seizures characterized by masticatory movements, salivation, facial and forelimb clonus, rearing, and loss of postural control. PN 30 animals showed stereotyped limbic-motor seizures a~d intense wet rat shakes between seizures. The acute EEG from PN 30 animals showed continuous spiking activity in the right hippocampi with eventual episodic spread into the contralateral hippocampi and frontal cortex electrodes. After 1 to 2 h the limbic seizures clustered and evolved into sustained bilateral limbic status lasting for another 3 to 4 h. 3.3. Neuron counts Fig. 1 illustrates the quantified granule cell, hilar, and CA3c neuron counts. For granule cells, neither KA- nor saline-hippocampi (lesion and non-lesion respectively) showed differences between naive controls and the PN groups (Fig. 1; top bar graphs). For hilar neurons, compared to controls reduced neuron counts were found in PN 12-14 and older animals only on the KA-side (Fig. 1; middle left graph). For CA3c pyramids, compared to controls there were fewer neuron~ in the PN 7-10 and older groups again only on the KA-side (Fig. 1; lower left graph). Regression analyses comparing cell counts with age at sacrifice showed that granule cell neuron densities in both KA and saline-injected hippocampi increased in older aged animals ( p =0.0001 and p = 0.012 respectively). In other words, the older the animal the greater the granule cell der~sities consistent with previous work showing that granule cell neurogenesis persists throughout the life of Sprague-Dawley rats [3,21,40]. Hilar and CA3c cell counts were not different in older aged animals ( p > 0.7).

22,.3

J.P. Le#e e~ al.. / Epi/ep.~y Researcf~ 26 ¢I996) 2t9-23~;

450000

CONTROL

~

PN 12-t4

~

PN 17o!8

PN7°I0

~

PN t5o18

~

PN30

(p = 0.0898)

I

(p = o.8~5)

@}

LESION S~DE

,2~i/./)//-&'-.\\\'~,..... NON LESION SiDE 0.1829)

Z 0

i

W Z ,,J i

l

LESiON SIDE

t~ Z 0

200180-

ILl o ¢,~

120£ 100-

t,,,} '<

60 80l

0 P

20-

:~

,.J

(p =0.0001)

NON LESION SiDE

(p=o.9712)

160140-

402 O't LESION SiDE

NON LESION SiDE

Fig. 1. For different postnatal days, quantification of hippocampal neuron counts with unilateral kaJnate injections (lesion side) compared to saline-injected hippocampi (non lesion) and naive controls. The ANOVA p values are shown above each graph and the significant post-hoc differences indicated with an asterisk. For granule cells (top row), neither the lesion nor the non-lesioned sides showed differences in neuron densities, and the same was found on the non-lesioned side for hilar (middle row) and CA3c counts (bottom row). For hilar counts, the kainate-injected side showed that the PN categories were different ( F = 26.78, p = 0.0001), and the post hoc tests found that PN ! 2-14, PN !5-16, PN 17-18 and PN 30 groups were less than controls ( p < 005, Games-Howell). In addition, PN 30 was less than all groups. The CA3c cell counts on the lesion side showed that the PN categories were different ( F = 35.08, p = 0.0001) and all PN categories were less tl~s, control and PN 30 less than all groups ( p < 0.05, Games-Howell). Note that hilar and CA3c neuron counts decreased with older ages a~ ~N injections.

224

J.P. Leiw et al./ Epilepsy Research 26 (1996) 219-231

3.4. Mossv fiber sprouting In summary, aberrant fascia dentata mossy fiber sprouting on the lesioned side increased as a function of older PN dates at injury and longer survivals following KA injection. Fig. 2 illustrates Nissl and neo-Timm staining from two animals given KA at different PN dates but with a similar sacrifice age. The rat on the left (Fig. 2A, C, and E) received

intrahippocampat kainate at PN 18 and was sacrificed 188 days later (sacrifice age 206 days), and the animal on the right (Fig. 2B, D, and F) was injected at PN 30 and survived for 147 days (sacrifice age 177 days). By neuron counts, the PN 18 animal had slightly more hilar (16%) and CA3c (13%) damage compared to the PN 30 rat (14 and 4% respectively; Fig. 2A and 2B); however, tl,,~ neo-Timm sections show that the PN 30 ammal had greater supragranu-

Fig. 2. Photomontage illustrating the a g e dependent mossy fiber supragranul~ reacfve synaptogenesis. Top row shows a Nissl stained dorsal section of a PN 18 injected rat (A) compared to a PN 30 animal (B). Note that the neo-Timm stained sections show that the PN 18 animal has less sprouting (C), higher magnification (E) than the rat injured at PN 30 (D), higher magnification (F). Both animals had long survival times after intrahippocampalKA (over 130 days). Calibrationbars: Low magnificationbar = 100/zm, high magnificationbar = 50 /zm.

,/.P.

0

! 20

-

Leile ~'," M./kpi/q~.V ~"6'.~v,,,,nh2e~ ¢/Vc~6~219-23!

CONTROL

[ ] PN 12-14

t'~ PN 17-18

PN7-!0

~

~

PN 15-t(;

225

PN30

II~ ~ 0 O2Ogt

100

A

,m tiE 8060-

,

4o

,

:!:~:i:i:i:~:i

.i;:i#ilil Z

LESION

S~DE

NON LESION SIDE

Fig. 3. The gray value differences from control and kainate-mjected hippocampi shm.,in,,_,, increased supragranuiar mossy fiber sprouting with older postnatal (PN) injections. Data presented as the average gray value (+SEM) differences between ,,he inner molecular layer (IMLt and outer molecular layer ffOML) the of KA- (left: lesion) and saline-injected (right; non lesion) hippocampi compared against naive comro~s, in summ:~.ry, on the KA-side there ~as increased s~aining in all PN groups after PN 7-10 compared to controls ( p = 0.0001; see asterisk). Evidence of some comralateral sprouting was also obser~,ed ( p < 0.02, bar with asterisk) ~n the non-lesioned hippocampus of PN 30 ra~s. Analysis of variance (ANOVA) for |ML-OML GV differences on the lesion side indicated that the PN categories were different (F = 21.00, p = O.O001), and the post hoc statistical tests showed that the PN 12-14, PN 15-16, PN 17-18 and PN 30 groups were more than controls { * p < 0.05, Games-Howel|). For the non-lesion side the PN categories were different (F = 2.87, p = 0.02), and the post hoc statistical tests showed tha~ fl~e PN 30 group was more than controls ( " p < 0.05, Games-Hov~ell).

8 ~-T..

"C 0

lar s p r o u t i n g than the a n i m a l injured at PN 18 (Fig. 2 C - F ) . T h e visual a s s e s s m e n t s w e r e m i r r o r e d in the q u a n t i t a t i v e g r a y value m e a s u r e m e n t s as s h o w n in Fig. 3. C o m p a r e d to controls, o n the K/k-side (lesion) I M L - O M L G V d i f f e r e n c e s w e r e greater in rats injected at P N 1 2 - 1 4 and i n c r e a s e d with injections in o l d e r P N a n i m a l s . O n the s a l i n e - s i d e ( n o n - l e s i o n ) o n l y P N 30 a n i m a l s s h o w e d s o m e m o s s y f i b e r sprouting that w a s greater t h a n controls. I n s p e c t i o n o f the P N 30 a n i m a l s with contralateral sprouting did not find significant neuron losses in the saline-side, but there w a s r o b u s t d a m a g e a n d sprouting on the K A - s i d e . H e n c e , the m o s s y f i b e r sprouting contralateral to the K A - d a m a g e m o s t likely represents local synaptic r e o r g a n i z a t i o n on the saline-side as a c o n s e q u e n c e o f lost c o m m i s s u r a l p r o j e c t i o n s . A s m i g h t b e expected, regression analyses comparing IML-OML g r a y v a l u e d i f f e r e n c e s f r o m the K A - s i d e with n e u r o n counts f o u n d an inverse c o r r e l a t i o n with C A 3 c a n d

Fig. 4. Micrographs illustrating neo-Timm staining and supragranular mossy fiber sprouting in three rats injected with kainic acid at age PN 7. The panels are oriented with the molecular layer on top, IbLowed by the stratum granulosum (SG), and the hilus. (A) This rat survived 145 days following injections. Note the strands of neo-Timm stained fibers in the stratum granulosum (arrowhead). (B) This animal survived 194 days and there was sprouting into the inner molecular layer (IML; arrowheads). This rat had a spontaneous recorded seizure during depth EEG recording beginning on the K.A-injected side. (C) Another animal, not monitored by EEG that survived 213 days. Again notice the neo-Timm's staining in the IML (arrowheads). All micrographs of equal magnification: bar equals 50 Ixm.

226

J.P. Leite et al. / Epilepsy Research 26 (1996) 219-231

Fig. 5. Example of Neo-Timm sections showing progressive mossy fiber staining with longer survival times. (A, B) Kainate-injected hippocampus of a PN 30 rat sacrificed 33 days after KA-injury. (C, D) Animal injected at the same PN age but with a longer survival time (190 days). Calibration bars: Low magnification 100 p,m, high magnification 50/,tm.

hilar neuron counts (r = 0.613 and r = 0.646, p = 0.0001, respectively). There were no correlations between neuron counts and GV differences in nonlesion hippocampi. Mossy fiber sprouting was observed in 2 of 7 animals injured at PN 7, and progressively increased with longer survival times in animals injured older than PN 15 (Figs. 4-6). For example, Fig. 4 shows fascia dentata neo-Timm's staining in three animals all lesioned at PN 7. The animal in the top panel (Fig. 4A) survived for 145 days after injection and there was essentially no sprouting except for single strands of puncta in the stratum granulosum (arrowhead). The rats in the lower two panels survived 194 and 213 days respectively and both show neo-Timm's staining in the supragranular molecular layer though not as much as other rats injected at later PN ages (see Figs. 2 and 5). The rat surviving 194 days (Fig. 4B) had a chronic seizure recorded during EEG telemetry (see below). Of note, in these PN 7 animals the sprouting was observed only on the

KA-injected side. As an example of progressive sprouting, Fig. 5 shows two animals both lesioned at PN 30 with similar amounts of hilar and CA3c neuron losses by cell counts (14 and 13%) but with 160

,m, 120I,k, 100-

o

PN 15-30 R = 0.544

a,,, 14D-

p = 0.0005

o o o

-&-

PN 7-14 R = 0.096

p = 0,6394

oo

80I Z

604020-

_=

0-

m

. . . .

I

50

'

'

'

I

'

100

'

'

[

. . . .

150

I

200

. . . .

I

'

'

250

S A C R I F I C E A G E (days)

Fig. 6. For KA-injected hippocampi, graph comparing mossy fiber sprouting (IML-OML GV differences) with sacrifice age. The data are divided into two groups, one with less hippocampal damage (PN 7-14) and the other with more hippocampal cell loss (PN 15-30). Note that the positive correlation between IML staining and survival times in the latter group (r = 0.544, p = 0.0005).

J.P. Leile e'f aL / E p i l e p s y R e s e a r c h 26 ( t 9 9 6 ) ~ ~o,_ ~ ~ ,~

120R ~ 0.828

100-

p -

(4) older age at sacrifice for PN ~5-30 animats, Of all these factors, multivariate analyses of covariance found that I[ML sprouting mo:;t significantly con'elated with PN category ( p = 0,000|) and sacrifice age ( p = 0.028), and not with CA3c ( p = 0.65) and hilar ( p = 0.48) neuron losses. [n other words, the PN age at i~ection and sacrifice age correlated best with increases in mossy fiber sprouting.

0.0005

80~.~ 6040-

Z

£

20-

<>o

~

o~.~7"~

o

ul 00

20

40

60

80

100

120

aMt--oONL " f i N N S G V DIFFERENCE

Fig. 7. Graph comparing the average imerictal spikes per hour in the kainate injected hippocampus ~o corresponding IML-OML GV differences. Spike frequencies correlated with supragranular mossy fiber sprouting ( r = 0.828, p = 0.0005).

different survival times and amounts of mossy fiber sprouting, The animal in the top row was sacrificed 33 days after injection (sacrifice age 63 days) and showed moderate aberrant neo-Timm's staining while the rat in the bottom row was perfused 190 days post injection (sacrifice age 220 days) and shows more sprouting compared to the top animal. This was similar to that found in adult lesioned animals [25]. Fig. 6 quantitatively illustrates our visual assessments by comparing me IML-OML GV differences in the K_A-injected hippocampus to the sacrifice age. The reader should note that for both Fig~. 6 and 7, GV differences above 10.5 indicate animals with significant mossy fiber spouting compared to naive animals (control mean plus 2 SD). As a group the PN 7 - 1 4 animals had less hilar and CA3c neuron losses (see Fig. 1), the amount of mossy fiber sprouting was less, and sprouting did not increase with age (dashed line; p = 0.64). For PN 15-30 animals the hilar and CA3c damage was greater, sprouting occurred in most animals, and mossy fiber sprouting positively correlated with an older age at sacrifice (solid line; r = 0.544, p = 0.0005). In other words, in the animals injured at PN 15 or older there was mossy fiber sprouting that progressively increased with longer survivals. Thus far our analysis of KA-injected hippocampi found that progressively greater IML-OML gray value differences, indicating axon sprouting, correlated with: (1) An older PN age at injection; (2) greater CA3c cell loss; (3) greater hilar cell loss; and

3°5. /nterictal spikes In summary, interictal spikes on the KA-side correlated with mossy fiber sprouting and did not correlate with neuron losses. For the 13 randomly chosen rats, Fig. 7 compares the average interictal spike freque:~eies from the KA-injected hippocampal electrodes with the IML-OML GV differences. The density of interictai spikes correlated with supragranular mossy fiber sprouting (r = 0.828, p = 0.0005). In looking at this graph, one might wonder if the correlation was being driven by the single data point in the upper right corner. Repeating the analysis without that data point again found a positive correlation (r = 0.652; p = 0.021). Interictal spikes were often observed from the saline-injected hippocampi especially if spikes were frequent on the KA-side. Sometimes synchronous with the KA-side but just as often asynchronously, and the same was true of the two neo-cortieal contacts. There were no correlations on the saline-side between spike-densities and GV d~t~ ferences with either the saline- or KA-injected hippocampi. Furthermore, interictal spike frequencies did not correlate with hilar or CA3c neuron losses on eifl~er side. Computerized EEG monitoring averaged 44.2 h per rat over the 21 day monitoring period. Spontaneous seizures were recorded eleetrographically from two animals (PN 7 and 30) and in another two rats (PN 17 and 18) behavioral seizures were noted while not being recorded. Behaviorally, the chronic seizures were characterized by an initial immobile blank stare, mastication, facial movements, forelimb clonic jerks, rearing, and finally a generalized clonic convulsion. The intraeerebral EEG recordings showed seizure onsets in the KA-injected hippocampal electrodes with rapid involvement of the opposite hippocampal and frontal electrodes.

228

J.P. Leite et ai. / Epilepsy Reseatz:h 26 (1996) 219-231

4. Discussion

This study found several new pathologic and electrophysiologic findings in rats studied as adults given intrahippocampal KA during early postnatal development. All of the KA-injected postnatal animals displayed the age-specific behavioral characteristics of generalized seizures indicating that both hippocampi experienced acute sustained seizure activity. However, as adults only the KA-injected side showed reductions in hilar and CA3c neuron counts and increases in mossy fiber sprouting compared to saline-injected hippocampi and naive controls. Furthermore, compared to controls, KA-injected hippocampi showed: (1) Reduced hilar and CA3c neuron counts beginning with injections at PN 7-10 that increased with older PN injury ages (Figs. 1 and 2); (2) mossy fiber sprouting in individual animals associated with injections at PN 7 that increased with an older PN injury age (Figs. 2-4); (3) mossy fiber sprouting that progressively increased with older sacrifice ages in animals lesioned between PN 15 to 30 (Figs. 5 and 6); and (4) in a random subset of 13 animals a positive correlation between in vivo interictal spike frequencies and mossy fiber sprouting (Fig. 7). Given that there were differences in hippocampal pathology between the KA- and saline-injected hippocampi and that both sides experienced postnatal seizures, our results indicate that the developing rat hippocampus was resistant to seizure-associated excitotoxicity relative to direct KA neurotoxicity. However, this study also shows that comparable doses of KA can induce neuron losses and mossy fiber sprouting beginning with KA-injuries as early as PN 7. These findings support the hypothesis that reactive generalized seizures during early development are not very excitotoxic and generally will not dramatically injure the hippocampus. However, the immature hippocampus can be excitotoxically damaged and once injured certain abnormal responses, such as mossy fiber sprouting, can progressively increase with longer survivals possibly contributing to eventual granule cell electrep'..i~,siologic hyperexcitability and chronic seizures from reorganized axon circuits. The reader should be aware of our asaumptions in methodological design when interpreting our results. For example, the hippocampal pathologic data were

obtained many weeks after the initial injury. We presume that most of the hippocampal neuron damage occurred with the KA-injection as has been found using similar methods in adults rats, and that mossy fiber reactive synaptogenesis was initiated as a consequence of hilar and CA3c neuron losses [25,26]. However, our experimental design did not study animals between injection and post-injury day 28, and it is possible that other variables, such as additional sub-acute seizures, could have contribvted to the neuron loss and sprouting observed in adult animals. Furthermore, we found a correlation between mossy fiber sprouting and interictal spike frequencies in KA-damaged hippocampi; but we can only infer that sprouting 'caused' the interictal spikes and chronic seizures. Lastly, we presume that the progressive increases in neuron losses and sprouting with older PN injuries were related to the postnatal maturation of kainate receptors [36]. However, more damage may be possible at earlier PN ages with other neurotoxins or forms of excitotoxic injury. Despite these assumptions, our study found several new findings related to the pathophysiology of hippocampal damage in developing rats.

4.1. Kainate neurotoxicity during hippocampal det,elopmenl Previous studies of neonatal rats have shown that large doses of intrahippocampal kainate can result in neuron loss, but have not shown progressive mossy fiber sprouting a~:d correlations with in situ electrophysiology. For example, Cook and Crutcher [ 11,12] previously showed that intrahippocampal KA doses between 2.5 to 5.0 /xg could visibly damage hippocampal pyramidal neurons as early as PN 5. Similarly, we found quantitative reductions of CA3c neuron counts with KA injections beginning on PN 7-10. Furthermore, we found progressive fascia dentata mossy fiber sprouting, but the reasons for the increases were unclear. One possible explanation may relate to postnatal granule cell neurogenesis. For example, Altman and Bayer [3] previously showed that only 15% of rat granule cells are present at birth and the remainder are 'born' mostly over the next two weeks but can continue for over 1 year. Our data showing progressively greater granule cell densities

.LP. Le#e e¢ aL,/ Lpdepsv Res,arch 26 (f996) 2 1 9 - 2 3 1 " .

with older sacrifice ages which would be consislent ~vith the notion that granule cells increased in number following KA injections. Likewise, immunocy~.ochemical studies with the embryonic neural cel~ adhesion molecule (NCAM-H), a molecule belie'~ed to participate in the migration and axogenesis of new neural connections, shows that NCAM-H is highly expressed in the stratum granulosum and infragranular region especially during the first three postnatal weeks probably indicating granule cell migration [21,40]. The development of mature mossy fiber staining patterns and [3H] KA binding sites also occurs during the same postnatal period supporting the idea that most granule cell maturation takes place postnatally [4,9,51] Taken together, these data suggest that one reason we found progressive mossy fiber sprouting maybe that as new granule cells form axon connections there are not enough dendritic sites available because of the previous KA injury [38] and as a result axon collaterals sprout into the molecular layer. Alternately, the initial KA injury may alter the biologic signals that guide axons toward their normal targets resulting in aberrant axon circuits. Whatever the mechanism, it would appear that following hippocampal neuron damage mossy fiber sprouting progressively increases with time even in animals injured during postnatal development.

4.2. Seizure-induced hippocampal damage Our data showed that the immature hippocampus was relatively resistant to seizure-induced damage as previously demonstrated in studies using intraperitoneal KA injections [36,43,44]. These studies showed with lower doses of KA that visible neuron losses were observed only after the end of the third postnatal week and usually only with very severe status epilepticus. In our study, we found no quantitatively determined cell losses in saline-injected hippocampi that experienced acute seizures with injections from PN 7 to 30. One possible reason for this difference might be because our rats were not in status as long as those given IP KA. For example, a recent study has demonstrated that the amount of neuronal loss may differ depending on the anesthesia used during KA injection which augments the status duration [231.

~

229

4.3. Massy fiber spromi~rg a~d kippocampa[ exci~,abili(v tn huma.~s, focal tempora~ and hippocampal interictal EEG spike discharges are hallmarks of complex-partial seizures [t7]. In our in vivo preparation interictal EEG spikes on the KA-side correlated with increased supragranular mossy fiber sprouting. In addition, EEG and behavioral spontaneous seizures were only observed in animals that had mossy fiber sprouting. These findings lend support to the hypothesis that mossy fiber sprouting can contribute to hippocampal epileptogenesis from increased granule cell hyperexcitability (for review, see [32]). For example, Tauck and Nadler [46] showed in hippocampal slices from chronic KA-treated rats that dentate granule cells generate monosynaptic multiple population spikes following single antidromic mossy fiber stimulation. This finding would be consistent with functional recurrent excitatory collaterals that could synchronize granule cell discharges. On the other hand, Sioviter [42] has shown in a similar animal preparation that the loss of granule cell recurrent inhibition from perforant path stimulation precedes mossy fiber sprouting and that recurrent inhibition is restored after sprouting. He has proposed that the initial degeneration of hilar neurons may denervate inhibitory neurons., rendering them dormant. According to this hypothesis, the granule cell excitability described by Tauck and Nadler would be associated with the dentate hilar cell loss rather than the mossy fiber sprouting itself. Our findings in the in vivo preparation contrasts with this suggestion since the increase in hippocmnpal interictal spikes frequencies did not correlate with either hilar or CA3c cell losses but with sprouting. It seems plausible that reorganized recurrent inhibitory and excitatory circuits are functionally operating in kainate-treated rats, and that the excitatory circuits may be normally suppressed, but emerge when synaptic inhibition is blocked [ 13,32].

Acknowledgements The authors wish to thank the UCLA Student Research Program undergraduate students for their

230

J.P. Leite et al./ Epilepsy Research 26 (1996) 219-231

assistance in EEG and behavioral monitoring and image analysis measurements. Thanks are extended to the UCLA Clinical Neurophysiology Laboratory for use of digital EEG equipment. This work was supported by NIH grants NS 31655 and NS 02808 (TLB), NIH-FIC, CNPq and FAPESP (Brazil) to JPL, and a CIDA (I(08 NS 1603) to GWM. Mafia Melendez kindly provided assistance in manuscript preparation. References [1] Abercrombie, M., Estimation of nuclear population from microtome sections, Anat. Rec., 94 (1946) 239-247. [2] Albala, B.J., Moshr, S.L. and Okada, R., Kainic acid-induced seizures: A developmental study, Dev. Brain Res., 13 (1984) 139-148. [3] Altman, J. and Bayer, S., Postnatal development of the hippocampal dentate gyros under normal and experimental conditions. In: R.L. Isaacson and K.H. Pribram, (Eds.), The Hippocampus, Structure and Development, Vol. 1, Plenum, New York, 1975, pp. 95-122. [4] Amaral, D.G. and Dent, J.A., Development of the mossy fibers of the dentate gyros: I. A light and electron microscopic study of the mossy fibers and their expansions, J. Comp. Neurol., 195 (1981) 51-86. [5] Annegers, J.F., Hauser, W.A., Elveback, L.R. and Kurkland, L.T., The risk of epilepsy following febrile convulsions, Neurology, 29 (1979) 297-303. [6] Anr,'~gers,J.F., Hauser, W.A., Shirts, S.B. and Kurland, L.T., Factors prognostic of unprovoked seizures after febrile convulsions, New Engl..L Med., 316 (1987) 493-498. [7] Babb, T.L. and Brown, W.J., Pathological findings in epilepsy, In: J. Engel, Jr. (Ed.), Surgical Treatment of the Epilepsies, Raven Press, New York, 1987, pp. 511-540. [8] Babb, T.L., Kupfer, W.R., Pretorius, J.K., Crandall, P.H. and Levesque, M.F., Synaptic reorganization by mossy fibers in human epileptic fascia dentata, Neuroscience, 42 (1991) 351-363. [9] Ben-Ari, Y., Tremblay, E., Berger, M. and Nitecka, L., Kainic acid seizure syndrome and binding sites in developing rats, Dev. Brain Res., 14 (1984) 284-288. [10] Cherubini, E., De Feo, M.R., Mecarelli, O. and Ricci, G.F., Behavioral and electrographic patterns induced by systemic administration of kainic acid in developing rats, Dev. Brahz Res., 9 (1983) 69-77. [11] Cook, T.M. and Crutcher, K.A., Extensive target cell loss during development results in mossy fibers in the regio superior (CA 1) of the rat hippocampal formation, Dev. Brain Res., 21 (1985) 19-30. [12] Cook, T.M. and Crutcher, K.A., Intrahippocampal injection of kainic acid produces significant pyramidal cell loss in neonatal rats, Neuroscience, 18 (1986) 79-92. [13] Cronin, J., Obenaus, A., Houser, C.R. and Dudek, F.E.,

Electrophysiology of dentate granule ce!'.s alter kainate-induced synoptic reorganization of the nn~ssy fibers, Brain Res., 573 (1992) 305-310. [14] de Lanerolle, N.C., Kim, J.H., Robbins, R.J. and Spencer, D.D., Hippocampal intemeuron loss and plasticity in human temporal lobe epilepsy, Brain Res., 49 ~; (1989) 387-395. [15] Dunn, D.W., Status epilepticus in children: Etiology, clinical features, and outcome, J. Neurol. Child., 3 (1988) 167-173. [16] Falconer, M.A., The pathological substrate of temporal lobe epilepsy, Guy's Hosp. Rep., 119 (1970) 47-60. [17] Gastaut, H., Toga, M., Roger, J. and Gibsen, W.C., A correlation of clinical eleetroencephalographic and anatomical findings in nine autopsied cases of temporal lobe epilepsy, Epilepsia, 1 (1959) 56-85. [I•! Ggtmano J., Automatic recognition of epileptic seizures in the EEG, Electroencephalogr. Clin. Neuroph~,siol.. 54 (1982) 530-540. [19] Holmes, G.L. and Thompson, J.L., Effects of kainic acid on seizure susceptibility in the developing brain, Dec. Brain Res., 39 (1988) 51-59. [20] 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) 267-282. [211 Leite, J.P., Babb, T.L., Mathern, G.W., Pretorius, J.K. and Seki, T., Axonal growth and terminal specialization of mossy fibers in developing rat pups: An NCAM-H and Neo-Timm's study, Soc. Neurosci. Abstr., 20 (1994) 1682. [22] Lorente de N6, R., Studies on the structure of the cerebral cortex. II. Continuation of the study of the ammonic system, J. Psych. Neurol., 45 (1933) 113-177. [23] Magloczky, Z.S. and Freund, T.F., Selective neuronal death in the contralaterai hippocampas following unilateral kainate rejections into the CA3 subfield, Neuroscience, 56 (1993) 317-336. [24] Margerison, J.H. and Corsellis, J.A.N., Epilepsy and the temporal lobes, Brain, 89 (1966) 499-530. [25] Mathero, G.W., Kupfer, W.R., Pretorius, J.K., Babb, T.L. and Levesque, M.F., Onset and patterns of hippocampal sprouting in the rat kainate seizure model: Evidence for progressive ceil loss and neo-innervation in regio inferior and superior, Dendron, l (1992)69-84. [26] Mathem, G.W., Cifuentes, F., Leite, J.P., Pretorius, J.K. and Babb, T.L., Hippocampal EEG excitability and chronic spontaneous seizures are associated with aberrant synaptic reorganization in the rat intrahippocampal kainate model, Electroencephalogr. Clin. Neurophysiol., 87 (1993) 326-339. [27] Mathern, G.W., Babb, T.L, Vickrey, B.G., Melendez, M., Pretorius, J.K., Traumatic compared to non-traumatic clin;cal-pathologic associations in temporal lobe epilepsy, Epilepsy Res., 19 (1994) 129-139. [28] Mathem, G.W., Pretorius, J.K. and Babb, T.L., Quantified patterns of mossy fiber sprouting and neuron densities in hippocampal and lesional seizures, J. Neurosurg., 82 (1995) 211-219. [29] Mathero, G,W., Pretorius, J.K. and Babb, T.L., Influence of

J.P. Le#e cg aL / Epiiepsy Research 26 (I995"* 219-23_1

[30]

[31]

[32] [33]

[34]

[35]

[36]

[37]

[38]

[39]

[40]

the type of initial precipilafing injury and al wha~ age it occurs o~ course and outcome in patients wi~h temporal lobe seizures, J. Neurosurg., 82 11995) 220-227. Mathern, G.W., Babb, T.L., Vickrey, B.G., Me]c~dez, M. and Pretorius, J.K., The dimcal-pathogenic mechanisms of hippocampa~ neuron loss and s~rgical outct~mes h~ ~emp~raf lobe epilepsy, Brain, 1t8 (1995) 105-118. Maytak J., Shinnar, S., Mosh~, S.L, and AIvarez, L.A., The low morbidity and mortality of status epiiepticus in children, Pediatrics~ 83 (] 98q) 323-331. McNamm'a, J.O., Cellular and molecular basis of epilepsy, J. Neuro~ci, 14 (1994) 3413-3425. Meldrum, B.S., Vigouroux, R.A. and Brierley, J.B., Systemic factors and epileptic brain damage, Arch. Neurol., 29 (1973) 82-87. Mosh6, S.L., S0erber, E.F. and Valisek, L., Critical i~,sues of developmentai seizure disorders, Physiol. Res, 42 (19931 145-154. Nelson, K.B. and Etlenberg, J.H., Predictors of epilepsy in children who have experienced febrile seizures, ,¥e~' Engl. J. Med., 295 (1976) 1I)29- 1{133. Nitecka, L., Tremblay, E., Charton, G., Bouillot, J.P., Berger, M.L. and Ben-Ari, Y., Maturation of kainic acid seizure-brain damage syndrome in the rat. H. Histopathological sequelae, Neuroscience, 13 (1984) 1(173-1094. P~rez-Clauseil, J. and Dansher, G., Intravesicular localization of zinc in rat telencephalic boutons A histochemical study, Brain Res., 337 (1985) 91-98. Ribak, C.E. and Navetta, M.S., An immature mossy fiber innervation of hilar neurons may explain their resistance to kainate-induced cell death in 15 day old rats, Dec. Brai~ Res., 79 (1994) 47-62. Sugar, H.J. and Oxbury, J.M., Hippocampal ne~ron loss in : mporal lobe epikpsy: Correlation with early childhood convulsions, Atm. Neurol., 22 (1987) 334-340. Seki, T. and Arai, Y., Highly polysialylated neural cell adhesion molecule (NCAM-H) is expressed by newly generated granule cells in the dentate gyrus of the adult rat, J. Neurosci., 1313 (1993) 2351-2358.

231

[4!] Sherwood, N.M. and 3imiras. P.S., A S~ere~ta.~.-ic,4f/a.~ ~¢ ~l~e,Dece~q-d,,~ R~a B~z~i~, UC Pre,~s. Los Angeles, ~970. [42] Slovi~er, R.S, Pos~dble ftmctionai consequences of synaptic re~)rgmizak)a in the dentate gyrus of kainale-m.oated ruth, Neuro,~i. Let~., 1~7 f1992~ 91-96. [~?] 'qperber E.V.. Haas. K.Z., Stan~on. P.K. and Mo,,h£ ~; k Resistance of the. immm:~re brain to seizure-induced synaptic reorganization, DeL'. Brain Pes., 60 ( 1991 ) 88-93. [44] Sperber, E.F., S~anton, P.K., Haas. K., Ackermann, R.F., MoshC S.L., Developmental d?fli~rences in the neurobiology of epileptic brain damage. In: J. Engel, Jr., C. Wasterlain, E.A. Cavalheiro, U. Heinema~n and G Avanzini (Eds.), Molecular N~'urobiology of Emlepsy. Epilepsy Res.. Suppl. 9, E~sevier, Amsterdam. 1992, pp. 67-81. [45] Sutula, T., Cascino, G., Cavazos, J., Parada, I. and Ramirez, L.. Mossy fiber synaptic reorganization in the epileptic human temporal lobe, Ann. Neurok, 26 (t989) 321-330. [46] Tauck, D.L., Nadler, J.V., Evidence of functional mossy fiber sprouting in hippocampal formation of kainic acid treated rats, J. Neurosci., 5 (1985) 1016-1022. [471 Thompson, K.W., Holm, AM. and Waster]ain. C.G., Loss of hippocampal intemeurons in a model of neonatal status epilepticus, Soc. Nemvsci. Abstr., 20 (1994) 1668. [4g] Trcmblay, E., Nitecka, L., Berber, M.L. and Ben-Art, Y., Maturation of kainic acid seizure-brain damage syndrome in the rat. 1. Clinicai, electrograpbic and metabolic observations, Neuroscience, 13 ( i 9841 1051 - 10,72. [49] Wasterlain, C.G. and Sankm', R., Excitotoxicity and the developing brain. In: G. Avanzini, R. Fariello, U. Heinemann and R. Mutant (Eds.). Epih,ptogenic and Krcitotoxic Mechanisms, John Libbey & Company Ltd., New York, 1993, pp. 135-151. [50] Wasterlam. C.G. and Shira:;aka, Y., Seizures, brain damage and brain development, Brain Dec., 16 (1994) 279-295. [51] Zimmer, J. and Haug, F.M., Laminar differentiation o1 the hippocampus, fascia dentata and subiculum in devCoping rats, observed with the Timm sulphide silver method, J. Comp. Neurol., 179 (1978) 581-618.