Effects of Intraventricular Locus Coeruleus Transplants on Seizure Severity in Genetically Epilepsy-Prone Rats Following Depletion of Brain Norepinephrine

Audiogenic seizures (AGS) in genetically epilepsy-prone rats (GEPR) of the moderateseizure substrain (GEPR-3s) were investigated to determine whether norepinephrine (NE) depletion induced by 6-hydroxydopalnine (6-OHDA) microinfusion into the locus coeruleus (LC) could alter the efficacy of intraventricular NE tissue grafts in promoting reductions in seizure severity in AGS. GEPR-3s were stereotaxically infused with 6-OHDA (4μg/side/rat), or vehicle into the region of the LC. Following 6-OHDA treatment all animals were subjected to 3 AGS tests. GEPR-3s seizure severities were increased in 39.5% of the animals after microinfusion of 6-OHDA into the region of the LC. Following the third AGS test, each rat was stereotaxicaily implanted with 17 gestational day rat fetal tissue obtained from the dorsal pons and containing the primordia of the LC or with tissue obtained from the neocortex or were sham-grafted. Subsequent to grafting, rats were subjected to 3 additional AGS tests. 53% (10/19) of 6-OHDA treated GEPRs showed a significant reduction in seizure severity following transplantation of fetal LC tissue. In contrast, only 20% (1/5) of GEPRs infused with saline rather than 6-OHDA showed, a reduction of seizure severity following fetal LC transplantation. NE content in the cortex and pons/medulla was decreased by 78% and 46% respectively following 6-OHDA microinfusion into the LC. Prominent grafts with numerous TH positive neurons and neurites were present within the third ventricle of grafted animals, while cortex grafts contained no TH immunostained structures. These findings suggest that the efficacy of fetal LC tissue to promote reductions in seizure severity in GEPRs is increased following depletion of central NE by microinfusion of 6-OHDA.

NE tissue grafts in promoting reductions in seizure severity in AGS. GEPR-3s were stereotaxically infused with 6-OHDA (4pg/side/rat), or vehicle into the region of the LC. Following 6-OHDA treatment all animals were subjected to 3 AGS tests. GEPR-3s seizure severities were increased in 39.5% of the animals after microinfusion of 6-OHDA into the region of the LC. Following the third AGS test, each rat was stereotaxicaily implanted with 17 gestational day rat fetal tissue obtained from the dorsal pons and containing the primordia of the LC or with tissue obtained from the neocortex or were sham-grafted. Subsequent to grafting, rats were subjected to 3 additional AGS tests. 53% (10/19) of 6-OHDA treated GEPRs showed a significant reduction in seizure severity following transplantation of fetal LC tissue. In contrast, only 20% (1/5) of GEPRs infused with saline rather than 6-OHDA showed a reduction of seizure severity following fetal LC transplantation. NE content in the cortex and pons/medulla was decreased by 78% and 46% INTRODUCTION Therapeutic potential in transplantation neurobiology is being explored in several animal models of neurological disorders including models of Parkinson's disease/17,28,31,39/, Huntington's disease /13,18/, Alzheimer's disease /33,35/, and epilepsy /1,10,11,25/. For Parkinson's disease, progression of research on the potential beneficial effects of neurotransplantation has entered, perhaps prematurely /15,34/, the human clinical arena /24,26/. The hypothetical potential of neurotransplantation as a therapeutic approach to several neurological disorders is appreciable.
The genetically epilepsy prone rat (GEPR) represents a natural animal model of epilepsy/19/. Currently, there are two independently derived strains of GEPR: GEPR-3s which consistently display an audiogenic seizure (AGS) score of 3, and GEPR-9s which consistently display an AGS score VOLUME 5,NO. 1, 1994 66 R.W. CLOUGH ET AL. of 9 according to the seizure-rating scale developed by Jobe /20/. GEPR-3s have running-bouncing donic seizures and GEPR-9s exhibit severe tonic seizures with full hindlimb extension. Seizure predisposition in these animals is manifested by enhanced susceptibility to a variety of seizureinvoking stimuli, including sound, heat, convulsive drugs and electroshock/9,32/. The basis of seizure predisposition and the determinants of severity of seizures experienced by GEPRs are presently unknown, but are hypothesized to result from regionally-specific aberrant neurotransmitter profiles in the brains ofthese animals/14,21/. The noradrenergic neurotransmitter system appears to serve as a determinant of seizure severity in the GEPR brain /20,23/ and pharmacological evidence supports the hypothesis that there is an inverse relationship between AGS severity and the concentration of norepinephrine (NE) in the synaptic clet. Treatments that decrease the availability of NE at its receptors enhance both seizure susceptibility and severity in the GEPR. For example, drugs which deplete brain NE by interfering with the vesicular storage (i.e., reserpine) are demonstrated to cause an increase in seizure severity /20/. Inhibition of NE synthesis plus displacement of NE from its storage vesicle can further intensify AGS/20/. Similarly, inhibition of NE synthesis coupled with cold stress to produce depletion of NE also will increase AGS severity in GEPRs /20/. We and others have reported that depletion of central NE using the catecholaminergic neurotoxin, 6 There is evidence which suggests that the degree of integration between fetal grafts and host animals may in part depend on a synaptic void in the host animal/29/. For example, NE axons from fetal LC grafts in the anterior chamber of the eye hyperinnervate co-grafts of hippocampal tissue when the host superior cervical ganglion (SCG) is lesioned/29/. Typically, axons from the host SCG invade and innervate intraocular-gratted hippocampal tissue; however, when the SCG is lesioned, there is a resulting synaptic void in the grafted hippocampus. Without competition from host NE fibers from the SCG, co-grafts of fetal LC innervate the hippocampal co-grafts with NE axons to a much greater extent/29/. Additionally, it has been shown that neurite outgrowth in vitro from catecholaminergic explant cultures is much reduced in the presence of exogenously administered Ldopa, a precursor of NE /36/. These findings support the hypothesis that there is a type of feedback inhibition on neurite outgrowth and axon aborization from fetal grafts in the presence of synaptic competition or neurotransmitter feedback from the host. Therefore, inasmuch as LC gratis have been shown to reduce seizure severity in limited numbers of GEPR-9s and GEPR-3s /11/, NE depletion has resulted in increased propensity for seizures in rats/5,8/and fetal gratis have been shown to exhibit greater neurite growth in the absence of synaptic competition /29/, the present study was designed to determine whether the incidence of reduced seizure severity following neurotransplantation of fetal noradrenergic neurons could be increased in GEPR-3s after depletion of brain NE.

Animals
Adult male genetically epilepsy-prone rats (GEPRs) were obtained from the colony at the University of Illinois College of Medicine, Peoria (Peoria, IL). After arrival, each rat was pretested on three separate occasions for audiogenic-induced seizures (AGS) to confirm that they were GEPR-3s.
The seizure severity was scored using the seizurerating scale described by Jobe/20/, which assigns scores from 0 (no seizure) to 9 (generalized tonicseizure with complete hindlimb extension) with each score above 0 representing an increasingly severe seizure. GEPR-3s display seizures characterized by two running phases culminating in clonus of the forelimbs and hindlimbs. Testing for AGS involves placing each rat into a round (40 cm) Plexiglass arena and exposing them to a bell tone of 100 db intensity for 60 seconds. This procedure induces consistent seizures of intensity 3 in GEPR-3s and 9 in GEPR-9s. Only GEPR-3s were used in the present study. The latency to onset of running behavior and convulsion was timed and the seizure severity was scored and recorded for each rat.

Experimental protocol
After preliminary AGS-testing of each rat for confirmation of seizure display, rats were divided into the following two groups" 6-OHDA microinfused (n=38), and saline microinfused (n=7). Animals from each of these two groups were subsequently divided into the following subgroups" 6-OHDA-infused, LC-grafted (n=19); 6-OHDAinfused, cortex-gratted (n=10); 6-OHDA-infused, sham-grafted (n=9); saline-infused, LC-grafted (n=5); saline-infiased, sham-gratted (n=2). The timing of pretests, microinfusions, post-microinfusion tests, grafting and post-grafting tests are shown in schematic form in Figure 1. Briefly, animals were pretested upon delivery to insure that they were GEPR-3s. After pretesting, rats were microinfused with either saline or 6-OHDA as described below. Atter microinfusions, animals were again tested on three separate occasions to determine whether seizures were more severe than the pretest seizure scores. After the third test, animals received grafts from either the fetal LC, the fetal neocortex, or were sham-grafted. Atter grafting, animals were again AGS tested three times to determine whether gratting would alter the severity of seizures atter central noradrenergic depletion. At the end of the study, animals were processed for either general histology (n=20) with morphometric measurement of graft sizes, neurochemistry (n=14) or immunocytochemistry of tyrosine hydroxylase (n 11) as described below.

Grafting
Following decapitation of pregnant Sprague-Dawley rats, uteri containing pups (17 days gestation) were removed and placed into a Petri dish containing a balanced salt solution (Hank's, GIBCO Labs, Grand Island, NY) on ice. Under a dissecting microscope, each fetal brain was removed VOLUME 5, NO. 1,1994 Experimental Seq .uen 6 Experimental Week No.
Temporal sequence of experimental procedures for the present study. First, a series of 3 audiogenic seizure pretests were evaluated in order to confirm that the animals were GEPR-3s. The means of these three tests are represented as AGS test number 1. Following these tests, animals were infused with 6-OHDA directly into the LC. A series of three AGS post tests were evaluated at 2, 3 and 4 weeks after 6-OHDA. After the fourth test, animals were subjected to ICV gralng of fetal LC or CTX tissue or were sham grafted. Subsequent to grafting, rats were again AGS tested three times at 4, 8 and 12 weeks after grafting. and the anterior lip of the pontine flexure on each side or small pieces of occipital cortex were excised similar to the surgical procedure described by Bj6rklund /3/. Fetal tissues were aspirated into spinal needles for subsequent transplantation. Adult GEPR-3 hosts were anesthetized with chloral hydrate (400 mg/kg, i.p.) for neurosurgery. After reflection of the scalp and drilling a small burr hole through the calvaria, a spinal needle containing fetal LC or cortex tissue was stereotaxically directed into the third ventricle using coordinates: AP=7.0 from IAL; L=0 from midline; V=9.0 from dura; incisor bar set at -3.3 mm/30/. Tissue was pressure ejected very slowly (over a 5 minute interval). Sham grafts consisted of an identical infusion of the balanced salt solution into the third ventricle. After infusion, the spinal needle was leg in place for 5 minutes before removal. The burr hole was then covered with gelfoam, and the scalp sutured with 9 mm wound clips. After surgery, GEPRs were allowed to recover under supervision. Immunohistochemistry Subsequent to final AGS testing, the brains of 11 rats from either the 6-OttDA/LC grafted, the saline/LC graRed or the 6-OHDA/ctx grafted groups were subjected to immunohistochemistry to determine whether noradrenergic neurons were present in the grafted tissue. Rats were deeply anesthetized with sodium pentobarbital (50 mg/kg, i.p., Butler labs, Columbus, OH) and transcardially perfused with phosphate buffered saline (PBS, pH 7.4) followed immediately by perfusion of 200 ml, ice-cold paraformaldehyde (4%) in PBS. Brains were removed and post fixed in fixative for 1.5 h, and transferred to cold 25% sucrose for 48 h. Brains were sectioned at 40 tm thickness on a cryostat and free-floating sections were immunohistochemically processed for localization of tyrosine hydroxylase (TH) using a biotin-avidin method (ABC kit, Vector Labs, Torrence, CA). A polyclonal primary antiserum for TH was obtained from Eugene Tech Int. (Allendale, NJ) and used at a dilution of 1/15,000 (in 0.1M PBS, 0.1% Triton-X, 0.1% BSA, pH 7.4). After 48 h incubation at 4C and subsequent processing with the ABC kit, the reaction was visualized using 3,3'-diaminobenzidine tetrahydrochloride in the presence of hydrogen peroxide. The presence of TH-positive neuron somata and fibers in the host and gaffed tissues was recorded for each animal so processed.

NE determination
Norepinephrine content was measured in order to assess the degree of depletion of central NE following 6-OHDA or saline microinfusion. In 14 animals not used for immunohistochemistry, the content ofNE in the brains of 6-OHDA treated and control rats was determined by high performance liquid chromatography (HPLC) with electrochemical detection. Brain regions were dissected on an ice-cold plate, wrapped in aluminum foil and frozen in liquid nitrogen until assay. The pons/medulla was isolated as a representative brainstem sample and the cerebral cortex was isolated as a representative forebrain sample. Brain samples were homogenized in 4 ml 0.2N perchloric acid (PCA) containing 80 ng/ml 3,4-dihydroxybenzylamine (DHBA) and transferred to glass centrifuge tubes. After centrifugation, 2.0 ml of the supematant from each sample was shaken with 10 ml of 0.5 M TRIS buffer (pH 8.6) containing 50 mg of acid-washed alumina and 10 gl of sodium metabisulfate. After centrifugation and removal of the supernatant, the alumina was washed with 1.0 ml TRIS buffer (pH 7.0). Finally, the NEwas eluted from the alumina by shaking in the presence of 0.5 ml 0.2N PCA for 20 min. The eluate was then centrifuged at 13,000 g for 3 min and 50 gl was injected into a rheodyne injector connected to a bioanalytical HPLC system. The concentration of NE was determined by comparing peak heights of NE in samples to those ofNE in standards (20 to 80 ng/ml) run through the assay simultaneously. The HPLC system consisted of a 5 lam biophase C18 reverse-phase column (15 cm by 4.6 cm, Beckman) and a glassy carbon electrode with the applied potential set at 460 mV vs a Ag/AgC1 reference electrode. The mobile phase contained citric acid (14.6 mM), Na2HPO4 (4.9 mM), octane sulfonate (100.8 mM) and ethylenediaminetetraacetic acid (EDTA, 55.3 mM) in 3.0% methanol. changes in seizure severity preceding and following gratting treatments was analyzed by a contingency table analysis. Differences in NE content in either the cortex or the pons-medulla of randomly selected animals were analyzed by Student's t-test. The presence of TH positive perikarya and fibers within grained LC tissues of selected animals was noted.
Comparisons of gratt sizes between Ctx and LC tissue grafts were made by Student's t-test.

RESULTS
NE content after 6-OHDA infusions NE content in the brainstem and cerebral cortex is shown in Figure 2. There was a 78% reduction of NE in the forebrain (p<0.05) and a 46% reduction of NE in the brainstem (p<0.05) of animals receiving an LC microinfusion of 6-OHDA. Statistical analysis of changes in seizure severity was carried out using non-parametric tests. The incidence of changes in seizure severity following microinfusions of 6-OHDA or saline was analyzed by a coded chi-squared test. The incidence of (Mean + S.E.M. *=difference between controls and 6-OHDA group in each respective brain area, p<0.05.) VOLUME 5, NO. 1,1994 Although only about 40% of the animals treated with 6-OHDA experienced a subsequent increase in seizure severity (see below), all of the examined brains of 6-OHDA treated animals were consistently depleted ofNE (Fig. 2).

Histology of grafts
An example of an LC graft in an animal of the present study is shown in Figure 3. Immunohistochemical examination of grafted brains revealed that LC grafts within the third ventricle contained numerous TH positive perikarya and dense plexuses of TH positive fibers while Ctx grafts were devoid of TH positive profiles. In some of the LC grafted animals, the cytoarchitexture of TH positive areas within the grafts resembled that of the LC in situ. LC or Ctx gratis were found in the third ventricle of all 11 rats analyzed with immunohistochemistry (ICC) and all the LC gratis analyzed by ICC were found to contain numerous TH-positive somata and neurites. A morphometric analysis was conducted on 5 cortical-tissue grafts and 8 LC-tissue grafts from gratted animals not processed for ICC or neurochemistry. Gratis derived from the occipital cortex were significantly larger than those from the region ofLC (Table 1).

Seizure scores
There were four typical behavioral responses across the treatment groups: 1) rats which displayed an increase in seizure severity following 6-OHDA microinfusion, and a subsequent reduction in seizure severity following grafting (Figure 4a). This response was called exacerbated/corrected; 2) rats which demonstrated an increase in seizure severity following 6-OHDA microinfusion with no reduction in seizure severity following grafting (Figure 4b).
This response was called exacerbated/noncorrected; 3) rats which showed no increase in seizure severity after 6-OHDA treatment, but who did show a reduction in seizure severity atter grafting ( Figure  4c). This response was called nonexacerbated/ corrected; and 4) rats which demonstrated no increase in seizure severity following 6-OHDA microinfusion and no reduction in seizure severity atter grafting (Figure 4d). This response was called nonexacerbated/noncorrected. The number of animals in each treatment group displaying each of these four responses are indicated in Table 2. With reference to change in seizure severity following microinfusions of 6-OHDA into the LC, 15 of the 38 animals (39.47%) displayed a significant increase in seizure severity following 6-OHDA treatment (p<0.0418). In the control group, 0 of the 7 had altered seizure severity following saline microinfusion. Figure 5 shows the percentage of animals in which grafting resulted in a reduction of seizure severity. As shown in Figure 5, only 1 of the 10 GEPR-3s microinfiased with 6-OHDA and transplanted with cortex tissue showed a decrease in seizure severity subsequent to grafting. Similarly, 1 out of 9 6-OHDA microinfused and shamtransplanted GEPR-3s showed a slight decrease in seizure severity atter surgery. The difference in the incidence of change of seizure severity between the cortex-and sham-transplanted groups was not significant (p=0.5957). In contrast, 53% (10/19) of the 6-OHDA treated animals transplanted with fetal LC tissue showed a significant decrease in seizure severity after transplantation. The incidence of change in seizure severity of the LC transplanted group compared to the cortex-and sham-treated groups was significant (p<0.0167). Of the 5 salineinfused GEPR-3s transplanted with LC tissue, no rats showed increased severity following saline microinfusion; however, 1 of 5 LC-transplanted rats showed a marked reduction in seizure severity following grafting of fetal LC. Finally, although the incidence of reduction in seizure severity in LC transplanted animals receiving 6-OHDA compared to those receiving saline was not significantly different (p>0.05), 53% (10/19) of the 6-OHDA treated animals showed post-LC-gratting reductions in seizure severity while only 20% (1/5) of saline treated animals showed post-LC-gratting reductions in seizure severity. Of the two rats microinfused with saline and sham transplanted, there were no effects of either treatment on seizure severity. Figure 6 shows the actual degree of the seizure reduction following LC gratting in 6-OHDA treated rats. Panel A shows the data for all animals grafted with fetal LC (n=19), while panel B shows data only from those animals that had a reduction in seizure severity following grafting (i.e., corrected animals only, n=10). In both groups, the pretest score

Seizure Test Number
Four graphs (A-D) each representing an individual showing one of the four typical behavioral responses of rats following 6-OHDA administration and fetal brain tissue grafting. The data in Panel A (representing animal GEPR-3, s2-#3) showed an increase in seizure severity from a score of 3 to a score of 9 by the third test after 6-OHDA. This represents an increase in severity from a moderate clonic seizure to a severe tonic-clonic seizure with full hindlimb extension. Subsequent to grafting fetal LC, the severity of seizures was again reduced from a score of 9 to the previous score of 3. Panel B shows the data from an animal (GEPR-3, sl-#2) who did not show an increase in seizure severity after 6-OHDA but who did show a reduction in seizure severity from a pregraffing score of 3 to a score of 0, representing a complete absence of seizure in response to the audiogenic stimulus. Panel C shows the data from an animal (GEPR-3, s4-#16) who showed an increase in seizure severity following 6-OHDA treatment, but who did not show any reduction in seizure severity up to 12 weeks following grafting of fetal cerebral cortex. Panel D shows the data from an animal (GEPR-3, s3-#24) who did not show an increase in seizure severity following 6-OHDA, nor was there a reduction in seizure severity following grafting. VOLUME 5, NO. 1,1994

Treatment groups
The percentage of animals who showed a reduction in seizure severity subsequent to grafting or sham treatment. The incidence of reduction in seizure severity was highest in the 6-OHDA, LC transplanted group (p<0.0167). that for the entire group (n=l 9) due to the fact that some of these animals were not exacerbated by 6-OHDA yet all these animals' seizure severity scores were reduced following grafting (i.e., some to below pretest values). Following grafting of LC, seizure scores were significantly lower than both the pretest and post 6-OHDA treatment scores (p<0.05). Latencies to the onset of running behavior and convulsions were not altered subsequent to brain infusions or transplantation in rats of any of the groups. Seizure severity scores (mean of three separate tests + S.E.M.) in all LC grafted animals (n=19, Panel A) and those LC grafted animals which showed a reduction in seizure severity following grafting (n=10, Panel B). In all animals, the pretest score was 3. Panel A: Following 6-OHDA, the severity of seizure in all animals increased significantly from a mean of 3 to 4.7 in the entire group (note, however, that only 39% of these animals were exacerbated following 6-OHDA). The effects of LC grafts on seizure severity across the entire group were not significant in that the scores after grafting were not significantly different from those after 6-OHDA; however, these scores were also not different from the pretest values (*=p<0.05 from pretest scores only, ANOVA [F2,18 5.06]). Panel B. In those 10 animals which showed reduction in seizure severity, 6-OHDA significantly increased the severity of seizure from 3 to 4.2 (note, however, that not all of these animals were exacerbated following 6-OHDA). After LC grafting, those 10 animals that were corrected had significantly lower seizure severity scores than animals at pretest and following 6-OHDA (*p<0.05 from pretest and post grafting; **p<0.05 from pretest and post 6-OHDA: ANOVA lF2, 9 10.37]).

DISCUSSION
It is well documented that seizure severity in the GEPR is inversely correlated with brain NE level /14,20,21,23/. Previous studies in this laboratory/8/ and others/5/have demonstrated that depletion of brain NE in normal rats and GEPR-3s using the catecholamine neurotoxin 6-OHDA results in an increase in seizure severity in a significant number of these animals. In the present study, analysis of NE content in the cortex, an area of the brain which receives the majority of its noradrenergic terminals from the LC, revealed that 6-OHDA microinfusion into the LC caused a 78% reduction in cortical NE. The 46% depletion of NE content in the brainstem as a result of microinfusion of 6-OHDA into the LC is also significant. It should be recognized that the remaining catecholamine nuclei in the brainstem, including those in the diffuse lateral tegmental system, remain intact. The effectiveness of the LC infusions is clearly demonstrated by the marked depletion of NE (78%) in the terminal regions of these neurons (i.e., the cerebral cortex). Although the extent of NE depletion in terminals of LC neurons innervating the brainstem following LC lesion could not be dissected out, it is presumed that it would be similar to the depletion of NE in the cortex which is also innervated by the LC. Thus, in light of the remaining catecholamine cell groups in the brainstem, the 46% depletion following LC microinfusion of 6-OHDA appears to represent a rather large depletion of NE in those brainstem terminal fields which originate in LC neurons. The present data demonstrate that depletion of NE in GEPR-3s following the microinfusion of 6-OHDA into the LC is associated with an increase in the severity of seizures in approximately 40% of GEPR-3s. However, other studies in our laboratory have shown a greater incidence of seizure exacerbation (i.e., 50-70% of animals) following LC infusions of 6-OHDA/8/. The reason for the lower incidence of seizure exacerbation in the present study compared to previous work/8/is unknown; however, the degree of NE depletion between the VOLUME 5, NO. 1,1994 studies was also different. In the present study, the intra-LC microinfusion of 6-OHDA reduced the content of NE in the cortex by 78%, while in the previous study /8/, cortical NE depletion was approximately 90%. The differences between the two studies in degree of central NE depletion after intra-LC microinfusion of 6-OHDA may account for the differences in the incidence of seizure exacerbation. Additionally, because of the small size of the LC as a target for microinfusion, differences between studies in central NE depletion may have resulted from subtle differences in location of the cannula tip within the surround of the LC. Interestingly, 6-OHDA-induced LC lesions have also been shown to significantly increase the severity of audiogenic-like seizures in normal rats rendered susceptible to sound-induced seizures by microinjection of N-methyl-D-aspartate (NMDA) into the inferior colliculus/7/. In that study/7/, as in the present one, some rats displayed a marked depletion of NE but failed to exhibit an increase in seizure severity following 6-OHDA treatment. The reason for this remains unknown, although it is conceivable that NE depletion in a crucial region not assayed per se is responsible for regulating seizure predisposition in the GEPRs and NMDA infused rats. Perhaps a better correlation between NE depletion and seizure exacerbation would be obtained should this hypothesized crucial region be examined.
A number of studies have reported beneficial effects of host NE or NE-tissue transplants on seizure activity in other models of epilepsy or epileptiform activity. Fetal LC cell suspensions transplanted into the hippocampal formation of adult rats suppress the onset and the progression of kindling-induced epilepsy in rats made hypersensitive to kindling by noradrenergic depletion /1,25/. Hippocampal tissue gratis in the anterior chamber of the eye are protected from epileptiform activity induced by penicillin, when adrenergic afferents to the gratis are activated /16/. Tonic inhibitory neurotransmission, as assessed by electrophysiology, can be reinstated in the hippocampal formation following placement of fetal LC gras into the hippocampus/4/.
Very few studies have examined gratting in GEPRs/11,3 7/. Stevens/3 7/reported that grats of fetal cerebellum containing gamma-aminobutyric acid (GABA) neurons transplanted to the surface of the inferior colliculus, or gratis of adrenal medulla transplanted to the lateral ventricle are, in general, ineffective in altering seizure expression in GEPRs. However, an identifiable cerebellar graft was found in only one of these animals upon autopsy and this graft was encapsulated and not integrated with the host brain. One of six rats with an adrenal medulla gratt in the lateral ventricle showed a decrease in seizure severity. Our laboratory has previously reported that intraventricular or intrahippocampal transplantation of fetal pontine tissue containing noradrenergic neurons can effectively reduce severity of seizures in a limited number of GEPR-3s and GEPR-9s /11/. Grafting of fetal neurons has also been demonstrated to suppress seizure development in kindling-induced /1,2,25/ and facilitate pentylenetetrazol-induced seizures in rats with fimbria-fornix lesions/10/. Thus, grafted fetal noradrenergic or cholinergic neurons can effectively alter seizure expression in at least three models of epilepsy. In the present study, immunohistochemical detection of tyrosine hydroxylase (TH) in tissue sections of the grafted brains demonstrated numerous TH-positive neuron somata as well as fibers within the grafted tissues and that TH positive fibers invaded the parenchyma of the host brain through the grat/host interface. Moreover, in some cases, the densely packed clusters of neurons within the grafted LC were similar to the cytoarchitecture of the in situ locus coeruleus of control rats. The present study demonstrates that, while Ctx gratis grow to significantly larger size than LC gratis in the third ventricle, grats containing the fetal LC are significantly more effective than grafts of neocortical tissue in reducing seizure severity in GEPRs depleted of NE. Additionally, although not statistically significant (likely due to the small number of animals), LC gras appear to be more effective in reducing seizure severity in NE-depleted animals than in saline-treated LC-grafted animals. Approximately 53% of animals receiving 6-OHDA microinfusions and LC grafts showed significant reductions in seizure severity. In animals that displayed increases in seizure severity following 6-OHDA, LC grats subsequently reduced the severity to at least the pre-infusion level. In animals which did not display an increase in seizure severity following 6-OHDA, LC grafts effectively reduced seizure severity to below the preinfusion level and in 1 case completely abolished the display of AGS. In contrast, of those rats receiving microinfusions of saline into the LC, only 20% showed a decrease in seizure severity following transplantation of fetal LC tissue. These data suggest that the effectiveness of noradrenergic neural gratis in promoting reductions in seizure severity in GEPRs may be increased when central NE is depleted by 6-OHDA. It is a tenable hypothesis that depletion of central NE with 6-OHDA creates a synaptic void in the GEPR brain which promotes a more extensive integration between the fetal LC grafts and the host brains, increasing the potential to reduce seizure severity. This hypothesis is not without precedence. For example, hippocampal gratis in the anterior chamber of the eye are more readily innervated by LC co-grafts when competition from the host SCG is removed by lesioning this structure in the host. Additionally, it has been shown that neurite outgrowth in vitro from catecholaminergic explant cultures is much reduced in the presence of exogenously administered L-dopa, a precursor of NE/36/. These findings support the hypothesis that there is a type of feedback inhibition on neurite outgrowth and axon aborization from fetal grafts in the presence of synaptic competition or neurotransmitter feedback from the host. However, the hypothesis that development of NE fiber interconnections between the host and the grafted tissues in animals with NE depletion has been examined with LC grafts in the hippocampus/27/. In this study, at two weeks following depletion of NE in the host, there was no apparent difference in graft-derived axon densities or synaptic profiles between control animals and animals previously depleted ofNE/27/ The observation that fetal LC grafts effectively reduce seizure severity in a limited percentage of animals raises questions about the mechanism of this effect. It is presently unknown whether an obligatory site for regulation of seizure severity is innervated by NE neurons from the gaff, or whether the release of NE into the ventricular fluid provides an anticonvulsant effect in the GEPR through a more general mechanism. A previous report found that gaffs of adrenal medulla tissue had no effect when transplanted onto the surface of the inferior colliculus, or into the lateral ventricle of GEPRs /37/. However, microinjection of noradrenergic agonists into the inferior colliculus also fails to reduce seizure severity in GEPRs /6/, suggesting that this is not the site at which NE produces an anticonvulsant effect in the GEPR. In other studies, we have shown that gratis of fetal LC placed into the parenchyma of the hippocampus can reduce seizure severity in limited numbers of GEPR-3s and that LC grafts into the third ventricle reduce seizure severity in both GEPR-9s and GEPR-3s/11/. Moreover, this effect may be more appreciable when the host brain is depleted of NE (present study). The reductions in seizure severities following gratting of fetal LC into the third ventricle support the findings of a seizure regulating site near the ventricular surface. In a report by Stull and colleagues /38/, near-ventricular terminal fields of host NE axons have been implicated in the regulation of electrically induced tonic extensor seizures in non-epileptic rats, and several nearventricular areas of the brainstem have been implicated in regulation of seizures in the GEPR /19/. Thus, although the mechanism of action of NE gratis in the modification of seizure severity requires further study, the apparent findings of facilitated efficacy of fetal LC grafts to reduce seizure severity following depletion of host NE provide additional support for the anticonvulsant effects of NE in the epileptic brains as well as the potential of fetal NE tissue gratis to modify the severity of seizures. Furthermore, these studies provide the impetus to continue to explore the feasibility of neuronal replacement with grafting in a genetic model of epilepsy.