Abstract
Background: Although seizures are very common in neonates and are often the harbinger of poor neurologic outcome, there is controversy regarding the degree of brain damage induced by seizures during early development. Here, we evaluated the effect of neonatal seizures on subsequent brain injury induced by status epilepticus.
Methods: Twenty-five seizures were induced by the inhalant flurothyl in neonatal rats during the first 5 days of life. Flurothyl reliably produced generalized seizures with concomitant electroencephalographic changes and a low mortality rate. During adolescence or early adulthood, animals were subjected to status epilepticus using either kainic acid or perforant path stimulation.
Results: Although flurothyl-induced neonatal seizures did not cause cell death, animals that had neonatal seizures had significantly more severe brain injury after both kainic acid and perforant path stimulation than did animals without a history of neonatal seizures.
Conclusions: Neonatal seizures increase the susceptibility of the developing brain to subsequent seizure-induced injury.
Seizures are very common in neonates and are often the precursor of poor neurologic outcome.1,2 Understanding of the neuropathologic processes responsible for neurologic sequelae in these infants is not clear. Although etiologic factors responsible for the seizures likely play a major role in the consequences of neonatal seizures, there are concerns that seizures themselves may have adverse effects on the developing brain. Seizures may perturb a range of phenomena that are activity dependent, including cell division, migration, sequential expression of receptors, formation, and probably stabilization of synapses.3
Despite theoretic concerns about the deleterious effects of seizures on brain development, a number of studies have demonstrated that the immature brain is less vulnerable to seizure-induced injury than the mature brain. In the adult animal, status epilepticus causes neuronal loss in hippocampal fields CA1, CA3, dentate granule cell layer, and the dentate hilus,4-6 leads to aberrant growth (sprouting) of granule cell axons (the so-called mossy fibers) in the supragranular zone of the fascia dentata7,8 and stratum pyramidale region of CA3,8 and results in long-term deficits in learning, memory, and behavior.9 However, a single prolonged seizure in an immature animal results in less cell loss,9-16 sprouting,17,18 and fewer deficits in learning, memory, and behavior than similar seizures in adults.9,19 The resistance to damage induced by status epilepticus appears to decrease at approximately 2 weeks of age. Sankar et al.20 have demonstrated clear cell damage in 2-week-old rats undergoing lithium–pilocarpine status epilepticus.
Although immature rats during the first 2 weeks of life may have less cell loss after status epilepticus than their adult counterparts, recent studies from our laboratory have demonstrated that repetitive seizures during the first 5 days of life in a rat alter subsequent brain development. We have found that neonatal seizures result in subsequent sprouting of mossy fibers to both the CA3 and supragranular regions without any accompanying cell loss.21,22 Furthermore, rats with neonatal seizures had reduced seizure thresholds and impairment in learning, memory, and activity level compared to control rats when tested as adults.21
We raised the question as to whether these seizure-induced changes in connectivity during early brain development alter subsequent risk for damage during seizures. The goal of this study was to determine whether neonatal seizures affect brain damage induced by status epilepticus at a later age. We subjected adolescent rats with prior seizures during the neonatal period to status epilepticus with either a chemical convulsant, kainic acid (KA), or electrical stimulation, perforant path stimulation. We chose to use two different models of status epilepticus because we wished to create a range of seizure-induced damage. KA typically results in major cell loss12 whereas perforant path stimulation causes less-severe cell loss.6 We found that flurothyl reliably produces behavioral and electrographic seizures in the neonatal period. Although these neonatal seizures did not lead to cell loss, the rats had significantly more cell loss after subsequent status epilepticus than did rats without a history of neonatal seizures.
Methods.
Neonatal seizure paradigm.
Male Sprague–Dawley rats were used in all experiments and were treated in accordance with the guidelines set by the National Institutes of Health and Children’s Hospital for the humane treatment of animals. Animals had access to food and water ad libitum and were housed with their litter until weaning at postnatal (P) day 21, when they were group housed in plastic cages under diurnal lighting conditions with lights on from 8:00 am to 8:00 pm.
The volatile agent flurothyl (bis-2,2,2-triflurothyl ether), a potent and rapidly acting central nervous system stimulant that produces seizures within minutes of exposure, was used to induce seizures.23 Rats were placed in a small cylinder plastic container (radius = 12 cm; height = 23 cm). Liquid flurothyl was delivered through a plastic syringe and dripped slowly onto filter paper in the center of the container where it evaporated. Experimental rats were exposed to flurothyl until tonic extension of both the forelimbs and hindlimbs was observed. Control rats were placed into the container for an equal amount of time without exposure to flurothyl. Each rat was subjected to five seizures per day with a minimum of 2 hours between seizures.
Effects of flurothyl seizures on cell number in the hippocampus.
To determine the effects of recurrent flurothyl seizures on cell number, 15 rats subjected to 25 flurothyl seizures from the day of birth (P0) through postnatal day 4 (P4) were compared with 9 control rats. At postnatal days 10 (P10), 20 (P20), and 60 (P60), five flurothyl-treated rats and three control rats were killed and cell counts were performed.
EEG changes with flurothyl seizures in neonatal period.
Another group of rats (n = 35) was used to characterize EEG changes during the flurothyl seizures in the early neonatal period. The pups were anesthetized by being placed on ice until they became immobile and failed to respond to pinching of the tail. Using the bregma and sagittal suture as landmarks, a burr hole was made and a recording electrode introduced into the right hippocampus. The electrode was secured with dental acrylic. A needle reference was placed subcutaneously above the right ear. After recovery from the cooling, the rats’ baseline EEG activity was obtained. Flurothyl was then dripped through a funnel to filter paper placed under the animals’ nostrils until behavioral seizure activity occurred. Flurothyl seizures were recorded on each day between P0 and postnatal day 7 (P7). A minimum of three rats were studied at each age. Rats received flurothyl only once. At the end of the recording, the animal was given a lethal dose of sodium pentobarbital (100 mg/kg), and the brain was dissected with the electrode present to determine electrode placement.
KA status epilepticus after neonatal flurothyl seizures.
To study the effects of KA-induced status epilepticus on the hippocampus, KA was given to two groups of male rats; Group A (n = 20) had received a total of 25 flurothyl seizures, beginning at P0, and Group B (n = 18) were control rats and were placed in the flurothyl chamber but not exposed to flurothyl.
At postnatal day 43 (P43), both the flurothyl-treated and control rats were anesthetized with ketamine 80 mg/kg and xylazine 20 mg/kg intraperitoneally and bipolar electrodes placed in the right dorsal hippocampus using the following coordinates: from bregma, 5.3 mm posterior; 5.3 mm lateral; and 6.0-mm depth from dural surface. At postnatal day 45 (P45), 10 of the rats treated with flurothyl and 12 of the control rats not treated previously with flurothyl were given a convulsant dosage of KA (12 mg/kg). Eight rats that had previously received flurothyl seizures and six control rats without previous flurothyl seizures received an equal volume of saline instead of KA.
All rats were observed behaviorally for 6 hours after the initial injection of KA. EEGs were obtained periodically to assess the morphology and frequency of the epileptiform activity. Samples of EEGs were obtained during the various behavioral stages of the status epilepticus.
Perforant path stimulation.
To evaluate the effects of perforant path stimulation-induced hippocampal damage, two groups of rats were stimulated. Group A (n = 20) had received a total of 25 flurothyl seizures beginning at P0 at a rate of five seizures per day; Group B (n = 18) were control rats and were placed in the flurothyl chamber but not exposed to flurothyl.
Between P60 and postnatal day 65 (P65), bipolar electrode animals were implanted after anesthesia with ketamine 80 mg/kg and xylazine 20 mg/kg intraperitoneally in both groups of animals. Coordinates for the perforant path stimulation were as follows: stimulation electrode in angular bundle (from bregma), 7.0 mm posterior; 4.5 mm lateral; depth determined from evoked potentials; recording electrode in CA1 (from bregma): 4.1 mm posterior; 2.6 mm lateral; depth determined by evoked potentials.24 During implantation of the stimulating and recording electrodes, evoked potentials were recorded using 2-mA, 0.5-msec pulses. Stimulus voltage greater than required to evoke maximum amplitude population spikes was used. Once a reproducible, well-defined waveform was recorded, input/output voltages were established to ensure that the responses were linear. Animals in which reproducible-evoked potentials could not be obtained were not used for further experimentation.
After a recovery period of 5 to 7 days, evoked population spikes were again elicited to verify that the stimulating and recording electrodes were functioning properly. The perforant path was then stimulated in freely moving rats with a 0.1-msec, 20-Hz stimulation for 60 minutes using a Grass S88 stimulator (West Warwick, RI). Intermittent EEG recordings during and for 60 minutes after the stimulation were obtained to verify that the animals were having electrographic seizures. Animals without behavioral and electrographic seizures were not used for further experimentation.
Histologic analysis.
The animals were killed 10 to 14 days after KA and perforant path stimulation using an overdose of sodium pentobarbital (100 mg/kg). Animals were rapidly perfused transcardially with 200 to 300 mL phosphate-buffered saline followed by 400- to 500-mL 10% buffered formaldehyde solution. The brains were postfixed in 10% formaldehyde solution for 24 hours and then placed in a 30% sucrose solution until the brains sank to the bottom of the chamber. Coronal sections through the entire extent of the hippocampus were cut at 20 μm on a freezing microtome, and sections were stored in phosphate-buffered saline.
Because rats receiving KA had significant cell loss, a scoring system was used to grade the degree of pathology. CA1, CA3, and the hilus were graded separately using the following scale: 0, no cell loss; 1, <25% cell loss; 2, 25% to 50% cell loss; 3, 50% to 75% cell loss; 4, no remaining cells. Because rats receiving perforant path stimulation had less-discernible cell loss, cell counting was performed.
Cell counts.
All hippocampi were evaluated using both visual microscopic inspection of the specimen and quantitative measurements. Unbiased stereologic cell counts were performed by using the optical dissector method.25,26 A 40× lens was used with a numeric aperture of 0.75 and a depth of field of 2.5 μm. A counting box of 50 × 50 μm divided into 100 equal squares (each with an area of 25 μm2) was used for cell counts. Cells were counted in a four-square (100 μm2) area. The bottom plane of the counting box was 5 μm above the bottom plane of the tissue section while the top plane of the counting box was 5 μm below the top plane of the tissue. Only cells with a nucleus coming into focus when focusing from the bottom to the top of the counting box were included.27 If any part of the cell touched the left or bottom edges of the box, that cell was not counted; cells touching the upper and right-hand border were counted.
Cells were counted from both hippocampi of each specimen. The hippocampal region analyzed was identical to the region with Timm stain density measurements. Counting was confined to the anterior portion of the hippocampus, where the two blades of the dentate gyrus were approximately equal in length (figure 20, and a section 1000 μm posterior, figure 22, Paxinos and Watson24).
In the dentate gyrus, the counting box was placed over the distal edge of the inferior blade so that the box covered the area of interest. Once the cells were counted, the counting box was moved systematically across the granule cell layer from the edge to the crest. The number of counting boxes that covered each dentate varied as a function of the region of the hippocampus so that only cells from eight counting boxes were used. If, for example, there were 10 boxes counted, cell numbers from the 5th and 10th boxes were not used in subsequent calculations.
In the CA3 region, the edge of the counting box was placed over the region bisected by an imaginary line connecting the two edges of the dentate granule cells. Once cells were counted, the entire box was advanced in a uniform method through the extent of the CA3 region to approximately 250 μm before the first CA1 cells were encountered. The CA1 region is characterized by smaller, densely packed cells. The counting box was placed over the first CA1 cells and advanced in a uniform fashion across the CA1 layer, ending at an imaginary line drawn perpendicular to the crest of the dentate gyrus. A maximum of 12 boxes were counted for the CA3 and CA1 regions with excessive counting boxes eliminated as described above for the dentate gyrus cell counts. In the hilus, total cell counts were obtained between the two blades of the dentate granular cell layer.
Reference volume measurements.
The same slides used for the cell counts were used for volume measurements. At 40× magnification, images were captured digitally to a monitor using an image analysis system (Media Cybergenics, Silver Spring, MD). Areas of CA3, CA1, hilus, and the dentate gyrus were obtained manually by tracing the histologic area. Volume was calculated by multiplying the mean area of two adjacent specimens by distance between the two specimens and adding the volumes for the entire length of the hippocampus analyzed.
Extrapolated cell number.
To estimate the total number of cells located in the region of interest, the mean cell density (cell number/mm3) of dentate granule cells, CA3, and CA1 neurons from two adjacent sections was multiplied by the volume of tissue between the two sections. The extrapolated total cell count was then obtained by adding all of these counts. Hilar cell counts were estimated by multiplying the means of the total cell counts for adjacent hippocampal sections by the length of the hippocampus measured.
Results.
Flurothyl seizure exposure.
Flurothyl resulted in seizures at each age group studied. The seizures were quite stereotyped, as described previously.21 Rats initially became quite agitated with the head bobbing or turning from side to side, followed by attempts at running, squealing, and loss of posture. The rats would then invariably develop tonic posturing with both the forelimbs and hindlimbs stiffly extended. Mild cyanosis, urinary and fecal incontinence, and salivation usually were noted. Rats were removed from the chamber when the tonic phase began and allowed to recover in room air. Typically, the rats returned to baseline behavior within 10 to 15 minutes. In the animals receiving 25 flurothyl seizures, there was a progressive increase in the duration and intensity of the tonic phase with serial seizures. Mortality rate was low, only 2 of 32 (6%).
EEG changes.
There were no clear differences in the behavioral responses to flurothyl in animals receiving the inhalant from P0 to P7. However, there were age-related changes in the EEG. Although there was some variability in the frequency and morphology of the waveforms among animals of the same age, in general the younger animals (P0 and postnatal day 1 [P1]) had slower and less-regular ictal discharges than the older animals studied (figure 1).
Figure 1. Representative examples of EEG during tonic phase of flurothyl-induced seizures. Marked changes occurred from baseline (not shown) during the seizures. With increasing age, there was an increase in frequency and rhythmicity of the discharges. Calibration = 2 s/1000 μV.
Flurothyl-induced histologic changes.
No cell loss was seen in any of the rats with recurrent flurothyl seizures, whether killed at P10, P20, or P60 (table). There were more cells in the dentate granule cell layer in all three age groups receiving flurothyl compared to the control rats, as well as an increase in the number of CA1 cells in the rats killed at P10.
Extrapolated number of cells (± standard error) in hippocampal region where cell counting was performed in rats with neonatal seizures and control rats killed at P10, P20, and P60
KA after neonatal flurothyl seizures.
All rats receiving KA had status epilepticus develop. During the first 15 minutes after KA injection, behavioral manifestations consisted of wet dog shakes, facial twitching, myoclonic jerks, head nodding, circling, scratching, and increased salivation. During the next 30 minutes, animals had intermittent forelimb clonus develop, later accompanied by rearing and falling. Intermittently, there were periods of immobility, ataxia, and agitation. This behavior lasted a minimum of 90 minutes and generally continued for 5 to 6 hours after KA administration.
In addition to behavioral seizures, all rats receiving KA had ictal discharges on the EEG. There were no apparent differences in frequency, waveform morphology, or duration of EEG ictal discharges in the rats with previous neonatal flurothyl seizures and in the control rats (figure 2).
Figure 2. Representative example of EEGs during various behaviors during kainic acid-induced seizures in a control rat (A) and a rat (B) subjected previously to a series of flurothyl seizures from the day of birth (P0) through postnatal day 4 (P4). Calibration = 2 s/1000 μV.
The mortality rate after KA in the rats with recurrent flurothyl seizures (3/10, 30%) was similar to that in the rats without prior flurothyl seizures (5/12, 42%). None of the rats receiving saline died. Considerable hippocampal damage was seen in both groups of animals receiving KA. As expected, no hippocampal damage was seen in either of the two groups (flurothyl-treated or control rats) that did not receive KA. There were clear differences in pathology scores after KA between rats with recurrent neonatal seizures and control rats, with the flurothyl-treated animals having significantly higher pathology scores (control rats followed by KA, 1.8; flurothyl-treated followed by KA, 2.7; control rats followed by saline, 0; flurothyl-treated followed by saline, 0; Kruskal–Wallis = 23.29, p < 0.001). The pathology scores between the two groups receiving KA were significantly different (Kruskal–Wallis = 4.94, p = 0.026).
Figure 3 provides examples of the difference in hippocampal pathology between the control rats and flurothyl-treated rats. Although KA caused considerable damage in all rats, the neuronal loss and gliosis in the rats with neonatal seizures were more pronounced.
Figure 3. Comparison of histologic damage in control rat without history of neonatal seizures receiving kainic acid (KA) (A, C, E) and rat subjected to KA after a previous series of flurothyl seizures during neonatal period (B, D, F). The cell loss in the rat with previous flurothyl seizures is considerably greater than in the control rats. Cell loss and gliosis were greater in CA3 (C, D) and CA1 (E, F) in the rats with neonatal seizures than in the control rats. Scale bar for A and B = 100 μm; for C–F = 50 μm.
Perforant path stimulation.
Twelve of the animals in Group A (animals with recurrent neonatal seizures) and 12 of the rats in Group B (control rats) had behavioral and electrographic seizures with perforant path stimulation. During the stimulation, animals had wet dog shakes, chewing, immobility, forelimb clonus, and periods of immobility. Ictal activity was noted during and after the stimulation stopped. One (8.3%) of 12 control rats and 4 (33.2%) of 12 rats with neonatal seizures died during or after the stimulation.
Both the control rats and the flurothyl-treated animals had modest hippocampal cell loss after perforant path stimulation. Figure 4 shows examples of mild cell loss in the CA3 region after stimulation. No clear differences were noted in the two groups on visual inspection. However, with cell-counting techniques, it was found that when compared to control rats, animals with recurrent flurothyl seizures had decreases of cell number in CA1 (t = 2.934, p = 0.011) and the hilus (t = 5.66, p < 0.001) (see figure 5). Conversely, in the dentate gyrus, there was an increase number of cells in the flurothyl-treated animals compared to the control animals (t = 4.916, p < 0.001).
Figure 4. Example of cell loss in CA3 subfield after perforant path stimulation in a control rat (A) and a rat with recurrent flurothyl seizures (B). No gross differences between the two groups were noted with visual inspection. Scale bar = 20 μm.
Figure 5. Comparison of extrapolated cell number in dorsal hippocampus after perforant path stimulation in rats with prior neonatal seizures (flurothyl-Rx) and control rats. Cell number was decreased in CA1 and the hilus and increased in the dentate gyrus in the rats with prior neonatal seizures (* p < 0.05).
Discussion.
We found that seizures during the neonatal period substantially increase the degree of seizure-induced damage later in life. In both the KA and perforant path stimulation models, animals with previous flurothyl seizures had significant increases in cell loss compared to control rats without prior neonatal seizures. This cell loss was a result of the status epilepticus induced by KA or perforant path stimulations since cell loss was not observed in rats with neonatal flurothyl seizures when examined at P10, P20, or P60. This study also demonstrates that neonatal seizures predispose the brain to further seizure-induced injury regardless of whether the second seizure is severe, as with KA, or mild, as with perforant path stimulation.
Flurothyl inhalation is a reliable method for inducing seizures with a very low mortality. Although the mechanism of the flurothyl seizures is not entirely clear,28 the behavioral manifestations suggest that the seizures are generalized. During the early neonatal period in which we administered flurothyl, ictal behavioral changes were relatively consistent. However, the EEG ictal pattern changed during the first few days of life, evolving from slow, irregular discharges to more rapid, regular epileptiform activity. As noted previously,23,29,30 flurothyl administration resulted in a kindling effect with a decrease in latency to seizure onset with serial exposures.
Prolonged flurothyl seizures in the adult rat have been associated with cell loss in the cerebral cortex, hippocampus, and thalamus.31 However, brief flurothyl seizures in the young rat do not appear to cause cell loss. In this study, as well as others from our laboratory, no cell loss has been observed after flurothyl seizures.21,30 Although it is not clear why flurothyl leads to an increase in granule cell number, there are now a number of studies demonstrating that recurrent seizures21,32 or status epilepticus33 can lead to neurogenesis in the dentate granule cell layer. An alternative explanation is that recurrent neonatal seizures reduce the amount of programed cell death that normally occurs during early development.
Why neonatal seizures that result in no apparent cell loss could “prime” the brain for later seizure-induced cell loss is not clear. However, recent studies in our laboratory have demonstrated that although neonatal seizures do not result in significant cell loss, they do result in alterations of neuronal circuitry.3,21,22 With both recurrent flurothyl 21and pentenlenetetrazol22 seizures during the neonatal period, there is sprouting of mossy fibers into the inner-molecular layer of the dentate and CA3 pyramidal cell layer. Therefore, one explanation for our findings might be that the excess number of granule cells has resulted in an overabundance of projections to target cells resulting in a hyperexcitable circuit.7,34
Because glutamate is the neurotransmitter of the mossy fibers, it is tempting to suggest that increased numbers of glutamatergic synapses could increase excitability, lower seizure threshold, and increase the amount of seizure-induced damage. However, sprouting of mossy fibers may not necessarily increase glutamate release in flurothyl-treated rats. Acsády et al.35demonstrated that terminals from dentate granule cells are more likely to innervate gamma-aminobutyric acid inhibitory interneurons than excitatory pyramidal cells.
This study presents additional evidence that despite lack of cell loss, neonatal seizures initiate a cascade of changes in the developing brain that are maladaptive and increase the risk of subsequent damage with a second insult. Whether children with neonatal seizures are at higher risk of neurologic damage after a second seizure later in life is unclear. It is also unclear whether medical intervention at the time of the neonatal seizures would alter the subsequent course.
Acknowledgments
Supported by the Emily P. Rogers Research Fund and grants from the National Institutes of Health (NS27984) (G.L.H.) and EpiFellows (R.S.).
- Received February 25, 1999.
- Accepted June 7, 1999.
References
Disputes & Debates: Rapid online correspondence
NOTE: All authors' disclosures must be entered and current in our database before comments can be posted. Enter and update disclosures at http://submit.neurology.org. Exception: replies to comments concerning an article you originally authored do not require updated disclosures.
- Stay timely. Submit only on articles published within the last 8 weeks.
- Do not be redundant. Read any comments already posted on the article prior to submission.
- 200 words maximum.
- 5 references maximum. Reference 1 must be the article on which you are commenting.
- 5 authors maximum. Exception: replies can include all original authors of the article.
- Submitted comments are subject to editing and editor review prior to posting.