Jennifer Hiscock

Kainic acid induces distinct types of epileptiform discharge with differential involvement of hippocampus and neocortex.

Systemic administration of kainic acid (KA), an excitatory amino acid agonist, provides a model of epilepsy due to increased neural excitation. We examined discharges using multi-channel EEG recording and spectral analysis in rats implanted with neocortical and hippocampal electrodes after intravenous infusion of KA (10 mg/kg), until and including the first convulsive seizure. Gamma activity (30-80 Hz) increased in hippocampus from 3-9 min after KA administration. Two types of preconvulsive bilateral rhythmic discharges were observed, both consisting of generalised high voltage sharp waves at low frequencies (<10 Hz) mixed with fast oscillations (<20 Hz): (1) generalised non-convulsive discharges (GNCD) occurred in all animals and (2) spike-wave discharges (SW), predominantly localised in neocortex, occurred in 45% of animals. Convulsive seizure evolved out of a GNCD. Spectral profiles of epileptiform discharges were characterised by an increase in power of low (<10 Hz) and high (beta and gamma range, 20-80 Hz) frequencies which were differently expressed in neocortex and hippocampus. Thus, in this model of convulsive epilepsy caused by increased excitation, there is an early increase in gamma activity, a process that might contribute to synchronisation, and two distinct types of bilateral discharges, hippocampal-neocortical (GNCD) and preferentially neocortical (SW). Neocortical, not hippocampal, changes in EEG power correlated with development of convulsive behaviours.

Fos Induction Following Systemic Kainic Acid: Early Expression in Hippocampus and Later Widespread Expression Correlated With Seizure

We determined the distribution of Fos protein expression in a model of generalised epilepsy caused by excessive neuronal excitation. Fos immunoreactivity was mapped in forebrain in unrestrained rats, previously prepared with an indwelling venous catheter, after the intravenous administration of kainic acid (10 mg/kg). We determined cerebral activation following various periods of exposure to kainic acid by using intravenous administration of pentobarbitone to prevent further activation. Within a few minutes, kainic acid caused episodes of staring, sniffing, wet dog shakes, nodding and chewing. Fos induction occurred initially and simultaneously in hippocampus, subiculum, septum and entorhinal cortex as early as 9.5 min after kainate injection. After up to 40 min of staring, sniffing, wet dog shakes, nodding and chewing, Fos induction was not further increased above levels present within the first 9.5 min. After 56 +/- 6 min a motor convulsion occurred, initially affecting the jaw, head and tail and variably extending to the forelimbs, trunk or hindlimbs. Following the convulsive event, additional Fos was expressed in hippocampus, thalamus, caudate-putamen and other subcortical structures and in the cerebral cortex. Fos induction was sometimes asymmetric in entorhinal, visual, piriform, cingulum, parietal and frontal cortices and in amygdala and dorsal endopiriform area. Electroencephalographic recordings after a few minutes exposure to kainic acid revealed an increased amplitude of fast frequencies in hippocampus which appeared to correlate with Fos induction in this structure. The findings are generally consistent with the reported distribution and slow development of kainic acid-induced seizure activity using electrophysiological and deoxyglucose methods. However, the Fos distribution suggests that (i) hippocampal, possibly dentate, activation precedes significant activation elsewhere, (ii) extensive involvement of other cerebral structures and cerebral cortex occurs simultaneously and correlates with motor seizures and (iii) brain structures can be recruited asymmetrically.

Distribution of Fos-positive neurons in cortical and subcortical structures after picrotoxin-induced convulsions varies with seizure type

The distribution of Fos protein was mapped in rat brain following a single non-focal convulsive seizure. Single seizures were induced with intravenous picrotoxin in unhandled animals housed in isolation. Different convulsive behaviours occurred unpredictably. The least severe seizures were predominantly localised to the face, head and forelimbs, without loss of posture control (restricted seizures). The most extensive seizures affected all limbs and trunk, sometimes with falling (generalised seizures). There was a correlation between seizure behaviour and distribution of Fos induction. After restricted seizures, Fos was induced at highest levels in neocortex and piriform cortex and was prominent in entorhinal cortex, caudal-ventral caudate-putamen and amygdala. Regions of thalamus were consistently and lightly labelled, but Fos induction did not occur in hippocampus. After generalised seizures, there was Fos induction in cortex but less than after restricted seizures and, in three of four animals, also in dentate gyrus, hippocampus and subiculum. There was occasional or variable labelling of thalamus, basolateral amygdala and caudate-putamen. One animal with generalised seizures showed no hippocampal Fos induction. The findings indicate that picrotoxin induces seizures with at least two different patterns of neuronal involvement. The cortex, part of the caudate-putamen, amygdala and thalamus are involved in restricted seizures while the hippocampus, cortex and thalamus are involved in generalised seizures. The results do not support the view that generalised seizures are a progression from restricted forms. Cortical Fos involvement is entirely consistent with the participation of cortex in non-focal epilepsy. In these non-focal seizures, the dentate-hippocampus may be a source of excitation to cortex in the generalised group while the cortex appears to be the predominant site of excitation in the restricted group.

Morphological characterization of substance P-like immunoreactive amacrine cells in the anuran retina.

Using substance P immunohistochemistry it was possible to demonstrate a class of morphologically homogeneous group of neurons in the inner nuclear layer (INL) of the retina of two anuran species: Xenopus laevis and Bufo marinus. The number of cells with substance P-like immunoreactivity (SP-LI) was about 250 and 800 in juvenile and 600 and 2500 in adult Xenopus and Bufo, respectively, SP-LI cells had a small soma with one primary dendrite having up to four slender branches, located in the vitreal sublamina of the inner plexiform layer (IPL). Mean dendritic field sizes were 0.12 and 0.30 mm2 in juvenile and 0.29 and 0.65 mm2 in adult Xenopus and Bufo, respectively. The density of SP-LI cells was 40/mm2 in juvenile and 24/mm2 in adult Xenopus compared with 20/mm2 in juvenile and 13/mm2 in adult Bufo. Nearest neighbour distance measurements indicated that SP-LI cells were randomly distributed across the entire retina in both species. The location and the morphology of SP-LI cells indicated that they correspond to a subclass of wide-field amacrine cells, similar to types 20 and 21 described by Golgi techniques in the cat.

Neuropeptide Y-like immunoreactive amacrine cells in the retina ofBufo marinus

Neuropeptide Y-like immunoreactive (NPY-LI) amacrine cells of the Bufo marinus retina were morphologically characterized, and their retinal distribution was established using immunohistochemistry on retinal wholemount preparations and sectioned material. The somas of NPY-LI amacrine cells were situated in the innermost part of the inner nuclear layer and their dendrites branched primarily in the scleral sublamina of the inner plexiform layer. A subgroup of the NPY-LI cells had dendrites in both the scleral and vitreal sublamina. All immunoreactive cells had large dendritic fields (average 0.5 mm2) that resulted in a high dendritic overlap across the retina. NPY-LI amacrine cells were evenly distributed across the retina, with an average density of 30 cells/mm2, although higher densities were observed at regions adjacent to the ciliary margin. The dendritic field size of the NPY-LI cells, together with the previously characterized substance P-like immunoreactive (SP-LI) amacrine cells, indicates that they belong to the class of wide-field amacrine cells. However, unlike the SP-LI neurons whose dendrites branch in the vitreal sublamina of the inner plexiform layer, the dendrites of the majority of the NPY-LI neurons branch in the scleral sublamina.

Neuropeptide Y-like immunoreactivity in neurons of the human retina

The distribution of neuropeptide Y-like immunoreactivity (NPY-LI) was investigated in wholemounts and in transverse sections of the human retina. NPY-LI was localized to the soma and axonal processes of large ganglion cells (GCs) and to the soma and dendritic arborization of amacrine cells (ACs). NPY-LI GCs were unevenly distributed across the retina, the highest density of 875 cells/mm2 was found in the fovea centralis and the lowest density of 15 cells/mm2 in the peripheral retina. The total number of NPY-LI GCs in the retina was estimated to be about 85,000. The soma sizes of NPY-LI GCs increased from 116 microns 2 +/- 23 (s.d.) in the retinal centre to 251 microns 2 +/- 57 in the retinal periphery. The soma size of NPY-LI ACs was in the range of 40 and 50 microns 2. In transverse sections NPY-LI was seen to be localized to the optic fibre layer, to the somata of GCs, to the scleral sublamina of the inner plexiform layer (AC dendrites) and to the innermost part of the inner nuclear layer (somata of ACs). The gradients of soma sizes and retinal distribution of NPY-LI GCs were taken as an indication that they correspond to the class of large to very large GCs, previously identified in the human retina by Golgi impregnation.

Picrotoxin-induced generalised convulsive seizure in rat: changes in regional distribution and frequency of the power of electroencephalogram rhythms.

OBJECTIVES It is unknown how generalised discharges in primary generalised epilepsy (PGE) develop from background brain electrical activity or how widespread these discharged are throughout the brain. Here we address this by determining which neural structures and rhythms lead to and participate in generalised discharges in the picrotoxin rat model of PGE. METHODS Rats with chronically implanted electrodes were infused with picrotoxin until a seizure occurred. This process we refer to as acute epileptogenesis. The electroencephalogram (EEG) was recorded and spectral analysis applied off-line to determine changes in the spectral power of contributing frequencies in 13 brain regions. RESULTS Two types of generalised discharge occurred, spindles and seizure, which were present in all brain regions studied. None of the frequencies (1-100 Hz) were significantly increased in background EEG before either spindles or seizure. Within the generalised discharges, power changes revealed significant increases in 6-8 Hz, most powerful in ventrolateral thalamus and neocortex. Gamma frequencies were increased significantly in neocortical structures during spindles with further increases in most structures at seizure onset. 1 Hz was significantly increased in parietal cortex during spindles with differential increases at seizure onset. CONCLUSIONS We conclude that gamma, 1 and 6-8 Hz frequencies do not appear to contribute to picrotoxin epileptogenesis but do play a role in generalised seizures. The distribution of these frequencies during discharges suggests that the spindles are thalamocortical events and that the seizure is a cortical event with downstream effects on other brain regions.

Post-metamorphic retinal growth in Xenopus.

SummaryThe postmetamorphic growth of the retina in Xenopus was studied using 3H-thymidine (3HT) autoradiography and quantitative morphometric assays. 3HT was administered to tadpoles at stages 58, 62 and 66 and the animals sacrificed between 3 weeks and 12 months after metamorphosis. Reconstructions were made from serial sections and the position of labelled cell groups in the retina were established. On the reconstructed retina, regions formed up to stage 58, between stages 58 and 66 and after metamorphosis were measured. The area of the dorsal, ventral, temporal and nasal retinal halves was also determined from stage 58 through to adult.The entire retinal area increased 10-fold from stage 58 to 12 months after metamorphosis, the fastest growing region being the retinal periphery due to continuous cell addition at the ciliary margin. Concommitant with the retinal area growth, the number of ganglion cells increased from 20,000 to 85,000 over the time of investigation. Asymmetric cell addition to the ciliary margin from stage 58 onwards resulted in a predominantly crescentic retinal growth along the nasoventral ciliary margin. Consequently, the optic nerve head became displaced away from the geometric centre of the eye into the dorso-temporal retinal quadrant.These results suggest that besides a sustained cell production exclusively at the ciliary margin, a passive area expansion contributes to the overall retinal growth from the metamorphic climax to adulthood. It is also apparent that the steady increase of the number of retinal ganglion cells and optic fibres necessitates a continuous remodelling of the retinotectal connections throughout the lifespan of the animal.

Temporo-nasal asymmetry in the accretion of retinal ganglion cells in late larval and postmetamorphic Xenopus

SummaryThe spatial pattern of cell production and retinal growth were studied in Xenopus between stage 60 and two months after metamorphosis using 3H-proline and 3H-thymidine autoradiography. The position and the number of the ganglion cells labelled with 3H-thymidine were determined. The area of the unlabelled retina due to growth since 3H-proline administration at stage 60 was measured. Both retinal area measurements and counts of labelled ganglion cells showed 30–40% higher values in the temporal than in the nasal retinal half. The greater cell production and area accretion were even more pronounced between the temporal and the nasal retinal quadrants. The results on the temporoventral crescentic retinal growth rule out the possibility that from midlarval stages onwards the retinal and the tectal growth patterns are matched.

The changing distribution of neurons in the inner nuclear layer from metamorphosis to adult: a morphometric analysis of the anuran retina

SummaryThe generation and changing distribution of neurons of the inner nuclear layer (INL) in the retina of two anuran species, Bufo marinus and Xenopus laevis, were studied from metamorphosis to adult. Morphometric studies were undertaken at six developmental stages in Bufo and four in Xenopus. The number and thickness of neurons in the INL were established in 29 predetermined retinal locations from serial sections of the eyes cut vertically or horizontally. The total number of neurons in the INL increased from metamorphosis to adult from 826000 ± 185 to 18760000 ± 562 (mean ± SD) in Bufo and from 308000 ± 25 to 877000 ± 31 in Xenopus. Over the same period the surface area of the INL increased about 50-fold from 2 mm2 to 96 mm2 in Bufo and 5-fold from 2.5 mm2 to 13 mm2 in Xenopus. In Bufo the difference between the highest cell number (centraltemporal retina) and the lowest cell number in a sample area (dorsal and ventral peripheral retina) was 2.1∶1 at metamorphosis. This ratio increased to 3.4∶1 in the adult. Both the cell number and cell density per sample area in the INL was found to be higher along the nasotemporal meridian of the eye overlying the visual streak of the ganglion cell layer (GCL) of the retina. The retinal distribution of neurons in the INL did not change significantly during postmetamorphic growth in Xenopus. At metamorphosis a 1.7∶1 difference was found between the highest neuron number (retinal ciliary margin) and lowest neuron number (retinal centre) decreasing to 1.5∶1 in the adult. Retinae were labelled with 3H-thymidine in 15 mm Bufos and examined 2, 6, 12 and 18 weeks later. Higher rates of cell addition to the nasal and temporal poles of the INL were found compared with that at the dorsal and ventral poles. The retinal radial growth at the ciliary margin of the dorsal, ventral, nasal and temporal poles between the time of isotope injection and 18 weeks survival was found to be uneven; more radial elongation occurred at the nasal, dorsal and ventral poles and less at the temporal pole. These observations suggest that (a) the neuron distribution of the INL in adult animals approximates that of the GCL and (b) the visual streak-like area of the INL in Bufo develops by a sustained differential cell addition at the temporal and nasal poles of the retina.