Alterations in glutamate but not GABAA receptor subunit expression as a consequence of epileptiform activity in vitro.

The consequences of epileptiform discharge on the expression of glutamate and GABA receptors were examined by in situ hybridization histochemistry after treatment of rat hippocampal slice cultures with convulsants. Application of 500 microM picrotoxin for two days led to decreases in the messenger RNA levels for the N-methyl-D-aspartate receptor subunits, NR2A and NR2B, and for the non-N-methyl-D-aspartate receptor subunits, glutamate receptors 1 and glutamate receptors 2, to about 50% of the levels seen in control cultures. Messenger RNA levels for the N-methyl-D-aspartate receptor subunit, NR1; the non-N-methyl-D-aspartate receptor subunits, glutamate receptors 3 and 4; the high-affinity kainate receptor subunits 1 and 2; and the GABAA receptor subunits, alpha 2, beta 2, gamma 2 were unchanged. Decreased levels of expression were no longer seen five days after removal of convulsant. The down-regulation could be prevented by co-application of both the non-N-methyl-D-aspartate and N-methyl-D-aspartate receptor antagonists, 6-cyano-7-nitroquinoxaline-2,3-dione (CNQX) and dizocilpine maleate, but not by applying each alone. Application of CNQX or dizocilpine maleate in the absence of picrotoxin also resulted in changes in glutamate receptor expression. We suggest that the convulsant-induced reduction in glutamate receptor expression leads to a decreased excitability in these cultures, and that this down-regulation represents a compensatory reaction of hippocampal pyramidal cells to enhanced excitatory input.

Selective degeneration of CA1 pyramidal cells by chronic application of bismuth.

The heavy metal bismuth induces a new type of selective neuronal degeneration that shares some common aspects with that seen following hypoxia and ischemia. Continuous application of 3 microns bismuth to organotypic cultures of rat hippocampus resulted after 2-3 weeks in selective degeneration of CA1 pyramidal cells, while CA3 pyramidal cells, dentate granule cells, and subicular neurons were resistant. With 10 microns bismuth, the majority of hippocampal neurons degenerated. The addition of 20 microns MK-801, a noncompetitive NMDA-antagonist, during the entire culturing period failed to prevent neuronal degeneration induced by 3 microns bismuth. GABA-immunoreactive interneurons were also affected by bismuth, but were generally less sensitive than CA1 pyramidal cells. Acute application of up to 100 microns bismuth did not change the electrophysiological properties of CA1 pyramidal cells.

Reversible loss of dendritic spines and altered excitability after chronic epilepsy in hippocampal slice cultures.

The morphological and functional consequences of epileptic activity were investigated by applying the convulsants bicuculline and/or picrotoxin to mature rat hippocampal slice cultures. After 3 days, some cells in all hippocampal subfields showed signs of degeneration, including swollen somata, vacuolation, and dendritic deformities, whereas others displayed only a massive reduction in the number of their dendritic spines. Intracellular recordings from CA3 pyramidal cells revealed a decrease in the amplitude of evoked excitatory synaptic potentials. gamma-Aminobutyric acid-releasing interneurons and inhibitory synaptic potentials were unaffected. Seven days after withdrawal of convulsants, remaining cells possessed a normal number of dendritic spines, thus demonstrating a considerable capacity for recovery. The pathological changes induced by convulsants are similar to those found in the hippocampi of human epileptics, suggesting that they are a consequence, rather than a cause, of epilepsy.

Cholinergic fiber growth in co-cultures of CNS tissue.

In co-cultures prepared from the septum and the hippocampus, cholinergic fibers originating in the septal slices grew into the neighboring hippocampal tissue and established functional cholinergic connections with pyramidal cells. To get further insight into the mechanisms governing cholinergic fiber growth, we have added TTX to the growth medium (2 x 10(-7) M) to block propagated electrical activity. Under these conditions, considerably fewer cholinergic cells appeared to survive. A few cholinergic fibers still invaded hippocampal target tissue, but their number was markedly reduced compared with control cultures. Simultaneous application of NGF together with TTX, however, not only increased enzyme levels and enhanced survival of cholinergic neurons, but also led to hippocampal ingrowth in virtually all septo-hippocampal co-cultures. These data, therefore, suggest, that in the absence of spiking activity, cholinergic fibers are capable of growing into a co-cultured target tissue. To test the specificity of growth of septal cholinergic fibers, we have co-cultured septal slices with slices of various brain areas which in situ lack a major cholinergic innervation, in particular the cerebellum. In the vast majority of such co-cultures, cholinergic fibers remained restricted within the septal slices, without innervating cerebellar tissue. This failure might in part be related to the lack of trophic factors released by the target tissue. We have, therefore, grown septo-cerebellar cultures in the presence and absence of NGF. Following application of 100 ng/ml NGF during the entire growth of the cultures, numerous AChE-positive fibers originating in the septal slices invaded the co-cultured cerebellar slices.(ABSTRACT TRUNCATED AT 250 WORDS)

Continuous presence of nerve growth factor is required for maintenance of cholinergic septal neurons in organotypic slice cultures.

In co-cultures of rat septum and hippocampus, cholinergic neurons, identified by immunocytochemical techniques using antibodies against choline acetyltransferase, were found to be exclusively located in septal tissue. The presence of nerve growth factor during the entire growth period of four weeks increased the activities of acetylcholinesterase and choline acetyltransferase about 10-fold and strongly increased the number of acetylcholinesterase-positive neurons. Application of nerve growth factor yielded different effects depending on the age of the cultures. During the first two weeks in vitro, nerve growth factor enhanced the number of acetylcholinesterase-positive neurons, an effect which was no longer observed following later applications of nerve growth factor. Nerve growth factor increased the activities of cholinergic enzymes during all phases of in vitro development, but the effects of one-week applications were always considerably smaller than those observed following continuous application of nerve growth factor. The results of different application schedules suggest that the continuous presence of nerve growth factor is needed for maximal increases in cholinergic enzyme activities and maintenance of cholinergic neurons in septohippocampal co-cultures.

Organotypic Co-Cultures of Rat Locus Coeruleus and Hippocampus.

Slices from the brainstem at the level of the locus coeruleus and from the hippocampus of 5 - 7 day old rats were co-cultured using the roller tube technique. After 2 - 6 weeks in vitro the co-cultures were examined with antibodies raised against tyrosine-hydroxylase (TH). The cultures derived from the brainstem consistently contained a bilateral cluster of TH-positive neurons with 3 - 5 long slender dendrites. These neurons typically gave rise to several fine varicose fibres reminiscent of catecholaminergic axons. A morphologically distinct group of TH-positive neurons was detected in the hippocampal slices. The vast majority of them were located in the subicular region and a smaller number in the CA1/CA3 region of the hippocampal explant. TH-positive neurons were also present in mono-cultures of hippocampus or brainstem. In the vast majority of co-cultures, a variable number of TH-immunoreactive fibres of neurons derived from the locus coeruleus grew over considerable distances to terminate finally within the co-cultured hippocampus where they branched to form a diffuse innervation plexus with club-like endings. Even after several weeks in vitro, TH-positive fibres could still be seen exploring sites which were not related to their target, including the cell-free areas surrounding the cultures. Fibres in these outgrowth areas formed whirl-like endings. TH-positive fibres arising from neurons located in the hippocampus, on the other hand, did not branch extensively and never projected over long distances. Nerve growth factor had no apparent trophic effect on TH-positive cells in the hippocampus, the locus coeruleus, or in the co-cultures.