The influence of long-term potentiation on the spatial relationship between astrocyte processes and potentiated synapses in the dentate gyrus neuropil of rat brain

Neurophysiological analysis of long-term potentiation in mammalian brain

Long-term potentiation (LTP) is a persistent increase in postsynaptic response following a high-frequency presynaptic activation. Characteristic LTP features, including input specificity and associativity, make it a popular model to study memory mechanisms. Mechanisms of LTP induction and maintenance are briefly reviewed. Increased intracellular Ca2+ concentration is shown to be critical for LTP induction. This increase is believed to be based on Ca2+ influx secondary to activation of N-methyl-D-aspartate (NMDA) subtype of glutamate receptors. Existence of other sources of Ca2+ increase and other critical factors is now becoming evident. They include voltage-dependent Ca2+ channels, Ca2+ intracellular stores, metabotropic glutamate receptors, 'modulatory' transmitters. An example of an involvement of voltage-dependent Ca2+ channels is potentiation induced by intracellular depolarizing pulses. LTP can be divided into decremental earlier (E-LTP) and non-decremental late (L-LTP) phases which explains some inconsistencies in studies of LTP mechanisms. E-LTP is suggested to be based on a transient increase in presynaptic release probabilities. A hypothesis is considered which explains L-LTP by suggesting that Ca2+ activates structural changes leading to an increase in the synaptic gap resistance. This enhances positive synaptic electrical feedback and augments release probability. The hypothesis predicts specific morphological changes, synchronous transmitter release of two or several quanta in some central synapses and the amplification of such synchronization following LTP induction. Data are discussed which maintain these predictions.

Permanent increase of immunocytochemical reactivity for gamma-aminobutyric acid (GABA), glutamic acid decarboxylase, mitochondrial enzymes, and glial fibrillary acidic protein in rat cerebral cortex damaged by early postnatal hypoxia-ischemia

A former study indicated that hypoxic-ischemic encephalopathy in rat sustained during early postnatal life may result in permanent epileptic activity in the baseline electroencephalogram. We, therefore, investigated whether the presumed higher firing frequency and metabolic activity of neurons in such hypoxia-damaged cortical areas would be reflected by an enhanced light microscopic immunoreactivity of gamma-aminobutyric acid (GABA), the two isoforms of glutamic acid decarboxylase (GAD67 and GAD65), the mitochondrial enzymes cytochrome c oxidase and ATP synthase, and/or glial fibrillary acidic, protein (GFAP). To that end rat pups, 12-13 days of age, were unilaterally exposed to hypoxic-ischemic conditions and, after a survival period of 2 and 6 1/2 months, respectively, killed by perfusion fixation. After dissection of the brain, coronal vibratome sections of animals showing cortical damage were immunostained for the presence of the above-mentioned antigens. Subsequent qualitative analysis revealed that the surroundings of cortical infarctions were unambiguously characterized by a disordered neural network containing numerous nerve cells, fibers and/or endings showing an enhanced immunoreactivity for GABA, both isoforms of glutamic acid decarboxylase, and cytochrome c oxidase and ATP synthase, while the astrocytes showed an enhanced immunoreactivity for GFAP. The diverse patterns of enhanced immunoreactivity suggested, furthermore, a wider low-to-high range of metabolic activities in both excitatory and inhibitory neurons.

Glutamate blocks astroglial stellation: effect of glutamate uptake and volume changes

Neurotransmitters which increase intracellular cAMP levels can cause cultured astroglia to change from a flat, polygonal shape to a stellate morphology. Little is known about how glial stellation can be regulated by other transmitters. In the present study, we demonstrated that L-glutamate blocked isoproterenol (ISO) or dibutyryl-cAMP induced stellation in astroglia. The glutamate inhibition was concentration dependent, with its maximal effect on > 90% of cells at 500 microM. Glutamate also reversed glial stellation within a short period (< 30 min). Glutamate uptake analogues, D-glutamate and D-aspartate, rather than receptor agonists, kainate and quisqualate, mimicked the glutamate effect. Likewise, the glutamate uptake blocker, D-thero-beta-hydroxyaspartate, blocked the glutamate effect. The glutamate inhibition was not a result of inhibition of cAMP formation, since norepinephrine, which inhibited 80% of ISO-stimulated cAMP, also caused glial stellation. Increases in extracellular K+ to 50 mM also reduced glial stellation, whereas 25 mM K+ had little effect. Since 25 mM K+ caused much greater depolarization than 400 microM glutamate, it was unlikely that the effects of both glutamate and high [K+] on glial stellation were due to membrane depolarization. Hypotonic treatment (120 mOsm) enhanced, whereas hypertonic treatment (520 mOsm) prevented, the glutamate reversal of glial stellation. Thus, glial swelling appeared to be a primary mechanism for the inhibitory effect of glutamate and high [K+] on glial stellation. This mechanism could also explain the observation that glutamate inhibited stellation induced by PMA, a PKC activator. Our data suggest that glutamate released from neurons during neuronal activity or pathology can be taken up by astrocytes and alter their morphology. Changes in glial morphology may in turn affect the volume and composition of the extracellular space and, as a result, neuronal activity.

Glial hypertrophy is associated with synaptogenesis following motor-skill learning, but not with angiogenesis following exercise

Rats reared from weaning in a complex environment have an increase in 1) glial surface area, 2) capillary volume, and 3) the number of synapses, per neuron. In that paradigm it has not been possible to determine whether the glial increase more closely correlates with the increase in synaptic numbers or with angiogenesis. More recently we have found that rats that exercised had an increase in the density of capillaries without an increase in the synaptic numbers, whereas rats that learned new motor skills had a greater number of synapses per neuron without an increase in the density of capillaries. Those findings provided the opportunity to investigate whether changes in glial volume in the cerebellum correspond to changes in the number of synapses or in capillary volume. Glial area fraction estimates were obtained using point counts on electron micrographs from the previous studies. The skill learning group had a greater volume of molecular layer per Purkinje cell, and also a greater volume of glia per Purkinje cell, than rats in either an inactive group or rats in two exercise groups. No significant differences were found in glial volume per synapse and glial volume per capillary across groups, although there was a tendency for glial volume per capillary to be lower in the exercise groups. The data indicate that glial volume correlates with synaptic numbers and not with capillary density.

Synaptic transmission in the hippocampus: critical role for glial cells

The importance of glial cells in controlling the neuronal microenvironment has been increasingly recognized. We now demonstrate that glial cells play an integral role in hippocampal synaptic transmission by using the glial-specific metabolic blocker fluoroacetate (FAC) to selectively inhibit glial cell function. FAC inhibits evoked intracellular postsynaptic potentials (PSPs; IC50 = 39 microM) as well as population PSPs (IC50 = 65 microM) in field CA1 of the guinea pig hippocampal slice. Spontaneous synaptic transmission is concurrently decreased. These effects are time and dose dependent. ATP concentrations in glial but not neuronal elements are also significantly reduced with FAC treatment. Simultaneous application of the metabolic substrate isocitrate with FAC prevents both the reduction in glial ATP concentrations and the decrease in evoked PSPs. Given that isocitrate is selectively taken up by glia, these data further support a glial specific metabolic action of FAC. Additionally, FAC has no postsynaptic effects as peak responses to iontophoretically applied glutamate are unchanged. However, the decay of both iontophoretic and evoked PSPs are prolonged following FAC treatment suggesting inhibition of glutamate uptake may contribute to the FAC-induced depression of synaptic potentials. These results show, for the first time, that glial cells are critical for maintenance of synaptic transmission and suggest a role for glial cells in the modulation of synaptic efficacy.

Increase in extracellular NCAM and amyloid precursor protein following induction of long-term potentiation in the dentate gyrus of anaesthetized rats

The extracellular concentrations of the amyloid precursor protein (APP) and neural-cell adhesion molecule (NCAM) in the dentate gyrus of the anaesthetized rat were assayed before and after the induction of long-term potentiation (LTP) in vivo. Levels of high molecular weight neurofilament protein and activity of the lysosomal enzyme arylsulphatase were measured as internal controls and indicators of neuronal damage. Ninety minutes after the induction of LTP, the concentrations of NCAM and APP increased, in an NMDA-dependent manner, in the absence of changes in neurofilament and arylsulphatase levels. The delayed changes in the extracellular concentration of these molecules may reflect events leading to morphological modifications following LTP.

A method for quantification of astrocytic processes reveals laminar differences in the dentate gyrus but not in CA1 of the hippocampal formation

The distribution of astrocytic processes in the dentate gyrus molecular layer and CA1 subregion of the hippocampal formation was examined in 8-microns frozen sections from adult male rats. The stereological cycloid intersection method (Baddeley et al., 1986) was used to estimate surface density of glial fibrillary acidic protein-immunoreactive astrocytic processes. A distinct lamination pattern was delineated in the dentate gyrus which corresponds to the afferent input to this region. No such pattern was detectable in s. radiatum of CA1. By using this relatively new stereological method which utilizes a sensitive sampling scheme, it was possible to increase the reliability in reporting the distribution of fine processes while at the same time greatly reducing the amount of time necessary to estimate the amount of astrocytic processes in a given region.

Astrocytic and synaptic response to kindling in hippocampal subfield CA1. II. Synaptogenesis and astrocytic process increases to in vivo kindling

Astrocytic glia are important for maintaining synaptic function during physiological activity. Recent hypotheses concerning epilepsy suggest a role for astrocytes in the control of neuronal excitability and in pathogenesis. This report provides morphological evidence that the periodic electrical stimulation used in the kindling model of epilepsy induces astrocytic hypertrophy and an increase in shaft synapse density in the CA1 region of the hippocampus. The Schaffer collateral pathway in the stratum radiatum of CA1 of five pairs of rats was kindled in vivo. Control animals received the same number of stimulations at a lower intensity and frequency. The animals were killed 24-48 h after reaching the criterion of five generalized seizures, and the brains were examined by electron microscopy. Kindling produced a 37% and 33% increase in the volume fraction of astrocytic processes in the middle and distal portions, respectively, of the s. radiatum in CA1. In the same tissue, the number (areal density) of shaft synapses was increased 25% in the s. radiatum of animals exhibiting generalized seizures. On the other hand, the areal density of degenerating synapses in both kindled and control animals was low and not significantly different. These results suggest that both synaptogenesis and hypertrophy of astrocytes contribute to an early stage of epileptogenesis when degenerative changes of the sort that might induce gliosis were not prominent in the tissue under study.

Glial glycogen stores affect neuronal survival during glucose deprivation in vitro

Glia perform several energy-dependent functions that may aid neuronal survival under pathological conditions. Glycogen is the major energy reserve in brain, and it is localized almost exclusively to astrocytes. Using murine cortical cell cultures containing both glia and neurons, we examined the effect of altered glial glycogen stores on neuronal survival following glucose deprivation. As previously reported, cultures exposed for several hours to media lacking glucose developed widespread neuronal degeneration without glial degeneration. If glial astrocyte glycogen content was increased to 2-3 times control levels by a 24-h pretreatment with 1 microM insulin or 0.5 mM methionine sulfoximine (MSO), glucose deprivation-induced neuronal degeneration was attenuated. These protective effects were blocked if glycogen levels were reduced back to control levels by a 30-min exposure to 1 mM dibutyryl cyclic AMP or 20 microM norepinephrine prior to glucose deprivation. Astrocyte glycogen stores may be an important factor influencing neuronal survival under conditions of energy substrate limitation.